full document - Social Policy Research Centre

Transcription

COMPETING VISIONS
Proceedings of the National
social Policy Conference,
Sydney, 4-6 July 2001
edited by Tony Eardley and
Bruce Bradbury
SPRC Report 1/02
Social Policy Research Centre
University of New South Wales
April 2002
For a full list of SPRC Publications see, www.sprc.unsw.edu.au or
Contact: Publications, SPRC, University of New South Wales, Sydney, NSW, 2052, Australia.
Telephone: +61 (2) 9385 7800 Fax: +61 (2) 9385 7838 Email: sprc@unsw.edu.au
ISSN 1446 4179
ISBN 0 7334 1914 3
April 2002
The views expressed in this publication do not represent any official position on the part of the
Social Policy Research Centre, but only the views of the individual authors.
Editorial Note
The papers in this collection are a selection of those presented at the National Social
Policy Conference, 2001, held at the University of New South Wales on 4-6 July 2001.
The selection of papers is based on reviewing by independent referees in accordance with
the Department of Education, Science and Training's E1 publication category.
Please note that the views expressed in all papers are those of the respective authors and
not necessarily those of the Social Policy Research Centre.
The editors would like to thank, Annie Parkinson and Diana Encel for their assistance
with text editing and Duncan Aldridge and Rosita Lang for report production.
Tony Eardley and Bruce Bradbury
i
Foreword
The seventh National Social Policy Conference was held at the University of New South
Wales on 4-6 July 2001. The Conference is a key event on the Australian social policy
calendar and brings together a broad range of researchers, practitioners and advocates
with an interest in the theory and practice of social policy. The conference has become an
important national forum for exchanging ideas about research on the issues facing the
nation and its policy makers. Its interdisciplinary character reflects the nature of social
policy and is designed to encourage an informed critical dialogue about the strengths,
weaknesses and directions of social policy in Australia.
The theme chosen for the 2001 conference - Competing Visions - reflects the state of flux
that social policy is currently in around the world as new policy responses emerge to
address problems that have proved unexpectedly resilient to past policies. How far these
new approaches should be embraced given the uncertainties surrounding them is at the
centre of a policy debate that lacks any single vision or organising theme.
The extent to which social policy has become caught between a decline in the confidence
attached to existing approaches and the emergence of a new market-based approach is
evident in the balance of papers presented at the conference and in the selection published
here. Many of the underlying issues are addressed in Anna Yeatman's Keynote Address,
which re-visits conflicts between social policy and the freedom of the individual that now
feature prominently on the policy agenda. Welfare reform has been an area of great
activity in Australia in recent years and we are delighted to include David Ellwood's
masterly account of the US welfare reform experience that contains many lessons for
Australia.
The remaining papers provide a unique snapshot on the concerns of Australian social
policy research at the start of the millennium. Many of the topics covered, including
welfare reform, poverty and inequality, homelessness and disadvantage and Indigenous
welfare issues highlight some of the constant themes of previous conferences. Others,
focusing on the new social contract, superannuation, volunteering and mutual obligation
reflect the current concerns of policy and the trends in theory and practice that are driving
them.
I would like to take this opportunity to thank my SPRC colleagues for their on-going
commitment to the goals of research rigour, open enquiry and transparency that are
reflected in the National Social Policy Conference. Without that professional dedication
this conference (like its predecessors) would simply not have taken place. A special word
of thanks is due to Bruce Bradbury and Tony Eardley for putting this collection together
in its new format. I trust that you will find the papers accessible and informative, but
above all a stimulating contribution to what is an increasingly important debate.
Peter Saunders
SPRC Director
ii
Table of Contents
Editorial Note
i
Foreword
ii
Anna Yeatman
Social Policy Freedom and Individuality
1
David Ellwood
The US Vision of Work Based Reform: Promise,
Prospects, and Pitfalls
12
Jon Altman
Indigenous Hunter-gatherers in the 21st Century:
Beyond the Limits of Universalism in Australian Social
Policy
35
Maureen Baker and David Tippin
Health, Beneficiaries and Welfare to Work: Competing
Visions of Employability
44
Janeen Baxter
Changes in the Gender Division of Labour in Australia,
1986-1997
64
Sue Casey
Snakes and Ladders: Women’s Pathways Into and Out of
Homelessness
74
Gillian Considine and Gianni Zappala
Factors Influencing the Educational Performance of
Students from Disadvantaged Backgrounds
91
Peter Davidson
Employment Assistance for Long-term Unemployed
People: Time for a Re-think
108
David De Carvelho
The Social Contract Renegotiated: Protecting Public
Values in the Age of Contracting
126
James Doughney
Socioeconomic Banditry: Poker Machines and Income
Redistribution in Victoria
136
iii
Brian Fleming
Targeting Aged Care to People who are Financially or
Socially Disadvantaged: Policy Evaluation Study about
the Social Gradient in Health of Older South
Australians
155
Paul Henman
The Poverty of Welfare Reform Discourses
180
Boyd Hunter, Steven Kennedy and Daniel Smith
Sensitivity of Australian Income Distributions to Choice
of Equivalence Scales: Exploring some Parameters of
Indigenous Incomes
192
Simon Kelly, Richard Percival and Ann Harding
Women and Superannuation in the 21st Century: Poverty
or Plenty?
223
Pamela Kinnear
Mutual Obligation: A Reasonable Policy?
248
Carmel Laragy
Individualised Funding in Disability Services
263
Rachel Lloyd, Ann Harding, and Harry Greenwell
Worlds Apart: Postcodes with the Highest and Lowest
Poverty Rates in Today’s Australia
279
Greg Marston
Fashion, Fiction, Fertile Inquiry? Struggling with the
Postmodern Challenge and Social Policy Analysis
298
David Martin
Reforming the Welfare System in Remote Aboriginal
Communities: an Assessment of Noel Pearson's
Proposals
317
Rose Melville
Volunteers and Community Legal Centres: A
Partnership Under Threat
326
iv
Wayne Vroman and Vera Brusentsev
Australian Unemployment Protection: Challenges and
New Directions
348
Ariadne Vromen
Community Activism and Change: The Cases of Sydney
and Toronto
387
Maggie Walters
Labour Market Participation and the Married to Sole
Mother Transition
411
Jocelyn Williams, Margie Comrie and Frank Silgo
Walking the Path With New Parents
421
Lou Wilson, Barbara Pocock, and Marg Sexton
Workplace Change in South Australian Local
Government: Doing More with Less
439
vi
Social Policy, Freedom and Individuality
Anna Yeatman
Sociology
Macquarie University
1
What is the Rationale for Social Policy?
What is the rationale for social policy? By what values and vision should it be
oriented? It would seem that almost everyone who has a role in leading public opinion
conspires to tell us that such a fundamental debate cannot be afforded at this time of
scarce public resources. That we do not need vision because we cannot afford it. This
is a strange conception of public policy. It is actually saying we should not have a
vision. It proposes that we allow the means to establish the ends of public policy,
rather than ask what ends should public policy follow, and with what consequences
for a public revenue base, and its justification.
Why is it that we are not encouraged to think about the ends of public policy at this
time? My answer is that it is because we live in an era of laissez-faire. Laissez-faire is
a framework for public policy that is designed to free up the market to be selfregulating. The rules for operation of a market economy of any kind derive from
government. Government policy of the kind that favours a self-regulating market
economy is no less interventionist than any other kind of government policy. Laissezfaire government policy leads to a particular approach to social policy. It is one that
privileges the freedom of property right over other kinds of right, especially the right
of each person to an equality of regard.
When laissez-faire is allowed to control the policy settings, we find ourselves in a
position where wealth-seeking, or economic growth, is established as a goal in and of
itself. ‘The economy’ comes to dictate our action. We lose policy literacy in being
able to discuss and determine how the economy can serve human development.
We also live at a time of unparalleled wealth where there is no sign of the dynamics
of capitalist growth becoming tired, or technical change slowing. It this context, it is
even more striking that the public policy agenda should have been highjacked in this
way because the simple logic of doing things for the economy makes sense only as
long as one does not ask the fundamental question: ‘what do we need wealth for'
(Levine, 1995: 19)?
Those who have thought about this question argue that the pursuit of wealth makes
sense only to the extent that it serves human freedom. Two contemporary political
economists eloquently make this argument. I am thinking of Amartya Sen (1999) and
David Levine (1995). Each of them is aware of their debt in thinking about this
question to the classical economists, especially Adam Smith who saw wealth as the
condition of what he called civilisation. Sen (1999: 3) says simply: ‘Development can
Yeatman, A. (2002), ‘Social policy: freedom and individuality’, in T. Eardley and B. Bradbury, eds,
Competing Visions: Refereed Proceedings of the National Social Policy Conference 2001, SPRC
Report 1/02, Social Policy Research Centre, University of New South Wales, Sydney, 1-11.
ANNA YEATMAN
be seen as a process of expanding the real freedoms that people enjoy’. Levine (1995:
19) proposes that economics has been much better at explaining how market
economies encourage the growth of wealth (a focus on means) than it has been in
answering the question ‘what we want wealth for’ (a focus on ends). He goes on to
say ‘We need to know this not because it might reinforce our commitment, if any, to
the system - capitalism - that brings about the expansion of wealth but because we
need to know when we have enough and when our quest for wealth might end'
(Levine, 1995: 19).
For Sen, wealth-seeking makes sense if it expands the freedoms that people may
enjoy. For Levine (1995: 177), wealth-seeking makes sense only to the extent that it
achieves two things: firstly, the ‘capacity to secure material well-being'; and,
secondly, the creation of ‘choice and opportunity'. Both of these achievements can be
ascertained only if we (to borrow Sen's idea of capability) examine what it is that
people can be and do. By people we mean individuals. Are individuals secure? Do
they have both choice and opportunity to explore their individuality and to act in selfdetermining ways? Security and self-determination go together. An individual cannot
risk self-determination if s/he does not have a reasonably secure access to means of
livelihood, health, and longevity (see Levine, 1995: 29-30). It is wealth that has
provided the material base of human security. It is also wealth that has opened up the
multiplication of wants and life choices that make the conception of living one's own
life possible (for instance, in the exploration of one's talents, choice of occupation and
choice of way of life). The ability to lead a civilised sort of life follows upon access to
the enjoyment of these opportunities for security and autonomy.
When we ask the question, ‘What is wealth for?’ social policy no longer appears as a
supplement to the main game, public policy in service to the self-regulating market
economy. Rather, social policy turns out to be the main game. Social policy is the
appropriate name for what we are doing by way of public deliberation when we
engage both rationally and inclusively in asking the question how can wealth serve
human development? Or, the alternative way of asking the question, how can wealth
serve individual self-determination? I return at the end of this address to the question
of what a social policy that serves human development and individual selfdetermination involves. I do this in dialogue with T.H. Marshall's idea of social
citizenship.
However, social policy that is oriented to making the economy serve human
development and freedom is not the favoured approach at this time. The public policy
of laissez-faire dictates an approach to social policy that adapts it to the needs of the
economy. This approach to social policy is expressed in the two ideas of protection
and reciprocity (mutual obligation). Let us briefly examine these ideas and think
about how they measure up in terms of the values of human development and
individual autonomy.
2
Social Policy as the ‘Protection’ of the Poor
The conception of social policy in terms of protection is social policy designed not for
everyone, but for a marginalised group, the poor. It is a very different approach from
social policy for human development and individual autonomy. The protection
approach is structured by an opposition between those who are self-reliant and who
2
SOCIAL POLICY, FREEDOM AND INDIVIDUALITY
do not need protection, and those who are not or cannot be self-reliant, and who
therefore need protection. Freedom is the privilege of those who are self-reliant, and
the idea of freedom narrows to the kind of contractual freedom that is associated with
market exchange. Freedom turns out to be the freedom associated with the exercise of
property right. For those who are not or cannot be self-reliant, they forfeit freedom by
becoming dependent on the largesse of property owners either directly or through the
taxation they contribute to the state. The provision of public welfare is designed so as
to punish the able-bodied non-working poor for their failure in self-reliance. The
punishment is the deprivation of their freedom of action, and their subjection to the
state's coercive controls in order to make the labour market a more attractive option
than welfare dependence. T.H. Marshall's (1977) remarks on the New Poor Law of
1834 could have been designed to apply to today's version of social policy as
protection.
The Poor Law treated the claims of the poor, not as an integral
part of the rights of the citizen, but as an alternative to them - as
claims which could be met only if the claimants ceased to be
citizens in any true sense of the word. (Marshall, 1977: 88)
Marshall also discerns the true aim of this type of social policy orientation. Speaking
of the first era of laissez-faire, the second half of the nineteenth-century, he remarks:
‘The common purpose of statutory and voluntary effort was to abate the nuisance of
poverty without disturbing the pattern of inequality of which poverty was the most
obvious unpleasant consequence’ (1977: 105). This conception of social policy social policy as the protection of the poor through offering them a ‘safety-net' belongs to a laissez-faire policy context where it is difficult for most people to think
about what should social policy do as distinct from what should the poor be made to
do.
When poverty is associated with subjection to state coercion, and when the only real
security for freedom of choice and opportunity resides in market-oriented action, then
all who can participate in the market economy will exhaust the major part of their
energies in so doing. Their mental and physical energy for doing anything other than
follow the demanding and exploitative work ethic of the self-regulating market
economy will be minimal. Most will be driven by an acute anxiety to work hard and
save enough to create a private investment fund to protect themselves against the
insecurities of ill health and old age. The answer to the question of what we need
wealth for will seem so self-evident to them as to make the question redundant. Yet
the question they are asking and answering is not what do we need wealth for. Rather,
it is what do I need wealth for given the context is one in which insurance against risk
has been privatised at least for those who are able to enjoy such freedom as is on
offer, that is, the freedom of property right. When people think like this, they are
unable to think about the society that they share. By society I mean that people are
interdependent with each other in many and different ways, and that when they think
about who they are as individuals, and what it is that they want and need as
individuals, they do so always in reference to others. Private decisions make sense as
long as private action effectively serves the integrity of individuals, but often this is
not possible when the private action of one impacts negatively on the opportunities or
well-being of another. It is when we are able to think about our interdependencies and
3
ANNA YEATMAN
about shared objectives and purpose that we enter into the sphere of public goods, and
into thinking about what should be the balance between public and private goods.
3
Social Policy as Reciprocity
The current policy set-up encourages us to behave as though we were rats on a
sinking ship: it is every man and woman for him or herself. In this context a particular
kind of ressentiment develops. Ressentiment is the righteous anger the good feel in
relation to those who seem to be taking advantage of their goodness, those who are
seeking a free ride at their expense. In this case, the good are those who see
themselves as having made the necessary degree of sacrifice, through hard work and
private investment decisions, to take care of their own futures by way of private
insurance. They readily resent those who do not seem to be taking responsibility for
themselves in this way. From this it is a small step to assume that if individuals are
able to take care of themselves, both now and in the future, it is because they have
worked hard, and they have earned their share of privately distributed wealth.1 They
forget about the unearned wealth of those who have inherited property assets, and
they are unable to critically assess the ease of acquisition of vast wealth by those who
have the assets, the connections, and the capacity to take advantage of de-regulated
money markets. Instead the focus turns to the easiest target of their rage and
depression: the non-working poor. At this point, the argument that the non-working
poor should do something ‘in return for’ public income support assumes a self-evident
quality. This is the argument that White (2000) calls ‘welfare contractualism’.
White’s defence of what he calls the reciprocity principle (or, in this country, mutual
obligation) is one of the better argued conceptions. The policy is directed against what
is seen as the free-riding of the non-working poor on those who are making a
contribution to ‘the social product’.
It is intrinsically important that citizens who share in the social
product make a reasonable effort to ensure that others also
benefit from their membership of the productive scheme and, as
the flip-side of this, that they do not burden their fellow citizens,
making others worse off than they would be in their absence,
when they can reasonably avoid it. (White, 2000: 513)
White's argument makes sense if we assume as given policy parameters a selfregulating market economy that is accompanied by a state that functions on behalf of
the freedom of private property. As we have seen, the freedom of private property
requires that the state coerces the non-working and able-bodied poor into work if they
are dependent on public income support. In this framework, the community to which
we belong is not a citizen community but an economic or productive community. We
have rights only to the extent that we participate in productive or wealth-creating,
action.
The assumptions of the reciprocity approach to social policy, then, are not
independent of the protection approach to social policy. Both follow from the policy
framework settings of laissez-faire. Contrary to the implication of the reciprocity
1
For a critique of the idea that earnings are equivalent to contribution, see Levine, 1995,
chapter 8.
4
approach that it is revisiting the idea of the Social Contract, it does not do this. If we
revisited the idea of the Social Contract, we would find that we start with the question
of what it is that government understood as the public authority should do for us in
order to secure the integrity of our personhood.
4
Social Policy and Freedom
Properly understood, the social contract is an account of how the public authority has
to be conceived or designed so as to secure the rights of everyone to be an individual.
In its classical liberal version, the public authority (Locke's phrase) is established to
secure the private property right or freedom of individual householders. By freedom
Locke (1960, 324, para. 87) meant the right to life, liberty, and estate of the
householder. In more contemporaneous and inclusive language, it is the freedom of
each individual to live his or her own life. In the classical liberal conception of private
property, the role of property is to make personality possible.
However, from Locke's time onwards, property becomes confused with investment in
capital. There thus develops a profound tension between the conception of property
that functions on behalf of individual autonomy and the conception of property as
privately owned capital.
The point to hold onto is that in classical liberalism property functions on behalf of
the autonomy of the citizen, his right to be a distinct and individual personality. This
starting point, and the limitation of individualised personality to male household
heads, makes it possible for us to open up a more general enquiry into the conditions
for autonomous or, as we might call it, individualised citizenship.
For the classical liberals such as Locke, autonomous citizenship rested on two
interlocking conditions. The first of these is that there be a public authority that
secures the freedom of individuals from the domination of other individuals.2 The
second of these is that the public authority itself be so constructed and monitored that
it functions impartially, without fear or favour, and, as we would put it nowadays, that
it adopt a principle of non-discrimination. Otherwise, if the public authority comes to
embody the private judgment of particular individuals or interests, it can be no longer
impartial, non-discriminatory and procedurally just.3 Instead, it will be all too likely
to represent private interests who use public power to advance their particular
interests and dominate those whom they consider inferior for one reason or another.
2
Locke makes it clear that freedom is not the liberty of individuals to do as they please
(choose) but is the freedom to live within and abide by the rule of law: ‘Freedom of
Men under Government, is, to have a standing Rule to live by, common to every one of
that Society, and made by the Legislative Power erected in it; A Liberty to follow my
own Will in all things, where the Rule prescribes not; and not to be subject to the
inconstant, uncertain, unknown, Arbitrary Will of another Man’ (Locke, para. 22: 284).
Philip Pettit (1998) elaborates the case for a liberal-republican polity that is so
constructed as to protect the freedom of especially vulnerable individuals from the
domination of others.
3
‘And thus by all private judgement of every particular Member being excluded, the
Community comes to be Umpire, by settled standing Rules, indifferent, and the same to
all Parties’ (Locke, para. 87: 324).
5
ANNA YEATMAN
In other words, autonomous personhood depends on government conceived as a
public authority. The social contract is an account of how the public authority is
conceived or designed so as to secure the rights of everyone to be an individual: That
is, a person who is conceived as the subject of his or her own life, who therefore has
the right to be treated as an autonomous centre of initiative (to use Levine's, 1995,
useful phrase). Private property right is valued but only as it secures individuality.
This point is easier to appreciate once we are talking about all individuals and not just
individual householders. When the classical liberal social contract theorists assumed
that the subject of freedom was the male head of a household, his individuality was
still embedded within the individuality of the household and that of his dependants. In
order to be an independent householder, the individual had to be a man of means, to
own property in land or something that counted as an equivalent. To free this
individual could seem to be the same thing as freeing up private property for marketoriented competition by getting rid of traditional and customary kinds of regulation
(see Laski's, 1936, historical account of the rise of liberalism). Thus, in the tradition of
liberal accounts of freedom, there is a profound confusion between the emergence of
individuality as a legally recognised unit of social action and the freeing up of private
property for modern market rules of competition. The consequence of which is that
instead of market action being required to serve individuality, individuality too often
is made to serve market action.
It is for this reason that in his historical narrative of the development of modern
citizenship, T.H. Marshall sees a tension between what he calls civil rights and social
rights. Leaving aside his gendered language, it is clear that for Marshall social
citizenship refers to the right of each individual to be accorded an ‘equality of status’
as a citizen. By this Marshall meant that each individual is conceived ‘of equal moral
worth’ and as entitled to live the life of a free person, that is someone who has access
to the civil, political and social rights that make self-determination possible.
In the historical context Marshall operated within, inherited class differences
combined with market-based inequalities of wealth to make it difficult for working
class people to get effective access to the civil and political rights of citizenship. For
instance, working class people did not have the means to buy the equal protection of
the law (civil rights) or to get the education that would build their capabilities for
freedom (Sen's language). In this context Marshall felt compelled to argue for social
rights as rights that would ensure working class people got to enjoy the same political
and civil rights as individuals of the propertied classes. The difficulty with this
argument is that it makes social rights necessary for the poor (the protection
conception again) but not for the propertied classes. Social rights thus become an addon to fundamental rights, rather than a fundamental constituent of the individuality of
all people. Not only do the propertied classes not need social rights (at best they are
an optional extra for them) but, as they will see the situation, it is the redistribution of
the income from their property that pays for the social rights of others.
We can understand why Marshall felt compelled to argue as he did. Social rights
justified the public provision of health, education, housing, and legal aid services.
Collective consumption of this kind could secure for individuals of the working class
access to the civil and political rights that propertied individuals had been able to
enjoy for some time.
6
However, Marshall argues as though equality were an add-on to freedom. This is
unfortunate, however necessary the argument was in its time. It is unfortunate because
it suggests that there is a conflict between the two values of freedom and equality
when in fact the opposite is true. Equality, in fact, is the value that secures the right of
all individuals to be free.
Freedom can be enjoyed only as it is practiced as a universal and inclusive value. That
is, freedom makes sense only as it is accountable to equality. Individuals can practise
their freedom to live their own lives as autonomous individuals only as they respect
this freedom for all others. If individuals cannot relate to other people as autonomous
individuals, but instead treat them as in some way instruments of what it is they want
to do and be, their individuality is symbiotic rather than autonomous.
When equality as seen as a supplement to freedom rather than as inherent in freedom,
it is all too easy for the social settlement Marshall celebrated in the name of equality
to be disestablished in the name of freedom. This is in fact what has occurred.
5
Freedom and Equality as Co-defining Values: the Emergence of
the Individual as the Unit of Social Life
How then do we develop social policy as functioning on behalf of the selfdetermination of all individuals where these two values of freedom and equality can
be seen as inherent in each other? Can we say that what made it so difficult for
Marshall to do this has changed in ways that might free us up to develop this
conception of social policy?
I think we can. Marshall (1977: 101) could see the way forward when he spoke of ‘the
conception of equal moral worth' and could see the difference between this idea and
the idea of natural right. The tradition of natural right ties freedom to householder
status (see Yeatman, 1994) whereas the idea of social rights opens up the possibility
of a personality that is not propertied in the householder sense.
What Marshall is groping for here is a post-liberal conception of individuality, but he
did not possess the intellectual and cultural resources to open up this conception. Do
we? I want to argue that we do. Firstly, the idea of human rights as I (Yeatman,
2000a) and others have interpreted it opens up a conception of a right to individuality
that belongs to each unit of humanity regardless of their age, gender, and degree of
property or lack of it. On this conception, property is not irrelevant, but it has to
follow upon what is deemed to be the human rights of each individual. Thus, the
argument may turn on what kind of economic security a child ought to have in order
to enjoy the freedoms a modern society offers.
This will not take us beyond the add-on conception of social rights or equality. There
is something else that has to happen. There has to be a disentangling of the
individuality of the one from the individuality of his or her others. Or to use language
I have already used, a symbiotic confusion between the personality of an individual
and that of those who in some sense depend on him (or, in these equal opportunity
times, her) has to become seen as unacceptable. The individual has to fully emerge as
the unit of social life.
7
ANNA YEATMAN
In this context we are able to turn our intelligence to reflect on the questions: what is
individuality? How does an individual come to be autonomous? What role do others
have to play in the development of an individual's autonomy? And, what role do the
institutions of the market and government have to play in the securing and
development of an individual's autonomy? We then can see the role of social policy in
providing for an institutional design for individuality.
6
Social Policy as Institutional Design for Individuality
Individuality means that each individual is valued and regarded as a centre of
initiative, and that rights, relationships, procedures, institutional practices and policies
are designed accordingly. Now many of us realise that an individual does not have to
be able-bodied, adult, or cognitively independent, for this individual to be accepted by
his/her relevant others as an autonomous centre of initiative. Nor do we find it
paradoxical that these others have to engage in action that makes it possible for the
child, or the intellectually disabled adult, or the physically disabled adult, to be
autonomous. For instance, in providing support for choice-making of this individual
or in providing assistance in getting around to an individual whose physical disability
makes it impossible for them to do this alone.
Individuality means also that an individual is secure enough in his/her relationships
with others to risk self-disclosure and thereby to explore who s/he is, a process that in
some respects is lifelong. It is only through the development of capacities for selfdisclosure, self-exploration, and self-reflection, that an individual can come to be
aware of what it is that s/he thinks or wants as distinct from others may think of want
for him or her (see Levine, 1995, Chapter Two).
What kind of institutional design is necessary for individuality understood in this
way? What does government as the public authority have to do in order to ensure that
each person is able to be an autonomous or self-determining individual?
The first thing that government in its role as the public authority has to do is to
constitute the rights of individualised personhood, and to secure for them
constitutional protection. We can term these rights the status of individualised
citizenship. Without a status entitlement to be an individual unit of social life, the
right of individuality does not exist.
If the status of individualised citizenship is to fully evolve, there is a good deal of
work to be done in developing a coherent conception of it. Moreover, it is important
that legislative specification of rights is not allowed to become an empty symbolic
gesture, and that a constitutional authority ensures that they are implemented.
Implementation also requires resources that are allocated to fund the services that
make such implementation possible.
The second thing that government has to do is to provide the infrastructure of
standards, professionalism, and services that underpin the recognition and the
development of individual capabilities for self-determination. We can term this aspect
of institutional design, the public provision of an infrastructure for human capability
building and preservation. Central to such an infrastructure is the provision of
education of a kind that cultivates in individuals the abilities to reason and think
autonomously, to listen to others and to one's inner being, to name and explore
8
feelings, to know what it is to take care of oneself, to respect the psychic and bodily
integrity of others, and to use imagination and thought in forming a conception of the
good.4
The third thing that government has to do is to ensure equality of opportunity with
regard to participation in the economy. Government can ensure that there are no
discriminatory barriers to market participation and to other kinds of socially valued
work, and that individuals have access to the training and labour market preparation
that enables them to enter the market economy or to pursue non-market-oriented
vocations.
Fourthly, the government has to provide what Levine calls economic rights. He
defines economic rights as ‘entitlements (to goods or money primarily) other than
those we claim by claiming our property right; rights to work or to income whether
we work or not, rights to welfare when we have no resources of our own' (Levine,
1995: 101). Levine argues that if welfare is not a matter of right, then it necessarily
becomes a matter of charity.
Charity … places the recipient in a subordinate and dependent
position because the initiative rests with the giver. It denies the
agency and autonomy of those in need. Welfare rights shift the
balance. The recipient can, in principle, demand his or her
rights. (Levine, 1995: 100)
Fifthly, the government has to model procedures, protocols and processes that
facilitate the recognition of autonomous individuality in the conduct of relationships.
There is a host of such practical norms and techniques for what we might broadly
term participative governance. These are norms and techniques for the conduct of
relationships that facilitate individual voice and choice within them, and when
necessary, exit from them ( these terms are Hirschmans, 1970).
The Five Aspects of the Role of Government in the Constitution of
Autonomous Individuality
4
•
The constitution of the status or rights of individuality, and the provision of
constitutional protection for these rights and their implementation
•
The provision of an infrastructure of individual capability building and
preservation
•
The provision of equal opportunity for participation in the economy and
socially valued work
•
The provision of economic rights (including the right to welfare)
•
The provision of norms and techniques for the participatory conduct of
relationships.
I am indebted for this last idea and for some other suggestions here to Nussbaum, 2000:
78-81).
9
ANNA YEATMAN
7
Freedom and Property Rights
T. H. Marshall was ambivalent about how to set up social rights. Are they in
contradiction to the civil and political rights that have been already historically
established? Or do civil and political rights need to be reconceived so that they
constitute along with social rights a coherent bundle of rights?
The problem he had is the problem we have: how to situate property rights in relation
to citizenship. Property rights ‘are the right to buy, sell, and use’ (Levine, 1995: 101).
If we made economic rights in Levines sense more foundational than property rights,
then we would conceive and limit property rights accordingly. They would have to
operate so that they do not undermine economic rights.
If we developed this argument, we can propose that individualised citizenship is to be
taken as both prior to and constitutive of market freedom. It follows that the exercise
of market freedom must not undermine individualised citizenship. Rather the exercise
of market freedom must be consistent with individualised citizenship.
This brings out the complexity of the individualised citizen’s identity. The individual
is both a citizen-individual and a market-individual. That is, his individual interest
turns out to be two different kinds of interest, his public interest as an individualised
citizen, and his private interest as a market-individual. These two interests can readily
conflict. The question turns on whether individuals are willing to be aware of this
conflict and to take responsibility for resolving it.
8
Reprise and Finale
You will recall that I proposed that classical liberalism joins property to propriety or
personality, the implication being that property right is to function on behalf of
individual autonomy. Classical liberals got it right even though we want of course to
revise their limited conception of who can get to be an individual. However, the
dilemma of modern capitalist societies is that property rights acquire a power of their
own. It is easy for wealthy individuals to forget that they owe their freedom of market
action to government, and to claim instead a freedom for their privately interested
action from government interference regardless of the harm it causes others and the
environment. In this context, it is easy for governments to be persuaded that the best
course of action is to design policy not so that property right functions on behalf of
individual autonomy but so that individuals function on behalf of property right. This
is not a freedom that all can enjoy, and therefore it is not a conception of freedom that
satisfies the test of ethical universality. Instead it authorises a social policy that
marginalises the poor and ‘undermines their integrity and their sense of full and equal
citizenship' (Levine, 1995: 100). If we are to call this situation into question we have
to re-develop our capability to discuss what is wealth for. If we agree that the goal of
wealth is the freedom of individuals to be autonomous and secure, then we must
develop a conception of individualised citizenship that is reconcilable with human
rights. This would offer a powerful vision for social policy.
10
References
Freeman, M. (1997), The Moral Status of Children: Essays on the Rights of the Child,
Martinus Nijoff, The Hague.
Hirschman, A. (1970), Exit, Voice and Loyalty: Responses to Decline in Firms,
Organizations and States, Harvard University Press, Cambridge, Massachusetts
and London, England:.
Laski, H. (1936), The Rise of Liberalism, Harper and Brothers, New York and
London.
Levine, D. (1995), Wealth and Freedom: An Introduction to Political Economy,
Cambridge University Press, Cambridge.
Locke, J. (1960), Second Treatise of Government, in Two Treatises of Government,
ed. Peter Laslett, Cambridge University Press, Cambridge.
Marshall, T.H. (1977), ‘Citizenship and social class,’ in Class, Citizenship and Social
Development, University of Chicago Press, Chicago and London.
Nussbaum, M. (2000), Women and Human Development: the Capabilities Approach,
Cambridge University Press, Chicago and London.
Pettit, P. (1997), Republicanism, Cambridge University Press, Cambridge.
Pocock, J.G.A. (1985), Virtue, Commerce and History, Cambridge University Press
Cambridge.
Polanyi, K. (1944), The Great Transformation: The Political and Economic Origins
of Our Time, Beacon Press. Beacon Hill, Boston.
Sen, A. (1999), Development as Freedom, Alfred A. Knopf, New York.
White, S. (2000), ‘Review article: Social rights and the social contract - political
theory and the new welfare politics’, British Journal of Political Science, 30,
507-32.
Yeatman, A. (1994), ‘Beyond natural right: the conditions for universal citizenship’,
in Postmodern Revisionings of the Political, Routledge, New York and London.
Yeatman, A. (2000a), ‘Who is the subject of human rights?’, American Behavioural
Scientist, 43(9), 1498-514.
Yeatman, A. (2000b), ‘What can disability tell us about participation?’ Law in
Context, 17(2), 181-203.
Yeatman, A. and K. Owler (forthcoming), ‘Contract and democratic service delivery’,
Law in Context.
11
The US Vision of Work Based Reform:
Promise, Prospects and Pitfalls
David T. Ellwood1
Harvard University
1
Introduction
The United States has undergone an almost revolutionary change in social policy in the
past decade. And so far, the public and many policymakers seem quite pleased with the
results. Public aid is down, work is up, poverty has fallen. There has even been a slowdown in the formation of solo parent families. Most of the attention has centred on a
single legislative act popularly known as ‘welfare reform’, but the truth is that the nature
of the change is more profound than that single legislative act. And the future risks are
much greater than is commonly understood as well.
In the terms of this conference, the US has adopted one vision: an almost total
preoccupation with market work. Social policy has been altered to drastically reduce
support and add administrative pressures for healthy individuals who do not work, even if
they are lone parents. Simultaneously, economic support for working families has
expanded considerably, effectively raising earnings and pay of parents who do work
outside the home.
This paper opens by briefly describing the nature of the policy changes that actually
occurred in the US over the past decade. It then examines the social, economic, and
political forces that led to those changes. Next, I move to a discussion of how these new
policies have changed incentives and what impacts they have had on behaviour. Finally
the paper examines the future risks and opportunities of this approach.
I have not had the opportunity to look in detail at the current structure of Australian
policies or recent reform initiatives; thus, this paper will primarily be US-centred, with
only a modest amount of international comparison. I have recently completed work on
the UK, and I will make some comparisons between America and that nation to illustrate
the main themes of this paper.
2 Social Policy in America: Before and After
Until the mid-1990s, healthy working age adults in the US who were not working could
1
I would like to thank Andrew Leigh for helpful information he provided and Pamela Metz
for her thoughtful comments.
Ellwood, D.T, (2002), ‘The US vision of work based reform: promise prospect and pitfalls’, in T. Eardley
and B. Bradbury, eds, Competing Visions: Refereed Proceedings of the National Social Policy Conference
2001, SPRC Report 1/02, Social Policy Research Centre, University of New South Wales, Sydney, 12-34.
DAVID T. ELLWOOD
have been covered by up to three layers of federally sponsored protection.
Persons with recent work experience who had been laid off their job due to economic
conditions could receive up to six months of Unemployment Insurance (UI). Poor singleparent families with children and some two-parent families which did not qualify for UI
could receive aid indefinitely under a means-tested program called Aid to Families with
Dependent Children (AFDC). And virtually anyone who was poor could receive Food
Stamps - coupons which could be used to purchase food.2
This system was significantly less generous than the Australian and most European ones.
Most non-disabled adults, even most two-parent families with children, could only
qualify for six months of unemployment insurance, and then only if they had previous
work experience and did not leave their job voluntarily and were not fired for good cause.
The benefits were not means-tested; they were only linked to personal unemployment,
but the restrictions on eligibility and the limits on duration meant that for much of the
1980s and 1990s, less than 40 per cent of those officially classified as unemployed
actually qualified for benefits. Beyond UI, food stamps was the only major source of
support available to most low-income families, and it provided benefits only in the form
of coupons which could be used to purchase food. The US has no universal children’s
allowance, no federal means-tested support for lone adults (though some states have
programs), no universal housing program (though here too there is housing aid available
in some cases), no guarantee of health coverage.
The only groups that got federally sponsored cash aid of unlimited duration generally
were lone parents with children, who were covered under AFDC, and the aged and
disabled, who were covered both with a social insurance program (Social Security) and
means-tested benefits (Supplemental Security Income).
The AFDC system was a strange federal-state hybrid. The federal government
determined what types of people were eligible - mostly single parents - and how their
income would be ‘counted’, but states were free to set benefits at any level they chose,
and states were responsible for administering the system. In 1995, for example,
Mississippi paid $120 per month for a family of three with no earnings, while
Connecticut would have paid the same family $636. (Committee on Ways and Means,
1996, Table 8-12: 437) The costs of whatever money was paid in benefits was shared
between the states and the Federal government. There was no limit on the duration of aid,
and no person who met the eligibility criteria in the state could be denied aid.
It was this relatively narrow public assistance system that became known in the public
mind as ‘welfare’. For reasons unknown to me, the American lexicon has converted this
perfectly wonderful word ‘welfare’ into a synonym for a particular form of means-tested
aid, and added a large measure of disdain to go with it. I’ll return to the disdain issue in a
moment, but for now, accept that when I talk of ‘welfare’, I really mean means-tested
support for single parents.
2
Anyone, that is, except strikers and students.
13
THE US VISION OF WORK BASED REFORM
3 Welfare Reform Arrives
Starting in the late 1980s, states began to use new flexibility that allowed for
experimentation to explore new initiatives ranging from voluntary training programs to
mandatory work requirements. By 1996, over half the states had gotten some form of
federal ‘waiver’ from national law to conduct some form of welfare reform. Finally, the
Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA) was
passed and signed in the summer of 1996 and labeled ‘welfare reform.’
‘Welfare reform’ abolished the Aid to Families with Dependent Children (AFDC)
program and replaced it with the Temporary Assistance for Needy Families (TANF)
block grant. States were now given a fixed amount of money, regardless of their
caseloads and they could largely spend it on anything ‘consistent with the purposes of the
act’. States were required, however, to either get half the aid recipients working, or
reduce caseloads equivalently. Time limits were imposed so that only 20 per cent of the
caseload could receive federal aid for more than five years.
Starting in 1992, the AFDC and later TANF caseloads began to plummet. Figure 1 shows
just how dramatic the fall has been. Caseloads are down nearly 60 per cent since their
peak in 1994. In some states, the fall has been closer to 80 per cent. The striking feature
of these drops is that they come well before one the most visible and controversial
elements of the reform had even begun to bind: few recipients have hit time limits yet.
Figure 1: Number of Cases on AFDC and TANF By Year
6 000 000
5 000 000
4 000 000
National Welfare Reform Passed
3 000 000
September 2000
2 000 000
1 000 000
0
1960
1965
1970
1975
1980
14
1985
1990
1995
2000
DAVID T. ELLWOOD
The economy in the US has been exceptionally strong, of course. But it is hard to believe
that the economy alone accounted for the lion’s share of change. Recall these are almost
entirely solo mothers, a group which is not likely to be highly sensitive to the economy.
And on Figure 1, the caseload did not vary dramatically over the economic cycle.
What really happened to the welfare system? Since the states were given almost complete
control, there is limited information available at this stage.
States clearly did change their economic incentives and a few altered benefits. But many
of the changes have been in what might be called ‘the culture of welfare offices’. Some
states have dramatically increased pressure to move off welfare. Their methods have
generally not involved dramatically lowering benefits. Rather they have sought other
means to divert people from getting aid or to move people off welfare quickly.
Consider two examples. In Georgia, before she can even begin the application process, a
woman seeking aid is required to get a form signed by six employers saying that she
applied in good faith for a minimum wage job and was turned down. (LaDonna Pavetti,
personal communication) Once enrolled, if she is penalised twice for failure to meet some
key administrative/work requirement, she is barred for life from seeking aid in the state.
In Wisconsin, no aid is provided unless the person is already working. When applicants
claim they really cannot find a job, the providers of TANF will, in some cases, provide a
subsidised job for a limited duration, but aid remains tied to working. In both states,
caseloads have dropped dramatically - nearly 80 per cent in Wisconsin.
It is probably not too much of an exaggeration to say that the US is effectively shutting
down the system of means-tested support for single parents.
4 Increased Support for Working Families: The Untold Story
In the popular mind, the change in means-tested aid for solo parents is virtually the entire
story of social policy reform. Indeed, even many administrators and scholars seem to
focus the bulk of their attention there. But the story in the US is just as much about rising
supports for parents who work outside the home.
Even while welfare caseloads were falling, though, supports for low-income working
families have expanded dramatically. Figure 2 shows that the aid available to low-income
working families jumped from about $11 billion in the late 1980s to over $70 billion in
real dollars.
About half of this growth can be traced to dramatic expansions in the Earned Income Tax
Credit (EITC). The EITC is an earnings credit with benefits tied to level of earnings and
the number of children. For each dollar earned up to a maximum, the parent gets a
refundable tax credit from the government. For a parent with two or more children, the
credit is currently 40 per cent of earnings up to a maximum credit of roughly $3800 roughly triple the maximum for such families in 1992. For a minimum wage worker, this
is roughly the equivalent of a 40 per cent pay raise.
15
Figure 2: Federal Outlays on Low Income Families Under Selected Programs
(Billions of 1999 Dollars)
$80
Children's Health Insurance Program
$70
Medicaid
$60
$70.2 billion
Child Care
Child Tax Credit
$50
Earned Income Tax Credit
$40
$30
$20
$11.0 billion
$10
$0
1988
1992
But as a family’s income rises above $13 000, the credit is gradually reduced (at a rate of
21 cents per additional dollar earned), and the credit is fully phased out for families with
incomes over $30 000. To put this in perspective, in 1995, the year before the passage of
PRWORA (and a peak year in welfare spending), the federal government spent a total of
$17.3 billion (1999 dollars) on the old Aid to Families with Dependent Children (AFDC)
program. The EITC increase alone was nearly as large as the total amount the federal
government ever spent annually on AFDC.
Another big change has been medical coverage. The US has nothing like a system of
universal coverage for health costs. Until recently, two groups had the most reliable
coverage: middle and upper income workers who got coverage through their employer,
and non-working cash aid recipients who got it automatically with AFDC or the program
for the aged or disabled. Low-wage workers often had no private coverage and were not
eligible for public coverage either. States are now required to provide coverage for all
children aged 18 or younger with family income at or below the poverty line. Many states
have chosen to cover children who are older and whose families are considerably above
the poverty line.
Still another area of expansion is support for child care. States were given more money,
allowed to use their new welfare block grant money for child care, and given much more
flexibility over its use.
Major new child tax credits have been passed. They now equal $1000 per child. But they
have been non-refundable and thus of no value to poor families who owe no taxes. But
just this month, the new tax cut bill passed by the Congress and signed by President Bush
made these credits partially refundable for families with earnings above $10 000.
16
DAVID T. ELLWOOD
Dramatic as these figures are, they surely understate the magnitude of increases in
spending for low-income working families. At least 10 states, including New York,
Maryland, Massachusetts, Minnesota, and Wisconsin, have EITC programs that add
between five and 25 per cent more credit on top of the federal one (Johnson and Lazere,
1998). And with welfare reform in 1996, states are now free to redirect their
AFDC/TANF monies to support of working families, and many have done so.
In sum, the change in social policy in America has not been a story of draconian cuts in
support for those who are not working outside the home. It is a story of drastic
redirection. Aid is being withdrawn from non-workers and added to workers. On net, the
nation is actually spending far more than before on low-income families, but it is
spending it very differently.
5 What Led to the Work Revolution in Social Policy?
To some observers, the change in policy was a morality play. Race has always been
perhaps the ultimate political cleavage in the US. Bigotry, a general disdain for the poor,
and a distrust of government seem nowhere in better focus than in the AFDC program.
The debate itself was often ugly and mean. In the Congress, welfare recipients were
occasionally compared to alligators and monkeys. The Democrat who had chaired one of
the most powerful committees in the House prior to the 1994 elections finally yelled,
‘Shut-up, shut-up all of you’, on the floor of the House, an event that made the evening
news across the country. In an article in the Atlantic Monthly entitled The Worst Thing
Bill Clinton Has Done, Peter Edelman, one of three senior administration officials to
resign in protest over Clinton's signing of the welfare reform bill, argued that Clinton
hastened the collapse of AFDC with his incendiary rhetoric about ending welfare and that
he ultimately sold out poor children (Edelman, 1997). Some estimates suggested that as
many as one million more children would be poor as a result of its provisions (Zedlewski
et al., 1996). Meanwhile, in the eyes of many political observers, by signing the bill he
assured his reelection (Morris, 1997).
The short-term political analysis of these particular events begs the larger question of
what led welfare to be so unpopular in the first place. Poll data showed that well before
the 1992 elections the vast majority of Americans felt that ‘welfare’ was broken. Only
one Senator up for reelection voted against the bill: Paul Wellstone, a man who is
unquestionably the most liberal member of the Senate and who represents what is
probably the most liberal state in the country. Even Wellstone had such a hard time
justifying his vote back home during his reelection fight (he was labeled ‘Senator
Welfare’ by his opponent) that he finally rescued his campaign with ads saying, in effect,
‘I know many people were disappointed with my vote on welfare and that this has hurt
me, but I had to stand by my principles’.
6 The Origins of Discontent
The evaporation of political support actually originates in part from the political and
institutional forces which first lead to the creation of AFDC. The program began during
17
the New Deal era of the mid-1930s. Edwin Amenta, in a fascinating book on the origins
of the American version of the welfare state, emphasises the critical role that American
institutions played, including the structure of government and origins and orientation of
the policymakers, the configuration of political forces in Congress, and the impact of
social protest movements and capitalist organisations. Yet even in his heavily
institutional story, AFDC stands out in its uniqueness. In contrast to UI with its strong
labour support, ‘the [AFDC] program did not have any movement of organized group
outside the state demanding it. Even labor stayed out of the fight’ (Amenta, 1988: 115).
The program was largely designed and pushed through the efforts of the Children's
Bureau, a long-standing government agency. ‘...Congress was more concerned with
“other programs and” ... A[F]DC's supporters were eager to avoid scrutiny’ (Amenta,
1988: 87).
Policymakers seemingly designed the program to be sold quietly as additional federal
support for an existing set of state programs supporting children, often through widows
pensions. Thus, from the start the program mostly offered federal financial backing for
state efforts. There was no serious push for the program to be uniform nationally, and a
conscious attempt to keep its visibility low. Throughout the entire saga of AFDC, this
lack of serious movement support, the control of policy experts, and the desire of
supporters to keep the visibility of reforms low is repeated over and over.
Still AFDC did not become a powerful political target until recently. Katz (1989) has
argued that means-tested cash benefits are limited to the ‘deserving’ poor: mostly people
who are unable to work through no fault of their own. Low-income aged and disabled
persons fit easily into this category, and Supplemental Security Income (SSI) is a
federally crafted system that provides them with means-tested aid. Single parents could
also be treated as ‘deserving’ poor back in the 1930s when AFDC was first crafted. Most
single mothers were widows, and with relatively few married mothers working in the
labor market, it seemed reasonable to allow support for single parents to remain home to
nurture their children.
But the nature and magnitude of single parenthood has changed radically since the 1930s.
By 1960, widows were already a minority of all single parents. In the 60s and 70s divorce
grew rapidly among parents. Figure 3 shows that starting in the early 1980s, divorce
leveled off; since 1980, there has been a large increase in the number of never married
women with children. Today over 30 per cent of families with children are headed by one
parent, almost always a woman.
At the same time, work patterns of married mothers changed. Figure 4 shows that during
the early 1960s only about a quarter of married mothers worked outside the home, while
over half of all single mothers were employed in the labour market. But over the
subsequent decades, employment of married mothers rose sharply, while work patterns of
single mothers remained stable. By the mid-1980s, employment rates of married mothers
exceeded those of single mothers, and by the late 1990s, nearly two-thirds of married
mothers were working at least part time.
18
DAVID T. ELLWOOD
Figure 3: Families with One Parent as a Percentage of All Families with Children
By Marital Status of Lone Parent
Precentage of familes with children
35
30
25
Never Married
20
15
Separated or Divorced
10
5
0
1965
Widowed
1970
1975
1980
1985
1990
1995
2000
Year
Figure 4: Proportion of Mothers Working by Marital Status
0.8
Proportion Working
0.7
0.6
Unmarried W omen W ith Children
0.5
0.4
M arried Women With Children
0.3
0.2
0.1
0
1960
1965
1970
1975
1980
1985
19
1990
1995
2000
The changing pattern of maternal work and the altered reasons for becoming a single
parent undoubtedly undermined the public perception of the ‘deserving’ status of single
mothers. Adjusted for inflation, welfare benefits and eligibility have actually been falling
since the 1970s. Real benefit levels in the median state have been cut in half since 1970
(Committee on Ways and Means, 2000). Indeed, Figure 5 illustrates that while the
absolute numbers of households on AFDC was flat or rising until the recent sharp
declines, the fraction of single mothers getting AFDC or TANF actually declined almost
continuously since 1973 except for a brief interruption during the early 1990s.
Figure 5: Cases on AFDC or TANF as a Fraction of Mother Only Families
0.900
0.800
0.700
0.600
0.500
Welfare reform passed
0.400
0.300
0.200
0.100
0.000
1960
1965
1970
1975
1980
1985
1990
1995
2000
By the 1990s, the notion that welfare should be a source of ongoing support for single
mothers who are at home with children was extremely unpopular.
In 1993 an overwhelming 93 per cent of voters favored requiring welfare recipients to
work (Garin, Molyneux, and Divall, 1993). Even a favourably worded statement: ‘since
young children are better off with their mothers, we should enable welfare mothers to
stay at home and care for them’ could garner even mild support from just 32 per cent of
Americans (Farkas, and Johnson, 1996).
Such findings may shock audiences who would understandably argue that being at home
nurturing children is so critical that it ought to be supported by the public sector. But in
Katz's language, single mothers on welfare had slid from the deserving into the
undeserving poor, and a highly visible group at that.
One particular vulnerability of the program was its overwhelming focus on poor single
mothers. It seemed to many Americans, even many mothers, that other women were
20
DAVID T. ELLWOOD
expected to go to work and pay taxes to give single mothers the chance to stay home with
their children, a luxury that other mothers could not afford.
Whatever connections had existed between various political forces and AFDC recipients
evaporated. Women's groups, made up chiefly of working women, were ambivalent at
best. AFDC was no longer a civil rights issue. Indeed, an increasingly vocal minority
argued that by keeping people out of the labour force and ‘dependent on the dole’,
welfare actually contributed to both perceptions and economic realities of minority
isolation from society. Labour had never been a particularly strong supporter of AFDC,
and when single mothers looked like unemployed workers with far better benefits,
organised labour was unlikely to provide much real support.
All that remained to hold AFDC in place was a small group of deeply committed
legislators and policymakers in key positions in government - committee chairs in the
Congress, program administrators in the Executive - who up until this time were able to
defeat increasingly vocal and often spiteful and bigoted calls to radically alter it, while
attempting to steer reform in a more moderate direction. The Republican sweep to power
all over the US eliminated those committee chairs, and administrators were often gone.
Combined with the already harsh words from a Democratic president, political support
for AFDC seemed to vanish entirely.
7 Support for Work
But the frustration with AFDC did not translate into frustration with all poor families or
even all single parents, only those who were not willing to provide some part of their own
support. One of the most fascinating legislative juxtapositions in recent memory came in
August of 1996 in the last legislative month before the fall elections. Republicans were
feeling politically vulnerable, and the President wanted to avoid the humiliation of two
years earlier. Political expediency was in the air. On Tuesday, 20 August, President
Clinton signed a bill to raise the minimum wage, an idea traditionally fought by
Republicans. On Thursday, he signed welfare reform. To many advocates, academics,
and observers of Congress, these bills seemed to indicate almost contradictory impulses.
To most citizens, both were an indication of the importance of emphasising and
supporting work.
By transforming the argument from whether to give more or less cash aid into an
argument about moving people from welfare to work, some of the meanness that has
surrounded this debate for years has been at least temporarily removed. A front-page
story in the New York Times opened with, ‘Welfare recipients, last year’s political
pariahs, are shedding their outcast status.’ The article recounts that with falling caseloads,
a strong economy, and a new focus on work, ‘...legislators have found themselves less
guided by ideology than by a new, nuts and bolts practicality’ (DeParle, 1997).
Plainly the moral legitimacy of a time-unlimited cash for non-working parents was
eliminated by the changing nature of family structures and the patterns of mother's labour
market work.
21
But the moral legitimacy for aid for working single parents was heightened with the new
focus on the enormous pressures they face in trying to nurture and provide for their
children.
It appears that most future efforts in the US at supporting the poor will be much more
closely linked to work, designed to move people quickly into work and support lowincome workers with tax credits, health coverage, and child care. That strategy is more
politically popular, in large part because it is more consistent with a set of American
values focused on work, independence, and responsibility.
By converting the welfare poor into low-income working families, a very different set of
coalitions emerged: low-income workers connected automatically with union workers
and working people at all incomes who fear the markets stresses and insecurities. And the
potential for protections which spread to a wider part of the population rises. There is not
much logic in providing supports to working single parents while ignoring the struggles
facing two-parent families.
The focus on the work-related struggles of the poor holds out the hope that the vindictive,
us versus them rhetoric which used to be common in the US will diminish, and the nation
will be forced to face the real day to day struggles of a group who cannot be accused of
not playing by ‘society's rules.’
Yet there appeared to be limits to how far the nation’s politicians were willing to go to
support work. The welfare reform bill that emerged was quite different from the one
Clinton originally proposed.
Those who designed Clinton’s original reform plan (including this author), introduced in
1994 but never passed, included time limits of a different sort. After people reached a
two-year time limit, they were guaranteed a subsidised job. This provision was seen as
necessary to ensure that when people could not find a job, they would still have a way of
supporting themselves and their family. This element had always been a part of the
original Clinton vision of work-oriented reform. Polls suggested such a plan would be
popular, but even within the administration there were doubts about the logic of such
publicly supported jobs. And when Republicans took control of the Congress in 1995, the
guarantee of a job for anyone who hit a time limit and could not find one was jettisoned
without any fight at all from the President.
So strong supports were enacted to help working families, cuts were made in aid to nonworking families, and no provision was made for people who could not find jobs.
Obviously significant risks attend such a strategy. Would single mothers be able to find
work? Would some be left destitute?
8 New Incentives, Altered Behavior
The cuts in welfare aid and expansions in support for working families inevitably
changed work incentives quite radically, particularly among single parents. For 1988 and
1999, Table 1 indicates how much a single mother of three could expect to receive if she
22
DAVID T. ELLWOOD
did not work and collected welfare versus if she worked full time at the minimum wage.
Welfare benefits vary state to state, and child care expenses vary across people, but the
table gives a sense of how a typical person might fare.3
The payoff to working has increased dramatically since 1988. A full-year full-time
minimum wage single parent who would otherwise be on welfare in 1988 would have a
net gain of only $2325, and she would likely lose her Medicaid benefits, which might
easily be worth more than that gain. By 1999 the same woman would gain over $7000 by
working, the children would keep Medicaid, and even the woman would be eligible to
keep it for a time.
Table 1: Income and Benefits for a Single Mother with Two Children When
Working and Not Working in 1988 and 1999
1988
1999
Working Full
Not Working Time Minimum
Wage $
Working Full
Time Minimum
Wage $
Not
Working
Total Earnings
0
$9813
0
$10 712
Payroll Taxes
0
- 737
0
- 819
$8612
$2630
$7967
$2310
0
$1231
0
$3816
Child Care Expense
0
- 2000
0
- 2000
Child Care Support
0
0
0
1000
Disposable Income
$8612
$10 937
$7967
$15 018
Yes
No
Yes
Children Under
16
TANF(AFDC) and
Food Stamps
Earned Income Tax
Credit
Government Paid
Health Insurance?
(Medicaid)
Net financial gain
from working
3
$2325
$7051
Welfare benefits are based on the weighted average of state benefits, weighted by the share
of all single parents who reside in the state.
23
A higher minimum wage and slightly lower welfare benefits contributed to this gain, but
it is mainly the result of a sizable rise in the EITC and greater child care support.
Table 1 understates the real change in incentives for work among single parents. By 1999
revised welfare programs were putting increasing pressure on welfare applicants and
recipients to eschew welfare and enter the workforce. Thus the option to not work and
collect welfare was increasingly unavailable.
Let me offer an alternative way to examine the changing patterns of aid. Ideally one
would like to explore how much aid persons in identical economic circumstances would
qualify for over time. Benefit levels alone are not very helpful since administrative
practices influence who gets aid, and these clearly have changed. But one can observe the
amount of combined aid persons at different levels of earnings actually received. One can
determine the average amount of combined aid persons with $0 earnings received;
similarly for those with $1-7500 in earnings, and $7501-15 000, and so forth. If, for
example, administrative procedures have tightened or stigma has increased, persons with
a given level of need may be less likely to get aid and thus will be observed to receive a
lower benefit.
Figure 6 offers this alternative method of looking at the combined impact of various
programs on single parents with different levels of earnings.
Figure 6: Benefits for Single Adult-headed Households in the US (Non-disabled,
Non-widowed) By Annual Earnings Category
$14,000
$12,000
Annual Benefits
$10,000
No Earnings
$8,000
$6,000
$4,000
$1-7500 Earnings
$7501-15 000 Earnings
$2,000
$15 001-$30 000 Earnings
$0
1978
1983
1988
Year
24
1993
1998
DAVID T. ELLWOOD
Average annual benefits received by those with zero earnings have fallen precipitously in
recent years, moving from an average of over $7000 to less than $5000. By contrast, aid
for those in the $7500-15 000 category, after falling in the early 1980s during the Reagan
era cuts, have grown dramatically in recent years from roughly $1500 in the mid 1980s to
nearly $3500 today. This latter change is almost entirely the result of expansions in the
EITC. Indeed, the difference in actual average benefits reported between those with zero
earnings and those in the $7500-15 000 category has narrowed from $7000 in the mid1980s to only about $1000 today.
And note that this analysis still does not include Medicaid expansions for low-income
working families, which would lower the gap even further.
With such dramatic changes in incentives, one would expect work patterns of single
mothers to change. And indeed they have, especially for women at the bottom. It is
helpful to classify all women into thirds based on what we would predict they would earn
if they went to work. In other words, highly educated women who could command a
better job and higher pay would be in the top third. Those with weaker educations would
be in the bottom third. We can then track these groups over time. In general, we would
expect single mothers whose education would have put them in the top third of wages to
work quite a bit. Mothers in the lower third would be expected to work less and instead
rely on benefits. What we are interested in, though, is changes over time. If the changed
incentives really mattered, in recent years, we should expect to see single mothers in the
lower third to work much more than they used to.
Figure 7 illustrates the level of work of single mothers by predicted wage quartile. After
roughly 15 years with almost no change in work patterns, starting in the early 1990s with the advent of welfare reform, sharp expansions in work supports, such as the EITC,
and a very strong economy - work by single parents began to rise sharply. And the
greatest increases have been among single parents in the lowest quartile. Their work
levels have risen from 40 per cent to 60 per cent.
These are unprecedented changes in behaviour. Of course, other factors might be at play,
such as the strong economy. There is a rather large and rapidly growing literature in the
US that examines the impact of these dramatic policy changes on work by single mothers,
which is nicely summarised by Hotz and Scholz (2001). They universally credit the
combination of policy changes with being the major influence on work.
This seems rather compelling evidence that dramatic changes in incentives can affect
behaviour. There are good data in the United Kingdom that allow one to take a similar
look at benefit and work patterns over time (All of these figures are taken from Dickens
and Ellwood, 2001). Figure 8 shows combined benefits for single parents in Britain. Note
that the data are based on weekly rather than annual earnings and benefits. The scales on
Figures 6 and 8 are chosen to be roughly equivalent after adjusting for exchange rates and
annual versus weekly periods so one can compare the level and trend in benefits in the
two nations simply by comparing the figures.
25
Figure 7: Percentage Working Among Singles with Children in the US by Predicted
Wage Third
100%
90%
Predicted Wages in Top Third
80%
Percent Working
70%
Predicted Wages in Middle Third
60%
50%
40%
Predicted Wages in Bottom Third
30%
20%
10%
0%
1978
1983
1988
1993
1998
Year
Benefits for single parents who are not working are much higher in the UK than in the
US, even though wages are considerably higher in America. Moreover, the pattern over
time in benefits is the opposite of the US. In the United Kingdom, benefits for nonworking lone parents have risen significantly over time.
If benefits for not working are higher and the pattern of change is the opposite of the US,
then work should have followed a different pattern. Figure 9 shows what happened to
work by single parents in Britain.
Also in contrast to the US, work by single parents has fallen over time in Britain as one
might have predicted. Still, the timing of declines in work does not mimic the timing of
changed incentives. However, in the UK the economy improved dramatically in the
1990s, so work would have been expected to rise. The fact that it did not probably
reflects the competing effects of reduced work incentives.
The comparison between the US and Britain should serve to highlight something more
than the notion that incentives seem to matter. The UK is trying to create a more workfocused set of reforms themselves. But according to Blundell et al. (2000) these have not
altered incentives much. In part this reflects the fact that raising benefits to workers often
leads to reduced payments for housing, effectively nullifying the effect. But the issue runs
deeper.
26
DAVID T. ELLWOOD
Figure 8: Benefits for Singles with Children in Britain by Weekly Earnings
Category
£180
£160
No Earnings
Weekly Benefits
£140
£120
£1-150 Earnings
£100
£80
£60
£151-300 Earnings
£40
£20
Over £300 Earnings
£0
1978
1983
1988
1993
1998
Year
Figure 9: Percentage Working Among Singles without Children in Britain by
Predicted Wage Third
100%
P re d ic te d W a g e s in T o p T h ird
90%
80%
Percent Working
P re d ic te d W a g e s in M id d le T h ird
70%
60%
50%
P re d ic te d W a g e s in B o tto m T h ird
40%
30%
20%
10%
0%
1978
1983
1988
1993
Year
27
1998
Clearly the US has successfully raised work among low-income families, notably single
parents. But the differences in the US and UK benefit systems are enormous. Single
parents in Britain with zero earnings get benefits equivalent to just 62 per cent of the
relative poverty standard. The US pays just 19 per cent! Two-parent families and single
adults without children get even less. To mimic the financial work incentives in the US,
benefits for non-working families would have to be cut enormously for all families while
maintaining aid for working ones. And, of course, cuts in benefits for non-workers will
surely raise poverty or increase hardship among those with little or no earnings.
I have not had the opportunity to do similar charts for Australia. But I suspect the above
conclusion holds. Whiteford (2000) shows that benefits for non-workers are much higher
than in the US. And these benefits are taxed away rapidly. Lowering benefits to the US
levels would raise serious questions about poverty and well-being. Reducing the effective
tax rate on earnings as much as the US has would be extremely expensive. This is not to
say that improvements in work incentives cannot make a difference. But anyone who
aspires to the kinds of work increases seen in the US needs to recognise how different the
benefit systems really are.
9 A Strong Beginning, A Risky Future
Many observers worried about the dangers of the new policies, particularly the welfare
reform elements designed to discourage use of means-tested aid and limit the duration of
help.
Opponents feared that even in good times, many of these mothers may not be particularly
employable. There were also fears that a major influx of welfare mothers could not be
absorbed into the economy. And even if the women were employed, several studies cast
doubt on whether they would be able to support themselves and whether their wages
would actually grow over time (Burtless, 1997).
One study cited above predicted up to a million more poor children after the time limits
were enforced. So far the news has been far less discouraging than many had expected.
Poverty (measured the US way as whether or not a family has income below a fixed
standard which varies by family size) is down.
While overall poverty declined slightly between 1993 and 1999, poverty rates among
blacks and Hispanics, as well as among female-headed households (those most affected
by welfare reform), declined more rapidly, and all three groups experienced their lowest
recorded poverty levels in 1999. This suggests that, at least in the short run, the altered
social policies generally improved the economic situation of households.
Schoeni and Blank (2000) estimate that the social policy reforms reduced poverty rates
by two per cent. But this is hardly definitive. Time limits have not been hit yet. The
hardest cases probably still remain on welfare, and if they are pushed off, the
consequences may be far more worrisome. Many people may be able to make short-run
accommodations that will fall apart in the long run. And most importantly, the reforms
28
DAVID T. ELLWOOD
just happen to coincide with the strongest US economy in almost 40 years.
The changes in social policy, as actually enacted, leave some very serious areas of
concern.
10 Recessions and High Unemployment Areas
A reform designed to support people who are working fails if people cannot find work,
even temporarily. So long as jobs are available, the strategy has appeal both politically
and economically. But what happens when there are not enough jobs?
The US is in the midst of an unprecedented recovery. Growth has remained solid and
inflation low since the early 1990s. But certain areas continue to face extremely high
level of unemployment. And even where the economy is now strong, when the next
recession comes, newly working welfare mothers seem likely to be among the first to lose
their jobs. All the child care, work-related health care, or job search assistance a state can
muster will not do much good if there are no jobs to be had.
One of the most disturbing features of the new reform legislation at both the federal and
state level is that the unemployment problem is simply ignored. Recall that the US has no
universal children's allowances or health benefits. Furthermore, because of the rules
regarding previous work experience and the reasons for job loss, the current UI system
actually covers only about one-third of the unemployed today. (rising to roughly 50 per
cent during recessions) Thus parents and their children are at considerable risk if they
lose benefits.
The new social policies have changed aid from a counter-cyclical form of support,
whereby people get the most benefits when they are without a job, to a pro-cyclical one.
Get a job, get added benefits. When times turn bad, some former recipients will be faced
with a loss of both their job and their benefits.
States will be hard-pressed to help people who lose jobs during a recession. The block
grants that states now receive are fixed regardless of the state of the economy and
regardless of their caseload. In the current economic climate, when unemployment is low
and caseloads falling, states have enjoyed the fixed block grant. But when the economy
turns sour, that fixed aid may add to their burdens.
11
The Hard To Employ
Even in the best of times, some recipients will be hard pressed to work. In 1995, nearly
40 per cent were dropouts, and only 16 per cent had any schooling beyond high school.
Some 45 per cent had children under the age of six. Almost 60 per cent were never
married mothers. Some 37 per cent were black and another 20 per cent Hispanic
(Committee on Ways and Means, 1998, Table 7-19: 441). At least 40 per cent had not
worked in the past 2 years (Bane and Ellwood, 1994). Danziger et al. (2000) report that a
quarter have major depressive disorders and 22 per cent have children with significant
health problems.
29
When time limits hit, these hard to employ mothers may face serious problems. In a timelimited, work-oriented system, when people have reached their limit without finding a
job, there are only three choices: terminate benefits anyway (that is what happens in UI),
allow people to continue to receive cash aid without working (relax the limit), or provide
some form of subsidised or government job.
Of course, the struggles of the hard-to-employ and the problem of recessions were far less
serious in the original Clinton vision of reform, with its guarantee of some form of
subsidised work for those hitting a time limit. But that plan never made it through the
Congress. In principle, individual states could create subsidised job programs. But only a
few states have so far opted to create jobs for recipients who say they cannot find work.
Implicitly more than explicitly, many states are choosing simply to cut people off when
they reach a time limit. Arguments to justify this position range from a claim that one or
two or five years is enough time to find a new job, to a simple claim that there are plenty
of jobs available so long as child care and health care are provided. But no one really
knows how many will be affected or what will happen to them.
This avoidance is possible in part because the hard-to-employ cases have not yet reached
any time limits. In the current US economy, low-skill jobs seem generally available in
most areas, though they pay very poorly. Thus it is much easier to argue that the problem
is low pay and benefits rather than jobs. And so long as times remain good, these hard-toemploy, those whose problems are not quite serious enough to qualify for disability
benefits, may find themselves even more isolated and stigmatised than before. States may
be reluctant to provide jobs for people who cannot hold down a private sector job. These
are the families where the levels of stress and abuse could rise dramatically.
Frankly, I think it is hard to see how a time-limited work-oriented reform strategy can
work without some form of long-term aid or last-resort subsidised jobs in cases where
people cannot find work. Otherwise, poor-single-parent families will face enormous
economic risks. So far, public jobs have proven unpopular in Congress in recent decades
in the US, even though government work was an important part of the support strategy of
the 1930s under FDR. Whether that will change with economic conditions remains to be
seen.
12
The Continuing Insecurity of Single-parent Families
The move from welfare benefits to work supports in combination with a strong economy
has reduced the overall poverty of single-parent families somewhat. Yet even with all the
new supports in place, even with the extraordinarily strong economy of recent years, and
even with the sharp increase in work by single parents, roughly one-third of single
parents still fall below the very low US poverty line.
The basic problem is that worker support policies still fail to address the biggest
underlying source of poverty and insecurity of single-parent families: the family must
generally rely on the earnings of one person - typically a low-skilled woman - for
support. These women typically have relatively low wages, and Burtless (1997) has
shown that wage growth over time is modest at best for former welfare mothers. And
30
DAVID T. ELLWOOD
Pavetti (1997) has shown that former welfare recipients who do move from welfare to
work have frequent spells of unemployment.
Full-time work may not be practical in many situations. While the majority of single
mothers are working at least somewhat, CPS data reveal that less than half of all single
mothers worked full time all year in 1999, in spite of the strong incentives to do so. An
even smaller share of married mothers worked fully. Recent studies tracking people who
leave welfare typically find that 40 per cent are not working. And work by Primus et al.
(1999) suggests that as TANF has been withdrawn and worker supports added, the
poorest decile of single parents have lower incomes than previously.
Perhaps even more importantly, the question inevitably arises about what full-time work
may mean for children if mothers would prefer to be caring for their children. The basic
problem is that one parent, often one with poor education and limited work experience, is
expected to both provide for and nurture her children. Without some form of additional
support, poverty rates among lone parents seem likely to continue at very high levels.
One place to look for additional support for single parents is non-custodial parents,
typically the fathers. Sorenson (1995), Bartfeld (1998), and others estimate that many
fathers pay little and most could pay significantly more. And indeed another major
element in recent welfare reform efforts has been a very large increase in child support
enforcement efforts. Though these are just beginning to take hold, there is evidence of
rising collections for both formerly married mothers and never-married mothers
(Sorenson and Halpern, 1999). These elements of support should grow fairly rapidly in
the years ahead. Still, on average, child support collections remain modest and sporadic
for single parents. Moreover, as a larger and larger share of single-parent families are
formed as a result of a birth to an unmarried woman, the difficulties of collecting grow
and the capacity of men to pay may shrink.
I continue to hope that the nation will experiment with an insured minimum child support
payment (also known as child support assurance) that mothers can count on even when
the state fails to collect the full child support, as suggested by Garfinkel (1992); however,
such plans remain controversial.
Of course, the best solution might be to find a way to reduce the incidence of singleparent families in the first place. The last few years have finally seen a leveling of family
change. Some credit welfare; many are skeptical. Social scientists have had little success
in explaining the rises of the past 30 years, and even less success in finding policies that
might conceivably reverse the trends. Ultimately, if the US does not find a way to reduce
the incidence of single-parent families, we may be forced to decide whether to accept
widespread poverty among children or to provide some modest source of additional
support for children in single-parent homes.
America is doing less to support the welfare poor and more for the working poor. The results
of this policy in the 1990s, in the context of a very strong economy, have been sharp
reductions in welfare use, much more work, and somewhat less poverty - though the most
disadvantaged single parents may be worse off. But the underlying challenges of assisting
the families of those who cannot find work because of weakness in the economy or their
31
own limits remain unresolved. The changing patterns of families and wage inequality will
pose still greater challenges in the opening decades of the next century.
13
Lessons That Might Hold Even Across A Very Large Ocean
The US is engaged in a bold but potentially risky strategy of orienting public support for
healthy adults around work. Much of what has transpired is inevitably peculiarly
American, much to the relief, I suspect, of many international observers. But there may
be some lessons for those who might want to increase work as an element in social policy
in places like Australia. Here is my list.
Shifting social policy in the direction of work can change the political climate for the
better, especially in a context where benefits are highly targeted and an ‘us’ versus ‘them’
mentality has developed, or where prevailing norms and expectations regarding work,
responsibility, and poverty seem inconsistent with current policies.
Incentives matter, and big incentives matter a lot. The US has instituted a rather
draconian set of policies which give long-term aid to virtually no adult who is able to
work in the market place but without a job. It provides increasingly generous aid for
those who do work. With incentives like that, people work more, sometimes a lot more.
The US started with quite low benefits to begin with. Dramatically reshaping a system
such as Australia’s or the UK’s to match these incentives would imply massive benefit
reductions and almost certainly an attendant rise in poverty. But increasing work
incentives through reducing effective tax rates could raise work at least somewhat.
Encouragement to work does not need to come entirely in the form of added financial
support for workers. Altering expectations of aid recipients and caseworkers, offering
training, even imposing time limits can push people into work even if it does not pay a
great deal more.
Any work-oriented support system will work best in a very strong economy. In a much
weaker one, the goal of encouraging greater work will be far more difficult to achieve,
both practically and politically.
Work can work to improve well-being in certain situations, but there will remain serious
risks, especially to less advantaged adults and their children, from programs that
emphasise work almost exclusively, as in the US.
So work can sometimes work. But it's not for everyone.
References
Amenta, Edwin (1988), Bold Relief. Institutional Politics and the Origins of Modern
American Social Policy, Princeton University Press, Princeton, N.J.
Bane, Mary Jo and David T. Ellwood (1994), Welfare Realities: From Rhetoric to
Reform, Harvard University Press, Cambridge.
32
DAVID T. ELLWOOD
Bartfeld Judi (1998), Child Support and the Post Economic Well-Being of Mothers
Fathers and Children, IRP Discussion Paper No. 1182-98, Institute for Research on
Poverty, Madison, WI.
Blundell, Richard, A. Duncan, J. McCrae and C. Meghir (2000), ‘The labour market
impact of the working families' tax credit’, Fiscal Studies, 21(1), 75-104.
Burtless, Gary T. (1997), ‘Welfare recipients' job skills and employment prospects’,
The Future of Children, 7(1), Spring, 39-51.
Committee on Ways and Means (1996), Green Book. Background Material and Data on
Programs Under the Jurisdiction of the Committee on Ways and Means, US
Government Printing Office, Washington, DC.
Danziger, Sandra, Mary Corcoran, Sheldon Danziger, Colleen Heflin, Ariel Kalil, Judith
Levine, Daniel Rosen, Kristin Seefeldt, Kristine Siefert and Richard Tolman (2000),
‘Barriers to employment of welfare recipients’, in R. Cherry and W. M. Rodgers III,
eds, Prosperity for All? Economic Boom and African Americans, Russell Sage, New
York.
DeParle, Jason (1997), ‘With the economy humming, welfare has its image polished’,
The New York Times, 11 September, A1.
Dickens, Richard and David T. Ellwood (2000), Whither Poverty in Great Britain and the
United States? The Determinants of Changing Poverty and Whether Work Will
Work, NBER Working Paper No. 8253, National Bureau of Economic Research,
Cambridge MA.
Edelman Peter (1997), ‘The worst thing Bill Clinton has done’, The Atlantic Monthly,
279, 3, March.
Farkas, Steven and Jean Johnson (1996), The Values We Live By: What Americans Want
From Welfare Reform, Public Agenda, New York.
Garfinkel Irwin (1992), Assuring Child Support: An Extension of Social Security, Russell
Sage Foundation, New York.
Garin Geoffrey, Guy Molyneux and Linda Divall (1993), Public Attitudes Toward
Welfare Reform: A Summary of Key Research Findings, American Viewpoint, Peter
Hart Research Associates, Washington, DC.
Hotz, V. Joseph and John Karl Scholz (2001), The Earned Income Tax Credit, NBER
Working Paper No. W8078, National Bureau of Economic Research, Cambridge,
MA.
Johnson, Nicholas and Ed Lazere (1998), Rising Number of States Offer Earned Income
Tax Credits In the Last Two Years, Five States Have Enacted New State EITCs or
Expanded Existing Credits, Center For Budget and Policy Priorities, Washington,
DC.
Morris, Richard S. (1997), Richard S. Morris, Winning the Oval Office: Winning the
Presidency in the Nineties, Random House, New York.
33
Pavetti, LaDonna (1997), How Much More Can They Work? Setting Realistic
Expectations for Welfare Mothers, mimeo, The Urban Institute, Washington, DC.
Primus, Wendell, Lynette Rawlings, Kathy Larin and Kathryn Porter (1999), The Initial
Impacts of Welfare Reform on the Incomes of Single-Mother Families, Center on
Budget and Policy Priorities, Washington, DC.
Schoeni, Robert F. and Rebecca M. Blank (2000), What Has Welfare Reform
Accomplished? Impacts on Welfare Participation, Employment, Income, Poverty,
and Family Structure, NBER Working Paper No. W7627, National Bureau of
Economic Research, Cambridge, MA.
Sorenson, Elaine (1995), Noncustodial Fathers: Can They Afford to Pay More Child
Support?, The Urban Institute, Washington, DC.
Sorenson, Elaine and Ariel Halpern (1999), Child Support Enforcement: How Well Is it
Doing? Assessing the New Federalism, Discussion Paper No. 99-11, The Urban
Institute, Washington, DC.
Whiteford, Peter (2000), The Australian System of Social Protection - An Overview,
Policy Research Paper No. 1, Department of Family and Community Services,
Canberra, Australia.
Zedlewski, Sheila, Sandra Clark, Eric Meier and Keith Watson (1996), Potential Effects
of Congressional Welfare Reform Legislation on Family Incomes, The Urban
Institute, Washington, DC.
34
Indigenous Hunter-Gatherers in the 21st
Century: Beyond the Limits of Universalism
in Australian Social Policy?
Jon Altman1
Centre for Aboriginal Economic Policy Research
The Australian National University, Canberra
1
Introduction
In the past two decades, Australia’s social policy approach to Indigenous people has been
based on the overarching goal of striving for equality of outcomes, while also
increasingly recognising difference. Such an approach is appropriate and allows for an
amalgam of Indigenous citizenship rights, fully recognised since the late 1960s, and an
expanding envelope of inherent rights, with the former being wedded to the equality
agenda and the latter to difference. A fundamental principle of Australian social policy is
universalism: ‘the idea that rules, written in general terms, can be equally and fairly
applied to all people whatever their social identity or background’ (Sanders, 2001: 1). A
problem with universalism as highlighted by Sanders is that it has difficulty recognising
difference; logically the more different, the greater the difficulty.
This paper sets out to highlight a problematic extreme for social policy making that is
committed to the universalism norm: Indigenous people whose cultural and economic
circumstances are very different from the dominant society’s, those people residing on
Aboriginal-owned land in the remotest parts of Australia and engaged in a fundamentally
different customary economy. This extreme is problematic, especially in the current
context of welfare reform with its emphasis on mutual obligation, because the ‘striving
for equality of measured outcomes’ goal might be unattainable and inappropriate in such
circumstances. Historical legacy aside, this unattainability is primarily due to two factors:
residence in extremely remote localities ‘beyond the market’ and the shortage in such
geographically-remote locations of ‘viable’ labour markets; and a concomitant high level
of engagement in the customary (non-market) economy facilitated by strong cultural
continuities, due in part to relatively late and benign contact histories.
Unlike much social and Indigenous policy debate, this paper explores some of the
difficult issues for outstation communities under the session rubric ‘New approaches for
1
The author would like to thank Melinda Hinkson, Robert Levitus, Ian Munro and John
Taylor for comments on an earlier draft of this paper.
Altman, J. (2002), ‘Indigenous hunter-gatherers in the 21st century: beyond the limits of universalism in
Australian social policy’, in T. Eardley and B. Bradbury, eds, Competing Visions: Refereed Proceedings of
the National Social Policy Conference 2001, SPRC Report 1/02, Social Policy Research Centre, University
of New South Wales, Sydney, 35-44
JON ALTMAN
Indigenous social policy’. This exploration is neither dilettantist nor romanticist and is
based on the following three observations:
•
a combination of land rights and native title and the possibility of enhanced
recognition of Indigenous rights could see more, rather than fewer, Indigenous
Australians living in such remote circumstances in the future;
•
my own research in Arnhem Land over two decades indicates that this mode of living
remains robust: people living in such circumstances are very adept at resisting the
penetration of the state, the market and globalisation; and
•
a great deal of recent social policy debate under the rubric of mutual obligation seeks
mechanisms to enhance a degree of social participation support that is already well
established at these remote and small Indigenous communities.
The paper ends with a discussion of some innovative new directions that social policy
might consider in relation to outstation communities. My motivation in providing these is
a firm belief that a greater cross-cultural awareness is required in social policy
development for the 21st Century.
2
The People and the Places
The outstations movement was first so labelled nearly 30 years ago in the early 1970s
(see for instance Coombs, Dexter and Hiatt, (1980) for a historical account). A
combination of a policy shift to self determination, the implementation of land rights law
and the failure of assimilation policies in Aboriginal townships resulted in an unusual
rural exodus. People went back to live on their remote traditional lands and resuscitated a
customary economy based on exploitation of renewable resources. By the mid 1980s
there was a recognition that this choice required some policy attention and action
(Commonwealth of Australia, 1987) A new hybrid Aboriginal economy was emerging,
one that was primarily state-based, with a sliver of the market but with a considerable
additional customary component. This customary component was, and remains,
unrecognised in policy discussions about Indigenous employment (Altman, 1991) or
about income, diet, health and well-being as well as other broader contributions: the
dominant market ideology appears to succeed in excluding such unconventional
alternatives.
Information about outstation populations is not readily available and is difficult to collect.
Standard and even modified Australian Bureau of Statistics practices are poorly
structured to collect relevant information about dispersed, mobile, remote and culturally
different outstation populations. In 1998, ATSIC estimated that there were about 12 000
people living at 1 000 outstations serviced by 100 Outstation Resource Agencies. These
outstations were primarily on Aboriginal land in the Northern Territory, but also in the
north of South Australia, Queensland and Western Australia. In a study that I participated
in recently, it was estimated that Australia-wide there could be between a minimum of 13
000 and a maximum of 32 000 people, representing between four per cent and nine per
cent of the Australian Indigenous population at that time, residing at as many as 1400
36
INDIGENOUS HUNTER-GATHERERS IN THE 21ST CENTURY
outstations (Altman, Gillespie and Palmer, 1999: 83–4).2 More recently, the 1999
Community Housing and Infrastructure Needs Survey estimated that there were 856
outstations with a total population of 12 739, with an outstation defined as a discrete
Indigenous community that had a population of less than 50 and that is linked to a
resource agency for the provision and maintenance of services such as housing, water,
power and sewerage (Roger Jones, personal communication).
The orthodox view is that when outstations evolved in the early 1970s, it was in large
measure out of the failure of Aboriginal townships to offer economic development
opportunities to Indigenous people, despite concerted effort. With land rights and access
to some welfare income, small family groups exercised a choice to return to their
traditional lands. A factor that influenced this movement was a desire to escape social
problems at polyglot townships located on other Indigenous people’s traditional lands. In
making this choice, people voted with their feet, there was no concerted effort by either
the state, academics or other social scientists to get people back out onto their lands
(contra Sandall, 2001). Indeed it was common bureaucratic practice to require groups to
demonstrate a degree of commitment prior to providing any financial support: six months
unassisted residence was the norm before a $10 000 establishment grant was provided to
allow provision of rudimentary infrastructure (Altman, 1987).
Outstations vary enormously from region to region owing to differences in ecological
zones and pre-contact social formations; but they share the following common features:
•
their residents had a relatively late and benign contact history and associated late
access to the provisions of the Australian welfare state;
•
structurally, outstations are located in remote areas where the market is small or nonexistent and where the state (either as a service provider or welfare agency) looms
relatively large but is physically distant;
•
the demographics of outstations are unusual: for example their populations are highly
mobile and there is limited mixed (inter-ethnic) family formation; and
•
social capital in these communities is well adapted to customary productive activity
where rights in land and resources are well defined, but this social capital is poorly
adapted to the market.
As a general rule, the outstations population is poorly represented in social policy
debates. Furthermore, there is very little understanding of the complex hybrid economies
of outstations, with a customary sector whose economic contribution is generally
unrecognised and unquantified; productive market activity such as art production that is
state-mediated and usually dependent on inter-cultural brokerage; and residents whose
major engagement with the wider economy is only as consumers.
2
This wide range is partly linked to seasonal fluctuation and partly to the absence of
established and standard mechanism to measure these populations. Altman, Gillespie and
Palmer (1999: 83) used four population estimates: minimum, maximum, usual and
effective with the last attempting to measure person nights spent at outstations.
37
JON ALTMAN
3
The Problems
There are some contradictions emerging in both social and Indigenous policy that need to
be addressed. Some of these problems have wider applicability than only for outstations,
but are starkest at these communities that are the most different.
First, at the broadest level, a number of laws and institutional mechanisms have been
established in the modern Indigenous policy era (post-1967) to assist Indigenous people
gain access to their traditional lands. It is now estimated that up to 18per cent of the
Australian continent is Indigenous owned, with most of this land in remote regions
(Pollack, 2001). At the same time, when Indigenous people occupy their land, much of
which is of marginal commercial value, they are criticised for living away from viable’
labour markets (see Reeves, 1998).
Second, there is a very high level of formal Indigenous unemployment in remote areas
and this is reflected in official statistics. The extent of formal unemployment is
ameliorated to a considerable extent by a high rate of Indigenous participation in the
Community Development Employment Projects (CDEP) scheme, a form of work-for-thedole where participants are classified as employed. At the same time there is an official
reluctance to either measure or acknowledge the levels of employment in, and
contributions of, the customary economy.
Third, there is a tendency to conflate the welfare dependency of outstations and
Aboriginal townships, as if they are no different. In reality there is frequently a great deal
more productive, even market-oriented, activity at outstations than at townships. And
there is also a tendency to ignore the extent of state dependency—in one guise or
another—of other sections of the non-Indigenous population in similar remote localities,
but that is another issue.
The fundamental policy problem at outstations is that in the relative absence of the
market and given the prevalence of the modern consumptive economy, there is a high
level of cash dependence on the state. This is generally met in one of two ways: via the
payment of CDEP scheme wages, on the assumption that people are working for their
dole equivalents; or via access to a range of welfare benefits like pensions and Newstart
Allowances (something of a misnomer in the outstations case). Both forms of payment
stretch the logic of universalism to its limits: outstation residents need a minimum cash
income, but not as a safety net until they find regular employment. Social policy is not
comfortable with the idea of paying a guaranteed minimum income to outstation residents
in case universalism dictates such minimum income should be paid elsewhere. Access to
welfare represents the citizenship rights of outstation residents; their opportunity to live
on their traditional lands is, arguably, a special Indigenous right—indeed returning land
to people via complex and expensive claims processes has no policy logic if it is not
accompanied by a right of residence.
4
Some Current Policy Debates
In the last five years, since the election of the Howard Government in 1996, there have
been heightened and generally implicit critiques of outstation communities. Much of this
38
has been ideological, based on a commitment to incorporate Indigenous people into the
mainstream economy. At the level of rhetoric there has perhaps been a greater emphasis
on equality of outcomes than on cultural difference, but this developmental perspective
has lacked realism, focusing far more on the role of the state, and the need for its
reduction, than on the structural absence of the market in remote regions. At times this
debate has been couched in terms of practical reconciliation versus symbolic
reconciliation, although again such rhetoric has limited meaning in explaining just how
development options might be delivered to remote localities.
A concerted critique of outstations was mounted in the review of the Aboriginal Land
Rights (Northern Territory) Act 1976 undertaken by John Reeves in 1998. Reeves (1998)
acknowledged that the Land Rights Act had been successful in returning vast tracts of
unalienated crown land to recognised Aboriginal traditional ownership—today, nearly 50
per cent of the Northern Territory is Aboriginal-owned. But he went on to argue that the
provision of this land to Indigenous groups is facilitating the maintenance, rather than
amelioration, of Indigenous economic marginality, at least as measured by standard social
indicators. Reeves’s ‘assimilationist’ view argues that the longer-term incorporation into
the mainstream economy is the preferred option for both future Indigenous generations
and Australia more broadly (Reeves, 1998). This position was not supported by rigorous
historical, cultural or economic analysis of the lived reality of Indigenous Australians, as
was highlighted by a range of researchers (Altman, Morphy and Rowse, 1999) and a
committee of federal parliamentarians (Commonwealth of Australia, 1999).
The naïve view that the standard route to economic integration is available to all has been
taken up by other commentators, who argue that the lifestyle typified by outstation people
and their engagement with the customary economy is either ‘separatist’ (Johns, 2001) or
a form of ‘romanticisation of the primitive’ (Sandall, 2001). Interestingly, none of these
neo-assimilationists offer any constructive strategy for Indigenous economic
development beyond the orthodox doses of state-provided health, education, training,
employment and housing, a risky strategy that was attempted without much success or
Indigenous acceptance in the 1960s and 1970s. Paradoxically, such a strategy would
require a great deal more state intervention and subsidisation than the market-oriented
ideology of its proponents might allow. Interestingly too, none of these views attempts to
engage with Indigenous perspectives or priorities beyond, when convenient, the position
put by Noel Pearson (2000). A markedly different recent polemic of cultural relativism,
which is realistic and informed, is provided by Folds (2001) in his analysis of the
divergence between Pintupi and wider Australian expectations of Indigenous policy in a
remote community. While Folds is able to capture well the divergent aspirations and
expectations of Aborigines and policy makers, he has difficulty articulating a more
productive policy prescription beyond the status quo.
Specifically in the social policy arena, Patrick McClure’s Reference Group on Welfare
Reform (2000) did not deal with the outstations case, or any similar community type.
Statistics on outstations might suggest intergenerational poverty and unemployment and
associated social exclusion, but such a view would not be accurate given the social
capital-intensive nature of such communities with kin-based relations of production (see
Altman (1987) for a case study). Similarly, Pearson’s depiction of ‘welfare as poison’
39
JON ALTMAN
(Pearson, 2000) while arguably of relevance to townships where current policy thinking
regards inactivity as excessive and a problem, is of limited relevance to outstation
communities. Indeed, one plank in Pearson’s development proposal for Cape York
Peninsula (rejuvenation of the customary economy) has been practised at outstations
elsewhere and especially in Arnhem Land for some 30 years (Meehan, 1982; Altman,
1987). It is likely however that the productivity of the customary economy is lower in
arid regions, although little recent evidence is available (see Cane and Stanley, 1985) and
lower productivity may be compensated by lower population density.
Both McClure and Pearson advocate a form of mutual obligation whereby the state
provides benefits and individual recipients reciprocate with community work, the
rationale being in part that state expenditure should be, if possible, partially offset and
that activity of benefit to the community should be generated by income support.
However, the McClure/Pearson prescription does not take into account situations where
people are actively engaged in productive activity and where community-oriented activity
would merely divert work effort from such activity. Nor is the issue of supervision of
reciprocal work and appropriate sanctions well considered: in small family-based groups
it is exceedingly difficult, if not undesirable, to allocate supervisory responsibility to any
one individual; such a role would reduce community social capital. On the other hand, to
get an outsider to undertake the supervisory task is not just expensive, but also counter to
the physical distancing from the state and wider society that many outstation
communities pursue (see Levitus, 2001).
5
Australians Working Together: The Government’s Response
The broad principles enunciated by McClure and also by Pearson have been largely
accepted by government and even championed by some members of parliament. In the
May 2001 Budget there were measures, most prominent in the package Australians
Working Together: Helping People to Move Forward that call for ‘A fair deal for
Indigenous Australians’ (Commonwealth of Australia, 2001). These measures however
represent only partial and cautious moves to implement the McClure recommendations.
As the title suggests they are also developmental, wanting people to ‘move forward’, to
help Indigenous people into work and to contribute to their communities. Interestingly,
the former goal seeks to move up to 10000 CDEP participants from what is now termed
‘CDEP work experience’ into paid jobs where job opportunities are available. This is an
ambitious target without a stated timeframe. The latter goal seeks to deliver Community
Participation Agreements and capacity building to up to 100 remote communities,
presumably townships.
From the outstations perspective (lumping for the moment outstations with remote
communities) much policy initiative is summarised in Fact Sheet 23 ‘Community
Participation Agreements and Capacity Building’ which combines elements of reality
with some ominous implications (Commonwealth of Australia, 2001). Let me quote and
then comment on and deconstruct two key statements:
There are few opportunities in some remote Indigenous
communities for people on income support to meet activity test
40
requirements. In others, people have been exempt from activity
testing. We want all Australians to be active and involved in
their communities, to be using their skills and potential and to
be looking for work when they can.
and
Remote Indigenous communities will be given more support to
develop their own practical solutions to the challenges they
face…Indigenous Australians will be able to contribute in a
practical way to their communities in return for their income
support payments.
First, if there are few existing opportunities, how will the new policy assist to create
additional opportunities for those looking for work? To do so would require massive
subvention beyond that allocated to proposed community participation and capacity
building options.
Second, it is unclear from the perspective of outstations how this initiative is different
from what is currently occurring - Newstart is exempt from activity testing and CDEP
payments are provided as income support to outstation residents. The major problem at
present is the inconsistency in income testing between the two regimes, with Centrelink
Newstart payments income tested and subject to a taper and ATSIC’s CDEP payments
having a far more generous income test and no taper (Altman and Johnson, 2000).
Third, if individual work effort in the customary economy for family or household
consumption (and benefit) is occurring, why is there any need for a re-orientation to
community contribution? There is also a danger that such re-orientation will facilitate the
substitution of the state provision of citizenship rights in remote communities with the
contributions of individual welfare recipients.
Some Australians Working Together strategies might result in decreased state subvention
in those situations where people actually move to paid employment, in situations where
there are viable labour markets; but the proposed community participation and capacity
building options for remote communities appear to be predicated on continuing state
support.
6
Sustainable Options for a Real Way Forward: A New Approach?
A major problem for Indigenous people and policy formulation is that much Indigenous
productive activity in remote Australia remains unrecognised and unvalued. While both
contemporary and historical case studies document the value of such contributions to
domestic economies, there is an urgent need to expand some of this research to more
widespread regional and national perspectives. Some such research is being undertaken
by scientists at the Northern Territory University’s Key Centre for Tropical Wildlife
Management and Tropical Savannas. Preliminary findings indicate that Indigenous
participation in the customary economy is generating local, regional and national benefits
41
JON ALTMAN
in land management and maintenance of species biodiversity.3 The potentialities to
convert sustainable harvesting from the customary to the commercial sector are being
investigated and trialed in joint ventures (Bawinanga Aboriginal Corporation, 2000;
Vardon, 2001).4 Similarly, the sustainability of the Indigenous production of material
culture for the market needs to be rigorously assessed and the extent of market demand
needs evaluation.
The value of Indigenous land management needs to be properly quantified. It would be
interesting, for example, to benchmark expenditure on land management in national parks
with the unrecognised value of land management undertaken on adjacent Aboriginal land.
Ecologically, in contiguous zones, land management outcomes in both are interdependent. Similarly, the potential national value of Indigenous land management,
especially in relation to the emerging global issues of greenhouse gas emission and
carbon sinks and credits, needs assessment. And also nationally, there are the more
mundane potential benefits of Indigenous infrastructure (roads, airstrips, etc) in remote
regions, most constructed with an again unquantified Indigenous labour input.5
How does a call for such speculative and properly quantified economic futures correlate
with social policy? This paper takes at face value stated government aims of improving
the socio-economic status of Indigenous Australians, including those living in the
remotest circumstances. While it is acknowledged that there is limited information about
the Indigenous customary economy and its wider spin-off benefits, indications are that it
is robust and ecologically sustainable at least in some parts of the country. Yet little
thought has been given to the relative, strategic and unrecognised contributions of
outstation people, nor how to ensure these are measured. Consequently, it is not
surprising that income support continues to be structured as a temporary measure to be
tolerated until a ‘proper job’ is secured. Despite decades of historical evidence to the
contrary, policy makers continue to maintain the rhetoric of such fiction, most recently in
the 2001 Budget. What is needed is some guaranteed minimum income scheme that will
provide signals for outstation residents to maximise their participation in the customary
economy and their prospects for reduced dependency. The formal recognition of such
work and income would alleviate the negative ‘welfare dependent’ public perception
accorded to those whose productive work is in the customary economy. But such
recognition is double-edged and potentially contentious, because it will ameliorate the
extent of recorded Indigenous disadvantage in remote regions and much Indigenous
political argument for additional resourcing is based on emphasising disadvantage rather
than successful adaptation. Such recognition though is an important element in
structuring social policy and income support to reflect lived reality.
3
4
5
Land management activities include burning of country, weed eradication, erosion control
and preservation of scarce breeding habitats for wildlife species.
Examples here range from harvesting of species like crocodile eggs and turtles to joint
venture safari hunting and recreational catch and release fishing ventures.
Again though it must be emphasised that too much of the emerging data base is focused on
the tropical savannas and wetlands, too little on arid regions.
42
7
Conclusion
Contemporary welfare reform embodied in the Australians Working Together package is
based on the universalist notion that welfare support for the able-bodied is only a
temporary measure prior to paid employment being secured. Such universalism needs to
be challenged if creative solutions to the economic development problems faced by
Indigenous people residing at outstations in remote locations are to be addressed.
This paper has argued that in such situations improvement can occur via the customary
and market sectors of the economy, but that this will require that state support is held
constant into the foreseeable future. An initial step in policy discourse might be to
describe and document the unique contemporary form of Indigenous hybrid economy
more accurately; to quantify the ‘unrecognised’ economic contributions of the customary
sector; and to seek to link this sector creatively with new income-generating market
activity. It is perhaps to these ends that discretionary additional resources for capacity
building can be targeted, to enable the development of regional solutions to some very
complex social and economic issues.
References
Altman, J.C. (1987), Hunter-Gatherers Today: An Aboriginal Economy in North
Australia, Australian Institute of Aboriginal Studies, Canberra.
Altman, J.C. (1991), ‘Conclusion’, in J.C. Altman, ed., Aboriginal Employment Equity by
the Year 2000, Research Monograph No.2, Centre for Aboriginal Economic Policy
Research, Australian National University, Canberra.
Altman, J.C., D. Gillespie and K. Palmer (1999), National Review of Resource Agencies
Servicing Indigenous Communities, 1998, Aboriginal and Torres Strait Islander
Commission, Canberra.
Altman, J.C. and V. Johnson (2000), ‘CDEP in town and country Arnhem Land:
Bawinanga Aboriginal Corporation’, CAEPR Discussion Paper No. 209, Centre for
Aboriginal Economic Policy Research, Australian National University, Canberra.
Altman, J.C., F. Morphy and T. Rowse (1999), eds, Land Rights at Risk, Evaluations of
the Reeves Report, Research Monograph No. 14, Centre for Aboriginal Economic
Policy Research, Australian National University, Canberra.
Bawinanga Aboriginal Corporation (2000), Annual Report 1999–2000, Bawinanga
Aboriginal Corporation, Maningrida.
Cane, S. and O. Stanley (1985), Land Use and Resources in Desert Homelands, North
Australia Research Unit, Darwin.
Commonwealth of Australia (1987), Return to Country: The Aboriginal Homelands
Movement in Australia, House of Representatives Standing Committee on
Aboriginal Affairs, Australian Government Publishing Service, Canberra.
Commonwealth of Australia (1999), Unlocking the Future: The Report of the Inquiry into
the Reeves Review of the Aboriginal Land Rights Act 1976, House of
Representatives Standing Committee on Aboriginal and Torres Strait Islander
Affairs, Canberra.
43
JON ALTMAN
Commonwealth of Australia (2001), Australians Working Together: Helping People to
Move Forward, Department of Family and Community Services, Canberra.
Coombs, H. C., B. G. Dexter and L. R. Hiatt (1980), ‘The outstations movement in
Aboriginal Australia’, Australian Institute of Aboriginal Studies Newsletter, 14, 16–
23.
Folds, R. (2001), Crossed Purposes: The Pintupi and Australia’s Indigenous Policy,
UNSW Press, Sydney.
Johns, G. (2001), ‘The failure of Aboriginal separatism’, Quadrant, May 2001, 9–18.
Levitus, R. (2001), Aboriginal Development and Self Determination: Policy, Distancing
and Articulation, unpublished manuscript, Centre for Aboriginal Economic Policy
Research, Australian National University, Canberra.
Meehan, B. (1982), Shell Bed to Shell Midden, Australian Institute of Aboriginal Studies,
Canberra.
Pearson, N. (2000), Our Right to Take Responsibility, Noel Pearson and Associates,
Cairns.
Pollack, D. (2001), ‘Indigenous land in Australia: a quantification of holdings in 2000’,
CAEPR Discussion Paper 222 (forthcoming), Centre for Aboriginal Economic
Reeves, J. (1998), Building on Land Rights for the Next Generation: Review of the
Aboriginal Land Rights (Northern Territory) Act 1976, Aboriginal and Torres Strait
Islander Commission, Canberra.
Reference Group on Welfare Reform (2000), Participation Support for a More Equitable
Society: Final Report of the Reference Group on Welfare Reform, Department of
Family and Community Services, Canberra.
Sandall, R. (2001), The Culture Cult: Designer Tribalism and Other Essays, Westview
Press, Boulder, Colorado.
Sanders, W. (2001), ‘Indigenous Australians and the rules of the social security system:
universalism, appropriateness and justice’, CAEPR Discussion Paper 212, Centre
for Aboriginal Economic Policy Research, Australian National University,
Canberra.
Vardon, M. (2001), ‘Aboriginal wildlife use in the Top End of the Northern Territory’,
paper presented at the Veterinary Conservation Biology: Wildlife Health and
Management in Australasia conference, Taronga Park Zoo, Sydney.
44
Health, Beneficiaries and Welfare to Work:
Competing Visions of Employability
Maureen Baker and David Tippin
Department of Sociology, University of Auckland, New Zealand
1
Introduction
Governments in many OECD countries have restructured their welfare programs to
encourage unemployed people to rely less on social benefits and more on their own
employment earnings. In this paper, we discuss the relationship between lone parenthood
and health, and poor health as a potential variable affecting the integration of lone
mothers into the workforce. The paper draws on our ongoing study at the University of
Auckland1 that focuses on the extent to which poor health of women, their children, and
their dependants constrains the transition from welfare to work for sole mothers on the
Domestic Purposes Benefit (the NZ counterpart to the Australian Sole Parent Benefit).
2
The DPB and Work Testing
In 1973, both Australia and New Zealand created similar social benefits for sole mothers,
called the Sole Parent Pension in Australia and the Domestic Purposes Benefit (DPB) in
New Zealand, to counteract rising rates of poverty among sole parent families. Although
this was not the first benefit for sole mothers2, the 1973 legislation was important in New
Zealand because it was gender neutral and offered income support to all parents with low
assets and low income caring for dependent children or other sick or infirm relatives at
home. It also embodied the government’s view, supported by organisations such as the
Council of Women of New Zealand, that the children of sole mothers were best cared for
at home by their mothers (Uttley, 2000: 444). The DPB was part of a package of social
programs to help families raise their children.
Australian and New Zealand policy discourse has emphasised the importance of
responsible parenting and supporting one’s children emotionally, physically and
1
This project is funded by a NZ Health Research Council Grant (from 2000-2002) to the Institute for
Research on Gender. Principal investigators are Maureen Baker (Department of Sociology), Heather
Worth (Institute for Research on Gender), and David Tippin and Tracey McIntosh (Department of
Sociology).
2
A widows pension was created in 1911 in NZ and 1942 in Australia. A benefit for deserted
wives was created in NZ in 1936 (Baker and Tippin, 1999: 32).
Baker, M. and D. Tippin (2002), ‘Health, beneficiaries and welfare to work: competing visions of
employability’, in T. Eardley and B. Bradbury, eds, Competing Visions: Refereed Proceedings of the
National Social Policy Conference 2001, SPRC Report 1/02, Social Policy Research Centre, University of
New South Wales, Sydney, 45-63.
MAUREEN BAKER AND DAVID TIPPIN
financially. Yet, compared to other OECD nations, neither country has offered substantial
material support for parenting or for combining work and family responsibilities. Korpi
(2000) compared 18 industrialised countries on various measures of class and gender
inequality from 1985 to 1990 (see Table 1). While further change has occurred in some
countries in the 1990s, the comparisons are revealing. Australia ranked 16th and New
Zealand 17th in terms of general family support (child allowances, family tax benefits
and childcare services for over 3s). Australia ranked 15th and New Zealand 18th in terms
of government support for dual-earner families (childcare for children aged 0-2,
maternity/parental benefits, home help for elderly).
Table 1: Countries Ranked According to Levels of General Family Support and
Dual-Earner Support: 1985-90
General Family Support (a)
Dual-earner Support (b)
1 Belgium
2 Germany
3 France
4 Norway
5 Italy
6 Austria
7 Denmark
8 Ireland
9 Sweden
10 Finland
11 Netherlands
12 Canada
13United Kingdom
14 Switzerland
15 Japan
16 Australia
17 New Zealand
18 United States
1 Sweden
2 Denmark
3 Finland
4 Norway
5 France
6 Belgium
7 Germany
8 Italy
9 Netherlands
10 Austria
11 Ireland
12 United Kingdom
13 Canada
14 Japan
15 Australia
16 Switzerland
17 United States
18 New Zealand
Notes:
Source:
a) General Family Support includes cash child allowances as a percentage of net average
wage of single worker, family tax benefits, percentage of public childcare spaces for
children aged three to school age as a percentage of number of children in that age group.
b) Dual-earner Support includes public childcare spaces for children up to two as a
percentage of age group, a measure of the value of paid maternity leave and paternity
leave, and public home help for the elderly (percentage receiving services at home).
Korpi, 2000.
One-parent families led by mothers have had the lowest incomes in the ‘liberal’ welfare
states compared to the ‘social democratic’ or ‘corporatist’ welfare states (EspingAndersen, 1990). Poverty rates after taxes and transfers are close to 50 per cent for sole
46
HEALTH, BENEFCIARIES AND WELFARE TO WORK
mothers in the liberal countries compared to about two to three per cent in social
democratic countries (Baker and Tippin, 1999: 22). This suggests that government
policies can make a difference in reducing poverty rates.
Frequently inspired by neo-liberal principles, governments have altered social benefits to
reflect political priorities, demographic changes and the rising cost of social programs. In
New Zealand, all social benefits (including the DPB) were cut in 1991 by the National
Government. Several amendments were also made to eligibility rules for the DPB,
including raising the minimum qualifying age to 18 in 1991, introducing new
employment expectations after 1995, a new abatement scheme in 1996, and enforcing
new work tests after 1997. After school and school holiday care was also expanded for
the children of DPB recipients (Wilson, 2000: 79). At present, DPB (sole parents)
recipients must be actively preparing for full-time paid work if their youngest child is
aged 14 or over. If children are between six and 13 years, the mother is subject to a parttime work test (at least 15 hours per week). Those with younger children are expected to
attend annual planning meetings with a case manager, (DWI, 1999).
From the beginning, most recipients of the DPB were female. In 1998, 87 per cent of
recipients of this benefit were women. Of all recipients, 94 per cent of recipients were
sole parents, two per cent were women alone and four per cent were people caring for
sick or infirm family members. Almost three-quarters of recipients of the DPB are
separated or divorced (Wilson, 2000). Only nine per cent are under the age of 20, while
three-quarters are between 20 and 40 years old (Wilson, 2000).
The Australian and British Governments both faced considerable opposition when they
attempted to tighten the work test requirements for lone mothers on benefit with schoolaged children. In the May 2001 budget, however, the Australian Government announced
that mothers whose youngest child is 12 years old will have to attend an annual interview
about future employment prospects. By July 2003, parents whose youngest child is 13-15
years old will be expected to seek part-time work or training (Allard, 2001). By New
Zealand (or North American) standards, however, these rules appear generous. The
length of time sole mothers may remain on benefit is considerably longer in Australia,
Britain and NZ than in Canada or especially the United States. Neither the US nor
Canada permits mothers to receive social benefits if all of their children attend school,
unless there are extraneous circumstances such as disabilities. Instead, they encourage or
coerce mothers to enter the labour force when their youngest child is under six years and
often under three years old, depending on state or province. Some American states and
the Canadian province of Alberta expect mothers with infants to find paid work rather
than rely on government benefits (Baker and Tippin, 1999).
3
Sole Mothers and Paid Work
Previous studies suggest that sole mothers experience particular problems fulfilling the
requirements of welfare to work programs (Shaver et al., 1994; Edin and Lein, 1996;
Evans, 1996; Hardina, 1997; and Mullaly, 1997). Typically, they have lower levels of
47
education and job skills than do married mothers or fathers (Dooley, 1995), and many can
not find jobs with adequate wages to support themselves and their children. The jobs they
do find seldom include flexible work hours, paid sick leave, or extended health benefits.
In addition, sole mothers are less likely than married mothers to have another family
member to care for their children while they work and are also less likely to be able to
afford to pay someone to look after their children (Baker and Tippin, 1999). The
international literature review of Bryan Rodgers and Jan Pryor (1998) reports that sole
mothers sometimes experience emotional problems arising from the breakdown of their
marital relationships, continuing disputes with the children’s father, and behavioral
problems with their children that interfere with finding and keeping paid work.
Official statistics from OECD countries consistently show that the full-time employment
rates of sole mothers and married/partnered mothers vary substantially, as Table 2
indicates. These rates are influenced by employment opportunities, cultural values, the
availability of childcare, and public policies. Since the mid-1970s the gap has grown
between the employment rates of partnered and sole mothers in New Zealand,
particularly between 1981 and 1991 when the employment rate of sole parents (especially
sole fathers) dropped while the rate for partnered mothers steadily increased (Goodger
and Larose, 1999). The job market became more competitive, the percentage of sole
mothers in the population increased with higher divorce rates, and their average age
declined as fewer sole mothers were widowed and more were never married or
separated/divorced. The restructuring of the NZ economy also reduced employment
opportunities for younger less skilled people, including some sole parents, Maori, and
Pacific Island people.
More women than men work part time or in low-wage jobs. Women’s wages are most
likely to decline relative to men’s after they become mothers, and poverty rates are
particularly high if they become sole mothers. Mothers are more likely than fathers to opt
out of the labour force when family responsibilities, such as the care for young children,
clash with paid work. When mothers remain in or re-enter the labor force, they are more
likely than fathers to work in low-wage or temporary jobs, and to make work
accommodations or to disrupt paid work in order to balance work and family
responsibilities (Drolet and Morissette, 1997; O’Connor, Orloff and Shaver, 1999; Baker,
2001).
The probability of earning low wages and living in poverty is augmented if women are
also members of certain ethnic or cultural groups (Torjman and Battle, 1999). In both
Australia and New Zealand, indigenous people and Pacific Island peoples tend to
experience lower incomes and higher unemployment rates than the general population
(Baker, 2000).
Despite these trends, liberal welfare states have strengthened work tests. Moira Wilson
(2000) attempted to find out if the New Zealand changes had been successful in
encouraging women beneficiaries to find paid work. She noted that accurate data on
hours of work are unavailable but more DPB recipients in New Zealand are now
declaring earned income, especially those with children over seven years old and
48
Table 2: Percentage of Sole Mothers and Partnered Mothers Employed Full Time:
Early to Mid-1990s
Country
Sole Mothers
Partnered Mothers
Australia
Austria
Belgium
Denmark
Germany (reunited) (a)
Finland
France
Italy
Japan
Luxembourg
Netherlands
New Zealand
Norway
Portugal
Sweden
United Kingdom
United States
23
43
52
59
28
61
67
58
53
61
16
17
44
43
41
17
47
24
28
36
64
21
62
49
29
17
32
13
31
40
48
42
21
45
Note:
Source:
a) 36 plus hours
Bradshaw et al., 1996b
especially those with ‘European’ backgrounds. The lower declared earned income by
Maori recipients may reflect the fact that they are less likely to have formal educational
qualifications, and more likely to live in high unemployment areas and to have
responsibility for younger children. Wilson also notes that higher levels of declared
earned income could result from more employment or could reflect new rules that created
financial penalties for undeclared income (Wilson, 2000: 90). Clearly, fewer people
received the Domestic Purposes Benefit from 1996 to 1998, but many moved to another
benefit rather than finding full-time work. This is consistent with the research conducted
by Shaver et al., (1994) in Australia.
4
Sole Mothers, Poor Health and Paid Work
An underlying assumption of many advocates of welfare-to-work programs is that
unemployment contributes to social exclusion and depression, and conversely that paid
work is associated with well-being. This assumption was examined by Baker et al.
(1999), who compared a British sample of 719 sole mothers with 8779 partnered mothers
in order to test the hypothesis that paid work would improve sole mothers’ health. The
study found that sole mothers were significantly more likely than partnered women to
report poorer well-being and to have a major depressive disorder. While employment
49
may improve a sole mother’s economic situation, no significant relationship existed
between paid work and better health. In fact, employed sole mothers were more likely
than those who remained at home to report minor respiratory symptoms such as coughs
and colds.
Results from the Australian Longitudinal Study on Women's Health (1995-2000) indicate
the women in paid employment report better health than women without paid
employment (Bryson and Warner-Smith, 1998). However, the authors note that this
association may reflect self-selection factors: women in poor health are unable to find or
retain paid work. Bryson and Warner-Smith also found that women with both mothering
and employment commitments experience health deficits, especially in mental health.
Other results from the same study indicate that mid-age caregivers are less likely to be
employed full time and report poorer health than other women their age, especially higher
levels of stress (Lee, 2001). In the same longitudinal study, satisfaction with leisure was
strongly associated with mental and physical health for mid-age women, and those in
part-time paid employment expressed the highest level of satisfaction with the amount of
time for leisure (Brown and Brown, 1999).
Until recently, few studies focused on the poor health of sole parent beneficiaries as a
constraint in the transition to paid work. Nevertheless, health has surfaced as one of
several factors. The research by Shaver et al. (1994), on women who were reaching the
end of their Sole Parent Pension eligibility found that 30 per cent of those not currently in
training said that their own health or disability prevented them from doing so. Of those
women looking for some or more work, 20 per cent said that their or their child's
disability made it difficult to do so. One-quarter of the women who had been terminated
from the SPP and were now on unemployment benefits cited their own disability or their
child's as a barrier to employment (Shaver et al., 1994: 67). Nine to twelve months after
termination, of those women who were on other benefits, 12 per cent were receiving the
Disability Support Pension and six per cent were on the Sickness Allowance.
Dorsett and Marsh (1998) reported that high rates of cigarette smoking augment the
financial problems and poor health of U.K. sole mothers on benefits, and that poor morale
acts as an obstacle to returning to paid work. They argue that ‘welfare-to-work’ needs to
begin with ‘welfare-to-health’. The Policy Studies Institute and the National Centre for
Social Research recently surveyed Britain’s lone parents and low-income families with
dependent children, focusing on effectiveness of work incentives and a number of family
welfare issues. They found that people currently not working identified three constraints:
childcare issues, ill health and disability, and a lack of relevant skills and work
experience. One-third of lone parents reported a longstanding illness or disability (Marsh
et al., 2001).
Whitehead, Burström and Diderichsen (2000) noted that sole mothers reported lower
health status than partnered mothers in both Sweden and Britain, despite more favorable
social policies in Sweden that protected women from poverty and labour market
insecurity. The British social welfare system is less effective in keeping sole mothers
healthy and helping them out of poverty, both of which are related to the relative non50
participation of British sole mothers in the labour force (58 per cent in Britain compared
to ten per cent in Sweden). In both countries, not being employed was associated with
poor health for both sole and couple mothers, but the relationship was stronger in Britain
and stronger for sole mothers.
Whitehead and colleagues also found that unhealthy women were slightly more likely to
become or remain lone mothers, suggesting that there is a selection factor in the
relationship between sole mothers and poor health. Hope, Power, and Rogers (1999), also
came to this conclusion for Britain using the 1958 National Birth Cohort Study.
Whitehead and colleagues (2000) pose three hypotheses about the relationship between
sole motherhood and poor health. Sole mothers (including those who are employed) may
suffer from ‘time poverty’, which elevates their stress levels and leads to illness. The
kinds of work they do compared to married mothers may be more stressful and
dangerous. And finally, they may suffer from lower social support than couple mothers.
All these factors may help to explain the lower levels of reported health of Swedish sole
mothers despite the low levels of poverty, high standards of housing, and relatively
generous social services.
An exploratory study (Cook, Raine and Williamson, 2001) interviewed nine single
mothers in Alberta, Canada, where the welfare to work program expects sole mothers on
social benefits to search for work when their youngest child is six months old. Worries
about unpaid bills and lack of childcare services, and a sense of lack of control over their
lives encouraged stress. The inadequacy of social benefits encouraged unhealthy choices,
such as skipping meals so their children could eat or buying less nutritious but filling
food. The study concluded that Alberta’s welfare to work program overlooks women’s
care-giving role by expecting them to search for work while not providing childcare
services. By keeping benefit levels very low, the government is compromising the
physical and mental health of these families.
Sarfati and Scott (2001) extracted a sample of 721 sole mothers from the 1996-97 New
Zealand Health Survey to compare the socio-demographic characteristics and mental and
physical health of sole mothers and couple mothers. Sole mothers were more likely to be
Maori, to have lower family incomes, lower educational qualifications, and to live in
more deprived areas. Sole mothers reported poorer physical and mental health than
couple mothers, but after controlling for socioeconomic variables, only mental health
differences remained.
A report by New Zealand’s National Health Committee reiterated other international
studies that ‘income is the single most modifiable determinant of health and is strongly
related to health and well-being’ (NHC, 1998: 8). The poor can less afford to purchase
more nutritious food, dryer and warmer accommodation, better hygiene, warmer clothes,
preventative health services, and better health care. Those in poverty are also more likely
to work in dangerous jobs, live in high crime areas, and to engage in lifestyles that can
have greater risks.
51
Canadian research indicates that children living in poor families have more than twice the
incidence of chronic illness and physical and developmental disability as children living
in non-poor families, and are more than twice as likely to drop out of school (Ross, Scott
and Kelly, 1996). Children in poor families are also more likely to experience social
impairments and psychiatric, emotional, hyperactivity, and conduct disorders (Lipman,
Offord and Boyle, 1994). British research (Roberts, 1997) reports similar findings, noting
that childhood poverty is linked to higher rates of sickness and premature death, but also
poor educational attainment, greater chance of delinquency, and higher rates of
unemployment in adult life.
In the past two decades, poverty and inequality have increased in many OECD countries,
and has disproportionately affected single parent and indigenous households (Baker and
Tippin, 1999; Torjman and Battle, 1999). These trends have negative consequences for
the health of marginalised groups such as sole parents - particularly lone mothers - at a
time when they are under greater pressure to enter paid work. Moreover, poor health can
interfere with their ability to secure and retain a job, which could further exacerbate
existing problems they may be having with welfare to work programs. Taken together,
the recent research suggests that poor health may be a more important factor in the
transition to employment than previously thought.
5
Health, Domestic Purposes Beneficiaries and the Transition to Paid
Work
We are currently studying, with the collaboration of the NZ Department of Work and
Income (DWI), the health constraints in the transition to paid work for sole mothers on
the Domestic Purposes Benefit. As the project is ongoing, the results reported below are
preliminary in nature and at this point in the research should be viewed with some
caution. Three areas deserve highlighting.
Health-related Aspects of the DPB (Sole Parents)
The DPB (sole parents) is a work-tested benefit. Exemptions or deferrals from a full-time
or part-time work test, the Yearly Planning Meeting, and/or Work Preparation Activity3
can occur for a variety of health-related reasons (see Table 3). In all cases, the assessment
of medical status (either of the recipient or dependant) is made by a recognised medical
practitioner/health professional, of the recipient’s choosing. The health practitioner may
be requested to provide relevant details such as: nature of the health-related problem; type
and length of time care may be needed; the risk (e.g. to a dependant child) if the
3
Participation in a Work Preparation Activity (WPA) generally occurs when the
beneficiary’s youngest child reaches the age of five. A WPA (which might be a DWI
course in budgeting, CV preparation, voluntary work, or externally-run courses or
activities that have a paid work focus) is designed to increase self-confidence, develop job
search skills, and generally prepare the beneficiary for the time when they will have formal
part-time work-test obligations.
52
beneficiary is sent onto training or another work-related activity outside the home; and a
recommendation concerning length of time a deferral should be granted. Case managers
have discretion to request a second medical opinion by a physician nominated and paid
for by DWI. Beneficiaries receiving an exemption or deferral from work testing are
obligated to report to DWI any significant changes in health status, including
hospitalisation. Although there are many grounds for a DPB (sole parent) recipient to
request an appeal or review of a DWI decision, no right of appeal is granted for medical
reasons.
Beneficiaries may also receive a range of health-related supplemental allowances. Among
these are: a Disability Allowance (income-tested) can be provided to an individual or a
financially dependent child under age 18 who has a disability of at least six months
duration and requires on-going medical care and help with everyday tasks. The allowance
assists with expenses such as doctor/hospital visits, medicines, extra clothing or travel.
The Disability Allowance can be combined with a benefit such as the Invalids Benefit. A
DPB (sole parent) recipient eligible for this allowance may be transferred to another
benefit, such as the Invalids Benefit. A Child Disability Allowance (not income-tested) is
provided to the main caregiver of a child less than 18 years of age who has a serious
physical or intellectual disability, and who needs constant care and attention for at least
one year. It is possible to receive both allowances for the same child (DWI, 2001).
The Health Status of Beneficiaries
Part of our research involved administering a questionnaire to a purposive and stratified
sample of over a thousand women currently on the DPB whose youngest child was
between seven and 14 years of age and who were subject to a work test (none were
exempt from the work test for disabilities or illness). The questionnaire included the SF36, an internationally recognised self-reported health status/health quality instrument
used by the New Zealand Ministry of Health in the survey Taking the Pulse (1999).
The SF-36 is constructed to represent two dimensions of health: physical health and
mental health. As well as asking questions about beneficiaries’ health and their use of
health services, our survey gathered information about family and household,
employment and education.
The sample drew from three very different North Island socio-economic and ethnic areas
served by DWI offices.
The north Auckland suburb of Brown’s Bay, which is disproportionately Pakeha or
‘European’ New Zealander, and rates three out of 10 on the NZ Deprivation Index (with
10 being most deprived);
53
Table 3: Possibilities for Exemptions/Deferrals from Work Testing for Healthrelated Reasons
Health-related issue
Beneficiary
sickness, injury or
disability (a) that
would qualify the
individual for an
Invalids Benefit
Caring full time for
a person (including
child or other
dependant) who is
sick, injured or
disabled
Caring full time for
a dependant child
with ‘special needs’
(b)
who lives at
home.
Pregnancy (c)
Notes:
Work Test
Yearly Planning
Meeting
Work Preparation
Activity
Exemption
or deferral
Exemption
or deferral
Exemption
or deferral
Exemption
Exemption
Exemption
Exemption
Exemption
Exemption
Exemption
Attendance required
Exemption
a)Temporary sickness, injury or disability (that would not require an Invalids Benefit)
can also result in a deferral being granted for the work test, Yearly Planning Meeting,
and Work Preparation Activity.
b)‘Special needs’ include: physical or intellectual disability; medical condition; chronic
or recurring illness; or a learning or behavioural difficulty.
c)Pregnancy exemptions apply from the date a pregnancy is confirmed by a doctor or
registered midwife, up until the birth due date.
Derived from Department of Work and Income (NZ), Manuals and Procedures 2001
(departmental intranet).
Source:
•
Otara, south of Auckland, rates 10 on the Deprivation Index and has a high
proportion of Pacific Island peoples; and
•
Kaitaia is rural area in the Far North, is predominantly Maori, and also rates
10 on the Index.
The questionnaire was returned by 244 women who were disproportionately Pakeha. We
compared our survey results to those obtained for women of comparable age in the New
Zealand health survey (Taking the Pulse). Three significant features can be identified.
54
First, we found that the women on the DPB reported poorer health than women of
comparable age in the general population. In the past 12 months, a higher proportion of
the DPB women had used public hospital emergency services than women in the general
population, and our sample also reported a higher rate of unmet health need. A higher
percentage reported that they have not gone to a general practitioner when needed or to
have not picked up a prescription when needed. The major reason given was cost (76.9
per cent of respondents who had not visited a GP when needed said this).
Second, the mean scores for DPB women for the eight different components of health
measured by the SF-36 questionnaire were lower on every measure than the comparable
general population group. These measures include scales of both physical and social
functioning, problems with work and other daily activities relating to both physical and
emotional health, bodily pain, vitality, social functioning, mental health and general
health, as Figure 1 indicates.
Third, beneficiaries reported health as a significant welfare to work obstacle.
Respondents were asked what issues prevented them from taking up paid work (see Table
4). Half our sample said that their responsibilities for children prevent them from finding
a job. Over a third reported ‘inability to find a job’ (38.1 per cent) as a constraint, 30 per
cent said ‘lack of training’, and 27.5 per cent said that ‘transport difficulties’ prevented
them from taking up paid employment. However, 29.1 per cent said that their health
prevented them from taking a job, which is quite striking when it is considered that these
women have not met the policy criteria for exemption from their work/training
obligations due to their health or disability. In addition, 21.7 per cent reported that their
children’s health prevented them from taking a job and 9.4 per cent said that the health of
other dependants was a constraint. Among those who said they could not find a job, a
disproportionate number were older, Maori and Pacific Island in ethnicity, low in formal
education, and had more children in the household. A higher percentage of those who
reported their health as a reason for not taking a job had lower educational attainment and
lower incomes.
These findings suggest that self-reported health status, as measured by the SF-36, is lower
in this NZ beneficiary population than in the general female population of comparable
age. This is consistent with international studies that cite a strong correlation between low
levels of self-reported health and sole motherhood status.
Services and Decision-making for DPB Recipients
Part of the overall mandate of the NZ Department of Work and Income includes
providing ‘services to provide benefit entitlements and obligations to working age
beneficiaries and to promote self sufficiency’, as well as reducing unemployment and
moving people into work. These objectives require ongoing assessments and ‘…
management of people’s need for income support and other assistance, including
responding to changes in people’s circumstances in a timely and appropriate manner’
(DWI, 2000: 14). These changes can, of course, include those attributable to health
issues.
55
Figure 1: Comparison of SF-36 Health Scores of DPB Women and NZ Women
100
90
DPB beneficiaries
80
NZ women
70
60
PF–Physical
functioning
RP – Role-physical
BP – Bodily pain
GH – General health
VS – Vitality score
SF – Social
functioning
RE – Role-emotional
MH – Mental health
50
40
30
20
10
0
PF
RP
BP
GH
VS
SF
RE
MH
Source: DPB Study by Maureen Baker, Heather Worth, David Tippin and Tracey
McIntosh, University of Auckland.
Table 4: Main Issues Preventing Women on the DPB Taking up a Paid Job
Percentage Reporting (a)
52.5
38.1
30.0
29.1
27.5
21.7
9.4
Issue Preventing Paid Work
The care of children
Inability to find a job
Lack of training
Poor health
Transport difficulties
Poor health of child(ren)
Poor health of other dependants
Note:
a) Multiple answers permitted, so percentages do not add to 100
Source:
DPB Study by Maureen Baker, Heather Worth, David Tippin and Tracey McIntosh,
University of Auckland. Calculations prepared by Karen McMillan (Assistant Research Fellow).
The findings and comments in this section are based on a small number of interviews and
should be regarded as preliminary and subject to modification as we accumulate further
56
data. We have done group interviews with about 20 New Zealand government policy and
operations staff (including some seconded case managers) responsible for aspects of the
Domestic Purposes Benefit. Their comments indicate that the dynamics of client/case
manager interactions and decision-making processes can be particularly complex as they
pertain to health issues. Two aspects deserve mention.
First, training and guidelines regarding health-related matters may sometimes be
exceeded by the complexity of the individual circumstances that case managers can
encounter. Initial health-related interactions with a client usually focus upon issues such
as identifying disabilities (which may qualify for other benefits or allowances), and
determining whether there are any ongoing costs related to a medical condition. DWI and
its case managers rely upon the NZ Ministry of Health to set broad policy directions with
respect to health-related matters. This means that DWI procedural manuals may not
encompass all of the detailed and ‘grey’ issues relating to broader trends in contemporary
health care delivery that can arise in interactions with clients. These include the
appropriate use of alternative medical practitioners or the movement towards community
care, and the relationship between health status gradations or fluctuations and the ability
to undertake part-time or full-time work.
There are a number of learning mechanisms available to case managers that can be the
means of increasing and transmitting knowledge of health-related issues situations. Case
managers participate in weekly training sessions in service centres and work in a ‘buddy
system’ with colleagues, and more informal exchanges of case histories and requests for
second opinions occur. A national ‘Helpline’ service (available to assist case managers)
can identify recurring issues that merit a change in operational policy.4 Formal forums
with beneficiary advocacy groups have occurred concerning health issues. For some case
managers, their decision-making draws upon their integration into the local community
and their knowledge of its health-related support services (such as counselling).
Second, interaction and decision-making can also occur in an environment of information
uncertainty and time pressure. Mental and emotional health issues - beyond those that
have been previously identified by qualified medical practitioners - are acknowledged as
difficult. Case managers may be reluctant to raise, or may overlook, issues of depression
or poor motivation, and clients may be unwilling to admit that a problem exists. The
open-concept physical structure of a typical DWI office may inhibit health discussions,
particularly those that pertain to mental/emotional issues. Similarly, the time constraints
of an initial hour-long interview, most of which is typically occupied by processing a
client to allocate the correct benefits, may leave little time to raise and address more
subtle health concerns. In addition, as one staff member observed, the number and detail
of tasks at the initial meeting can ‘be too much for the client to take in themselves’
particularly if she is in a state of personal distress. Clients may not recognise that they are
4
Discrepancies between forecast and actual expenditures for particular benefits and
allowances are also monitored by DWI to identify health-related gaps/issues that may
merit further attention.
57
stressed or depressed, and may confound mental health issues with questions of low selfesteem and lack of confidence.
The full picture of the health circumstances of a work-tested client and their dependants
may not be apparent until after further encounters and after a rapport has built up between
case manager and client. Indeed, it was suggested by DWI staff that health issues can
sometimes become quite salient only after work training or paid work is started,
whereupon questions of what to do about sudden onsets of illness of children or
availability of sick leave become more urgent. That is, it is the transition from the relative
predictability and security of the DPB to the world of paid work that can change, often in
stressful ways, many aspects of a beneficiary’s life and challenge her coping skills.
While we are still early in the process of conducting interviews, initial indications from
organisations involved in providing support, information and advisory services to
beneficiaries suggest that several health-related dimensions are of concern. Advocacy
groups’ experience confirms the SF-36 survey findings in the sense that a portion of their
caseload includes assisting beneficiaries with health-related issues, such as being put on
the correct benefit for the health circumstances of themselves and/or dependants. Health
status is also cited as an important factor in both going on and getting/staying off the
DPB. In the former instance, the physical and emotional stress of the combination of lowpaid work and childcare – a general feeling of not being able to cope – can result in a
decision to apply for the DPB. For a beneficiary who is in paid work, the sudden onset of
acute illness for children, or the recurrence of a chronic childhood illness such as asthma,
can lead to the use of all sick leave (if any is provided), meaning that future illness can
place the job in jeopardy. Advocacy groups are also concerned that the health-related
benefits of the Community Services Card (which provides beneficiaries and their
dependants with subsidised physicians’ fees and prescriptions) can be reduced or lost to a
work-tested DPB recipient if too much income is earned.
Advocates also felt that the level of health-related incapacity was set too high, and that
gradients of incapacity should be integrated further with work-testing requirements and
the benefit structure. This is linked to the perception that the capacity of case managers to
invite, detect and act upon health-related issues is variable – which confirms the
perceptions of some DWI staff. This situation can also be reinforced by beneficiaries’
own behaviour. For example, as one advocate put it, the ‘discourse of poverty’ of some
beneficiaries includes a reluctance to speak about mental health issues that are recognised
by themselves and others to be socially stigmatised and poorly understood.
What emerges from the work we have done thus far on this project is that the complexity
of the lives of many DPB recipients extends to the health circumstance of themselves and
their dependants. Health-related issues become particularly acute in the successful and
sustained transition from welfare to work, especially when the thoroughness of health
assessments – in the context of other social factors – remains in doubt.
Future work on this project will include casting further light on the dynamics of how
health-related matters are raised and addressed in decision-making. Qualitative interviews
58
will be conducted with 100 sole mothers on the DPB, to obtain more details about factors
related to their health and their views on how these issues are recognised and
incorporated into work-related assessment and decision-making by their case managers.
We will also interview case managers based in DWI service centers to ask them how they
view and assess beneficiaries’ health and how they consider health status and health
issues in case management strategies and practices. Finally, more interviews will be
conducted with national government staff responsible for operations, policy planning and
development, as well as more representatives from beneficiary advocacy groups.
6
Conclusion
For some time, social scientists have known that there is a strong link between poverty
and poor health. In the past few years, more researchers have examined the connection
between sole motherhood and poor health, attempting to identify various pathways to
poverty. Yet few studies have tried to link poor health and the implementation of welfareto-work programs.
We already know that the transition to paid work is difficult for many low-income
mothers (of whom DPB sole parents comprise a subset) for a variety of reasons: they tend
to have low education and poor job skills and are likely to obtain only short-term, lowpaid jobs. Many have recently experienced separation or divorce, and some continue to
deal with emotional stress, children’s behavioural problems, and disputes with former
partners, which interfere with finding and holding a job. Low-income mothers have few
economic resources to pay for childcare services and they are justifiably unwilling to
leave their children with unreliable caregivers. In addition to these issues, many report
health problems, either for themselves or for their children, which make returning to and
staying in paid employment difficult.
The preliminary findings of our current study raise some important policy issues. In one
sense, welfare-to-work policies focus on the broad issue of reducing the number of
citizens who rely upon social benefits. At a deeper level, the success of these policies
depends partly upon matching beneficiaries’ skills and capabilities to particular jobs,
while providing other social supports, such as affordable childcare and transport to and
from work. For the majority of beneficiaries, the types of low-waged job for which they
will be eligible will typically involve tasks that require some physical stamina (such as
standing for long periods of time or walking about). The health status component of our
research project suggests that beneficiaries are less robust physically than the general
female population of comparable age. Thus, although other supporting mechanisms may
be put in place to facilitate the transition to paid work, the process of finding and keeping
a job may founder on the poor physical and mental health of the beneficiary.
A compounding factor is the comparatively high self-reported rate of mental health
problems. Were these mental health problems part of the reason for becoming sole
mothers or staying as sole mothers, or do the impoverished circumstances and strains of
living in poverty contribute to poor health? As we delve further into the health status and
59
perception issues, the transition from welfare to work appears even more complex than
previously thought. If ways are not found to support a reasonable level of physical and
mental health for beneficiaries as they move towards paid employment, then the
sustainability of welfare-to-work programs could be compromised.
If we want to assist mothers on social benefits to make a successful transition to paid
work, they clearly need institutional supports in place, especially affordable childcare and
paid maternity or parental benefits. They also need to be assured that when they return to
paid work they will not lose access to subsidised health services or experience so many
stresses that retaining paid work is unhealthy for them.
The women, DWI staff and advocate group representatives to whom we have spoken thus
far in the research seem to acknowledge the numerous barriers to sole mothers entering or
returning to paid work, including health factors. The research suggests that there may be
at least two frameworks governing health and employability in New Zealand. One
framework consists of a bureaucratic, rules-based system reliant on judgements of
medical professionals, leading to authoritative decisions on medical fitness for work
testing. A second perspective is that ‘good health’ for beneficiaries is embedded within a
complex and changing web of social circumstances, that paid work itself can have health
consequences, and that the universe of ‘health-related issues’ is a broader one than
suggested by the bureaucratic framework. This leads to a more contingent view of what
constitutes ‘employability’. We need to complete our research in order to understand the
ways in which these visions of employability may be both competing and
complementary, and how they are incorporated into both social policy and case
management.
References
Allard, Tom (2001), ‘Payment tests tightened but child care boosted’, Sydney Morning
Herald, 23 May 2001, online, www.smh.com.au.
Baker, Maureen (2000), ‘Labor market poverty, gender and the global economy:
comparing Canada, Australia, and New Zealand’, Social Work and Globalization,
Special Issue of Canadian Social Work Review, 17, July, 88-104.
Baker, Maureen (2001), Families, Labor and Love, Allen & Unwin, Sydney, and
University of British Columbia Press, Vancouver.
Baker, Maureen and David Tippin (1999), Poverty, Social Assistance and the
Employability of Mothers: Restructuring Welfare States, University of Toronto
Press, Toronto.
Baker, D., K. North and The ALSPAC Study Team (1999), ‘Does employment improve
the health of sole mothers?’ Social Science and Medicine, 49, 121-31.
Bradshaw, Jonathan, Steven Kennedy, Majella Kilkey, Sandra Hutton, Anne Corden,
Tony Eardley, Hilary Holmes and Joanne Neale (1996a), Policy and Employment
of Sole Parents in 20 Countries, The EU Report, Commission of European
Communities.
60
Bradshaw, Jonathan, Steven Kennedy, Majella Kilkey, Sandra Hutton, Anne Corden,
Tony Eardley, Hilary Holmes and Joanne Neale (1996b), The Employment of Sole
Parents: A Comparison of Policy in 20 Countries, Family Policy Studies Centre,
Paper No. 13, London.
Brown, P. and W. Brown (1999), ‘Women and leisure: does all work and no play make
Jill unwell?’, World Leisure and Recreation, 41, 11-4.
Bryson, Lois and P. Warner-Smith (1998), ‘Employment and women's health’, Just
Policy, 14, 3-14.
Cook, Kay, Kim Raine and Deanna Williamson (2001), ‘The health implications of
working for welfare benefits: the experiences of single mothers in Alberta,
Canada’, Health Promotion Journal of Australia, 11(1), 20-6.
Dooley, Martin (1995), ‘Sole-mother families and social assistance policy in Canada’, in
Dooley et al., Family Matters: New Policies for Divorce, Sole Mothers, and Child
Poverty, CD Howe Institute, Toronto, 35-104.
Dorset, Richard and Alan Marsh (1998), The Health Trap: Poverty, Smoking and Lone
Parenthood, Policy Studies Institute, London.
Drolet, Marie and René Morisette (1997), ‘Working more? What do workers prefer?’,
Perspectives on Labor and Income, 9(4), 32-8.
DWI (NZ Department of Work and Income) (1999), Work Test: Work Services
Handbook. 7/6/99 Internet.
DWI (NZ Department of Work and Income) (2000), Purchase Agreement between the
Minister of Social services and Employment and the Chief Executive of the
Department of Work and Income (1 July 2000 to 30 June 2001).
DWI (NZ Department of Work and Income) (2001), Disability Allowances: a Guide for
People with Disabilities, Information Pamphlet.
Edin, Kathryn and Laura Lein (1996), ‘Work, welfare and single mothers: economic
survival strategies’, American Sociological Review, 61, February, 253-66.
Esping-Andersen, Gøsta (1990), The Three Worlds of Welfare Capitalism, Polity Press,
Cambridge.
Evans, Patricia (1996), ‘Single mothers and Ontario's welfare policy: restructuring the
debate’, in Janine Brodie, ed., Women in Canadian Public Policy, Harcourt Brace
and Company, Toronto, 151-71.
Goodger, Kay and Peter LaRose (1999), ‘Changing expectations: sole parents and
employment in New Zealand’, Social Policy Journal of New Zealand, 12, July,
53-70.
Hardina, Donna (1997), ‘Workfare in the US: empirically-tested programs or ideological
quagmire?’, in Eric Shragge, ed., Workfare: Ideology for a New Under-Class,
Garamond, Toronto, 131-48.
Hope, S., C. Power, and B. Rogers (1999), ‘Does financial hardship account for elevated
distress in sole mothers’, Social Science and Medicine, 49(12), 1637-49.
Korpi, Walter (2000), ‘Faces of inequality: gender, class, and patterns of inequalities in
different types of welfare states’, Social Politics, 7(2), 127-91.
Lee, C. (2001), ‘Family caregiving: a gender-based analysis of women's experiences’, in
S. Payne and C. Ellis Hill, eds., Chronic and Terminal Illness: New Perspectives
on Caring and Carers, Oxford University Press, Oxford, 123-139.
61
Lipman, E.L., D.R. Offord and M.H. Boyle (1994), ‘Relation between economic
disadvantage and psychosocial morbidity in children’, Canadian Medical
Association Journal, 151 (4), 431-7.
Marsh, A., S. McKay, A. Smith and A. Stephenson (2001), Low-Income Families in
Britain: Welfare and Social Security in 1999, Department of Social Security
Research Report No. 138, UK Government, London.
Mullaly, Robert (1997), ‘The politics of workfare: NB works’, in Eric Shragge, ed.,
Workfare: Ideology for a New Under-Class, Garamond Press, Toronto.
National Health Committee (NHC), (1998), The Social, Cultural and Economic
Determinants of Health in New Zealand: Action to Improve Health, A Report
from the National Advisory Committee on Health and Disability, Wellington.
New Zealand Ministry of Health (1999), Taking the Pulse: The 1996/97 New Zealand
Health Survey, Ministry of Health, Wellington.
O'Connor, Julia S., Ann Shola Orloff and Sheila Shaver (1999), States, Markets,
Families. Gender Liberalism and Social Policy in Australia, Canada, Great
Britain and the United States, Cambridge University Press, Cambridge.
Roberts, Helen (1997), ‘Children, inequalities and health’, British Medical Journal,
314(7087), April, 1122-4.
Rodgers, Bryan and Jan Pryor (1998), Divorce and Separation: The Outcomes for
Children, Joseph Rowntree Foundation, York.
Ross, David, K. Scott and M. Kelly (1996), Child Poverty: What are the Consequences?,
Canadian Council on Social Development, Ottawa.
Sarfati, Diana and Kate Scott (2001), ‘The health of lone mothers in New Zealand’, The
New Zealand Medical Journal, 114(1133), 257-60.
Shaver, Sheila, Anthony King, Marilyn McHugh and Toni Payne (1994), At the End of
Eligibility: Female Sole Parents Whose Youngest Child Turns 16, SPRC Report
No. 117, Social Policy Research Centre, University of New South Wales, Sydney.
Torjman, Sherri and Ken Battle (1999), Good Work. Getting it and Keeping it, Caledon
Institute of Social Policy, Ottawa.
Uttley, Stephen (2000), ‘Sole mothers and policy discourse in New Zealand’, Journal of
Social Policy, 29(3), 441-58.
Whitehead, Margaret, Bo Burström and Finn Diderichsen (2000), ‘Social policies and the
pathways to inequalities in health: a comparative analysis of sole mothers in
Britain and Sweden’, Social Science and Medicine, 50(2), 255-70.
Wilson, Moira (2000), ‘The policy response to the employment task force and changing
patterns of domestic purposes benefit receipt: a cohort analysis’, Social Policy
Journal of New Zealand, 14, July, 78-103.
62
Changes in the Gender Division of Household
Labour in Australia, 1986-1997
Janeen Baxter
School of Sociology, The University of Queensland
1
Introduction
Recent years have seen increased sociological attention focused on trends in domestic
labour patterns and the gender gap in men’s and women’s contribution to child care and
housework (Gershuny and Robinson, 1988; Shelton, 1992; Bittman, 1995; Bianchi et al.,
2000). The impetus for this increased attention stems, in part, from other trends which
focus attention on possible associated changes in the way men and women organise their
household responsibilities. These include the increased participation of married women in
paid employment, a much smaller, but nevertheless significant decline in men’s labour
force participation rates, the decline in levels of childbearing, the delay in entering a
marital relationship, increased rates of de facto cohabitation, and increasing divorce rates.
In other words, the expectation has been that changes in patterns of family formation and
dissolution, in conjunction with the changing gender distribution in paid work, would
lead to changes in the distribution of work between men and women in the home
(Hochschild, 1989; Baxter, 1993; Brines, 1994; Western and Baxter, 2001). The results of
the research are far from clear-cut. Most research tends to suggest that women’s hours on
housework are declining, but there are mixed views about whether men’s hours on
housework have changed. Some research has found that men’s housework contribution
has increased (Gershuny and Robinson, 1988; Robinson and Godfrey, 1997; Bianchi et
al., 2000), while others have found no change in men’s contribution (Coverman and
Sheley, 1986; Bittman, 1995). This paper examines men’s and women’s participation in
housework using repeated national cross-sectional Australian data from 1986, 1993 and
1997. The aim is to examine change over time in men’s and women’s levels of
participation in child care and housework.
2
Previous Research
There is remarkably little research that directly examines change over time in men’s and
women’s involvement in housework. In Australia the only research is that of Bittman
(1995). Bittman uses the 1992 Australian Bureau of Statistics Time Use Survey in
conjunction with pilot time use surveys conducted in 1987 and 1974. While these data are
quite detailed, as is the case with time diary methods, they are somewhat limited by being
partially based on pilot work in a major urban centre.
Baxter, J. (2002), ‘Changes in the gender division of labour in Australia, 1986-1997’, in T. Eardley and B.
Bradbury, eds, Competing Visions: Refereed Proceedings of the National Social Policy Conference 2001,
SPRC Report 1/02, Social Policy Research Centre, University of New South Wales, Sydney, 64-74.
JANEEN BAXTER
Bittman (1995) reports a number of key findings.
•
•
•
•
•
Between 1987 and 1992 women reduced their time on housework by nearly three
hours per week.
This comes primarily from a reduction in the amount of time spent on cooking, and a
reduction in time spent on laundry.
In contrast women spent an increased amount of time (a 21 minute increase) on home
maintenance and car care.
There were no significant changes in men’s time on housework.
For both men and women, there was a small but significant increase in the amount of
time devoted to child care.
Bittman concludes that while there has been some improvement in the gender sharing of
domestic labour, the changes are not the expected ones (i.e. men doing more) but rather
are due to women reducing the time they spend on unpaid labor. Moreover, he finds that
the core housework activities of cooking, cleaning and laundry are still mainly the
responsibility of women.
In the United States, the most recent and comprehensive analyses of change in domestic
labour patterns are the work of Bianchi, Milkie, Sayer and Robinson (2000). Using time
diary data from repeated cross-sectional representative samples of American adults
between 1965 and 1995, they find that women have reduced their time on housework by
about 12 hours per week (down from about 30 hours per week in 1965 to about 17.5
hours in 1995). During this period men’s housework time doubled (from about five to ten
hours per week), with the largest increases occuring prior to 1985. Their findings differ
somewhat from Bittman’s results in that they find that men have increased their time on
housework. But overall, like Bittman they conclude that the declining gender gap in
housework has more to do with a decline in women’s hours on housework than an
increase in men’s hours on housework. And like Bittman, and all other studies of
housework, they find that the core housework activities of cooking, cleaning and laundry
are still the main responsibility of women. They also report that cooking is an area that
shows the greatest convergence in men’s and women’s time, with ’women’s reported
hours 8.8 times higher than men’s in 1965 but only 2.8 times men’s hours in 1995‘ (2000:
207).
3
Data
The data for this paper come from three national cross-sectional studies – the 1986 Class
Structure of Australia Project, the 1993 Class Structure of Australia Project, and the 1997
Negotiating the Lifecourse Project. All of the surveys are based on nationally
representative surveys, although all had differing parameters. The 1986 survey was based
on a representative sample of the employed Australian population, the 1993 survey was
based on a sample of the entire population, while the 1997 survey was restricted to the
population between 18 and 59 years of age. In order to generate comparable estimates
over time I have restricted all of the samples at each point in time to employed
respondents (in either full or part-time employment) aged between 18 and 59 years.
Additionally, in the 1986 survey, the questions on housework were only asked of married
or cohabiting respondents. I therefore confine all of the analyses on housework for each
65
CHANGES IN THE GENDER DIVISION OF HOUSEHOLD LABOUR IN AUSTRALIA 1986-1997
survey to this subgroup. Similarly the questions on child care were only asked of
respondents with children under 10 years of age living in the household and hence for all
of the child care analyses I restrict the samples to this subgroup.
As Table 1 shows, each of the surveys adopted a different method of data collection. The
1986 survey was conducted via face-to-face interviews, the 1993 survey was a selfcompleted mail out survey, while the 1997 one was conducted using telephone interviews
employing Computer Assisted Telephone Interviewing (CATI). The table also shows the
total sample size and the sample size once the above restrictions relating to marital status,
employment, age and age of children are imposed. Note that in each survey only one
member of the household was interviewed
Table 1: Data, Sample Size and Data Collection Method
Project
Year
Total
Sample
The Class
Structure of
Australia Project
The Class
Structure of
Australia Project
1986
Negotiating the
Lifecourse
4
1195
Married /
Cohabiting
Sample
754
Married /
Cohabiting With
Children Sample
329
Method of
Collection
1993
2780
1081
789
Self complete
mail-out
questionnaire
1997
2231
1116
596
Telephone
interviews
Face-to-face
interviews
Variables
The dependent variables in these analyses are a set of measures of the domestic division
of labor – some relating to child care and others concerning housework activities. In some
of the surveys the range of tasks included was more comprehensive than in others. For
reasons of comparability however, I have restricted my analyses to those items that are
common across each of the surveys. For child care I focus in this paper on the
proportionate contribution of men and women to child care tasks. For housework
however, I examine both the proportionate contribution of husbands and wives to
particular activities, in addition to the amount of time spent doing them. This is important
since some activities require very little time to complete, while others are more time
consuming. Often the tasks undertaken by men, for example taking out the garbage and
mowing the lawn, are tasks that are undertaken only once a week or less, whereas the
tasks routinely completed by women, such as cooking and cleaning up after meals, tend
to be daily activities and often more time consuming. It is possible therefore to have an
equal division of labor where both husband and wife undertake 50 percent of the
domestic work, but in which women spend considerably more time than men on domestic
labor.
66
JANEEN BAXTER
The child care tasks were: bathing and dressing children, getting children to bed and
helping children with homework. The housework tasks were: preparing meals, cleaning
up after meals, grocery shopping, cleaning the house, taking out the garbage, washing,
ironing, mowing lawns, gardening and home maintenance/repairs.
The response categories were ’I do most‘; ’I do more‘; ’we share this equally‘; ’my
partner does more‘; and ’my partner does most.’ These responses were subsequently
coded as percentages and then summed to create a scale ranging from 0 to 100 per cent
reflecting the relative contribution of each spouse. For example, a respondent who
reported doing most of a particular task was coded as 100 indicating that they take full
responsibility for this task, while a respondent who reported that their partner had most
responsibility for a task was coded as 0.
I do most = 100 per cent
I do more = 75 per cent
We share this equally = 50 per cent
My partner does more = 25 per cent
My partner does most = 0 per cent
Respondents were also asked to indicate how much time they spent on each of these
activities in an average week.
There are at least two ways of examining change over time in the domestic division of
labour: first in terms of changes over time in the gender gap between men’s and women’s
contributions; and second in terms of changes over time in men’s and women’s individual
contributions. I will be examining both kinds of change in this paper. The analyses
reported below examine the means for men and women at each time point on each of the
dependent variables. In order to test gender differences and differences over time, I
estimate a series of multiple regression models in which the independent variables are
time coded as a dummy variable with three categories and gender coded as a dummy
variable with two categories. In the tables below, I report the least squares means from
these models in addition to the significance of differences between these means.
5
Results
Table 2 shows the percentage distribution of child care tasks for men and women at the
three time points. Overall women report more responsibility for child care than men,
doing 66 per cent to 69 per cent compared to between 37 per cent and 34 per cent for
men. All of the gender differences are statistically significant. Over time there is no
indication that the gender gap in responsibility for child care is diminishing. Rather if
anything, it appears that the gender gap is getting larger, particularly in relation to helping
children with homework, which increases from a 29 per cent difference in 1986 to a 35
per cent difference in 1997.
67
Table 3 shows the significance of change over time in men’s and women’s contribution
to child care. The arrows indicate the direction of change. The table only reports
statistically significant change as indicated by the results of significance tests of
differences in least squares means. Overall, there is no change when the child care tasks
are combined into a single scale, but as indicated in the previous table, there is a
significant change in the distribution of men’s and women’s responsibility for helping
children with homework. Women’s percentage responsibility for this activity has
increased by seven per cent while men’s has declined by six per cent. These results differ
from other research which has generally found that men’s responsibility for child care has
increased over time (Bittman, 1995). However, this increase is generally associated with
an increase in time spent playing with children. The current research suggests that in
terms of the day-to-day activities associated with very young children, men are not doing
an increased share. If anything, these data suggest that men’s involvement with schoolage children may have declined.
Table 2: Least Squares Means for Percentage Distribution of Childcare Tasks for
Men and Women, 1986, 1993, 1997
Men
1986
Women
Men
1993
Women
Men
1997
Women
Bathing and
dressing children
31
69
38
68
30
69
Getting children
to bed
43
65
44
65
43
62
Helping children
with homework
39
68
43
67
34
72
Total
38
67
42
66
35
66
N
230
99
455
373
316
278
Table 4 shows the percentage distribution of housework tasks. Again there is a clear
gender division of labour with women doing the bulk of the work. And once again, all of
the gender differences are statistically significant. The results show a clear gender
division in the kinds of tasks that women do and the kinds of tasks that men do. Women
are still doing the bulk of the traditional routine everyday housework activities such as
cooking, cleaning and washing, while men still have most responsibility for the
traditional male outdoor activities such as mowing lawns, garbage removal and home
maintenance. However the gender gap is not as large as it was in relation to child care,
which suggests that there is more equality in relation to housework than in relation to
child care. I would not want to push this conclusion too far though, since the range of
child care tasks that are comparable across the surveys is very limited.
68
JANEEN BAXTER
Table 3: Results of Significance Tests for Comparisons of Least Squares Means for
Percentage Distribution of Childcare Tasks for Men and Women, 1986 to 1997
Men
Women
Bathing and dressing
children
No change
No change
Getting children to bed
No change
No change
Helping children with
homework
↓ 5%
No change
Total
No change
No change
Unlike with child care, there is evidence of a decline in the gender gap over time, down
from a 16 per cent difference in 1986 to a mere three per cent in 1997. This decline is
consistent across each of the individual tasks. But for different activities, it is the result of
different kinds of change, either men doing more, women doing less or no change on the
part of one partner and a substantial change on the part of the other.
Table 4: Least Squares Means for Percentage Distribution of Housework Tasks for
Men and Women, 1986, 1993, 1997
Men
1986
Women
Men
1993
Women
Men
1997
Women
Preparing meals
23
74
31
73
35
70
Cleaning up after
meals
35
62
42
61
43
63
Grocery shopping
27
72
32
72
31
74
Cleaning the house
18
76
28
75
25
78
Taking out garbage
75
27
66
35
76
31
Washing
14
83
25
80
21
83
69
Table 4: Continued
1986
Men
Women
1993
Men
Women
1997
Men
Women
Ironing
11
81
20
82
21
83
Mowing lawns
87
17
81
24
89
19
Gardening
58
46
59
46
58
48
Home maintenance /
repairs
87
23
80
28
88
24
Total
44
56
47
57
49
57
N
471
325
618
546
545
572
Table 5 reports the statistical significance of these changes. The big area of increase for
men is in preparing meals, increasing by 12 per cent between 1986 and 1997. Women on
the other hand report doing less of this activity, a decline of four per cent over the same
period. The other areas of increase for men are cleaning up after meals, cleaning the
house, washing and ironing. There is no change for men in other areas, but overall these
changes in specific tasks add up to a five per cent increase in the total proportionate
contribution of men.
Women report increased responsibility for some of the traditional male activities, taking
out the garbage, gardening and home maintenance/repairs. Interestingly, the only area
where women report a decrease in responsibility is in preparing meals. Hence, although
men report increasing responsibility in a number of the traditional female activities, this
does not appear to be associated with a corresponding decrease for women. Overall, there
is no change for women in their total contribution to these housework activities.
Table 6 shows the number of hours per week spent on three main areas of housework.
Although some of the surveys collected data on many more tasks, only these three areas
are comparable across all three surveys. While they do not give us a total picture of the
amount of time men and women spend on domestic labour, they are the regular, routine,
everyday activities that all households need to accomplish in order to survive. Again we
see quite a clear gender division, with women spending almost three times as much time
on these activities as men.
70
JANEEN BAXTER
Percentage Distribution of Housework Tasks for Men and Women, 1986 to 1997
Men
Women
Preparing meals
↑ 12%
↓ 4%
Cleaning up after meals
↑ 7%
No change
Grocery shopping
No change
No change
Cleaning the house
↑ 7%
No change
Taking out garbage
No change
No change
Washing
↑ 7%
No change
Ironing
↑ 10%
No change
Mowing lawns
No change
No change
Gardening
No change
No change
Home maintenance /
repairs
No change
No change
Total
↑ 5%
No change
Table 6: Least Squares Means for Hours per Week on Housework Tasks for Men
and Women, 1986, 1993, 1997
Men
1986
Women
Men
1993
Women
Men
1997
Women
Preparing meals and
cleaning up after
meals
7
14
6
16
5
10
Grocery shopping
1
2
1
3
1
2
Cleaning the house
and washing
3
10
3
12
3
8
Total
11
26
10
30
9
20
N
471
325
618
546
545
572
71
Table 7 shows the results of the significance tests for the least squares means for men’s
and women’s hours on these tasks. Overall, women are spending six fewer hours on these
activities in 1997 compared to 1986, and again the biggest decrease is in time spent
preparing meals and cleaning up after meals. There is a very slight increase in women’s
time spent shopping. Bittman’s (1995) work suggests that this is due to an increase in
time spent getting to and from shopping. There is also however, a decline in men’s time
on some of these activities. Even though we see an increase in the proportion of time men
spend preparing meals, there is also evidence that they are spending less time preparing
and cleaning up after meals. This is not a contradictory finding. If we take into
consideration the trend toward greater consumption of pre-cooked, pre-prepared foods
and take away food, it may well be that men are doing a greater proportion of an activity,
but time overall on the activity is declining. Hence the patterns reported here may
indicate broader changes in consumption patterns that are leading to changes in the
distribution of work between men and women as well as a decline in time spent doing
domestic labour. Note however, that these patterns do not appear to extend to men’s time
spent on grocery shopping, cleaning the house or washing.
Hours per Week on Housework Tasks for Men and Women, 1986 to 1997
Men
Women
Preparing meals and
cleaning up after meals
↓ 2 hours
↓ 4 hours
Grocery shopping
No change
No change
Cleaning the house and
washing
No change
↓ 2 hours
Total
No Change
↓ 6 hours
6
Conclusions
The gender division of labour, both in terms of the gender gap in time spent on child care
and housework, and gender differences in the kinds of tasks men and women do in the
home, is still very evident. Women do about two- thirds of childcare tasks, at least threequarters of the routine, everyday, indoor housework tasks, and spend about three times as
many hours as men on the latter. However, if we consider the full range of domestic
tasks, including the outdoor activities that are the traditional tasks that men do, then we
72
JANEEN BAXTER
see a more equal division of labour. There is little evidence here of change over time in
the division of child care tasks, if anything the trends are in the direction of men doing
less and women doing more. This differs from earlier research, but it may be that the very
restricted number of tasks included here is influencing the results. For example, I have
not included questions about playing with children, or taking them on outings, which
might be the areas where men are doing more of the labour.
There is evidence that men are doing a greater share of housework, particularly preparing
meals, but this does not seem to translate into greater time spent on housework. Rather
the data indicate that both men and women are spending less time on housework. Men’s
hours have declined by about two hours per week overall and women’s by about six
hours per week overall. While these results for women are in line with earlier studies,
they are different to the findings of some earlier studies for men. For example, Bittman
(1995) found no change in men’s housework hours, while Bianchi et al. (2000) found a
marked increase in time on housework for men. It should be noted however that Bianchi
et al. found that most of this change had occurred prior to 1985. The current data begin in
1986 so it is possible that men did increase their time on housework prior to this period
and since then it has leveled off and even declined.
Broadly speaking then, it seems that we are witnessing changes in the domestic division
of labour, but as others have noted they are not the ones that we might have imagined.
The gender gap is getting smaller, but mainly because women are doing much less, rather
than men doing much more. And it may be that these changes have more to do with
changes in the other characteristics of women and men than in any deliberate change in
their domestic labour work. Although we have undoubtedly witnessed changes in men’s
and women’s attitudes to gender roles in the home, we have also witnessed other changes
that might affect the gender division of labour in the home. For example, women’s
reduced responsibility for preparing meals may have more to do with women’s increased
involvement in paid work and the fact that they have less time to spend cooking meals. At
the same time, the increased reliance on the service economy to provide the goods
formerly produced in the home may also lead to less time on certain activities,
particularly meal preparation, as well as a re-organisation of men’s and women’s
responsibility for certain tasks. For example, men may be more inclined to do the cooking
if it involves heating up a jar of pasta sauce than if it requires more elaborate or time
consuming preparation. Thus, we may be witnessing quite substantial changes in the
kinds of work to be done in the home and these changes may be helping to lead to
changes in who does the work and how much time is spent doing it.
References
Baxter, J. (1993), Work at Home: The Domestic Division of Labour. University of
Queensland Press, St Lucia.
Bianchi, S. M., M. A. Milkie, L. C. Sayer and J. P. Robinson (2000), ‘Is anyone doing the
housework? Trends in the gender division of household labour’, Social Forces, 79
(1), 191-228.
73
Bittman, M. (1995), ‘Changes at the heart of family households: family responsibilities in
Australia 1974-1992’, Family Matters, 40, 10-15.
Brines, J. (1994), ‘Economic dependency, gender and the division of labor at home’,
American Journal of Sociology, 100 (3), 652-688.
Gershuny, J. and J. P. Robinson (1988), ‘Historical changes in the household division of
labor’, Demography, 25 (4), 537-552.
Hochschild, A. (1989), The Second Shift: Working Parents and the Revolution at Home,
Avon, New York.
Robinson, J. P. and G. Godbey (1997), Time for Life: The Surprising Ways Americans
Use their Time, Pennsylvania State University Press, Pennsylvania.
Shelton, B. A. (1992), Women, Men and Time: Gender Differences in Paid Work,
Housework and Leisure, Greenwood, New York.
Western, M. and J. Baxter (2001), ‘The links between paid and unpaid work: Australia
and Sweden in the 1980s and 1990s’, in J. Baxter and M. Western, eds,
Reconfigurations of Class and Gender, Stanford University Press, Stanford, 81104.
74
Snakes and Ladders: Women’s Pathways
Into and Out of Homelessness
Sue Casey 1
University of Melbourne
1
Introduction
Women-headed households make up over 70 per cent of the world’s homeless (Mulherin,
1996). Homelessness is not, however, unique to societies with third world economies. It
is increasingly a phenomenon experienced by women in societies with advanced capitalist
economies. It is estimated that in Victoria between 16 150 and 17 300 women experience
homelessness each year (Horn, 1995).2
Single homeless women are often described as the hidden homeless, whilst homelessness
itself has been described as advanced marginality (Passaro, 1996) in a risk society (Beck,
1992; Winter and Stone, 1999).
This research provides an analysis of the pathways into and out of homelessness of single
women aged 25-45 years without children in their care. The personal experiences of
eleven women interviewed for this study are considered within a broader systemic
framework.
A reconceptualisation of the categories of homelessness in urban Australia is proposed,
utilising three categories: situational, long-term and chronic homelessness. These
categories are used to examine different experiences of homelessness: pathways into and
out of homelessness and potential points for early intervention.
1
2
The research for this paper is part of a larger work undertaken for a thesis completed for a
Master of Arts (Research), with the School of Social Work, University of Melbourne. This
research would not have been possible without the active participation of women who
participated in the interviews for this study. Their input and assistance is gratefully
acknowledged. I would also like to thank the housing and support agencies who assisted
me in contacting women for interviews and my supervisor, Dr Fiona McDermott.
It is to be noted that the number of homeless single women was not reported in the more
recent ‘Counting the Homeless’ based on the 1996 Australian Bureau of Statistics (ABS)
census collection and Supported Accommodation Assistance Program (SAAP) data
(Chamberlain, 1999). Horn’s (1995) estimate is based on 1993-94 SAAP data and Hanover
Welfare services data of 1992-94. It is an estimate only, based on the available data.
Casey, S. (2002), ‘Snakes and ladders: women’s pathways into and out of homelessness’, in T. Eardley and
B. Bradbury, eds, Competing Visions: Refereed Proceedings of the National Social Policy Conference
2001, SPRC Report 1/02, Social Policy Research Centre, University of New South Wales, Sydney, 75-90.
SUE CASEY
2
Definition of Home and Homelessness
Definitions of homelessness are inextricably linked with the meaning and reality of home,
as distinct from a house. In Australia, the very definition of home is strongly linked with
home ownership and the nuclear family (Paris, 1993).
Kendig and Paris (1987) describe typical housing types over a lifetime as the ‘Housing
Career Ladder’ that commences with living with parents as a child through to outright
home ownership as an adult, generally considered to be within the context of a nuclear
family. The metaphor of a ladder highlights the fact that the climb up the housing career
ladder may not always be successful and may include sliding down ‘snakes’ or moving
back down the ladder (Paris, 1993: 51).
Some women interviewed for this study who had experienced homelessness since
childhood were marginalised to the extent that they have never been on the housing career
ladder, others moved off the ladder for extended periods of time and had difficulty reestablishing a home once it had been lost, whilst others moved off the housing career
ladder for only short periods of time.
The definitions of homelessness used for this research relate to the physical and social
aspects of homelessness. The following two definitions are used.
Homelessness has two interconnected dimensions: a lack of
secure affordable accommodation; and the fracturing of
relationships with families and communities of origin. (Driscoll
and Wood, 1998: 2)
The SAAP3 Act 1994 defines a person as being homeless ‘if and only if, he or she has
inadequate access to safe and secure housing’.
Homelessness is caused by housing which:
(a) damages or is likely to damage the person’s health; or
(b) threatens the person’s safety; or
(c) marginalises the person through failing to provide access to:
(i) adequate personal amenities; or
(ii) the economic or social supports that a home normally
affords;
or
3
The Supported Accommodation Assistance Program (SAAP) is funded jointly by the
Commonwealth, States and Territory Governments. It is the primary funding source for
accommodation and support services for homeless people in Australia.
76
WOMEN’S PATHWAYS INTO AND OUT OF HOMELESSNESS
(d) places the person in circumstances which threaten or
adversely affect the adequacy, safety, security or affordability of
that housing. (SAAP Act, 1994: 4)
Figure 1: The Housing Career Ladder
O u trig h t O w n e r
M o rtg a g e
P a id
S ec o n d T im e b u yer
T w o in c o m es
H ig h w a g e s
S a v in g s
s u b s id y
U n e m p lo y m e n t
D iv o rc e s
P erso n a l
p r o b le m s
F irst T im e b u y e r
P u b lic T e n a n t
M o d e ra te
in c o m e
P riv a te T e n a n t
F irs t jo b
L iv in g w ith p a re n ts
Source: Kendig and Paris, 1987: 30.
Chamberlain (1999) describes three levels of homelessness, with a primary focus on
accommodation type and tenure.
Three categories of homelessness are described.
Primary Homelessness: People without conventional
accommodation, such as people living on streets, sleeping in
parks, squatting in derelict buildings, or using cars or railway
carriages for temporary shelter.
77
SUE CASEY
Secondary Homelessness: People who move frequently from
one form of temporary shelter to another. It covers: people using
emergency accommodation (such as hostels for the homeless or
night shelters); teenagers staying in youth refuges; women and
children escaping domestic violence (staying in women’s
refuges); people residing temporarily with other families
(because they have no accommodation of their own); and those
using boarding houses on an occasional or intermittent basis.
Tertiary homelessness - People who live in boarding houses on
a medium to long-term basis. Residents of private boarding
houses do not have a separate bedroom and living room; they do
not have kitchen and bathroom facilities of their own; their
accommodation is not self-contained; and they do not have
security of tenure provided by a lease. (Chamberlain, 1999)
Chamberlain’s categories have an emphasis on descriptions of the type of accommodation
and as such are more useful for quantifying the number of people who are homeless at
any particular time and the types of accommodation or lack of accommodation in which
they reside. This is particularly useful for counting the homeless.
In contrast, Brown and Zeifert (1990) provide more dynamic descriptors of homelessness,
which are more useful for understanding homelessness longitudinally and to assist in an
assessment of the types of support services women may or may not require. Three groups
of homelessness experiences are described by Brown and Ziefert (1990): chronically
homeless, episodically homeless and situationally homeless.
Chronically homeless women are described as being on a low income or no income.4
They frequently have characteristics that hamper their ability to search for scarce lowincome housing, such as substance abuse issues, psychiatric disability and/or volatility.
Their lives often revolve around daily survival. A survivalist orientation makes it difficult
for them to focus on longer term goals. In order to maintain any level of improvement in
the quality of their lives, women in this group require a continuing level of care.
Episodically homeless women are described as being highly motivated to find permanent
housing, however, they are typically unprepared for independent living. Due to a range of
factors such as substance abuse, mental illness and volatility it is difficult for them to find
and maintain independent housing in a market of scarce resources. They often have
histories of unresolved crises that have led to their minimal functioning. Frequently they
present a picture of resignation and hopelessness. The service commitment needs to be
4
It is of note that ‘chronic homelessness’ is often used as a generic term to describe longterm homelessness (Horn, 1995; Chamberlain and Johnson, 2000). Chamberlain and
Johnson describe an operational measure of adult chronic homelessness as six months or
more.
78
lengthy, with periods of intensive activity as new crises appear. These women have
identifiable strengths, including a personal support network.
Situationally homeless women are described as lacking shelter because of an acute crisis
and are usually homeless for the first time. These women respond well to short-term crisis
intervention, including emotional support, help setting priorities, advocacy for community
support, and concrete help with basic needs. Once they find affordable housing, they
move quickly to re-establish independent living.
Brown and Ziefert (1990) argue that differentiating between these groups of homeless
women enables one to anticipate and provide the services that particular women need and
to better structure these services. Their categories have been developed based on the
experiences of homelessness in the United States of America. They are used in this
research as the basis for developing categories of homelessness more relevant to urban
Australia.
Chamberlain and Johnson (2000) propose a more universal homelessness career specific
to adults. In this homelessness career, households can spend extended periods of time at
risk of homelessness. There is a critical point where accommodation is lost and
homelessness begins. This compares to the homelessness career of youth, where there can
be an extended period of being at risk where the young person moves in and out of the
family home.
3
Causes of Homelessness
This research is premised on the understanding that the primary causes of single women’s
homelessness relate to economic disadvantage caused by low income and inadequate
affordable housing (Horn, 1995; DHS, 2000; Watson and Austerberry, 1986). However,
some women are more vulnerable to poverty and homelessness than others.
In the later part of the 20íÜ century and beginning of the 21ëí century, major changes to
economic policy, the impact of globalisation and major social changes have led to what
can be described as ‘manufactured uncertainty’ in a ‘risk society’ (Giddens, 1996; Beck,
1992; Winter and Stone, 1999).
‘Manufactured uncertainty’ is described as the level of uncertainty in people’s lives in
contemporary societies which is not inherent to the human condition, but has direct links
with external factors such as deregulation of the labour market and the economy
(Giddens, 1996; Winter and Stone, 1999).
Within this ‘risk society’, individuals who are well paid, well educated and have good
material and emotional supports from family or friends are buffered against the extreme
outcomes of risk taking. Indeed risks may present as a range of exciting opportunities. For
those without buffers, such uncertainty of employment, housing or family and social
relationships can very easily lead to insecurity and a loss of opportunity and ultimately
homelessness (Beck, 1992; Winter and Stone, 1999).
79
SUE CASEY
Within this context, poverty and homelessness may be described as the endpoint of a
process that comes about following a temporal chain of events, conditions and other
stressors that push individuals into areas of risk or aggravate their social marginalisation
by making it a permanent condition (Mingione and Morlicchio, 1993; Milburn and
D’Ercole, 1991).
The systemic causes of women’s homelessness within the ‘risk society’ may be further
understood using Young’s (1990) framework of injustice and oppression as a heuristic
device. She describes powerlessness, violence, cultural imperialism, marginalisation and
exploitation as faces of oppression. These have resonances with the experiences of
homeless women interviewed for this study. However, a detailed discussion of these
themes are beyond the scope of this paper.
The outcome of increasing numbers of people on low and uncertain income has led to
delays in home purchase and increasing numbers of people seeking out low-cost private
rental. The extent of this demand has meant that it is unable to be met. In the face of a
relatively unregulated private rental market, the high cost of entry into home ownership
and the diminishing availability of public housing, there are increasing numbers of people
who are experiencing housing stress, housing related poverty and ultimately homelessness
(Burke, 1999; Moriarty, 1998; Winter and Bryson, 1998).
4
Methodology
The primary research involved in-depth interviews with eleven women who had
experienced homelessness. The women were between the ages of 25-45 years at the time
of interview, considered themselves to be single whilst homeless and without dependent
children in their care. Women interviewed were not in crisis and were in a position to
reflect back over their experiences of homelessness. At the time of interview women were
living in a variety of settings including transitional housing, rooming houses, a residential
service, with relatives and in private rental.
Women were asked questions not only about their experiences of homelessness and their
pathways into and out of homelessness, but also their views about service provision and
what should be done to address homelessness.
5
Categories of Homelessness
Three distinctive groups of experiences of homelessness and the pathways into and out of
homelessness, were identified using a qualitative thematic analysis of the interviews.
These groups are described as situational, long-term and chronic homelessness. These are
a variation of Brown and Ziefert’s (1990) groupings based on urban Australian, rather
than the United States experience.
These categories group experiences of homelessness longitudinally, in contrast to
Chamberlain’s (1999) categories that have a greater emphasis on accommodation type at
80
a particular point in time. Many women in this study moved between primary, secondary
and tertiary homelessness over time.
These groupings are useful in understanding in more detail:
•
•
•
women’s pathways into homelessness;
women’s experiences of homelessness; and
women’s pathways out of homelessness.
These categories of homelessness - chronic, long-term and situational homelessness –
will now be described as they relate to the pathways into and out of homelessness and
experiences of homelessness of women interviewed for this study.
6
Pathways into Homelessness and Experiences of Homelessness
Chronic Homelessness
The term chronic homelessness is well and truly part of the lexicon of the homelessness
field in Australia. It is often used to simply describe long-term homelessness (see for
example Horn, 1995; Chamberlain and Johnson, 2000). However, based on this study, it
appears to be too broad to adequately inform service development and case work practice.
Based on the findings of this research it is proposed that the term chronic homelessness
be separated into two categories, described as chronic homelessness and long-term
homelessness.
The term chronic homelessness is used for those women who have been homeless since
they were children and had little or no significant experiences of home as adults. Their
pathways into homelessness occurred as children and teenagers, rather than as adults and
are consistent with the youth homelessness career, rather than adult homelessness careers,
described by Chamberlain and Johnson (2000).
Women in this group were Jacinta, Rosie and Annie. None of the women in this group
had had paid employment. Only Jacinta, who was in her twenties, talked about gaining
the skills necessary for employment in the future. Both Rosie and Annie, who were in
their thirties and forties respectively, still needed to establish somewhere to live after a
lifetime of homelessness.
Annie and Jacinta spoke about experiencing abuse and neglect as young children that
resulted in their going into the care of the state. Rosie spoke of the death of her father as
the precipitating factor that led to her and her brothers and sisters going into care. None of
these three women could describe a place that they thought of as home except for very
short-term experiences. They had all had a lifetime of moving from institutions to
temporary accommodation and moving in and out of contact with support services.
Not being able to stay put in one place for any length of time was another key theme for
this group. All three women had really struggled to achieve any stability of housing over
many years that had started with the inability of their families to be able to support them
81
SUE CASEY
as children. They all talked about needing to learn the practical skills necessary to
maintain a home. This was a distinguishing factor compared to the other two groups. It
went as far as the women identifying difficulties in having a home in comparison to being
homeless. This was particularly the case for Rosie and Annie, who were older than
Jacinta and had consequently experienced longer periods of homelessness and appeared to
have had fewer supports as teenagers.
Brown and Ziefert (1990) included having a survivalist orientation that makes it difficult
for women to establish long-term goals as typical of chronically homeless women. Both
Jacinta and Annie spoke about long-term goals: Annie getting public housing and Jacinta,
who was younger, going back to school. Schooling was something Jacinta had been able
to achieve in the past. However, Annie had a number of steps to go before even applying
for public housing. Rosie had no plans, although she talked vaguely about having options
this time around.
Consistent with Brown and Ziefert’s (1990) chronic homelessness group, Rosie and
Annie had significant characteristics that made it difficult for them to search for scarce
long-term housing. In Rosie’s case it was her volatility and for Annie it was substance
abuse. This may also have been the case for Jacinta. She had had a substance abuse
problem in the past and described herself as having a ‘short fuse’. At the time of
interview, she had been off drugs for three months that corresponded to the period when
she had stopped living on the streets.
Well it’s hard, I’m just learning now to settle down, and to stay
in one place for a length of time, unfortunately this place doesn’t
really suit me, so I’m trying to stick it out here until the absolute
most, until I’m almost ready to commit suicide or something.
(Jacinta)
These factors were further compounded by chronic physical health problems, mental
illness and/or disability and being victim to sexual assault and/or violence as adults.
The experiences of these three women, Jacinta in her twenties, Rosie in her thirties and
Annie in her forties, provides an insight into the failure of the provision of social services
for homeless children across the decades which led to their being homeless as adults. This
is consistent with Burdekin’s findings (HREOC, 1989), that care of the state is a key
predisposing factor for homelessness as an adult.
In contrast to Brown and Ziefert’s (1990) chronically homeless group, no women in this
study experienced primary homelessness indefinitely. This may well reflect the
differences in the extent of homelessness in Australia compared to the United States. In
Australia, there is a more comprehensive range of services and, in relative terms, greater
access to low-cost accommodation (McDonald, 1994). This minimises the number of
people who experience primary homelessness on a permanent basis.
82
Long-term Homelessness
Long-term homelessness is identified as a separate category, rather than episodic
homelessness (referring to Brown and Ziefert, 1990) or as a sub-category of chronic
homelessness. This group includes Karen, Shaun, Helen, Mary, Brooke and Belinda. All
the women in this group, except for Mary, had at some stage, if not for an extended
period of time, lived independently in the private rental market. Mary had been able to
maintain a room in a rooming house for most of her adult life. With the exception of
Mary, the women in this group had all participated in the paid workforce.
The women in this group became homeless following a series of critical events. For
example, Karen was in full-time employment, married and renting a house, not far away
from purchasing a home. Her first child died soon after birth and she reported that
unresolved grief led to undiagnosed post-natal depression following the birth of her
second child. What was occasional amphetamine use became a habit once she stopped
breast-feeding. Her marriage broke down and her husband left with their child. She felt so
unworthy due to her depressive illness that she was unable to accept assistance from her
husband’s family. She eventually had to move out of her home, she was too ashamed to
accept social security payments and she survived by shoplifting.
Consistent with the adult homelessness career proposed by Chamberlain and Johnson
(2000) Karen and the other women in this group were at risk of homelessness for an
extended period of time before becoming homeless. Women in this group had
experienced primary homelessness for extended periods of time or had lengthy
psychiatric hospital admissions or series of admissions, whilst their housing was tenuous.
Consequently, the experience of homelessness itself contributed to the perpetuation of
their homelessness.
In contrast to Brown and Ziefert’s episodically homeless group, the long-term
homelessness group appeared to go on a downward spiral, which resulted in primary
homelessness or longer term psychiatric in-patient admissions. Once they had moved too
far off the housing career ladder they had great difficulty getting back on.
Going back two or three years ago, what would have helped at
the time? (Sue)
Well if I had money of course ... nothing really … it just all led
in that direction; veering off the path; it was just basically a path
I got on; that I knew that I would become homeless. (Brooke)
The women in both the long term and chronically homeless groups, when they ‘slept
rough’ often slept alone and for reasons of safety in places where they would not be
found. This included the verandah of a suburban school, in cars, in people’s sheds and in
empty houses in the suburbs. This contributes to the hidden nature of single women’s
homelessness.
83
SUE CASEY
Situational Homelessness
Consistent with Brown and Ziefert’s (1990) categories, this term is used to describe the
group in this study who became homeless in response to a particular crisis. This was the
case for Lindy and Aba. In Lindy’s case a mental illness led to her losing her job and
accommodation (that was associated with the job). Aba was a relatively new arrival to
Australia and when she had to move from living with her extended family, could not
afford private rental. She had limited work opportunities due to poor English skills and
limited access to English language classes.
Neither woman considered herself to be homeless. They sought support to find short-term
or low-cost housing when they were unable to access appropriate low-cost housing
through the private rental market. They were both able to access accommodation support
services and referral to medium to long-term low-cost accommodation before needing to
use emergency accommodation services. Both women required other support services.
However, their other support needs did not affect their ability to maintain a home.
7
Pathways Out of Homelessness
The women’s pathways out of homelessness, the types of supports they required and the
time it took to get back to having a home or establishing a home were related to their
pathways into homelessness and experiences of homelessness itself.
Chronic Homelessness
The group who had experienced chronic homelessness had little or no experience of
home. Resiliency was highlighted by the women, who had experienced chronic
homelessness, as being the key to simply surviving all that they had experienced.
The women in the chronically homeless group all reported using multiple services over
time. This included housing support, health, welfare, mental health, drug treatment
services, counselling and educational services. They also required additional support to
the other two groups in needing to learn practical skills associated with maintaining a flat
or other accommodation.
The women in this group had accessed many services over time and were able to provide
valuable insights into the sorts of models that they found most helpful. The most
successful model reported by women who had experienced chronic homelessness were
services that were respectful and allowed women to contact and engage and re-engage
over time.
Only Rosie had successfully engaged with a housing support service over a period of
time. This was a service that provided a residential service moving onto transitional
housing. She had moved from the residential service to transitional housing, only leaving
to move to the country with her new partner. When that didn’t work out she was
confident in re-contacting and moving back to the residential service. The only place she
identified as home was when she lived in transitional housing for a period of nine months.
84
Annie had applied for transitional housing at the time of interview. Jacinta was living in a
rooming house but was not happy and was hoping to move back to another rooming
house where she had lived in the past. Neither Jacinta nor Annie were able to talk about
anywhere as having been home.
Long-term Homelessness
Women in this group had invariably experienced primary homelessness or extended
psychiatric in-patient admissions whilst tenuously housed. Their pathways out of
homelessness were not simple. A number of women were homeless for long periods of
time before accessing services. For example, Karen spent three years sleeping rough and
in other circumstances consistent with primary homelessness. During this period she only
approached a crisis service once; as the only immediate option was a private hotel she
returned to sleeping out until she eventually went to gaol where she was referred to a
housing support service. A number of the other women had very similar experiences. It
was only women who had long periods of psychiatric hospitalisation who avoided
primary homelessness.
Women in this group, were of particular concern in relation to access to the service
system. They had all experienced long periods of homelessness, yet the majority talked
about not knowing where to go or how to access services or feeling ashamed of needing
to access services or that they should be able to cope, given their past lifestyles where
they were able to live independently of welfare type services.
Women in the long-term homelessness group highlighted opportunities and lost
opportunities for referral to appropriate services, more timely intervention and potentially
shorter period of homelessness. The women’s responses suggest that generalist services
need to be more heavily targeted as referral sources for homeless women. This included
Centrelink, general practitioners, psychiatric in-patient services, gaols, police, needle
exchange services and crisis services. In many ways these were services that women often
could not avoid contacting due to their circumstances. When they did access services,
they reported often not being eligible due to their age or sex. Alternatively there were
long waits for accommodation or the accommodation was unsafe, for example referral to
private hotels.
The women in the long-term homeless group all required a range of support services in
order to re-establish themselves. All the women utilised more than one service in order to
move on from being homeless. These services included housing support, health, welfare,
mental health, drug treatment services, counselling and education and training services.
Once women in the long-term homelessness group accessed a housing support program
including accommodation they appeared to stay with that service through to finding a
home. However, this remains a hypothesis based on a limited sample that requires further
investigation. Nevertheless, this did contrast to women in the chronically homeless group
who had generally accessed more than one housing support service over time.
85
SUE CASEY
A number of women in the long-term housing group commented that the primary cause of
their extended period of homelessness was lack of affordable housing. Given that all
women in this group had the experience of maintaining their own home, most could
probably have managed their own place far earlier, if affordable accommodation had been
available more easily.
This was the case for Belinda for example, who managed her own private rental for a
period of time whilst still going in and out of a psychiatric hospital for treatment. She
moved back with her parents due to an inability to pay the rent.
Situational Homelessness
Women in the situationally homeless group were able to access housing and support
services before their situations deteriorated to ‘sleeping rough’ or extended periods of
institutionalisation. Lindy was living in a women’s rooming house and Aba was in
transitional housing. In the longer term, Lindy considered the private rental market an
option as she expected to be able to work part time teaching music. Aba was applying for
public housing as she simply could not afford private rental on social security payments.
8
Pathways Out of Homelessness: Common Themes
Some significant differences in experiences of homelessness have been described using
three categories of homelessness. However, the women interviewed for this study had
many similar reflections and comments to make about the causes of homelessness and
what needs to be done to address homelessness and related issues.
Comment has already been made on the need for women to have access to safe,
affordable housing and adequate income as a key to addressing homelessness. In addition,
women utilised a range of health, welfare, mental health, specialist drug treatment
services, education and training, counselling and other support services to assist them in
addressing the myriad of issues that they faced as a result of their experiences of
homelessness and pathways into and out of homelessness.
The most successful housing and support model, based on feedback from the interviewed
women, was an integrated women-only service that included residential, transitional and
rooming house accommodation and support services. In addition, women also accessed
drug and alcohol, mental health and counselling services. There was a strong emphasis on
referral to employment and training programs, support for education and/or other lifestyle
activities.
In terms of transitional housing women expressed a preference for ‘women-only’ services
with a one bedroom flat being the accommodation of choice for the majority of women,
although two women who lived in women’s rooming houses found their accommodation
suitable for their needs. However, other women who lived in rooming houses reported
finding it difficult, particularly because of the number of residents who had significant
86
behavioural problems that impacted on other residents. One woman commented that she
thought some women were being accommodated in a rooming house, when they required
accommodation with a higher level of support. This also appeared to reflect a tension
between the needs of short-term and long-term residents in rooming house
accommodation.
The issue of not being able to access crisis and housing support services in a timely
manner was highlighted by a number of women. This included long waiting periods and
not being eligible for a number of crisis and housing support services due to age and/or
sex.
In Victoria, at the time of interview, the turn-away rate for SAAP services was higher for
women than men.5 There were only three relatively small SAAP-funded crisis
accommodation services specifically for single women.6 An additional, small service has
recently been funded. There are also three SAAP-funded cross-target crisis
accommodation services (AIHW, 2000, Table 2.6). However, a number of women in this
study were concerned about violence in cross-target services and/or had experienced or
witnessed violence in cross-target services so did not use or attempt to use these services.
The extent of the demand for women’s crisis and housing support services may well be
under-represented as unmet demand is only recorded for potential clients who meet
eligibility criteria. Therefore if a service for a particular target group does not exist in an
area or as suggested by the findings of this study, potential clients do not know of the
service then the extent of demand would be under-represented.
A related issue is that women’s services often are not advertised with a sign on the front
gate due to security concerns. This serves as a barrier to access that provides a challenge
in terms of promoting women’s services in other ways.
Inadequate access to services was an issue for women not only when they were actually
homeless, but the period when they were at risk of homelessness and experiencing crises.
The findings of this research suggest that prevention of homelessness and early
intervention includes the provision of a range of social supports and health and welfare
services before women become homeless, in order to support them in staying in
employment and in maintaining their housing. In this study, that included access to
5
6
The SAAP data (AIHW, 2000) shows that of 10 450 people who were accommodated by SAAP
services in 1998-99, 5900 were women and 4550 were men (AIHW, 2000, Table 3.3). However,
women were more likely than men to be turned away from a service. In a two week census period,
there were 680 people who sought accommodation from a SAAP service who were unable to be
assisted, of these 384 were women and 296 were men (AIHW, 2000, Table 5.39). Of these ‘potential
clients’, women and men had similar turnaway numbers, 181 and 174 respectively, for crisis
accommodation. There were significantly more women (203), than men (122) seeking medium/longterm accommodation who were unable to be assisted (AIHW, 2000, Table 5.39).
This excludes women’s refuges that are specifically targeted for women and women with children
escaping domestic violence. There are 15 women only services, of which three are crisis services and
12 men only services of which three are crisis services. However, the average annual cost per agency
for ‘women only’ services is $146 384 compared to $349 080 for ‘men only’ services.
87
SUE CASEY
counselling, mental health services, specialist drug services, education, employment and
training services. This is in addition to the housing related supports identified by
Chamberlain and Johnson (2000) that included substantial emergency relief to assist
households to maintain their housing when facing housing related poverty.
The other areas commented on by women as keys to their pathways out of homelessness
were relationships and personal strengths. Women’s own personal strengths were often
highlighted by women as being the key to their moving on from being homeless.
Relationships and support particularly from extended family were important. The other
key relationship commented on by a number of women was with a particular support
worker. This often appeared to be based on the women’s perceptions of genuine concern
by that worker.
In summary women’s pathways out of homelessness appeared to rely on the interplay of a
range of factors that included:
•
inner strength to begin to seek out support;
•
access to affordable accommodation/housing in the short and long term;
•
access to housing support services which included individual support from workers
and referral to other services;
•
adequate income in relation to housing costs;
•
the development of political awareness, through experiences such as working with the
‘Big Issue’, participation in SAAP advocacy service courses and participating in
feminist counselling;
•
personal counselling, particularly regarding past trauma, including sexual assault
counselling and services such as domestic violence outreach;
•
high quality mental health services and drug treatment services;
•
support from family including assistance with temporary accommodation and
personal support;
•
education (including English language classes) and employment and training
opportunities; and
•
support in pursuing leisure activities, such as athletics and art.
9
Conclusion
In conclusion, this research highlights the fact that homelessness occurs when there is
insufficient low-cost housing to meet the needs of people on low incomes. However,
within this context the systemic causes of homelessness are not limited to housing
injustice and income inequality, but include a range of factors which contribute to the
oppression of particular groups leaving them more vulnerable to poverty and
88
homelessness without the necessary buffers required for survival within a ‘risk society’
(Beck, 1992).
The pathways into and out of homelessness, are not linear as suggested by the typical
housing career but are better represented by a game of snakes and ladders (Kendig and
Paris, 1987; Paris, 1993). In this case, the snakes are the multiple traumatic events and
overlapping oppressions and consequent disadvantage faced by women who experience
homelessness. The ladders on the pathways out of homelessness are access to a range of
services, adequate income and affordable housing. Women also highlighted the
importance of their own personal strengths and resiliency and sometimes the role of
support from extended family.
There is a need to understand in greater detail the pathways into and out of homelessness
in order to provide more timely assistance based on the different needs and requirements
of different women and so prevent and/or minimise the effects of homelessness.
This is a relatively small study, but indicates the need for further research in this area. A
greater understanding of the issues raised by this research could be gained through a
longitudinal study, involving a greater number of women, who are contacted over a
period of one to two years.
This research provides only part of the picture of women’s pathways into and out of
homelessness, but I hope makes some contribution towards the amelioration of women’s
homelessness.
References
Australian Institute of Health and Welfare (AIHW) (2000), Supported Accommodation
Assistance Program (SAAP) National Data Collection Annual Report 1998/99,
Victoria, Catalogue No. HOU 41, AGPS, Canberra.
Beck, U. (1992), Risk Society: Towards a New Modernity, Sage Publications, London.
Brown, K.S. and M Ziefert (1990), ‘A feminist approach to working with homeless
women’, Affilia, 5(1), Spring, 6-20.
Burke, T. (1999), The State We Are In, Paper prepared for VCOSS conference, 26 March
1999, Melbourne, Institute of Social Research Swinburne University, Melbourne.
Chamberlain, C. (1999), Counting the Homeless: Implications for Policy Development,
Australian Bureau of Statistics, Occasional Paper, 2041.0, ABS, Canberra.
Chamberlain, C. and G. Johnson (2000), Early Intervention: A research paper prepared
for the Victorian Homelessness Strategy, Department of Human Services, Victoria.
Department of Human Services (DHS) (2000), Victorian Homelessness Strategy Focus
Group on Single Women and Homelessness: Background Paper, Victorian
Government.
Driscoll, K. and L. Wood (1998), A Public Life: Disadvantage and Homelessness in the
Capital City, Department of Social Science and Social Work, RMIT for City of
Melbourne.
89
SUE CASEY
Giddens, A. (1996), ‘Affluence, poverty and the idea of a post-scarcity society,’ in C. De
Alcamatara, ed., Social Futures, Global Visions, Blackwell Publishers, UK, 151-63.
Horn, M. (1995), Women Alone… Stepping Forward, Hanover Welfare Services,
Melbourne.
Human Rights and Equal Opportunity Commission (HREOC) (1989), Our Homeless
Children, AGPS, Canberra.
Kendig, H. and C. Paris, with N. Anderton (1987), Towards Fair Shares in Australian
Housing, IYSH National Committee of Non-government Organisations, Canberra.
McDonald, P. (1994), Expanding Horizons: Responses to Chronic Homelessness,
Salvation Army Crossroads Housing and Support Network.
Milburn, N. and A. D’Ercole (1991), ‘Homeless women: towards a comprehensive
model’, American Psychologist, 46(11), 1161-9.
Mingione, E. and E. Morlicchio (1993), ‘New forms of urban poverty in Italy: risk path
models in the north and south’, International Journal of Urban and Regional
Research, 17, 413-27.
Moriarty, P. (1998), ‘Inequality in Australian cities’, Urban Policy and Research, 16(3),
211-8.
Mulherin, T. (1996), Ten Years on … Still no Place for Women, Council to Homeless
Persons Conference, Melbourne.
Paris, C. (1993), Housing Australia, Macmillan Education Australia P/L, Melbourne.
Passaro, J. (1996), The Unequal Homeless: Men on the Streets, Women in their Place,
Routlege, New York.
Watson, S. with H. Austerberry (1986), Housing and Homelessness: A Feminist
Perspective, Routledge and Kegan Paul, London.
Winter, I. and L. Bryson (1998), ‘Economic restructuring and state intervention in
Holdenist suburbia: understanding urban poverty in Australia’, International
Journal of Urban and Regional Research, 22(1), 60-75.
Winter, I. and W. Stone (1998), Social Polarisation and Housing Careers: Exploring the
Interrelationship of Labour and Housing Markets in Australia, Working Paper
No.13, Australian Institute of Family Studies, Melbourne.
Winter, I. and W. Stone (1999), Reconceptualising Australian Housing Careers, Working
Paper No. 17, Australian Institute of Family Studies, Melbourne.
Young, I.M. (1990), Justice and the Politics of Difference, Princeton University Press,
New Jersey.
90
Factors Influencing the Educational
Performance of Students from
Disadvantaged Backgrounds1
Gillian Considine
Australian Centre for Industrial Relations Research and Training
University of Sydney
Gianni Zappalà
Research and Social Policy Team
The Smith Family
1
Introduction
The relationship between family socio-economic status (SES) and the academic
performance of children is well established in sociological research.2 While there is
disagreement over how best to measure SES, most studies indicate that children from
low SES families do not perform as well as they potentially could at school compared
to children from high SES families (Graetz, 1995). Most studies, however, compare
students from across all SES backgrounds to reach the conclusion that low SES
adversely affects a range of educational outcomes. Another important dimension,
however, is the factors that may influence educational outcomes within particular SES
bands. This paper presents data on the educational performance of children from
financially disadvantaged backgrounds and examines its variation as affected by
traditional measures of SES as well as by a range of other family, individual and
contextual factors.
1
2
We thank Ian Watson for comments and assistance. Ben Parker, Vanessa Green,
Rosemary Dean, Rob Simons, and Anne Clark also provided comments and
assistance. The usual disclaimers apply.
We acknowledge that the academic performance of children is also related to their
innate ability. The degree of individual variance in academic performance accounted
for by variation in genetic factors, however, is the subject of intense debate (Sparkes,
1999). More recent research suggests that the traditional ‘nature versus nurture’
debate is somewhat misleading and that both a child’s heritable characteristics and
their environment are related and co-exist in complex and significant ways. For
instance, ‘hereditary vulnerabilities establish probabilistic, not deterministic,
developmental pathways that evolve in concert with the experiential stressors, or
buffers, in the family, the neighborhood, and the school’ (Shonkoff and Phillips,
2000: 55). Thanks to Vanessa Green for bringing this reference to our attention.
Considine, G. and G. Zappala (2002), ‘Factors influencing the educational performance of students
from disadvantaged backgrounds’, in T. Eardley and B. Bradbury, eds, Competing Visions: Refereed
Proceedings of the National Social Policy Conference 2001, SPRC Report 1/02, Social Policy
Research Centre, University of New South Wales, Sydney, 91-107.
GILLIAN CONSIDINE AND GIANNI ZAPPALA
2
Factors that Influence Educational Performance
Socioeconomic status
Socioeconomic status can be defined as ‘a person’s overall social position…to which
attainments in both the social and economic domain contribute’ (Ainley et al., 1995:
ix). When used in studies of children’s school achievement, it refers to the SES of the
parents or family. Socio-economic status is determined by an individual’s
achievements in: education; employment and occupational status; and income and
wealth. Several comprehensive reviews of the relationship between SES and
educational outcomes exist (Amato, 1987; Williams et al., 1991; Mukherjee, 1995;
Ainley et al., 1995). These studies and reviews make it clear that children from low
SES families are more likely to exhibit the following patterns in terms of educational
outcomes compared to children from high SES families:
•
have lower levels of literacy, numeracy and comprehension;
•
have lower retention rates (children from low SES families are more likely to
leave school early);
•
have lower higher education participation rates (children from low SES families
are less likely to attend university);
•
exhibit higher levels of problematic school behaviour (for instance truancy);
•
are less likely to study specialised maths and science subjects;
•
are more likely to have difficulties with their studies and display negative attitudes
to school; and
•
have less successful school-to-labour market transitions.
These results remain the same irrespective of how SES is measured and whether the
studies are based on individual or aggregate level data (Graetz, 1995: 32-35).
Similarly, studies of children’s educational achievements over time have also
demonstrated that ‘social background remains one of the major sources of educational
inequality’ (Graetz, 1995: 28). In other words, ‘educational success depends very
strongly on the socio-economic status of one’s parents’ (Edgar, 1976, cited in Graetz
1995: 25).
The effect of parental SES on children’s educational outcomes may be neutralised,
strengthened or mediated by a range of other contextual, family and individual
characteristics. Parents may have a low income and a low-status occupation, for
example, but nevertheless transmit high educational aspirations to their children.
What family members have (material resources, for instance) can often be mediated
by what family members do (for example parental support, family cohesion). The
social and the economic components of socio-economic status, in other words, may
have distinct and separate influences on educational outcomes. While both
components are important, social factors (for instance, parents’ educational
attainments) have been found to be more significant than economic factors, such as a
family’s capacity to purchase goods and services, in explaining different educational
outcomes. It is argued that families where the parents are advantaged socially,
92
FACTORS INFLUENCING THE EDUCATIONAL PERFORMANCE OF STUDENTS
educationally and economically, foster a higher level of achievement in their children.
They also may provide higher levels of psychological support for their children
through environments that encourage the development of skills necessary for success
at school (Williams et al., 1980; Williams, 1987; Williams et al., 1993).
Family Structure
Socio-economic status may therefore also be linked to family structure. As sole parent
families on average have lower levels of income, are headed by parents with lower
educational attainment and are less likely to be in the labour force, children from these
families are likely to have lower educational performance (Rich, 2000). Other factors
in sole parent families that are likely to adversely affect educational outcomes of
children compared to those from two-parent families are said to include:
•
reduced contact between the child and non-custodial parent;
•
the custodial parent having less time to spend with children in terms of
supervision of school-work and maintaining appropriate levels of discipline;
•
the lack of an appropriate role model, especially for males;
•
increased responsibilities on children such as childcare roles, domestic duties
which impede the time available for school work; and
•
the nature of parent-child relationships in sole parent families may cause
emotional and behavioural problems for the child (Buckingham, 1999; Rich,
2000).
The influence of family structure has been found to be only weakly associated with
educational attainment, however, once controlling for other variables (Machin, 1998).
It is more detrimental when children in sole parent families also experience a range of
other risk factors such as low income (Sparkes, 1999).
Type of School
As well as socio-economic status, research has shown the importance of the type of
school a child attends in influencing educational outcomes. While research in the US
has found that SES variables continue to influence educational attainment even after
controlling for different school types, the school context tends to affect the strength of
the relationship between SES and educational outcomes (Portes and MacLeod, 1996).
Similarly, research in Britain shows that schools have an independent effect on
student attainment (Sparkes, 1999). While there is less data available on this issue in
Australia, several studies using the Longitudinal Surveys of Australian Youth have
found that students attending private non-Catholic schools were significantly more
likely to stay on at school than those attending state schools (Long et al., 1999; Marks
et al., 2000). Students from independent private schools are also more likely to
achieve higher end of school scores (Buckingham, 2000a). While school-related
factors are important, there is again an indirect link to SES, as private schools are
more likely to have a greater number of students from high SES families, select
students with stronger academic abilities and have greater financial resources. The
school effect is also likely to operate through variation in the quality and attitudes of
teachers (Sparkes, 1999). Teachers at disadvantaged schools, for instance, often hold
93
low expectations of their students, which compound the low expectations students and
their parents may also hold (Ruge, 1998).
Absences
Also related to poor educational performance is the level of truancy or unexplained
absence among students. Truancy can be modelled both as an educational outcome
and as a causal factor in explaining educational performance. Truancy tends to be
higher among students from low SES backgrounds. Truancy, even occasional, is
associated with poorer academic performance at school (Sparkes, 1999). Having high
levels of unexplained absence at school has also been found to be associated with
poorer early adult outcomes in the labour market for instance higher probability of
being unemployed and poorer adult health relative to non-truants (Sparkes, 1999).
Gender
Educational performance at school has also been found to vary according to the
student’s sex (Horne, 2000). In particular, reviews of the evidence suggest that boys
suffer an educational disadvantage relative to girls, especially in terms of performance
in literacy (Buckingham, 1999; 2000b). There are several explanations for this
increasing gender gap which include: biological differences; gender biases (such as
reading being seen as ‘not masculine’); teaching, curricula and assessment (for
instance less structured approaches to teaching grammar may have weakened boys’
literacy performance); and socioeconomic factors (Buckingham, 1999: 5). The last
explanation is of particular interest in the context of this paper, especially the finding
that the gender gap continues within each socio-economic level (Teese et al., 1995).
That is, girls have been found to out-perform boys within high or low socio-economic
groups. Furthermore, the performance of boys deteriorates more rapidly than the
performance of girls as they move down the socio-economic scale (Teese et al.,
1995). As was noted above, the relationship between the performance of boys and
socio-economic status is often mediated or partially explained by family structure
(Buckingham, 1999: 9-10).
Ethnicity
The ethnic background or immigrant status of parents is also an important mediating
variable on the influence of SES on children’s educational performance. Studies of the
academic performance of second-generation school students in the US have found that
while their performance is also influenced by the SES of their parents and type of
school, ‘their national background plays a significant independent role’ (Portes and
MacLeod, 1996: 270). The authors found that some first-generation immigrant parents
(e.g., Cuban, Vietnamese) through the process of migration and subsequent
incorporation in the host society, come to see education as a key means of upward
mobility for their children, despite their own low levels of education and income
(Portes and MacLeod, 1996). Children from these communities did well despite
coming from low SES backgrounds whereas the negative effects of SES were not
ameliorated in the academic performance of children from immigrant communities
with low levels of social capital (e.g., Haitian, Mexican).
Similar findings have emerged within the Australian experience. While the children of
immigrants were seen to be at a disadvantage up until the mid-1970s (Martin, 1978),
the gradual introduction of multicultural policies in the classroom from that time may
94
have improved the situation (Cahill, 1996). A series of studies based on Census data
suggests that the second generation (especially those of European, Indian and Chinese
origin) have achieved substantial educational mobility in terms of staying on at
school, compared to those from British, German, Dutch and Australian origin (Birrell
and Khoo, 1995; Khoo, 1995). As a consequence, higher percentages of children from
non-English speaking background (NESB) achieve tertiary qualifications compared to
those from English-speaking background (ESB) (Birrell and Khoo, 1995; Dobson et
al., 1996). As with the US research, however, there is a great deal of variation
between different ethnic groups. Studies have found that it is more likely that people
from Vietnamese, Chinese, Eastern European and Korean backgrounds are in higher
education than people from ESB. Whereas those whose language group was Arabic,
Khmer and Turkish were half as likely to be in higher education than those from ESB
(Dobson et al., 1996; Cahill, 1996; see also Marks et al., 2000).
Geographical Location
Students from non-metropolitan areas are more likely to have lower educational
outcomes in terms of academic performance and retention rates than students from
metropolitan areas (Cheers, 1990; HREOC, 2000). Despite an adequate number of
educational facilities in rural and remote Australia, school children from these areas
remain disadvantaged by other factors. Issues affecting access to education in regional
areas include costs, the availability of transport and levels of family income support.
In addition, inequity exists with regard to the quality of the education that rural
students receive, often as a result of restricted and limited subject choice.
Furthermore, students may also have limited recreational and educational facilities
within their school (HREOC, 2000: 12).
Housing Type
Lower educational attainment has also been found to be associated with children
living in public housing compared to those in private housing (Sparkes, 1999). This
may be due to the effects of overcrowding, poor access to resources and a lack of
social networks, and in this sense, housing type may also be a measure of
neighbourhood influence. A recent Australian study based on 171 Year 12 students
from 10 state schools, found that neighbourhood effects ‘were an important influence
on students’ educational plans…’ to continue further post-secondary education, after
controlling for a range of individual and family socioeconomic characteristics (Jensen
and Seltzer, 2000: 23). Measures of the neighbourhood included the level of
neighbourhood income, the unemployment rate, an index of educational attainment
and the percentage employed in professional fields. This study was unable to identify,
however, the precise transmission mechanisms for such neighbourhood effects.
Whether, for instance, they were due to spillover effects such as peer group influence,
the presence or lack of job networks and role models or whether the neighbourhood
variables were acting as proxies for school quality or housing type.
3
Data Description
The data for this study comes from a sample of 3 329 students who were on the Smith
Family’s Learning for Life (LFL) program in 1999. The LFL program is an
intervention aimed at assisting families and children by providing financial and nonmaterial support to families and children in financial disadvantage (almost 90 per cent
95
of families on the program are dependent on social security payments as their main
source of income).3 The sample includes students from Year 1 to 12 from state
schools in NSW, Victoria, Queensland, South Australia and the ACT. The sample
equates to 48 per cent of all school-level students who were on the program at the end
of 1999.4
The measure of educational performance (the dependent variable) used for the
analysis is based on the LFL program staff’s assessment of their students’ overall
grades (on all subjects). The assessment was based on the 1999 mid-year school
report results of the student. 5 Student academic results were classified as: i) below
average ii) average to good iii) outstanding. Student codes allowed the matching of
students’ academic results to family background information contained in the LFL
database. Using school report data is limited in that there may be variation between
schools in assessing and reporting student results. While it would be preferable to use
standard literacy and numeracy test scores of students, such as those now conducted
in several state jurisdictions, this data was not available at the time student results
were collected.
One option was to restrict the sample to students in senior secondary school (Years 11
and 12) where a greater degree of standardisation in school reporting procedures
exists. This would lead to another perhaps more serious limitation, namely, the
censoring problem that may occur due to the fact that poorer performing or
disadvantaged students generally leave school earlier than better performers (see
Portes and MacLeod, 1996: 272). In order to test for a possible drop-out bias, we also
ran separate models for senior secondary and junior/secondary students.
The data allowed us to control for the influence of several factors on student
educational performance. With respect to individual characteristics, these included
sex, age (as proxied by school level), unexplained absences and ethnic background.
3
4
5
Each Learning for Life Education Support Worker has a caseload of 235 students
from between 110 to 115 families. The Support Worker’s role covers a wide range of
duties and at any given time he/she may act as: assessor; advocate/mediator;
supporter/adviser; referral/information source; research information source. For
further details on the history and operation of the program, see Zappalà and Parker
(2000).
At the end of 1999, there were 6 937 students on the program, of whom 41 per cent
were in primary school, 49 per cent in junior secondary, nine per cent in senior
secondary and one per cent at university. Data on academic results for 1 416 students
was not available as some students joined the program after the data collection
exercise in August, and some of the recently established LFL centres did not
participate in the data collection exercise. Some background data (e.g., parental
education) for an additional 2 192 students was missing or not recorded. The results
reported in this paper reflect the findings for those students (3 329) on whom data was
complete. Preliminary analyses indicated no systematic differences between these two
groups of students with regard to the proportion of students getting outstanding school
results or with regard to the individual, family, and contextual characteristics of the
students (e.g.. gender, family composition, parental education level).
Students on the program are required to send copies of their school reports to their
Education Support Worker twice a year. This data is collected for administrative
purposes. As well as data on academic performance, data were also collected on
student application, absenteeism rates and the degree of contact with LFL
caseworkers. For further details see Zappalà and Parker (2000).
96
In order to test for any variation in the educational performance of students from
NESB, students were grouped according to the following regional categories based on
the birthplace of parents: Australian born; English-speaking country; Southern
Europe; other Europe; Asia; Middle East or North Africa; Other Africa; Central and
South America and Pacific Islands (see Marks et al., 2000: 24). Higher percentages of
students from all NESB groups, with the exception of Middle East or North Africa
and Other Africa, achieved outstanding results relative to the Australian born or those
from English speaking backgrounds. With the exception of those from Middle
Eastern/African backgrounds, small cell sizes for the other individual ethnic groups
precluded analysing each ethnic group separately. Those from Middle Eastern and
African backgrounds (n = 217) were re-coded as a new variable for the multivariate
analysis with all other non-English speaking background groups collapsed into a
combined NESB variable (n = 363). (See Table 1).
Family structure controls included whether the student was from a one or two-parent
household. Even though all students were from financially disadvantaged
backgrounds, socio-economic status variables included whether the main source of
parental income was from employment or social security, and the educational
attainment of parents.6 Controls were also included for whether the student/family
lived in a metropolitan or non-metropolitan area, as well as housing type whether
public or private housing. As all students in the sample were in non-selective state
schools, school type was controlled for.
Table 1 presents a profile of the students in the sample. They were almost evenly
divided between male and female students. The majority of students was in either
primary school (40 per cent) or junior secondary school (52 per cent), of English
speaking background (83 per cent) and living in a metropolitan area (55 per cent).
Overwhelmingly, these students lived in households that were dependent on social
security payments (88 per cent) as the main source of income and had parents who
had not completed Year 12 (66 per cent). There was a slightly higher tendency for
students to come from one-parent families (58 per cent) as opposed to two-parent
families (42 per cent) and to live in private (51 per cent) rather than public (49 per
cent) housing.7 With respect to school performance, the majority of students attained
average results (70 per cent). For the remaining students, 13 per cent attained poor
results while 17 per cent achieved outstanding results.
4
Odds Ratio Analysis
The extent to which particular defining characteristics affected the likelihood of
attaining outstanding results is presented in terms of odds ratios in Table 2. Girls, for
instance, were 1.7 times more likely to achieve outstanding results compared to boys.
6
7
Data on income levels were not available so it was not possible to test whether
household income levels differed significantly according to main source of income.
Information on parental education level was only available for the highest educated
parent. Confidentiality requirements precluded identification of this parent’s sex.
Discussions with program staff, however, indicated that for one-parent families (58 per
cent of the sample) parental education level data is for the mother in the majority of
cases.
The majority of those in private housing were in private rental accommodation.
97
Table 1: School Results by Various Characteristics
Total students in sample
Characteristics
Sex
Male
Female
School level
Primary
Junior secondary
Senior secondary
Absences
No unexplained absences
Had absences that were
unexplained
Ethnic Background
Australian born/English
speaking
Non-English speaking
Middle East/African
Location
Metropolitan
Rural
Family Composition
Two-parent family
One-parent family
Parental main source of income
Social Security
Employment
Parental education level
Less than year 10
Completed year 10, not year 12
Year 12
TAFE/other post-secondary
University
Housing Type
Public housing
Private housing
Poor
%
13
School
Average
%
70
Results
Outstanding
%
17
N
3329
Total
%
100
16
9
71
70
14
21
1612
1717
48
52
11
13
12
74
67
65
15
18
23
1321
1722
286
40
52
9
12
13
69
80
19
7
2871
458
86
14
13
71
16
2752
83
9
9
58
78
32
12
360
217
11
6
14
10
67
75
19
15
1834
1495
55
45
13
13
69
71
19
16
1403
1926
42
58
13
11
70
69
17
20
2921
408
88
12
18
13
10
11
5
73
72
70
73
51
9
15
20
16
44
580
1637
413
440
259
17
49
12
13
8
15
9
70
71
15
20
1633
1696
49
51
Older students had improved odds of performing better in school than younger
students, with students in senior secondary school being 1.5 times more likely than
primary school students to achieve outstanding results. Students in junior secondary
students were only 1.1 times more likely to achieve outstanding results compared to
students in primary school. The difference between senior and primary school
students suggest that a selection bias is occurring with poorer performing or
98
disadvantaged students leaving school earlier than better performers. The level of
schooling was controlled for in the multivariate analysis and is discussed below.
Dichotomous comparisons between those with no unexplained absences and those
with unexplained absences, revealed that those students with no unexplained absences
were three times more likely to attain outstanding results compared to those students
with at least one unexplained absence. The odds of achieving outstanding results also
significantly increased for students with parents who had high education levels.
Compared to students whose parents had not completed Year 10, students whose
parents had completed Year 12 were 1.2 times more likely to achieve outstanding
results and students whose parents had a university education were 4.5 times more
Table 2: Odds Ratio of Getting Outstanding Results
Characteristics
Odds Ratio
Sex
Female
1.7
School Level
Junior secondary
1.1
Senior secondary
1.5
Unexplained absence
None
3.0
Ethnic Background
Non English speaking
2.6
Middle East/African
.66
Location
Metropolitan
1.3
Family Composition
One-parent Family
.84
Parental income source
Employment
1.2
Completed Year 10, not Year 12
.83
Year 12
1.2
TAFE/other post- secondary
.83
University
4.5
Housing Type
Private Housing
1.4
Notes: 1) Students on the Learning for Life Program who received outstanding
results.
2) An odds ratio of 1 means there is no difference between the
comparison groups.
Based on the categorisation of ethnic background described earlier, students from
NESB (excluding Middle Eastern/Africa) were almost three times as likely as other
students to gain outstanding results. Students from Middle-Eastern/African
backgrounds were less likely than other students to achieve outstanding results.
99
For students living in a household where the main source of income was from
employment earnings the odds of achieving ‘outstanding’ results was only 1.2 times
higher than for students from households where the main source of income was from
social security benefits. Of slightly more importance than income source appeared to
be the type of housing that students lived in, with those living in private housing 1.4
times more likely to achieve outstanding results compared to those living in public
housing.
Similarly, students from metropolitan areas were 1.3 times more likely to achieve
outstanding results compared to those living in non-metropolitan areas. Students from
one- parent families were slightly less likely to achieve outstanding results compared
to those from two-parent families.
5
Multivariate Analysis
In order to determine the extent to which particular factors influence the achievement
of outstanding results while keeping the effects of other variables constant, we ran a
binomial logistic regression on academic results (1 if the student attained outstanding
results and 0 otherwise). This approach allows us to estimate the ‘pure’ effects of, for
instance, student sex, adjusted for the effects of other variables. The interpretation of
the model is based on the value at which the non-linear independent variables are set.
The independent variables were set at their mean (standard convention).8
Table 3 displays the logistic regression results estimating the extent to which
individual, family, behavioural and socio-economic factors contribute to students’
achieving outstanding results. The model Chi-square statistic was highly significant
(χ2 = 257.404, df = 14, p <.0001) while the Hosmer and Lemeshow test revealed a
good fit between the data and the model (Goodness-of-Fit χ2 = 6.1638, df = 8, p =
.6289).
With respect to the independent variables, the Wald test of significance showed the
coefficients were statistically significant for sex, ethnicity (with the exception of those
students from Middle Eastern/African backgrounds, students from NESB were more
likely to achieve outstanding results), unexplained absences, parent’s educational
attainment, housing type, and student age as reflected by school level. Family
structure (i.e. two-parent vs. one-parent family), the main source of family income,
and geographical location did not significantly predict school performance outcomes.
We also tested for any interactive effects between some of the variables of interest.
No interactive effects were significant, for example, the predicted probability of
achieving outstanding results was the same for boys from a one-parent family as it
was for girls from a one-parent family. In other words, the findings do not support the
argument that one-parent households may have relatively more detrimental effects on
boys than girls.
To facilitate interpretation of the logit model, the Beta coefficients in Table 3 were
converted into predicted probabilities. With all variables held constant at their mean,
an ‘average’ student would have a 15 per cent predicted probability of achieving
outstanding results. Table 4 shows the change in predicted probabilities of attaining
outstanding results based on each defining characteristic. With all other variables held
8
Using a different value (e.g. the median) may alter the interpretation.
100
constant at their mean, female students had a significantly higher predicted probability
of achieving outstanding (18 per cent) results than male students (12 per cent).
Similarly, at a predicted probability of 19 per cent, senior secondary students
performed better than younger students (predicted probabilities were 16 per cent for
junior secondary and 13 per cent for primary school students). Those with no
unexplained absences had higher predicted probabilities (17 per cent) compared to
those with unexplained absences (14 per cent). Students from NESB also had a
significantly higher predicted probability (23 per cent) of outstanding results than
students from either an English-speaking (14 per cent) or Middle Eastern/African
background (eight per cent).
With regard to other significant characteristics, a student who lived in private housing
had a four percentage point higher predicted probability (17 per cent) of attaining
outstanding results than a student who lived in public housing (13 per cent).
Generally, as the parental level of education increased a student’s predicted
probability of attaining outstanding results also increased. A student whose parents
were university educated had a 39 per cent predicted probability of achieving
outstanding results while the predicted probability of attaining outstanding results was
only nine per cent for a student whose parents had not completed Year 10. In the case
of parents with a Year 10 education, a student’s predicted probability of achieving
However, for those students whose parents had completed a TAFE qualification, their
predicted probability of achieving outstanding results dropped to 14 per cent.
In the preceding analysis the influence of individual, family and socio-economic
characteristics have been interpreted with respect to the average student (i.e. with all
other variables set at their mean). However, setting the values of the independent
variables to that of an actual student allows for a more meaningful interpretation of
the influence these variables have on achieving outstanding results. In the following
hypothetical examples, the independent variables have been set to reflect the
individual and family characteristics typical of large groups of actual students in the
sample. These examples illustrate the influence that multiple factors have on a
student’s predicted probability of achieving outstanding results.
Example 1: A student in senior high school from a two-parent family
financially dependent on employment income, from a non-English
speaking background, who has no unexplained absences, lives in
private housing in a metropolitan area, with one parent having a
university education.
Predicted probability of achieving outstanding results: male 63 per cent; female 74 per
cent.
Example 2: A student in senior high school from a two-parent family
financially dependent on employment income, from an English
speaking background, who has one unexplained absence, lives in
public housing in a metropolitan area with one parent who
completed TAFE qualifications.
101
Table 3: Logistic Regression Equation Predicting Outstanding Results
Variables
Female
NESB
Middle East/Africa
Unexplained absences
Private housing
Employment income
Metropolitan location
One-parent family
Parental education
Completed Year 10, not
12
Year 12
TAFE/other
postsecondary
University
School level
Junior secondary
Senior secondary
Constant
.537
.587
-.639
-.201
.283
.111
.196
.038
Sign
SE
Wald [p]
.098 30.0
.000
.148 15.8
.000
.230 7.7
.006
.046 19.3
.000
.102 7.7
.005
.149 0.6
.456
.105 3.5
.061
.105 .1
.716
Odds Ratio
[Exp (β )]
1.711
1.799
.528
.818
1.33
1.12
1.22
1.04
.601
.162
13.7
.000
1.82
.711
.509
.195
.201
13.3
6.4
.000
.011
2.04
1.66
1.857
.202
84.6
.000
6.40
.250
.425
-2.94
.104
.168
.202
5.8
6.4
212.6
.016
.012
.000
1.28
1.53
β
Notes:
1) Students on the Learning for Life Program who received outstanding results.
2) No interactive effects were significant.
3) A second model was run with senior secondary students excluded from analyses to determine
the extent to which level of schooling impacted on the results. The beta co-efficients and standard
errors held constant across both models.
4) Another model was run using ‘unexplained absences’ as a categorical variable. This model
Chi-square was significant but the Goodness-of-fit χ2 was not. For the purposes of multivariate
analyses, therefore, ‘unexplained absences’ was entered as a continuous variable. In order to test
for a possible early school leaver bias (i.e. poorer performing or disadvantaged students generally
leave school earlier than better performers), we also ran separate models for senior secondary and
junior/secondary students. The beta coefficients and standard errors held constant across both
models suggesting that the factors influencing academic performance were consistent across the
different school levels.9
outstanding results was 15 per cent. This increased to 17 per cent for those students
with parents who had completed Year 12.
cent.
9
These results are available on request from the authors.
102
Table 4: Effects of Student Characteristics upon the Predicted probability of
Attaining Outstanding Results
Characteristics
Sex
Male
Female
School level
Primary
Junior secondary
Senior secondary
Unexplained absences
No unexplained absences
One unexplained absence
Ethnic Background
Non English Speaking
Middle East/Africa
Australian born/English speaking
Location
Metropolitan
Rural
Family Composition
Two-parent family
One-parent family
Parental main source of income
Social Security
Employment
Less than Year 10
Completed Year 10, not year 12
Year 12
TAFE/other post -secondary
University
Housing Type
Public housing
Private housing
Predicted
probability
(%)
Percentage
Difference
12
18
6
13
16
19
3
6
17
14
-3
23
8
14
-15
-9
16
14
-2
15
15
0
15
16
1
9
15
17
14
39
6
8
5
30
13
17
4
Notes:1) Students on the Learning for Life Program who received outstanding
results.
2) All other variables are set at mean values.
Example 3: A student in primary school from a one-parent family
financially dependent on Social Security benefits, from an English
speaking background, with three unexplained absences, lives in public
housing in a metropolitan area with one parent who did not complete
Year 10.
103
cent.
Example 4: A student in junior secondary school from a one-parent
family financially dependent on Social Security benefits, from an
English speaking background, with three unexplained absences, lives
in private housing in a rural area with one parent who did not
complete Year 10.
Predicted probability of achieving outstanding results: male five per cent; female eight
per cent.
6
Discussion and Policy Implications
The above results raise several implications for public policy with respect to
education and human services more generally. A key finding is that even within a
group with considerable financial disadvantage, socio-economic status as reflected by
the level of parental education, was a key predictor of student academic achievement.
This finding lends support to the notion advanced by some studies that the social and
the economic components of the socio-economic status equation may have distinct
and separate influences on educational outcomes. While both components are
important, social factors, such as parents’ educational attainments, have been found to
be more significant than economic factors in explaining children’s educational
outcomes and among the most replicated results in child development studies
(Shonkoff and Phillips, 2000). Higher status families, some researchers suggest, foster
a higher level of achievement and provide higher levels of psychological support for
their children (Williams, 1987; Williams et al., 1993). Our findings suggest that this is
the case even within financially disadvantaged families.
Most government approaches to addressing the effects of low SES in education are
aimed at the economic redistribution of resources and direct financial assistance
(Graetz, 1995). Some have therefore concluded that ‘there is some danger of
recommending cures for socio-economic disadvantage when the malady is social
rather than economic’ (Williams et al., 1993: 23). While not wishing to discount the
importance of financial assistance (whether to schools or families), policies and
programs that also assist parent/s in providing appropriate psychological and
educational support for their children should therefore be encouraged. The level of
parental education, for instance, has been found to be strongly associated with factors
such as the home literacy environment, parents’ teaching styles and investing in
resources that promote learning such as quality child care, educational materials and
visits to museums (Shonkoff and Phillips, 2000). While families with low income face
greater hurdles in achieving effective parenting which in turn often harms their
children’s development and educational achievement (Berk, 1997: 549), our findings
support the thesis that low income alone is not the only factor.
The findings are therefore consistent with the current policy direction at state and
federal level with respect to early intervention programs such as Families First
(NSW) and the recent Stronger Families and Communities Strategy initiated by the
federal Department of Family and Community Services. While not aimed solely at
educational issues, these programs are designed to assist parents in preparing their
104
children for school and provide appropriate community based services for a range of
parenting based needs, support and advice (Calvert, 2000; Newman, 2000).
In contrast to much publicised recent research and media comments on the negative
effects of one-parent families on children, however, our findings do not support such a
conclusion. Neither do our findings support the argument that one-parent households
may have relatively more detrimental effects on boys than girls. Consistent with other
studies, however, our findings do confirm the existence of a significant gender gap in
educational achievement among students from low socioeconomic status. This
suggests that the current policy focus on improving the performance of boys at school
should also ensure that any programs cater for boys from low SES families (Kemp,
2000; Arndt, 2000; West, 2000). Also consistent with other recent studies, our
findings suggest that some ethnic groups seem to be more disadvantaged than others
in terms of educational outcomes (Cahill, 1996; Dobson et al., 1996).
Finally, while geographical location was not a significant predictor of academic
achievement, whether children live in private or public housing was found to be
significant even after controlling for other factors. Whether housing type is acting as a
proxy for neighbourhood influences, such as lower levels of local social capital in
some public housing estates or lower levels of economic resources, is unclear. Using
the Index of Relative Socio-Economic Disadvantage developed from the 1996
Census, 36 per cent of students in the sample, for instance, lived in areas that ranked
in the bottom quartile of this index, with just under a quarter in the next quartile
(ABS, 1998). In any case, the significance of housing type suggests that public policy
approaches to addressing disadvantage which are neighbourhood and community
based, such as place management initiatives should be encouraged (Zappalà and
Green, 2001).
7
Conclusion
Although there is a vast literature on the relationship between family socio-economic
status and the academic performance of children, the factors that may influence
educational outcomes within particular SES bands have not been as exhaustively
examined. This paper presented new data on over 3 000 students from financially
disadvantaged backgrounds to estimate the extent of socio-economic, family,
individual and contextual factors on school educational performance. Despite some
data limitations, our model specification and results were robust. The results from the
logistic regression indicated that sex, unexplained absences, ethnicity, parental
educational attainment, housing type and student age as reflected by school level were
all statistically significant variables and predictors of academic performance. In
contrast, family structure, the main source of family income and geographical location
did not significantly predict variation in school performance once other factors were
controlled for.
The finding that, even within a group with considerable financial disadvantage, socioeconomic status as reflected by the level of parental education was a key predictor of
student academic achievement raises several policy implications.
In brief, it supports the notion that the social and the economic components of the
socio-economic status equation may have distinct and separate influences on
educational outcomes. While financial assistance to schools and families in need is
105
important, policies and programs that assist low-income parent/s in providing
appropriate psychological and educational support for their children should also be
promoted. Low income is not the only factor that harms effective parenting and
children’s development and educational achievement. The findings also lend support
to policies that address issues of boys’ behavioural problems, the needs of some
groups from NESB, as well as neighbourhood and community based frameworks for
addressing disadvantage.
References
Ainley, John., Brian Graetz, Michael Long and Margaret Batten (1995), Socioeconomic Status and School Education, DEET/ACER, Canberra.
Amato, P. (1987), Children in Australian Families: The Growth of Competence,
Prentice Hall, Sydney.
Arndt, B. (2000), ‘Attention boys.’ Sydney Morning Herald, 28 October, 35.
Australia Bureau of Statistics (1998), Socio-Economic Indexes for Areas, Catalogue
No. 2039.0, ABS, Canberrra.
Berk, L. (1997), Child Development (4th edition), Allyn and Bacon, Massachusetts.
Birrell, R and S. Khoo (1995), The Second Generation in Australia: Educational and
Occupational Characteristics, BIMPR, Canberra.
Buckingham, J. (1999), ‘The puzzle of boys’ educational decline: a review of the
evidence’ Issue Analysis, No.9, Centre for Independent Studies, Sydney.
Buckingham, J. (2000a), ‘The truth about private schools in Australia’, Issue Analysis,
No.13, Centre for Independent Studies, Sydney.
Buckingham, J. (2000b,) Boy Troubles: Understanding Rising Suicide, Rising Crime
and Educational Failure, CIS Policy Monograph 46, Centre for Independent
Studies, Sydney.
Cahill, D. (1996), Immigration and Schooling in the 90s, BIMPR, Canberra.
Calvert, G. (2000), ‘Valuing children, young people and families’, Family Matters,
56, 28-33.
Cheers, B. (1990), ‘Rural disadvantage in Australia’ Australian Social Work, 43 (1),
5-13.
Dobson, I., B. Birrell, and V. Rapson (1996), ‘The participation of non-Englishspeaking background persons in higher education,’ People and Place, 4 (1), 4654.
Graetz, B. (1995), ‘Socio-economic status in education research and policy’ in John
Ainley et al., Socio-economic Status and School Education DEET/ACER
Canberra.
Horne, R. (2000), ‘The performance of males and females in school and tertiary
education’, Australian Quarterly, 72 (5/6), 21-26.
Human Rights and Equal Opportunities Commission (2000), Emerging Themes:
National Inquiry into Rural and Remote Education, HREOC, Sydney.
Jensen, B. and Seltzer, A. (2000,) ‘Neighbourhood and family effects in educational
progress’ The Australian Economic Review, 33(1), 17-31.
Kemp, D (2000), ‘Opening address – Educational attainment and labour market
outcomes, factors affecting boys and their status in relation to girls’, Australian
Quarterly, 72(5/6), 15-20.
Khoo, S. (1995), ‘The second generation in Australia’, BIMPR Bulletin, No.14, 11-18.
Long, Michael, Carpenter, P and Hayden, M (1999), Participation in Education and
Training 1980 – 1994, LSAY Research Report No.13, Australian Council for
Educational Research, Camberwell, Victoria.
106
Machin, S. (1998), ‘Childhood disadvantage and intergenerational links’ in Atkinson,
A.B and J. Hills, eds, Exclusion, Employment and Opportunity, CASE Paper
No. 4, Centre for Analysis of Social Exclusion, London School of Economics,
London.
Marks, G. et al. (2000), Patterns of Participation in Year 12 and Higher Education in
Australia: Trends and Issues, LSAY Research Report No.17, Australian
Council for Educational Research Camberwell, Victoria.
Martin, J. (1978), The Migrant Presence, Allen and Unwin, Sydney.
Mukherjee, D. (1995) ‘The relationship between socio-economic background and
participation in education’ ACEE Research Monograph No.1, Darlinghurst
Newman, J. (2000). ‘Minister opens AIFS conference’, Family Matters, 56, 54-57.
Portes, A and D. MacLeod (1996), ‘Educational progress of children of immigrants:
the role of class, ethnicity and school context’, Sociology of Education, 69, 25575.
Rich, A. (2000), Beyond the Classroom: How Parents Influence their Children’s
Education, CIS Policy Monograph 48, Centre for Independent Studies, Sydney.
Ruge, J. (1998), ‘Raising expectations: achieving quality education for all’, DSP
Discussion Paper, NSW Department of Education and Training, Sydney
Shonkoff, J. P and D. A. Phillips, eds, (2000), From Neurons to Neighborhoods: The
Science of Early Childhood Development, National Academy Press, Washington
DC.
Sparkes, J. (1999), Schools, Education and Social Exclusion, CASE Paper 29, Centre
for Analysis of Social Exclusion, London School of Economics, London.
Teese, R., Davies, M., Charlton, M., & Polesel, J (1995), Who Wins at School? Boys
and Girls in Australian Secondary Education, Macmillan, Sydney.
West, P. (2000), ‘Raising boys: who will we listen to?’ Australian Quarterly, 72 (5/6),
35-38.
Williams, N. Penelope, R.W. Connell and V.M. White (1991), ‘Australian research on
poverty and education, 1979-1987’, in R.W. Connell, V.M. White and K.M.
Johnston, eds, ‘Running Twice as Hard’: The Disadvantaged Schools Program
in Australia, Deakin University Press, Geelong.
Williams, T. (1987), Participation in Education, Australian Council for Educational
Research, Hawthorn.
Williams, T., J. Clancy, M. Batten and S. Girling-Butcher (1980), School, Work and
Career: Seventeen-Year-Olds in Australia, Australian Council for Educational
Research, Hawthorn.
Williams, T., M. Long, P. Carpenter and M. Hayden (1993), Year 12 in the 1980s,
AGPS, Canberra.
Zappalà, G. and B. Parker (2000) ‘The Smith Family’s Learning for Life program a
decade on: poverty and educational disadvantage’ Background Paper No. 1,
Research and Advocacy Team, The Smith Family, Sydney.
Zappalà, G. and V. Green (2001), ‘Addressing disadvantage through place
management: is there a role for non-profit organisations?’ Working Paper No.3,
Research and Advocacy Team, The Smith Family, Sydney.
107
Employment Assistance for Long-term
Unemployed People: Time for a Rethink
Peter Davidson
Australian Council of Social Service
1
Introduction
In 1998, employment assistance for long-term unemployed people was restructured
and funds were cut. The job compact - a guarantee of a temporary job after 18 months
of joblessness, was replaced by intensive assistance - a more open-ended system of
support provided through the newly-established Job Network.
The theory behind the Job Network funding model is appealing: employment
assistance providers are best placed to judge what assistance each job-seeker requires,
and appropriate, cost effective assistance will be offered if funding is tied to
employment outcomes rather than programs. However, the result has been a reduction
in the level of support provided to most long-term unemployed people, leading to
poorer employment outcomes than the most effective (though not all) Job Compact
programs raise.1
Competitive tendering was effective in driving costs down but it weakened service
quality. Further, the system of outcome-based payments shifted the risk of investing in
substantial employment assistance from Government to providers, who have been
reluctant to risk their own funds on more costly interventions.
The solution is to shift some of the risk and political responsibility back to
Government, which should guarantee substantial employment assistance to all longterm unemployed people. This could be implemented within the Job Network model
by making a pool of funds available to providers so that they can offer the assistance
each long-term unemployed job-seeker needs, while retaining the system of outcomebased payments.
2
What Should Employment Assistance Do?
Employment assistance for jobless people should do at least three things.
•
improve the efficiency of job matching;
•
lower workforce barriers for disadvantaged job-seekers; and
•
reduce long-term unemployment and joblessness.
1
For a more detailed treatment of these issues, see ACOSS (2000) and ACOSS (2001).
Davidson, P. (2002), ‘Employment assistance for long-term unemployed people: time for a re-think’, in
T. Eardley and B. Bradbury, eds, Competing Visions: Refereed Proceedings of the National Social
Policy Conference 2001, SPRC Report 1/02, Social Policy Research Centre, University of New South
Wales, Sydney, 107-125.
PETER DAVIDSON
There are three reasons for a focus on long-term unemployment and joblessness. The
first is equity. The longer a person remains unemployed, the greater the risk of
poverty, poor health, family conflict and social isolation. The second is to reduce
unemployment. Lowering long-term unemployment helps improve the functioning of
the labour market, thereby reducing the overall level of unemployment. Long-term
unemployment is a major factor contributing to our high level of structural
unemployment - that part of unemployment that has ratcheted upwards in every
business cycle in Australia since the 1970s.
The third is cost effectiveness. Most short-term unemployed people will get a job with
minimal help, so it makes sense to target the most expensive employment assistance
towards long-term unemployed people and those who are identified as being at high
risk of long-term unemployment.
The focus of this paper is therefore on the effectiveness of employment assistance for
long-term unemployed people over the 1990s. We compare two programs that
comprised part of the Working Nation package in the mid 1990s, Jobstart and
Jobskills, with two present-day schemes, Intensive Employment Assistance and Work
for the Dole.
We begin with a brief description of the two radical experiments undertaken in
employment assistance in Australia during the 1990s: Working Nation and the Job
Network.
3
Working Nation and the Job Compact
The Working Nation strategy was introduced in 1994 in response to a sharp rise in
long-term unemployment following the recession of the early 1990s. Its centrepiece
was the Job Compact. The Job Compact guaranteed all unemployed people a
temporary job for around six months after they had been out of work for 18 months.
This was intended as a circuit breaker to provide employment experience and training
that would improve their employment prospects.
The guarantee was a very important commitment for the Government to make. Until
then, Australian governments guaranteed unemployed people access to social security
payments and basic job matching assistance through the CES, but more intensive
employment assistance was rationed and only a minority of long-term unemployed
people benefited in any given year.
Working Nation was implemented through a series of employment assistance
programs. They were of three types:
•
paid employment experience programs;
•
training and personal support programs; and
•
programs that combined employment experience and training.
110
EMPLOYMENT ASSISTANCE FOR LONG-TERM UNEMPLOYED PEOPLE
Access to programs was facilitated by a system of case management. Case managers
were responsible for counselling each disadvantaged job-seeker, working out a plan of
action to secure employment and referring them to an appropriate mix of programs.
Case managers, but not programs, were mainly funded according to the employment
and training outcomes they achieved. This funding was in accordance with a fixed
schedule of fees that was meant to reflect the degree of labour market disadvantage of
each job seeker.
Two of the more effective programs were Jobstart and Jobskills. We describe them
below.
Key Features of Two Working Nation Programs
Jobstart
•
Targeted towards job ready long-term unemployed people
•
Six months paid employment
•
Roughly half a full-time wage subsidised by government
•
Most placements in the private sector
Jobskills
•
Targeted towards more disadvantaged long-term unemployed people
•
Six months subsidised employment and training
•
Almost 100 per cent of the wage subsidised by government
•
Three-fifths work experience two-fifths training
•
Placements in the community sector and local government
The Job Compact achieved positive employment outcomes for many long-term
unemployed people, but it did not meet expectations in reducing the overall level of
long-term unemployment. One reason for this was the economic slow-down of the
mid-1990s. Another reason was flaws in its design and implementation.
One major design flaw was the assumption that all long-term unemployed people
needed the same kind of assistance - a temporary subsidised job. A related flaw was
that the program-based delivery system meant that the case manager role was largely
reduced to slotting people into available places in programs, often inappropriately.
Thirdly, there were problems with the implementation of the Job Compact. Program
places were expanded too rapidly, there was too much focus on very long-term
unemployed people (out of work for five years or more) at the outset, and too little use
of the most effective programs such as Jobstart and Jobskills.
Nevertheless, these programs were effective despite being implemented on a large
scale, and the overall effect of the Job Compact was to reduce long-term
unemployment significantly.
111
PETER DAVIDSON
4
The Job Network and Work for the Dole
The Job Compact was abolished from 1997 and replaced by a second radical
experiment, the Job Network. Ostensibly, this was done because the Job Compact had
failed. However, a major consideration was the $1 billion per annum in expenditure
savings achieved by abolishing the Job Compact and introducing competition in the
provision of public job matching services. In 1997-98, employment assistance outlays
for jobless people were cut by approximately 50 per cent.
The Job Network provides three tiers of employment assistance: job matching, job
search training and intensive employment assistance. These services are delivered by a
combination of community, private sector and government agencies, who tender for
Job Network contracts.
Intensive assistance is the highest level of assistance and it attracts the greatest level of
funds. It is targeted towards the most disadvantaged jobseekers, using an assessment
tool known as the Job Seeker Classification Instrument. However, there is no
guarantee of access to intensive assistance for long-term unemployed people. In the
September quarter 2000, 25 per cent of long-term unemployed social security
recipients were denied access to Intensive Assistance following assessment, mainly on
the grounds that they were not sufficiently disadvantaged (DEWRSB, 2001a). This is
a curious definition of disadvantage, when long-term unemployment is widely
recognised both here and overseas as a key indicator of labour market exclusion.
Intensive Assistance is not program based. Government funding is linked to
employment outcomes rather than the services provided. In many ways, this is a better
model because it concentrates the minds of service providers on outcomes rather than
securing places in programs. However, a pure outcomes-based model has major
drawbacks, as we explain below.
Work for the Dole is the other key employment assistance scheme for long-term and
other disadvantaged job seekers. This provides part time temporary employment on
community projects, in return for unemployment benefits.
The irony of the introduction of Work for the Dole is that it shares features in
common with old-fashioned job creation programs such as the New Work
Opportunities program within the Working Nation strategy. Yet it was introduced just
as these previous programs were abolished, on the grounds that employment
assistance should be more flexible and outcomes-based.
Work for the Dole is job creation on the cheap. Payment for participants is limited to
unemployment benefits plus a small supplement, and the service providers receive
very little financial assistance to provide supervision or training. Improving
employment outcomes is not one of the official objectives of this scheme. Rather, it
was introduced as part of a broader strategy to re-orient employment assistance to
meet welfare compliance goals, on the assumption that unemployment is to a large
extent a behavioural problem. While compliance is important, an excessive focus on
compliance goals has in recent years distorted employment assistance priorities from
their original purpose of positively assisting people to obtain employment (ACOSS,
2001).
112
Key features of Two Working Nation Programs
Intensive Employment Assistance
•
Service providers are usually paid up to $4 500 or $8 500 per client (depending on
degree of labour market disadvantage) to achieve employment outcomes
•
After taking account of overheads and those clients who don’t gain employment,
this falls on an average of about $200 to $400 per client
•
The services provided are not specified by Government, but these are mainly
confined to job search support (though a minority of clients receive subsidised
employment or vocational training).
‘Work for the Dole’
•
Six months work on community projects
•
Usually for two days a week
•
Payment is Newstart or Youth Allowance plus a $10.40 per week supplement
5
What Happened to Long-term Unemployment?
Trends in unemployment and long-term unemployment over the 1990s should provide
some clues as to the relative effectiveness of the above employment assistance
schemes. If employment assistance is working well, we would expect that, during an
economic recovery, long-term unemployment would decline in line with reductions in
overall unemployment levels, with a one year delay. This would indicate that
employment assistance is overcoming the relative disadvantages confronting longterm unemployed people, such as skills deficits or a lack of recent work experience.
If, on the other hand, there are signs of persistence in the long-term unemployment
data, this would suggest that employment assistance is not working effectively.
Our two Working Nation programs formed part of the Job Compact from 1994 to
1997. Intensive Employment Assistance and Work for the Dole operated from 1998 to
the present.
Figure 1 outlines trends in the numbers of unemployment benefit recipients2 from
1990 to 2000. We use unemployment benefit statistics rather than ABS unemployment
data because the latter is a very limited measure of labour market disadvantage.3
2
Newstart Allowance and Youth Allowance (Other)
3
One of the drawbacks of using unemployment benefit data for this purpose is that
changes in eligibility conditions (as distinct from labour market conditions) have an
influence on the number of recipients. However, a detailed examination of these
trends over the 1990s suggests that changes in eligibility conditions over the 1990s
did not have a major net impact on the numbers. See Warburton, Opoku and Vuong
(1999).
113
PETER DAVIDSON
Figure 1: Unemployment Benefit Recipients
950
Persons ('000s)
750
Unemployment
benefits
550
Long-term
unemployment
benefits
350
150
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
Year
Sources: Centrelink (various years); Warburton, Opoku and Vuong (1999)
Note: Newstart Allowance plus Job Search Allowance plus Youth Allowance (unemployed)
recipients.
There are two main reasons for this:
•
First, people who obtain an hour's employment in a given week are classified by
the ABS as employed. This means that the growing number of under-employed
people who are either employed for very short part-time hours or cycle between
unemployment and casual work are excluded from the ABS unemployment data;
•
Second, people with a disability or temporary illness that prevents them from
working in the short term are also excluded because they are unlikely to be
actively seeking and available for employment at the time
Another reason for using unemployment benefit data is that the Government has set
itself the objective of reducing the number of people reliant on social security
payments.
The above data can be broken down into four periods:
•
the recession in the early 1990s, which led to a sharp rise in both unemployment
and long-term unemployment;
114
•
a brief period of strong employment growth from 1993 to 1995, in which both
unemployment and long-term unemployment began to fall;
•
a period of sluggish employment growth from 1995 to 1998, in which
unemployment remained flat and long-term unemployment tracked upwards;
•
a subsequent period of strong employment growth.
The graph indicates that long-term unemployment tracked the reduction in
unemployment during the Working Nation period, but there is evidence of persistence
in the period following the introduction of the Job Network, despite strong
4
employment growth.
On the face of it, this suggests that the Working Nation programs were more effective
in reducing long-term unemployment than the present schemes. However, it is
difficult to assess program performance using macro-economic data because many
factors other then employment assistance influence the results. In addition, there is an
element of churning in employment assistance programs that often distorts these
unemployment statistics.5 Nevertheless, it is of great concern that the number of
people on unemployment benefits long-term has hardly shifted over the past three
years of solid growth in employment opportunities. This should sound alarm bells for
policy makers.
6
Comparing Effectiveness
There is little point in comparing the overall outcomes of the Working Nation
programs and present employment assistance schemes. Both systems contain flaws. It
would be undesirable and unlikely for a future government to restore the previous
regime in its entirety. However, if we aim to achieve best practice in the delivery of
employment assistance, we should learn from past experience. It is regrettable that
Working Nation was abandoned in the mid-1990s without an adequate evaluation of
its effectiveness (and that of its component parts). If evolutionary change had been
pursued instead of tearing down the old foundations to construct the new (the Job
Network), we would not have lost the program and service delivery infrastructure that
had been built up over many years. We might also have retained those elements of the
previous system that were working well. To put this another way: if some of the
Working Nation programs were more effective than present employment assistance
services, there is a strong case for offering this kind of assistance within the Job
Network framework.
4
The ABS data suggests that long-term unemployment fell more sharply in the late 1990s, with
the significant exception of people unemployed for two years or more. The main reason for this
discrepancy between ABS and social security data is probably growth in casual employment
among unemployment benefit recipients in recent years.
5
While people are participating in employment programs they may go off benefits. They might
no longer be classified as long-term unemployed afterwards, even if they fail to secure a job.
115
PETER DAVIDSON
We therefore compare two of the most effective Working Nation schemes - Jobstart
and Jobskills, with the two major contemporary schemes for disadvantaged job
seekers (Intensive Assistance and Work for the Dole).6
The best way to compare the employment outcomes of different labour market
programs is to conduct a net employment impact study, in which the employment
outcomes following participation in a program are compared with those achieved by a
matched sample of non-participants.
Unfortunately no such study has been published in respect of the two existing
programs, despite the fact that they have operated now for three years.7 The
Employment Department has released net benefit impact studies for both the above
schemes, but these studies ask a different question. They ask what effect these
programs have on unemployment benefit receipt. For example, an off-benefit outcome
in these studies might include the transfer of a job seeker from unemployment benefits
to a Disability Support Pension.
A second-best option for the information-poor is to use the official Post Program
Monitoring data series to compare the proportions of program participants who were
in an unsubsidised job three months after they left a program.
Four factors should be considered when making this comparison:
•
The characteristics of program participants
In this paper, we use the proportion of program participants who are unemployed
long-term as a proxy for labour market disadvantage. This is not ideal, but duration of
unemployment is a strong indicator of labour market disadvantage. It should be kept
in mind that this approach advantages the Jobstart program to some extent by
comparison with the other three programs, especially Jobskills and Intensive
Assistance. This is because those selected to participate in Jobstart were relatively job
ready. Job seekers who were harder to place were streamed into other Job Compact
programs, such as Jobskills and New Work Opportunities. Further, access to Intensive
Assistance is based on an assessment that a job seeker is at risk of long-term
unemployment, even though she may only have been unemployed for a short time.
•
The state of the labour market
We use data for 1994-95 and 1999-00 for our comparisons on the grounds that
employment growth was strong, and unemployment was falling, in both these years.8
6
It might be argued that this comparison is unfair because the previous schemes were only part
of the Job Compact package. However, these schemes were offered on a very large scale in the
mid 1990s. During 1994-95, more than 100,000 job seekers were placed in one or the other.
Since our objective is to assess how the present regime matches the best practice of the past,
and not to compare the overall effectiveness of the Job Compact and Intensive Assistance, this
seems a reasonable approach.
7
I understand that studies along these lines are now under way and that they will be published
before the end of 2001.
8
A number of official studies use 1995-96 data to assess the effectiveness of Working Nation
programs. They do not take sufficient account of the fact that employment growth was sluggish
116
•
The sustainability of employment outcomes
A valid criticism of the use of Post Program Monitoring data is that employment
outcomes achieved three months after completion of a program might not be sustained
over time. However, previous net impact studies of the two Working Nation programs
(Jobstart and Jobskills) indicate that their positive employment outcomes were
sustained in most cases over a period of at least 12 months (DEETYA, 1997) We have
no data on the sustainability of employment outcomes (as distinct from off-benefit
outcomes) for the current programs.
•
The average cost per place in each scheme;
The relative cost of each program should also be taken into account, and weighed up
against their employment outcomes and the labour market disadvantages faced by
their participants. Characteristics of Participants Figure 2 below compares the
participant profiles (with regard to duration of unemployment) of the four schemes.
Figure 2: Percentage of Clients who were Long-term Unemployed
Sources: DEET (1994-95); DEWRSB (2001b)
100%
90%
80%
93%
70%
83%
79%
60%
62%
50%
Work for the
Dole (1999-2000)
Intensive
Assistance
(1999-2000)
Jobskills (199495)
Jobstart (199495)
Figure 2 suggests that:
in that year, and well below the levels attained in the late 1990s and in 2000. See, for example,
DEETYA (1997).
117
PETER DAVIDSON
•
the proportion of long-term job-seekers in both Working Nation programs was
higher than in the present schemes. This reflects the targeting of the Job
Compact to long-term unemployed people;
•
as the above discussion and these data suggest, Jobskills was targeted towards a
more disadvantaged group of long-term unemployed people than was Jobstart;
•
the proportion of long-term job-seekers in Work for the Dole was higher than
for Intensive Assistance.
Employment Outcomes
Figure 3 below compares the post program employment outcomes of each of the
schemes.
Figure 3: Percentage of Participants employed Three Months Later
60%
50%
40%
59%
30%
33%
41%
36%
20%
Work for the
Intensive
Jobskills (1994- Jobstart (1994Dole
(1999Assistance
95)
95)
2000)
(1999-2000)
% of all clients in jobs 3 months after assistance
Sources: DEET (1994-95), Annual Reports; DEWRSB (2001c)
•
On the face of it, both Working Nation programs were more effective than the
current schemes;
•
Jobstart appears to have been the most effective (though its targeting of job ready
long-term unemployed people should be kept in mind);
118
•
Intensive Assistance appears to be more effective than Work for the Dole
(although it assists a lower proportion of long-term job-seekers).9
Like Jobstart and Jobskills, Work for the Dole offers paid employment experience
over a period of roughly six months. Its relatively poor outcomes probably reflect the
quality of the employment experience provided, and the greater distance between a
Work for the Dole placement and mainstream employment.
Costs
Figure 4 below compares the average cost per placement in each scheme.10
Figure 4: Average Cost of Place in Program
$10,000
$660
$8,000
$6,000
$8,200
$4,000
$660
$2,000
$2,000
$2,200
$1,500
$0
Work for the Dole
Intensive
Jobskills (1995- Jobstart (1995(1999-00)
Assistance (199996)
96)
00)
program costs
case management
Sources: ESRA (1994-96); DEETYA (1997); DEWRSB (2001a); DEWRSB (2001b).
•
9
10
the costs of Work for the Dole, Intensive Assistance and Jobstart are similar;
Note that, for comparative purposes, these figures exclude people who moved on to
another program afterwards. If they were included, the outcome rate for Work for
the Dole would fall by around five percentage points to 27 per cent. Intensive
Assistance outcomes would not be significantly affected. On this alternative
measure, therefore, Intensive Assistance has significantly better outcomes than
Work for the Dole. Outcomes for the Working Nation programs would fall by an
unknown amount. See DEWRSB (2000a).
It should be kept in mind that the figures for Jobskills and Jobstart are net of
unemployment benefit offsets (that is the savings in unemployment benefits while
people participate in the program and draw a wage instead). There are no such offsets
for the two existing schemes.
119
PETER DAVIDSON
•
the cost of a Jobskills placement was significantly higher (but its targeting of
more disadvantaged long-term unemployed people should be kept in mind).
7
Flaws in the Job Network Funding Model
These data suggest that if Job Network providers or the Government offered long-term
unemployed people paid employment experience for six months, along the lines of the
previous Jobstart or Jobskills programs, better employment outcomes could be
achieved in the robust labour market conditions that prevailed in 2000.
Taking account of what we know about the characteristics of clients of Intensive
Assistance and Work for the Dole, their employment outcomes are poor by
comparison. Moreover, Jobstart would probably have achieved better outcomes for at
least some long-term unemployed people at a net cost that is not significantly greater
than the existing schemes. It might be argued that these outcomes would only have
been achieved for the more job ready long-term unemployed job seekers.
Nevertheless, on the basis of past experience, wage subsidies along Jobstart lines
could still assist large numbers of long-term unemployed people into employment.
Jobskills-type assistance (in effect a short traineeship for long-term unemployed
people) would have been more costly, but more effective for harder-to-place long-term
unemployed people.
This raises a critical question: if these kinds of assistance are cost-effective, why don't
Intensive Assistance providers offer them? There are three likely reasons:
•
The cost of Jobstart-style wage subsidies for Intensive Assistance providers is at
least twice the amount indicated in Figure 4 because, unlike the Government,
they do not derive any direct benefit from unemployment benefit offsets;
•
Current funding arrangements leave Intensive Assistance providers with only
small amounts of public funding to spend on assistance such as wage subsidies.
Beyond this, they have to risk their own funds;
•
The risk premium for those who do this is far too low. That is, the outcomes
payments do not adequately compensate providers for the risk that a positive
employment outcome might not be achieved.
Why are outcome payments too low? The main reason is that when providers bid for
Intensive Assistance tenders in 2000 their primary concern was that their competitors
would undercut them on price. Although they could reasonably anticipate that a higher
investment in each client would yield better employment outcomes, they could not be
certain that the funding body would view a high bid in this light.
This raises a serious problem with competitive tendering arrangements in a monopoly
market such as the Job Network.11 Providers are caught between the risk of over-
11
That is, a market dominated by a single purchaser, in this case the Government.
120
bidding, and being undercut by competitors, and under-bidding and failing to achieve
strong enough employment outcomes.
Under the Job Network funding arrangements, the Government has shifted the risk of
investing heavily in employment assistance for long-term unemployed job seekers to
the providers. In many ways, government is better placed to take this risk, since it can
pool the risk across all long-term unemployed job seekers, it has more resources, and
it benefits directly from the savings in social security payments arising from
successful employment outcomes.
At the very least, the Government should offer a substantially higher risk premium in
the form of outcome payments. However, this would shift much of the risk back to
government, and in more ways than one. Employment assistance schemes are always
vulnerable to political attack because only a proportion of their participants obtain
employment. The Intensive Assistance arrangements are especially vulnerable because
providers are not directly accountable for the services they provide. Some might, for
example, offer minimal assistance on the basis that many of their clients are going to
achieve positive employment outcomes in any event. This limited form of direct
accountability for public funds has already attracted much controversy. Yet this very
feature of the system is also one of its greatest advantages. It means that providers
have the flexibility to decide what form of assistance they should offer each job
seeker.
The basic problem with the Job Network funding model is that as a pure outcomesbased model it fails to give providers sufficient resources and incentives to offer
substantial help to the most disadvantaged job seekers. If some providers offer limited
assistance, or concentrate their efforts on the easiest to place clients, they can hardly
be blamed for behaviour that is generated by the funding model itself. Most providers
do attempt to offer long-term unemployed people the best assistance they can afford
within these constraints. However, unless they draw on their own resources, this is not
a great deal.
There is very little information available to the public on the kinds of services
Intensive Assistance providers offer their clients. However, official evaluations of the
Job Network (DEWRSB, 2000a; 2001b) indicate that:
•
In the first funding round (up to 2000), less than 25 per cent received vocational
training (and the evaluation report made no mention at all of wage subsidies);
•
In regard to the second round (from 2000), the evaluation indicates that those jobseekers who did not secure a job within three months of participation in Intensive
Assistance had little chance of getting one.12 This suggests that after the first three
months (and once the easier-to-place clients have secured a job) the providers had
run out of ammunition to assist those with more intractable employment barriers.
12
In fact, less than a 20 per cent chance. Most of these people were unemployed for
more than two years.
121
PETER DAVIDSON
8
An Alternative Model
One of the key lessons from the two radical employment assistance experiments of the
1990s is that it is better to build on existing foundations than tear them down, even if
radical change is needed. Another lesson is that it is a grave mistake to make shortterm budgetary savings at the expense of long-term unemployed people, once the
economy recovers. If long-term unemployment is not brought under control in one
business cycle, then unemployment will be higher in the next. A third lesson is that we
must find ways to marry the flexibility of Job Network model with the greater
effectiveness of previous interventions such as wage subsidies and traineeships for
long-term unemployed people.
With this in mind, ACOSS is developing an alternative model of employment
assistance for long-term unemployed people. It has four steps:
•
guaranteed access to Intensive Assistance for long-term unemployed people
All long-term unemployed people, as well as others who are assessed as facing a high
risk of long-term unemployment, should be referred for Intensive Assistance after 12
month's unemployment, unless they need specialised help such as disability
employment programs.
All referrals to employment assistance, apart from job matching, should be made
following an interview with an employment assistance specialist officer at Centrelink.
This helps people understand how the system works and how to make a genuine
choice of service provider. The present system of automated referral (referral by mail)
to some employment programs should be abolished to improve their very low take-up
rates;
•
personal advisers
Intensive Assistance providers should be required to assign each client a personal
employment adviser, to assist them for the duration of their placement.
The Government's Australians Working Together package recognises the critical
importance of personal advice and support and allocates resources for this purpose.
However, the proposed system is very complex. The same client may have to deal in
quick succession with three or more personal advisers or service brokers.13 Personal
advisers should be in a position to assess a client's needs over time and mobilise
resources to meet them. This suggests that they should either be located within a
service that is providing employment assistance directly, or a brokerage service such
as the Jobs Education and Training program.
The proposed personal adviser and brokerage arrangements should therefore be
rationalised to improve their efficiency and make the system simpler for job seekers.
•
The 'gateway'
13
One located in Centrelink (where there would be a number of layers of personal
advisers for the same client groups such as sole parents), one with a Community Work
Coordinator and one in the Job Network.
122
Once a client is referred to Intensive Assistance, their personal adviser should
undertake an extended assessment over a three-month period while assisting them
with job search. This should enable the adviser to assess more precisely their
employment assistance needs. It would also help service providers to ration more
expensive forms of employment assistance, since many job seekers will obtain
employment within the three-month period. 14
• Employment assistance guarantee
Those who have not left Intensive Assistance by the end of the gateway period should
be guaranteed employment assistance of a substantial nature to overcome their
workforce barriers. This should extend well beyond low level assistance such as
personal advice, coaching and job search assistance.
Personal advisers within the Intensive Assistance service should decide the form of
assistance to be offered to each client under the guarantee, in consultation with them.
This is a further development of the Intensive Assistance Support Plans already in
place. This is a more flexible approach than referring people to programs or
introducing supplementary funding such as the recently-proposed training credits that
can only be used for a fixed purpose.
•
Employment assistance guarantee funding pool
A separate pool of funds should be established to assist providers to meet their
obligations under the guarantee. In effect, the employment assistance guarantee would
be a government guarantee. This is very important to ensure that Governments are not
able to side step their employment assistance obligations to long-term unemployed
people by passing the buck to service providers. Each provider would be accountable
to spend any funds drawn from this pool to provide the services they offer their clients
under the guarantee. This is a further development of the present declarations of
intent. The existing system of up-front and outcome based payments would remain in
place.
The implementation of the guarantee, together with regulation of the employment
assistance market, the setting of basic service standards and the evaluation of
outcomes, would be over-sighted by an independent statutory body (separate from the
funding body).
•
Abolish the Work for the Dole program
The $200 million per annum that would be saved in this way could be redirected to
the Employment Assistance Guarantee funding pool, although more than this amount
would be needed.
14
This is the case in the United Kingdom, where a similar system (also called the
gateway) operates. A large proportion of clients leave intensive assistance before the
end of this period. See Finn (2001).
123
PETER DAVIDSON
9
The Expenditure Commitment to Employment Assistance
Unemployment and long-term unemployment will not be reduced in Australia unless
governments are prepared to make a sustained investment in employment assistance
for long-term unemployed people.
Over time, the employment assistance guarantee should extend to other jobless people
who want a job and face severe labour market disadvantage, but whose unemployment
is hidden. This includes sole and married parents who are on Parenting Payment and
people who are on Disability Support Pension, for prolonged periods. There are long
queues for the voluntary employment assistance schemes available to these groups.
Although the up-front cost of such investment may be high, this must be weighed up
against the social and fiscal cost of doing nothing. While other countries expanded
their employment assistance effort in the 1990s, Australia dropped the employment
assistance ball in the middle of the decade (see Figure 5) and we will pay a high price
for this in the future.
Figure 5: Expenditure on Employment Assistance (percentage of GDP)
0.7
0.6
0.5
0.4
0.3
1993/94
1994/95
1995/96
1996/97
1997/98
1998/99
Source: Federal Budget Papers
References
ACOSS (2000), Is the Job Network Working?, ACOSS Paper No. 108, Australian
Council of Social Services, Sydney.
ACOSS (2001), Does Work for the Dole Lead to Work for Wages?, ACOSS Info
Paper No. 223, Australian Council of Social Services, Sydney.
124
1999/00
Centrelink (various years), ’Labour market and related payments’, Centrelink,
Canberra.
Department of Employment, Education and Training (1994-95), Annual Reports,
AGPS, Canberra.
Department of Employment, Education, Training and Youth Affairs (DEETYA)
(1997), The Net Impact of Labour Market Programs, AGPS, Canberra.
Department of Employment, Workplace Relations and Small Business (2000a), Job
Network Evaluation, Stage One: implementation and market development,
EPPB Report 1/2000, Evaluation and Program Performance Branch, DEWRSB,
Canberra.
Department of Employment, Workplace Relations and Small Business (DEWRSB)
(2000b), Employment Assistance Outcomes, December 2000, DEWRSB,
Canberra.
(2001a), Net Impact of Work for the Dole, DEWRSB, Canberra.
Department of Employment, Workplace Relations and Small Business (2001b), Job
Network Evaluation, Stage 2: progress report, EPPB Report 2/2001, Evaluation
and Program Performance Branch, DEWRSB, Canberra.
(2001c), Labour Market Assistance Outcomes, June 2001, DEWRSB, Canberra.
Employment Services Regulatory Authority (ESRA) (1994-96), Annual Reports,
ESRA, Melbourne.
Finn, D. (2001), ‘Britain's work-based welfare state - the impact of the New Deals'
(unpublished paper), University of Portsmouth
Warburton, M., A. Opoku and L. Vuong (1999) 'Long-term unemployment: a
statistical analysis of FaCS customers', Australian Social Policy, 1999 (2), 3352.
125
The Social Contract Renegotiated:
Protecting Public Values in the Age of
Contracting
David de Carvalho
In some ways this paper can be seen as a response to the call to analytical action made by
Anna Yeatman in the most recent edition of the Australian Journal of Public
Administration.
Now is the time to discuss and debate different conceptions of
contracting out…If the economic model of contracting out is
now under sustained critical scrutiny, the more urgent task is to
work on institutional design for publicly oriented and
democratic contracting out. (2001: 73)
My purpose therefore is to suggest, tentatively, a conceptual framework for, and a
practical approach to, the development of a new type of contractual arrangement between
governments and civil society organisations, including religious and other value-based
bodies, involved in the delivery of publicly funded social services.
First I need to make a pedantic point. The term ‘contracting out’ is often used when
simply ’contracting’ is more accurate. It is not uncommon to read commentators
describing government contracts as ’legal agreement[s] to regulate the private provision of
government services‘ (Bridgman and Davis, 2000: 19). However, the contractualism that
characterises the new public management movement involves not only the contracting out
of services previously provided by government, but also the transformation of funding
arrangements for social services that had been previously and continuously provided by
the non-government sector, whereby grants to assist with input costs were transformed to
performance-based contracts through which governments purchased specified outputs.
In 1995, the then Secretary of the Victorian Department of Health and Community
Services, John Paterson, described this transformation in the department’s annual report.
He said,
In Victoria, we have moved from a situation in which
government funded providers to do what they want, to a
situation in which government funds them to do what
government wants. But this is still a long way from the ultimate
De Carvalho, D. (2002), ‘The social contract renegotiated: protecting public values in the age of
contracting’, in T. Eardley and B. Bradbury, eds, Competing Visions: Refereed Proceedings of National
Social Policy Conference 2001, SPRC Report 1/02, Social Policy Research Centre, University of New South
Wales, Sydney, 126-135.
DAVID de CARVALHO
goal, where providers will compete to do what consumers
want. (Paterson, 1995)
Thus the take-up of contracts by non-government (including church) agencies does not
necessarily represent any particular enthusiasm for market mechanisms, but simply
reflects the historical fact that non-government social service agencies previously funded
by grants must now be funded by contract or not at all.
Contract is a political term, and the term social contract is self-consciously political. In his
1651 work, Leviathan, Thomas Hobbes wrote that the life of the individual, without the
social order imposed by the state, is ’solitary, poor, nasty, brutish, and short’. He, along
with Locke and Rousseau, developed the basic idea of the social contract as an agreement
between individuals, by which they give to the state power to enforce the rule of law so
that they can, to the greatest extent possible, exercise their liberty in the pursuit of their
own interests, in particular through the operations of the free market.
In the twentieth century, the Keynesian Welfare State was adopted by most Western
democracies as a means of smoothing out the inequalities generated by the free market.
Marshall’s concept of social citizenship and social rights underpinned that model of state
welfare.
Figure 1: The Individual and Society
L eg a l F r a m e w o r k
Citizenship
State Power
Public
administrative
law
Indivi
dual
Private contract
law
Market Forces
127
Social
Citizenship:
Social rights to
publicly
provided goods
and services
Market
Citizenship:
Consumer
rights and
market power
THE SOCIAL CONTRACT RENEGOTIATED: PROTECTING PUBLIC VALUES
The legal framework governing the relationship between individuals and the state is public
law, specifically administrative law, which aims to ensure fairness, accountability and
transparency in government decisions affecting individuals. The legal framework
governing the relationship of the individual with the market is private law, specifically the
law of contract. It is less concerned with fairness, accountability and transparency, than
with simply ensuring that market transactions are fulfilled according to the terms agreed
by the contracting parties.
In schema illustrated in Figure 1, the individual’s engagement with society is conceived as
a set of constant interactions with the market and the state. Something is missing from
this diagram: those non-state, non-market organisations to which individuals belong
voluntarily, such as unions, churches, co-operatives, sporting clubs, ethnic communities
and families, from which individuals derive their sense of personal identity. Figure 2
portrays this richer set of relationships.
Figure 2: The Individual - in Community and Society
State Power
Burke’s ‘little
platoons’
Voluntary Associations
Unions
‘Social capital’
builders
‘Buffers’ between
the individual on
the one hand and
the state and the
market on the
other
Churches
I
I
I
I
I
Cooperatives
I
Families
I
Ethnic communities
Sporting Clubs
Market Forces
Edmund Burke referred to these organisations as little platoons. They play a key role in
the development of social capital. They can be seen as buffers between the individual on
the one hand and both the state and the market on the other. Individuals who are members
128
DAVID de CARVALHO
of such community or civil society organisations can be seen as having multiple
citizenships and hence multiple accountabilities in addition to their national citizenship.
However, the nature and role of those community organisations that provide social
services have been contested as a result of pro-market reforms in the public sector over the
last fifteen years.
Figure 3: ‘New’ Public Management Economic Model of Contracting
Arms of the
State and/or
commercial
businesses
State Purchaser Power
Private
Non-state
provider
Contracts
Non-state
provider
Non-state
provider
Mediators of
state power
and market
forces
The State
(principal)
steers
while the
community
sector
(agent)
rows
Individual Consumer Power
Their traditional role and character has come under pressure from both the state and the
market as they have been squeezed to become more like government agents on the one
hand and more like commercial businesses on the other (Figure 3). Instead of acting as a
buffer between the individual and the power of the state or market forces, some of these
organisations are in danger of losing their character as they become mediators of that
power and those forces.
For example, the Job Network not only is regulated by legal contracts between the state
and non-state providers, but these legal contracts give expression to the current
government’s conservative communitarian notion of the social contract as mutual
obligation. This takes the form of a contract between the individual and the state whereby
the state provides income support in return for the individual fulfilling certain personal
responsibilities in respect of job search. The content of these responsibilities and the terms
of this social contract are defined by the government, acting on behalf of the community.
129
Through the Job Network, government attempts to enlist the credentials of community
organisations to its particular social agenda.
But these organisations do not always act out the role assigned to them. While their
contracts require them to report breaches of mutual obligation requirements, many nonprofit Job Network providers, including some auspiced by church-based organisations
with long histories of local involvement and a strong tradition of social justice advocacy,
have additional accountabilities. They have an accountability to their value-based mission,
which in many cases means they are willing to put more time into the 'hard cases' - those
who are seriously marginalised from the labour market.
In practice, this means that while their funding contract gives them no official discretion
about whether or not to report breaches or possible breaches, they do exercise a de facto
discretion of this kind. They may make decisions about whether to report breaches or not,
and in doing so, they are acting as a buffer between that individual and the power of the
state. Such agencies have to balance their different accountabilities – contractual
obligations to the government and moral obligations to their value-based mission. These
accountabilities do not always pull them in the same direction. Their contract may say
‘breach em’ but their value base may say ‘don’t’.
From the perspective of the public interest and the public sector, as opposed to the
community sector, the economic model of contracting out has additional difficulties.
Financial information that was and should be in the public domain becomes secret as the
result of ’commercial-in-confidence‘, and access to the remedies provided by public
administrative law for the citizen who is provided with services becomes more difficult.
As the Administrative Review Council noted in its 1998 report on contracting out
The delivery of government services by contractors, and the
consequent ‘privatising’ of the relationship between service
providers and members of the public, has the potential to result in a
loss of the benefits which the administrative law system provides
for individuals. In turn, this may affect the efficiency and quality
of government administration. Further, since a contractor’s
connection with government will be governed by contract, the
accountability mechanisms traditionally provided by ministerial
responsibility and Parliamentary oversight may no longer be as
effective. (ARC, 1998: vi)
First, there is a loss in the ability of individual citizens and the Parliament to hold
government accountable through administrative law for the implementation of public
policy. Secondly, there is a loss in the ability of community organisations to be responsive
and accountable to their own ethos.
Can we develop a form of contract that enhances both the ability of the public and
Parliament to hold governments accountable for contracted services and the ability of civil
society organisations to be faithful to their own ethos and accountable to their own
mission? Or are these two goals mutually exclusive?
130
DAVID de CARVALHO
We need to re-conceptualise the nature of the contractual relationship between the state
and community organisations, from one based on private contract to one which recognises
the mixed nature of the activity as having both private and public elements, and which
recognises the dual accountabilities faced by the state’s community sector partners. I
Figure 4 shows how the economic contract regime has created a double barrier.
Figure 4: The Economic Contract as a Double Barrier
G overnm ent
C o m m u n ity
agency
P u b l i c a c c o u n t a b i l ity
P r iv a t e a c c o u n t a b i lit y
C
O
N
T
R
A
C
T
C itiz e n s a n d
P a r lia m e n t
V a lu e s a n d
m is s io n
suggest that we should conceive of these contracts as social contracts. These social
contracts would incorporate a higher degree of public accountability than currently exists
in many government contracts, but also more flexibility to allow the contracted
organisation freedom to be itself in the contract, that is, to be faithful to its own value
base.
This is consistent with developments in social policy theory and in particular the emerging
discourse of governance as opposed to the discourse of new public management, and it is
also consistent with new understandings of citizenship and shared sovereignty.
Anna Yeatman (2000: 63) has written about shared sovereignty in the following way:
The conception of modern democratic citizenship in terms of
the one and indivisible nation has increasingly lost legitimacy
for a number of reasons. First, the claim of indigenous people
to self-determination usually has to be made in a context where
it is a question of how an established state accommodates this
claim. The only way of working with both indigenous selfdetermination and state authority is to suggest that sovereignty
can be divisible. Second, there have been a number of social
movements that have insisted on the heterogeneity as distinct
from the ‘oneness’ of the nation. Multicultural movements, for
example, have demonstrated the many differences of ethnicity,
131
culture and history that are caught up within the jurisdiction of
any particular state.
I would add that differences in religious tradition and heritage are another example of the
way in which communal citizenships overlap with national citizenship and invite
consideration of what shared sovereignty means.
This notion of shared sovereignty helps us distinguish between new public managerialism
and governance. New public management is about making things happen through state
power via tightly specified contracts, whereas governance is about managing relationships
and a diversity of approaches to a common problem. The distinction can be stated another
way – it is the difference between power and influence. Rhodes (1997: 43) argues that in
governance-style relationships, ’although the state does not occupy a privileged, sovereign
position, it can indirectly and imperfectly steer networks.’
Taking the steering/rowing metaphor a step further, we should shed the model in which in
which the state steers while the community sector rows, and adopt a model in which the
state acts like a flagship in a sea-battle, this is shown in Figure 5.
Figure 5: Democratic Communitarianism – A Social Model of Contracting
Democratic Communitarianism – A Social Model of Contracting
Voluntary
agencies
participate
in policy
development and
coordination;
collaborate
rather than
compete with
one another
Empowering State
(Flagship)
Social
Contracts
C
Governments
influence
heterogeneous
agencies through
engagement and
empowerment,
rather than
control
through
contract
Governments, Voluntary agencies and their clients co -produce improved social outcomes
Agreed Social Goals / Diversity of Approach
132
DAVID de CARVALHO
The flagship communicates direction to and coordinates the activities of a fleet of ships as
best it can, while recognising that each individual ship has its own capabilities and each
captain has their own particular personality, character, attitudes and approaches. Each ship
has the freedom to move around its section of the battle area in order to meet the particular
circumstances it is confronting. That is, civil society organisations, each with their own
sets of values and interests work with government to achieve commonly agreed social
outcomes. In acting as a flagship rather than a steerer, governments recognise that their
social partners have other accountabilities in addition to their role as a government
contractor.
Is the legal contract the appropriate form for governance-style, sovereignty-sharing
relationships? Yeatman argues that it can be.
Contract, like all terms of political discourse is a contestable
concept and it can be reframed to become a dialogic
contractualism where the contractual partnership between
government and non-government service-providers is open, as
appropriate, to constitutional conversation with key
stakeholders. (Yeatman, 2000: 69)
So rather than private economic contracts that reduce both accountability to the public
interest and accountability to an organisation’s value-based mission, we need a new kind
of contract and a new kind of law to regulate it. Social contracts would share sovereignty
and acknowledge both public and private accountabilities. Social contracts would see the
’state engaging and supporting heterogeneous associations in a way that is closer to
democratic and identity politics’ (Butcher and Mullard, 1993). To return to Anna
Yeatman’s call to focus our attention on institutional design for publicly oriented and
democratic contracting out, an initial practical step could be taken in this direction by any
Australian government that has contracts with community sector providers. That would be
to initiate a consultation with a view to establishing a social compact or an accord
between the government and the community sector on how they will collaborate in the
development of social policy and the delivery of social outcomes.
We do not have to reinvent the wheel here. In Canada, the government and the sector are
working towards the development of an Accord and in Britain there is the Compact on
Relations between Government and the Voluntary and Community Sector in England
which provides an example of the kind of thinking that is required. Presented to
Parliament in November 1998, the Compact
provides a framework which will help guide our relationship at
every level. It recognises that Government and the sector fulfil
complementary roles in the development and delivery of public
policy and services, and that the Government has a role in
promoting voluntary and community activity in all areas of our
national life. (Blair, 1998)
133
The Compact applies to government departments and to the range of organisations in the
voluntary and community sector, and sets out a number of principles and undertakings on
both sides. While it is not a legally binding document, it does provide the basis from
which a legal framework might evolve.
The next steps in that evolution could be the passage of legislation, (say, a Social
Contracts Act), that confers special legal status on contracts between government and
community sector providers of publicly funded social services. These contracts would
have the status of social contracts. Such contracts might have at least the following
characteristics:
•
the state could not avoid vicarious liability, by virtue of privity of contract, for
failure of service-delivery by the contractor;
•
‘commercial in confidence‘ could not be invoked to exempt social contracts from
Freedom of Information requests;
•
social contractors would not be constrained from making public comment on any
aspect of government policy;
•
non-government contractors would be required to have high standards of
governance and conduct.1
If this proposal or something like it was implemented, a jurisprudential tradition would
develop over time which would recognise and value the special nature of such social
partnerships.
The regulation of social contracts through the Social Contracts Act would require a Social
Tribunal to provide a forum for the adjudication, at first instance, with a right of judicial
review, of disputes about the non-fulfilment of social contracts.
So the elements of legal institutional design for publicly oriented and democratic
contracting might be as follows:
•
a Compact between Government and Community Sector;Social Contracts Act
which confers special legal status on contracts between government and non-state
organisations delivering publicly funded social services; a Social Tribunal which
adjudicates disputes about non-fulfillment of social contracts, at which serviceusers would have ’standing‘; and aJudicial Review of tribunal decisions’
To conclude, the rapid multiplication and diversification of private-public partnerships in
recent years has seriously tested the ability of the traditional public-private legal
dichotomy to balance issues of public accountability on the one hand and the
organisational autonomy of social partners on the other. The recognition that there is a
public element about such partnerships is not enough to justify the full scrutiny of public
1
This list is not exhaustive. The British Compact document provides a list of 'undertakings' by both
parties, which might be incorporated into such contracts.
134
DAVID de CARVALHO
administrative law to the decisions and actions taken by community organisations; but
neither is the recognition that there is also a private, non-public element to these
partnerships enough to justify the complete removal of public law scrutiny. Replacing the
economic model of contracting with a social model of contracting would enhance both
public accountability and the ability of community organisations to make their distinctive
contribution to a democratic, pluralist society.
References
Administrative Review Council (1998), The Contracting Out of Government Services,
Report No.42, ARC, Canberra.
Blair, A (1998), 'A message from the Prime Minister', Compact: Getting it Right Together
– Compact on Relations between Government and the Voluntary and Community
Sector in England, National Council of Voluntary Organisations, London,
http://www.ncvo-vol.org.uk/main/gateway/compact.html
Bridgman, P. and G. Davis (2000), ’Implementation‘, The Australian Policy Handbook,
Allen and Unwin, Sydney.
Butcher, H. and M. Mullard (1993), ’Community policy, citizenship and democracy‘, in
Butcher, H. et al., eds, Community and Public Policy, Pluto Press, London, cited
by Paul Smyth and Michael Wearing (2001) in ’After the welfare state: state and
civil society in the delivery of social services‘, unpublished paper presented at
Institutional Dynamics of Australian Economic Governance Conference, Brisbane
February 2001, University of Queensland.
Paterson, J. (1995), ’Foreword to the 1994-95 Annual Report of the Victorian Department
of Health and Community Services‘ Momentum 3 (11), 4.
Rhodes, R. (1997), 'From marketisation to diplomacy: it’s the mix that matters', Australian
Journal of Public Administration, 56 (2), 40-54.
Yeatman, A. (2000), ’Social policy and the global economy‘, in I. O’Connor et al, eds,
Contemporary Perspectives on Social Work and the Human Services, Longman,
Sydney, 60-72.
Yeatman, A. (2001), 'Contracting out and Public Values', Australian Journal of Public
Administration, 60 (2), 71-73.
135
Socioeconomic Banditry: Poker Machines
and Income Redistribution in Victoria
James Doughney
Victoria University
1
Introduction
Victoria introduced poker machine gambling in 1992. There were ostensible reasons,
and there were reasons. Underlying the policy shift was the then state Labor
government’s desire to widen its own taxation base. Underlying this were the
Commonwealth’s fiscal restraints on the states, which had become more severe during
the 1980s. Also important were the increasing demands from the state opposition and
‘the markets’, an entity at once diffuse and yet omnipotent, for balanced state budgets
and reduced borrowing and debt. Underlying all of these reasons had been the victory
of the markets’ ideological expression, neo-liberalism, at least in the English speaking
world.
By an unpleasant, though predictable, twist of fate the new pokies were ‘rolled out’
alongside the new Liberal-National Kennett Government. Victoria would be
transported from the 1990-92 recession by a gambling led recovery, we were told.
What the new Crown Casino would do for the city, poker machines would do for the
suburbs and regional Victoria. There would be 30 000 machines: 2500 at Crown, a
maximum of 80 per cent of the remaining 27 500 in metropolitan Melbourne and an
even share distributed between pubs and ‘clubs’. A ‘duopoly’, Tattersall’s and
Tabcorp, would operate the machines. The ‘industry’ would be regulated by the
Victorian Casino and Gaming Authority, which would also have, thanks to the
Kennett Government, the express mission to promote gambling in and for Victoria.
The game was afoot: everyone would be a winner!
Of course, the ‘everyone’s a winner’ phrase, used in advertising by one or another of
the ‘industry players’, is nonsense. At best, when there is no ‘house’ and no
transaction costs, some win and others lose. It is a net zero sum game. When there is a
gambling industry the calculation is just a bit more complicated: some win, some lose
and losses provide the industry its income. Such net losses, or ‘expenditure’ as it is
known, in turn become the profits of the owners and operators, the wages of their staff
and income for their suppliers. Net losses also give the state government significant
Doughney, J. (2002), ‘Socioeconomic banditry: poker machines and income
redistribution in Victoria’, T. Eardley and B. Bradbury, eds, Competing Visions:
Refereed Proceedings of the National Social Policy Conference 2001, SPRC Report
1/02, Social Policy Research Centre, University of New South Wales, Sydney, 136154.
JAMES DOUGHNEY
taxation revenue. In Victoria this is set at 33 and one-third per cent of every net dollar
lost in clubs and 41 and two-thirds of every net dollar lost in pubs.1
Since 1992, net losses (henceforth losses) have grown enormously, both in their own
right and as a proportion of total gambling losses. The state government has also
become much more dependent on gambling taxes as a result. It is worth remarking
also that gambling losses overall have grown dramatically Australia wide. In the
financial year 1999-2000 gamblers lost a total of $13 341 554 000 000, or $13.34
billion, to the ‘industry’ in Australia. Of this $7.65 billion, or 57.37 per cent, was on
poker machines. Table 1 presents the raw gambling data for 1999-2000 of each type
of gambling in each state and Australia. Table 2 presents percentages for each type of
gambling to the state total and the Australian total. To put this into perspective,
$13.34 billion represents 3.5 per cent of total Australian household disposable
income.2
Figure 1 compares the trends in real gambling losses for Australia, NSW and Victoria
from the mid-1970s. Figure 2 presents this comparison in per capita terms.3 Both the
bars and the lines in Figure 1 are ordered: Australia, NSW and Victoria. The order
differs in Figure 2, which shows per capita losses. In this figure the broken lines
represent total gambling losses per capita for NSW, Victoria and Australia, while the
solid lines represent the poker machine gambling losses per capita of each. The
heavier lines show Victoria’s data. Both figures depict the acceleration in gambling
that occurred from the late 1980s onwards and the contribution of poker machine
gambling to it. The sharp increases in Victoria, especially in real per capita terms, and
their contribution to the national trends are also evident. Figure 3 presents the
proportionate contribution of poker machine gambling losses to the sum of all
gambling losses in NSW, Victoria and Australia. In each it is a growing proportion,
but in Victoria and Australia as a whole it has grown more because of legalisation in
the former and the spread of poker machines in Queensland and South Australia in the
1990s.
1
2
3
Hotels are taxed an additional eight and one-third per cent to contribute to the Community
Support Fund. Disbursement of this fund’s revenues has been an issue of concern, but that is a
subject for another paper.
All data in this paragraph are from Australian Gambling Statistics 1974-75 to 1999-2000
(Tasmanian Gaming Commission (TGC), 2001).
Real data are used in these charts. That is, the data are deflated to remove the effects of changes
in the general price level (inflation). This means that the charts focus exclusively on changes in
the magnitude of gambling. All data are again from TGC (2001).
138
.Table 1: Net Australian Gambling Losses by State and Type: 1999-2000 ($million)
NSW
Victoria
Queensland
SA
WA
Tasmania
ACT
NT
Australia
Racing
654.420
458.483
250.900
106.555
159.486
28.285
19.304
22.465
1,699.898
Lottery
42.067
5.185
1.320
-
-
0.347
0.925
0.782
50.626
286.176
283.345
205.560
72.588
148.736
16.305
12.492
11.722
1,036.924
66.418
25.614
101.051
10.165
33.425
2.368
2.469
1.517
243.027
4.151
1.101
1.807
0.282
0.779
0.078
0.096
0.031
8.325
486.400
823.869
530.700
75.831
288.560
77.070
17.700
62.385
2362.515
-
-
-
22.758
24.312
9.017
-
-
56.087
91.450
6.783
54.160
15.611
-
15.176
-
-
183.180
Gaming machines
3882.199
2170.560
871.303
485.987
-
60.773
156.835
26.474
7654.131
Interactive gaming
-
-
-
-
-
-
-
5.387
5.387
12.397
7.881
2.300
0.767
1.229
0.190
0.000
16.689
41.453
5525.678
3782.821
2019.101
790.544
656.527
209.609
209.821
147.452
13341.554
Lotto
Instant lottery
Pools
Casino
Minor gaming
Keno
Sports betting
Total gambling
Source: Australian Gambling Statistics 1974-75 to 1999-2000 (Tasmanian Gaming Commission, 2001)
Table 2: Net Australian Gambling Losses by State and Type: 1999-2000 (% of column totals)
NSW
Victoria
Queensland
SA
WA
Tasmania
ACT
NT
Australia
Racing
11.84
12.12
12.43
13.48
24.29
13.49
9.20
15.24
12.74
Lottery
0.76
0.14
0.07
0.00
0.00
0.17
0.44
0.53
0.38
Lotto
5.18
7.49
10.18
9.18
22.65
7.78
5.95
7.95
7.77
Instant lottery
1.20
0.68
5.00
1.29
5.09
1.13
1.18
1.03
1.82
Pools
0.08
0.03
0.09
0.04
0.12
0.04
0.05
0.02
0.06
Casino
8.80
21.78
26.28
9.59
43.95
36.77
8.44
42.31
17.71
Minor gaming
0.00
0.00
0.00
2.88
3.70
4.30
0.00
0.00
0.42
Keno
1.66
0.18
2.68
1.97
0.00
7.24
0.00
0.00
1.37
Gaming machines
70.26
57.38
43.15
61.48
0.00
28.99
74.75
17.95
57.37
Interactive gaming
0.00
0.00
0.00
0.00
0.00
0.00
0.00
3.65
0.04
Sports betting
0.22
0.21
0.11
0.10
0.19
0.09
0.00
11.32
0.31
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
Total gambling
Source: Australian Gambling Statistics 1974-75 to 1999-2000 (Tasmanian Gaming Commission, 2001)
Figure 1: Real Gambling and Poker Machine Losses: Australia, NSW and Victoria: 1974-75 to 1999-2000
16,000
14,000
12,000
8,000
6,000
4,000
2,000
4
19 -75
75
19 -76
76
19 -77
77
19 -78
78
19 -79
79
19 -80
80
19 -81
81
19 -82
82
19 -83
83
19 -84
84
19 -85
85
19 -86
86
19 -87
87
19 -88
88
19 -89
89
19 -90
90
19 -91
91
19 -92
92
19 -93
93
19 -94
94
19 -95
95
19 -96
96
19 -97
97
19 -98
98
19 -99
99
-0
0
0
19
7
$ MILLION
10,000
Lines: Real net poker machine losses
Bars: Real net total gambling losses
(In order of size, highest to lowest: Australia,
NSW, Victoria)
YEARS
Figure 2: Real per Capita Gambling and Poker Machine Losses, Australia, NSW and Victoria: 1974-75 to 1999-2000
1,200
1,000
800
600
400
NSW poker machine
Australia total gambling
Victoria total gambling
200
Victoria poker machine
Australia poker machine
0
19
74
19 -75
75
19 -76
76
19 -77
77
19 -78
78
19 -79
79
19 -80
80
19 -81
81
19 -82
82
19 -83
83
19 -84
84
19 -85
85
19 -86
86
19 -87
87
19 -88
88
19 -89
89
19 -90
90
19 -91
91
19 -92
92
19 -93
93
19 -94
94
19 -95
95
19 -96
96
19 -97
97
19 -98
98
19 -99
99
-0
0
$
NSW total gambling
YEARS
POKER MACHINES AND INCOME DISTRIBUTION IN VICTORIA
Governments are now correspondingly more dependent on poker machine taxes. The
factors presented in the first paragraph are the fundamental cause. All governments
have endeavoured to give business lower taxes. The Commonwealth squeezed the
states and, inter alia, introduced a GST. The states themselves competed to lower the
supposed ‘business tax burden’ and to reduce borrowing and repay debt. As the
community lost more on poker machines, correspondingly higher pokie tax revenues
were locked in to state budgets. Victoria’s accelerated integration into the pokie
partnership makes its dependency appear all the more stark. Indeed the combination
of a high rate of pokie tax and reduced state revenues from other sources means it is
now the state government most dependent on gambling taxes
Before moving on to discuss this result let us, first, restate what we already know
about poker machine losses and, second, introduce an apparently small but significant
finding from other recent research. Two important facts have already emerged from
independent research: machines are concentrated in municipalities that have a lower
than average socio-economic status; and average losses per adult are concentrated in
these municipalities. Figure 4 shows how losses are distributed across metropolitan
Melbourne’s local government areas (LGAs), which are ranked from lowest to highest
by socioeconomic status.4 The Productivity Commission referred extensively to the
relationship depicted in the figure in its Draft Report: Australia’s Gambling
Industries (PC, 1999). It noted:
It remains the case that, in Victoria at least, gaming machine
densities are higher in socially and/or economically
disadvantaged areas and that, in turn, this is likely to mean
that people in those areas spend [i.e. lose] more on gaming
machines than people in other areas ...
Whatever the reasons, where socially and economically
disadvantaged areas do have a high density of gaming
machines, there will be implications for the local community
...
Further, it is possible that, in communities that already suffer
from significant socio-economic disadvantage, overlaying an
additional source of socio-economic stress
4
Socioeconomic status here is defined by the Australian Bureau of Statistics (ABS)
socioeconomic indexes for areas (SEIFA) index of disadvantage (DIS). This index is based on a
number of variables related to the economic and social characteristics of families and
households, as well as personal education qualifications and occupation. High status areas
obtain high index numbers and vice versa. Rankings for the Melbourne metropolitan LGAs
range from Maribyrnong at 887.680 to Boroondara at 1133.77. The SEIFA(DIS) average for
Melbourne is 1024.839. It should be noted that we have defined Melbourne here and in all
subsequent data to exclude the City of Melbourne proper. This eliminates the distorting effects
of venues in the Central Business District. Charles Livingstone was the first to draw attention to
the association represented in Figure 1.
143
Table 3: Government Own Account Tax Revenues, Victoria, NSW and Australia: 1999-2000
$million
Victoria
Payroll taxes
Property taxes
Financial transactions
Stamp duties etc.
Gambling taxes
Lotteries
Poker machines
Casino
Racing
Other
Total gambling taxes
Insurance (fire)
Third party
Other insurance taxes
Motor vehicle taxes
Fuel taxes
Tobacco taxes
Liquor taxes
All other
Total
NSW
% of column total
Total all states
2,356
500
611
1,759
3,769
924
924
3,269
8,942
2,427
2,237
7,420
291
933
155
137
4
1,520
189
80
310
863
528
727
226
39
9,708
293
955
126
191
5
1,570
287
17
598
1,468
614
1,083
323
345
15,191
964
2,463
497
486
10
4,420
540
180
1,418
3,900
1,631
3,318
973
414
37,820
Source: Australian Bureau of Statistics 2001, Taxation Revenue (Cat. No. 5506.0), released 12 June 2001.
Victoria
24.27
5.15
6.29
18.12
0.00
3.00
9.61
1.60
1.41
0.04
15.66
1.95
0.82
3.19
8.89
5.44
7.49
2.33
0.40
100.00
NSW
Total all states
24.81
6.08
6.08
21.52
0.00
1.93
6.29
0.83
1.26
0.03
10.34
1.89
0.11
3.94
9.66
4.04
7.13
2.13
2.27
100.00
23.64
6.42
5.91
19.62
0.00
2.55
6.51
1.31
1.29
0.03
11.69
1.43
0.48
3.75
10.31
4.31
8.77
2.57
1.09
100.00
Figure 3: Poker Machine Losses as a Proportion of Total Gambling Losses: 1974-75 to 1999-2000
80
70
60
%
50
40
30
20
10
0
00
99
98
97
96
95
94
93
92
91
90
89
88
87
86
85
84
83
82
81
80
79
78
77
76
75
4- 75- 76- 77- 78- 79- 80- 81- 82- 83- 84- 85- 86- 87- 88- 89- 90- 91- 92- 93- 94- 95- 96- 97- 98- 997
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
YEARS
NSW
Victoria
Australia
Figure 4: Poker Machine Losses per Adult in Metropolitan LGAs Ranked by SEIFA Index of Socioeconomic
Disadvantage: 1998-99
25.00
Maribyrnong
20.00
$ Per Adult (18+) Per Week
Greater Dandenong
Whittlesea
Knox
Monash
Moonee Valley
Darebin
15.00
Wyndham
Brimbank
Maroondah
Hobsons Bay Frankston
Moreland
10.00
Glen Eira
Kingston
Casey
Hume
Banyule
Yarra
Manningham
Whitehorse
Port Phillip
Stonnington
Bayside
5.00
Boroondara
0.00
0
5
10
15
20
SEIFA Ranking (most disadvantaged to least disadvantaged)
25
30
JAMES DOUGHNEY
may have significant community-wide impacts. That is, social
and economic stresses may have compounding impacts ...
In the Commission’s view, the potential for disadvantaged
communities to suffer more adverse social problems from
expansions in gambling has important implications for
government policy. (PC, 1999, paras, 9.38-9.41)
A similar pattern exists in regional Victoria. Indeed three of the most disadvantaged
local government areas (LGAs), Bass Coast, La Trobe, and East Gippsland, have the
highest numbers of poker machines per adult. Not surprisingly Bass Coast and La
Trobe also have the highest losses per head of the major regional LGAs. Other aboveaverage losers are East Gippsland, Mildura, Greater Shepparton, Greater Geelong,
Ballarat, Greater Bendigo, and Warrnambool, all of which are in the lower half of the
SEIFA rankings for regional Victoria. Figure 5 shows the regional distribution of
losses, with the trend line once again clearly demonstrating that the less well off
regional areas bear the heaviest burden.
As is also true in metropolitan Melbourne, the better off areas have the lowest number
of machines per head and, consequently, the lowest average losses per head.
However, there is a difference. The average SEIFA ranking for Melbourne LGAs is
about 1025, whereas in regional Victoria the average ranking is much lower, at about
993 for the LGAs shown in Figure 5 (the larger regional LGAs with poker machines).
Only one of these, Macedon Ranges, has a higher ranking than the Melbourne average
(1061.82). It is hardly surprising that it is among the lowest three losers and has the
lowest number of machines per head. It has barely more than one-third of the losses
per head in La Trobe and Bass Coast. Moreover only the metropolitan cities of
Maribyrnong, Greater Dandenong, Darebin, Brimbank, and Moreland have lower
socio-economic rankings than La Trobe, Bass Coast, East Gippsland, Mildura, and
Hepburn. Of these only Hepburn has below average pokie losses for regional Victoria.
That the high-loss areas are those in which people can least afford to lose also violates
an important principle expressed by John Maynard Keynes many years ago. Keynes
emphasised that gambling was fine if it were ‘frivolous’. By this he meant that losses
should be small relative to a person’s budget. Keynes’s main objection to gambling
practices in the 1930s in Britain
… was that they led to losses of large amounts of money. Its
‘evil’, he said, lay not in the hope of obtaining unearned
money, but ‘in the fact that the indulgence of this hope will
cause you to lose a great deal of money’
Gambling always results in financial losses for the great
majority of gamblers, the only beneficiaries being the small
number of winners and the gambling industry. If the financial
losses to individuals are small and outweighed by the
enjoyment obtained, gamblers derive
147
Figure 5: Poker Machine Losses per Adult in Rural and Regional LGAs by SEIFA Index of Socioeconomic
Disadvantage: 1998-99
18.00
16.00
Bass Coast
La Trobe
14.00
Warrnambool
Ballarat
$ per Adult (18+) per week
Greater Geelong
12.00
East Gippsland
10.00
Greater Shepparton
Greater Bendigo
Mildura
Wellington
Horsham
Mitchell
Northern Grampians
8.00
Glenelg Swan Hill
Colac-Otway
Wodonga
Delatite Wangaratta
Baw Baw South Gippsland
6.00
Hepburn
Campaspe
Macedon Ranges
4.00
2.00
0.00
0
5
10
15
20
SEIFA Ranking(most disadvantaged to least disadvantaged)
25
30
JAMES DOUGHNEY
net benefits. If the losses are large, however, the situation is
reversed. The financial damage is considerable, psychic
income falls and probably becomes negative, and the gambler
faces a double misfortune. (O’Donnell, 1999: 13; citing
Keynes’s evidence to the 1932 Royal Commission on Lotteries
and Betting, Keynes, 1982: 404)
The second preliminary point that is important to introduce comes from the most
recent VCGA community impact study (KPMG, 2000). It supported what many
researchers into gambling have suspected: the seemingly trivial finding that most
people who use pokies do so close to where they live. In fact it said that most play
within 2.5 kilometres of their homes. While prima facie this might seem insignificant,
or even obvious, it does suggest that research should also focus on the immediate
vicinity of a venue and not only on the much larger LGA catchment. We will do this
below by analysing the socio-economic statuses of smaller areas surrounding all
Melbourne venues. In this way we test the hypothesis that poker machines are located
in areas of lower socio-economic status than had been demonstrated previously using
crude LGA-level indexes.
2
Analysis Using ABS Collection Districts
The Local Area Pokie Impact Software Tool (LAPIST) data allow researchers to
shrink the socio-demographic area of interest down to that of the Australian Bureau of
Statistics (ABS) census collection district (CCD or CD) in which a venue is located.
However, these districts contain only about 200 or so households. Therefore we will
look at a larger collection of CDs: the CD in which the venue is located plus the 10
closest CD numbers on either side (or the CD +/-10). LGAs in the Melbourne
metropolitan area, as defined here, have an average adult population of about 75 000
to 80 000. This means that we will be examining areas approximately between 10 and
15 per cent of an LGA.
Unfortunately it is not possible to obtain loss data for venues or for smaller
geographical units than a municipality. Even for a municipality itself we know only
how much is lost in total in its venues. To turn this into an estimate of how much is
lost by the LGA’s adult residents we have to assume that they gamble as much outside
the LGA as non-residents do within it. It is a reasonable assumption, especially in the
light of the community impact study finding that most people who use pokies do so
within 2.5 kilometres of their homes (KPMG, 2000). Therefore we will assume for
convenience here that the average loss prevailing in the municipality can be applied
uncontroversially to the smaller areas.5
When we make this assumption it is possible to create an additional best fit chart
similar to Figures 4 and 5. Figure 6 presents the relationship between losses per adult
per week and the weighted SEIFA indexes of disadvantage of the smaller areas
5
As it happens the assumption is conservative. If people gamble in venues that are closer to home
it is, therefore, reasonable to expect that losses in the areas closer to the venue would be higher
than average.
149
POKER MACHINES AND INCOME REDISTRIBUTION IN VICTORIA
surrounding all Melbourne poker machine venues. It might help to imagine these
areas as spreading out around the CD of the venue, with about 10 000 people in each.6
The line in Figure 6 exhibits the downward slope of the lines in Figures 4 and 5,
demonstrating that losses are concentrated on average in less well off areas. The
scatter plot of the CD +/-10 trend is also shown.7 To help to understand the
relationship between this and the LGA level data it is useful to compare averages or
means. The mean of the SEIFA index of disadvantage for the Melbourne metropolitan
area, including surrounding growth areas such as Mornington Peninsula Shire, is
approximately 1025. The mean for the CD +/-10 area is about 1005. The decline is
statistically significant.8 Were we to have used LGA-level averages for each venue the
mean would have been 1012. The decline is again statistically significant. Given that
we were forced to apply the LGA level loss per adult data to each venue, the loss
levels shown have not changed. What has changed with the specific CD level data is
the positioning of the scatter points. They have shifted leftwards by comparison with
those in Figure 4. It is as if the shift of scatter points (representing venues) to lower
socioeconomic positions, ‘clustering’ on average in and around the 1000 SIEFA mark,
has caused the CD +/- 10 line to be pulled leftward.
This means that, on average, losses are higher in areas of lower socio-economic status
than of Melbourne as a whole and they are located also, on average, in areas of lower
socio-economic status than of their municipalities. However, it should be repeated that
this section so far has used only a ‘first approximation’ approach to identify the socioeconomic status of areas around poker machine venues in Melbourne and surrounding
growth areas. It is approximate because the surrounding collection districts were
chosen for their numerical proximity rather than their immediate geography.
A closer analysis is possible if we use ABS maps and identify the best-fit CDs within
a certain radius of a venue.9 This, however, is labour intensive, so a brief analysis of
poker machine numbers in the City of Yarra only is given here as an example. A one
kilometre diameter circle around each venue was used in this case study because of
the relatively high density residential structure of the area. The average population in
the two kilometre diameter areas is about 3125 people. In total they account for about.
6
7
8
9
T he actual areas designated by numerical proximity to the CD number are rather
ragged at the edges and may be regarded more as being a reasonable first attempt. A
closer geographic proximity is obtained by using maps, as will be explained regarding
the City of Yarra case study below.
Note that the R-squared figures for these lines are low. It should be kept in mind that
the broad trends interest us here and that the variance about these trends would be
expected, even intuitively, to be high. This points to the need for closer and more
specific local area analysis. See footnote 6.
Using a single-sample t-test against the Melbourne average with both 95 per cent and
99 per cent confidence intervals.
It should be noted also that this ‘second approximation’ approach is also only an
estimate of the venue’s immediate catchment. If, by the latter, we mean the natural
location of users then a multiplicity of factors (roads, traffic flows, train lines, paths,
proximity to other venues etc.) needs to be assessed.
150
JAMES DOUGHNEY
Figure 6: Poker Machine Losses per Adult by SEIFA Index of Socioeconomic
Disadvantage: 1998-99 (at ABS CD +/- 10 level)
25
Venue CD +/- 10 average (approx.
Melbourne SEIFA average (approx. 1025)
$ Loss per Adult per Year
20
15
10
5
0
800
900
1000
1100
1200
SEIFA (most to least disadvantaged)
half of Yarra’s population. The analysis illustrates the general proposition above that machine
locations target the poorer parts of the community
Yarra is a variegated inner city area comprising the suburbs of North Carlton, Fitzroy, Clifton
Hill, Abbotsford, Collingwood, Richmond and parts of Northcote. It contains both high
density housing estates and gentrified pockets of affluence. There are ten venues in Yarra: the
Albion Inn in Collingwood, the Baker’s Arms in Abbotsford, the Park View in North Fitzroy,
the Prince of Wales, Royal Oak, Vine, Vaucluse and Richmond Tavern in Richmond, the
Tankerville Arms in Fitzroy and the Collingwood Football Club at Victoria Park. The average
SEIFA ranking for the areas around these venues is 946, compared with that for Yarra as a
whole of 984.
A drop of 38 in SEIFA ranking from the LGA average of 984 to 946 is quite dramatic. Recall
also that the Melbourne metropolitan ranking is about 1025. Even more worrying is that four
of the venue areas with the lowest SEIFA rankings (the Albion with 912, the Baker’s Arms
with 725, the Prince of Wales with 769 and the Richmond Tavern with 774) also have four of
the highest populations within their one kilometre diameter circles (4084, 5320, 6532 and
3845). All are located near high density public housing estates. This is worrying first because
most people gamble close to home. It is unlikely that very large numbers of people travel
from outside Yarra or even from too far away within it to gamble at these venues. Second, as
Tabcorp representative Tricia Wunsch (1998) said in evidence to the Productivity
Commission gambling enquiry, the ‘industry’ regards poker machine gambling as being a
‘blue collar’ form of recreation.
151
Put all of these factors together and we can obtain a clearer picture socio-economically of who
lost the sum of more than $33 million, or about $555 per adult, in Yarra’s pokies in 19992000. That represented a 12.7 per cent increase over the previous year. Moreover it is
generally estimated that only about 40 per cent of adults use poker machines in any year. The
‘industry’ itself acknowledged, again in evidence to the Productivity Commission enquiry,
that 80 per cent of its revenues (net losses) come from 20 per cent of pokie gamblers. Thus it
is possible that heavy gamblers in Yarra might be losing as much as 10 times the average of
$555 per adult, or $5500 each in 1999-2000. More worrying still is that Yarra is a mid-range
loser among LGAs. Maribyrnong has more than double Yarra’s level of losses per adult.
Greater Dandenong, Brimbank, Hume and Darebin, for example, are other metropolitan
municipalities with high levels of disadvantage with higher levels of loss than Yarra’s. A
number of poor rural and regional LGAs are also in a similar position.
3
Victoria’s ‘Reforms’: Testimony to the Legacy of Neo-liberalism
The Victorian Government has recognised formally that socio-economic status is important in
poker machine gambling policy. Indeed, the best part of its new policy on ‘regional capping’
of poker machine numbers is the principle that the level of losses, numbers of machines and
socio-economic status must be considered in determining numbers. Despite this the policy
suffered dramatically in implementation: only five areas will be ‘capped’, and the cap will
only reduce the numbers of machines in these areas to the level in the 9th highest municipality
in the state. This level is 11 machines per 1000 people. There are 78 municipalities in
Victoria. Only 2.5 per cent or so of Victoria’s total non-casino poker machines will be
affected.
There are strong reasons to believe that, even in the capped areas, the policy will do nothing
to alleviate gambling problems and related socio-economic harm to communities. First it
applies to only five of the State’s heavy gambling areas and to a relatively small number of
machines in them. Indeed, the drawing of the areas around the city LGAs to be capped –
‘Maribyrnong plus’, ‘Greater Dandenong plus’ and ‘Darebin plus’10 – prima facie seemed to
be a harm minimisation policy for the pubs, clubs and the duopoly operators. ‘Darebin plus’
was rewarded with a ‘zero’ reduction. Second, the capping policy will be phased in over a
three year period. Third, Tattersall’s and Tabcorp may succeed in shifting the machines to
other low-income parts of the state outside the five designated areas. Fourth, they may reduce
machine numbers in smaller and lower performing venues, rather than bigger venues where
people lose more per machine, and try to shift the remaining machines in the areas towards
bigger, high-take venues. Fifth, they may work machines more intensively and/or reduce the
return to gamblers towards the current legislated minimum of 87 per cent.11
10
11
The non-metropolitan LGAs of Bass Coast and Latrobe are the remaining members of the
five.
That is a minimum of 87 per cent of turnover in Victoria must be returned to players on average.
Currently more is actually returned, and the gambling operators may have about 3-4 per cent leeway to
cushion the effect of the caps if they choose. It is always an option, of course, to opt for a perverse form
of utilitarianism: the greatest good for the greatest number of machines. Why tempt action against the
97.5 per cent of non-casino machines untouched by the caps by pushing the capped machines to the limit?
Of course, greed is its own reward: why worry about changes in government policy when it is clear that
the state is as hooked on pokie revenues as you are?
152
JAMES DOUGHNEY
Therefore overall losses per head of population in the five regions may not fall. Only when
the operators are forced to make larger cuts in machine numbers in poorer high gambling
areas will we really see falls in per capita gambling. Only significant reductions in machine
numbers and severe limits on shifting machines to ‘green fields’ sites in areas of similar
socio-economic status (e.g. parts of Hume, Casey, Wyndham etc.) will reduce harm. Industry
representatives went through a mandatory process of perfunctory wailing, but ever revealing
Tabcorp spokeswoman Tricia Wunsch (1998) let the cat out of the bag when she said the
cutbacks would not be effective.
A question for the policy analysis is whether the Government’s approach is cynical or just
incompetent. It could, of course, be both. In fact I am inclined to think that there are elements
of blundering and sleight of hand in what the Government is doing. Surely it would be to their
advantage, electorally and to cement relations with the three independents, to be more
forthright on gambling policy? In fact, in a breathtaking volte face, the state opposition is
giving hints that it might use poker machine policy to campaign against the Government.
Perhaps the lead was shown when former Crown head Lloyd Williams made a public but
scarcely believable mea culpa about the harm caused by the spread of pokies throughout the
community. What is it that explains the Labor Government’s inaction and worse?
The answer goes deeper than mere electoral politics. We have seen from Table 3 that the
Victorian Government raises about one billion dollars each year in pokie tax. Sixteen or so
per cent of its own account revenues comprise gambling taxes. The pokie tax is the biggest
component of these by far. It is irrelevant that this is a regressive tax par excellence,
disproportionately paid by people on low incomes. Quite simply the Government is locked
into a neo-liberal agenda of tax reduction for business in the state, while receiving
circumscribed grants from a Commonwealth even more driven to do likewise while
simultaneously reducing PAYG taxes and factoring the GST into future budgets.12
This article has given data to support its original hypothesis that poker machines are located
in areas of lower socio-economic status than had been previously demonstrated. When venues
are located disproportionately in less well off areas the corollary is that less well off people
are frequenting them disproportionately and losing disproportionately. If these were
alternative recreational venues, cinemas say, then we might not think it important. Indeed we
might even be pleased. However, when the subject is poker machine gambling we must raise
an ethical question. Is it right for our society to license, literally, a form of institutional
banditry against those less well off? Why is it banditry? The answer is that there is no equality
of risk. This is not a game of ‘toss the coin’. Tattersall’s and Tabcorp’s are guaranteed
duopoly profits. The government is guaranteed a tax take equal to the operators’ profits (or
rents). These rents and taxes are regressive: more regressive than we had earlier been able to
establish. Income is effectively, very effectively, being redistributed away from low income
areas, and its flow back effects are at best marginal. (It is probably even wrong in the
aggregate to speak of poker machine ‘gambling’: there is no gamble when the ‘house’ bears
no risk and sets its winnings by computer program)
What is to be done? An even redistribution of poker machines – such that machines were
capped regionally at the statewide average per adult or thereabouts – would be a major
advance. However, it would only be half of the social policy remedy. If the redistribution does
12
There may be a lesson in the Victorian experience for a future Labor government. It is not so
easy to roll back indirect taxes while remaining faithful to the low-tax fundamentals of neoliberalism.
153
not take account of the fact that a dollar lost in Toorak hurts less than a dollar lost in
Braybrook then poorer areas will remain disadvantaged. Equality in the proportion of the
average person’s budget in Balwyn and Springvale, as Keynes might have put it,13 must also
be a guiding principle. The redistribution therefore must be social not merely geographic,
taking into account both an area’s losses and its capacity to pay.
What else is to be done? Three additional issues must be discussed and worked on by
researchers, policy makers and the community. First, we now know that we must narrow the
focus so that we can understand the policy implications of the distribution of venues within
municipalities. That LGA-level data are a rougher approximation has to be acknowledged.
Yet even the sorts of finer data presented here are limited. Until the ‘commercial in
confidence’ veil of secrecy is lifted from venues, and the duopoly is forced to accept
independent research to answer the great unanswered question of ‘who gambles’, then we will
be left with first, second and third approximations regarding gambling socio-economics.
The second issue I will pose as questions. What are the ethical conditions that might allow
continued government taxation of an activity like poker machine gambling? Is the guaranteed
monopoly rent levied by Tattersall’s and Tabcorp justified in any sense at all? It might seem
strange, but the Victorian community has yet to debate these questions seriously.
Third, we must now raise a question that most parties to this debate, this author included,
have ducked, dodged or, in more academic language, eschewed. Policy hitherto has been
shaped with the limit of 27 500 non-casino machines taken for granted. It is about time – and
in the certain knowledge that this government will do little, short of an epiphany – that the
policy of reducing aggregate machine numbers is argued strongly as a matter of equitable
redistributive social justice.
References
Australian Bureau of Statistics, (ABS) (2001), Taxation Revenue, Cat. No. 5506.0, ABS,
Canberra,
Doughney, J. and T. Kelleher (1999), Preliminary Local Area Gambling Research –
Economic Effects, Workplace Studies Centre Victoria University, Melbourne.
Keynes, J.M. (1982), Social, Political and Literary Writings, Vol. 28 of The Collected
Writings of John Maynard Keynes, Macmillan, London.
KPMG Consulting (2000), Longitudinal Community Impact Study, Victorian Casino and
Gaming Authority, Melbourne.
O’Donnell, R.M. (1999), Keynes on Gambling, paper presented to the History of Economic
Thought Society of Australia Conference, ANU, Canberra, 15 July.
Productivity Commission (1999), Australia’s Gambling Industries, Part 1, Part 2, Productivity
Commission. Canberra.
Tasmanian Gaming Commission (2001), Australian Gambling Statistics 1974-75 to 19992000, TGC, Hobart.
Wunsch, T. (1998), ‘Evidence to the Productivity Commission enquiry into Australia’s
Gambling Industries’, Productivity Commission, Canberra, http://www.pc.gov.au
13
Perhaps he might have spoken of the increasing marginal utility of income to the less well off.
154
Targeting Aged Care to People who are
Financially or Socially Disadvantaged:
Policy Evaluation Study About The Social
Gradient in Health of Older South
Australians.
Brian Fleming1
Department of Health, South Australia
1
Introduction
There is considerable research interest in inequalities in health, or inequities in health, or,
as I prefer, the social gradient in health. Much of the published research is descriptive
and there is interest in moving to policy approaches to the issue and to its attenuation.
This study lies at the beginning of that path as it examines existing health services
funding policy by asking how a particular policy might be different if evaluated from the
perspective of a social gradient in health. Health services’ funding occupies the great
bulk of the effort and attention of the Department of Health and Aged Care in Australia.
A change in health services’ policy is ‘after the event’ of health differences, and therefore
does not contribute to the attenuation of the social gradient in health. However, this
policy evaluation study is a contribution to a more general advancement of social-gradient
policy formulation.
2
Background
Responsibility for aged care in Australia is shared between national and state jurisdictions
in a federal system. The Commonwealth has responsibility for aged care services and the
supply of these places is regulated by a population-based formula. Aged care services, for
this paper, include high care (formerly nursing home care) low care (formerly hostel care)
and community aged care packages (whose origins are as a home-based substitute for
N=
I would like to acknowledge the helpful comments of Dr. Neville Hicks on an earlier
draft and my employees assistance with study through Mr David Kemp.
Fleming, B. (2002), ‘Targeting aged care to people who are financially or socially disadvantaged: policy
evaluation study about the social gradient of health of older South Australians’, T. Eardley and B.
Bradbury, eds, Competing Visions: Refereed Proceedings of the National Social Policy Conference 2001,
SPRC Report 1/02, Social Policy Research Centre, University of New South Wales, Sydney, 155-179.
BRIAN FLEMING
hostel care)2. The current formula for the supply of Commonwealth aged care services,
or ‘places’, at one hundred per thousand persons age 70 years or more, (100/1,000 70+)
was established in 1986 by the Nursing Homes and Hostels Review (Commonwealth of
Australia 1986: 129).
The reasoning for selecting 100/1000 70+ was that this ratio would maintain the then
existing average national supply while delivering ‘equity’ over time between the states
and territories (Commonwealth of Australia, 1986).3 Since that time the ratio has also
been used to deliver population-based parity of services within states, down to areas
defined as Statistical Local Areas (SLAs), which corresponded closely to local
government boundaries, particularly in South Australia, until about 1998.4 So, while it
was not created as an indicator of ‘need’, the then existing national average ratio was
adopted, and has been used since, as a benchmark for planning and providing additional
services to SLA level. The department routinely obtains population projections by SLA
from the Australian Bureau of Statistics (ABS), computes the benchmark, and uses the
results to inform decisions about both the overall supply and distribution of additional
services. There is a growing population over 70. The department controls the distribution
of aged care places by advertising areas of under-supply, informed by the benchmark, and
selecting providers of care from applicants for funding.
This study is not about the total national supply of places, which is decided by the
Minister for Aged Care, informed in part by the ratio. Instead, this study is about the
distribution of those places, particularly the intra-state distribution effects.
There are three main distributional issues with the benchmark formula. First, it requires
an overlay to consider access; SLAs have large variation in geographic area, with some
large sized rural SLAs having relatively small 70+ populations. So the distribution of
services within an SLA becomes important. Secondly, the formula distributes services to
SLAs in a way that is affected by spatial settlement patterns. Some SLAs, settled at
particular times, have a cohort effect where a relatively large proportion can reach the
planning age together. This paper is about a third issue: the formula delivers
proportionately more services to SLAs that have longer lived populations. An argument is
2
The majority of data presented is for residential aged care, which includes high care and low care.
3
See page 42 of the Nursing Homes and Hostels Review 1985. A comparison with international
supply at the time showed that Australia’s total was similar to 11 other ‘high provision’ countries but
more heavily oriented to nursing care places. Another aim of the review was therefore to shift the
balance between nursing home care and hostel care, so the formula of 100/1000 70+ was divided
into 40 nursing home and 60 hostel care places, and later, 40 high care, 50 low care and 10
community aged care packages
4
There was a concerted effort to amalgamate local governments, which resulted in boundary changes.
156
TARGETING AGED CARE TO PEOPLE WHO ARE DISADVANTAGED
made for a distributive method based on life expectancy, so it is relevant to examine some
features of life expectancy calculations.
3
Life Expectancy
Life expectancy is calculated from life tables using current age-specific death rates, that
is, using the current rate of deaths per population between exact age ‘x’ and exact age
‘x+1’.5 Life expectancy, at age ‘x’ can be regarded as the probable mean length of
additional life beyond age ‘x’ of all the people alive at age ‘x’ if they had the overall
population experience. Life tables are produced based on a hypothetical starting
population of 100 000 persons at exact age zero, with Australian tables ceasing at age
99years.6 Males and females experience markedly different death rates, so life tables are
produced separately by sex. Deaths in each year reduce the population at the next age, and
if plotted, produce a curve of ‘survival’. Life tables are useful for comparing one
population’s experience with another’s as they standardise for different age structures.
A visual example of the 1998 South Australian lifetable appears at Figure 1. It has a curve
typical of developed countries, with high survival until older ages. Males have death rates
higher than females from birth and this affects their surviving population, so that the male
survival curve is below that of the female.
In Figure 2, the same 1998 life-table information is presented as deaths by single age,
rather than survivors, for these representative 100 000 populations. Here it can be seen
that, compared with female deaths, male infant deaths are higher, there is a rise in male
deaths in late teens, and deaths in older age begin to rise earlier for males. These
circumstances combine to give males a five-year reduction in the mean expectation of life
at birth compared to females. A formula that delivers more services to an increasingly
aged population size, as the current benchmark approach does, is a valuable and
accountable feature of the aged care program. Use of health services, however, is much
more closely related to time until death than time from birth (Fuchs 1984; van Weel and
Michels, 1997; Himsworth and Goldacre, 1999). One estimate is that health service
expenditure in the last year of life is 6.6 times (in the next to last year, 2.3 times) as large
as for those who survived at least two years (Fuchs, 1984). So aged care services may be
better targeted if based on years from death than years from birth, as is now the case.
5
Note this is the ‘current’ life table method. An alternative is the cohort life table method,
where the death rate experience of birth cohorts is used, see for example Armitage and
Berry, 1994: 470. Current life tables can also be produced for age groups and these are
referred to as abridged life tables. Life tables in this paper were produced by the Australian
Bureau of Statistics.
6
There are smoothing formulae applied in official tables to adjust for annual fluctuations. In
Australia that has meant using three years of data to produce life tables for the middle year.
157
BRIAN FLEMING
Figure 1: Lifetable Survival Curves, South Australia, Male and Female
100000
90000
80000
70000
60000
females
50000
Survivors
males
40000
30000
20000
10000
0
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
10
0
10
5
11
0
Age
Source: ABS, 1999, Catologue No. 3311.4
Figure 2: Lifetable Deaths Curves, South Australia, Male and Female
6000
5000
4000
females
Deaths
3000
males
2000
1000
0
1
6
11
16
21
26
31
36
41
46
51
56
61
66
71
76
81
86
91
96
10
1
10
6
11
1
Age
Source: ABS, 1999, Catalogue No. 3311.4
An approach based on years from death, or life expectancy, is supported by evidence that
the expected length of stay in residential aged care, in Australia, varies very little
compared with age at entry, see Table 1, (Australian Institute of Health and Welfare,
1999). Years to death is relatively constant at about one year for males and two years for
158
females. If anything, the trend is that the younger the age at entry the shorter the length of
stay, contrary to what might be expected.7
Table 1: Expected Length of Stay in Residential Care by Age, Compared with Life
Expectancy at that Age
Age ‘x’
Males
Females
65
80
90
65
80
90
Life expectancy at Expected length of stay
age ‘x’
(years) at age ‘x’
16.1
7.2
4.0
19.8
9.0
4.6
0.7
0.9
1.1
1.9
2.2
2.1
Source: AIHW Australia’s Welfare 1999, Chapter 6; Gibson 1996.
Life Expectancy and Residential Aged Care
In this section I establish that information about the lifetable model of deaths offers a
reliable surrogate indicator for the pattern, by age, of entry to residential aged care.
The aim was to gain an understanding of relationships, between life expectancy and
residential care use. I obtained data from the Department of Health and Aged Care on the
current age, of all persons in residential aged care in South Australia, at 30 June 2000.
The age distribution of actual use of residential aged care in South Australia in June 2000
is shown in Figure 3 and Figure 4. Note that few people aged 70 years, the ‘planning age’,
use residential aged care. The median age at entry to residential care in South Australia is
about 82 years, so using the population 70+ includes the majority of people who use aged
care but the planning age itself is not a representative descriptor of the population using
aged care, in a statistical sense. This has led to suggestions that raising the planning age to
80+ years would make more sense, as this would be closer to the median age at entry and
an age more representative of the pattern of use – this idea is criticised later in the paper.
The lifetable deaths for South Australia, are the product of rising age-specific death rates
by age and a diminishing population by age. Similarly the population in aged care is a
product of rising age-specific risk of admission and a diminishing population by age. It is
likely that these are related as life expectancy in aged care is limited, see Table 1, and
most separations are due to death. Data on residents in aged care services were overlayed
with deaths data, and are discussed separately by sex.
For females the data were for current residents. According to Gibson, (see Table 1 1996,
5), female length of stay is about two years, independent of age. So, assuming an average
of one year in care for current female residents the deaths data were translated forward
7
While most departures will be due to death, this may be affected by transfers.
159
BRIAN FLEMING
one year so that the number of deaths relates to the age less one year. This puts the
‘deaths in two years’ data roughly in line with the resident data at entry to aged care, for
females. This results in Figure 4. The data are a very good visual match. Linear regresion
also confirms a very strong relationship: r-squared = 0.9623p<0.0001 (female) and rsquared = 0.9286 p<0.0001 (male – see Figure 6).
Figure 3: The Distribution of Females and Males in Residential Aged Care in South
Australia by Age, June 2000
SA current residents by single age (high and low care) 30 Jun 2000. Males (n=3380)
Females (n=9612)
600
500
400
female residents
male residents
Freq in residential aged
care
300
200
100
0
1
6
11
16
21
26
31
36
41
46
51
56
61
66
71
76
81
86
91
96
101 106 111
Age
Figure 4: The Distribution of Females in Residential Aged Care in South Australia
by Age, June 2000, and Deaths from South Australian Lifetable: 1997-99
SA females by single age on two axes: 1. residents (high and low care) 30 Jun
2000 (n=9612) and 2. Deaths 19971999)
600
6000
500
5000
400
4000
Freq in residential
aged care
300
3000
200
2000
100
1000
0
1
7
13 19 25 31 37 43 49 55 61 67 73 79 85 91 97 10 10
3 9
Age
Source: ABS, (1998), Catalogue No. 3311.4
160
0
Deaths between age
(x-1) and x
female residents
female deaths
In order to compare a years-from-death approach, with the current planning approach, the
South Australian lifetable was used to describe the pattern of survivors at aged 70 or
more. Given that the lifetable is based on current age-specific death rates, the shape of
this curve reflects the South Australian experience of survival at these ages; the actual
numbers are not relevant at this stage. It can be seen from Figure 5 that the current
method of counting the population aged 70 years or more is a poorer fit than deaths data,
r-squared = 0.3158, p<0.0001.
Figure 5: The Distribution of Females in Residential Aged Care in South Australia
by Age, June 2000, and Survivors 70 Years or More from South Australian
Lifetable: 1997-99
SA females scatter plot by age: 1. Survivors= >70 1997-9 (schema life table)
and 2.residents (high and low care) 30 Jun 1999 (n=9612)
600
100000
90000
500
80000
70000
400
60000
50000
300
residents
survivors
40000
200
30000
20000
100
10000
0
0
5
10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 10 10 11
0 5 0
0
Source: ABS (1999) Catalogue No. 3311.4
The exercise was repeated for data on males in residential care in South Australia,
without shifting the deaths data by one year. This is because the average length of stay is
only one year for males and the current age will already be roughly six months older than
age at entry (see Figure 6). The 70+ figure for males is similar to that of females but is not
shown; r-squared for deaths is 0.9285 p<0.0001 and for 70+ is 0.4575.
It is clear that age specific deaths by sex are very strongly correlated with current patterns
of residential aged care use by sex. Since summary statistics for the lifetable pattern of
deaths and residential care population are very similar, particularly at entry to aged care,
information about deaths is confirmed as a tool to describe residential aged care use.
The Social Gradient in Health and Aged Care Planning
Life expectancy is related to material circumstances, and, if there is substantial spatial
segregation by measures of material circumstances then life expectancy will vary spatially
or geographically.
161
BRIAN FLEMING
In Australia, and South Australia, a series of social atlases illustrate a variation of health
outcome by area, along with the spatial distribution of material and other resources, in
line with the expected pattern –the higher the material indicators the better the health
indicators, including mortality (Glover, Shand et al,. 1996; Glover and Tennant, 1999a
and 1999b).
Figure 6 The Distribution of Males in Residential Aged Care in South Australia by
Age, June 2000, and Deaths from South Australian Lifetable.
SA males by single age on two axes: 1. residents (high and low care) 30 Jun
2000 (n=3380) and 2. Deaths 1997-9 (life table n=100,000)
200
5900
180
4900
160
140
3900
120
100
2900
male residents
male deaths
80
1900
60
40
900
20
0
0
5
10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 10 10
0 5
-100
If one area has lower life expectancy than another then its population will experience the
last two years of life at a lower age; the survivors at a given age will be fewer. So a
funding formula based on a particular age, say 70 years, will deliver greater resources to
areas that have a higher survival, other things, such as age structure, being equal.
For example, take two areas A and B with about a five-and-a-half year difference in life
expectancy; 81.5 and 75.9 years respectively.8 Typical survival curves for the two areas
are shown at Figure 7. If the populations required care in the last two years of life, then on
average, population A would need care from 79.5 to 81.5 years and population B 74 to 76
years, with similar numbers of people requiring care. The population aged between 81
and 82 in Area A would be 62285 and between 79 and 80 in area B 62545. However,
under the present benchmark formula of 100/1000 70+, the difference in the suviving
population at age seventy is 85238 – 74697 = 10541. Under the benchmark formula for
aged care the Commonwealth would supply 8524 places in area A compared to 7470 in
area B, a difference of 1054 places.
8
This is about the size of the expected difference between the longest and shortest lived regions.
162
Figure 7: Survival Curves for Two Populations Whose Life Expectancy is about 5
Years Different
Number surviving at
100000
age x
Life Table survival curves
Survivors at
age = 70.0
90000
Life expectancy
at birth = 81.5
80000
70000
60000
Life expectancy
at birth = 75.9
50000
40000
30000
20000
10000
0
x
Source ABS 3302.0
There are 11 South Australian planning regions, four metropolitan and seven nonmetropolitan, used by the Commonwealth Department of Health and Aged Care. The four
metropolitan regions are, conveniently, North, South, East, and West. The consistent
overall pattern is that the East region experiences the best health and socioeconomic
circumstances, with the North the poorest, and West and South in between (Glover and
Woollacott, 1992). Life expectancy is therefore expected to reflect these general patterns.
However, regional and SLA life tables are not currently produced nationally. An
indication of what regional and SLA tables might reveal is given by some research in
Victoria; a page extracted from the publication is at Figure 8. (Victoria, 1999, Burden of
Disease Study, http://www.dhs.vic.gov.au/phb/9903009/index) This shows the expected
regional variation in life expectancy, in this case for males by local government area. It
also shows a clustering of LGAs that are below, equal to or above the State averages,
particularly in the metropolitan area.
In the following pages I propose a method to distribute aged care resources using
information about the spatial distribution of life expectancy and show how it would
compare to the distribution based on the benchmark 100/1000 70+. In order to
demonstrate how a method based on deaths information might work, I have used median
age at death by SLA as a proxy measure for life expectancy.9 The median age at death is
similar to life expectancy at birth, in most jurisdictions of Australia, but does not adjust
9
The ‘median age’ statistic does not standardise for the population structure, as life expectancy does,
and so will deliver a younger age where there is a younger age structure.
163
BRIAN FLEMING
Figure 8: Extract of Australian Example of Life Expectancy by Local Area
Source:Victoria, (1999) Burden of Disease
for age structure see Table 2 For example, younger age structures in the Northern
Territory and the ACT affect the size of the difference between life expectancy and
median age at death.10 ABS supplied data on the average annual number of female
deaths, male deaths, male, female and person median age at death by SLA for the years
1997-99.11 A wider range of years was not readily obtainable due to boundary changes.
Aggregates of those SLAs, however, were in current use in aged care planning aged care
through 1999.
10
The Northern Territory has a younger population and a relatively large indigenous population, which
has high standardised- mortality (2.2 for males and 1.5 females in 1997-98). The Australian Capital
Territory, the home of the national government, has a healthy, educated and young worker inmigration effect.
11
SLAs were based on 1996 profiles.
164
Table 2 Compare median age at death with expectation of life at birth, Australian
jurisdictions.
NSW
Vic
Qld
SA
WA
Tas
NT
ACT
Aust
Median age death
e00 96/8
Males
74.4
74.8
73.7
75.3
73.6
75.1
53.7
72.6
74.3
Male
75.8
76.3
75.6
76
76.1
75.1
70.6
77.5
75.9
Females
81
81.6
80.3
81.7
80.8
80.5
57.7
78.8
81
delta (LE-median)
Female
81.6
81.7
81.5
81.6
81.9
80.4
75
81.6
81.5
Male
1.4
1.5
1.9
0.7
2.5
0.0
16.9
4.9
1.6
Female
0.6
0.1
1.2
-0.1
1.1
-0.1
17.3
2.8
0.5
Source: ABS, (1998:12) Table 1.1, 12 Catalogue No. 3302.0
Only SLAs with more than an average of 5 female deaths per year were used in the
subsequent analysis.12 A scatter plot of median age at death by SLA appears at Figure 9.
Figure 9: Median Age at Death, Males, Females and Persons (ranked), by SLA
Median age at death, males, females, persons(ranked), by SLA deaths(f)>4
90.0
85.0
80.0
75.0
70.0
65.0
60.0
Persons
Males
Females
55.0
50.0
45.0
40.0
In Table 3, I have set out some basic features of the data. When separated into the
planning regions used by the Commonwealth, the metropolitan regions show the expected
relationship; the East has the highest group of median ages and the north the lowest, see
Figure 10.
12
The decision to use five deaths was based on discussion with ABS. Five female deaths was chosen as
the population in aged care is mainly female and it is a more conservative approach as male deaths
usually exceed female deaths.
165
BRIAN FLEMING
Table 3: Median Age at Death Data by SLA - Selected Features
SLA
Avg.
annual
male
deaths
Lowest median age
at death persons
18
Highest median
age at death
persons
6
With most (male
and) female deaths 478.5
With largest
population
(111 910)
287
Median
Median
Median
age death age death age death
male
female
persons
Avg.
annual
female
deaths
5.5
40.2
52.5
56.6
8
80.0
86.0
85.3
404.5
76.5
80.8
78.5
226.5
70.0
76.8
72.9
Figure 10: Median Age at Death, Persons, and Metropolitan Areas South Australia
(SLAs) Grouped into Commonwealth Planning Regions
Median age at death, deaths 1997-98, SLAs grouped into four urban planning regions, more than 5 female
deaths.
86.0
Kensington
St Peters
Walkervill
Unley
Payneha
Adelaide
Burnside
84.0
Glenelg
Mitcham
Brighton
82.0
West
THenley&Grang
80.0
78.0
Gawler
Prospect
Cambelltow
Hindmarsh&Woodvill
Onkaparing
Marion
Port
Ad l id
Thebarto
Noarlunga
76.0
Enfield
Stirling
T T Gully
EnfieldB
74.0
72.0
Salisbury
Elizabeth
70.0
Munno
P
North
Happy
V ll
Willunga
East Torrens
South
West
Source: ABS (1999), Catalogue No. 3302.0
The non-metropolitan data is displayed at Figure 11.
166
East
Figure 11: Median Age at Death, Persons, Non-metropolitan Areas South Australia
(SLAs) Grouped into Commonwealth Planning Regions
Median age at death, deaths 1997-98, SLAs grouped into seven non- urban
regions, more than 5 female deaths. Source ABS
90.0
85.0
Quorn
80.0
Loxton
77.1
76.8
76.3
75.0
70.0
Cleve
80.9
79.0
78.1
77.5
76.9
76.8
Tatiara
78.3
77.5
76.8
74.5
72.5
70.5
Berri
Mt Gambier
Tanunda
82.9
81.9
81.4
81.3
81.1
80.8
79.5
79.1
78.7
78.4
78.2
75.4
72.1
71.8
70.0
Mallala
Pinnaroo
Orroroo
80.9
80.1
78.9
78.3
77.3
76.4
76.3
76.2
75.0
74.4
Yankalilla
80.0
79.5
78.0
76.6
76.1
Pt Broughton
65.0
Ceduna
60.0
55.0
Uninc far north
Whyalla &
Riverland
Far North
Eyre
South
East
Yorke &
Barossa
Hills
Mallee
Mid North
In order to arrive at a representative median age at death by region the medians were
treated as means and population-weighted to arrive at the ‘medians’ in Table 4. I am well
aware of the arithmetical leap here but, while it would be a relatively straightforward
exercise for the ABS to make the calculations based on the original deaths data, it is
expensive for an individual to obtain this data, and not warranted as the exercise is
illustrative. The result is consistent with other spatial information and indicates clearly a
range of median age at death, by locality, for both Adelaide and the rest of South
Australia.
The analysis of deaths data and residents by age suggests that one possibility would be to
count the numbers of deaths and use the subsequent proportions by area to distribute aged
care. Note that the very strong correlation, between deaths data and resident data, is for
these two sets to move together by age. At SLA level, deaths are quite small in number,
and, while adequate to determine median age at death, are too unstable by single year to
discriminate well between SLAs for planning purposes. A surviving population measure
is more stable. What is needed is a distribution that matches the residential aged care
residents’ profile by SLA. The curve of survival above the life expectancy age is the
same shape as the last half of the curve of residents in aged care. If that last part of the
survival curve, by SLA, were mirrored about the vertical from the life expectancy age, the
resulting distribution would more closely match the residential distribution. This would
167
BRIAN FLEMING
Table 4: Median Age at Death, Persons, South Australian Areas (SLAs) Grouped
into Commonwealth Planning Regions
South Australian Region
Median age at death
(yrs)
Whyalla and far north
Riverland
Eyre
South East
Yorke &Barossa
Hills Mallee
Mid north
Non-metropolitan
planning regions
70.3
76.3
76.4
76.6
77.4
77.4
77.8
North
South
West
East
Metropolitan regions
73.7
78.0
78.4
80.3
have the effect of doubling the population survival count at a given age. Since the
resulting count is to be used relatively, as it is distributional, the effect of doubling the
count and allocating proportions is the same as counting the original half and allocating
proportions.
ABS data was obtained on the population single ages from the 1996 census13. The
calculation is done for regions and the method is illustrated in Figure 12. Take the region
in the top row and the median age at death from the corresponding column in the second
row. The population greater than or equal to the median age at death is read from the first
column at the bottom of the table. The existing benchmark process reads the population
of each region from the 70+ row. So, for example for the northern metropolitan region,
the median age at death is 74 and the population is 9845, shown shaded in dark, rather
than the 70+ population of 15 410, unshaded.
A comparison of the proportion of places allocated to different regions using the
population over compared with the population older than the median age is shown at
Figure 13. The proportion of places that would ‘move’ is about 15 per cent; given 13 000
total places in South Australia, about 2 000 would move, summarised in Table 5.
13
ABS projected populations, for 1997-99, were only available in 5-year age groupings and
individual ages are necessary to plan at different ages. This is not important for the
comparison as the same population base is used, and individual age projections can be
sought from ABS.
168
Figure 12: Determining the Population by Using a Life Expectancy Approach
Planning region
North South West East
Eyre
Med age at
death - person
= 'X'
74
78
78
80
76
Hills
Mallee
Whyalla Yorke Lwr
Flinders North
Mid NorthRiverlandSouth EastFar NorthBarossa Totals
77
78
76
77
70
77
4065
1302
1603
2292
3043
3684
63964
6.31%
2.24%
2.25%
3.50%
2.20%
5.59%
1
6.36%
2.04%
2.51%
3.58%
4.76%
5.76%
1
abs (med -70+) 4.26% 1.77% 1.93% 3.83% 0.35%
0.04%
0.20%
0.26%
0.08%
2.56%
0.17% 0.154688133
Population 65 + 24099 44037
66+
22208 41648
67+
20367 39199
68+
18671 36704
69+
17009 34134
70+
15410 31541
71+
13887 29014
72+
12490 26566
73+
11128 24177
74+
9845 21734
75+
8631 19376
76+
7453 17095
77+
6341 15097
78+
5594 13438
79+
4876 11792
80+
4239 10223
12711
11917
11085
10302
9524
8738
7964
7220
6518
5874
5215
4645
4065
3611
3153
2755
4586
4277
3987
3686
3377
3101
2824
2589
2315
2080
1870
1639
1452
1302
1140
992
4610
4321
4032
3703
3410
3113
2839
2612
2349
2078
1833
1603
1432
1266
1113
972
7129
6658
6156
5710
5277
4850
4422
4029
3657
3271
2943
2596
2292
2029
1778
1548
4881
4447
4099
3696
3362
3043
2736
2469
2219
1954
1743
1520
1339
1186
1033
861
Population 'X' + 9845
13438 11030 12141
1521
Region share
total based on
70+ pop
11.13% 22.78% 19.17% 22.81% 2.02%
Region share
based on
median age
"X"+
15.39% 21.01% 17.24% 18.98% 2.38%
36962
34927
32988
30899
28756
26547
24294
22197
20038
18011
16033
14051
12319
11030
9754
8543
42721
40544
38290
36058
33886
31591
29256
27096
24865
22703
20630
18591
16731
15198
13660
12141
4133
3842
3559
3298
3070
2802
2573
2335
2122
1911
1727
1521
1370
1194
1052
916
169
11274
10534
9814
9110
8393
7734
7020
6390
5841
5233
4704
4166
3684
3256
2866
2469
138470
BRIAN FLEMING
Figure 13: Comparison of the Proportion of Aged Care Services by Region Using
Median Age at Death (+) versus 70 Years from Birth(+).
Comparison between proportions of places by region using 70+ vs median age at death, SA planning
i
25.00%
20.00%
15.00%
10.00%
5.00%
Sources
ABS data
1996 and 0.00%
1997/8
70 +
median age
North
South
West
East
Eyre
11.13%
15.39%
22.78%
21.01%
19.17%
17.24%
22.81%
18.98%
2.02%
2.38%
Hills
Mallee Mid North Riverland South
E t
6.31%
2.24%
2.25%
3.50%
6.36%
2.04%
2.51%
3.58%
Whyalla Yorke Lwr
Flinders
North
Far North Barossa
2.20%
4.76%
5.59%
5.76%
Source: ABS 1996 Census
Table 5: The Change in Service Distribution Using Median Age at Death
Pop 70+
North
South
West
East
Eyre
Hills Mallee
Mid North
Riverland
South East
Whyalla Flinders Far North
Yorke Lower North Barossa
Median age Change
Percentages
14.10
2.97
24.66
1.88
15.93
-3.24
19.83
-2.98
2.48
0.46
5.90
-0.41
1.86
-0.38
2.99
0.74
3.74
0.24
3.19
0.99
5.32
-0.27
0.00
11.13
22.78
19.17
22.81
2.02
6.31
2.24
2.25
3.50
2.20
5.59
100.00%
Absolute change
2.97
1.88
3.24
2.98
0.46
0.41
0.38
0.74
0.24
0.99
0.27
14.56
The results are consistent with what one would expect from a distribution of services
based on life expectancy given the spatial segregation of South Australia shown in social
atlases (Glover, Shand et al., 1996; Glover and Tennant, 1999). The North Metropolitan
region ‘share’ of a given resource would rise, as the last two years of life would be
experienced at a younger age. The demonstration has been with median age, as a proxy
170
for life expectancy, and so would distribute relatively more services to structurally
younger populations. A more sophisticated approach could be envisaged by calculating
male and female provision separately, based on a one- year expected residence for males
and two years for females, from Table 1. For policy purposes however, it may be
preferable to say simply that services are planned on known patterns of use.
It should be noted that, in South Australia, based on the current planning benchmark
100/1000 70+, there is a large ‘oversupply’ of places in the East region, a residential aged
care variant of the inverse care law.14 Life expectancy, if calculated by region, or by
Statistical Local Area (SLA), uses deaths data and it is known that the calculation is
affected by the recorded ‘usual residence’ at death. So, if people from other regions, who
are likely to die younger, are unable to receive care, in the last two years of life, in their
home region and move to the East, to access the high supply, the life expectancy of the
East will reduce as an artefact. The effect will be to attenuate the differences between
regional life expectancy. So the method, if anything, will underestimate the size of the
redistribution required, at least at first.
Supporting Data on Actual Use
In the next section the theory, that there are spatial differences in the age at which
populations use aged care services, consistent with mortality experience, is confirmed in
two ways. One is that potential residents are assessed at different ages by region. Another
is that the actual use of residential aged care shows a variation of age by region.
Assessment
Assessment for entry to aged care is carried out independent of services by an Aged Care
Assessment Team (ACAT) in each planning region. The ACAT evaluation team report
shows a picture of assessments at higher ages in the east compared with other regions
(Basso, Pretreger et al., 1999: 22).
While the variation in age is small, it should be noted that ACATs assess for more than
residential aged care and the variation in age at assessment is consistent, for example,
with higher life expectancy in the east15.
The median age is higher than the mean age in each case indicating some skewness in
assessment age in the distribution.
14
‘The availability of good medical care tends to vary inversely with the need for it in the population
served … This ... operates more completely where medical care is most exposed to market forces,
and less so where such exposure is reduced’ Hart, J. T. (1971), 405-12, 'Commentary: Three decades
of the inverse care law' Hart, 2000: 18-19.
15
A caution must be noted that this is assessment of eligibility for, not the same as admission to aged
care.
171
BRIAN FLEMING
Table 6: ACAT Client Age by Team(s) 1998
ACAT team(s)
Median
Mean
Northern
Western
Southern
Eastern
All Metropolitan
All Country
82
82
83
83
83
82
80.4
80.6
81.8
82.1
81.2
81.4
Source: ACAT Evaluation Unit Report Jan-Jun 1998.
Residential Aged Care
Data were obtained on the age at entry to residential aged care. A potential confounder is,
as noted above, that there is high supply in the East region and low supply in the North
region, against the benchmark. This means that the age of existing residents will not truly
reveal any pattern in the predicted direction, to the extent that younger residents move
from North to East to gain services. Accordingly, data were extracted on a cross section
of residents in permanent residential aged care at 31 December 2000 and divided into
regions based on the address recorded when the aged care assessment had been made.
The assessment form asks for ‘preferred contact address’ and it is possible that a
proportion of these will have the address of a friend or relative not in the same home
region as the resident. This is a limitation of the data collection. The results are shown in
Table 7 by sex, separately for metropolitan and non-metropolitan regions and ranked by
female mean age, females being the main users of residential aged care.
The age differences are notable. For comparison, a two-year difference in life expectancy
is the estimated impact of eliminating all cancer.
The use of contact address is unlikely to have an impact. There would have to be younger
aged care residents from higher socioeconomic areas systematically giving lower
socioeconomic area contact addresses, or older people from lower socioeconomic areas
systematically giving higher socioeconomic area contact addresses, for the result to be in
doubt. This may happen if the individual circumstances of the person were not that of the
area, the ecological fallacy, in which case the underlying association with socioeconomic
circumstances would, ironically, be confirmed. It seems more likely that the impact
would be to attenuate the real differences by younger people from lower socioeconomic
areas giving a higher socioeconomic area as the, say, child’s address.
Comparing The Ranking of Regions Using Median Age at Death and Age at Entry.
The ranking of the regions in Table 4 and Table 7 is similar, with an outlier of Eyre. The
median age at death of persons in Eyre is affected by the size of the indigenous population
relative to the non-indigenous population, with indigenous deaths at younger ages. The
172
Table 7: Age at Admission by Region: SA Residents in Permanent Care by
Preferred Contact Address and Age at Date of Admissions: 31 December 2000
Ranked by female mean age.
Female
Metro
Mean Age
Male
Median Age
Mean Age
Median Age
MetroNorth
83.6
84
80.3
81
MetroWest
84.7
86
81.5
82
MetroSouth
85.3
86
82.3
83
MetroEast
85.4
87
81.5
83
WhyallaFlindersFarNorth
81.4
83
78.8
80
Riverland
84.3
85
82.1
84
SouthEast
84.4
85
82.0
83
MidNorth
84.6
86
79.5
82
YorkeLwrNthBarossa
84.7
86
81.1
82
HillsMalleeSouthern
85.0
86
81.0
82
Eyre
87.4
88
80.3
81
Non-Metro
residential aged care data are from non-indigenous providers obtained via the payments
system for standard, aged care services. Data were not obtained on the ages of people in
indigenous aged care programs in the Eyre region, which have a different funding
mechanism, where age is not recorded.
With Eyre excluded, the association between the rank of the planning region based on
median age at death (persons) and the rank of the region based on the mean female age at
entry to aged care is significant.16 With Eyre included the rank correlation two tailed p=
0.117, and with Eyre excluded p = 0.009.
4
Discussion
This paper is not concerned with the overal quantum of places of aged care, rather their
distribution. Some caution would be necessary in the use of life expectancy for other than
16
Note that the deaths data were for persons as it is intended to use general population data as the
planning tool. The age at entry is for females, the main users of aged care.
173
BRIAN FLEMING
Table 8: The Ranking of South Australia Regions by Median Age at Death (Persons)
Compared with Age of Female Residents at Entry to Residential Care by Contract
Address
Whyalla, Flinders, Far North
North
Eyre
Riverland
Hills Mallee
South East
Yorke, Lower, North Barossa
South
West
Mid North
East
Including Eyre
Excluding Eyre
Rank a
1
2
3
4
5
6
7
8
9
10
11
Rank a
1
2
Rank b
1
2
3
4
5
6
7
8
9
10
3
8
4
6
9
7
5
10
Rank b
1
2
11
3
8
4
6
9
7
5
10
distributive effects. Life expectancy has increased since the 100/1000 70+ formula was
introduced in 1986, but the absolute period of life with a disability is either not changing
or may be slightly decreasing (Mathers and Robine 1998). 17 The disability-free years of
life are increasing, and increasing as a proportion of life, for disability requiring
residential care. People were not ‘sicker’ or more frail, for longer periods of their life, in
1997 than they were in 1985, when the formula was introduced. So if the formula,
100/1000 70+, described the ‘correct’ proportion of the 1985 population requiring
services, it now describes a greater proportion of the population, yet there is increasing
demand.
Table 9: Expectation of Life at Birth and at 65 Years: 1985 and 1997
c1985
1996-8
Expectation of life
Male
Female
Male
Female
At birth
71.2
78.2
75.86
81.52
At age 65
13.7
17.9
16.32
20.01
Drivers of demand have had some scrutiny. High levels of personal care can be met by
committed and skilled carers, and carer support, or lack of it, is a critical factor in
admission to residential aged care, separate from clinical measures of the individual. Care
needs can be met at home; the majority of carers are spouses and the children of people
17
The compression of morbidity thesis was advanced by Fries in (1989: 205-28).
174
entering aged care services are themselves nearing retirement ages (Gibson 1996). Gibson
and Liu (1995) suggest that structural ageing has an effect. This is to say that the need for
aged care increases with age, such that the portion of the population 80+ has a higher risk
of admission to aged care, and so an increasing proportion of people 80+ increases
demand, or ‘need’. This point was reiterated in an Australian Institute of Health and
Welfare (2000) paper on disability and ageing, which suggested a one in three chance of a
person over 65 years ever needing aged care compared with a population in aged care at
any one time aged 65 years being about three per cent (Australian Institute of Health and
Welfare 2000). However, it was acknowledged that this conclusion was drawn from
examining the existing population in residential care, and so could not be regarded as a
statement about any objective measure of need. It also did not take increases in life
expectancy into account. Gibson’s papers point to use of disability as a guide to demand,
but these measures are not available at SLA level so it is not yet feasible to use them for
these purposes
The fact that use of aged care services increases with age has contributed to calls for the
80+ population to be used as a better indicator of need. However, as has been
demonstrated, age-based formulae deliver a higher proportion of services to longer lived
populations, that is, to higher socioeconomic areas. The higher the age the greater that
distribution. To illustrate this effect, consider the following graph, Figure 14. It shows the
proportion of the metropolitan population above a particular age by region against the
proportion of the existing metropolitan aged care supply by region. The large movements
by age are in the North and East. The North share of population declines as the age
increases, and the East share rises. This is expected with lower life expectancy in the
North; the higher the age the fewer survivors. The aged care supply stays as horizontal
lines by age. The gap between the two lines is regarded as the ‘need’ - indicated by age, or
population share minus supply share. At lower ages the East supply substantially exceeds
population but this narrows as the age increases. In the North, supply is below population
at age 70+ but not at age 80+. A planning benchmark age of 80+ would see the current
distribution of supply as being appropriate, perversely, because there would be fewer
survivors in the North.18 The main point from the graph is that the volatility of age based
planning makes it a poor indicator of need.
Certain conditions associated with residential aged care use may not be consistent with
planning based on mortality. The prevalence of dementia increases with age, and there
may not be as strong an association of that condition with socioeconomic circumstances
as there is of mortality. However, life expectancy with dementia, beyond age 65, for both
18
Note this is not adjusted for any cohort settlement effects.
175
BRIAN FLEMING
Figure 14: The Regional Proportion of the Metropolitan Population Above Ages
Compared with the Regional Proportion of Metropolitan Aged Care Supply in
South Australia
0.4
0.35
Eastern
0.3
Northern
0.25
Southern
Western
0.2
Eastern supply
0.15
Northern supply
Southern supply
0.1
Western supply
0.05
0
Sum of P65+
Sum of P70+
Sum of P80+
Sum of P85+
Source: ABS (middle projection of year 2001 population) and Department of Health and
Aged Care Supply tables 31 January 2001
sexes, has been estimated at one year, so the condition is feeding in to lifetables via agespecific death rates (Ritchie 1994).
The current planning method, based on years from birth, also distributes services to areas
that were settled at particular times. As Australia’s population expanded, particularly over
the last 50 years, with post World War II migration, new housing areas were established
from greenfields. If, for example, most people settled in an area in 1955, when they were
25, they will turn 70 together in 2000 and be counted for planning purposes, while not
having the age distribution of longer settled areas. Planning based on mortality would
overcome this weakness.
There is a spatial distribution of the sex ratio, the number of males to 100 females. In
metropolitan areas the number of females 65+ outnumbers males but this reverses in
remote and some rual regions in Australia (Haberkorn, Hugo et al. 1999, 116). This is
illustrated in Figure 15. Since men die younger than women, a fixed planning age delivers
more services to longer lived women and therefore disadvantages remote Australia for
aged care services, albeit at low numbers. Planning based on life expectancy would
incorporate adjustments for such populations.
176
Figure 15 Sex Ratio, Australia Population 65+, 1996
Source: National Key Centre for Social Applications of GIS
5
Conclusion
The social gradient in health persists into old age, and there is a clear social gradient in
age at death. There is a spatial distribution of socioeconomic circumstances and therefore
of life expectancy. An age-based planning mechanism, that is, years from birth, therefore
distributes finite resources spatially to longer lived populations. Perversely, the older the
planning age, the greater the distribution of services would be to higher socioeconomic
areas. Choosing a planning age closer to the median age at entry to aged care services, say
80 years or more, an apparently more meaningful age statistically, would have a
distributive effect towards higher socioeconomic areas.19
19
There are also cohort effects of areas settled at particular times.
177
BRIAN FLEMING
Aged care legislation includes the capacity to consider financial or social disadvantage
and there are no current mechanisms to address this special needs group. An approach
based on life expectancy, such as years to death or equal probability of death would better
predict the distributed use of residential aged care services. It would therefore better target
aged care resources.
There may be other distributive advantages. A life expectancy method can be applied to
particular populations other than spatially. So, for example, there need not be a different
planning approach to distributing services for indigenous populations.20 Similarly for
rural areas, as the method adjusts for the sex ratio change that is experienced by remote
Australia, given the lower life expectancy of males. The method also adjusts for different
age structures.
The method requires the calculation of life tables by region and SLA and populations
projected by single age. While these are not routinely available, they do not require any
additional data collection by the ABS to be produced. Home address mobility, a potential
problem for such calculations, is lowest at age 65 years (Bell and Hugo 2000: 172).
Data on the age structure of persons in residential aged care confirms a spatial pattern of
use of aged care services consistent with the association between the expectation of life
and socioeconomic circumstances.
References
Armitage, P. and G. Berry (1994). Statistical Methods in Medical Research. Blackwell
Science Ltd, Oxford.
Australian Bureau of Statistics (ABS) (1999), Demography, South Australia, Catalogue
No. 3311.4, ABS, Canberra
Australian Bureau of Statistics (ABS) (1999), Deaths, Australia, Catalogue No. 3302.0
ABS, Australia.
Australian Institute of Health and Welfare (1999), Australia's Health 1999, Australian
Institute of Health and Welfare, Canberra.
Australian Institute of Health and Welfare (2000), Disability and Ageing: Australian
Population Patterns and Implications, Australian Institute of Health and Welfare,
Canberra
Basso, P., L. Pretreger et al. (1999), Aged Care Assessment Program South Australia
Evaluation Unit Report, Department of Human Services, South Australia, Adelaide.
Bell, M. and G. Hugo (2000), Internal Migration in Australia 1991-1996, Australian
Population, Immigration and Multicultural Research Program, Commonwealth of
Australia, Canberra.
20
Tacit acceptance of planning for aged care bearing some relationship to life expectancy has been the
different planning age used by the Commonwealth for Aboriginal and Torres Strait Islander
populations, that is, the population aged 50+ rather than 70+.
178
Commonwealth of Australia (1986), Nursing Homes and Hostels Review, Department of
Community Services and Health, Canberra.
Fries, J. F. (1989), 'The compression of morbidity: near or far?, The Milbank Quarterly,
67 (2): 208-28.
Fuchs, V. R. (1984), '"Though much is taken": reflections on ageing, health, and medical
care', Milbank Memorial Fund Quarterly/Health and Society, 62 (2), 143-67.
Gibson, D. (1996), 'Reforming aged care in Australia: change and consequence', Journal
of Social Policy, 2, 157–79.
Gibson, D. and Z. Liu (1995), 'Planning ratios and population growth: will there be a
shortfall in residential aged care by 2021?' Australian Journal on Ageing, 14 (2),
57-62.
Glover, J., M. Shand, et al. (1996), A social health atlas of South Australia, South
Australian Health Commission, Commonwealth of Australia, Adelaide,.
Glover, J. and S. Tennant (1999a), A Social Health Atlas of Australia, Public Health
Information Development Unit, Commonwealth of Australia, Adelaide.
Glover, J. and S. Tennant (1999b), A Social Health Atlas of Australia, Public Health
Information Development Unit, Commonwealth of Australia, Canberra.
Glover, J. and T. Woollacott (1992), A Social Health Atlas of Australia, Commonwealth
of Australia, Adelaide.
Haberkorn, G., G. Hugo et al. (1999), Country Matters Social Atlas of Rural and
Regional Australia, Commonwealth of Australia Bureau of Rural Sciences,
Canberra.
Hart, J. T. (1971), 'The inverse care law', Lancet, i, 405-12.
Hart, J. T. (2000), 'Commentary: three decades of the inverse care law' British Medical
Care Journal, 320, 1 January, 18-19.
Himsworth, R. L. and M. J. Goldacre (1999), 'Does time spent in hospital in the final 15
years of life increase with age at death? A population based study', British Medical
Journal, 319, November, 1338-39.
Mathers, C. D. and J. M. Robine (1998), 'International trends in health expectancies: a
review', Australasian Journal on Ageing, 17, 1 Supplement, 51-5.
Ritchie, K. (1994), International comparisons of dementia-free life expectancy: a critical
review of the results obtained, Advances in Health Expectancies, Australian
Institute of Health and Welfare, Canberra.
van Weel, C. and J. Michels (1997), 'Dying, not old age, to blame for costs of health care',
Lancet, 350 (9085) 1159-60.
Victoria
(1999),
Burden
of
Disability
Study,
http://www.dhs.vic.gov.au/phb/9903009/index
179
The Poverty of Welfare Reform Discourse
Paul Henman
Department of Sociology, Macquarie University
1
Welfare Reform Discourses1
During the last decade, the need for welfare reform has been an often repeated refrain
in the First World, but especially in the English-speaking countries of Australia, the
United Kingdom, the United States of America, New Zealand and Canada. The
various reasons for exhorting welfare reform reflect a wide political range of ideas
about the economy and society.2
In Australia, the former Minister for Family and Community Services, Jocelyn
Newman (1999), argued that there had been a large increase in welfare dependency
that needed to be reduced. Alongside this, the Howard Government’s clear interest in
extending the ‘mutual obligation’ principle in welfare presents a view that the
previous welfare system lacked obligation. The Work for the Dole program also
emphasises the acquisition of a work ethic that welfare recipients supposedly lack.
In response, the Reference Group on Welfare Reform (RGWR) (2000a and 2000c)
championed the notion of social and economic participation as a way to interpret the
need for welfare reform and to shape the form it might take. The language implies that
welfare recipients do not participate in Australian society or the economy.
In response, and in its earlier social security and unemployment reforms, the
Australian Labor Party has continually championed labour market training programs
to address the perceived lack of skills of unemployed persons, a lack which is
supposedly self-evident due to the very status of being unemployed (Prime Minister,
1994). Buttressing these programs has been the utilisation of a ‘reciprocal obligation’
rhetoric and advocacy of income tax credits to improve incentives.
The rational actor framework espoused by economists leads them to emphasise either
the apparent lack of skills to enable the unemployed to compete in the labour market
or a lack of incentives to take up paid work.
Australian Third Way advocates seem to have absorbed ideas from a range of
viewpoints, often referring to a ‘paid inactivity’, a lack of reciprocity, ‘negative
welfare’ and the passive welfare state (Latham, 2001a; 2001b; Pearson, 2001).
1
2
Although I personally prefer the term social security to welfare, to avoid confusion
and to maintain a level of consistency with current terminology, this paper will use
the latter term.
For a more comprehensive comparison of welfare reform that reflects the great range
of welfare regimes see, for example, Esping-Andersen (1996) Giersch (1997) and
Lødemel and Trickey (2001).
Henman, P. (2002), ‘The poverty of welfare reform discourse’, in T. Eardley and B. Bradbury, eds,
PAUL HENMAN
This diverse range of Australian welfare reform discourse is repeated in other
countries. Lawrence Mead’s (1986; 1997) call for obligation has been influential in
the USA; former American President Clinton sought to improve incentives to ‘make
work pay’; British New Labour has also negotiated a discursive and policy framework
involving obligation and labour market skills in its New Deal (see Grover and
Stewart, 2000); and the OECD has for many years called for the creation of more
active welfare state policies (2001). In a similar way, Anthony Giddens (1998), the
academic Third Way advocate, argues that the welfare state has created moral hazard
and welfare dependency and that it needs to be restructured to enhance obligations
and investment in human capital.
2
Welfare Reform Analytics
Despite the diversity of these welfare reform discourses, they significantly share a
similar analytical focus, a comparable ‘welfare reform analytics’. Each of these
discourses begins by treating the behaviour of welfare recipients and the welfare
recipient population as problematic. In each case, recipients lack something:
obligation, initiative, skills, motivation, entrepreneurial zeal, incentives, work ethic,
normalcy, and so on. The task is then to restructure welfare so that it provides a new
regime of carrots and sticks which then changes the problematic behaviour.
As a consequence of this welfare reform discourse, the problem which welfare reform
seeks to address is clearly identified as being located in individual welfare recipients,
the recipient community and/or the welfare system that co-constructs welfare
recipients and communities. The analytical gaze of the welfare reform analytics is
accordingly blind to the structural nature of the welfare realities it seeks to understand
and respond to and to the activities and behaviour of non-recipients3. In contrast, this
paper shows that structural changes in Australia’s economy and society, not recipient
behaviour nor the welfare system explain much of the increase in recipient numbers
over the last 35 years. .
A further consequence of the welfare reform analytics is that the policy ‘solutions’
that are identified focus on changing welfare recipients, in their behaviour or skill
levels. The efficacy of policy aims is apparently self-evident.4 In contrast, the
questions of whether changing the structural realities would be more effective or
whether changing recipient behaviour without addressing the structural realities will
be ineffective are not even raised, let alone seriously considered.
As a consequence of the focus of the welfare reform analytics, policy change almost
universally involves increasing the governmental requirements on recipients (or some
section of them) with some approaches adding in a dose of government expenditure
for good measure. Under the threat of their already inadequate level of income,
recipients are obliged to be more active and jump more hurdles. The resulting policies
3
4
This is assuming that a dichotomy between recipients and non-recipients can be
clearly established, which it cannot. For example, many recipients combine part-rate
benefit with part-time wages, welfare is provided to low-income families as in-work
benefits, and many others receive tax benefits in lieu of welfare benefits.
To be sure, the policy objectives of much welfare reform discourse are unclear.
Although in some cases this probably reflects lack of intellectual clarity, in other
instances the ambiguity is intentional as part of a political tactic.
181
THE POVERTY OF WELFARE REFORM DISCOURSE
thus reinforce the social and economic disparities between those in full-time paid
work and those that are not.
3
The Structural Realities of Welfare
Despite the unerring focus of welfare reform discourse on problematising and
changing the behaviour of welfare recipients, this paper demonstrates that the
structural realities of social and economic change explain much of the increase in
recipient numbers. This will be shown through an analysis of the change in recipient
numbers since 1965.5
Figure 1: Proportion of Population of Workforce age Receiving Income Support
Payments, 1965-1998
20%
18.4%
17.5%
18%
18.4%
15.8%
15.1%
Proportion of workforce age
people in receipt of Social
Security payments
Unemployment rates (August
original ABS)
16%
14%
13.0%
11.0%
12%
10.5%
9.4%
10%
9.9%
8%
8.1%
8.0%
7.8%
5.0%
6%
3.2%
3.5%
3.7%
5.7%
5.9%
5.7%
4%
2.4%
2%
1.2%
0%
1965
1.6%
1.7%
1968
1971
1974
1977
1980
1983
Source: Newman, 1999
5
This section of the paper draws heavily on Henman (2001).
182
1986
1989
1992
1995
1998
PAUL HENMAN
The implications of this finding are that policies that seek to change recipient
behaviour without structural change are misplaced and will be marginally effective at
best, and that improved labour market outcomes for recipients can only be obtained
through cross-portfolio policy change involving multiple levels of government and
actively engaging the community and business sectors.
Presenting its case for welfare reform, both the Government and the Reference Group
on Welfare Reform repeatedly presented graphs depicting a considerable increase in
recipient numbers since 1965 (see FaCS, 1999; Newman 1999; RGWR 2000b).
Furthermore, levels of recipient numbers were often graphed alongside official
unemployment or ‘jobless’ statistics to allegedly show that the increase in recipients is
only partially due to increases in unemployment. (See, for example, Figure 1
reproduced from Newman, 1999). The apparent implication is that the discretionary
behaviour of recipients to withdraw from the labour market accounts for the
difference. Unfortunately, the arguments presented and conclusions reached by the
Government and the Reference Group are fallacious.
Rather, as the Australian Bureau of Statistics has recently observed,
[t]his increase [in income support receipt by the workforce-age
population] is largely associated with: sustained declines in full-time
employment and increased levels of unemployment, particularly
among young people and those just below retirement age; increasing
proportions of people without partners in general, and lone parents in
particular; and increasing levels of education participation among
young people (ABS, 2001: 106)
Thus, in interpreting changes in the recipient population, a number of important
factors need to be isolated and assessed. These are economic cycles, artefactual
distortions resulting from policy change, labour market transformations and
demographic changes.
Economic Cycles
It is well known that economic cycles affect aggregate levels of employment. The
economic downturns of the late 1970s, early 1980s and early 1990s led to reductions
in employment and increases in unemployment (see Figure 1). These cyclical
increases are also evident in recipient numbers (see Figure 1). The dynamics of the
economic cycle would suggest that unemployment numbers should return to prerecession levels when growth returns. However, there has been a ratcheting-up of the
‘natural rate of unemployment’ which is largely related to structural changes in the
labour market and economy.
Artefactual Distortions
The first problem with a straightforward examination of recipient numbers is that
policy changes, including the increasing generosity and coverage of payments, alter
recipient numbers without a change to recipient behaviour or social and economic
realities. In interpreting changes in recipient statistics, these represent artefactual
distortions. This is because it is policy change, rather than underlying changes in
recipient behaviour or social and economic realities, that is being measured. Figure 2
illustrates changes in recipient statistics when the two largest distorting factors are
183
accounted for, namely, the need to include Service Pensions to war veterans, their
wives and widows (Adjustment 1), and the removal of Parenting Allowance to lowincome households (Adjustment 2). Service Pensions need to be included, as they
have been to some extent a substitute for Disability Support Pension in the past,
whereas recipients of Parenting Allowance for low-income households must be
removed because the policy change is a form of in-work benefit and it involves an
increase in benefit generosity.
Although these adjustments do not challenge the large historical upward trend in
recipient numbers, they do significantly suggest that the rise in recipient numbers
from the recession peak of the early 1980s to the peak in the mid 1990s is less
pronounced than at first suggested. Furthermore, the adjusted data indicate that the
measured rise in recipient numbers from 1995 solely results from the introduction of
the in-work benefit of Parenting Allowance to low-income households. Once this
distortion is removed, there has been a decline in recipients since 1995.
Figure 2: Income Support Recipients as a Percentage of the Workforce age
Populations
% of workforce age population
20%
15%
10%
5%
0%
1965
1970
1975
1980
Income Support Recipients
Adjustment 2
1985
1990
1995
Adjustment 1
Adjustment 3
Source : Author’s calculations from published statistics from the Department of
Social Security and of Family and Community Services.
Labour Market Transformations
Since the 1960s, there have been significant transformations in the labour market.
Part-time work, particularly for women has grown, while full-time employment
especially in male manufacturing jobs has declined. Female participation in the labour
market has dramatically increased, while the youth labour market has been decimated.
Deregulation, downsizing and outsourcing have also increased the existence of
precarious – that is, casual, part-time and contract – employment.
The enormous growth in part-time work has especially impacted on recipient
numbers. Because the growth in part-time work has been accompanied by a loss in
184
2000
PAUL HENMAN
full-time work, this has involved an increase in persons combining part-time wages
with part-rate benefits, rather than a transfer of full-rate recipients to part-rate
recipients. In short, growing part-time work has exacerbated recipient numbers, rather
than providing avenues for people to partially move off benefits. Figure 2, Adjustment
3 indicates the effect of removing from the recipient population people who combine
part-time work with part-rate benefits. Recognition of the existence of such people is
extremely important when comparing recipient statistics with unemployment statistics
as was done by the Government (see Figure 1). This is because a person who is
employed for one hour a week or more is regarded as employed and not unemployed.6
A further problem concerning the comparison of recipient statistics with
unemployment statistics is that the latter are not a good reflection of the number of
people who would like work but are unable to find it. In particular, people who would
like work but who are not currently looking for work (such as discouraged jobseekers)
are not included in the statistics, nor are those who want work but were not able to
begin work immediately (for instance, those who are sick, are full-time students or
need to make prior work arrangements). The Australian Bureau of Statistics defines
such people as having ‘marginal labour force attachment’.
For these reasons, unemployment statistics do not provide an accurate picture of the
extent of structural joblessness – that is, the number of people who want work but
have not been able to obtain it. Figure 3 shows that since 1977, when the first
statistics were collected, the number of people marginally attached to the labour
market has almost consistently been greater than the number of official unemployed
persons, and has oscillated between six and eight per cent of the workforce age
population. In addition, people who want work but who are not available to start work
in the next four weeks are regarded as not in the labour force. These individuals have
been separately counted by the ABS since 1983 (the dotted area)
Figure 3 compares the total number of jobless persons who want work with the
number of full-rate income support recipients (reproduced from Adjustment 3, Figure
2). Part-rate recipients have been excluded from the comparison because as they have
part-time work, they are officially regarded as employed. The comparison shows a
remarkable level of concurrence between the number of jobless income support
recipients and the number of jobless who want work.
These statistics, however, must be interpreted with caution. In particular, not all of the
jobless who want work are welfare recipients. The former category includes, for
example, people with partners in the labour force and full-time students.
Counterbalancing this effect is the observation that many income support recipients
do not currently want work, such as disabled adults or full-time carers of children or
frail/disabled adults. Despite the need for caution when interpreting the statistics in
Figure 3, they strongly support the view that many income support recipients want
work, but are unable to find it. This means that structural factors in the labour market
accounts for most of the increase in welfare receipt rather than lack of incentive to
work or welfare dependency.
6
Australia’s unemployment statistics are based on a national survey on labour force
participation and not on benefit numbers. The survey defines a person as unemployed
if they are 15 years and over, currently available and looking for work and did not
work for one hour or more in the week prior to being surveyed (ABS, 2000).
185
Demographic change
The foregoing analysis highlights the importance of structural unemployment and
underemployment as primary reasons for the long-term increase in welfare recipients.
However, this outcome is not simply a result from falling (male) full-time
employment. Overall, labour force participation has increased over the last few
decades, particularly by married women. Changes in recipient numbers and structural
unemployment are thus outcomes from a complex set of changes in the labour market
interacting with changes within households.
People tend to partner others with a similar educational level, socio-economic status
or class. From after World War II to the mid 1970s, the breadwinner model was the
dominant form of household organisation.7 The male was employed full-time while
his female partner was responsible for maintaining the household and rearing their
children. Because husbands were assumed to provide financial support for their
wives, these households did not receive income support. (Schematically, see the lefthand side of Table 1.) Since the mid 1970s recession that ended full employment, the
key changes in the labour market include a loss of low-skilled, low-paid ‘male’ jobs,
especially in manufacturing, and an increase in high-skilled, high-paid jobs. This has
resulted in a loss of jobs for men in low-skilled or working class households and an
increase in employment by women, particularly in high-skilled households. Even if
the number of (full-time) jobs had remained constant, the net result would be an
increase in welfare recipients due to the loss of employment in low-skill households.
In short, there has been a net transfer of jobs to high-skilled or middle class
households to the cost of low-skilled, working class households, resulting in increased
welfare numbers. (See right-hand side of Table 2.)
To further complicate matters, since the mid 1970s, there has been a shift from twoadult to single-adult households. This change is due to a number of reasons, including
marriage and relationship breakdown and divorce, a greater social acceptance of sole
parents, delayed partnering due to extended education and an increase in people
deciding not to partner. The shift to single adult households by itself is likely to
increase welfare receipt. This is because of the statistical reality that single adult
households are more likely to have no adults in work – and therefore require welfare –
than two adult households are. This is schematically represented by the dotted lines on
the right-hand side of Table 1.
In summary, most of the increase in recipient numbers from the mid 1960s occurred
during the period 1971-1983. The key reasons for this growth include reductions in
low-skilled (male) full-time employment and the increase in single-adult households.
7
As Australia’s benefit system only recognises heterosexual relationships, same sex couples
continue to be treated as independent individuals.
186
Figure 3: Full-rate Income Support Recipients and Jobless Who Want a Job as a Percentage of
Workforce Age Population
20
15
% of workforce age
10
5
0
1965
1970
Unemployed
1975
Marginal labour force
1980
1985
1990
Want work - not immediate start
1995
Full-rate recipients
Source: Author's calculations from published social security statistics and from ABS statistics (Catalogue Nos. 2203.0, 2220.0).
.
Table 1: Changes in Household Structure, Employment and Income Support Receipt
Pre mid-1970s
Post mid-1970s
High skilled household
♂ full-time job
♂ full-time job
♀ unpaid domestic work
no income support
no income support
♀ full-time or part-time job
no or part income support
Low skilled household
♂ full-time job
♂ unemployed
1 income support payment
no income support
1 income support payment
To further complicate matters, since the mid 1970s, there has been a shift from two-adult
to single-adult households. This change is due to a number of reasons, including marriage
and relationship breakdown and divorce, a greater social acceptance of sole parents,
delayed partnering due to extended education and an increase in people deciding not to
partner. The shift to single adult households by itself is likely to increase welfare receipt.
This is because of the statistical reality that single adult households are more likely to have
no adults in work – and therefore require welfare – than two adult households are. This is
schematically represented by the dotted lines on the right-hand side of Table 1.
In summary, most of the increase in recipient numbers from the mid 1960s occurred
during the period 1971-1983. The key reasons for this growth include reductions in lowskilled (male) full-time employment and the increase in single-adult households.
4
Policy Implications
The identification of the structural changes outlined above is commonplace. However,
their location in welfare reform discourse is varied. For example, former Minister
Newman directly linked these structural changes to increasing welfare receipt:
as a result of these pressures [of labour market restructuring,
increased labour market participation of women, population ageing
and early retirement and the unequal distribution of work], the past
20 years has seen a large increase in the proportion of workforceage people on social security payments (Newman, 1999).
188
PAUL HENMAN
In contrast, the Reference Group on Welfare Reform refers to such structural change to
argue that the current welfare system is anachronistic (RGWR, 2000c), a view also evident
in the Third Way literature.
Even when the enormous effect of such structural transformations on recipient numbers is
recognised, welfare reform advocates proceed to espouse policy changes that completely
ignore such transformations. Indeed, given that these transformations account for virtually
the entire increase in welfare receipt, any attempt to reduce recipient numbers without
responding to these transformations is simply playing around the edges and doomed to fail
to produce any substantial outcome.
Without job growth, increasing the activity or skills of the unemployed, the obligations of
welfare recipients and incentive regimes, will only exacerbate a sense of failure and
worthlessness, the fragility of their income and divisiveness between the haves and the
have-nots. The best that can be hoped for is a reduction of the time people are out of work
by rotating a greater pool of people in and out of work.
A further illustration of the poverty of the welfare reform discourse is its failure to identify
key policy initiatives that could be undertaken even within the reformers’ narrow focus.
In particular, unemployment can occur when there is a skills mismatch - that is, a shortage
of appropriately skilled persons to take available jobs.8 No Australian welfare reform
document identifies the areas of skills mismatch nor analyses the extent of this problem.
Accordingly, training policies can not be devised to address this problem, and so labour
market programs are unable to maximise their impact. Similarly, geographical mismatch,
whereby skilled unemployed persons are located in different areas from job vacancies that
might use their skills, has not been analysed nor have remedial policies been identified.
Welfare reform deliberations have not considered the ways in which employers may
discriminate against the long-term unemployed, older workers, Indigenous persons or
those with non-English speaking backgrounds, thereby creating barriers for people trying
to enter or return to the labour force. Furthermore, welfare reform discourse fails to
engage seriously in the question of the adequacy of welfare benefits and to recognise that
lack of income can undermine people’s jobsearch and reskilling activities, their health and
thus their capacity to rejoin the labour market. Regardless of the failure of welfare reform
discourse to recognise that the structural realities are a primary policy concern, these
shortcomings suggest that welfare reform has been stifled by a framework emphasising the
inactivity and inability of the jobless and the passiveness and lack of incentives of existent
welfare policy.
It is of course possible, that some advocate welfare reform not in an effort to reduce
welfare expenditure or the poverty associated with joblessness, but as a political tactic.
Wedge politics is alive and well in Australia (and elsewhere) and marginalising and
demonising welfare recipients is likely to gain support from tabloid newspapers or from
voters with downward envy (Gunn, 1996). To the extent that this is the case, then policy
outcomes are irrelevant.
Assuming that policy, rather than political, outcomes are desirable, when identifying
policy responses to the current state of welfare, it is first necessary to understand the
nature of the policy problem to be addressed. In much of the welfare reform discourse, the
significant increase in recipient numbers is taken as prima facie evidence that the welfare
8
Despite the focus on a lack of skills, there is some evidence to suggest many people are overqualified for their jobs.
189
state is not working, that it is broken and in need of repair; that welfare has created moral
hazards, fostered dependency and delivered rights without obligations. However, nothing
can be further from the truth. As has been shown, it is the structural social and economic
transformations that have created the increase in recipient numbers. Consequently, by
financially supporting those who have been involuntarily excluded from the labour
market, the welfare system is working and is far from broken. Put another way, the
increase in recipients is the consequence of labour market policies of deregulation and
downsizing, public service disinvestment and modest fiscal and monetary policies.
Indeed, the safety net of the welfare state enables capitalism to become more laissez faire
and to go global. As a consequence, the current crisis is not to do with an allegedly
antiquated and inappropriate welfare state, but stems from the rise of poverty, inequality
and social exclusion that has resulted from decades of structural change.
This analysis, therefore, highlights the importance of identifying policies that seek to alter
the structural transformations to create greater opportunities for people to move from
welfare to work. In contrast to welfare reformers’ concentration on the supply side, we
need to give primary attention to the demand side of the labour market. This also means
that the policy focus must widen from the isolated treatment of welfare policy to consider
the interconnectedness of policies and institutions that constitute the nature of the labour
market. Such policies must involve monetary and fiscal settings, taxation, industrial
relations, employment, public services and infrastructure investment, research and
development, regional development and education. It will necessarily involve what British
New Labour calls ‘joined up government’. It will necessarily involve a critical rethinking
of neo-liberal economic policies to ensure a greater mutuality of social and economic
objectives. It also means that instead of accepting the globalisation rhetoric that says
nation states are helpless to act, we actively engage in a politics that can shape economic
globalisation to achieve more equitable and democratic outcomes Mishra, 1999; Kerr,
2001). Welfare reform discourse therefore needs to widen its horizons, as William
Beveridge sought to do, to actively construct a social and economic environment that can
realistically achieve the aims of welfare reform.
Fortunately, there are many well-qualified authors who have already identified and
advocated policies that can create jobs and improve the conditions for the most
disadvantaged, and are appropriate for today’s realities (for instance, Rees et al., 1993;
Langmore, 1994; Morris, 1996; Galbraith, 1998; Stilwell, 2000). The task now is to see
them to fruition.
References
Australian Bureau of Statistics (2000), Labour Force, Australia, Catalogue No. 6203.0, ABS,
Canberra.
Australian Bureau of Statistics (2001), Australian Social Trends 2001, Catalogue No.
4102.0, ABS, Canberra.
Australia, Department of Family and Community Services (1999), ‘Trends in pension and
benefit receipt’, Research FaCS Sheet, No. 2, Canberra.
Esping-Andersen, G., ed (1996), Welfare States in Transition: National Adaptions in
Global Economies, Sage, London.
Galbraith, J. K. (1998), Created Unequal: The Crisis in American Pay, Free Press, New
York.
Giersch, H., ed (1997), Reforming the Welfare State, Springer, Berlin.
Giddens, A. (1998), The Third Way: the Renewal of Social Democracy, Polity Press,
Cambridge.
190
PAUL HENMAN
Grover, C. and J. Stewart (2000), ‘Modernising social security? Labour and its welfare-towork strategy’, Social Policy and Administration, 34(3), 235-252.
Gunn, M. (1996),‘Downward envy: why middle Australia hates the poor’, The Australian:
Weekend Review, 2-3 November, 1-2.
Henman, P. (2001), ‘Deconstructing welfare dependency: the case of Australian welfare
reform’, Radical Statistics, No. 79.
Kerr, D. (2001), Elect the Ambassador!: Building Democracy in a Globalised World, Pluto
Press, Sydney.
Langmore, J. (1994), Work for All: Full Employment in the Nineties, Melbourne
University Press, Melbourne.
Latham, M. (2001a), ‘Stakeholder welfare’, presented at the International Conference on
Asset-Based Welfare, January, Institute of Public Policy Research and Centre for
Social Development, London.
Latham, M. (2001b), ‘Making welfare work’, in P. Botsman and M. Latham, eds, The
Enabling State: People Before Bureaucracy, Pluto Press, Sydney, 115-131.
Lødemel, I. and H. Trickey (2001), ‘An Offer You Can’t Refuse’: Workfare in
International Perspective, Policy Press, Bristol.
Mead, L. (1986), Beyond Entitlement: the Social Obligations of Citizenship, Free Press,
New York.
Mead, L. ed (1997), The New Paternalism: Supervisory Approaches to Poverty, Brookings
Institution Press, Washington, D.C.
Mishra, R. (1999), Globalization and the Welfare State, Edward Elgar, Cheltenham.
Morris, P. ed, (1996), Jobs and Justice: a Public Service Agenda for Full Employment,
Lawrence and Wishart, London.
Newman, J. (1999), The Challenge of Welfare Dependency in the 21st Century, Discussion
Paper provided to the Welfare Reform Reference Group, Department of Family and
Community Services, Canberra.
Organisation for Economic Cooperation and Development (2001), Labour Market Policies
and the Public Employment Service, OECD, Paris.
Pearson, N. (2001), ‘Rebuilding Indigenous communities’ in P. Botsman and M. Latham,
eds, (2001), The Enabling State: People Before Bureaucracy, Pluto Press, Sydney,
132-147.
Prime Minister (1994) Working Nation: Policies and Programs, Australian Government
Publishing Service, Canberra.
Rees, S., G. Rodley, and F. Stilwell, eds, (1993), Beyond the Market: Alternatives to
Economic Rationalism, Pluto Press, Sydney.
Reference Group on Welfare Reform (2000a), Participation Support for a More Equitable
Society: Interim Report, Department of Family and Community Services, Canberra.
Reference Group on Welfare Reform (2000b), Interim Report of the Reference Group on
Welfare Reform: Technical and Other Appendices, Department of Family and
Community Services, Canberra.
Reference Group on Welfare Reform (2000c), Participation Support for a More Equitable
Society: Final Report, Department of Family and Community Services, Canberra.
Stilwell, F. (2000), Changing Track: A New Political Economic Direction for Australia,
Pluto Press, Sydney.
191
Sensitivity
of
Australian
Income
Distributions to Choice of Equivalence
Scale: Exploring Some Parameters of
Indigenous Incomes
Boyd Hunter (Centre for Aboriginal Economic Policy Research, Australian
National University)
Steven Kennedy (Research School of Social Sciences, Australian National
University, Australian Bureau of Statistics)
Daniel Smith (Australian Bureau of Statistics)
1
1
Background
Poverty and inequality studies almost always use an equivalence scale to adjust raw
income to account for the cost of maintaining households and families. These costs
are said to vary with household size and composition, and sometimes the number of
employed in the household and other household characteristics. Unfortunately,
variations in the assumptions about the relevant costs, and the relative complexity of
the transformations involved, mean that measures of equivalent income are difficult
to compare directly. For example, different groups will be classified as poor
depending on which equivalent income is used. The purpose of this paper is to
understand better the calculation of equivalent income and to identify whether a
particular group, Indigenous persons, are being re-ranked by several widely used
measures. The paper also examines to what extent Indigenous people move along the
overall distribution of Australian income when different measures of equivalent
income are applied.
One of the most widely cited international studies of poverty claims that
‘equivalence scales have in general no great effect on the rank order of measured
inequality across countries as long as average family size is not extremely large’
(Buhmann et al., 1988). Since the rationale for choosing a specific scale is rather
vague, the importance of testing the sensitivity of income inequality estimates to the
choice of equivalence scale has long been acknowledged especially where there are
substantial differences in family size and composition (Aaberge and Melby, 1998;
Coulter, Cowell and Jenkins, 1992; De Vos and Zaidi, 1997; Lancaster and Ray,
1998; Schiepers, 1992). The qualification about family size by Buhmann et al., is
1
The views expressed in this paper are those of the authors and do not necessarily
represent the views of the Australian Bureau of Statistics. Where quoted or used,
they should be clearly attributed to the authors
Hunter, B., S. Kennedy And D. Smith (2002), ‘Sensitivity Of Australian Income Distributions To
Choice Of Equivalence Scales: Exploring Some Parameters Of Indigenous Incomes’, in T. Eardley and
B. Bradbury, eds, Competing Visions: Refereed Proceedings Of The National Social Policy Conference
2001, SPRC Report 1/02, Social Policy Research Centre, University Of New South Wales, Sydney,
192-222.
BOYD HUNTER, STEVEN KENNEDY AND DANIEL SMITH
clearly not valid in the context of Indigenous people who live in family groupings
which tend to be very different from those of other citizens.
Daly and Smith (1995) suggest that Indigenous families are experiencing substantial
and multiple forms of economic burden in comparison with other Australian
families. Indigenous people live in larger households than do other Australians and
have smaller incomes, which, once corrected for household size, are much smaller
incomes. Indigenous households are more likely to be comprised of more than one
family compared with other Australian households. They are also more likely to be
multi-generational with older Indigenous people living with younger people in
extended family households. Adult mortality is another important factor driving
family formation (and dissolution) among the Indigenous households, with many
children forced to live with other relatives or friends (Gray, 1990). The complexity
of extended family formations is matched by equally complex definitions of
parenting and related child care arrangements.
These household characteristics have economic implications: for example,
suggesting that older generations are not having to survive independently but remain
ensconced within an extended family network (Smith and Daly, 1996). This can
have benefits for a household, especially as aged adults are often in receipt of
reliable sources of pension income and provide childcare and stability to household
membership. However, there may also be economic disadvantages to these social
arrangements with many Indigenous households being overly dependent on kin with
low welfare-based incomes. These welfare recipients may, in turn, be under
substantial economic pressure from other adults and children who are dependent
upon their incomes (Finlayson, 1991; Rowse, 1988). Such dependants may not
qualify as dependents under Centrelink criteria and may be particularly at risk in
these economically vulnerable households.
Since extended family formations are the norm in many Indigenous communities, it
is necessary to consider the level at which income analysis should be conducted. The
household and family are notoriously difficult to define, especially in a cross-cultural
context. The Australian Bureau of Statistics defines a ‘household’ in the national
population census as:
[A] group of people who reside and eat together (in a single
dwelling) … as a single unit in the sense that they have common
housekeeping arrangements - i.e. they have some common
provision for food and other essentials of living. (ABS, 1991:
60)
Persons living in the same dwelling, but with separate catering arrangements, can
therefore be classified as separate households. However, the identification of such
households can be particularly problematic where people: are living in improvised
dwellings, share domestic resources across dwellings, are highly mobile or have
large flows of visitors (Daly and Smith, 1995; Finlayson, 1991; Martin and Taylor,
1996; Smith, 1992; Smith and Daly, 1996). One innovative attempt to capture the
dynamic complexities of Indigenous households is being implemented in surveys of
193
SENSITIVITY OF AUSTRALIAN INCOME DISTRIBUTIONS TO EQUIVALENCE SCALES
two remote communities, Kuranda and Yuendumu (Smith, 2000).2 Notwithstanding
future developments in data collection, this paper is necessarily confined to existing
ABS data sources. The focus on the non-remote Indigenous population in this paper
may somewhat reduce the impact of these factors on our results.
Income and poverty studies often use ‘income units’, which are defined by the ABS
as:
One person, or a group of related persons within a household
whose command over income is assumed to be shared. Income
sharing is considered to take place between partners in a
couple relationship, and between parents and their dependants
(ABS, 1995: 95)
Given the ethnographic evidence about resource sharing in (and even across) large
Indigenous households, the assumption of a nuclear or coupled family is
inappropriate. The nature of the definitions of income units and families means that
it is better to study Indigenous economic activity at the household level.
The distinct nature of Indigenous families and households has led Smith and Daly to
claim:
The census indicator of household income is a more reliable
measure of Indigenous income and status than family income,
given that the census concept of household at least has the
potential to capture extended kin formations via the multifamily household type, than does the discrete family concept.
(Smith and Daly, 1996: 6)
While we endorse this judgment, we also examine the other family grouping that is
widely used in poverty analysis, income units. However, given our preference for the
household concept in the Indigenous context, the analysis will predominantly focus
on the broader grouping of people.
The style of analysis in this paper is deliberately rudimentary, with its focus on the
extent of re-ranking along Australia’s income distribution, in order to understand
better equivalent income calculations. However, before any empirical analysis is
possible it is necessary to introduce some basic theory underlying equivalent income
calculations
2
Kuranda is a small hinterland town with a population of 600 people, about half an hour’s drive
from the urban and tourist centre of Cairns in North Queensland. Yuendumu is a discrete,
remote and predominantly Aboriginal town of 900 people located about four hours drive from
Alice Springs.
In Yuendumu, censuses were taken nightly of all persons staying overnight in households over a
twelve-month period. The dynamic nature of Indigenous household composition can be
illustrated in one of the four-bedroom houses surveyed. A total of 27 different adults and 15
different children slept at the house over the fortnight, totalling 42 different persons. Out of this
flow of 42, a core of 11 persons (seven adults and four children) slept at the house for the whole
two-week period (Musharbash, 2000).
194
A Brief Introduction to Equivalence Scales
To compare economic well being one needs to adjust for the income needs of
households with different characteristics. As indicated above, this can be achieved
by applying equivalence scales to adjust raw household income. Buhmann et al.
(1988) describe four types of equivalence scales: scales developed for statistical
purposes only (for example counting people relative to a given standard of living);
program-based scales; those based on analysis of consumption surveys (indirectly
measuring utility through consumer revealed preference); and subjective scales
which attempt to measure directly utility associated with different income levels or
family characteristics. While each type of scale has different strengths and
weaknesses and a distinct theoretical basis, they can all be represented in a
parametric form to facilitate comparisons between scales.
Equivalence scales are usually applied to raw household income in order to derive
equivalent (or equivalised) household income as follows:
IE =
IH
(1)
Si
where I E is equivalent income, I H raw household income and Si the equivalence
scale. When Si is set equal to one, the scale does not vary with households size or
composition and equivalent income I E equals raw household income I H . When Si
equals the number of persons in the household, I E is per capita income. Equivalence
scales typically result in measures of equivalent income that lie between raw
household income and per capita income. A very useful parametric form of Si is that
proposed by Buhmann et al. (1988):
Si = H θ
(2)
where H is household size and θ a parameter representing ‘the elasticity of scale
with respect to household size’. (Jenkins and Covelly, 1994) Variations in θ
between 0 and 1 result in equivalent income measures that lie between raw
household income and per capita income. For example, Buhmann et al., (1988: 120)
estimated the scale predominantly used in Australian poverty studies (i.e. the
Henderson equivalence scale) had a θ of 0.55.
Whilst this functional form is a very convenient way of capturing a range of
equivalence scales with different inherent economies of size, it does not capture
variation in the composition of households. One simple extension of equation 2 that
captures differences between adults and children is:
Si = ( A + η K )θ
(3)
where A is the number of adults in the household and K the number of children.
The parameter η can be thought of as a relative weight for a child. Variations in θ
between 0 and 1 again capture differences in economies of size. The differential
weighting for children, as opposed to adults, permit the analyst to capture differences
in household composition as well as household size.
195
Of course, other functional forms can also be used that capture both composition
effects and economies of size. A linear alternative to equation 3 is:
SO = 1 + α ( A − 1) + β K (4)
where A and K are defined as above and α and β represent the cost of additional
adults and children. Variation in α and β between 0 and 1 will produce measures of
equivalent income that lie between raw household income and per capita income.
In this paper, equation 4 is used to represent particular equivalence scales and to
generalise the results across a range of possible scales. The two equivalence scales
used in this paper are the Henderson and Organisation for Economic Cooperation
and Development (hereafter OECD) scales.
The Henderson scale arose out of Professor Henderson’s (Commission of Inquiry
into Poverty, 1975) report, Poverty in Australia, and has been used in a wide array of
circumstances. The scale applied in this paper is a modified version of the original
scale and ignores adjustments for housing costs. This modified version of the scale
was used by the ABS in publishing estimates from the 1995 National Health Survey
(NHS). The Henderson equivalence scale was designed in such a way as to be
applicable to income units which, by definition, comprise at most two adults plus
children.
The modified Henderson scale can be represented as follows:
S H = 0.76 + .24( A − 1) + .2 K − 0.14WNotWorking (5)
where WNotWorking is a dummy variable that equals 1 if no adults in the income unit are
working and 0 otherwise. This version of the Henderson scale is normalised (i.e. the
scale equals 1) for income units where there are two adults, no children, and at least
one adult is working.
The OECD equivalence scale used in this paper is a version of equation 4 where α
and β equal 0.5 and 0.3 respectively (i.e., the ‘modified’ OECD scale used in De
Vos and Zaidi 1997, 321). One advantage of the modified OECD scale is that, unlike
the Henderson scale, it can easily be applied to various levels of analysis.
In this paper, we are interested in examining Indigenous and non-Indigenous
populations using both income units and households. As noted, the Henderson scale
was designed to apply primarily to income units. However, it is possible to represent
the relative weight for children (compared to adults) and the underlying economies
of size assumptions embodied in the Henderson scale using the various parametric
forms. For example, if one removes the Henderson scale working adjustment and renormalises the scale so that single adult income units have a scale value of 1, then it
is possible to derive a linear representation (equation 4) of the Henderson scale. We
undertook this estimation for the (normalised) Henderson scale using Ordinary Least
Squares (OLS) and obtained parameter values for α and β of 0.316 and 0.263
respectively. These estimates are useful to consider particularly when examining reranking across Henderson and OECD equivalent income.
196
Though it was not designed to do so, the Henderson equivalence scale can be applied
to households. In this case, we have to decide how to account for the working
adjustment that the Henderson scale applies to income units. If we use the same
assumption (i.e. if no adults in the household are working, then reduce the scale by
0.14), then few households, as opposed to income units, would meet this condition.
Consequently, an incremental working adjustment was deemed appropriate for
households whereby the scale is reduced by 0.07 for each adult that was not working
in the household.
Sensitivity testing of a variety of equivalence scales is a common practice in research
(Burniaux, et al., 1998). For example, Atkinson, Rainwater and Smeeding (1995)
considered income inequality using raw family/household income (that is, no
adjustment for family size: a θ in equation 2 of 0) and per capita income ( θ of 1).
They found that the level of income inequality was higher with these assumptions
than when income was equivalised using the square root of the number of people in
the family ( θ of 0.5).
Sensitivity analysis is usually conducted because of uncertainty about the implicit
assumptions about economies of size and composition embedded in the equivalence
scales. However, there are additional source(s) of uncertainty when measuring
Indigenous income, arising from the appropriate level of analysis and heterogeneity
in the Indigenous population. The question of the level of the analysis is partially
addressed by conducting the analysis at both the household and income unit level.
The issue of heterogeneity arises because of cultural differences within the
Indigenous population. For example, some Indigenous people, especially those in
cities, may live in income units, but others may live in households with resource
commitments to their extended families living elsewhere. Consequently, the search
for a single best measure of equivalent income for Indigenous Australians is
probably a flawed exercise. Notwithstanding, it is useful to highlight the likely
biases involved in the current measures of equivalent income.
2
Data
The 1995 National Health Survey (NHS) was conducted over the period February
1995 to January 1996. Households were randomly selected using a stratified multistage area sample. (see ABS, 1995) The survey obtained information from over 54
000 Australian residents of private and non-private dwellings. The overall response
rate for households was 91.5 per cent.
In addition to the main NHS sample, a supplementary sample of 1 034 Indigenous
persons was obtained. This resulted in a total sample of 2 099 Indigenous persons.
There were 52 763 non-Indigenous persons in the sample.
In this study, we have excluded households where a response to income questions
was not obtained for all persons in the household and/or information about the
composition of the household was incomplete. Visitors to the household and
households in remote areas were excluded because of questions about data
197
availability and quality.3 Remote areas were defined as Statistical Local Areas
(SLAs) where the dwelling density was less than 57 dwellings per 100 square
kilometre.4 After applying these exclusions to income unit data, there were 44 890
non-Indigenous and 1 541 Indigenous persons in the sample (Table 1). Applying the
exclusions to household data further reduced the sample sizes to 44 562 and
1 451 respectively.
At the time of the 1996 Census, about 76 200 Indigenous people lived in remote
areas, defined by using a similar criterion to that above (authors calculations based
on Hunter, 1998). While the remainder of the excluded Indigenous observations in
the NHS arose largely from the lack of adequate income data within income units,
the population represented by the sample is still a substantial fraction of the total
population. For example, the Indigenous population represented by the NHS sample
was about 73 per cent of the overall 1996 Census count.
Table 1: National Health Survey Samples Sizes (after exclusions): 1995
Non-Indigenous
Indigenous (original)
Indigenous (supplement)
Weighted
Non-Indigenous
Indigenous (original)
Indigenous (supplement)
Weighted
Indigenous
Non-Indigenous
Persons in income units
44 890
859
682
265 600
14 833 200
Persons in households
44 562
778
673
260 600
14 807 200
The measure of income on the 1995 National Health Survey was gross personal
income from all sources. The estimates of family or household income are based on
two standard ABS definitions. The broader definition of households outlined above
includes all persons who are usual residents of the particular dwelling. Under the
ABS definition of income units, dependants are defined as
All persons under 15 years, and persons aged 15-24 years who
are full-time students, live with a parent, guardian or other
relative and do not have a spouse or offspring of their own
living with them. (ABS 1997: 51)
For the purposes of this study, we define children in a manner that is consistent with
the definition of dependants in income units. That is, any full-time student aged
between 15 and 24 years is classified as a dependant. This assumption, which is an
attempt to facilitate comparisons between income measures, can be contrasted to the
3.
The data on visitors was frequently incomplete because the ABS did not collect data on
people who were not usual residents. The data quality and comparability of data in
remote areas has been questioned, especially in the context of the NHS (Altman and
Hunter 1998; Gray 1997; Hunter 2001).
4.
In 1995, there were 156 000 persons living in remote areas of whom 68 400 were
Indigenous persons, see ABS (1999).
198
OECD scale where only those aged 14 years or less are treated as dependent
children.
The household sample contains fewer people than the income unit sample because
the exclusion restrictions remove people living in households where someone did not
complete the income question, even when some income units in the household
provided all the required information. Problems arising from the income question
within households were negligible among the non-Indigenous sample and only
reduced the Indigenous population represented in the NHS by 1.9 per cent. The
higher non-response of the Indigenous population is probably driven by their larger
and more complex household structures which increase the probability that at least
one person failed to provide all the information required.
Describing Indigenous and non-Indigenous Households in the Context of
Income Measurement Issues
The distinct nature of Australian Indigenous households and families is well known.
Despite an extensive literature on household composition, there is little discussion of
how the different structures of Indigenous and non-Indigenous households can be
related to income measurement issues. One essential precursor to such a discussion
is a summary of the joint distribution of adults and dependents in the respective
household types using the most recent census data. Surprisingly, this has not yet
been done in published studies, which instead focus on overall household size or
ethnographic case studies. Table 2 redresses this gap in the literature.
While non-Indigenous households are twice as likely as Indigenous households to
have one adult and no dependants (10.6 per cent and 5.2 per cent respectively), they
are actually less likely to live in households with only one adult. The difference
arises, of course, because single adult Indigenous households are substantially more
likely to have dependants than other households with only one adult in them. This
observation is particularly pronounced among single adult households with four or
more dependants (2.9 per cent and 0.5 per cent respectively).
The second salient point to arise from Table 2 is that Indigenous households are
much less likely than other households to only have two adults in them (45.6 per
cent and 66.9 per cent respectively). This confirms that Indigenous people do not
conform to the dominant paradigm of a nuclear family ‘model’ where there is a
couple living in a single dwelling.
The only two adult households where Indigenous people are more likely to live than
other Australians are those with four or more dependants. This is consistent with the
third stylised fact of Indigenous households that they are more likely to have large
numbers of dependants. The only households with dependants where Indigenous
households were less likely to be concentrated than non-Indigenous households were
those with only two adults in them (i.e. those with between one and three
dependants). Indeed, Indigenous households were only less likely to have two
dependants than non-Indigenous households (20.7 per cent and 20.4 per cent
respectively), but this minor difference is driven solely by the concentration of nonIndigenous dependants in two adult families.
Indigenous people are disproportionately concentrated in households with three or
more adults. More than one-third of Indigenous people live in such households (35.9
199
per cent). In contrast, less than one-sixth of the non-Indigenous population live in
these ‘atypical’ households (16.5 per cent). The substantial part of this difference
arises from particularly large households where there are more than three adults
taking care of four or more dependants: 6.9 per cent and 0.2 per cent respectively.
To recapitulate, Indigenous people are more likely to live in larger households where
there are three or more adults and large numbers of children. One important
exception to this observation is that Indigenous people are substantially more likely
to live in single parent households.
Table 3 presents analogous data for the NHS in order to gauge how representative
the sample was, at least in terms of household composition. Household
characteristics in the NHS reveal that it tends to slightly under-sample the large
households with three or more adults and over-sample single adult households. For
example, the percentage of the Indigenous population living in NHS households with
three or more adults was 32.1 per cent compared to 35.9 per cent in the 1996 Census.
The converse of this is that the NHS was more likely than the census to have
Indigenous households with only one adult (23.2 per cent and 18.5 per cent
respectively).
While the overall differences between the census and NHS data are relatively minor,
the differentials become exaggerated when the necessary data exclusions are
imposed. That is, focusing on the households for which there is valid income data
and who do not live in remote areas further reduces the percentage of Indigenous
households where there are three or more adults from 32.1 per cent to 26.2 per cent.
Presumably, this is because it is harder to collect income information from all adults
in large households and because remote areas
Table 2. Percentage of Population Living in Households by Composition of
Adults and Dependants, 1996 Census (a)
Number
of adults
Number of dependants
0
1
2
3
4 or more Total
Non-Indigenous
10.6
2.1
2.3
1.1
0.5
27.4
10.5
16.3
8.9
3.8
4.3
1.5
1.5
0.8
0.3
6.0
0.8
0.7
0.2
0.2
48.4
15.0
20.9
11.0
4.7
Indigenous
1
5.2
3.3
4.1
3.0
2.9
2
10.0
7.7
10.7
8.8
8.4
3
3.3
2.5
2.9
2.5
2.3
4 or more
8.2
2.5
2.7
2.1
6.9
Total
26.7
16.0
20.4
16.4
20.5
Note. a) At the time of the 1996 Census there were 17 536 300 non-Indigenous people
and 352 800 Indigenous people counted. These counts are adjusted for nonresponse and other factors in deriving final estimates of the resident population.
1
2
3
4 or more
Total
200
16.7
66.9
8.5
8.0
100.0
18.5
45.6
13.5
22.4
100.0
Table 3. Household Composition of the National Health Survey, 1995
Number of adults (a)
0
1
2
3
4 or more Total
Percentage of non-Indigenous households (after exclusions)
1
10.6
1.8
2.6
1.0
0.4
16.5
2
25.0
9.5
15.8
9.3
3.8
63.5
3 or more
11.2
4.3
3.0
1.1
0.6
20.1
Total
46.8
15.6
21.4
11.4
4.8
100.0
Percentage of Indigenous households (after exclusions)
1
5.3
4.7
7.2
4.1
4.1
25.5
2
10.4
9.7
12.8
9.1
6.4
48.4
3 or more
7.0
4.9
1.6
7.7
4.9
26.2
Total
22.7
19.3
21.7
20.9
15.4
100.0
Percentage of Indigenous households (before exclusions)
1
4.9
4.6
6.8
3.4
3.5
23.2
2
9.3
9.0
12.0
7.6
6.9
44.7
3 or more
7.4
5.6
2.9
7.1
9.1
32.1
Total
21.6
19.2
21.7
18.0
19.5
100.0
Notes. a) The weighted populations represented in the three panels are 14 810 000;
260 600; and 345 000 respectively. The third panel relates to households where
exclusions were not enforced. Households with three or more adults are
aggregated because of the small cell sizes in NHS.
have larger households. There are few differences between the distribution of nonIndigenous households in the 1996 Census and the NHS, either with or without
exclusions.
Taken as a whole, the NHS sample is broadly representative of the Indigenous
population. Unfortunately, the exclusions necessary to conduct an analysis of income
lead to a particular sample of Indigenous people, which is not necessarily
representative of the Indigenous population. As a consequence, the role of larger
Indigenous households will be understated by the NHS data. Conversely, the role of
single adult households may be overplayed and this should be taken into account in
the analysis of income re-ranking observed in the NHS.
Other features of the NHS indicate that it is representative of the Indigenous
population, irrespective of whether the exclusions are applied. The Indigenous age
distributions in the NHS and the 1996 Census are virtually identical, with 40 per cent
being less than 15 years of age. At the other end of the distribution, the census and
NHS indicate 3.0 per cent and 2.4 per cent respectively are aged over 65 years. The
non-Indigenous age distribution in the census is virtually identical with that in the
NHS.
While the census does not include any information on income units, but rather on
families, it is worth reinforcing the above story by examining the differential in
composition of Indigenous and non-Indigenous income units. Table 4 illustrates that
the differences between income units replicated the basic differentials identified
above between households. However, while the differences are still substantial they
tend to be compressed for income units, by definition, given that income units are
201
forced to have a maximum of two adults in them. That is, Indigenous income units
are closer to non-Indigenous income units because the non-Indigenous are more
likely to live in two adult households while the Indigenous are disproportionately
concentrated in households with three or more adults. Consequently, the act of
breaking up larger Indigenous households into smaller income units creates more of
both single adult and dual adult income units—thus distorting both. The increase in
the number of single adult units with dependants, relative to the household
distributions is a distinct feature of the Indigenous sample. This is consistent with
ethnographic evidence that indicates Indigenous sole parent families tend to live in
multi-adult households with their extended families.
Table 4. Composition of Income Units in National Health Survey, 1995
0
1
2
3
Number of adults
Non-Indigenous income units
1
25.6
2.8
2.9
1.2
2
24.9
11.1
17.2
9.8
Total
50.5
13.9
20.1
11.0
Indigenous income units
1
19.7
8.9
8.1
5.3
2
10.4
9.5
11.8
13.6
Total
30.1
18.4
19.9
19.0
4 or
more
Total
0.4
4.1
4.5
33.0
67.0
100.0
5.2
7.4
12.6
47.3
52.7
100.0
Another dimension of some equivalence scales is that they attempt to correct for
costs associated with employment. Table 5 provides useful background by
describing the proportion of the population living in households categorised by the
number of people working in them. Indigenous people are much more likely to be
living in households where either no one works or only one person is employed. That
is, in spite of the fact that Indigenous people are more likely to live in multi-adult
households, they are less likely to live with more than one employed person. This
follows, more or less directly, from the disproportionate level of joblessness in the
Indigenous community and has important implications for the extent of re-ranking
from Henderson equivalent income.5
Table 5. Percentage of Population by Number of People Working in the
Households, 1995 NHS
Number of people working in households
0
1
2
3
4
Total
Population
5.
Indigenous
Non-Indigenous
27.1
41.7
23.9
5.8
1.5
100.0
260 600
18.7
29.2
39.2
9.4
3.6
100.0
14 807 200
See Hunter and Gray (1998) for details of employment disadvantage of Indigenous Australians.
202
3
Overview of Indigenous Income Distributions
The relatively low-income status of Indigenous Australians is widely documented
(Altman 2000; Altman and Hunter, 1997; Hunter, 2001). This section presents
similar evidence using the 1995 NHS to set up the analysis of income deciles in the
rest of the paper. Indigenous people are clearly concentrated in the bottom half of the
Australian income distribution for both households and income units, with the
largest cluster tending to be in the bottom two deciles (Figures 1 and 2). The
dominance of the non-Indigenous population in the overall Australian income
distribution mean that their distribution is basically a flat line at the 10 per cent
mark.6
Overall, the distributions are similar, but per capita income tends to classify more
Indigenous people in the bottom decile. This is consistent with the fact that this scale
assumes that larger households have higher costs than other scales and hence such
households are more likely to fall into a lower decile. The differences between the
distributions of raw and per capita income tend to be the most pronounced with the
Henderson and OECD distributions tracking each other reasonably closely.
The distribution of Indigenous income for income units is more variable. This is
consistent with the artificial nature of many income units in the Indigenous context.
That is, re-ranking among the deciles is more pronounced for income units than
households, thus providing prima facie evidence that Indigenous income will be
more sensitive to the assumptions about the economies of size and composition of
the relevant social grouping. Consequently, the failure to get these assumptions right
is likely to lead to higher measurement error in Indigenous income.
Another inconsistency between Figures 1 and 2 is that all measures have relatively
low numbers of Indigenous income units in the bottom decile. It is particularly
unusual that the per capita, OECD and Henderson measures all have fewer
Indigenous in the bottom decile than the second decile. For raw income, there are
actually more people in the third decile income units than in each of the bottom two
deciles. This is unlikely, given that all existing studies of Indigenous disadvantage
point to an overwhelming concentration of economic problems in the bottom two
deciles (especially Altman and Hunter, 1997; Hunter 2001). This suggests that
equivalent income calculated for Indigenous income units does not reflect
Indigenous circumstances and provides further rationale for preferring households to
income units.
Notwithstanding such problems in measuring income unit income, the overall
distributions are reasonably similar. However, analysing the distribution can hide
important issues such as whether the people in the various deciles are the same.
Indeed, the following section will show that a substantial percentage of the
population moves between deciles even when the net changes to the distribution are
quite small.
6.
Given that Indigenous people only comprise two per cent of the total Australian population, the
numbers of non-Indigenous in each decile is more or less ten per cent (plus or minus 0.3).
203
Figure 1. Distribution of Indigenous household equivalent income, 1995
Per cent of Indigenous population
30
25
20
15
10
5
0
1
2
3
4
5
6
7
8
9
Income decile
Total
Henderson
OECD
Per Capita
Ssource: 1995 NHS.
The Extent of Re-ranking Using Several Widely Measures of Income Need
To examine the extent of re-ranking in the NHS, we re-estimate equivalent income
using the various equivalence scales. Given that we are using the same raw income,
the only thing which varies between the measures of equivalent income are the
equivalence scales and their assumptions about the economies of size and
composition of households and income units
Households: Table 6 illustrates the effects of assumptions about economies of size
and composition within households. The first three rows take the extremes of the
assumptions about equivalence scales, raw income and per capita income. By
moving to per capita income and hence assuming the maximum feasible cost for
each extra person, large Indigenous families are likely to be moved down the
distribution into lower deciles. Indeed, over three-fifths of the Indigenous population
were in households that changed to a lower rank (62.0 per cent). In contrast, 42.2 per
cent of the non-Indigenous households moved down the distribution when per capita
income is calculated. This is not surprising given the disproportionate number of
Indigenous people living in large households with many dependants and adults.
204
10
.
Figure 2. Distribution of Indigenous Income Unit Equivalent income, 1995
Per cent of Indigenous population
30
25
20
15
10
5
0
1
2
3
4
5
6
7
8
9
Income Deciles
Total
Source:
Henderson
OECD
Per Capita
1995 NHS
Many Indigenous people live in households whose rank improves after per capita
income is calculated (19.2 per cent). The reason for this is that a substantial
proportion of the Indigenous population live in households where there is only one
adult (although often with dependants). Non-Indigenous people are two-and-a-half
times more likely than Indigenous people to live in households where there is only
one adult and no dependants, and hence it is not surprising they are more likely to
be re-ranked up by the per capita calculation (34.4 per cent). This differential in the
propensity to be ranked higher in per capita income is completely explained by the
different distribution of Indigenous and non-Indigenous population in such
households (Tables 2 and 3).
The net effect on re-ranking of these competing tendencies is that Indigenous people
are slightly more likely to live in households which are re-ranked than other
Australians, with only 18.8 per cent compared to 23.4 per cent remaining in the same
decile in the respective distributions. Clearly, the net effect of re-ranking along the
distribution hides a substantial degree of (gross) distributional change, which in turn
is driven by large differences in household size and composition. This is a major
finding which implies that equivalent income and derived poverty measures may
hide substantial changes in the ranks of the poor. The fact that only one-fifth of
people, Indigenous or otherwise, stay in the same decile when using per capita as
opposed to raw income illustrates the possible extent of the problem.
205
10
Table 6. The Extent of Re-ranking among the Various Equivalence Scales
Households
Income units
Indigenous
NonIndigenous
NonIndigenous
Indigenous
Raw same decile as per capita
18.8
23.4
17.8
21.4
Raw higher decile than per capita 62.0
42.2
57.0
40.8
Raw lower decile than per capita 19.2
34.4
25.2
37.8
Raw same decile as OECD
Raw higher decile than OECD
Raw lower decile than OECD
35.7
48.0
16.3
40.8
31.1
28.1
28.1
50.2
21.8
38.2
32.4
29.4
OECD same decile as per capita 35.3
OECD higher decile than per
capita
48.9
OECD lower decile than per
capita
15.8
41.6
33.3
35.5
32.3
46.5
35.0
26.1
20.2
29.5
85.3
72.9
76.5
7.3
11.6
11.9
7.4
15.5
11.7
Henderson same decile as OECD 85.4
Henderson higher decile than
OECD
6.9
Henderson lower decile than
OECD
7.7
Henderson scale without working adjustment
Henderson same decile as OECD 84.7
Henderson higher decile than
OECD
9.1
Henderson lower decile than
OECD
6.2
85.5
7.2
7.3
These stylised facts are replicated when the analysis focuses on income units. The
fact that households are often artificially broken down into smaller income units
tends to generate more upward re-ranking and slightly less downward re-ranking
than was seen in households. However, the overall observations about the relativities
between Indigenous and non-Indigenous populations and the fact that net changes
tend to hide many (gross) distributional changes remain valid.
In order to disaggregate the extent of re-ranking between raw and per capita income,
the changes between raw and OECD income and OECD and per capita are
presented. Given that per capita assumes higher costs for additional household
members than OECD income, which in turn assumes higher costs than raw income,
one should expect that the tendencies identified above are replicated, although
probably with smaller magnitudes. This expectation is borne out with Indigenous
people more likely to move down the distribution, and less likely to move up, than
non-Indigenous people. Therefore, the phenomenon of gross re ranking along the
distribution appears to be monotonic in the assumption about economies of
household size and composition. Since the variation in assumptions about household
costs is not as large for each of these pair-wise comparisons, it is not surprising that
more people remain in the same decile of the distribution for both populations.
206
Given that raw and per capita incomes represent the extremes of assumptions about
household costs, it is useful to examine widely used measures of equivalent income,
that is, using the Henderson and OECD scales. The Henderson scale implicitly
assumes higher economies of size and composition and is closer to raw income
measures than per capita, with the OECD income being closer to per capita than the
Henderson measures.
There is less difference between Henderson and OECD measures than was apparent
between other types of equivalent income examined. The reason for this is that the
underlying assumptions about the economies of size and composition are reasonably
similar. Among the Indigenous population, 84.7 per cent remain in the same decile,
which is actually higher than the stability indicated in the non-Indigenous
population.
Another notable observation is that the OECD scale is likely to re-rank Indigenous
household incomes, as opposed to other Australian incomes, higher than in the
Henderson distributions (7.7 per cent and 7.4 per cent respectively). At face value,
this result is surprising to the extent that the distribution and composition of
Indigenous households would tend to reverse this relativity. However, the
adjustment for the costs of working, included in the Henderson but not the OECD
scale, means that the Henderson measure of equivalent income for Indigenous
people will already be depressed relative to non-Indigenous incomes because of the
ongoing low levels of Indigenous employment (Hunter and Taylor, 2001).
It is possible to gauge the effect of the working adjustment by examining how the
results change when it is excluded from the Henderson calculations. This
hypothetical can also be rationalised on the grounds that the OECD scale does not
have any equivalent adjustment, thus isolating the effect of the scales attributable to
assumptions about household size and composition. The exclusion of the working
adjustment scale appears to have no or little effect on the extent of re-ranking in the
non-Indigenous population, but it has a significant effect on the Indigenous results,
increasing the extent of movement up the distribution by 2.2 percentage points (with
9.1 per cent, rather than 6.9 per cent, of the Indigenous population being in a higher
decile in the OECD distribution). Furthermore, the Henderson scales that exclude an
adjustment restore the expected relativities in the distribution of household income,
with Indigenous households being slightly less likely than other Australians to be
ranked higher when using OECD measures (6.2 per cent and 7.3 per cent
respectively). Overall, there is relatively little re-ranking between the Henderson and
OECD income distributions with similar percentages of the populace moving up and
down the distribution for both Indigenous and non-Indigenous population,
irrespective of whether the adjustment for the costs of having a job is applied.
Table 7 examines the extent of income re-ranking in more detail by examining where
on the distribution a household is re-ranked after applying per capita income
adjustment. For example, Indigenous households at the extremes of the distribution
are actually more likely to stay in the same decile, with 54.0 per cent staying in the
first decile compared to 39.8 per cent of the non-Indigenous population in the lowest
decile of raw income. At the other extreme, 87.9 per cent of the 5 000 or so
Indigenous householders in the top decile of raw income remained in that decile after
applying the per capita equivalence scales. It is important to note the underlying
small sample sizes for Indigenous estimates in the upper deciles. The fact that
Indigenous people are slightly less likely than other Australians to remain in the
207
same decile overall is explained solely through the greater instability in the middle
deciles.
Both Indigenous and non-Indigenous Australians appear to have distinct clusters of
people some way from the main diagonal in Table 7 (i.e. the entries which indicate
the extent to which people remain in the same decile). Such clusters can be identified
by casting your eyes across the rows and identifying where the percentage figures in
the row increase then fall then increase again. For example, the extent of re-ranking
in the second non-Indigenous decile of raw income increases to the main diagonal
(the second decile of per capita income), declines to zero per cent in the fifth decile
of per capita income before increasing to 5.2 per cent in the seventh decile. This
cluster of people who are re-ranked up four or five deciles from their original place
in the distribution is likely to consist of smaller single adult households, presumably
most of who are without dependants. Such households will be ranked higher in the
per capita income because they do not incur the extra costs of larger households
assumed in this scale. A similar phenomenon is apparent for Indigenous Australians,
albeit with smaller clusters. This is consistent with the substantial number of single,
adult, Indigenous households although it is smaller than in the non-Indigenous
population.
A relatively large number of Indigenous Australians is re-ranked lower in the per
capita distribution. For example, a remarkable 66.4 per cent of Indigenous people in
the second decile of raw income are ranked in the lowest decile of the per capita
distribution. In contrast, only 23.6 per cent of non-Indigenous people in the second
decile are re-ranked downwards. This, as was pointed out above, is because many
Indigenous households are very different from the Australian average household.
Consequently, it is not surprising that there is a large cluster of people who are reranked lower in the per capita distribution. because large households are assumed to
have higher costs than other households; there is a cluster of people, especially nonIndigenous people, who are re-ranked down about four or five deciles. For example,
about one-eighth (12.4 per cent) of Indigenous householders in the ninth decile of
raw income are ranked in the fourth decile of per capita income.
Relatively little re-ranking was evident when comparing Henderson and OECD
income in detail (Table 8). While the re-ranking is likely to include the same groups
of households re-ranked between raw and per capita income, the extent of movement
along the respective income distributions is constrained by the similarity of the
assumptions about household costs. As a result, re-ranking is constrained within a
relatively narrow band within one decile of the original place in the distribution.
Even if there are clusters similar to those identified in Table 7, the focus on deciles is
too broad to identify them. Consequently, Table 6 provides an adequate summary of
the changes in the distributions when using Henderson rather than OECD income.
Appendix One further breaks down the overall re-ranking among households in
Table 7. For example, Tables A1.1 and A1.2 shows the percentage of each decile of
OECD household income that was classified in the various deciles of raw income
and per capita income.
In the first decile of OECD income, almost twice as many people were originally
classified in the bottom decile of per capita income (85.8 per cent compared to 44.5
per cent). In a sense, this decile is closer to the per capita measure than to raw
208
Table 7. Percentage of Households whose Income was Re-ranked between Raw Income and Per Capita Scales, 1995
Decile of Per capita household income
Raw income 1
2
3
4
5
6
7
8
9
10
Total
deciles
Indigenous households
1
54.0
24.7
6.5
8.6
6.1
0.1
0.0
0.0
0.0
0.0
100.0
2
66.4
17.5
11.7
2.9
0.0
0.0
1.4
0.0
0.0
0.0
100.0
3
42.4
33.2
9.7
2.5
4.7
2.6
0.0
5.0
0.0
0.0
100.0
4
25.9
18.1
22.8
14.1
6.7
5.6
4.3
0.0
2.5
0.0
100.0
5
0.2
38.9
15.9
7.9
10.8
6.4
4.8
10.7
0.3
4.1
100.0
6
0.0
12.1
16.0
13.8
31.9
2.4
4.2
10.3
0.6
8.6
100.0
7
0.0
5.3
10.2
34.4
7.2
18.2
17.9
1.7
5.1
0.0
100.0
8
0.0
0.0
0.0
20.3
0.6
17.5
30.0
8.7
10.8
12.1
100.0
9
0.0
0.0
0.0
12.4
7.1
35.9
16.8
5.7
17.3
4.8
100.0
10
0.0
0.0
0.0
0.0
0.0
2.7
5.7
3.7
0.0
87.9
100.0
Total
26.7
18.9
11.2
11.3
7.1
7.6
6.6
3.9
2.8
3.9
100.0
Non-Indigenous households
1
39.8
11.3
12.7
24.1
9.6
2.5
0.0
0.0
0.0
0.0
100.0
2
23.6
23.5
36.4
9.1
0.0
2.2
5.2
0.0
0.0
0.0
100.0
3
24.4
20.2
11.3
11.2
17.7
4.4
0.5
8.6
1.7
0.0
100.0
4
8.9
23.1
12.8
15.5
6.1
15.6
8.6
0.0
9.4
0.0
100.0
5
0.7
16.5
15.9
17.7
13.1
8.4
13.2
7.5
1.0
6.0
100.0
6
0.3
4.2
8.5
10.7
26.2
14.5
8.7
21.6
0.6
4.7
100.0
7
0.0
0.3
2.1
9.0
16.2
27.9
15.3
4.1
22.4
2.6
100.0
8
0.0
0.0
0.1
1.8
9.9
15.8
25.1
21.9
19.2
6.2
100.0
9
0.0
0.0
0.0
0.8
2.2
8.5
21.2
26.8
19.2
21.2
100.0
10
0.0
0.0
0.0
0.0
0.0
0.5
2.5
10.3
27.3
59.4
100.0
Total
9.7
9.9
10.0
10.0
10.1
10.0
10.1
10.1
10.1
10.1
100.0
Population
in each decile
26803
44402
40404
32385
30277
18818
30034
13611
18529
5324
260600
1480101
1493214
1444090
1466009
1486799
1478998
1477487
1492320
1487077
1501114
14810000
Table 8. Percentage of Households whose income was Re-ranked between OECD and Henderson Scales, 1995
Decile of household income adjusted using Henderson scales
OECD
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
Total
99.7
9.9
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
19.8
0.3
87.9
13.9
0.0
0.0
0.0
0.0
0.0
0.0
0.0
16.0
0.0
2.2
73.4
10.0
0.0
0.0
0.0
0.0
0.0
0.0
13.1
0.0
0.0
12.7
79.9
3.4
0.0
0.0
0.0
0.0
0.0
12.5
1
2
3
4
5
6
7
8
9
10
Total
94.5
6.5
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
10.0
5.5
82.3
10.8
0.0
0.0
0.0
0.0
0.0
0.0
0.0
9.8
0.0
11.2
79.7
9.5
0.0
0.0
0.0
0.0
0.0
0.0
9.9
0.0
0.0
9.4
81.1
9.5
0.0
0.0
0.0
0.0
0.0
10.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
10.1
0.0
0.0
86.5
10.1
0.0
28.1
69.8
2.2
0.0
4.4
86.1
0.0
0.0
0.8
0.0
0.0
0.0
0.0
0.0
0.0
12.9
6.1
7.6
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
9.4
0.0
0.0
78.3
12.0
0.2
12.0
77.6
10.5
0.0
10.7
82.3
0.0
0.0
7.2
0.0
0.0
0.0
0.0
0.0
0.0
10.0
10.1
10.0
Total
Population
in each decile
0.0
0.0
0.0
0.0
0.0
0.0
9.5
90.6
0.0
0.0
5.9
0.0
0.0
0.0
0.0
0.0
0.0
0.0
8.5
78.1
2.6
2.9
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
21.9
97.4
3.3
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
47 804
40 779
40 691
33 025
29 299
17 280
22 361
14 494
7946
6908
260 600
0.0
0.0
0.0
0.0
0.0
0.0
7.1
87.8
5.0
0.0
10.1
0.0
0.0
0.0
0.0
0.0
0.0
0.0
5.0
91.8
3.0
10.1
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
3.2
97.0
10.1
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
1 460 127
1 466 935
1 464 697
1 474 032
1 478 288
1 487 916
1 484 292
1 492 412
1 503 701
1 494 809
14 810 000
income. However, with the possible exception of the top two deciles of OECD
income, a substantially greater number of people originates from lower deciles of per
capita income. This is because per capita income assumes lower economies of
household size and composition than either the OECD or raw income.
Notwithstanding, the fact that the whole distributions are recalculated means that
some household types move up the distribution between raw income and OECD
income—with the top two deciles of OECD being more likely to originate from a
lower decile of raw income than a lower decile of per capita income. These
observations are consistent with the differences in Tables A1.1 and A1.2 for the nonIndigenous populations, although these tend to be substantially smaller than the
Indigenous differentials. This confirms that the extent of re-ranking is much greater
for the Indigenous population than for other Australians in every part of the
distribution.
4
Who is Being Re-ranked?
The above analysis is speculative because equivalence scales embody many factors
that can vary simultaneously. While the changes in the decile re-ranking provide
information about where the changes in distribution are occurring, it is useful to
tease out what factors are driving the changes. Consequently, an appropriate multivariate technique is required to capture the extent of re-ranking along the deciles of
the income distribution. The ordered probit procedure is commonly used where there
is a natural ordering (ranking) of a categorical variable (Greene, 2000: 434 - 6). This
regression procedure assumes there exists a latent (unobserved) variable, which
determines whether the rank of a person falls, stays the sameor increases. This
variable is correlated with regressors and hence changes in the values of regressors
may lead to a latent variable crossing a threshold (determined empirically within the
model). The value of the latent variable relative to the thresholds determines the
probability that an income will be re-ranked by various equivalence scales.
The specification of the regressors is relatively easy since it is largely suggested by
the formulae used to derive equivalence scales. The number of adults and children in
a household should obviously be taken into account. While labour force status is
used to define the Henderson scales, its omission from the OECD and other
scalesmeans that the role of employment and other labour force states in driving the
sensitivity of equivalent income can be identified. Basic demographic characteristics
of individuals are also included because differing sources of income may be apparent
at various stages of the life cycle and this may affect the extent of re-ranking. For
example, the reliance of many elderly Australians on fixed sources of income, may
mean they are more likely to be re-ranked as income thresholds change relative to
some benchmark for example pension rates. Finally, information on whether a
person has identified as Indigenous is included in order to test whether there are
residual differences between Indigenous and other Australians after other factors are
taken into account (especially, the substantial differences in household structure and
age profiles; see Taylor, 1997).
Table 9 reports the results of an ordered probit regression model that predicts
changes in the probability, arising from the correlates of re-ranking, between the
Henderson and OECD equivalence scales (see Appendix Table A2.1). Such changes
211
are called marginal effects (ME), and are measured relative to a reference person
defined as a non-Indigenous female aged between 25 and 64 years who is outside the
labour force and lives in a household where there are two adults and no dependants.
Note that the MEs for each row sums to zero because each regressor can only re-rank
a household one of three ways: into a lower decile, into the same decile or into a
higher decile.
Table 9. The Predicted Probability of Re-ranking between Henderson and
OECD Measures for a Reference Person (a) (Marginal Effect in Percentages),
Australian Households: 1995
One adults
Three adults
Four or more adults
One dependant
Two dependants
Three dependants
Four or more
dependants
Aged 24 or less
Aged 65 or more
Male
Employed
Unemployed
Indigenous
Henderson versus OECD
In a lower decile
In the same decile In a higher decile
-11.7
7.5
4.3
11.8
-11.3
-0.6
21.5
-20.8
-0.7
-6.9
5.7
1.2
-8.8
6.8
1.9
-10.2
7.4
2.8
-9.7
7.2
2.4
-9.4
14.4
2.0
-12.4
-13.5
2.7
7.1
-13.8
-1.8
7.2
5.6
-2.5
2.3
-0.6
-0.2
5.2
8.0
-0.2
Notes: a. The reference person is defined as the omitted variable in the ordered
probit regression in Appendix Table A2.1. Omitted variables include: two
adults; no dependants; aged between 25 to 64; female; not in the labour force;
and non-Indigenous.
The pattern of MEs confirms the earlier analysis. Living in a single adult household
means that a household’s Henderson decile rank is likely to be lower than it is in the
OECD ranking relative to the reference person (11.7 per cent). Similarly, such
people are also more likely to be either in the same decile or in a higher decile of the
OECD distribution. The probability of being in a lower decile for OECD income
increases as the number of adults in the house gets larger with the ME peaking at
21.5 per cent in households with four or more adults. The MEs of staying in the
same decile fall commensurately with few of the people who live in large households
(at least large in terms of the number of adults) that shift into a higher decile.
The MEs for the number of dependants in the house are not strictly consistent with
the economies of household size and composition implicit in the respective scales. A
larger number of more dependants in a house tends to be associated with being less
likely to be in a lower OECD decile ranking relative to the Henderson distribution:
people in houses with three or more dependants are about 10 percentage points less
likely to be in a lower decile of OECD income. The explanation for this anomaly is
that there is an interaction between the number of children and the working
adjustment in the Henderson equivalence scale. For example, adults in households
with large numbers of dependents are generally less likely to be working than those
in other households. That is, the effect of the adjustment for differences in working
costs is swamping the fact that the OECD scale assumes that extra children cost
more to maintain than does the Henderson scale.
One revealing aspect of the analysis is that demographic factors can be particularly
important in affecting the extent of re-ranking. The MEs for respondents aged 65 and
over being in a lower decile is only surpassed by the MEs for households with four
or more adults. The likely reason for this is that the OECD decile cut-offs are close
to the relevant entitlements for the age pension. Again, very few in such households
move into a higher decile of OECD income.
The MEs for the labour force status variables reflect the fact that the Henderson
scale includes an adjustment for the cost of working. Accordingly, those employed
are less likely to be ranked in a lower decile of OECD income and more likely to
either stay in the same decile or move to a higher decile.
Once the differentials in household size and composition and other variables are
controlled for, there is relatively little re-ranking of Indigenous people between
Henderson and OECD measures. Only 2.7 per cent of Indigenous people are reranked in a lower decile of OECD income with a commensurate fall in the
probability of remaining in the same decile.
The truncated nature of the dependant variable at the extremes of the distribution
may affect the estimated results in this section. For example, given that bottom
decile households cannot be placed in a lower rank, the characteristics of such
households could potentially bias the estimates. Similarly, households in the top
decile cannot move into a higher decile. One way around this is to estimate the
results for Table 9 excluding the top and bottom deciles. A sensitivity analysis
revealed that the above results were not substantively affected by excluding the
extremes of the distribution.
An analysis of MEs was also conducted to determine whether the correlates of reranking differed for the other measures of equivalent income (Appendix Table
A2.2). The probability of being re-ranked between raw and per capita incomes was
dominated by the effect of household size and composition. Demographic variables
and labour force status were only weakly correlated with re-ranking, if at all. Most
importantly in the context of this paper, Indigenous status is not significantly
correlated with re-ranking; that is, the difference between Indigenous and nonIndigenous populations is being explained almost entirely by household size and
composition.
While the importance of household factors is replicated between raw income and
OECD deciles and OECD and per capita deciles, the other factors tend to be more
important. With regard to the extent of re-ranking between raw income to OECD
income, older people are relatively more stable than persons aged between 25 and 64
years and less likely to be in a higher decile (MEs of 11.6 per cent and -12.3 per
cent). However, these older people tend to be more likely to be in a lower decile of
OECD, and less likely to be in the same decile, compared to the per capita
distribution. That is, being aged 65 and over is associated with a shift down the
distribution at each stage between raw and per capita income, but these MEs tend to
balance out in the overall re-ranking between the extremes of equivalent income.
213
This observation can be made for most of the demographic and labour force status
variables.
The examination of the re-ranking of each stage of the move from raw income to per
capita measures revealed one striking anomaly, the result for the Indigenous
variable. Indigenous status was significantly correlated in Appendix Table A2.1
between per capita and OECD deciles, but not between raw income and OECD
deciles. However, the MEs associated with this association were relatively small.
Given that there was no correlation between the probability of re-ranking and
Indigenous self-identification, this means that the influence of Indigenous variable is
overwhelmed by the influence of household size and compositon.
5
Policy Implications
This paper argues that households should be used in preference to income units (or
even families) as the basic unit of income analysis, especially statistical or policy
analysis of Indigenous Australians. While households are slightly less stable than
income units, at least in terms of measured income, they provide a better basis for an
analysis that directly relates to Indigenous circumstances. In addition to being
culturally inappropriate, income units frequently generate distributions of equivalent
income, which do not accord with widespread consensus about Indigenous poverty.
For example, Indigenous incomes appear to be overstated, with more Indigenous
people being in the second decile of per capita, using the Henderson, and OECD
measures of equivalent income than there were in the bottom decile. In any case,
there is little loss of information for other Australians in shifting the focus towards
the household with non-Indigenous households looking relatively similar to nonIndigenous income units, at least in terms of size and composition. Few nonIndigenous households have more than two adults living in them and hence the
overall income distributions will be relatively unaffected by moving towards a
household level analysis.
Indigenous incomes are harder to characterise and less stable than non-Indigenous
incomes, even if household data are used. This finding is robust and holds for the
various types of equivalence scales. This is consistent with the greater diversity of
household types being more likely to be both single adult and exceptionally large,
(i.e. with three or more adults and/or many dependants).
The large differences between the size and composition of Indigenous and nonIndigenous households mean that the issue of economies of scale in household
production (and consumption) has important implications for the measurement of
Indigenous poverty. Equivalence scales should accurately reflect the real cost of
raising large households. Unfortunately, the range of equivalence scales used by
poverty researchers becomes significantly wider as the number of children increases
(Whiteford, 1985: 13, 106-7). For example, while the values of the Henderson
equivalence scales do not differ markedly from other scales, they do diverge for
large households (Saunders, 1994: 251). Consequently, the choice of equivalence
scale is not trivial and there are no unique solutions to determining the extent of
income inequality and poverty.
Given that an important source of variation in the inequality of equivalent income is
from the number of dependants, it is useful to reflect on the extent to which the
definition accords with household requirements for resources. Clearly, there is a
need for consistency in the treatment of dependants when comparing equivalence
scales. While this consistency was imposed in this paper, more discussion is required
in the literature over which definition of dependants is appropriate. If older
dependants’ cost as much as adults, then the OECD treatment of children should be
preferred to others, such as the Henderson scales, which assume that full-time
students as old as 24 cost the same as children. The OECD definition of dependants
appears to offer a more realistic assessment of resource requirements than the
Henderson definition.
The policy importance of the ‘measurement’ error in income, indirectly identified in
this paper, is that it questions the efficacy of income-based means tests. If income
cannot be accurately used to identify needy families and households, then it is
impossible to guarantee that welfare is being directed to the right people. However,
given that an overwhelming majority of Indigenous Australians experiences multiple
disadvantage across a range of indicators, there is probably no real question about
overall Indigenous need. Notwithstanding, Indigenous families could miss out on
welfare entitlements if their needs are understated by inappropriate definitions of
income.
While this paper is based on national (or almost national) averages, the results
probably hold for a diverse range of Indigenous groups, those living in urban, rural
and remote areas, respectively. That is, the sorts of issues identified between
Indigenous and non-Indigenous Australians are likely to be similar to those between
Indigenous sub-populations. It is not necessary to assume a monolithic Indigenous
culture when evaluating equivalence scales, except perhaps because there are no
adequate data of sufficient quality against which particular sub-populations could be
benchmarked. In the near future, this qualification will be a binding qualification on
the ability to distinguish between groups of Indigenous Australians.
Future work will generalise the results in this paper. This will be achieved by using
single and dual parameter models of equivalence scales, which examine the
continuum of possible equivalence scales. Such models will facilitate the analysis of
the implications of measurement error which is induced by potentially inappropriate
characterisations of equivalence scales for measured Indigenous poverty, especially
Indigenous poverty relative to other Australian poverty.
215
Appendix One
Table A1.1: Percentage of Households whose Income was Re-ranked between Raw Income and OECD Scales, 1995
Decile of household income adjusted using raw household income
OECD
1
2
3
4
5
6
7
8
9
10
Total
1
2
3
4
5
6
7
8
9
10
Total
44.5
10.2
3.4
0.0
0.0
0.0
0.0
0.0
0.0
0.0
10.3
44.9
43.8
10.9
0.7
1.4
0.0
0.0
0.0
0.0
0.0
17.0
8.0
36.2
39.1
10.6
5.1
5.4
0.0
0.0
0.0
0.0
15.5
2.7
9.6
22.4
37.6
11.9
8.2
2.2
2.1
0.0
0.0
12.4
1
2
3
4
5
6
7
8
9
10
Total
71.2
22.5
7.6
0.0
0.0
0.0
0.0
0.0
0.0
0.0
10.0
19.8
52.4
21.8
5.3
2.5
0.0
0.0
0.0
0.0
0.0
10.1
8.9
19.5
32.6
24.9
4.7
6.5
1.3
0.0
0.0
0.0
9.8
0.1
5.0
28.2
27.4
23.8
5.7
5.7
3.5
0.0
0.0
9.9
0.0
0.0
0.0
0.2
0.0
0.0
18.6
5.6
0.0
28.9
11.3
10.9
18.4
23.6
27.3
17.9
10.0
41.1
14.4
7.6
37.7
9.2
5.7
17.1
0.0
20.5
5.3
0.0
0.0
0.0
11.6
7.2
11.5
0.0
0.0
0.0
0.7
0.0
0.0
7.9
1.8
0.1
32.2
9.3
0.8
26.6
27.0
13.2
17.7
28.8
27.3
8.2
17.5
28.8
4.4
10.4
20.8
2.9
4.8
5.8
0.0
0.0
2.6
10.0
10.0
10.0
Population
in each decile
0.0
0.0
0.0
0.0
4.5
9.7
20.6
19.7
39.7
0.0
5.2
0.0
0.0
0.0
0.0
7.9
7.6
16.8
44.1
31.9
32.3
7.1
0.0
0.0
0.0
0.0
0.0
0.0
0.6
2.1
2.5
67.7
2.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
47 804
40 779
40 691
33 025
29 299
17 280
22 361
14 494
7946
6908
260 600
0.0
0.0
0.0
0.0
2.0
11.5
28.9
28.2
27.4
2.0
10.1
0.0
0.0
0.0
0.0
0.2
2.6
9.5
29.7
37.2
20.2
10.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
3.0
22.0
75.2
10.1
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
1 480 101
1 493 214
1 444 090
1 466 009
1 486 799
1 478 998
1 477 487
1 492 320
1 487 077
1 501 114
14810000
Table A1.2. Percentage of Households whose Income was Re-ranked between Per Capita Income and OECD Scales, 1995
Decile of household income adjusted using per capita scales
OECD
1
2
3
4
5
6
7
8
9
10
Total
1
2
3
4
5
6
7
8
9
10
Total
85.8
59.8
10.2
0.0
0.0
0.0
0.0
0.0
0.0
0.0
26.7
8.5
29.6
63.4
22.4
0.0
0.0
0.0
0.0
0.0
0.0
18.9
3.7
6.6
14.8
52.1
5.1
0.0
0.0
0.0
0.0
0.0
11.2
2.0
3.3
8.4
12.9
48.7
30.9
0.0
0.0
0.0
0.0
11.3
1
2
3
4
5
6
7
8
9
10
Total
68.4
24.2
5.8
0.0
0.0
0.0
0.0
0.0
0.0
0.0
9.7
8.5
30.9
46.7
13.5
0.0
0.0
0.0
0.0
0.0
0.0
9.9
12.8
26.2
19.6
35.5
6.3
0.0
0.0
0.0
0.0
0.0
10.0
10.3
14.1
20.1
22.6
30.8
2.5
0.0
0.0
0.0
0.0
10.0
0.0
0.0
0.0
0.7
0.0
0.0
3.3
0.0
0.0
8.8
3.1
0.7
31.7
9.4
1.4
27.1
18.3
18.3
0.0
44.9
26.0
0.0
20.0
51.2
0.0
0.0
1.5
0.0
0.0
0.0
7.1
7.6
6.6
0.0
0.0
0.0
4.7
0.0
0.0
5.1
2.7
0.0
20.3
5.3
2.8
31.7
22.0
6.9
36.3
32.0
22.7
2.7
35.0
33.4
0.0
3.1
32.2
0.0
0.0
2.2
0.0
0.0
0.0
10.1
10.0
10.1
Population
in each decile
0.0
0.0
0.0
0.0
3.8
5.4
26.9
9.0
10.0
0.0
3.9
0.0
0.0
0.0
0.0
0.0
0.0
2.2
11.2
46.1
20.8
2.8
0.0
0.0
0.0
0.0
0.0
0.0
0.0
8.6
42.4
79.2
3.9
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
47 804
40 779
40 691
33 025
29 299
17 280
22 361
14 494
7946
6908
260 600
0.0
0.0
0.0
0.0
2.3
6.0
22.0
39.7
29.1
1.0
10.1
0.0
0.0
0.0
0.0
0.0
0.4
7.0
22.0
54.2
16.5
10.1
0.0
0.0
0.0
0.0
0.0
0.0
0.0
3.0
14.5
82.5
10.1
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
1 480 101
1 493 214
1 444 090
1 466 009
1 486 799
1 478 998
1 477 487
1 492 320
1 487 077
1 501 114
14 810 000
SENSITIVITY OF AUSTRALIAN INCOME DISTRIBUTIONS TO CHOICE OF EQUIVALENCE
Appendix 2
Table A2.1: Ordered Probit Regressions of the Extent of Income Re-ranking to
a Lower, in the Same, to a Higher Decile
Henderson vs Per Capita vs Total vs
Total vs
OECD
OECD
OECD
Per Capita
Adults 1
0.764 (39.4) -0.708 (-39.2) 0.837
(47.0) 1.053
(49.9)
Adults 3
-0.412 (-18.7) 0.650
(33.2) -2.083
(-86.6) -1.886
(-80.3)
Adults 4+
-0.683 (-22.9) 1.282
(43.3) -3.489
(-93.1) -3.950
(-89.1)
1 Child
0.352 (16.2) 1.868
(75.6) -1.004
(-52.4) -1.726
(-80.3)
2 Children
0.481 (22.7) 3.384 (114.7) -2.467 (-104.7) -4.129 (-136.7)
3 Children
0.599 (22.7) 3.910 (113.2) -3.849 (-122.0) -4.574 (-119.5)
4+ Children
0.552 (15.5) 3.872
(91.3) -4.218
(-92.2) -4.596
(-89.7)
Age 24
0.532 (21.8) 0.197
(9.1) 0.246
(11.3) 0.042
(1.7)
Age 65+
-0.489 (-17.7) -0.364 (-13.4) -0.309
(-12.2) 0.162
(5.2)
Male
-0.080 (-5.7) -0.010
(-0.8) -0.048
(-3.8) -0.048
(-3.2)
Employed
0.849 (40.2) 0.278
(15.5) 0.427
(23.6) 0.079
(3.9)
Unemployed
1.047 (26.4) -0.018
(-0.5) -0.263
(-7.3) -0.162
(-4.0)
Indigenous
-0.109 (-2.7) -0.167
(-4.5) 0.002
(0.1) -0.005
(-0.1)
Log
-20 335
-26 504
-27 235
-19 715
likelihood
No of
46 013
46 013
46 013
46 013
observations
Notes: Omitted Variables: Adults 2; Kids 0; Age 25 to 64; nNot in the labour force
Not applicable, Query author, is this order right? non-Indigenous.
218
BOYD HUNTER, STEVEN KENNEDY, AND DANIEL SMITH
Table A2.2: The Predicted Probability (a) of Re-ranking between Various
Income Measures for a Reference Person (ME in Percenatges)
Probability of being re-ranked (%)
In a Lower Decile
In the Same Decile In a Higher Decile
Total versus per capita
Adults 1
-0.2
-17.5
17.7
Adults 3
15.0
50.0
-65.0
Adults 4+
84.8
-5.7
-79.1
1 Child
11.5
49.6
-61.1
2 Children
88.6
-9.4
-79.1
3 Children
95.0
-15.8
-79.2
4+ Children
95.2
-16.0
-79.2
Age 24
0.0
-1.2
1.2
Age 65+
-0.1
-4.3
4.3
Male
0.0
1.4
-1.4
Employed
0.0
-2.2
2.2
Unemployed
0.1
4.8
-4.9
Indigenous
0.0
0.2
-0.2
Total versus OECD
Adults 1
-0.4
-27.6
28.0
Adults 3
30.1
22.9
-53.0
Adults 4+
81.1
-25.5
-55.6
1 Child
5.1
31.1
-36.2
2 Children
44.6
10.0
-54.6
3 Children
89.1
-33.5
-55.6
4+ Children
94.3
-38.7
-55.6
Age 24
-0.3
-9.2
9.4
Age 65+
0.6
11.6
-12.3
Male
0.1
1.8
-1.9
Employed
-0.4
-15.5
15.9
Unemployed
0.5
9.9
-10.5
Indigenous
0.0
-0.1
0.1
Per capita versus OECD
Adults 1
23.7
-23.6
-0.1
Adults 3
-25.4
24.6
0.7
Adults 4+
-44.2
40.4
3.9
Kids 1
-53.6
41.5
12.1
Kids 2
-58.5
-5.1
63.6
Kids 3
-58.5
-22.3
80.8
Kids 4+
-58.5
-21.2
79.8
Age 24
-7.8
7.7
0.1
Age 65+
13.4
-13.3
-0.1
Male
0.4
-0.4
0.0
Employed
-11.0
10.9
0.2
Unemployed
0.7
-0.7
0.0
Indigenous
6.4
-6.3
-0.1
Note: a) Omitted Variables: Adults 2;Children 0; Age 25 to 64; Not in the Labour
force and Not applicable; non-Indigenous.
219
References
Aaberge, R. and I. Melby (1998), 'The sensitivity of income inequality to choice of
equivalence scale', Review of Income and Wealth, 44 (4) 565–9.
Altman, J.C. (2000), The Economic Status of Indigenous Australians, CAEPR
Discussion Paper No. 193, Centre for Aboriginal Economic Policy Research,
Australian National University, Canberra.
Altman, J.C. and B.H. Hunter (1997), Indigenous Poverty Since the Henderson
Report, CAEPR Discussion Paper 127, Centre for Aboriginal Economic Policy
Research, Australian National University, Canberra
Altman, J.C. and B.H. Hunter, (1998), ‘Indigenous poverty’, in R. Fincher and J.
Nieuwenhuysen, eds, Australian Poverty: Then and Now, Melbourne University
Press, Melbourne.
Atkinson, A. B., L. Rainwater. and T. Smeeding (1995), Income Distribution in
OECD Countries: Evidence from the Luxembourg Income study, Social Policy
Studies No. 18, OECD, Paris.
Australian Bureau of Statistics (1991), 1991 Census Dictionary, Catalogue No.
2901.0, ABS, Canberra.
Australian Bureau of Statistics (1995),. National Health Survey Users’ Guide,
Catalogue No. 4363.0, ABS, Canberra.
Australian Bureau of Statistics (1997), Survey of Income and Housing: User Guide,
Catalogue No. 6553.0, ABS, Canberra.
Australian Bureau of Statistics (1999), 1995 National Health Survey: Aboriginal and
Torres Strait Islander results, Cat. no. 4806.0, ABS, Canberra.
Buhmann, B., L. Rainwater, G. Schmaus,. and T.M. Smeeding (1988), 'Equivalence
scales, well-being, inequality, and poverty: sensitivity estimates across ten
countries using the Luxembourg Income Study (LIS) database', The Review of
Income and Wealth, 34 (2), 115–42.
Burniaux, J., T. Dang, D. Fore, M. Forster, M. D'Ercole and H. Oxley (1998), Income
Distribution and Poverty in Selected OECD Countries, Economics Department
Working Paper No. 189, OECD, Paris.
Commission of Inquiry into Poverty, R.F. Henderson, Chair, (1975), Poverty in
Australia, First Main Report, AGPS, Canberra.
Coulter, F. A. E., F.A. Cowell and S.P. Jenkins (1992), 'Equivalence scale relativities
and the extent of inequality and poverty', Economic Journal, 102. 1067-82.
Daly, A. and D. Smith (1995), The Economic Status of Indigenous Australian
Families, CAEPR Discussion Paper No. 93, Centre of Aboriginal Economic
De Vos, K. and M.A. Zaidi (1997), 'Equivalence scale sensitivity of poverty statistics
for the member states of the European Community', Review of Income and
Wealth, 43 (3), 319–33.
Finlayson, J. (1991), Don't Depend On Me: Autonomy and Independence in an
Aboriginal Community in Northern Queensland, unpublished PhD thesis,
Gray, A., ed, (1990), A Matter of Life and Death: Contemporary Aboriginal
Mortality, Aboriginal Studies Press, Canberra.
220
BOYD HUNTER, STEVEN KENNEDY, AND DANIEL SMITH
Gray, B. (1997), Indigenous data quality in the National Health Survey 1995:
analysis of selected questions from the unedited, unweighted file, ABS Internal
Discussion Paper 97/1, National Centre for Aboriginal and Torres Strait Islander
Statistics, Australian Bureau of Statistics, Darwin.
Greene, W. H. (2000), Econometric Analysis, Prentice Hall, New Jersey.
Hunter, B. H.(1998), 'Assessing the validity of intercensal comparisons of Indigenous
Australians, 1986–96', Journal of the Australian Population Association, 15 (1),
51–67.
Hunter, B. H (2001), ‘Tackling poverty among Indigenous Australians’, in R. Fincher
and P. Saunders, eds, Creating Unequal Futures, Allen and Unwin, Sydney.
Hunter, B.H. and M.C. Gray (1998), ‘Recent changes in the Australian workforce: a
focus on the structure of Indigenous employment’, Australian Bulletin of Labour,
24 (3), 220–40.
Hunter B.H. and J. Taylor (2001), ‘The reliability of Indigenous employment
estimates’, Agenda, 8 (2), 113–28.
Jenkins, S. P. and F.A. Cowell(1994). 'Parametric equivalence scales and scale
relativities', Economic Journal, 14, 891–900.
Lancaster, G. and R. Ray (1998), 'Comparison of alternative models of household
equivalence scales: the Australian evidence on unit record data', Economic
Record, 74, 1–14.
Martin, D. F. and J. Taylor (1996), 'Enumerating the Aboriginal population of remote
Australia: methodological and conceptual issues', Journal of the Australian
Population Association, 13 (1), 17–33.
Musharbash, Y. (2000), 'The Yuendumu community case study' in D. E. Smith, ed,
Indigenous Families and the Welfare System: Two Community Case Studies,
CAEPR Monograph No. 17, Centre of Aboriginal Economic Policy Research,
Rowse, T. (1988), 'From houses to households? The Aboriginal Development
Commission and economic adaption by Alice Springs town campers', Social
Analysis, 24, 50–66.
Saunders, P. (1994), Welfare and Inequality, Cambridge University Press, Cambridge.
Schiepers, J. (1992), 'On the choice of equivalence scale', Statistical Journal of the
United Nations, 9, 111–124.
Smith, D. E (1992), 'The cultural appropriateness of existing survey questions and
concepts'. in J. C. Altman, ed, A National Survey of Indigenous Australians:
Options and Implications, CAEPR Research Monograph No. 3, Centre of
Aboriginal Economic Policy Reasearch, Australian National University,
Canberra.
Smith, D.E. (2000), Indigenous Families and the Welfare System: Two Community
Case Studies, CAEPR Research Monograph No. 17, Centre of Aboriginal
Economic Policy Research, Australian National University, Canberra.
Smith D.E. and A.E. Daly (1996), The Economic Status of Indigenous Australian
Households: A Statistical and Ethnographic Analysis, CAEPR Discussion Paper
No. 109, Centre of Aboriginal Economic Policy Research, Australian National
University, Canberra.
Taylor, J. (1997), 'The contemporary demography of indigenous Australians', Journal
of the Australian Population Association, 14 (1), 77–114.
221
Whiteford, P. (1985), A Family's Needs: Equivalence Scales, Poverty and Social
Security, Research Paper No. 27, Development Division, Department of Social
Security, Canberra.
222
Women and Superannuation in the 21st
Century: Poverty or Plenty?
Simon Kelly, Richard Percival, Ann Harding
National Centre for Social and Economic Modelling
1
Introduction
The ageing of the population has been identified as one of the most critical policy
challenges facing governments in the industrialised world (OECD, 1996).
Governments everywhere have become deeply concerned about the ageing of the
‘baby boomers’ and the enormous fiscal strains that they may generate during the next
50 years. A key issue is the likely private incomes of future retirees and the extent to
which governments will be willing and/or able to support the living standards of
retirees, through the provision of the age pension, subsidised health care, and so on.
The need for governments to urgently make decisions in this regard is highlighted by
the fact that the first of the baby boomers are turning 55 this year. From this year on,
they may begin to enter their retirement and access their superannuation funds.
In such debates, the future fortunes of women loom large. Women are widely
acknowledged to be a particularly vulnerable group, because:
•
women are more likely than men to work part year, part time or in casual
positions (e.g. in October 2000, 43 per cent of female employees worked part
time, compared with only 13 per cent of male employees);
•
even when they work full time, women’s average earnings amount to only 84
per cent of male average full time earnings; and
•
women are more likely to have interrupted labour force careers, due to child
bearing and rearing. Women currently spend 17 years on average in the paid
labour force, compared to 39 years for men (Clare, 1994).
The result is that for many women, as contributions are usually linked to earnings,
they will accumulate substantially less superannuation than men. The male members
of superannuation funds, for example, already enjoy assets almost three times greater
than the female members (Goward, 1997: 3). There is also concern that occupational
superannuation and social security policy are not well integrated, with the
Superannuation Guarantee Charge (SGC) possibly being of marginal benefit to
women if it simply replaces the age pension (Knox, 1994). An additional major issue
is the treatment of superannuation upon divorce where, typically, divorce has a very
detrimental effect upon a women’s future retirement income. Many women who are at
home raising children or working in low-paid part-time jobs assume that they will
enjoy a comfortable retirement because of their husband’s superannuation. But now
Kelly, S., R. Percival and A Harding (2002), ‘Women and superannuation in the 21st century: poverty
or plenty’, in T. Eardley and B. Bradbury, eds, Competing Visions: Refereed Proceedings of the
National Social Policy Conference 2001, SPRC Report 1/02, Social Policy Research Centre, University
of New South Wales, Sydney, 223-247.
SIMON KELLY, RICHARD PERCIVAL AND ANN HARDING
that one-third of marriages end in divorce, for a very substantial proportion such an
assumption may not prove correct.
The financial situation of women in retirement is thus a major public policy issue.
There is concern that a significant proportion of female retirees this century will
receive age pensions, and possess little or no private assets or income. There is
concern about the likely cost of such support to the taxpayer, particularly as the ageing
baby boomers can be expected to also generate significant public health costs. If fiscal
pressures upon the government increase, the extent to which any shift from public to
private support is feasible will depend critically upon the future incomes and assets of
the aged, and the willingness of the aged to draw upon them to meet their own needs.
Given its recognised importance to public policy, it is not surprising that the past
decade has seen continuous policy reform in this area, most notably the introduction
of the compulsory SGC, which has spread superannuation across the workforce. The
past decade has also seen the development of a range of highly sophisticated models
for predicting future retirement incomes, particularly by the Retirement Income
Modelling (RIM) Task Force (now RIM Unit) established within the Commonwealth
Treasury.
Those models established to date do not, however, provide the best tools to answer
questions about the future retirement incomes of women. In particular, they are unable
to cope with the impact of divorce. In addition, while they provide a very general
overview of the future distributional picture - by age, gender and decile of career
income - they are not well able to capture diversity or the experiences of smaller
population sub-groups. Thus, they are not able to answer questions about the likely
retirement incomes of women who take up part-time jobs for long periods of time
because they are sole parents.
This paper discusses the initial results coming from a dynamic microsimulation model
developed at NATSEM. The model, which has been in development for a number of
years, has recently added a superannuation module to enable questions like those
above to be answered and to shed light on the future financial situation of women (and
men) in retirement. This dynamic model, DYNAMOD, provides detailed
distributional estimates of the impact of policy and other changes and of the likely
degree of dispersion in future retirement wealth. The development of the
superannuation module is being done at NATSEM with support from an Australian
Research Council grant (A79906127).
2
The Model
Possible Modelling Techniques
There are a number of modelling techniques that could be used to project the financial
situation of retired women. These include hypothetical models, group models and
dynamic microsimulation models.
Hypothetical Models: Hypothetical models analyse the situation of an individual,
whose particular characteristics are defined by the model user. Thus, a hypothetical
model might look at the future superannuation of a single woman earning average
224
WOMEN AND SUPERANNUATION: POVERTY OR PLENTY?
earnings and working full time for 20 years and part time for 20 years. Such models
are very useful for analysing policy outcomes for particular types of individuals and
can incorporate enormous detail. However, such models cannot be used to obtain
estimates of aggregate budget outcomes or distributional outcomes for the whole
population.
Group Models: Group models divide the population into a set of groups defined by
characteristics such as year of birth, sex, marital status and labour force status
(Rothman, 1996, 1997). The extent of diversity is thus effectively constrained by the
number and type of groups specified. Moreover, there is usually no scope for people
in the model to move from one group to another. As a result, such models cannot cope
with divorce. They also cannot cope with fluctuating individual earnings over a
lifetime, with people being assigned to a lifetime earnings group and then
experiencing the average outcomes for their group. Thus, group models provide
details of the average experience for each of the groups specified within the model.
Dynamic Microsimulation Models: Dynamic microsimulation models are a
relatively new type of model that can provide both detailed distributional answers and
profiles for small population sub-groups (such as sole parents), as well as aggregate
budgetary estimates. The key to this type of modelling is that individuals form the
basis of the model and that such individuals make up a representative sample of the
population. This means that the individuals within the model retain the diversity of
experiences apparent in the real world – but that conclusions can still be drawn for all
of Australia. Such models thus embody the key advantages of the hypothetical and
group models, but do not suffer from their major disadvantages.
DYNAMOD
NATSEM has developed a dynamic microsimulation model to show the implications
of policy outcomes on individuals while being able to provide aggregate and average
results. The dynamic model, DYNAMOD, begins with the 1986 Australian Bureau of
Statistics (ABS) one per cent Census sample file. Onto the individuals on this file are
imputed additional characteristics (such as more detail on the level and type of
education being undertaken, the state in which the person lives, the value of the family
home, etc). The result is a base population of 150 000 synthetic individuals with over
80 characteristics each. The model takes this base population and simulates their lives
from 1986 onwards, by stepping month-by-month through time until the middle of the
21st century. As the individuals step through life, they experience a range of life
events, in line with Australian data about the probabilities of those events happening
to real Australians. The life events include death; fertility; couple formation and
dissolution; emigration and immigration; primary, secondary and tertiary education;
labour force changes (full and part-time employment, unemployment and not in the
labour force, NILF), disability onset and recovery, and the earning of income.
These individuals are the ‘micro’ part of the simulation. Analysis of any aspect of
their lives can be done at this micro level. Additionally, individuals in the model are
linked to simulate couple formation and the links broken to simulate couple
dissolution. Links are maintained between parents and their children and between
current and ex-spouses. In this manner, the model maintains the diversity in the
225
population, copes very well with couple formation and dissolution, and, importantly,
allows generation of a longitudinal history (or future) for individual cases in the
model. This longitudinal element is essential in order to capture the accumulation
processes that underlie growth in superannuation and financial assets. Alternative
futures can be simulated by varying particular probabilities, such as the chance of
becoming full-time employed or of having a first child, given certain characteristics.
The model also contains a capacity to make the summed outcomes for all individuals
within the model fit some exogenously imposed total. For example, the proportion of
all women of a certain age range that are to be employed full time in 2010 can be set
by the user.
A wide range of technical papers provides detail of the processes used in DYNAMOD
(detailed overviews are provided in Antcliff, 1993 and King, Baekgaard and
Robinson, 1999). The latest version of the dynamic microsimulation software is
DYNAMOD-3. This version adds life events relating to paying income tax,
accumulating and diluting personal and family wealth assets and contributing to
superannuation. It is this version and its outcomes that are discussed throughout the
remainder of the paper.
The Superannuation Module
Superannuation is a recent addition to DYNAMOD and this module promises to
provide very detailed estimates of those who have superannuation and the amounts
that are held. The accuracy of the modelling, naturally, depends on the availability of
good quality data. For dynamic microsimulation there is an additional need and that is
for very detailed and preferably longitudinal or panel data. Unfortunately the data
available on superannuation holdings in Australia are limited. A significant step
forward will be the release later this year of the confidentialised unit record or micro
data from the 2000 ABS superannuation survey. At the present time, the major source
of superannuation asset data is a survey of major private superannuation funds
conducted by RIM in the early 1990s (Brown, 1994). This data collection, combined
with ABS survey data, provides the foundation of superannuation asset modelling in
DYNAMOD.
Superannuation is simulated from 1993 onwards. There is both RIM and ABS data
available for that year. RIM, through its data collection and research, has estimated the
average amounts held in superannuation in 1993 while ABS undertook a survey of
superannuation coverage in that same year. The ABS data shows who had a
superannuation account in 1993, while the RIM data indicates the balance of the
account at that time. The simulation uses this coverage and account balance
information to impute an initial superannuation amount onto individuals.
From 1993 onwards, a variety of transactions take place:
•
•
•
contributions equivalent to employer contributions are deposited into a
person’s account as income is earned;
a person may choose to make voluntary contributions into their superannuation
account;
interest is paid on the account balance; and
226
•
accounts can be closed and the balance transferred into another type of account
on retirement.
For reasons of practicality, the full complexity of superannuation could not be
modelled and some areas were simplified. In some other areas assumptions had to be
made as actual data were not available. The assumptions and simplifications include:
•
•
•
•
•
•
the reduction of fund types to only one: the Accumulation Fund. This is the
most common type of fund (87 per cent of superannuation membership is of
this type);
employer contributions are made for all employees at the superannuation
guarantee charge (SGC) rate. In line with the legislation, the following are the
only exceptions: individuals who are paid less than $450 per month; those
employed but over 70 years old; and those employed but under 18 years old
and not working full time;
employee and voluntary contribution percentages are based on the 1995 ABS
survey data for 45-74 year-old employees. For others in the labour force an
employee contribution of either zero or five per cent is assigned;
each person has only one account;
withdrawal of funds before retirement is not permitted; and
all funds pay the same rate of return.
Each of these assumptions and their impact on the model is described in more detail in
Appendix One.
Contributions to superannuation are made throughout the simulated working life of a
person and interest is earned annually on funds invested in superannuation until
retirement is reached, when the funds are transferred to the person’s cash account. The
decision of a person to retire is done randomly but is based on probabilities derived
from ABS data.
In the event of a divorce, the model currently includes superannuation in the division
of family assets. After the division, the respective amounts are deposited back into
each individual’s superannuation fund. In effect, DYNAMOD is enacting (a
simplified version of) the legislation that is currently before the parliament.
3
Results
In General
Before using DYNAMOD to project outcomes in regard to superannuation, it may be
appropriate to show that aggregate projections produced by the model agree with
benchmark data. Unless a reasonable fit is achieved at this level, the credibility and
usefulness of the model in regard to superannuation will be undermined.
Figure 1 is an example of the extent of concordance that is currently achieved between
DYNAMOD projections for Australia’s population and ABS actual data and
227
projections. The ABS projections are Series II1. As can be seen at this aggregate level,
the net effect of the simulation of mortality, fertility and migration is to reproduce a
population total that closely matches the ABS data.
Figure 1: Australian Population: 1986-2050
27
Population (millions)
25
23
21
19
Dynamod Estimate
17
ABS actual &
projected (Series II)
15
1986
1996
2006
2016
2026
2036
2046
Source: ABS, Cat No. 3222.0, Population Projections
The alignment at this aggregate level is remarkable when you consider the different
methodologies. ABS use a form of group modelling2 while dynamic microsimulation
looks at the circumstances of every individual, every month, and decides whether they
will become pregnant or not; whether they will continue to live or not; and whether
they will emigrate or not. These are very different approaches and the closeness of the
simulated aggregates with the ABS data suggests that the underlying modelling is
capturing individual behaviour. A major advantage of dynamic microsimulation
modelling is that we can now ‘drill down’ into this aggregate data or we can group
individuals together to look at collective or group outcomes. We could look at the
retirement assets of an individual who marries at 25, divorces at 35, is diploma
qualified, has two children, works in part-time employment for four years, and is
employed full-time for ten years; we could compare women and men at various ages;
or we could project the number of marriages per 1000 population (Figure 2). It is this
flexibility that gives dynamic microsimulation a significant advantage over other
techniques.
1
ABS Projection Series II most closely reflects prevailing trends. The Series II
assumptions are (i) a declining mortality rate; (ii) fertility rates falling to 1.75 by
2005/06 and then constant; and (iii) a net annual immigration gain of 70 000.
2
‘The ABS uses the cohort-component method which begins with a base population for
each sex by single years of age and advances it year by year by applying assumptions
regarding future mortality and migration’ (ABS 3222.0: 125).
228
Superannuation
Aggregate Projection: The flexibility of dynamic modelling is seen again in an
analysis of superannuation. Comparisons can be made at the aggregate level with
exogenous benchmark data; comparisons by age, gender or marital status can be
made; or the circumstances of a specific group of individuals can be examined.
Figure 2: Australian Marriage Rate: 1986-2050
Marriages per 1,000 population
8
7
6
5
Dynamod
ABS Actual
4
1986
1996
2006
2016
2026
2036
2046
Source: ABS, Cat. No. 3310.0, Marriages and Divorces 1999
These comparisons could have an outcome in terms of superannuation coverage,
annual changes in account balances, or balance at retirement. In Figure 3,
DYNAMOD aggregate estimates for superannuation assets are compared with
Australian Prudential Regulation Authority (APRA) and RIM estimates. The
DYNAMOD estimates are in concordance with those estimated by RIM. Around
2030, the last of the baby boomers will celebrate their 65th birthday. A good
percentage of those still working will celebrate their birthday by ceasing employment
and hence stop contributing to superannuation. The impact of this can be seen in the
levelling of the growth in superannuation assets from this time onwards.
Coverage Projection: The level of superannuation coverage for both men and women
has increased dramatically since the introduction of the three per cent industrial award
superannuation and then Superannuation Guarantee Charge (SGC) in 1992. The latter
of these, with its compulsory employer contributions for every employee earning more
than $450 per month, is having a dramatic impact on the coverage of superannuation.
The ABS estimated in 1993 that 50.3 per cent of all Australians aged 15 to 74 were
covered by superannuation and 80.5 per cent of those employed were covered. In
2000, ABS estimated 87.7 per cent of people aged 15 to 54 with one or more jobs had
superannuation coverage. While not directly comparable it does indicate the strong
229
Figure 3: Aggregate Superannuation Assets: 1986-2050
3000000
Total value ($, 1999)
2500000
2000000
1500000
1000000
Dynamod Estimate
RIM Estimates
APRA
500000
0
1986
1996
2006
2016
2026
2036
2046
Source: RIM estimates: Rothman, 1998; APRA estimates: Superannuation Trends June Qtr 2000
upward trend. Superannuation coverage simulated using DYNAMOD shows growth
from an estimated 45.9 per cent in 1993 to 84.4 per cent in 2000 for all people aged 15
to 74 - again a very strong upward trend. Figure 4 shows the projected growth in
coverage over the period 1993 to 2030. These figures cover all people aged 15 to 74
except full-time school students, rather than just those employed. A significant feature
of these projections, done using DYNAMOD, is the estimate that by 2000 almost all
people in some age groups were covered by superannuation. While the ABS estimates
for 2000 are a little lower than this (the ABS estimates were in the low 90s), it is clear
that SGC is well on its way to ensuring that everyone is covered by superannuation.
The definition of ‘covered’ used in DYNAMOD means that anyone who has any
amount in a superannuation fund is covered. It does not exclude those not currently
making a contribution. (For example a person may contribute to superannuation while
employed, then leave the labour force for a period to care for children.) Under the
DYNAMOD definition the person is covered by superannuation. To qualify a person
only needs to make one contribution (or have their employer make one contribution)
and they are then regarded as ‘covered’ until retirement. In the real world, it is quite
possible that a person has some superannuation from past employment but is not
aware of it. This may have lead to some underreporting in the ABS data.
The compulsory nature of SGC makes the probability of achieving old age without
ever earning more than $450 in one month remote and this is reflected in the figure
below. It may not occur in 2000, but we are not far away from the day when every
person will retire with some superannuation assets.
Average Superannuation Amounts
In a previous section the aggregate projected superannuation amounts were presented.
In this section a more detailed look is taken at individual holdings. DYNAMOD
begins simulating superannuation in 1993. In July of that year it allocated every
230
Figure 4: Superannuation Coverage: Selected Years 1993 -2030
People Covered by Superannuation
Proportion with Superannuation (%)
100
80
60
40
20
1993
ABS 1993
2000
2030
0
15-19
20-24
25-34
35-44
45-54
55-64
65-74
Source: ABS, 1993, Cat. No. 6319.0, Superannuation Australia, 1995; Other values: DYNAMOD
calculations
person with superannuation cover an initial balance based on age and gender. In Table
1 this alignment by age and gender with the RIM data can be seen. The higher ‘All
Ages’ value in DYNAMOD reflects the slight differences in spread of superannuation
over the ages.
All values in the table are in real 1999 dollars. The final four columns of the table
show the estimated superannuation assets in 2030 and the percentage change
1993-2030. While the average superannuation amounts for women are well below
those for men, there is a marked improvement over the period 1993 – 2030. The
average female superannuation assets increase from 45 per cent of the average male in
1993 to 70 per cent in 2030.
Women in Retirement
In the introduction the ageing of the population and the future fortunes of women were
discussed. An analysis of ABS population projections highlights both the ageing of
the population and its impact on women.
Ageing: In Figure 5 and Figure 6 significant change to the Australia age profiles can
be seen. In the two comparisons, the population in 1991 (dark colour) is compared
with 30 and 60 years later, while the percentage of the population is presented in five
year age groups from 0 to 85 and 85+. In 1991 the largest groups were the three
groups covering ages 20 to 35 (i.e. born from 1956-1971), with a fairly clear decrease
in the percentages above that age. The age group 85+ represented only 0.9 per cent of
the population (0.25 per cent males and 0.63 per cent females). By 2021 the peak age
groups will have flattened and become broader. The 1991 Christmas tree shape has
transformed into a more rectangular one. There is now a fairly even distribution of
231
Table 1: Average Superannuation Assets: July 1993 and 2030 (in 1999 dollars)
Age
Group
15-19
20-24
25-29
30-34
35-39
40-44
45-49
50-54
55-59
60-64
65 +
RIM
DYNAMOD
DYNAMOD
1993
1993
2030
Males Females Males Females
(dollars)
2237
3911
7798
12 930
19 588
26 606
35 646
44 712
55 688
71 731
75 424
All Ages 21 957
1488
2767
4739
6093
7792
9761
12 709
16 905
23 993
35 194
25 976
DYNAMOD
Change
1993-2030
Males Females Males Females
(Percentages)
2237
3911
7798
12 930
19 588
26 606
35 646
44 712
55 688
71 731
75 424
1488
2046
2767
8030
4739 25 220
6093 52 149
7792 91 403
9761 139 752
12 709 201 778
16 905 281 115
23 993 352 794
35 194 389 926
25 976 397 204
1422
5319
15 987
30 064
48 877
74 269
109 832
155 452
196 930
193 374
194 837
-9
105
223
303
367
425
466
529
534
444
427
-4
92
237
393
527
661
764
820
721
449
650
7993 21 419
9647 128 235
89 591
499
829
Source: Brown, 1994 and DYNAMOD, 2001.
people between the ages of 20 to 65. The average age has increased and the proportion
aged 85+ has more than doubled to 2.0 per cent.
By 2051, as seen in Figure 6, the age groups over 40 contain the higher proportions of
people and the rectangular shape is beginning to be replaced by an inverted pyramid.
The 85+ population has now become quite significant, representing 4.5 per cent of the
population, with almost two out of three of those aged 85+ being women (64.2 per
cent). In the space of only 60 years, from 1991 to 2051, a population with the majority
aged below 40 years will be supplanted by a population dominated by people aged
over 40, while the proportion aged over 85 will have increased fivefold.
The fivefold increase in people aged 85+, along with the significant increase in the
proportion aged 65 and over (increasing from 11 per cent in 1991 to 24 per cent in
2051), will result in greatly increased medical costs and nursing requirements. These
costs and requirements are going to continue to grow and place an increasing burden
on people’s own financial resources, their relatives and/or on the government. The
fivefold increase in women aged 85+ makes them vulnerable to bearing a high
proportion of this cost.
232
Figure 5: Australian Population in June 1991 and June 2021
Males 1991
Males 2021
Females 1991
Females 2021
80-84
70-74
60-64
50-54
40-44
30-34
20-24
10-14
0-4
5 Males4
3
2
1
0
1
2
3
% of population
4
Females
5
Source: ABS, Cat. No. 3222.0, Population Projections
Figure 6: Australian Population in June 1991 and June 2051
Males 1991
Females 1991
Males 2051
Females 2051
80-84
70-74
60-64
50-54
40-44
30-34
20-24
10-14
0-4
6 Males
4
2
0
2
4 Females
6
% of population
Source: ABS, Cat. No. 3222.0, Population Projections
In summary, the proportion of people aged 85+ will increase fivefold by 2051 and
women will represent two-thirds of the people in this age group. The proportion aged
65+ will increase to 24 per cent of the population.
233
Trends in Marital Status: The ageing of the population will impact on women but a
number of other changes will also have an impact. One of these is the trend towards
later marriage. Figure 7 shows the forecast decrease in the proportion of females
married in the two age groups up to age 35 and the higher proportion of females
married in their 40s and beyond. The ABS estimate that the median age for women at
first marriage has increased from 24.0 years in 1988 to 26.2 in 1998. The median age
for women at remarriage has also increased, from 35.3 to 38.4 over that period (ABS,
Cat. No. 4102.0). These trends are reflected in the declining proportion of younger
women who are married.
There are two possible reasons for the increased proportion of women married at ages
40 and above. The first is that since marriage started at an older age, it will finish at an
older age. The assumption here is that marriages are of the same duration. In fact, the
duration of marriages increased slightly over the period 1988-98, by 0.5 years to 7.8
years (ABS, Cat. No. 4102.0). The end result of getting married two to three years
later and the marriage lasting six months longer would have the impact of moving the
age distribution curve to the right as we see in Figure 7. A second or alternative reason
for the movement in the curve is the ageing discussed in the previous paragraph.
While on average women still outlive men, the ageing of both genders causes wives to
become widows at a later age. This trend can be seen in Figure 8. The proportion of
females who are widows decreases between 1990 and 2030/2050.
Figure 7: Estimated Proportion of Females Married by Age Group, Selected
Years 1993-2030
Proportion of Females (%)
100
80
60
40
20
0
15-24
25-34
35-44
1993
45-54
2010
55-64
65+
2030
Source: DYNAMOD
An alternative view of the changing marital status of women around retirement age is
presented in Figure 9. The figure shows the outcomes produced by DYNAMOD for
women aged 55 to 64 over the period 1986 to 2030. Each colour represents the
proportion of women in this age group with that particular status. The proportions are
cumulative, with the proportion not shown being married women (de jure or de facto).
As an example, in 1986, 28 per cent of women aged 55 to 64 have a marital status
other rather ‘Married’, while the remaining proportion (72 per cent) are married.
234
Figure 8: Estimated Proportion of Females Widowed by Age Group, Selected
Years 1993-2030
Proportion of Females (%)
60
40
20
0
15-24
25-34
35-44
1993
45-54
2010
55-64
65+
2030
Source: DYNAMOD
The figure therefore shows an increased proportion married over the period 19862030. A decrease in the ‘widowed’ proportion is also evident.
Figure 9: Estimated Marital (a) Status of Women aged 55-64: 1986-2030
Proportion of females Aged 55-64 (%)
40
30
20
Never Married
Separated
10
Divorced
Widowed
0
1986
1996
2006
2016
Note:
2026
2036
2046
a) ‘Married’ women are the residual group on this chart. That is, if the proportion in the four
other marital states identified totals 25 per cent, then 75 per cent of women at the age are
estimated to be married (de jure or de facto).
Source: DYNAMOD
235
Women in the Labour Force
The labour force participation of men and women is markedly different. Women have
always been more likely to be employed part time. In 1999, the proportion of women
employed part-time was 43.5 per cent, while for men it was only 12.5 per cent (ABS,
Cat. No. 4102.0: 108). This wide disparity was evident throughout the 1990s, despite
a significant growth overall in part-time employment for both genders (the overall
proportion employed part time rose from 20.9 per cent in 1990 to 26.0 per cent in
1999). Labour force participation rates are also quite different when compared on a
gender basis. The female participation rate rose by to 2.0 percentage points to 53.9 per
cent in the 1990s while the male participation rate fell 2.7 to 72.8 per cent over the
same period. In some age groups the difference was even greater.
For example, the participation rate trend for 55-59-year-old men is slowly downwards
(from 75.0 per cent to 73.2 per cent in the 1990s), while the trend for women of the
same age is rapidly upwards (up from 32.4 per cent at the start of the 1990s to 43.6 per
cent at the end) (ABS, Cat No. 4102.0: 109).
Overall, female labour force participation and the type of employment are changing.
The proportion not in the labour force is falling and the proportion in part-time
employment is rising. Both DYNAMOD projections and ABS actual data suggest that
some of the part-time rise may be at the expense of full-time but most is coming as a
result of choosing to participate in the labour force on a part-time basis rather than not
participate at all.
An advantage of dynamic microsimulation is that these differences can be easily
handled. The decision to participate in the labour force or to change employment
status is done at the individual level - and hence the influences on the decision are
applied at this level and can vary over time.
Figure 10: Estimated Labour Force Status of Females Aged 45-54: 1986 -2050
45-54yo Females
60
50
40
30
20
10
0
1986
1996
2006
Employed-FT
2016
Employed-PT
Source: DYNAMOD
236
2026
2036
Unemployed
2046
NILF
A different way of looking at the labour forces changes is to consider a particular
group over a period of time. This is done for females aged 45-54 in Figure 10. The
decreasing proportion not in the labour force is evident, along with the increasing
proportion in part-time employment. The growth in part-time comes mainly from
females entering or re-entering the workforce (the decreasing proportion NILF) but
there is a small impact on the full-time participation as well. A slight decrease in
females employed full time can be seen from 2000 onwards which is perhaps a
reflection of the desire to be in the labour force yet still have time for other things.
Women and Superannuation
To accumulate superannuation assets a person must have superannuation coverage. As
discussed in the section on Coverage Projection, almost everyone will have had
superannuation contributions made on their behalf by retirement because of the
government’s SGC policy. The amount that will be held in this superannuation
account will depend on three factors: the rate of return of the superannuation fund, the
contributions made into the account and taxation on the superannuation. The last of
these, taxation, is not part of NATSEM’s modelling at this stage. The SGC employer
contributions are a percentage of earnings, the higher the earnings and the longer the
time in employment, the greater the superannuation assets. In most cases a working
life spent in full-time, high salary employment and not interrupted by any periods of
unemployment or NILF would produce the highest superannuation. This set of
circumstances is most likely achieved by males, and it is for this reason that average
female superannuation amounts were only 45 per cent of those of males in 1993.
However, the increasing age of marriage and childbirth, reduced periods of NILF for
child raising purposes, and the increased use of part-time employment are having a
positive effect. Female superannuation is increasing as a proportion of male
superannuation, and is estimated by DYNAMOD to reach 70 per cent by 2030. Unless
there is complete equality in the labour force roles, the female proportion will remain
lower than the male, due partly to lower female earnings and different workforce
participation patterns in the child bearing and rearing years.
The estimated amounts of superannuation held by various age groups of women are
shown in Figure 1. The impact of compound growth is evident in the figure. In 1993
few women had contributed to superannuation in the early part of their working life
and hence the balance in their superannuation account did not grow through interest
being paid and added each year. In comparison, by 2030 most women will have made
superannuation contributions at the start of their working life. For the next 30 to 40
years that contribution will have been added to each year through interest.
This, combined with more years of employment than the previous generation’s, results
in exponential growth in the value of the asset.
As Table 1 suggests, women’s accumulated superannuation upon retirement is
expected to increase very rapidly over the next three decades. For example, the
average account balances of women aged 55 to 59 are forecast to increase to $196 930
by 2030, up from only $23 993 in 1993. This is an increase of more than 700 per cent.
Similarly, women aged 60 to 64 are forecast to have average account balances of just
over $190 000 in 1993, up from about $30 000 in 1993, a 449 per cent increase. Table
2: shows the superannuation balances for females aged 55 to 64 at three points across
237
the superannuation spectrum: the 10th percentile, the median (or 50th percentile) and
the 90th percentile. Eighty per cent of females with superannuation in this age range
have a balance between the 10th and 90th percentiles.
Figure 11: Estimated Accumulated Superannuation by Age Group, Females,
Selected Years 1993-2030
Value of Superannuation ($'000s)
300000
200000
100000
0
15-19
20-24
25-29 30-34
1993
35-39 40-44 45-49
2000
2010
50-54 55-59
2030
60-64
Source: DYNAMOD
Table 2: Superannuation Balances for Females with Superannuation, Aged 5564, Selected Years 2000-2030
Percentile
90%
Median
10%
2000
($)
47 042
29 459
3850
2010
($)
103 110
61 674
27 226
2020
($)
179 705
109 600
55 032
2030
($)
290 632
183 638
93 414
Source: DYNAMOD
A very encouraging feature of this table is the huge increase for the 10th percentile. In
real terms it increases by almost $90 000 over the 30 years. Restating this, by 2030
DYNAMOD estimates that 90 per cent of women aged 55 to 64 will have
approximately $93 000 or more in superannuation.
Adequacy: Most people estimate that they will need around 60 per cent of their
annual pre-retirement income to be comfortable in retirement (ASFA, 2000). At the
same time the government’s Age Pension provides a maximum income of 25 per cent
of the average total male earnings. If we make the not unreasonable assumption that
the average total male earnings is the average pre-retirement income, those with only
the age pension will have an income of 25 per cent when they would like 60 per cent
to be comfortable. We assume people would like their superannuation to make up the
other 35 per cent required reach the comfortable level. It does not appear this will be
the case for the majority of women. Consider women aged 55 to 64 in 2030 on the
10th percentile, if they invested their $93 000 superannuation at five per cent, it will
238
produce an annual income of $4650 - approximately 11 per cent of average earnings
(in May 1999). The same calculation for those in this age group with the median $183
600 would add 23 per cent. However, for those at the median, an asset of this size is
certainly sufficient to begin affecting the age pension entitlements, via either the
income or assets tests.
From these estimates it appears that at least half of all women will be relying, at least
in part, on the Age Pension to support them for some time to come. By 2030
superannuation will be providing a substantial supplement to the Age Pension for
these women but even with the income from superannuation they will be struggling to
reach the accepted income level for a comfortable retirement.
4
Conclusion
In the future, will women be able to live comfortably in their retirement? Will they
have plenty because they have contributed to and accumulated large amounts in
superannuation up until their retirement? Or will they be no better off than those
retiring at present?
NATSEM has developed a dynamic microsimulation model to project the lives and
circumstances of individuals out until the middle of this century. This paper presents
the first preliminary findings from this model using its new superannuation module.
The addition of this module allows questions such as those above to begin to be
answered. The superannuation modelling is unique in Australia in that it models
behaviour at the individual level rather than on a group or overall level. This type of
modelling allows greater detail than using other techniques.
The proportion of the population aged 65+ is projected to increase from 11 per cent in
1991 to 24 per cent in 2051 and the population aged 85+ will increase proportion
fivefold. Two-thirds of the 85+ population will be women. Statistics also show us that
the proportion of women participating in the labour force is increasing; the proportion
working part time is increasing; and the median age for marriage has increased two
years in the past ten.
These demographic and labour force changes impact on the earnings, and consequent
superannuation accumulation, of women. Fortunately, changing employment status
(full-time or part-time) - and the disruptions to labour force careers due to child
bearing and rearing which are not often found in men’s careers - are easily
accommodated within the dynamic model. Other factors that influence earnings and
superannuation accumulation such as the increasing age of first marriage and
childbirth, early retirement, and divorce are also accommodated.
The simulated results suggest that soon, if not already, the proportion of people who
retire without at least some superannuation will be reduced to zero. This is a result of
the government’s introduction of the SGC in 1992. Under the SGC, every employer
must make superannuation contributions for employees earning more than $450 per
month. The model suggests that very few people will not earn more than $450 in a
month at least once during their working life.
239
While everyone will have some superannuation, the amounts are not likely to take the
average women from near ‘poverty’ to ‘plenty’ in the near future. In 1999-dollar terms
the average balance in superannuation is estimated to increase from $9647 in 1993 to
$89 591 in 2030. While the eightfold increase seems good, it really only highlights the
poor current situation. For women getting close to withdrawing their superannuation
(aged 55-59) the balances go from an estimated average of $24 000 in 1993 to $197
000 in 2030. A very encouraging aspect of the growth in superannuation assets is the
growth in the bottom end. By 2030 NATSEM estimate 90 per cent of women aged 55
to 64 with superannuation will have accumulated more than $93 000 in their accounts.
Another area of improvement is in the gender balance. NATSEM estimates that the
average female superannuation balance will lift from 45 per cent of the average male
balance in 1993 to 70 per cent by 2030
The estimates suggest that women’s accumulated superannuation upon retirement will
increase very rapidly over the next three decades and the growth is across the
superannuation spectrum. Consider 55-64-year-old women in the middle of the
superannuation range (on the median or 50th percentile), they will see their average
superannuation balance rise from $29 459 to $183 638 over 30 years from 2000.
While this is clearly a very dramatic rise, the rise for women at the bottom is perhaps
more gratifying. For women of the same age but on the 10th percentile the increase is
from $3850 in 2000 to $93 414 in 2030. We estimate that 90 per cent of women in
this age group will have more than $93 000 in superannuation in 2030.
While these increases are dramatic and welcome, they will not move women to
‘plenty’. The Age Pension provides a retirement income of 25 per cent of average
male earnings, the amounts held in superannuation discussed above will supplement
this pension with an additional 11 to 23 per cent. Unfortunately as this
supplementation rises, income and asset means tests will start to reduce the Age
Pension. Most people feel 60 per cent of pre-retirement income is required to live
comfortably in retirement. A total income of at most 36 to 48 per cent of average
weekly earnings for more than half of all women, achieved through superannuation
and the Age Pension is still well below the comfortable or ‘plenty’ level.
The results found using this dynamic microsimulation model are preliminary. The
dynamic nature of the model means that significant resources need to be put into
validating every part of the model before the outcomes are truly reliable. While this
validation is well underway, the process is not complete. However, the results to date
are very promising and the level of detail unparalleled.
Appendix One: Assumptions
In the ideal world, superannuation balances would be known for each of the 150 000
synthetic people at the beginning of the simulation. Unfortunately, there are limited
data available on superannuation holdings in Australia and superannuation holdings
for every person in 1986 are certainly not available. Superannuation funds, in general,
are private organisations and hence details on individual holdings are not generally
available. The Australian Bureau of Statistics has periodically collected data on
superannuation, but the majority of their data relates to coverage and annual
240
contributions rather than actual holdings. The reason for ABS not collecting data on
actual holdings is clearly based on the unreliability of such data. Until the late 1990s,
most people did not know the value of their accumulated funds and they often had
more than one account.
A major exception to this general lack of detailed data on superannuation holdings is
the data collected from major private superannuation funds by the Retirement Income
Modelling (RIM) Unit of Treasury in 1993. RIM collected information on the
distribution of superannuation assets and contributions from three major life
companies. The study covered about 15 per cent of total superannuation assets and
around 1.8 million people (Brown, 1994). This data collection and the resulting
research and analysis by RIM provide a valuable source of data for DYNAMOD-3.
Defined Benefits and Accumulated Funds
Historically there have been three types of superannuation funds. The first is the
‘defined-benefit’ scheme where the pension entitlement is a complex calculation
involving the number of years in the scheme and the salary at the time of retirement.
This type of fund is predominately found in the public sector and in large white collar
industries. In general these funds are heavily subsidised by the employer and very
generous in their benefits. However, most defined-benefit schemes are now closed to
new entrants.
The second type of scheme is the ‘accumulation’ scheme (or defined-contribution
scheme). With the accumulation scheme, contributions (by the individual and/or their
employer) and interest build up in a personal account over time. The balance of this
account at the time of retirement represents the entitlement. This is the most common
type of scheme in the private sector. The introduction of the three per cent industrial
award superannuation, and then the Superannuation Guarantee Charge (SGC) in 1992,
have resulted in accumulation funds becoming the dominant form of scheme. In June
2000, 87 per cent of superannuation members were in an accumulation scheme (see
Table A1).
Table A1: Types of Superannuation Schemes
Fund Type
Accumulation
Defined Benefit
Hybrid
TOTAL
Members
(000’s)
Members
(%)
Assets
($ billion)
Assets (%)
18 885
87%
257
60%
499
2%
24
6%
2344
11%
146
34%
21 726
100%
427
100%
Source: APRA Superannuation Trends – June Qtr 2000
A third type of fund is a ‘hybrid’ scheme, representing a combination of the
defined-benefit and accumulated funds.
241
Given the predominance of the accumulation scheme, in DYNAMOD all
superannuation entitlements are treated as if they were of the accumulated type. RIM
noted that members of defined-benefit schemes in 1993 had an older age distribution
and had a heavier concentration of higher salary members than members of
accumulated schemes. This resulted in the average entitlement in defined-benefits
accounts being 3.8 times higher than in accumulation schemes.
The omission of defined-benefit funds may therefore under-report aggregate
superannuation in the early years of the simulation.
Employer Contributions
Under industrial award superannuation and then under the SGC, employers have been
required to make minimum contributions for most employees to a complying
superannuation fund. The minimum contributions under SGC are shown in Table A2.
Superannuation contributions do not have to be made for some categories of
employees. These categories include employees who are:
•
•
•
•
•
paid less than $450 in a month;
aged 70 or over;
under 18 years of age and working for not more than 30 hours a
week;
performing work of a domestic or private nature for not more than
30 hours a week for a non-business employer; and
who have vested benefits in excess of their pension RBL and who
elect not to receive any more employer contributions.
Table A2: Superannuation Employer Obligations
From
Percentage of Salary
July 1996
6
July 1998
7
July 2000
8
July 2002
9
Based on the exclusions above, the simulation assumes that everyone receives an
employer contribution equal to the rate defined in Table A2. The only exceptions are
those with gross annual earnings less than $5400 (450 x 12); those under 18 years and
not working full time; and those over the superannuation age limit (currently set at
70). Again there may be some under-reporting in the early years as some organisations
pay more than the SGC amount. This is particularly true in the public sector.
242
Employee and Voluntary Contributions
In addition to the employer contributions to an individual’s superannuation, a person
may either be required to contribute part of their salary into a superannuation fund or
may voluntarily contribute to their own superannuation fund. In DYNAMOD,
employee contributions are added to the same account as the employer funds. Whether
a personal contribution is made and, if so, the amount contributed, is based on two
sources: a RIM conference paper (Tinnion, 1998) for non-employees and those aged
under 45; and the 1995 ABS publication Superannuation Australia for employees
aged 45+.
For people who are currently unemployed, not in the labour force or doing unpaid
work, it is assumed they make no contribution to their superannuation. If a person is
employed then the chance of them contributing is based on a probability distribution.
The distribution is shown in Table A3 below. The amount contributed is five per cent
of their salary. This rate is one of the most common default rates used by
organisations that require employees to contribute to a superannuation scheme.
For employees aged 45 to 76, a more complex algorithm is used based on ABS data.
ABS Cat. No. 6319.0, Superannuation Australia, November 1995, Table A3 shows
employees aged 45 to 76 making various personal contributions (see Table A4). On
the basis of this table, an annual employee contribution rate is imputed for all
employees aged 45 to 76.
Table A3: Employed People Making own Contributions to Superannuation
Schemes
Age
Male
Female
(Percentages)
Under 18
2.4
4.8
18 – 24
16.1
19.3
25 – 34
27.5
39.8
35 – 44
32.4
46.9
45 – 54
32.5
46.4
55 – 59
24.8
35.6
60 – 64
10.4
20.7
0.6
1.4
65+
Source: Calculated from RIM Conference Paper 98/1.
The reason for using these income-based employee data is to attempt to capture
changing rates of contribution. There is evidence that the favourable taxation rates on
superannuation contributions may induce some individuals to contribute large
243
amounts to their superannuation fund. This is particularly true for high income
individuals. Evidence of this can be seen in Figure A1.
The use of the ABS data shown in Table A4 attempts to model the relationship
between increasing contributions and increasing income.
Table A4: Employee Contributions to Superannuation Schemes
MALES (000’s)
45-76 years
Employee Contribution
Average
10Weekly Earnings Zero
< 3% 3-4% 4-5% 5-6% 6-10% 15%
Under $200
16.3
0.3
0.4
0.0
0.6
0.6
1.1
200-400
47.8
4.2
3.7
1.6
2.0
4.5
3.6
400-600
124.4
20.9 18.4
24.3 30.2
35.2 11.2
600-800
78.5
16.1 16.1
21.3 29.2
36.7 10.0
800-1000
35.5
12.0
9.3
19.0 24.7
28.7
7.5
1000+
49.3
14.8
9.4
22.1 29.5
33.3 14.3
Not reported
17.0
0.3
0.3
0.4
1.9
1.3
1.3
Total
368.8
68.6 57.6 88.7 118.1 140.3 49.0
FEMALES
(000’s)
45-76 years
Average
10Weekly Earnings Zero
< 3% 3-4% 4-5% 5-6% 6-10% 15%
Under $200
56.7
0.9
0.4
0.9
1.8
2.4
3.6
200-400
133.9
11.0
8.4
6.5 10.1 14.5
7.2
400-600
114.7
15.5 11.1 16.9 22.4 32.5
9.2
600-800
40.6
7.0
8.0 11.7 13.6 18.7
5.3
800-1000
8.7
4.2
2.1
6.7
6.0 14.0
2.4
1000+
6.0
2.0
0.0
1.9
3.9
2.5
1.7
Not reported
7.9
0.0
0.3
0.2
0.2
1.2
0.1
15+%
3.6
5.2
7.8
4.9
1.7
10.4
0.0
33.6
Total
23.9
75.9
284.8
224.4
147.7
195.0
37.4
989.1
15+%
5.8
6.3
4.6
2.1
0.8
1.5
0.2
Total
73.5
203.9
237.8
114.7
48.4
21.9
17.0
Total
368.5
40.6 30.3 44.8
58.0 85.8 29.5 21.3 717.2
Source: ABS, Cat. No. 6319.0, Superannuation Australia, November 1995, Table 13.
Number of Accounts
The existence of multiple accounts for individuals is well known and is discussed in
Rothman (1995). The reasons for the multiple accounts include:
•
•
•
•
people holding more than one job;
requirements that some award payments be paid into a different account;
timing difficulties, that is, old accounts being reported as current when in
reality they are inactive (for example the person may have changed jobs
and started in a new scheme/account and data relating to the old account
is still being reported); and
other unusual arrangements, such as where superannuation associated
with performance pay for public servants was put in a separate account.
DYNAMOD calculations are based on individual data rather than account data and
DYNAMOD assumes there are no differences between superannuation funds. These
assumptions mean that there is no difference between a person having three accounts
244
with $1000 in each and a person having $3000 in one account. The existence of
multiple accounts does not impact on the model.
Early Withdrawals
In general, funds contributed to superannuation schemes cannot be withdrawn until
retirement (and being aged over 54). However, the superannuation legislation did
allow access to funds in certain circumstances - namely times of severe financial
hardship or emigration from Australia. There was some abuse of these guidelines,
with the result that the legislation for gaining early access was tightened and the
preservation age increased to 60 years for those born after June 1964. From 1 July
1999, all earnings and contributions can only be paid out if a person dies, becomes
disabled, or APRA approves an early release. Permanent emigrates have to wait for
their funds until retirement. These changes have had such an impact as to make early
withdrawals a minor, short-term case which can be ignored. Early withdrawals are not
modelled in DYNAMOD.
Rate of Return on Invested Funds
An average rate of return for superannuation funds is very difficult to obtain. Firstly,
there are some 213 747 superannuation funds in Australia (APRA, June 2000).
Of these, 211 175 have less than five members. This number of small funds makes
calculation of an overall rate of return for superannuation funds difficult and perhaps
meaningless, as it is possible that a number of these small funds are not motivated by
maximising returns over the long term. Exclusion of these small funds reduces the
total assets by only 15 per cent (down $68b to $409b) and should allow the rate of
return of the remaining major funds to be calculated. However, a second problem with
measuring rates of returns then becomes evident: rates of return can still not be
Figure A1: Male Employees Contributing 10-15% to a Superannuation Fund
Proportion
10%
Employee Contributions
Males Contributing 10-15%
8%
6%
4%
2%
0%
Under
$200
$200-$400 $400-$600 $600-$800
$800$1000
$1000+
Weekly Income
Source: Calculated from ABS, Cat. No. 6319.0, Superannuation Australia, November 1995.
Accurately obtained due to the superannuation industry’s ‘reserving’ practice.
Reserving involves paying a smaller rate of return in good years and putting the extra
245
funds aside for use when the actual rate of return is negative or low. A number of
superannuation funds use reserving to implement a policy of not passing on negative
returns. In years when the actual return is negative or zero, the reserve is used to pay a
small positive return. The practice of using reserves results in a smoothing of rates and
makes measurement of actual real returns very difficult to obtain.
An InTech survey (2000) shows an asset weighted average for a range of major retail
growth funds of 6.3 per cent in real terms over the last seven years, and 8.4 per cent
over the last five. The Reserve Bank of Australia (RBA), in Occasional Paper No 8,
Historical Economic Statistics, show average ten year growth for a balanced fund of
13.0 per cent to 1995, but inflation was an average 5.4 per cent pa over the period
giving a real rate of 7.6 per cent. APRA and Treasury use eight per cent nominal as an
upper bound.
ARPA suggest that forward projections above five per cent or six per cent real are
likely to be unsustainable. The consensus view on these figures would suggest a real
rate of six per cent for 1986-94, eight per cent for 1995-99 and five per cent for 2000
onwards. These are rates used in DYNAMOD.
References
Antcliff, S. (1993), An Introduction to DYNAMOD, DYNAMOD Technical Paper No.
1, NATSEM University of Canberra.
ASFA (Association of Superannuation Funds of Australia) (2000), Super Facts Fact
Sheet #2, Sydney.
Brown, C. (1994), The Distribution of Private Sector Superannuation Assets by
Gender, Age and Salary of Members, RIM Conference Paper 94/2, Department of
the Treasury, Canberra.
Clare, Ross (1994), Women and superannuation, in Office of Economic Planning and
Advisory Committee (EPAC), Women and Superannuation, Background Paper
No. 41, Economic Planning and Advisory Committee and the Office of the Status
of Women, AGPS, Canberra, 3-28.
Foster, R. (1996), Australian Economic Statistics 1949-50 to 1996-7, Occasional
Paper No. 8, Reserve Bank of Australia, Canberra.
Goward, Pru (1997), Golden Eggs or Empty Nests – Reflections on Making
Retirement Better for Women, paper delivered to Women in Superannuation
Seminar, Canberra, 3 December.
King, A., H. Bækgaard, and M. Robinson (1999), DYNAMOD-2: An Overview,
Technical Paper No. 19, NATSEM, University of Canberra.
Knox, D. (1994), ‘The relationship between the age pension and superannuation
benefits, particularly for women’, in Economic Planning Advisory Committee
(EPAC), Women and Superannuation, Economic Planning and Advisory
Committee and the Office of the Status of Women, AGPS, Canberra, 49-60.
OECD (Organisation for Economic Cooperation and Development) (1996), Ageing in
OECD Countries: A Critical Policy Challenge, OECD, Paris.
Rothman, G. P. (1996), Aggregate and Distributional Analysis of Australian
Superannuation using the RIMGROUP Model, RIM Conference Paper 96/3,
Department of the Treasury, Canberra.
246
Rothman, G. P. (1997), Aggregate Analyses of Policies for Accessing Superannuation
Accumulations, RIM Conference Paper 97/2, Department of the Treasury,
Canberra.
Rothman, George P. (1998), Projections of Key Aggregates for Australia’s Aged, RIM
Conference paper 98/2, The Sixth Colloquium of Superannuation Researchers,
Department of the Treasury, Canberra.
Tinnion, J. (1998), Provision for Retirement Who Does What? Distribution of
Superannuation Contributions, Lump Sum Payments and Capital Income of Tax
Filers 1995-6, RIM Conference Paper 98/1, Department of the Treasury,
Canberra.
247
Mutual Obligation: A Reasonable Policy?
Pamela L. Kinnear
The Australian Institute
1
Introduction
According to Fox’s (2000) recent history of the Great Depression in Victoria, efforts
to assist the unemployed were characterised by an overarching desire to ‘remoralise’
unemployed people and promote a ‘reinvigorated work ethic’. Book reviewer, Brian
Dickey summarises Fox’s argument saying that ‘the notion that the unemployed had
lost their moral standing and hence their rights to full participation [was] the central
proposition guiding the responses of the middle class people who wanted to be
involved. It justified blame, coercion, management, rationed resources, and permitted
them to treat the unemployed as other’ (Dickey, 2001: 244).
This statement could be mistaken as a contemporary critique of modern welfare
reform, echoing sentiments expressed by academics and others in contemporary
debates about the policy of Mutual Obligation.1 Even during the Great Depression, it
seems, when major economic collapse left vast numbers of people facing
insurmountable structural barriers to employment and economic participation, the
argument that unemployment was a function of the failure of individual responsibility
was dominant.
There is an eerie similarity in the beliefs of the ‘re-moralisers’ during the 1930s with
the ‘re-moralisers’ of today. They are linked by a common denominator: the belief
that poverty and unemployment is a result of the failure of personal morality.
However, this basic idea has become increasingly veiled by more high-minded
supporting arguments over recent years. The shift of modern Western countries away
from ‘entitlement’ based social security systems and towards ‘conditional’ welfare
programs is based on a range of arguments:
1
•
that they are consistent with fundamental and longstanding principles of
liberal democratic traditions;
•
that entitlement systems counterproductively create ‘welfare dependency’, and
over-emphasise citizens’ rights at the expense of their responsibilities; and
•
that they are just plain common sense: that if the community has an obligation
to provide income support for some members, then those who are being
supported have corresponding obligations to the community.
The capitalised form of the term ‘Mutual Obligation’ will be used when referring to the
specific Government program that applies to certain groups of social security recipients.
Otherwise the lower case form is used to refer to the more generic concept of mutual
obligation.
Kinnear, P. (2002), ‘Mutual obligation a reasonable policy?’, in T. Eardley and B. Bradbury, eds,
PAMELA L. KINNEAR
In its simplest form, conditional welfare, manifest most clearly in the Australian
context in the policy of Mutual Obligation, is based on the idea that people should not
get something for nothing – an idea that the Howard Government consistently
maintains is ‘simple yet compelling’ (Newman, 1999: 3).
The popular appeal of these arguments has meant that in recent times the concept of
mutual obligation as applied to welfare policy has attained ‘motherhood’ status and is
accepted largely uncritically as a reasonable basis for social security policy; as one
policy commentator has said, mutual obligation is ‘here to stay’ (Curtain, 1999: 4).
The consequence is that the concept and policy framework of mutual obligation now
defines the boundaries within which those committed to better policy must operate; it
seems to be a step too far for agencies or individuals to criticise the concept of mutual
obligation itself. This is despite the fact that, when the philosophy and ‘common
sense’ that support the contemporary idea of mutual obligation are stripped away, it
becomes clear that the policy relies on the belief that a selected group of social
security recipients are not trying sufficiently hard to be self-reliant and, when left to
their own devices, will ‘free-ride’ on the backs of the rest of the community.
2
Reform and the Policy Process
In part, the ‘motherhood’ quality of mutual obligation is generated from its intuitive,
surface appeal and the success of arguments about its links to democratic tradition,
issues which are tackled in the remainder of this paper. However, it has also come
about because, whilst criticism of how the idea is implemented in government policy
has been vigorous, going further to critique the concept itself places critics ‘outside
the tent’ where they run the risk of being characterised as non-cooperative ideologues
and subsequently shut out of the policy process.
Social policy reform is an uneasy melting-pot of ideology, empiricism, compromise
and critique. To make a difference – to be ‘in the tent’ as it were – academics,
advocates and community workers find that they must often engage in uncomfortable
compromise. As the chair of the Reference Group on Welfare Reform, Patrick
McClure, has recently said in response to criticism in the media, often those in the tent
have to be satisfied that ‘half a loaf is better than no loaf at all’ (McClure, 2001; see
also the critique; Kinnear, 2001).
But it is increasingly the case that the ‘tent’ is becoming more and more crowded.
This is particularly so as governments require charitable agencies to be more
accountable for their public funds and as they pass more government-based work
through charities (such as providers of Job Network services) and university
departments (program evaluations and policy analysis). Moreover, those standing on
the outside looking in are a diminishing force.2 Whilst there is undoubtedly a role for
2
Traditionally the responsibility for social critique has been accepted by academics who, with the
protection of academic freedom, have been able to speak strongly and in an informed manner
about contentious social and political issues. Many, in fact, still do this, but tend to confine their
often strident and well-argued critiques to academic journals rather than placing their critique in
the public arena and help to swell the ranks of the so-called ‘side-line’ critics. However, as
recent work on the demise of academic freedom amongst Australian social scientists has
demonstrated, academics are experiencing considerable constraints on their freedom to say
uncomfortable things (Kayrooz, Kinnear and Preston, 2001).
249
MUTUAL OBLIGATION: A REASONABLE POLICY?
compromise and cooperation, the task of careful but strident critique based on
independent research and observation is an equally, if not more, important role in the
melting pot of social policy. The outsider is able to say the things that those engaged
in active policy compromises cannot, or choose not, to say. Often this will involve
challenging not just specific details of social policies, but also going a step further to
examine policies in their wider social, political and economic context as well as the
presuppositions about the nature of society and human nature that underpin policy
options.
There is little doubt that it is possible to make the policy of Mutual Obligation fairer,
less intrusive and less punitive. To some extent, the Welfare Reform process which
resulted in the new welfare regime now called ‘Australians Working Together’ has
achieved some improvements in the operation of Mutual Obligation, making it (at
least on paper) a little fairer, a little less intrusive and a little less punitive. There is
also no doubt that even more can be done in terms of such improvements. But the
reform process stopped short of challenging the idea of mutual obligation as a
reasonable basis for welfare policy.
This paper issues such a challenge. The following sections argue that despite its
popular appeal, the idea of mutual obligation is neither simple nor compelling. Rather,
the idea as applied to welfare policy is built on loose philosophical foundations and on
a number of popular misconceptions about the nature of unemployment and poverty.
3
Philosophical Premises: Mutual Obligation as a ‘Social
Contract’
The placing of specific obligations on recipients of income support is frequently
justified by the language of ‘social contract’: that the government of a society is based
on an actual or implied contract between citizens and the state. Indeed, the idea of the
‘social contract’ is the basis of mutual obligation’s popular appeal. Proponents argue
that the policy is fair and reasonable, because it has its origins in liberal democratic
philosophical traditions.
The OECD has noted that ‘the principle of mutual obligation as it applies in Australia
can be viewed as part of the implicit social contract that underlies the income support
system’ (cited in Curtain, 1999: 4). Others have noted the contractual nature of the
Mutual Obligation scheme and have argued that, as such, it constitutes a ‘reworking
of the idea of social contract’ (Yeatman, 1999: 255). Proponents of the Third Way
approach to welfare also champion the idea of renewing the social contract (Latham,
1998, 1999).
The idea of the social contract is particularly useful for proponents of conditional
welfare, as it is able to generate support across traditional ideological divisions. Not
only does this idea tap beliefs that ‘dole bludgers’ are cheating or abusing the
goodwill of taxpaying citizens, but it also appeals to those who are critical of
governments for not providing adequate assistance to people who may be trapped in
cycles of poverty and long-term reliance on income support.
So, does a social contract exist between citizens and the state? If so, what are the
obligations on the part of citizens, especially those who are disadvantaged? To answer
250
PAMELA L. KINNEAR
these questions, we must return to the original social contract philosophers and their
contemporary followers.
The Social Contract and Political Obligation
The idea that society is constituted by actual or implied contractual obligations has its
origins in 17th and 18th century philosophers such as Locke, Hobbes and Rousseau.
In the pursuit of a democratic alternative to the monarchical State, social contract
theory was principally concerned with the duty of citizens to obey the authority of the
State. Philosophers argued that because of the precarious and dangerous competition
in pre-political society (the ‘State of Nature’), individuals rationally decide to enter
into political society and consent to the rule of the State, agreeing to sacrifice certain
liberties in return for the State’s protection of their lives and property. Citizens’ duties
to each other are acknowledged by the principle that if individuals sacrifice certain
freedoms which yield mutual advantages, then others have a moral obligation to do
the same.
Although it is undoubtedly true that the idea of a social contract has ‘provided perhaps
the single most influential image of societal government in the history of the modern
West’ (Hindess, 1997: 15), it is nevertheless far from a straightforward idea. Indeed, it
is widely acknowledged to have significant problems and has never enjoyed a
consensus amongst philosophers. Because of this, linking contemporary debates about
welfare with this philosophical tradition is fraught with difficulty.
Despite centuries of debate, contract theorists have not been able to convincingly
identify how, if at all, individuals express such consent. Nor have theorists been able
to identify realistic alternatives for those who do not consent to political society.
Locke’s attempt to resolve the problem that people rarely, if ever, give express and
clear consent to the rule of the state, was to develop the idea of tacit consent. Tacit
consent is indicated simply by the fact that people reside in, use the resources of the
state and accept its benefits (Locke [1704], 1984: 177). Moreover, he argued, people
were free to leave a given state if they did not agree to its authority. However, as
Hume (1711-1776) pointed out, this does not resolve the problem as consenting
implies the possibility of doing otherwise, which is rarely available. Hume maintained
that
… such an implied consent can only have place where a man
imagines that the matter depends on his choice … [c]an we
seriously say that a poor peasant or artisan has a free choice to
leave his country, when he knows no foreign language or
manners and lives from day to day, by the small wages which he
acquires? (Hume [1748] 1947: 155-6)
In summary, it is a fairly weak contract in which the time, place and nature of the
initial agreement cannot be identified and in which no realistic alternatives to
contractual membership are available.
251
The Social Contract and Moral Obligation
For traditional social contract theorists, the problem of inequality was largely
irrelevant. Traditional theorists quite openly assumed that the principal parties to the
‘social contract’ were free, rational, property-owning men, for whom the motivation
to enter political society was to ensure the protection of property from the dangerous
and lawless ‘State of Nature’. Women, slaves, children (themselves considered
property) and non-property-owning members of society were thus irrelevant to the
philosophical debates.
Contemporary contract theorists, however, have been more interested than their
traditional counterparts to develop principles of social justice from the idea of the
social contract. Theorists of this tradition have questioned the extent to which those
who do not share equally in social benefits can be said to participate in a ‘social
contract’, and on what basis they can be said to incur obligations. In doing so, the
most prominent of contemporary contract theorists, John Rawls, placed important
qualifiers and limitations upon the types of obligations that apply to disadvantaged
people in an unequal society (Rawls, 1973). According to Rawls’ ‘principle of
fairness’ obligations can only be said to arise when two main conditions are met:
when institutions are just and when individuals are able to freely accept social benefits
in a context of meaningful alternatives (Rawls, 1973: 112).3
The implication of this is that once the principle of mutual obligation is translated into
specific policy, vital questions arise about the equity, proportionality and distribution
of obligations as well as how and under what circumstances such obligations can be
enforced.
As the following section will discuss, with trends such as widening inequality and
stubborn levels of structural unemployment, it is far from clear that Australian
institutional arrangements are sufficiently just to generate specific and enforceable
obligations on marginalised people. Nor do social security recipients freely accept
benefits in a context of meaningful alternatives. Although mutual obligation is based
on the belief that recipients choose to accept welfare benefits over paid employment,
in reality choice is limited and few meaningful alternatives to welfare benefits exist.
In summary, because of its primary focus on political obligation, traditional contract
theory is of marginal relevance to current debates about how to make fair welfare
policy. Moreover, citizens’ obligations to ‘do their part’ in the pursuit of mutual
advantage is a non-specific ethical principle and is of little help in determining the
nature and distribution of specific obligations amongst citizens in unequal societies
(Hindess, 1997: 18). In the light of this, we need to seriously question the extent to
which disadvantaged people who are dependent upon the community for basic
provisions should have significant, legally binding obligations in return.
Even devotees of the social contract tradition admit that ‘no fully satisfactory
prescriptive form of contractarianism has been generated’ (Hampton, 1995: 389).
Until it is able to overcome its many difficulties, the idea of the social contract holds
3
I wish to acknowledge the contribution of Dr Jeremy Moss in identifying and
elaborating the relevance of Rawls’ ‘principle of fairness’ to the policy of Mutual
Obligation.
252
PAMELA L. KINNEAR
more promise than practical usefulness as the basis of social policy in a complex
world (Hampton, 1995: 389).
Contemporary Contractualism
Despite the problems with the philosophical idea of the ‘social contract’, at the most
practical level contemporary societies are increasingly reliant upon contractual
arrangements for the governance of economic and social affairs (Hindess, 1997: 22).
As a result, contractual arrangements are being used in welfare policy to regulate the
behaviour of income-support recipients. The Australian Mutual Obligation scheme is
based upon a direct contract wherein the government agrees to provide income
support and job-search assistance in return for which the unemployed person agrees to
undertake a series of activities designed to maximise their employment chances.
A fundamental requirement of a contract is that the potential for exploitation is
checked by a broadly equal balance of duties and obligations. Parties entering a
contract are usually strongly encouraged by legal advisers and others to protect
themselves from exploitation by ensuring an approximate equality of power and
symmetricality of dependence and vulnerability. Thus, for true mutuality to exist,
dependency and vulnerability must be approximately equally shared between
contracting parties and each must have the option to withdraw from, or not to enter, a
contract if the possibility for exploitation is present. Under these conditions, contracts
are morally unobjectionable. However, a contractual relationship in which one party is
dependent upon the other for the provision of basic needs, and does not have the
realistic option to withdraw from the relationship because of this dependence, is one
of exploitation. Goodin argues that exploitation can be defined by a relationship in
which the relations of power are ‘asymmetrical’ and the dependence is ‘unilateral’
(Goodin, 1985: 196). He argues
the most morally objectionable dependency or vulnerability
relationship would exist where one party has discretionary
control over resources that the other needs and cannot obtain
elsewhere, yet no such dependency exists on the part of the first
party to the second. (Goodin, 1985: 201)
The Mutual Obligation contract is at once asymmetrical – the individual is dependent
upon the government to supply basic needs – and unilateral: the government has no
corresponding dependency.
There is little doubt that the situation in which most welfare recipients find themselves
in relation to the Mutual Obligation contract is of the exploitative type. People are
‘assisted’ through an inappropriate exercise of power (i.e. withholding of the means to
satisfy basic needs) in an unequal relationship (Hindess, 1997: 24-5).
A strange logical inconsistency is also present in the idea that unemployed people take
on obligations by signing a contract setting out the activities that they will undertake
in the search for work. As will be discussed in the next section, according to
advocates of conditional welfare, long-term welfare dependency causes a form of
incapacitating demoralisation that prevents people from acting in their own best
interests. As a result, it is the responsibility of government to reverse this situation by
forcing people to act responsibly (Giddens, 1998; Latham, 1998; Mead, 1997a;
253
Yeatman, 1999). However, contractualism relies heavily upon ‘an internalised,
subjective sense of obligation undertaken by those who are in a position to act upon
this sense’ (Hindess, 1997: 18). In order to participate in a contract, people must be
able to make rational choices about their own interests and have the capacity to adhere
to contracts. At the same time as acknowledging that people do not possess the
characteristics necessary to freely enter a contract, advocates of mutual obligation
argue that people will attain these characteristics if compelled to enter a contract that
prescribes socially responsible behaviour. Those who do not adhere to the terms of the
contract are penalised. Penalties are applied despite the fact that their difficulty in
complying may be directly due to their acknowledged incapacities.
Contractual welfare agreements, therefore, have all the hallmarks of a ‘sink or swim’
approach, i.e. people learn how to be contractual partners by participating in a
contract. In other words, in the full knowledge that they cannot swim, they are thrown
in the ‘deep end’ and told they must learn. If they sink, they are then penalised by
being thrown in again. Instead, in a fair arrangement, the existence of incapacity
should, in fact, negate the basis of the original contract.
Some argue, however, that the Mutual Obligation contract, although unequal, is
defensible, as the function of the imbalance is for the benefit of the less powerful
party – to assist him or her to realise their ‘deeper preference’, even though he or she
may not yet be conscious of what this is. Moreover, the contract really is reciprocal,
since the service provider is under an obligation to provide the service (Yeatman,
1999: 264). However, as long as compelling people under threat of the withdrawal of
their means of support remains the strategy by which disadvantaged people are to find
the path to their ‘deeper preference’, the contract with income support recipients,
despite its intentions, remains coercive and inconsistent with principles of justice.
There are a number of other assumptions underpinning policies of Mutual Obligation
that are not so directly related to traditional philosophical arguments, but are equally
influential in the public debate. They are equally loosely argued. The next section
analyses these assumptions.
4
Mutual Obligation and Social Structure
Mutual Obligation policies are based on a number of popular misconceptions about
the nature of unemployment and social structure. These misconceptions are largely
induced by the focus on individual responsibility that underpins the policy of Mutual
Obligation.
Structural Unemployment and the Issue of Choice
The policy of Mutual Obligation gives little recognition to the fact that the way our
economy is structured limits the extent to which the exercise of personal
responsibility is able to produce better outcomes for marginalised people. In fact, over
recent decades, the Australian economy has relied on the creation of joblessness
through microeconomic reform and trading off employment growth for inflation
control.
Since the downturn in the global economy that began in the 1970s, governments have
experienced difficulty in living up to the goal of full employment (Langmore and
254
PAMELA L. KINNEAR
Quiggin, 1994). Indeed, in practice Australian governments essentially abandoned the
commitment to full employment in the mid-1970s, favouring low inflation in the
trade-off between price stability and unemployment (Argy, 1998). As a result, during
the 1980s and 1990s, because Australian governments accepted quite high levels of
unemployment for the sake of what they considered to be ‘good economic
management’, long-term unemployment in particular has remained stubbornly high.
The policy of deliberately slowing employment growth when the economy is at risk
of overheating is based on the belief that pursuing full employment can be poor
economic management.
Moreover, it is precisely under conditions of structural unemployment that
entitlement-based benefits are more reasonable than conditional benefits. It is rather
ironic that the entitlement-based unemployment benefit was at its uncontested height
when it was least needed, that is, when the economy was never very far from full
employment, while conditions have been placed on benefits when unemployment is
high and jobs are not available for all. Under full-employment conditions, expecting
unemployed people to improve their chances of obtaining jobs that really existed
would be much less controversial and a truly ‘mutual’ exchange.4
When viewed in this light, it is evident that unemployed people have made an
involuntary sacrifice for the economic well-being of employed people. Until a new
commitment to full employment is made, it is fair to argue that the obligations are
reversed: Australians in positions of advantage should feel gratefully obliged to those
less advantaged for their considerable contribution to the well-being of the economy.
Although justified on the basis that contemporary culture is characterised by ‘too
much taking and not enough giving’, the call for disadvantaged people to make social
repayments under threat of their only means of support may itself be evidence of the
moral crisis of taking without giving.
The recognition that unemployment and other causes of welfare recipiency are rarely
a result of the failure of individual motivation, and are more usually the result of
economic management and structural impediments to economic participation,
undermines the belief so fundamental to policies of Mutual Obligation: that people are
obliged to make social repayments in return for their acceptance of benefits. But as
Rawls has pointed out, for the acceptance of benefits to generate obligations, such
acceptance must be freely chosen within a context of meaningful alternatives. The
reality of structural unemployment means that unemployed people have only a limited
degree of choice. Few meaningful alternatives to welfare benefits exist in
contemporary societies.
4
Although this contradiction may be partly due to affordability (i.e. entitlement based
welfare was affordable when few were claiming it), it is also explained by the fact that
the state ‘must be seen to be fulfilling its part of the work/welfare bargain’ even though
it acknowledges its inability to do so (Kerr and Savelsberg, 1999: 239).
255
Demoralisation
It is regularly asserted that long-term poverty and ‘welfare dependence’ have a
‘demoralising’5 effect. It is argued that despite being well-intentioned, passive
entitlement-based programs ‘kill with kindness’, paralysing individuals and
preventing them from being able, or sufficiently motivated, to act in their own best
interests: to continue the search for work, to seek out training options or to otherwise
work towards improving their life circumstances. This belief justifies compulsory
activity testing regimes, the application of penalties for non-compliance and, in some
cases, the withdrawal of benefits. The most influential justification for this can be
found in what has become known as ‘new paternalism’, most clearly and formally
articulated by US academic Lawrence Mead (see Mead, 1997a and 1997b). Mead
provides a succinct summary.
People who live without limits soon sacrifice their own interests
to immediate gratifications. To live effectively, people need
personal restraint to achieve their own long-run goals. In this
sense, obligation is the precondition of freedom. Those who
would be free must first be bound. And if people have not been
effectively bound by functioning families and neighbourhoods in
their formative years, government must attempt to provide the
limits later, imperfect though they must be. (Mead, 1997: 23)
According to Government policy, conditional welfare regimes are designed to make it
‘harder for demoralised job seekers to give up’ the search for work (Abbott, 2000) and
sends a positive message to unemployed people that they can control their situation,
counteracting the stifling hopelessness of being told that the problem is beyond their
control (Abbott, 2000).
However, research on unemployed people has quite consistently found that depression
and ‘demoralisation’ sets in after a considerable period of unsuccessful job-searching.
(see for example, Scholzman and Verba, 1979; Jahoda, 1982; Feather, 1982; Fineman,
1987). People may begin their period of unemployment believing that their personal
efforts will eventually succeed, but this optimism fades with repeated failure.6
Although there is ongoing debate about the exact causes of ‘demoralisation’, the
research literature has repeatedly demonstrated the psychological impact of constantly
frustrated job-searching efforts and the hopelessness that unemployment creates, that
‘drains energy and initiative [and] makes effort seem pointless’ so that eventually lack
of effort becomes a rational response (Allan, 1997: 8).
5
6
The word ‘demoralisation’ is quite ambiguous. On the one hand, it may be taken to
mean a sense of frustration and disappointment to the point of a lack of desire to
persevere; on the other hand, it may mean a degeneration of moral values. Current
political rhetoric as well as some academic writing on the subject appears to play on the
ambiguity, attributing both meanings, in undefined ways, to the term.
See especially, Jahoda, 1982: 92, 97; Turner, 1983; Lane, 1991: 170. Although some
may continue to believe that this is due to their own personal deficit, others attribute the
cause to factors outside themselves. The research literature has yet to settle on whether
internal or externalisation of blame has greater effect. Furnham provides a useful
discussion of the literature relating to the internal/external attribution of blame and
social psychological responses to unemployment (Furnham, 1988: 170, 629). Either
way, damage is initiated by the personal experience of frustrated hopes and efforts.
256
PAMELA L. KINNEAR
‘Making it harder to give up’ the demoralising, frustrating and depressing search for
work under threat of the reduction or withdrawal of basic income support sends a very
clear message to unemployed people that they only have themselves to blame for their
personal situation: that unemployment is caused by the failure of individual
motivation. Doing so ignores and dismisses the very real frustrations that unemployed
people face and shows a distinct lack of targeted or creative policy-making.
Discriminatory Application
The policy of Mutual Obligation imposes specific obligations enforceable by
compulsion on certain categories of welfare recipients but does not apply similar
standards to others who have benefited from public resources. Unemployed people are
singled out as the group required to ‘give something back’ whilst others who rely
upon public resources do not. This is nowhere better illustrated than in the 2001-02
Federal budget which announced the extension of some Mutual Obligation programs
and requirements to people up to the age of 49 and the application of new conditions
under a mutual obligation framework to people over 50 but not yet of pensionable
age.7 These changes were announced concurrently with a raft of new advantageous
policies for pensioners and pension-aged Australians. In the words of the Treasurer,
older Australians ‘deserve’ such benefits because of their lifetime’s contribution to
nation-building. Thus, Australians of pensionable age are deemed to have already
made their contribution and thus ‘deserve’ their benefits, whilst Australians only a
few years younger are deemed to have not yet made their contribution. On reaching
pensionable age however, the ‘undeserving’ suddenly become ‘deserving’.
Others who receive community assistance in times of need are also not required to
‘give something back’ in return: accident victims who require emergency treatment in
a hospital are granted assistance without incurring a debt to society; those who rely
upon public health because they cannot afford private cover are not required to give
something back in return for their assistance; the right of adventurers to expensive,
publicly funded rescue operations is rarely questioned, despite the fact that in such
cases, although the vagaries of nature are blamed, the decision of individuals to place
themselves at risk is a particularly personal one.
In addition, public subsidies are extended to the corporate world to encourage
investment or to promote and sustain certain industries, with very few obligations or
accountability mechanisms in return. So extensive is this subsidisation, and so
minimal are the associated obligations that such activities have been dubbed
‘corporate welfare’ (Baragwanath and Howe, 2000; Whitfield, 2001).
Whilst the idea of mutual obligation implies that all citizens have obligations to ‘do
their part’, under the policy of Mutual Obligation, income support recipients are the
only citizens who are selected out for the enforcement of obligations. Other citizens
7
From July 2003, claimants over 50, but not yet of pensionable age, will be placed on
Newstart instead of the current Mature Age Allowance or Partner Allowance. Whilst in
many respects, this is a step forward in that it allows older unemployed people access to
Jobsearch services not previously available to them, it also means that these individuals
become subject to requirements under a mutual obligation framework: specifically,
Mature Age people on Newstart will be required to submit Jobsearch diaries every three
months. At the same time, new claimants for the Widows Allowance will be required to
attend an annual interview with a Centrelink Personal Adviser.
257
are either left free to exercise discretion in relation to their social obligations, or are
provided with significant incentives to encourage them to do so. The selection of the
‘stick’ rather than the ‘carrot’ as a regulatory device in social security policy contrasts
with neo-liberal regulatory philosophy applied elsewhere. Self-regulation through
persuasion and incentive (the ‘carrot’) rather than punishment (‘the stick’) is the
strategy of choice for most governments in many contemporary situations. In
occupational health and safety, or environmental protection, for example, government
policy generally favours strategies of self-regulation or persuasion over immediate
resort to direct penalties. Regulatory theory has developed sophisticated ‘enforcement
pyramids’ which move through a large number of steps, beginning with persuasion
and ending, as a last and reluctant resort, with the full enforcement of penalties
(Braithwaite and Ayres, 1992). Moreover, voluntary schemes are increasingly relied
upon in environmental regulation and in proposals to replace corporate tax with
corporate philanthropy.
In summary, the broad ethical ideal of mutual obligation – that citizenship is a shared
journey in which everyone does their part for the mutual benefit of all – is distorted
when attempts are made to match benefits with a corresponding obligation in a quid
pro quo manner. This is especially true when this is selectively enforced in specific
and intrusive ways for income support recipients.
In the light of these difficulties, one must conclude that the motivation behind
imposing mandatory conditions upon unemployed people is the assumption that they
are disproportionately irresponsible and liable to abuse the benefit system.
False Dichotomy
The idea that people on unemployment benefits should give something back to the
community in return for their benefit is based on a false distinction between ‘givers’
(working taxpayers) and ‘receivers’ (non-working, non-taxpayers). It is hard to go
beyond Senator Jocelyn Newman’s description in her 1999 address announcing
Welfare Reform of an ‘indulgent welfare mentality’ that the ‘hard working men and
women of this country cannot be expected to underwrite’ (Newman, 1999: 5).
In fact, however, not only have income support recipients paid income taxes in the
past and are likely to do so in the future, they are also current tax-payers due to
various forms of indirect taxation and the GST, as well as through direct taxation of
some benefits. The historical decision to finance Australia’s social security system
from public revenue rather than from wage-based contributions was an explicit
recognition of these essential interdependencies.
The inconsistency of the idea that society is made up of contributors and noncontributors is again illustrated by the benefits provided to older Australians in the
2001-02 budget. The budget focused explicitly on the contribution of older
Australians as a group: their collective contribution to national well-being. Regardless
of their individual circumstances, Australians of pensionable age - rich or poor,
employed or unemployed, single parents or not, able bodied or not - have been
recognised for collectively making Australia what it is today. In contrast, however,
Mutual Obligation policies focus on the extent of individual contributions of workingage citizens who currently rely on income support. These groups are not considered to
be contributing to the collective effort of nation-building; instead, it is assumed that
258
PAMELA L. KINNEAR
they are ‘taking’ rather than ‘giving’. Because of this, these groups are forced to ‘give
back’ through the activities required by Mutual Obligation.
5
Where To From Here?
In reviewing Fox’s (2000) historical account of unemployment in 1930s Victoria,
outlined at the opening of this paper, Dickey (2001) concludes by saying that the
book’s polemic against the ‘return of the modern re-moralisers’ is thoroughly justified
and is a ‘timely call to be ever vigilant in the defense of a widely shared citizenship’.
Stripping away the surface plausibility to the policy of ‘Mutual Obligation’ is an
important step in this endeavour. Far from being a simple and compelling idea based
on shared values of personal responsibility and fairness, the principle of mutual
obligation applied to welfare is complex, replete with inconsistencies and ethical and
logical flaws. The basis of the idea in the ‘social contract’ tradition is weak and is far
from as ‘simple’ or ‘compelling’ as we are led to believe. Nor is the idea of the
‘contract’ a useful basis for establishing obligations under situations of unequal
power. Moreover, the policy of Mutual Obligation does not take account of structural
component of unemployment; it is applied in highly selective and discriminatory ways
to welfare recipients; it is founded on a false distinction between taxpaying and nontaxpaying citizens; and it uncritically relies on compulsion as a regulatory device
without considering other strategies that may be more creative, more just and more in
keeping with broader principles of democratic freedom.
Rather than simply making policies of Mutual Obligation fairer and less punitive,
these difficulties are of sufficient magnitude to warrant the removal of the idea of
mutual obligation from its central place in modern welfare policy.
So what would serve as a better basis for fair welfare policy? The problems identified
here suggest a number of guiding principles.
•
Policies should assume that income support recipients are honest and upright
citizens, a few of whom (as in society generally) may abuse the system.
•
Policies should explicitly acknowledge that the need for support arises from the
failure of society to provide opportunities for all, rather than the personal
failings of individual recipients.
•
Policies should acknowledge that those in work have benefited from the
disadvantaged situation of the unemployed because unemployment results
mainly from the failure of the economy and economic management.
•
The mutuality should be balanced. Governments should undertake, on behalf of
society, programs and policies designed to overcome structural disadvantage.
•
Consent to any ‘contract’ between income support recipients and the
Government can only be assumed where realistic alternatives to income support
are widely available and accessible.
•
Systems of monitoring income support recipients should focus on non-punitive
methods of compliance management. Penalties should genuinely be an option of
last resort.
259
•
6
Accountability for, and enforcement of, obligations should be applied
consistently to all parties to the contract.
Conclusion
It is hard to disagree with the argument that society needs to maintain a careful
balance between rights and responsibilities. Allan argues that, however unpalatable it
may be for those of liberal persuasions, the conservative critics have been right … in
detecting something amiss in the moral culture of the West … some unbalancing of
rights and entitlements to the neglect of social obligations and civic virtues. (Allan,
1997: 15)
Moreover, the father of modern communitarianism, Amitai Etzioni has claimed that
‘to take and not to give is an amoral, self-centred predisposition that no society can
tolerate’ (Etzioni, 1995: 10).
These ideas go to the heart of our morality, whether secular or religious, and are a
foundation stone of political philosophy. Indeed, it is difficult to imagine wishing to
live in a society in which these qualities and commitments were absent. They speak of
a just, caring and responsible society where self-interest is not the driving force,
where people behave in ways that consider their own individual interests and desires
in the context of the well-being of others and the collective good.
But it is difficult to see that the current policy of Mutual Obligation extends these
principles or generates a social environment in which such ideals can find fulfilment.
It could even be argued that the call of those in a position of advantage for those less
advantaged to nevertheless make social repayments under threat of the withdrawal of
their only means of support is itself evidence of the moral crisis of ‘taking without
giving’.
References
Abbott, T. (2000), ‘Job seekers with attitude’, http://www.dewrsb.gov.au/
ministers/abbott/ speeches/2000/sp2march_00.htm, 2 March.
Allan, R. (1997), ‘Money for nothing: the ethics of social welfare’, Policy
Organisation and Society, 14, Winter, 1-22.
Argy, F. (1998), Australia at the Crossroads: Radical Free Market or Progressive
Liberalism?, Allen & Unwin, St Leonards.
Baragwanath C. and J. Howe (2000), Corporate Welfare: Public Accountability for
Industry Assistance, Discussion Paper No. 34, The Australia Institute,
Canberra.
Braithwaite, J. and I. Ayres (1992), Responsive Regulation: Transcending the
Deregulation Debate, Oxford University Press, New York.
Curtain, R. (1999), Mutual Obligation: Intention and Practice in Australia Compared
with the UK and the USA, Dusseldorp Skills Forum, Sydney.
Dickey, B. (2001), Book Review in Labour History, 80, 242-4.
Etzioni, A. (1995), The Spirit of Community: Rights, Responsibilities and the
Communitarian Agenda, Fontana Press, London.
260
PAMELA L. KINNEAR
Feather, N. (1982), ‘Unemployment and its psychological correlates: a study of
depression symptoms, of self-esteem, Protestant ethic values and attributional
style and apathy’, Australian Journal of Psychology, 34, 309-23.
Fineman, S., ed., (1987), Unemployment: Personal and Social Consequences,
Tavistock Publications, London.
Fox, Charlie (2000), Fighting Back: The Politics of the Unemployed in Victoria in the
Great Depression, Melbourne University Press, Melbourne.
Furnham, A. (1988), ‘Unemployment’ in, W.F. Van Raaj, G.M.Van Veldhoven and
K.E. Warneryd, eds, Handbook of Economic Psychology, Kluwer Academic
Publishers, Dodrecht.
Giddens, A. (1998), The Third Way: the Renewal of Social Democracy, Policy Press,
Cambridge and Massachussetts.
Goodin, R.E. (1985), Protecting the Vulnerable: A Reanalysis of our Social
Responsibilities, University of Chicago Press, Chicago.
Hampton, J. (1995), ‘Contract and consent’, in P. Pettit and R.E. Goodin, eds, A
Companion to Contemporary Political Philosophy, Blackwell Publishers Ltd,
Oxford, 379-93.
Hindess, B. (1997), ‘A society governed by contract?’, in G. Davis, B. Sullivan and
A. Yeatman, eds, The New Contractualism, Macmillan, South Melbourne.
Hume,D. (1711, 1980) ‘Of the original contract’, in E. Barker, ed., Social Contract:
Essays by Locke, Hume and Rousseau, Greenwood Press, Westport,
Connecticut (reprint of a version by Oxford University Press, 1969).
Jahoda, M. (1982), Employment and Unemployment: A Social-Psychological
Analysis, Cambridge University Press, Cambridge.
Kayrooz, C., P. Kinnear and P. Preston (2001), Commercialisation and Academic
Freedom: Results of a Study of Australian Social Scientists, Discussion Paper
No. 37, The Australia Institute, Canberra.
Kerr, L. and H. Savelsberg (1999), ‘Unemployment and civic responsibility in
Australia: towards a new social contract’, Critical Social Policy, 19(2), 23355.
Kinnear, P. (2001), ‘Welfare groups left out in the cold’, The Canberra Times, 5 June.
Lane, R. (1991), The Market Experience, Cambridge University Press, Cambridge.
Langmore, J. and J. Quiggin (1994), Work For All: Full Employment in the Nineties,
Melbourne University Press, Melbourne.
Latham, M. (1998), Civilising Global Capital: New Thinking for Australian Labor,
Allen & Unwin, St Leonards.
Latham, M. (1999), From Cape York to Campbelltown: Rebuilding Communities,
Speech for the Brisbane Institute, Customs House, Brisbane.
Locke, J. (1704, 1984), Two Treatises of Government, J.M. Dent and Sons, London.
McClure, P. (2001), ‘Half a loaf is better than no loaf at all’, The Canberra Times, 8
June.
Mead, L.M., ed., (1997a), The New Paternalism: Supervisory Approaches to Poverty,
Brookings Institution Press, Washington, DC.
Mead, L.M. (1997b), ‘Welfare employment.’ in L.M. Mead, ed., The New
Paternalism: Supervisory Approaches to Poverty, The Brookings Institution
Press, Washington DC.
Newman, J. (1999), The Future of Welfare in the 21st Century, National Press Club,
Canberra, 29 September, http://www.facs.gov.au/internet/new…/
Rawls, J. (1973), A Theory of Justice, Oxford University Press, Oxford.
261
Schlozman, K. and S. Verba (1979), Injury to Insult: Unemployment, Class and
Political Response, Harvard University Press, Cambridge.
Turner, M. (1983), Stuck! Unemployed People talk to Michele Turner, Penguin Books
Ltd, Ringwood, Victoria.
Whitfield, D. (2001), Public Services or Corporate Welfare? Rethinking the Nation
State in the Global Economy, Pluto Press, London.
Yeatman, A. (1999), ‘Mutual obligation: what kind of contract is this?’, in S. Shaver
and P. Saunders, eds, Social Policy in the 21st Century: Justice and
Responsibility, Reports and Proceedings No. 141, Social Policy Research
Centre, University of NSW, Sydney.
262
Individualised Funding in Disability
Services
Carmel Laragy
Department of Human Services
Victoria
1
Introduction
This paper reviews individualised funding and the principles of self-determination
that underpin it. The extent to which these principles have been applied in a Victorian
post-school options program, the Futures for Young Adults (hereafter FFYA)
program, will be examined. This program commenced in 1997 to assist young people
with a disability to move from school to adult day service options. Its intention was to
provide young people with services that meet their individual needs as well as to
promote the development of new services as necessary. The lessons learnt from this
project will be considered in light of the general principles of individualised funding.
2
Self-determination and Individualised Funding
Self-determination is formally defined as occurring when a person has control over
their life and the choices made (Wehmeyer, Agran and Hughes, 1998: 32). A less
formal definition is that it allows people to follow their dreams and to form meaningful
relationships (Bach, 2000: 8). Central to self-determination is the principle that all
people, including those with severe developmental disabilities, can be taught to take
some degree of control over their own lives (Wehmeyer, 1999: 11). It is acknowledged
that there may be limits to the degree of control people with an intellectual disability
can exercise (Joseph Rowntree Foundation, 1995), but Wehmeyer argues that all people
can be educated to enhance their decision making ability and become more selfdetermining. Individualised funding is one of a range of mechanisms that is used to
assist people to become more self-determining and it is generally seen as a means to an
end and not an end in itself. A description of individualised funding is provided below.
Shaddock (2000: 5) outlined the progression of attitudes towards people with a
disability that has led to the current interest in self-determination. Until the 1960s, the
medical model was dominant and when this gave way to deinstitutionalisation and
independent living during that decade, professionals still provided top-down support
and people with disabilities still had little power and influence. The following decade
saw the ‘normalisation principle’, a major catalyst for reform, and moves towards
individualised planning. Individualised planning came to prominence in the 1970s
and, under an umbrella of ‘management by objectives’ bequeathed by the business
world, it involved interdisciplinary coordination, accountability, integration and
consumer participation. This came to be criticised because it still retained the pivotal
role of the professional, had a bureaucratic style for life planning and emphasised
quality program delivery and not quality lifestyles. When these limitations were
recognised, person-centred-planning was introduced. This focused on the person and
their wishes and aspirations, but in turn this was found wanting because it often
Laragy, C. (2002), ‘Individualised funding in disability services’, in T. Eardley and B. Bradbury, eds,
INDIVIDUALISED FUNDING IN DISABILITY SERVICES
simply repackaged old practices in a new vocabulary. The progress to selfdetermination is summarised by Shaddock as moving from ‘planning for’ to ‘planning
with’ to ‘planning by’ people with a disability.
Many innovative programs have developed around the globe aiming to increase the
involvement and self-determination of people with disabilities. One mechanism that
has attracted much interest and debate is individualised funding. This term covers a
wide range of funding mechanisms based on the principle that funding is put under
the control of the individual (Dowson and Salisbury, 2001: 35; Belli, 2000: 1). As
Dowson and Salisbury described, individualised funding means that funds no longer
go from the funder (government or its representative) to service providers, but go
directly to the individual who requires assistance. Individualised funding gives control
of the funds to the person so they can purchase the services they require, sometimes
with the assistance of a broker or other agent. The intention is that the person will
determine the services needed and their needs will shape the service system.
… the effect of this switch in the flow of funding is to shake the
foundations of the entire service system. The relationship between
the providers and receivers of support is turned upside down
(Dowson and Salisbury, 2001: 35).
A wide variety of mechanisms has developed to implement individualised funding
and assist people with the responsibilities it imposes. These include service brokers,
personal agents, fiscal intermediaries and vouchering mechanisms, and they are all
designed to assist with budgeting, selecting services, managing payments and
accountability (O’Brien, 1999: 3). Across Europe, Canada and the United States of
America (US) support organisations have developed to help people with this new
funding structure because it is a radical departure from traditional services and it has
brought many challenges.
Individualised funding developed in two areas during the 1970s (Dowson and
Salisbury, 2001: 36; Bach, 2000: 1). In British Columbia, Canada, the Woodlands
Parents Group assisted their children with intellectual disabilities as they moved from
an institution to individually designed and funded programs. Around the same time in
California in the US, a group of people with physical disabilities formed Personal
Assistant Services and demanded control over the funding previously given by the
government to support agencies. They were able to employ their own assistants and
felt they gained real control over their own lives (Dowson and Salisbury. 2001:36).
There is wide agreement that service systems need to change and be better focused on
people’s needs. This requires better relations with clients and a wider range of flexible
alternatives (Moseley, 1999: 2; Kendrick, 2000: 15; Bigby, 1999: 56).
Individualized funding and planning systems must be flexible and
responsive to the culture, values and preferences of each person and
their family (Dowson and Salisbury, 2001: 37).
However, Dowson and Salisbury (2000: 1) joined others in cautioning that
individualised funding is complex, and even though it brings benefits to some, it
should be implemented incrementally and cautiously. There are service systems that
offer the opportunity for people with disabilities to choose the level of self-sufficiency
they require, ranging from traditional agency based services to the opportunities for
264
CARMEL LARAGY
self-management. Tilly, Wiener and Evans Cuellar (2000) argued that offering this
range of opportunities is the ultimate in self-determination and should be the goal of
all services.
There are concerns that the difficulties in implementing individualised funding have
obscured the original intention of giving people with disabilities greater control and a
better quality of life (Bach, 2000: 2). Bach reported instances where people had
received funds, but they did not gain increased power and control over their lives. He
found instances where adults and their parents or carers were even not aware they had
individualised funding and support workers, service providers, social workers and
case managers continued to make the decisions about how funds were spent. On the
other hand, there have been additional concerns expressed in discussions about
individualised funding. These relate to the removal of safeguards previously provided
by professional involvement which may leave people more subservient to the wishes
of their parents/carers. Support groups called ‘Circles of Friends’ and ‘Microboards’
developed in the United Kingdom and in Canada respectively, to provide a wide base
of support and protection to the person with a disability.
Amongst the concerns about individualised funding is the fear that existing services
will no longer be viable and the choices available to people with disabilities will be
diminished. There are some who argue that a demand driven service system, based on
a market economy model, is the best way to address needs. However, others argue
passionately that a market economy does not, and cannot, address the needs of the
most vulnerable (Ripper, 2000; Simons, 2000). This latter group states that
government regulation and support for community development will always be
required. For those who accept market-based models, splits between purchasers and
providers, stringent accountability and accreditation are considered essential (Dutton,
2000; Tilly et al., 2000: 18).
Some writers are concerned that governments will abdicate all responsibilities for
individual support and service development once they give direct payments, and
research has vindicated this concern (Bach, 2000: 4). There are also concerns, often
from unions, that individualised funding will destroy hard won benefits and securities
for workers that have been gained through group bargaining, and will leave them
exploited as they individually negotiate each job (National Union Canada, 1998).
There is some evidence in the literature, but not a lot, to justify these concerns (Tilly
et al., 2000: 18). The unions were criticised as being self-serving and narrow by
Bleasdale (2000: 1), who suggested that their social justice ethos positions them well
to work cooperatively with people with disabilities.
3
Rights
Underpinning the moves to self-determination and individualised funding is an
increased focus on the rights of the person with a disability. In recent decades the
rights of people with a disability have been increasingly recognised and this is evident
through legislation, individualised services and increased consultation with people
with a disability. The 1975 United Nations Declaration of Rights of Disabled Persons
provided a foundation and a guide for many nations to formulate their own legislation
and implementation strategies. A commitment to the rights of young people with a
disability underpins the FFYA program.
265
The Government of Victoria incorporated these principles into its Intellectually
Disabled Persons’ Services Act (IDPSA) 1986, where it included a Statement of
Principles giving people with an intellectual disability the same rights as other
members of the community to access services in order to support a reasonable quality
of life; a right to individualised educational and developmental opportunities; and a
right to exercise maximum control over every aspect of their life. These principles
were extended to all people with a disability in the Disability Services Act 1991 (DSA)
which followed the 1991 Commonwealth/State Disability Agreement (CSDA). The
changes in Australia parallel similar changes in other countries.
Attitudes towards people with a disability in the Western world have shown major
changes over the past hundred years (Wehmeyer et al., 1998: 32). At the turn of the
20th Century, people with a disability were viewed as a menace and locked away
from the rest of the community. In the 1920s, the eugenics movement recommended
the eradication of the ‘feeble-minded’ and this led to segregation and sterilisation.
Following World War II, the many veterans with disabilities were shown more
humane and sympathetic attitudes. Hand in hand with the development of science and
medicines, where the medical model was dominant, people with disabilities were seen
as patients to be cured. The medical profession in particular, but also psychology,
social work and teaching dominated decision-making. People with disabilities were
viewed as ‘victims’ to be pitied. It was in this context that the disability rights and
advocacy movement began with parents playing a more outspoken role (Wehmeyer et
al., 1998: 42).
The recognition of the right to be included in general community services has driven
many changes in service delivery. The strength of the inclusion movement is shown
internationally by the June 2001 Executive Order from The White House (USA)
directing that people with a disability be included in community services. This
followed a court decision which decided that the norm is to be community inclusion
and exceptions have to be justified. In Victoria, the One Community initiative is based
on the principle of inclusion and RuralAccess is an example of a program working
towards inclusion by strengthening community services so that supports do exist for
people.
4
Futures for Young Adults Program
The Futures for Young Adults (FFYA) program assists young people with all types of
disabilities in Victoria as they move from school to adult service options. Services
were provided to approximately 3 250 students between the years 1997 and 2000. The
need for transition programs to assist students with a disability to move to post-school
options has been well recognised in Victoria and elsewhere (Community Services &
Ministry of Education, Report of the Working Party on Students with Intellectual
Disability Aged Over 18, 1989:5). The 18+ Transfer Project, existing from 1990 to
1994, was the predecessor to the FFYA program. The 18+ Transfer Project was not an
individually planned transition program and was referred to as the whole of school
transition as whole schools of students were moved together to one adult service
(Schofield 1998: 7).
In November 1996, the Honourable Dr Denis Napthine, Minister for Youth and
Community Services, announced the new Futures for Young Adults program in the
Victorian State Parliament (Hansard, Victoria, 1996). This was in response to
266
CARMEL LARAGY
the 1 270 young people with an intellectual or physical disability over the age of 18
years who were in Victorian state schools.
Important features of the FFYA program announced by Dr Napthine were the
emphasis on individual needs, self-determination and the development of new options
to meet these needs. He stated:
The government has committed $17 million in recurrent funding to
provide services for young adults and an additional $10 million in
capital and start-up costs to facilitate the program. The program
must recognise the individual skills, abilities and needs of young
adult persons. …I advise the house that we will treat each of them
as an individual because we recognise that people with disabilities
are individuals and no one size fits all in the services provided.
…… The service was designed specifically to meet the needs of
individuals and their carers. …… In consultation with the young
person, with his or her family and with the school he or she is
currently involved with, we will develop specific programs for
each of these individuals to ensure that the most appropriate adult
option is available. …. We have the option of providing further
day programs, particularly for young adults with multiple
disabilities or severe disabilities or challenging behaviour. We will
also look at employment options, whether in the open work force
or in the assisted employment work force (13 November 1996,
Hansard).
The processes identified for the FFYA program relating to self-determination were:
•
Individual transition planning in which the young adults and their families are full
participants. This process is to include a formal review of each young person's
initial choice, and a review of their needs over time;
•
Recurrent and ongoing client-centred funding which is portable and travels with
the young adult as they move between service providers;
•
Service sector development with the involvement of existing service providers and
the introduction of new service providers and new models of service delivery
(DHS, FFYA Stage 1 Implementation, 1998:6, internal document).
5
Futures for Young Adults Program Funding
Dowson and Salisbury, (2001: 35) defined individualised funding as funds going from
the government funder to the person or their representative so they can purchase
services directly from service providers. The term individualised funding is currently
referred to in relation to the FFYA program but it is not present in the original
documentation about the program. Using the definition given above, individual
funding was only found in the FFYA program for less than ten young people, who
were self-managing through the assistance of their families. The overwhelming
majority of young people receives services negotiated and paid for directly by
government transition workers or their agents. The program does provide
transportable funding, in that participants can choose to move from one service
provider to another and the funds go with them. When necessary, services can use the
267
funds to create new places. This is a radical departure from previous programs where
service providers were funded by government to offer a set number of places and
service users remained on waiting lists until a vacancy arose. The FFYA program has
created two categories of service user, those on the FFYA program who can take their
funds to another provider and those outside the accepted age range who have to wait
for a vacancy. Not surprisingly, there are equity concerns about this discrepancy.
The FFYA includes an assessment of the support needs of the young adult using the
Support Needs Assessment (SNA) which has six levels of funding. There has been
some confusion about these funds and the level of control that participants and
families have over their use. Parents who wanted to be closely involved in decisionmaking and those who wanted to self-manage using an individualised funding model
wanted to know their entitlement. However, it was reported that some regions did not
reveal the amount of funds available. The program administrators viewed the SNA as
a means of distributing funds from the central office to each of the nine Department of
Human Services (hereafter DHS) regions, and transition workers are expected to
assess individual planning needs, negotiate with service providers and budget
accordingly. Funds not expended individually have been used to support service
development within the region. Additional service and equipment grants were also
provided to regions.
Overview of the Program Evaluation
The Futures for Young Adults (FFYA) evaluation examined the impact of the
program on all stakeholders. As stated by Minister Napthine in 1996, the program
intended to assist students in their move from school to further education, training,
employment and recreational activities and to facilitate the development of a flexible,
consumer- responsive service system. However, there were no clearly stated aims and
objectives for the program that could be used for its evaluation.
The first stage of the evaluation, the Destinations Evaluation (DHS, 2000) analysed
all existing DHS data files and presented a profile of the adult options attended by the
young people. The second stage, the Program Design Evaluation, consulted with
young people on the program, their families or carers and a range of service providers
and other stakeholders. A sample of young people on the FFYA program was
randomly selected based on the nine regions, three broad levels of assessed need,
gender and the first four years of the program under review. Some sample cells were
empty, giving an actual sample of 180 instead of the 216 that was theoretically
possible. One hundred and one young people or their families responded to letters sent
to participants and their families or to phone calls from the Department of Human
Services requesting their participation. Some declined involvement and the names of
those who agreed were passed to the university researchers. Seventy-nine participants
and their families/carers were interviewed.
Invitations to submit written submissions or have discussions with the researchers
were advertised through peak bodies and disability networks to additional
participants, families/carers, schools, further education bodies, service providers and
advocacy groups. A total of 122 interviews were conducted or submissions received
from these invitations. Additional consultations were undertaken with government
staff administering the program around the state. A large number of the young people
interviewed had communication impairments. To understand their expectations and
268
CARMEL LARAGY
aspirations, the researchers spent time visiting their current placements and talking to
the young people.
The Destinations Evaluation profiled the service types that students moved to, across
the nine Victorian DHS regions during the first four years of the program from 1997
to 2000, and showed the changes they made. The findings considered relevant to a
discussion of individualised funding in disability services are the profile of services
used and the numbers who moved to a second service option. It should be noted there
were differences in data collection across the nine regions that limited the accuracy of
the data and limited the findings to an approximation of the actual numbers. Despite
this limitation, the major trends are considered valid.
Statistical Analysis of Placement Options
The analysis of the placements chosen by the young people showed the majority of
young people attended disability specific services. The data did not specify type of
disability, but it was known the majority had an intellectual disability. Approximately
40 per cent attended traditional government funded day services (ATSS) for people
with an intellectual disability. A further 16 per cent attended TAFE disability- specific
courses and 10 per cent attended supported employment. Other options attended were
community based programs (sometimes part of an ATSS service) 10 per cent; open
employment eight per cent; open TAFE five per cent; other three per cent; university
two per cent; VCE two per cent; traineeship two per cent; adult education two per
cent; apprenticeship 0.2 per cent; no placement four per cent; no support requested
three per cent. (Total exceeds 100 per cent because of rounding.)
The movement from first to second placement option was analysed and only 16 per
cent of the young people were shown to have moved to a second option. This figure
may be an underestimate because some moves might not have been recorded, but the
consultations confirmed that relatively few people changed options. It was also found
that while the proportion of young people in each of the service categories remained
much the same after the first transition, there was a tendency for those leaving the
specially created TAFE disability courses to move to disability day services.
Unfortunately, there are no figures available to compare the choice of options with
those prior to the commencement of the FFYA program.
Overview of FFYA Service System
The consultations with young people and their parents/carers covered wide-ranging
issues and only those relevant to a discussion of individualised funding will be
reported. Dowson and Salisbury (2000: 65) presented a table detailing different
mechanisms for implementing individualised funding and how they assisted the
different stakeholders. Aspects of their table relevant to the FFYA have been adapted
and are presented below with additional comments on the FFYA program.
269
Table 1: Individualised funding structures
STRATEGY
INTENDED
TO ASSIST
PURPOSE
IMPACT
Dowson and Salisbury 2000
Banding
recipients
into broad
categories of
need
State fundholders
Control overall
spending
APPLICABILITY TO FFYA
PROGRAM
FFYA Findings
May undermine
true recognition of
individual
funding
requirements
FFYA program uses an assessment
tool with six categories of need.
Funding allocations were linked to
level of need and funds go from the
central government office to
regional offices. Individual plans
are made at the regional level and
amounts spent may vary from the
initial allocation. As Dowson and
Salisbury suggested, the bands
serve the purpose of allocating
funds within the bureaucracy.
The tool was criticised by many
service providers and some parents
on two accounts. They said it
focused on physical support needed
in the school and did not take into
account relevant factors that
influence future support needs,
such as aspirations, levels of
motivation, potential, emotional
state, challenging behaviour, social
skills, family background, literacy
and numeracy needs. Also, some
criticised the level of funding
allocated to the bands.
270
CARMEL LARAGY
STRATEGY
INTENDED
TO ASSIST
PURPOSE
IMPACT
Approved
provider list
State fundholders
Ensure
minimum
range and
quality of
services
PROGRAM
FFYA Findings
May undermine
development of
market
responsive to
consumer
demand
271
Expressions of interest were
invited from existing disability
services to provide additional
options for FFYA participants. The
majority of the young people
attended services that had a service
agreement with the government
prior to the program commencing.
Providers were accountable to
government and not the young
people. It was not possible to
establish in this study whether new
services had developed in response
to demand, which had been one of
the original intentions. While some
service providers reported their
services had become more
innovative, others said their
viability was under threat and this
made them less responsive and
innovative.
STRATEGY
INTENDED
TO ASSIST
PURPOSE
IMPACT
Crisis
contingency
funds
State fundholders
Provide fast
and flexible
responses
PROGRAM
FFYA Findings
If used
excessively,
may deny
recipients
ability to plan
from guaranteed
funding
allocation
'Special needs funding' is applied
for in a crisis and can be re-applied
for every six weeks (some regions
have extended this to longer
periods).
Service
providers
reported that they regularly
depended upon it to provide
necessary support for some young
people with high support needs
because funding was inadequate.
The inadequacy of the general
funding level for these young
people is obscured. It also raises
planning problems as the funds
were not assured, a point made by
Dowson and Salisbury.
The 'special needs funding' in the
FFYA program does assist the
recipients, although Dowson and
Salisbury do not include them as
beneficiaries.
Case
(or Care)
managers
Recipients
(ostensibly,
though
covertly may
serve state
fund-holders)
Plan, select,
and manage
supports
Severe conflicts
of interest if
accountable to
either state
fund-holder or
service provider
Transition workers did the basic
planning and their responsibilities
varied across regions. Some
worked exclusively within the
FFYA program while others also
carried
case
management
responsibilities in other areas.
Having access to a (familiar)
worker was a critical factor in
determining satisfaction with the
FFYA program. In all but two
regions, where transition workers
were employed by independent
services, they were government
employees and their primary
accountability was to the funding
body. As Dowson and Salisbury
suggest, this could lead to strong
conflicts of interest.
272
CARMEL LARAGY
STRATEGY
INTENDED
TO ASSIST
PURPOSE
IMPACT
PROGRAM
Service
brokers
Recipients
Negotiate
funds; plan,
select, and
manage
supports
FFYA Findings
Undermines the
control of the
recipient if the
broker has
loyalties or
obligations to
providers or
fund-holders
The FFYA program was fully
contracted through independent
service brokers in two of the nine
regions.
In
other
regions,
contracted services sometimes
brokered to additional services.
Many respondents in the evaluation
insisted that brokerage had to be
independent of service providers to
ensure impartiality.
One of the regions with
independent brokerage recorded
some of the most positive client
comments in the study, while
comments about services in the
other region were on par with the
moderately
highly
rated
government services.
Providers
given
guarantees of
minimum
income/busin
ess by state
funding
agency
Providers
Survive the
uncertainties of
demand-led
market
(especially
during
transition from
block-funded
system).
May obstruct
development of
demand-led
market.
The government had service
agreements with traditional service
providers
prior
to
the
commencement of the FFYA
program and they were invited to
submit an expression of interest to
participate in the program. Many
respondents held similar views to
Dowson and Salisbury and said the
close relationship between the
government and some providers
resulted in the promotion of a
restricted number of services with
young people being directed to
preferred services.
Table adapted from work of Dowson and Salisbury, (2000: 5)
Control
A significant feature of individualised funding, as previously discussed, is
empowering of service users. The aim of self-determination, the principle that
underpins individualised funding, is to give people as full and rich a life as possible
based on their decisions, to the extent that they are able to choose. Rigid service
273
systems were thought to limit people’s options. The intention of portable and
individualised funding is to make services responsive to people and not the reverse
where people have to fit into existing services.
Two findings from the Destinations Evaluation considered relevant to assessing the
degree of control that people had over their choice of option were the extent to which
non-traditional options were included, and also the degree of movement from one
option to the next over time. In both areas the results were conservative, with
traditional service options predominating and little movement between services.
Neither of these measures on their own were definitive in determining the degree of
self-determination that was afforded to young people and their parents/carers, as
people might prefer traditional services, and indeed many did. However, in
conjunction with the interviews conducted, they do give an indication that people
were very restricted in their choice. Some parents/carers reported they wanted more
choice regarding their services, but they found it difficult to achieve this in their
negotiations with regional offices. Also, it was frequently mentioned that information
was difficult to access. The nine government regions across the state differed widely
with regard to the degree of control they shared with families.
The researchers became aware of only a small number of people who had payments
made directly to them, or to a brokering agency acting on their behalf. Less than ten
families were mentioned in the study as being funded in this way, and three families
participated in the consultations. These three families said they had control and could
purchase the services they wanted and expressed great satisfaction with these
arrangements. The workload was difficult when families self-managed, but they
appreciated the choice and flexibility they achieved.
It should be noted that many of the families who did not control their own funds were
also satisfied with the degree of control they experienced and the services received,
and often did not want any further responsibility in their lives. However, according to
definitions of individualised funding, it cannot be said they were part of an
individualised funding program.
… a system which merely tells people what their service costs are,
without handing over control of funds or allowing any
renegotiation, has no resemblance to IF (individualised funding)
(Dowson and Salisbury, 2000: 65).
There were varying levels of satisfaction reported in the study regarding the level of
control and services experienced. Many expressed their appreciation of having
portable funding. This gave the potential to move to new services, even if the statistics
showed that only a minority actually did so. Other factors that led to people feeling
they had some involvement and control over the choices made were: access to
transition workers; access to information; access to an appeal body (a Consumer
Advocacy Reference Group was disbanded in the first year of FFYA and its loss
decried by some who knew of it); and a welcoming attitude from workers (some
parents/carers feared they would be labelled trouble makers if they complained and
'rocked the boat').
The study found many parents/carers who did not know that the program was
intended to be responsive to their needs, nor that it offered portable funding. Bach
274
CARMEL LARAGY
(2000: 2) reported similar findings where the workers determined funding allocations
and made major decisions.
Service Innovation
One of the intentions of the FFYA program was to be responsive to individual needs
and develop new programs if necessary. This intention is consistent with other
individualised funding programs attempting to break away from the limitations of
traditional services and give people the financial power to direct service development
through their choices (Dowson and Salisbury, 2001: 35).
The study found that portable funding allowed new places to be created in existing
services and it fast tracked access so that people moved off waiting lists. However, the
study was not able to determine the impact of the FFYA program on the development
of new services at a time when political directives resulted in major changes to the
services system. There were anecdotal reports that the imposition of competition
policy forced the closure of smaller disability services, and also that the remaining
services developed programs that were more innovative and responsive. However, the
extent of these changes and the contribution of the FFYA program cannot be gauged
by this study. It can be determined, however, that few FFYA participants moved
outside the established service system. It was also evident from the study that the
infrastructure within government and the contracted services remained focused on
group activities and it was difficult to cost and administer individual programs that
differed from the norm. Payments from government to services were processed
around annual service agreements and computer technology was designed for bulk
payments and not individual flexibility.
An important factor to be considered when judging the extent of change to the service
system associated with the FFYA program is the slow place at which change typically
occurs in complex administrative systems. Pollitt and Bouckaert (2000: 33) argued
that existing procedures reflect organisational history and large-scale change is always
difficult, requiring considerable investment in staff training, new information
technology and new accounting systems. Change in each of these areas is critical for
the full implementation of the FFYA program and both the government department
and service providers were grappling with the profound changes needed to provide
individually focused and funded services. The slow rate of change observed is
understandable from this perspective, but it also gives impetus to putting resources
into the areas identified. Without adequate technology and accounting systems, and
without educating staff in a new philosophy and new practices, the potential of the
program may not be realised.
Some service providers mentioned that individualised funding threatened their service
viability. They feared that portable funding could lead to a sudden loss of people
attending and this was of particular concern to some small providers. There was
acknowledgement from some service providers that they sometimes retained young
people at their service to bolster their numbers when it may have been appropriate to
refer them on to other options. Even though the actual number of young people in the
FFYA program who move is small, the loss of one or two was sometimes seen as a
serious threat.
275
The individualised funding model requires service users to act like consumers in a
marketplace. However, the study found parents/carers did not want to have to
negotiate with service providers about extra costs for activities and transport that is
now required in many services. Parents/carers were unaccustomed and uncomfortable
with this role. As discussed previously, many writers recommend a flexible service
system that accommodates those who want to manage their own funds as well as
those who want a well-defined service. The findings of this study support the need for
this varied approach.
The study was able to confirm the importance of positive relationships between young
people and their families/carers with service providers in determining satisfaction
with services, but it did not examine whether this led to a change in service delivery.
Kendrick (1999: 9) observed that the hoped for innovation in service reform does not
automatically flow from individualised funding. He suggested that individualised
funding is only a financing method and the critical factor in achieving innovation lies
in relationships between people and in the potential of staff to be responsive and
creative. While pockets of innovation were found, there were also services that
appeared to be unresponsive to individual needs.
7
Lessons
The cautions in the literature about the move to individualised services and
individualised funding being complex and needing to be implemented incrementally
were born out in this study (Dowson and Salisbury, 2000: 1). The FFYA evaluation
provides an opportunity to learn from the experiences of the past four years and
develop further opportunities. The program was a bold initiative that produced gains
for the young people for whom new places have been created because of portable
funding. A handful of families have full control of their funding and have directly
purchased their own services. This opportunity has been greatly appreciated by the
families concerned.
There are many challenges when a large service system tries to change from block
funding arrangements to individual responsiveness. Every person in the service
system including recipients, parents/carers, providers, funders and policy makers has
to travel the path of significant change. An ongoing commitment to implementing a
changed philosophy is required. The FFYA program has examples where people have
been empowered to follow their dreams and services have changed, and these can
serve as models for the whole program. However, it is clear that many people need
further support for changes to occur. Procedural changes are needed if people with a
disability are to have greater input into decision-making to the extent that they want
and are able. Disability advocacy and peak bodies demonstrated a depth of
understanding of the overall service system in their submissions, and they could
possibly be supported to play an important role in bridging the divide between
government funders and service users. Most importantly, staff in all government and
non-government areas will need support to adapt to the radical changes that are being
attempted.
276
CARMEL LARAGY
References
Bach, M. (2000), ‘Individualised funding and self-determination: making sure the
means does not become the end’ unpublished paper presented at First
International Conference on Self-determination and Individualised Funding,
Seattle, USA, July, http://members.home.net/directfunding/Materials.htm.
Belli, R. (2000), ‘Individualised funding: one key to self-determination’, unpublished
paper presented at First International Conference on Self-determination and
Individualised
Funding,
Seattle,
USA,
July
http://members.home.net/directfunding/Materials.htm
Bigby, C. (1999), ‘International trends in intellectual disability P\policy in the L\late
1990s’, in Ozanne, E. et al., Reframing Opportunities for People with an
Intellectual Disability, School of Social Work, The University of Melbourne,
Victoria.
Bleasdale, M. (2000), ‘Regulating IF - the case for union involvement’, unpublished
paper presented at First International Conference on Self-determination and
Individualised
Funding,
Seattle,
USA,
http://members.home.net/directfunding/Materials.htm.
Community Services and Ministry of Education, Victoria (1989), Report of the
Working Party on Students with Intellectual Disability Aged Over 18, Victoria.
Department of Human Services (2000), Futures for Young Adults Program, DHS.
Melbourne.
Department of Human Services (1999), Futures for Young Adults Program, Business
Rules and Guidelines, DisAbility Services Branch, DHS, Melbourne (internal
document).
Dowson, S. and B. Salisbury (2000), ‘Individualised funding – emerging policy
issues’, Abstract, Vol 4, Issue 2, 61-75.
Dowson, D. and B. Salisbury, (2001), ‘Foundations for Freedom: international
perspectives on self-determination and individualised funding’, based on First
International Conference on Self-determination and Individualised Funding,
Seattle, USA.
Dutton, D. (2000), ‘Can we trust consumerism, or is regulation necessary?’,
unpublished paper presented at First International Conference on Selfdetermination and Individualised Funding, Seattle, USA, July.
Hansard, 13 November (1996), House of Assembly, Victoria, Questions without
notice, Honourable Dr Denis Napthine
Joseph Rowntree Foundation (1995) ‘Increasing user control in social services: the
value of the service brokerage model’, Social Care Summary 4 - May,
http://www.jrf.org.uk/home.asp.
Kendrick, M. (1999), ‘Formal individualisation systems: their potential and
limitations’, Crucial Times, July, 9-10.
Kendrick, M. (2000), ‘When people matter more than systems’, keynote presentation
at The Promise of Opportunity Conference, New York State Developmental
Disabilities Planning Council, Albany, USA.
Moseley, C. (1999), Making Self-Determination Work, Robert Wood Johnson
Foundation, http://www.self-determination.org/links1309/links.htm .
National Union Canada (1998), ‘The Hard Truth about Individualised Funding’,
published on Individualized Funding Information Resources web site.
http://members.home.net/tsalisbury/
277
O’Brien, J. (1999), ‘Community engagement: a necessary condition for selfdetermination
and
individualised
funding,
http://members.home.net/directfunding/Materials.htm.
Pollitt, C. and G. Bouckaert (2000), Public Management Reform, a Comparative
Analysis, Oxford, New York.
Ripper, P. (2000), 'IF in Australia', unpublished paper at The First International
Conference on Self-determination and Individualised Funding, Seattle, USA,
July.
Schofield, G. (1998), Parent and Carer Perceptions of the Needs of Young Adults
with Multiple and Severe Disabilities in Transition from School to Adult
Settings, unpublished Masters thesis, Deakin University, Victoria.
Shaddock, A. (2000), ‘Person directed planning: where is it going and is it a good
place to go?’, unpublished paper presented at the 8th National Joint Conference
of the National Council on Intellectual disability and the Australian Society for
the Study of Intellectual Disability, Fremantle, Western Australia.
Simons, K. (2000), ‘Can we trust consumerism to make IF work?’, unpublished paper
presented at First International Conference on Self-determination and
Individualised
Funding,
Seattle,
USA,
July.
The White House 2001, United States of America, Executive Order, CommunityBased Alternatives For Individuals With Disabilities, 18 June 2001.
http://64.14.118.139/topnews/Current_Releases/0619-105.html
Tilly, J., J. Wiener and A. Evans Cuellar (2000), ‘Consumer-directed home and
community services programs in five countries: policy issues for older people
and government’ unpublished paper, The Urban Institute, Washington, DC,
USA.
Wehmeyer, M., M. Agran and C. Hughes (1998), Teaching Self-Determination to
Students with Disabilities, Paul H. Brooks Publishing Co., Baltimore, Maryland,
USA.
Wehmeyer, M. (1999) ‘A bridge to where? Transition, self-determination, and the
new millennium’, unpublished paper presented at the 1999 annual OSEP Project
Directors' Meeting, Washington, DC.
278
Worlds Apart: Postcodes with the Highest
and Lowest Poverty Rates in Today's
Australia
Rachel Lloyd, Ann Harding and Harry Greenwell1
NATSEM, University of Canberra
1
Introduction
This paper aims to add to existing research on poverty and regional diversity by
exploring the extent of poverty in small regional areas. Poverty analysis on a regional
basis has previously been severely hampered by a lack of suitable data. The unit
record files from the Australian Bureau of Statistics (ABS) household expenditure and
income surveys allow analysis at state and very broad regional levels. However, such
disaggregation results in small sample sizes and the two Territories are usually
collapsed together by the ABS, so that results cannot be derived for either the ACT or
the Northern Territory. While the national population Census provides data for small
regions (down to Census Collector District level – about 200 households), income
data are limited to gross household income ranges. Gross household income is not
generally regarded as the best income measure for poverty analysis, with most
analysts preferring an after-tax income measure, adjusted by an equivalence scale to
take account of varying needs due to differences in household size and composition
(for example, ABS, 1998; Saunders, 1996; Harding and Szukalska, 1999). A second
problem with the Census data is that, because the income data are in ranges, the
‘poverty line’ has to be set at the boundary of one of the income ranges.
NATSEM has recently developed the capacity to estimate poverty at a detailed
regional level. Marketinfo is a cutting edge synthetic regional data model that
provides sociodemographic, income and expenditure data for each Census Collectors
District (CCD). The Marketinfo model blends the 1996 Census CDATA and the
Household Expenditure Survey (HES) unit record file, so as to effectively provide a
synthetic HES unit record file for every CCD in Australia.
In 2000 NATSEM used Marketinfo/99 to provide a snapshot of poverty in the ACT,
including the demographic characteristics and spending patterns of people in poverty
1
Rachel Lloyd is a Senior Research Fellow at NATSEM. Ann Harding is Professor of
Applied Economics and Social Policy at the University of Canberra and inaugural
Director of NATSEM. Harry Greenwell is a Research Officer at NATSEM. Aspects of
this work were supported by Australian Research Council Grant No. A79803294. The
authors would like to thank Otto Hellwig of MDS Market Data Systems for producing
the postcode weights used in this analysis.
Lloyd, R., A. Harding and H. Greenwell (2002), Worlds apart: postcodes with the highest and lowest
poverty rates in Australia’, in T. Eardley and B. Bradbury, eds, Competing Visions: Refereed
Proceedings of the National Social Policy Conference 2001, SPRC Report 1/02, Social Policy
RACHEL LLOYD, ANN HARDING AND HARRY GREENWELL
(Harding, et al., 2000). The information at CCD level was aggregated to provide
estimates of the number of people living in poverty in each of the statistical
subdivisions (roughly equivalent to the town centres) of the ACT.
With the release of the 1998-99 Household Expenditure Survey, a new version of
Marketinfo – Marketinfo/2001 – has recently been developed. The postcode weights
from Marketinfo/2001 were combined in this study with the income information from
the HES to give preliminary estimates of poverty rates by postcode in 1998-99 and to
look at the characteristics of people in poverty in the postcodes in each state with the
highest and lowest poverty rates. We also examined the characteristics of each of the
selected postcodes to see what factors were driving particularly high and low poverty
rates. This is a new field of research and this paper should be seen as a presentation of
preliminary results and our first attempt at using techniques that will become more
sophisticated over time.
2
Data and Methodology
Data Source
This study uses income and demographic information from the 1998-99 Household
Expenditure Survey, combined with postcode weights for 2000 derived from
Marketinfo/2001, to derive regional poverty estimates.2 It is hoped that future work on
this project will allow us to use income information from STINMOD/01A, a HESbased version of NATSEM’s static microsimulation model, rather than the 1998-99
tax and income figures recorded in the original 1998-99 HES. STINMOD simulates
income tax and the major cash transfer programs administered by the Department of
Family and Community Services and the Department of Veterans’ Affairs. The latest
version of STINMOD/01A models the major changes to Australia’s tax and transfer
systems introduced on 1 July 2000. The ABS has not yet conducted an income survey
since the tax reforms so STINMOD will provide a unique opportunity to estimate how
individual families fare under the New Tax System (TNTS).
Marketinfo/2001 is a synthetic data set created by combining the 1996 Census
CDATA with the 1998-99 Household Expenditure Survey (HES) unit record file. The
Census surveys the whole population and provides detailed sociodemographic data on
a street block (Census Collector District – CCD) level. The HES provides more
detailed income information than the Census and it also includes extensive
expenditure data. Marketinfo uses sociodemographic variables common to the Census
and the HES to merge the two surveys. The resulting micro data set contains
sociodemographic, income and expenditure information for each CCD in Australia.
This method overcomes the sample size problems of using sample surveys such as the
HES directly for analysis at a state or broad regional level. It also allows analysis at a
detailed regional level, such as CCDs, postcodes, statistical local areas, statistical
subdivisions or electorates. In this study, the information was aggregated to postcode
2
Marketinfo weights for 1998-99 were not available for this study, so 2000 weights
were the best possible match. A possible improvement would be to uprate the
incomes from the HES from 1998-99 to 2000, but this is not likely to change the
overall results.
280
WORLDS APART: POVERTY RATES IN TODAY’S AUSTRALIA
level and poverty rates were estimated for each postcode so as to identify the
postcodes in each state with the highest and lowest poverty rates.
Ageing the Population and Data: The 1996 sociodemographic profiles shown for
each Census Collectors District in the 1996 Census have been uprated to estimated
2000 population levels using ABS data on dwelling commencements and estimates of
demolitions for each CCD, along with labour force survey data on labour force
characteristics by region. The incomes are as recorded in the HES unit record file.
Unit of Analysis: Marketinfo is derived from the Census and the HES and hence
provides data at the household level. Consequently, this paper uses households as the
unit of analysis. This effectively assumes that there is complete income sharing within
households. The HES person file was also used to derive some variables.
Validity: The results of this study are based on simulated data and the techniques are
at the cutting edge of poverty research. However, the model has solid foundations in
terms of the original data and the techniques. Its predecessors have been used for
market research purposes since 1993 and have been benchmarked against other data
sources. Policy makers frequently make us of simulated data from high quality model
when actual data are not available. Nonetheless, it should be appreciated that this is
among the first of NATSEM’s attempts to use the new synthetic regional income
database for poverty analysis and that subsequent efforts will no doubt embody more
sophisticated techniques and more recent data.
In a major report for the Smith Family written in 2000, NATSEM estimated the
before-housing Henderson half-average income poverty rate in 1999 to be 13.3 per
cent (Harding and Szukalska, 2000). These results were derived using the 1997-98
Survey of Income and Housing Costs (SIHC) with incomes uprated to 1999. The
Australian average poverty rate in this study, which uses the 1998-99 Household
Expenditure Survey with Marketinfo weights, is estimated to be 10.3 per cent.
Part of the difference can be attributed to the fact that this study uses households as
the unit of analysis while the Smith Family report was based on income unit analysis.
An analysis of the poverty rates at income unit and household level in the 1997-98
SIHC showed that household level poverty rates were approximately 2.5 percentage
points less than income unit level rates.
Another possible source of difference is the varying income distributions between the
HES and the SIHC – with our initial explorations suggesting that the HES income
distribution may be more equal than that shown in the SIHC. For these reasons, this
paper should be seen as a work-in-progress and the results as indicative rather than
definitive.
Defining Poverty
Australians generally do not suffer the severe material deprivation evident in some
developing countries. This affects our definition of poverty. In this study, poverty
applies not only to individuals without food or shelter, but also to those whose living
standards fall below some overall community standard. This relative poverty
definition underpins most estimates of the number of Australians in poverty (ABS,
1998).
281
There is no universally accepted measure of poverty. All of the decisions made by
analysts in defining and measuring poverty are subject to heated debate. In this report
we analyse the number of people living in poverty using the half-average income
poverty line allied with the Henderson detailed equivalence scales. This Henderson
half-average poverty line is defined for a benchmark household type, such as a couple
with two children, and then the Henderson equivalence scales are used to determine
comparable poverty lines for other types of households. If a household’s disposable
(that is, after income tax) income falls below the poverty line, we deem that they are
in poverty. The poverty rate (or risk) is the proportion of all households of a particular
type that fall below a given poverty line. In future work we hope to examine the
consistency of our results by using other poverty measures, such as the Henderson
poverty line, the Henderson half-median, the modified OECD half-median and the
‘International scale’ half-median poverty lines (see Harding and Szukalska, 2000, for
more information on these various poverty lines).
Equivalence Scales: The financial circumstances of a household are dependent not
only on the income of the household but also on its composition. For example, a
single person with a disposable income of $19 000 is unlikely to suffer from the same
degree of poverty as a couple with four children on the same income. Equivalence
scales provide a way of defining poverty levels for families of different composition.
Results can vary greatly depending on the equivalence scale used. The detailed
Henderson equivalence scale, which was used in this study, was derived from a
survey of household budgets and costs in New York in the 1950s. Despite this, it has
been widely used in Australia as a standard method for equivalising incomes. The
detailed Henderson equivalence scale takes account of the gender, age and labour
force status of the head, the age and labour force status of the spouse and other adults,
and the ages of dependent children. The original Henderson approach assigned higher
‘working’ points to people who were either working full time or unemployed and
looking for full-time work. In this study, the ‘working’ points have also been assigned
to those who are working part time and to those who are unemployed and looking for
part-time work.
The Henderson equivalence scale has been applied at the household level. Because of
this, we have given all non-dependent adults who live in the household the same
points as a spouse. In the case of a household consisting of three unrelated single
people we assign the reference points to the person deemed to be the household head
and points equivalent to spouse points to the other two. Other studies using the family
or the ABS income unit as the unit of analysis would, in contrast, have assigned head
points to all three adults. This is one key reason why our results vary from other
poverty estimates (for example, in Harding and Szukalska, 2000). The 1996 Census
and consequently Marketinfo/2001 do not, however, allow easy analysis at any level
other than the household.
Poverty Lines: Poverty lines are levels of income and are different for each type of
income unit: in our study the income unit is the household. If a household’s income
falls below the poverty line for that type of household then the household is
considered to be in poverty.
The Henderson half-average poverty measure sets the standard poverty line at half of
the average equivalent disposable household income for a standard household. The
282
standard household consists of a couple both under 40 years old with the husband
working and the wife not in the labour force with two children, a boy aged 6-14 and a
girl aged under 6 years old. The Henderson half-average poverty line was $400 per
week. (This is looking at all people in Australia, not just those living in the selected
postcodes.) Poverty lines for other family types are derived using the Henderson
detailed equivalence scales.
There is no consensus about whether the median or the average is to be preferred as
the poverty benchmark, with available studies using both (for example, Layte, Nolan,
and Whelan, 2000; Strengmann-Kuhn, 2000). The median has the advantage of being
less affected by extreme values than the average. For example, large increases in the
highest incomes will cause the average to increase but alone will not have an effect on
the median. On the other hand, during an era of rising income inequality, there is
concern that the incomes of those at the top end might increase substantially, while
still leaving median incomes - and thus the poverty rate - unaffected if the poverty
line is set at half median income. For that reason we have chosen to use the halfaverage income poverty line in this study.
3
Poverty Rates in Poor and Rich Postcodes
Poverty Rates
Using the methods described above, before housing Henderson half-average poverty
rates were estimated for each of the postcodes in Australia. From these, we chose the
postcodes with the highest and lowest poverty rates in each state. We excluded
postcodes with fewer than 1000 households as we chose to focus on postcodes of
reasonable size rather than those that were small and often special in their nature. For
example, Kapooka had a low poverty rate but consisted of only 92 households
associated with an army base. Conversely, Brim in Victoria had a low estimated gross
average household income of only $23 168 and was assigned one of the highest
recorded poverty rates in our study. However, we estimated that less than 100
households lived in Brim, so we excluded it from our analysis, as such a small size
increased the possibility of sampling or other errors. Table 1 lists the number of
postcodes in each state and the number that were excluded because they contain less
than 1000 households.
Table 1: Number of Postcodes and Number with less than 1000 Households, by
State
New South Wales
Victoria
Queensland
South Australia
Western Australia
Tasmania
Total number of
postcodes
589
625
392
321
289
108
Number of postcodes with
less than 1000 households
181
305
154
178
171
62
Source: 1998-99 HES and Marketinfo/2001 weights
Tables 2 to 7 provide an overview of the poverty rates in the top and bottom
postcodes in each state compared with the relevant state and Australian averages.
283
NSW Postcodes with Highest and Lowest Poverty Rates (a)
Table 2:
Postcode number
Postcode name
Poverty rates
People
Adults
Children
Note:
Source:
Highest poverty
rate
Lowest poverty rate
2834
2088
Lightning Ridge
Spit Junction
%
25.9
22.4
39.6
%
0.7
0.9
0.2
NSW
average
Australian
average
%
9.8
8.8
12.5
%
10.3
9.3
12.9
a) Using the Henderson half average poverty line. Poverty rates are measured at the
household level, which means they are not directly comparable to most other poverty
studies. Only includes postcodes with over 1000 households.
1998-99 HES and Marketinfo/2001 weights
The postcode with the highest poverty rate is Ferryden Park, a suburb of Adelaide,
where almost one-third of people are estimated to live in poverty. This contrasts with
the mining community of Roxby Downs, 560 km north of Adelaide, which has a
poverty rate of only 1.4 per cent.
The 2088 postcode on Sydney’s north shore, which includes the suburbs of Mosman
and Balmoral, has a poverty rate of just 0.7 per cent. While New South Wales has the
lowest average poverty rate of all the states (9.8 per cent), the postcode of Lightning
Ridge in the west of the state has over one-quarter of its people and 40 per cent of its
children living in poverty.
Victoria has poverty rates below the national average, with the lowest poverty rate in
the Melbourne bayside suburb of Brighton and the highest in Carlton South, close to
the University of Melbourne.
Table 3: Victorian Postcodes with Highest and Lowest Poverty Rates (a)
Highest poverty Lowest poverty Victorian
rate
rate
average
Postcode number
3053
3186
Postcode name
Carlton South
Brighton
Poverty rates
%
%
%
People
25.2
1.4
10.1
Adults
22.4
1.3
9.2
Children
38.5
1.8
12.8
Note:
Source:
Australian
average
%
10.3
9.3
12.9
studies.
284
Table 4: Queensland Postcodes with Highest and Lowest Poverty Rates (a)
Highest poverty Lowest poverty Queensland
rate
rate
average
Postcode number
Postcode name
Poverty rates
People
Adults
Children
Notes:
Source:
4671
Gin Gin
%
21.6
20.0
25.9
4069
Kenmore
%
3.4
2.9
4.6
%
10.6
9.6
13.3
Australian
average
%
10.3
9.3
12.9
studies.
Tasmania has the highest average poverty rate of the states, at 13 per cent, but there is
less diversity between the top and bottom postcodes. Taroona, an outer suburb of
Hobart, has Tasmania’s lowest poverty rate of 6.4 per cent (although this is
significantly higher than the lowest poverty rate in any other state). St. Mary’s has
Tasmania’s highest poverty rate, with about one-fifth of its residents in poverty.
Most of the postcodes with the lowest poverty rates are in metropolitan areas - the
exception is Roxby Downs – but the postcodes with the highest poverty rates are more
diverse. Lightning Ridge, Gin Gin and St. Mary’s are rural, and Carlton South,
Ferryden Park and Perth City are metropolitan.
Table 5: South Australian Postcodes with Highest and Lowest Poverty Rates (a)
Highest
poverty rate
Postcode number
Postcode name
Poverty rates
People
Adults
Children
Notes:
Source:
Lowest
poverty rate
5010
Ferryden
Park
%
29.8
27.2
36.9
5725
Roxby
Downs
%
1.4
0.9
2.6
South
Australian
average
%
12.1
11.2
14.6
Australian
average
%
10.3
9.3
12.9
studies.
285
Table 6: West Australian Postcodes with Highest and Lowest Poverty
Rates (a)
Postcode number
Postcode name
Poverty rates
People
Adults
Children
Note:
Source:
Highest
poverty rate
6000
Perth City
%
19.0
18.5
23.3
Lowest poverty West Australian
rate
average
6015
City Beach
%
%
2.8
10.3
2.5
9.4
3.5
12.7
Australian
average
%
10.3
9.3
12.9
studies.
Table 7: Tasmanian Postcodes with Highest and Lowest Poverty Rates (a)
Highest poverty Lowest poverty
rate
rate
Postcode number
7215
7053
Postcode name
St. Mary’s
Taroona
Poverty rates
%
%
People
20.5
6.4
Adults
18.5
6.2
Children
26.4
7.0
Note:
Source:
Tasmanian
average
Australian
average
%
13.0
12.3
14.9
%
10.3
9.3
12.9
studies.
Children face a higher risk of being in poverty than adults and the national figures
show that child poverty is some three percentage points higher than adult poverty. The
state averages show similar patterns. However, in the postcodes with the highest
poverty rates there is generally a much greater difference in the rates of poverty for
children and adults. For example, in Lightning Ridge, the child poverty rate of almost
40 per cent compares with an adult poverty rate of 22 per cent. In three out of the top
six poverty postcodes examined in Tables 2 to 7, almost two out of every five children
were in poverty, compared with only one in every eight children nationally.
Characteristics of those in Poverty in the Poorest Postcodes
What types of households are in poverty in the poorest postcodes in each state? Table
8 shows the characteristics of households in poverty compared with the national
averages for households in poverty. It is clear that the composition of poor households
can be very different in varying localities, even when the total poverty rates within
postcodes are fairly similar.
Postcode 2834 - Lightning Ridge: Poor households in Lightning Ridge are more
likely than poor Australian households generally to have government cash benefits as
286
their principal source of income. While 57 per cent of poor households in Australia
rely on government benefits as their main income source, almost seven in every 10
households in Lightning Ridge do. Conversely, while 16 per cent of poor Australian
households have wages and salaries as their main income source, less than one-tenth
of households in Lightning Ridge do. This parallels the fact that the proportion of
people in Lightning Ridge living in a household where the head is unemployed is
significantly greater than the national average, while the proportion where the head is
a full-time employee is considerably lower. Poor households in Lightning Ridge are
more likely than Australian poor households generally to have a head born in
Australia (80 per cent compared with 65 per cent) and almost none have a head born
in Asia. Poor households in Lightning Ridge are less likely than the Australian
average to live in public housing and more likely to be single person households.
Postcode 3053 - Carlton South: The majority of poor residents in the postcode of
Carlton South live in a household headed by a man. Seventy per cent live in a
household where the head is not born in Australia, which compares with 35 per cent
of all poor Australians. The head of the household is more likely to have never
married and be under 25 than for poor households nationally. Figure 1 shows that
almost one-quarter of the poor in Carlton South live in a household where the head is
unemployed and almost six in 10 live a household where the head is not in the labour
force. Only 18 per cent have a head who is working. These figure are quite different
to the national average. About half of all poor Australians are not in the labour force
and another 14 per cent are unemployed.
Table 8: Characteristics of People in Poverty in Poorest Postcodes and Australia
Postcode
Age
of
the
reference person
Less than 25 years
25-34 years
35-44 years
45-54 years
55-64 years
65 years or more
2834
3053 4671 5010 6000 7215 Australia
Percentage of total in poverty
household
Sex of the reference person
Male
Female
Labour force status of the
reference person
Employee – full-time
Employee – part-time
Self employed
Unemployed
Not in the labour force
2.1
35.2
22.6
19.9
13.7
6.6
100
10.3
24.7
32.5
22.1
3.4
7.0
100
0.8
21.6
35.8
17.9
15.4
8.6
100
9.5
31.6
22.9
19.8
3.9
12.3
100
23.2
28.4
26.0
14.0
4.6
3.9
100
4.8
22.9
32.5
18.6
13.8
7.3
100
4.6
25.9
30.7
17.1
10.7
11.1
100
43.3
56.7
100
52.6
47.4
100
43.0
57.0
100
42.1
57.9
100
51.4 36.9
48.6 63.1
100 100
37.6
62.4
100
1.9
17.0
14.3
24.5
42.3
100
3.9
6.3
7.7
24.3
57.7
100
4.7
16.2
18.9
16.3
43.8
100
3.8
8.5
3.5
28.7
55.6
100
1.2
16.9
4.7
26.6
50.7
100
5.6
13.8
15.0
13.8
51.7
100
287
3.8
15.0
15.6
12.2
53.4
100
Table 8: Continued
Postcode
Principle Source of Income
for the household
Wage and salary
Self-employed
Other
Government cash benefits
Zero or negative income
Occupation of the reference
person
NA (ie not working in
occupation)
Managers and professionals
Tradespersons
Clerical, sales and service
Labourers, production and
transport workers
Tenure type
Owner
Purchaser
Public housing
Private renter
Other, rent-free
Marital status of the
reference person
Never married
Separated/divorced/widowed
Married
Household type
Single person
Couple only
Couple with children
Sole parent
Multiple families
2834
3053 4671 5010 6000 7215 Australia
9.9
7.0
11.2
68.3
3.5
100
7.6
3.5
28.6
60.2
0.1
100
17.1
9.2
8.7
59.6
5.4
100
10.6
0.3
11.8
74.3
3.0
100
13.4
5.1
19.1
56.8
5.5
100
13.1
10.0
8.7
63.6
4.7
100
16.3
7.6
12.7
57.3
6.0
100
66.9
9.0
4.7
10.8
82.0
12.6
0.0
0.9
60.2
17.6
3.1
9.8
84.3
2.9
1.6
4.8
77.3
8.7
0.1
6.7
65.6
13.9
3.9
9.8
65.6
13.7
3.1
9.8
8.7
100
4.5
100
9.4
100
6.4
100
7.2
100
6.8
100
7.8
100
43.8
15.6
2.1
28.8
9.7
100
13.9
10.1
36.9
38.4
0.7
100
42.4
26.4
22.7
8.4
0.0
100
10.9
7.7
78.4
2.3
0.7
100
18.6
7.3
14.6
55.0
4.5
100
43.2
23.8
3.7
21.9
7.4
100
34.9
22.8
10.3
27.1
5.0
100
13.1
26.6
60.3
100
30.9
19.0
50.1
100
6.1
16.7
77.2
100
30.1
20.0
49.9
100
46.9
14.0
39.0
100
8.8
17.1
74.1
100
13.0
23.1
63.9
100
23.7
14.1
42.7
14.5
5.0
100
21.6
7.8
39.2
20.3
11.1
100
12.3
20.4
49.6
10.4
7.3
100
20.6
13.2
32.8
28.3
5.1
100
47.1
9.3
24.2
9.5
10.0
100
16.1
16.2
53.8
9.9
4.0
100
15.4
15.4
40.7
20.0
8.5
100
288
Table 8: Continued.
Postcode
2834 3053 4671 5010 6000 7215 Australia
Number of dependants in the
household
None
One
Two
Three
Four
Five or more
Country of birth of the
reference person
Australia
Other
Europe/former USSR
Asia
46.0
6.6
15.8
20.0
11.5
0.1
100
39.0
20.2
39.7
0.0
1.0
0.0
100
39.8
13.0
15.5
21.2
8.1
2.4
100
37.3
18.7
31.7
1.0
11.3
0.0
100
81.7
4.4
13.0
0.8
100
29.4
12.8
24.6
33.2
100
76.2
5.2
13.7
5.0
100
41.7
28.4
18.0
11.9
100
68.2
18.0
13.5
0.3
0.0
0.0
100
37.9
12.7
20.4
21.5
7.2
0.2
100
51.4 86.6
10.3 2.7
19.6 10.1
18.6 0.6
100 100
38.3
13.4
22.2
18.6
6.4
1.0
100
65.4
8.2
14.6
11.7
100
Note: a) Using the Henderson half average poverty line
Source: 1998-99 HES and Marketinfo/2001 weights.
Figure 1: Labour Force Status Of Household Reference Person In Poor
Households In Carlton South (and percentage point difference from national
average)
Employee -FT
3.9%
(1.7% < Aust av)
Not in Labour Force
57.7%
(6% > Aust av)
Employee -PT
6.3%
(7.5% < Aust av)
Self-employed
7.7%
(7.3%< Aust av)
Unemployed
24.3%
(10.4% > Aust av)
Data source: 1998-99 HES and Marketinfo/2001 weights
Over three out of 10 poor Australians have some sort of employment. Thus the poor
of Carlton South are much more likely to live in a household where the head is not
working. In Carlton South, 37 per cent of poor households live in public housing –
about 3.5 times the Australian average. Overall, therefore, poverty in Carlton South
289
seems to be due to large numbers of students, unemployed, migrants and a
concentration of public housing.
Postcode 4671 – Gin Gin: The picture of poverty in Gin Gin is quite different from
that of Carlton South. Less than one per cent of poor residents live in a household
headed by a person under 25. The head is more likely to be middle-aged – seven in 10
poor households are headed by a person aged 35-65 compared with 58 per cent of all
poor Australian households – and are much more likely to be married. Compared with
the Australian average, there are more poor households that are couples, either with or
without children. A significant proportion of the poor in this postcode live in a
household where the head is employed: 4.7 per cent are employed full time, 16.2 per
cent part time and 18.3 per cent are self-employed. Correspondingly, over one-quarter
of households rely on income from wages and salaries or self-employment as their
principal source of income. Compared with the national average, a greater proportion
of the poor in Gin Gin live in public housing or are home owners/purchasers, and they
are more likely to be Australian born.
Postcode 5010 – Ferryden Park: The poor of Ferryden Park tend to live in
households where the head has a very high chance of being unemployed (28.7 per
cent compared with the national average of 13.8 per cent), never married (30.1 per
cent compared with 13.0 per cent generally) and not born in Australia (58.3 per cent
compared with 34.6 per cent) (Figure 2). Most poor households in this postcode live
in public housing (a striking 78.4 per cent compared with the national average of 10.3
per cent) and almost three-quarters have government cash benefits as their principal
income source.
Figure 2: Selected Household Characteristics of Poor Residents of Ferryden
Park and all Poor Australians
Head unemployed
Never married
Ferryden
Park
Not born in Australia
In public housing
Australia
Main income source GCB
0
10
20
30
40
50
60
70
80
90
Percentage of poor
Data source: 1998-99 HES and Marketinfo/2001 weights
Postcode 6000 – Perth City: The picture of poverty in the Perth City postcode is one
of young, single people without dependants. Over 23 per cent of poor people in this
postcode live in a household headed by a person under 25 years of age. This compares
290
with 4.6 per cent of poor Australians. Almost half have never married (compared with
13 per cent nationally) and a similar proportion lives in single person households
(compared with 15 per cent of Australia’s poor). Over 68 per cent live in a household
without dependent children. The majority of the poor in Perth City live in households
headed by a male (51 per cent, compared with 38 per cent for Australia). Over onequarter is unemployed and 55 per cent are private renters.
Postcode 7215 – St Mary’s: The poor in St Marys in Tasmania have a profile more
like the Australian average than any of the other postcodes profiled here. Poor
households in St Marys are more likely than the poor Australian households generally
to have government cash benefits as their main income source and less likely to rely
mainly on wages and salaries. An Australian-born person heads almost nine in 10
poor households in St Marys and only 3.7 per cent of poor households there live in
public housing, compared with one-tenth nationally. Poor households in this postcode
are more likely to have a head who is married and more likely to be a couple with
children.
Characteristics of the Poor and Rich Postcodes
What causes a postcode to have high or low poverty rates? Table 9 looks at some of
the key characteristics of those living in each of the postcodes with high poverty rates
while Table 10 looks at the characteristics of those living in postcodes with low
poverty rates.
Table 9: Characteristics of the High Poverty Postcodes
2834 3053 4671 5010 6000 7215 Australia
Average total income
(annual, $)
Average disposable income
(annual, $)
Age of the household
reference person
Less than 25 years
25-34 years
35-44 years
45-54 years
55-64 years
65 years or more
Male
Female
Labour force status of
reference person
Self-employed
Unemployed
25164 40011 28955 22480 36995 27568
45574
21999 31740 25471 20600 29188 23993
36581
3.5
22.1
22.8
19.8
18.0
13.9
21.7
28.1
21.5
14.5
7.3
6.9
1.9
14.1
32.1
23.1
15.3
13.6
7.5
22.4
26.6
16.3
11.3
15.9
16.2
27.9
21.4
17.9
7.3
9.3
3.8
18.3
29.8
19.6
14.2
14.2
4.2
20.2
29.1
22.9
11.4
12.2
55.9
44.1
55.7
44.3
58.2
41.8
53.1
46.9
64.0
36.0
56.3
43.7
63.0
37.0
22.3
14.8
8.8
9.3
44.8
41.0
12.7
2.8
7.0
36.6
291
29.5
12.6
12.0
5.0
40.8
26.4
15.9
1.7
9.0
47.0
44.8
11.0
3.5
6.9
33.8
28.0
14.6
10.7
3.5
43.1
49.6
10.6
7.8
2.4
29.6
Table 9: Continued
2834
Principal source of income
for the household
Wage and salary
Self-employed
Other
person
NA (i.e. not working in
occupation)
Tradespersons
transport workers
Tenure type
Owner
Purchaser
Public housing
Private renter
Other, rent-free
Marital status of the reference
person
Never married
Married
Household type
Single person
Couple only
Sole parent
Multiple families
Number of dependants in the
household
None
One
Two
Three
Four
Five or more
3053 4671 5010 6000
34.4
6.9
8.5
49.3
0.9
50.4
2.0
12.6
35.0
0.0
Australia
54.7
3.9
15.9
24.5
1.1
38.9
10.2
8.3
41.7
1.0
61.4
6.8
6.6
24.6
0.6
54.1 43.6 45.9 56.0 40.7
12.3 36.4 18.3 7.0 35.3
6.3 2.2 6.9 7.3 3.2
12.1 10.6 9.9 6.1 11.2
46.6
20.7
10.3
9.8
32.0
32.3
10.1
13.1
15.2
9.6
12.6
12.6
14.9 45.8 11.2 19.6
10.6 32.1 12.9 16.7
21.0 17.5 68.0 4.8
52.1 4.6 7.6 56.5
1.3 0.0 0.4 2.3
45.0
29.8
2.8
18.7
3.6
36.9
33.8
5.0
22.2
2.2
14.2 36.0 5.0 21.0 38.1
24.1 17.4 14.8 22.5 16.1
61.7 46.6 80.2 56.5 45.8
7.8
15.6
76.6
10.4
16.4
73.2
20.6
23.5
32.6
12.1
11.2
17.0 7.5 14.2 33.1
14.6 26.7 17.1 22.9
26.1 45.5 34.1 20.5
14.0 9.1 24.5 7.4
28.3 11.2 10.2 16.1
12.2
27.6
45.4
8.9
5.9
9.6
19.7
45.4
11.5
13.7
60.3 59.5 46.9 46.1 73.0
10.7 17.8 11.6 19.2 19.2
13.6 21.5 19.1 23.3 7.3
11.4 0.3 13.4 3.0 0.4
3.8 0.8 6.0 7.1 0.0
0.2 0.2 3.0 1.3 0.0
49.1
13.0
20.8
12.8
3.7
0.6
45.3
16.5
21.7
11.9
3.5
1.1
57.8
10.9
1.9
22.9
6.4
42.1 37.2
10.7 0.6
7.4 4.5
38.7 56.9
1.2 0.9
7215
7.2 19.0 23.6
292
Table 9: Continued
2834 3053 4671 5010 6000 7215 Australia
reference person
Australia
Other
Europe/former USSR
Asia
75.7
4.9
19.0
0.3
47.6 78.0 50.0 55.7 87.4
12.3 5.6 13.9 12.2 3.1
17.3 14.2 16.9 16.4 9.2
22.8 2.2 19.2 15.7 0.4
69.2
7.3
16.1
7.3
Table 10:
Characteristics of the Low Poverty Postcodes
2088 3186 4069 5725 6015 7053 Australia
Average total income
(annual, $)
Average disposable income
(annual, $)
Age of the household
reference person
Less than 25 years
25-34 years
35-44 years
45-54 years
55-64 years
65 years or more
Male
Female
Labour force status of the
reference person
Self-employed
Unemployed
Principal source of income
for the household
Wage and salary
Self-employed
Other
97 677 87 880 79 073 87 412 88 544 51 203
45 574
67 953 62 773 58 338 63 243 63 901 40 572
36 581
4.5
21.5
24.9
25.5
12.0
11.6
2.4
10.8
28.1
30.9
12.6
15.3
3.0
9.7
27.9
33.6
15.4
10.4
8.4
37.2
34.5
13.9
4.0
1.9
2.3
7.4
25.3
32.6
17.3
15.2
2.1
8.3
29.3
32.5
13.0
14.9
4.2
20.2
29.1
22.9
11.4
12.2
68.7
31.3
73.3
26.7
75.0
25.0
90.1
9.9
80.2
19.8
64.2
35.8
63.0
37.0
69.0
7.0
6.3
0.4
17.2
65.4
5.0
6.3
0.9
22.5
66.8
9.3
5.9
0.8
17.1
85.8
3.2
6.4
0.3
4.3
63.4
6.7
8.2
0.3
21.4
54.5
10.3
6.0
1.6
27.6
49.6
10.6
7.8
2.4
29.6
73.3
7.6
10.0
9.1
0.0
71.5
7.3
12.3
8.9
0.0
75.9
5.8
9.2
9.1
0.1
88.3
5.6
2.0
4.0
0.0
70.5
9.1
14.7
5.7
0.1
64.3
5.4
10.0
19.8
0.5
61.4
6.8
6.6
24.6
0.6
293
Table 10 : Continued
2088 3186 4069 5725 6015 7053 Australia
person
NA (i.e. not working in
occupation)
Tradespersons
transport workers
Tenure type
Owner
Purchaser
Public housing
Private renter
Other, rent-free
Marital status of the
reference person
Never married
Married
Household type
Single person
Couple only
Sole parent
Multiple families
in the household
None
One
Two
Three
Four
Five or more
reference person
Australia
Other
Europe/former USSR
Asia
17.6
67.7
3.7
7.6
23.4
67.1
2.0
6.3
17.9
64.5
5.1
9.3
4.6
33.1
21.5
6.1
21.7
66.1
2.2
6.8
29.2
54.5
4.5
9.8
32.0
32.3
10.1
13.1
3.3
1.2
3.2
34.7
3.2
2.1
12.6
40.7
21.0
0.9
35.9
1.5
51.7
25.1
0.2
20.9
2.0
46.6
38.1
0.1
13.8
1.4
23.0
36.7
0.0
38.8
1.5
58.9
28.7
10.1
2.3
0.0
45.1
39.9
2.2
11.7
1.2
36.9
33.8
5.0
22.2
2.2
17.9
14.5
67.6
10.1
15.3
74.7
6.7
11.0
82.3
5.0
7.6
87.5
6.8
9.3
83.9
7.8
17.2
75.0
10.4
16.4
73.2
16.6
25.0
35.0
7.5
15.9
11.4
22.2
47.1
7.1
12.3
4.3
18.5
56.5
9.0
11.8
5.0
17.9
64.3
4.2
8.5
6.3
24.1
54.8
6.1
8.7
8.7
20.9
48.6
10.8
11.0
9.6
19.7
45.4
11.5
13.7
54.7
15.1
20.6
8.3
1.2
0.0
48.3
17.3
21.4
12.6
0.4
0.0
38.4
18.1
25.1
13.7
3.7
0.9
34.7
12.0
34.0
18.5
0.8
0.0
42.4
15.7
22.2
18.9
0.8
0.0
44.0
15.9
24.5
11.6
3.0
1.0
45.3
16.5
21.7
11.9
3.5
1.1
57.5
12.1
14.6
15.9
73.5
7.8
13.6
5.0
66.1
6.8
17.8
9.3
82.2
7.9
9.7
0.2
67.2
9.4
18.0
5.4
71.7
5.5
19.0
3.8
69.2
7.3
16.1
7.3
Income: As we are using an income-based measure of poverty, it is not surprising to
find that postcodes with high poverty rates have relatively low average household
294
incomes and postcodes with low poverty rates generally have high average household
incomes. For example, the estimated average 1998-99 household disposable income
in Ferryden Park is just $20 600 per annum, compared with an average household
disposable income of $68 000 in Spit Junction.
However, it is worth noting that some of the postcodes that have the highest and
lowest poverty rates do not have particularly low or high average incomes. Both Perth
City and Carlton South have average incomes much closer to the Australian average
than other poor postcodes, while the average disposable income in the low poverty
suburb of Taroona is just over $40 000, less than $4000 greater than the Australian
average. Because poverty lines are based on equivalent income, household
composition is also an important factor in determining poverty rates. In addition, the
degree of income inequality within a postcode is also important in determining
poverty rates. For example, two postcodes may have the same average income, but
one might have all households with income close to the average, while the other
might contain some households with very high incomes and some households with
very low incomes. The latter would have a higher poverty rate. This seems to be one
of the factors underlying the high poverty rates in Carlton South, where professionals
on relatively high incomes co-exist with poor young students and unemployed.
Age: Postcodes with high poverty rates tend to have one of two age profiles.
Lightning Ridge, Gin Gin, Ferryden Park and Taroona have older age profiles than
the Australian average, with a greater proportion of households with a head aged over
55. Perth City and Carlton South have a much younger age profile with a large
proportion of households headed by a person aged less than 35. This suggests a large
student population.
Moving to postcodes with low poverty rates, while Roxby Downs has a young age
profile the other postcodes with low poverty rates have a greater than average
proportion of households with a head aged 45-64, and in Brighton, City Beach and
Taroona, a greater proportion of households with a head aged 65 or over.
Sex of Household Reference Person: The high poverty postcodes are typified by a
greater than average proportion of households headed by a female. Postcodes with
low poverty rates are all above the Australian average of 63 per cent of households
headed by a male. In Roxby Downs, 90 per cent of households have a male head.
Labour Force Status of the Household Reference Person: Postcodes with high
poverty rates all have a smaller than average proportion of households headed by a
full-time employee and more headed by a part-time employee, self-employed person,
unemployed person or someone not in the labour force. There is some variation
between the different states, however. Ferryden Park has a strikingly high percentage
(56 per cent) of people in households where the head is not working. Lightning Ridge,
Gin Gin and St. Mary’s have a low proportion of households with the head working
full time, but with a higher than average proportion of part-time employees and selfemployed people. Lightning Ridge has a large proportion with an unemployed head.
Postcodes with low poverty rates all have a significantly greater proportion of
households headed by a full-time employee and a lower proportion headed by an
unemployed person or someone not in the labour force.
295
Principal Source of Income of the Household: Principal source of income is closely
linked to the labour force status of the household head. High-poverty postcodes have a
large proportion of households relying on government cash benefits. In Ferryden
Park, 57 per cent of households have government benefits as their principal income
source. Conversely, the proportion of households relying on government cash benefits
in the low-poverty postcodes is significantly less than average and most households
have wages and salaries as their main income source.
Tenure Type: Three of the high-poverty postcodes, Ferryden Park, Carlton South and
Gin Gin, have high levels of public housing; in Ferryden Park 68 per cent of
households are government renters. Perth City has a very high proportion of private
renters, as does Carlton South. However, Lightning Ridge, Gin Gin and St. Mary’s
have about average numbers of households that are owners/purchasers. Similarly
there is no clear trend among the low-poverty postcodes. There is a tendency for there
to be an above average proportion of owners/purchasers, but in Spit Junction and
Roxby Downs there are a large proportion of private renters and City Beach has an
above average proportion of public housing tenants.
Marital Status of the Household Reference Person and Household Type: Again
there is no clear pattern among high-poverty postcodes. In Perth City, Ferryden Park
and Carlton South, a greater than average proportion have a head who has never been
married and a greater than average number of single person and multiple family
households. Lightning Ridge and Ferryden Park have more heads who are separated,
divorced or widowed and Ferryden Park has almost one-quarter sole-parent
households. However, a married person heads eight out of 10 households in Gin Gin.
Among low-poverty postcodes, generally a greater than average proportion are
headed by a married person and the proportion of households headed by someone
who has never been married or is separated, divorced or widowed is less than the
Australian average. In all of the low-poverty postcodes, the proportion of sole-parent
households is less than the Australian average.
4
Conclusions
This report examined postcodes with the highest and lowest poverty rates, after
removing postcodes with less than 1000 households within them so as to reduce the
impact of outliers and small suburbs with highly specialised circumstances. The
poorest postcodes generally had poverty rates that were two to three times the
Australian average. In contrast, the postcodes with the lowest poverty rates generally
had poverty rates that were about one-tenth to one-fifth of the Australian average.
Rates of poverty among children at both the national and state level tended to be
about three to four percentage points higher than among adults. However, in the
poorest postcodes, the difference between adult and child poverty rates was often
much more pronounced. In some of the poorest postcodes, almost four in ten children
were estimated to live in poverty, 16 percentage points greater than for adults.
Many of the postcode characteristics associated with high levels of poverty were, not
surprisingly, the same factors traditionally identified in national studies as being
296
related to poverty. Factors likely to be associated with a high poverty rate within
postcodes included an above average proportion of:
•
household heads who were unemployed or not-in-the-labour force;
•
households headed by young people;
•
renters, particularly public renters; and
•
households with government cash benefits as the main income source.
However, one of the important findings of the study was that there was considerable
variation in the characteristics of postcodes with very high poverty rates.
Consequently, it seems that the factors causing poverty vary greatly throughout
Australia and it is important for policy makers to understand the characteristics of a
region in developing an appropriate response to combat poverty.
Factors likely to be associated with a low poverty rate within postcodes included an
above average proportion of full-time workers and a below average proportion of
sole-parent households.
References
Australian Bureau of Statistics (ABS) (1998), Australian Social Trends 1998, Cat.
No. 4102.0, ABS, Canberra.
Harding, A., R. Lloyd, O. Hellwig and G. Bailey (2000), Building the Profile: Report
of the Population Research Phase of the ACT Poverty Project, ACT Poverty Task
Group Paper No. 3, Canberra.
Harding, A. and A. Szukalska (1999), Trends in Child Poverty in Australia: 1982 to
1995-96, Discussion Paper No. 42, National Centre for Social and Economic
Modelling, University of Canberra, Canberra.
Harding, A. and A. Szukalska (2000), The Unlucky Australians? Financial
Disadvantage in Australia 1999, The Smith Family, Sydney.
Layte, R., B. Nolan and C. Whelan (2000), Applying the Irish National Definition of
Poverty Across 12 European Union Countries: The Structure and Determinants of
Low Income and Deprivation, Paper presented to 26th General Conference of the
International Association for Research in Income and Wealth (IARIW), 27
August – 2 September 2000, Cracow, Poland.
Saunders, P. (1996), ‘Poverty and deprivation in Australia’ in Australian Bureau of
Statistics, Year Book Australia 1996, Cat. No. 1301.0, ABS, Canberra.
Strengmann-Kuhn, W. (2000), Theoretical Definition and Empirical Measurement of
Welfare and Poverty: A Microeconomic Approach, Paper presented to 26th
General Conference of the International Association for Research in Income and
Wealth (IARIW), 27 August – 2 September 2000, Cracow, Poland.
297
Fashion, Fiction, Fertile Inquiry?
Struggling With the Postmodern Challenge
and Social Policy Analysis
Greg Marston
School of Social Work and Social Policy, University of Queensland
1
Introduction
Until relatively recently, a profound silence could be heard in core social policy as to
the postmodern, conveying a notion of isolation from key trends in social theory
(Carter, 1998: 102). While some policy academics and practitioners have seized on
the possibilities offered by poststructural understandings of power, identity,
knowledge and discourse, many question its exploratory and explanatory potential
(Yeatman, 1994). It is perhaps not surprising that ‘orthodox’ social policy research
remains sceptical and even hostile towards the intellectual perspectives offered by
postmodernism. Social policy research, that which is understood as the academic
study of welfare, has an unavoidably structural element; its whole history as a subject
(its causal logics, its research tools; its statistical headings) has been concerned with
patterned disadvantage (Carter, 1998: 110; Deacon and Mann, 1999). To relinquish
that is to apparently relinquish its core goals and purposes. As Taylor-Gooby (1994:
403) argues, 'The implications for social policy are that an interest in postmodernism
may cloak developments of considerable importance'.
While some academics in the social policy community may choose to stand firm and
argue that social policy does not bend to the dictates of fashion, the risk is that new
lines of inquiry, issues and questions will be overlooked, and opportunities for a richer
and more sophisticated account of policy practices may be missed. Advocates of
postmodern approaches to social policy issues claim, among other things, that this
diverse body of theory draws attention to local struggles and transformations, to the
dialectical relationship between language and material practices and to the centrality
of identity, difference and culture in social change – practices and processes that were
previously obscured (Fairclough, 1995; Penna and O’Brien, 1996; Mann, 1998). Of
course, these claims should not be taken at face value.
This paper addresses the question of the usefulness of postmodernism for social
policy, focusing attention on social policy analysis1 and starting from the assumption
1
Typically the social policy tradition is distinguished from the broader category of public policy analysis.
The dominant perspective in public policy analysis is to look at political decision making, exploring such
phenomena as the operations of power and the influence of interest groups, while social policy centres
on the normative bases and distributional consequences of particular policies (Petersen et al., 1999: 7). In
this section I will be drawing on and evaluating both forms of overlapping policy literature, as both are
relevant to the concerns of social policy given that the processes of policy making are what constitutes
policy outcomes.
Marston, G. (2002), ‘Fashion, fiction, fertile inquiry? Struggling with the postmodern challenge and
social policy analysis’, in T. Eardley and B. Bradbury, eds, Competing Visions: Refereed Proceedings
of the National Social Policy Conference 2001, SPRC Report 1/02, Social Policy Research Centre,
University of New South Wales, Sydney. 298-316.
STRUGGLING WITH THE POSTMODERN CHALLENGE
that poststructuralism is no different to other new schools that tend to overemphasise
the distinctiveness of their own approach and to dwell at some length on the
deficiencies of what went before (Petersen, et al., 1999: 7). In the first part of this
paper, I briefly summarise the academic social policy debate about the value of
postmodernism for social policy analysis, and in so doing introduce the concepts that
will be further explored later. In the second part, I outline some of the key theoretical
tensions that distinguish positivism, critical theory and postmodernism. In the third
and final part of the paper, I seek to demonstrate the value of postmodernist
approaches to social policy analysis, to show how a ‘critical postmodernism’
(Yeatman, 1994; Carter, 1998) offers productive ways of thinking about
contemporary social policy issues. In illustrating these possibilities, I focus on critical
discourse analysis as one site where a critical postmodernism can be made operational
in studies of the policy process. In summary, this paper argues for a constructive
dialogue between postmodernism and social policy research, where the task is one of
augmentation and adding – not replacing existing research paradigms.
The Debate So Far
In the last three decades of the twentieth century, all the disciplines
in the social sciences have experienced a fundamental reappraisal
of their basic assumptions, theories and methods. The most
significant common feature of this reappraisal is the recognition
that ‘culture’ deserves much more serious attention as an object of
study in its own right and this has produced a reassessment of the
linguistic, discursive and cultural conditions of social research
(Smith, 1998: 231).
The above quote captures trends in epistemology, postmodern social theory and
analytical frameworks that are reassessing the importance of language and culture.
Missing from the above excerpt, however, are the differentiated effects that the socalled ‘cultural turn’ has had on social science disciplines that remain resistant to
alternative epistemologies. In the area of social policy analysis for instance, there are
policy academics that remain sceptical about the postmodernist critique of
conventional social science (Taylor-Gooby, 1994; O’Neill, 1995), while others are
more optimistic about the possibilities offered by these intellectual perspectives
(Bacchi, 1999; Clark, 1998; Carter, 1998; Leonard, 1997).
There are many reasons that may explain an ambivalent attitude towards
postmodernism. The subject of social policy is under a massive weight of history and
orthodox visions of what is proper are deeply embedded in educational institutions –
at present it is somewhat like asking a currently unadapted car to run on unleaded fuel
(Carter, 1998: 110). On another level, the debate stems from a question about whether
the study of culture, discourse and meaning should replace the traditional concerns of
social policy, or whether this fluid body of theory is able to sit alongside established
research paradigms.
Claim and Counter-Claim
Over the past five to ten years, social policy research has started to engage with the
question about the usefulness of postmodernism for policy research. In the mid 1990s,
the Journal of Social Policy published a debate about whether postmodernism
represented a ‘great step backwards’, or ‘a small step forwards’. Here the debate was
299
GREG MARSTON
polarised. On one hand, there was Taylor-Gooby (1994) who viewed postmodernism
as depoliticising the agenda of social policy. On the other hand, there was Penna and
O’Brien (1996) who believed the different strands of poststructuralism and
postmodernism offered alternative theoretical frameworks and a more inclusive
political agenda to address contemporary welfare state restructuring.
Taylor-Gooby (1994: 385) argued that postmodernism contributes to depoliticising
social inequality and disadvantage, by discrediting structural analysis and
emphasising plurality and relativism:
Postmodernism ignores the significance of market liberalism and
the associated trends to inequality, retrenchment and the regulation
of the poorest groups. From this perspective, postmodernism
functions as an ideological smoke screen, preventing us from
recognising some of the most important trends in modern social
policy.
In this argument, the value of postmodernism is being evaluated against a modernist
understanding of ‘truth’ and ‘reality’, articulated within a meta-narrative about market
liberalism. In response, Fitzpatrick (1996: 306) states that Taylor-Gooby largely bases
his reservations about postmodernism on assertions, rather than argument, and as a
result he does not ask whether postmodernist theory could be used to develop a
critique of economic liberalism. In other words, setting up unitary conceptions of
modernism and postmodernism as ‘binary opposites’ does nothing to clarify that
which is useful in recent theoretical debates (Taylor, 1998: 31; Thompson and
Hoggett, 1996).
Taking a more positive interpretation in their article in the Journal of Social Policy,
Penna and O’Brien (1996: 60) believe that social policy analysis can and should
benefit from postmodernist theorising:
At the level of political culture and at the level of political
economy the capacity of normative policy analysis to deliver
effective solutions to the multiple lines of exclusion and
marginalisation characteristic of contemporary society is at best
uncertain. Postindustrial, postfordist, poststructuralist and
postmodernist writings, separately and together, provide the
discipline of social policy with alternative theoretical perspectives
and a political agenda with which to address the complexity of
current restructuring.
However, for these ideas to influence social policy, they need to appear as operational
and utilisable constructs for the study of welfare (Carter, 1998: 104). This empirical
challenge has started to be taken up in recent years.
A Practical Engagement
In the late 1990s, through a series of edited collections and journal articles, the debate
about the utility of postmodern thought moved toward a practical engagement with
contemporary social policy issues (see Carter 1998; Mann, 1998; Ferge, 1997; Taylor,
1998; Leonard, 1997, Fraser, 1997; Culpitt, 1999; and Bosworth, 1999). In the
Australian context, among others, Dean (1997, 1999) has developed Foucault’s work
300
on governmentality into an analytical technique to interrogate the construction of the
unemployed citizen and labour market programs in general. Using a theory of
discourse, Bacchi (1999) and earlier, Watts (1993), have employed theories of
discourse to understand policy as discursive practice. Yeatman (1994) assesses the
implications of postmodern feminist theorising for questions about the state and
community. Petersen (1999, et al.) have employed poststructural accounts of power
and knowledge to discuss questions of citizenship, technology and public welfare in
Australia. These writers have all engaged with issues that are of current importance to
the social policy field.
The critical interests of social policy – an analysis of power relations, the state and an
overriding concern with the distribution of resources – have not been lost in these
accounts. However, the way in which these constructs are conceptualised is very
different to how they have been discussed in a modernist discourse. In this respect,
postmodernist thinking stands in a critical, but not necessarily an oppositional,
relation to modernist ideas. Postmodernism acts as an analytical aid by opening up
different ways of thinking about rights and justice. ’Postmodern emancipatory vision
does not offer a utopian future, but works to develop contestatory political and public
spaces, which open up in relation to existing forms of governance‘ (Yeatman, 1994:
ix).
This brief summary of the debate about the value of postmodernism for social policy
analysis suggests a particular account of what postmodernism signifies. In this paper,
for example, I am not using the terms modernism and postmodernism to describe two
distinct historical eras, or to assert that we have moved into a postmodern age.
The ‘post’ prefix is used here to denote a different intellectual and philosophical
orientation to the project of modernity. ’The term ”post” suggests that it is possible
now to bound the project of modernity, to discern its features, to understand the kind
of paradigm it comprises‘ (Yeatman, 1994: 8). In exploring the discourse of
modernity and its influence on social policy research, I want to consider briefly some
of the distinguishing characteristics of traditional and postmodern approaches to
policy analysis, which have been touched on in the above discussion. In the following
section, I have been careful to avoid simply rehashing the somewhat tired
philosophical binary between positivism and post-positivism (Mann, 1998). Yet, at
the same time, and as a starting point for a dialogue, it is important to have a sense of
what defines these different paradigms before discussing how a ‘critical
postmodernism’ (Yeatman, 1994; Carter, 1998) might be envisaged and practised in
studies of the policy process, and social policy research more generally.
2
Evaluating the Differences Between Positivist, Critical and
Postmodern Social Policy Analyses
This section explores some of the competing epistemologies that have informed social
policy analysis, including positivism, critical theory and postmodernism. A detailed
evaluation of these paradigms is beyond the scope of this paper. In the following
discussion I have focused on a brief appraisal of the major features of each position.
What will be identified and defined in this discussion are the conceptual tools that can
be used to constitute a form of critical social policy research, which is informed by
poststructural theorising.
301
GREG MARSTON
Positivism and Social Policy Analysis
In short, positivism is based on a notion that the laws of understanding and studying
the ‘natural’ world also apply to the social world. ’Positivism assumes an
unproblematic relationship between conventional knowledge of the nature of the
world and its actual nature‘ (Hastings, 1998: 193). Despite recent challenges to
positivist philosophy, a rational, value-free science continues to have significant
hegemonic power in normative analyses of the policy process and welfare regimes. In
the post-war period in the West, a form of positivism emerged that emphasised the
importance of value-freedom, hard-facts and prediction as the basis for developing
‘objective’ policy proposals for governments, businesses and other private institutions
(Smith, 1998: 77). The following definition succinctly captures the positivist nature of
orthodox social policy research. ’Social policy research is concerned to generate high
quality objective knowledge that can be deployed in social planning‘ (Taylor-Gooby
1994: 387).
This epistemological position has inspired much of policy research’s most used
methods, such as survey, the questionnaire and statistical models (Hughes, 1990: 16).
Despite emerging epistemologies that are calling into question the founding principles
of the modern welfare state, orthodox social policy research continues to be captured
by a positivist tradition that has in turn inspired large numbers of population based
studies aimed at improving social planning (Mann, 1998: 84). In general, orthodox
policy research relies upon a positivist epistemology where the main task of the
researcher is one of discovering objective facts and presenting them in a descriptive
format in the expectation that policy makers will take notice and act accordingly
(Jacobs and Manzi, 2000: 35). Arguably, this paradigm has had mixed success in
influencing policy-makers (see Dalton et al., 1996).
Rational Actors and Empiricism
The positivist approach remains limited, partially due to its conceptual reliance on a
rational-actor understanding of the policy making process and its continuing faith in
the transformative potential of empirical facts (Ham and Hill, 1993: 77; Watts, 1993).
Rationality in policymaking can be defined as the process where, when faced with a
range of policy options, the decision makers choose the alternative most likely to
achieve the stated outcome (Ham and Hill, 1993: 77). Within this paradigm, policy
language is accepted as a neutral medium in which ideas and ‘objective’ worlds can
be quantified, represented and discussed. From the rational perspective, the state is
regarded as an empirical agent of governance and policy making, reliant on empirical
processes of problem discovery, and secondly as an entity whose history, processes,
and institutions can be understood empirically (Watts, 1993). The reliance on
‘objectivity’ has meant that policy analysis has largely ignored policy language,
particularly the implications of both the passing of certain words from policy
discourse and the introduction of new terms that feature prominently in policy debates
(Jacobs and Manzi, 1996: 544).
Moreover, the rational representation of policy making remains problematic for its
insufficient account of the political context, insufficient emphasis on the participants
in the process and the ideal type nature of the model itself (Dalton et al., 1996: 17).
While not seeking to deny the place of positivist policy analysis, this approach has its
limitations, particularly when it comes to understanding issues of power and social
302
relations between actors in policy change. As Yanow (1996: 6) states: ’Positivist
knowledge does not give us information about meanings made by actors in a situation.
When we read a policy we see more than just marks on a page, we hear more than just
sound waves‘. In other words, the positivist paradigm does not invite the policy
researcher to be reflexive about their practices of truth production in policy
development. In order to capture the cultural, political and contested dimensions of
policymaking it is important to consider what the critical theory tradition offers.
Critical Theory and the Argumentative Turn
Critical social theory, with its classical roots in Marxism and Hegelian philosophy, is
principally concerned with liberatory social transformation via political struggle to
overcome oppressive structures (Healy, 2000). This stands in contrast to positivism
where the role of social science is merely to understand and investigate a given social
order. Despite its theoretical advancements, the understanding of culture and ideology
in critical social science has tended to retain the Marxist theory of ‘false
consciousness’, where oppressed social groups internalise dominant ideologies, and
are thus prevented from transforming themselves and society into an ideal utopian
state until this falsity is recognised (Bottomore, 1984). This definition of ideology
rests on the assumption that the oppressed have greater access to the truth, where the
agency of the oppressed indicates the truth of humanity and the path to emancipation
(Dean, 1999: 123). The assumption in this version of critical social science is that
there is a fixed system of true representation and false representation. In this account,
critical theory is conceived of as a mental construct used to guide emancipation, with
little acknowledgement that emancipation is always relative to an established
discursive order (O’Brien and Penna, 1999: 52; Yeatman, 1994: 7).
In the case of critical theory and public policy making, there has been some work
done on transforming this body of theory into forms that are relevant for everyday
policy practice, particularly in the case of second-generation critical theorists such as
Habermas. For example, Forester (1993), in his work on the relevance of critical
theory for public policy, argues that in using Habermas’s theory of communicative
action we can conduct policy research that focuses on the situated, performative
qualities of the policy analysts’ work – their conversations and their texts – and realise
how power, class, culture, ethnicity and control manifest themselves in their speech,
writing and gesture. Similarly, Jackson (1999) uses Habermas’s work on ‘public
spheres’ to discuss the communicative processes involved in a Council tenants’
forum, where he observed the clash of discourses of the local government, social
housing providers and the ‘lifeworld’ discourse of tenants.
These examples of critically informed policy making studies constitute what Fischer
and Forester (1993) refer to as the ‘argumentative turn in policy analysis’. Diverse
theoretical perspectives, ranging from French poststructuralism to the Frankfurt
school of Critical Social Theory, have informed the growing concern with
argumentation in policy practice (Fischer and Forester, 1993). From these rich sources
research questions can be formed about what policy analysts and decision makers do,
how language and modes of representation enable and constrain their work and how
policy rhetoric includes and excludes policy actors.
The argumentative turn in policy analysis acknowledges various understandings of
power, it emphasises the constructive nature of policy language and it draws attention
303
GREG MARSTON
to the complex relationship between agency, identity and institutional structures. This
style of policy analysis challenges the positivist assumption that all human action is
literal, measurable and instrumentally rational, and seeks to reveal how this
assumption overlooks the extent to which policy issues are enunciated as the outcome
of power relations, identity formation, ideological contestation and political conflict
(Yanow, 1996; Jacobs and Manzi, 1996). It follows that much of the emphasis in
critical accounts of the policy process is on communicative approaches to achieving a
more democratic policy process through diverse forms of participation and
negotiation.
In terms of social planning, Healey (1993: 249) argues that what is being invented
through these types of studies is a new form of planning that is both appropriate to our
recognition of the failure of modernity’s conception of ’pure reason‘; yet is searching,
as Habermas does, for the continuation of the Enlightenment project of democratic
progress through reasoned argument among ‘free citizens’. In doing so, Habermas
seeks to recuperate the ‘public sphere’ of modernity (Steele, 1997: 50). However,
holding onto a theory of reasoned argument among free citizens suggests that
participants in the policy process are momentarily able to step outside practices of
domination. This possibility is problematised by the postmodern insistence that
domination and resistance are embedded in all practices of power (Leonard, 1997).
The next section explores this claim further and in a broader sense considers the
contribution of postmodernism for critical social policy research and policy making
studies.
Postmodernism and Social Policy Research
Postmodernism is based on an epistemology, or more precisely an ontology that
acknowledges multiple meanings within any given discursive context. In this way,
postmodernism shares the foundations of a social constructionist epistemology, where
meaning resides in individual interpretations within specific contexts (Spivey, 1997).
Postmodernism and poststructuralism, separately and together, provide the discipline
of social policy with a range of pathways that assist in theorising ‘societal
restructuration’ at the discursive level. The multiple strands of postmodernism and
poststructuralism dispute totalising meta-narratives of western reason by arguing that
social progress is particular to specific contexts, which brings greater attention to
cultural regulation at the local level (Agger, 1991: 121; Penna and O’Brien 1996: 40).
In considering the implications of these claims for social policy analysis, I have
focused on how social power, the state and universalism are represented, followed by
a discussion about poststructural understandings of the policy process itself.
Social Power, the State and Universalism
In terms of understanding power and policy analysis, postmodernism offers an
important contribution, however, not in the ‘top-down’ or authoritative sense that
state-political power has been understood in social policy research (Ham and Hill,
1987: 22). For postmodernists, power is not principally repressive, and it must be
practised, rather than possessed. Foucault believes that individuals are always in the
position of simultaneously undergoing and exercising power (Jacobs and Manzi,
1996: 549). In other words, power is both multi-dimensional and complex (Jacobs and
Manzi, 1996: 550).
304
There are a number of implications of this position for researching and theorising
policy change. First, postmodernism argues that the perceptions of individuals are a
central dimension for understanding power relationships. Second, language as an
expressive act is a form of power, and language is a creative and controlling force
(Jacobs and Manzi, 1996). This dual account of power means that research attention
must be given to both ‘top down’ and ‘bottom up’ understandings of power, to
identify practices of resistance that disrupt domination. Postmodern theory provides
an important contribution for understanding micro-processes, precisely because these
intellectual perspectives reject grand theories that seek to offer only one interpretation
or one explanation of social phenomena. Attention to resistance and other microrebellions at the local level of policy practice help to overcome a totalising, repressive
and functional view of social power and the state.
Social policy research generally treats the state as a unitary object with its own
rationale, motivations and interests. In this normative account, the welfare state is
positioned as some kind of ‘thing’ that thinks and responds; in contrast, postmodern
narratives dismember the state, emphasising the various and inconsistent practices,
which shape its manifold components (Petersen et al., 1999: 8). The position of the
state as the main seat of power is also downplayed in poststructural and postmodern
approaches. In the Foucauldian governmentality literature, the state emerges as one
segment of a much broader play of power relations involving professionals, schools,
families, leisure organisations and so forth (Dean, 1999). In this respect, among this
network of state practices, it is not possible, nor it is desirable to develop welfare
programs based on grand narratives divorced from local contexts.
The Welfare Subject
This decentred account of the state also problematises conceptions of the universal
welfare subject. In his critique of positivism, Foucault (1983) argues that the
dominant discourses of modernity transformed human beings into universal subjects
through ‘modes of objectification’, where the individual was created as an object of
study and intervention to be regulated and divided through various binaries and
exclusions (for example sane/insane, heterosexual/homosexual, bad/good). Social
policy, for example, constructs its subjects, such as the ‘needy’, ‘disabled’ and
‘homeless’. Bodies, minds, values, interests and behaviours are then inscribed onto
them (Gibbins, 1999: 37). The subject within this modernist discourse had to possess
a fixed identity, able to be placed in the social order as either one of ‘us’, or one of
‘them’ (Leonard, 1997: 17).
Modern welfare inspires a belief in the universal human subject, which exists in spite
of differences across time, place, culture, gender and ethnicity (Fuery and Mansfield,
2000: 6). Basically, there is an inherent universalism in the project of modernity that
has little place for particularist knowledge and identities (Thompson and Hoggett,
1996; Ellison, 1999). In postmodern feminist theorising, critiques of positivism have
continued to build (Yeatman, 1994; Weedon, 1997). In the wake of feminist and postcolonial criticism, the debate has sometimes polarised between those who cling to the
tradition based on a model of universal humanity, and those who see human society as
a field of endless and irreducible differences (Fuery and Mansfield, 2000: 6). In the
field of social policy, this is by no means a new debate. Over the past twenty years
there has been a sustained critique of universalism from various sources, including
305
GREG MARSTON
feminists, single parents, disability groups, anti-racist groups and many others (Mann,
1998: 99).
There is a dichotomy here that runs to the heart of social policy analysis. On one side
stands the universalist plea for greater social justice and equality, a ‘fair’ allocation of
social goods to mitigate the inegalitarian effects of the market; on the other, demands
for the recognition of diversity and ‘difference’ sustain the view that universalism can,
paradoxically, be socially exclusive (Ellison, 1999: 59). The question about
reconciling particularism and universalism is obviously central to debates about the
value of postmodernism for social policy; however, there is not sufficient space to
address this issue here. The basic point is to recognise that the search for solidarity on
basic economic and social issues, while retaining a commitment to responding to
diversity, will continue to be of considerable political importance (Leonard, 1997:
176). Dismembering the state and the goals of social policy are not the only objects
that are subject to the postmodern critique. Postmodernism also raises valuable
questions for studying the policy process, particularly in relation to the category of
discourse.
Poststructuralism, Discourse and the Policy Process
In the case of policymaking research, French poststructuralism, and particularly
Foucault’s work on discourse, offers an important contribution. Using Foucault’s
work, one author who has been at the forefront of the study of the symbolic and
linguistic aspects of policy making since the 1960s is Murray Edelman (Parsons,
1995: 179). In one of his later books, Constructing the Political Spectacle Edelman
(1988) argues that the real power in policy-making resides in the process whereby
problems are constructed and articulated, since it is through language that we
experience politics. Similarly, in the Australian context, Bacchi (1999) has developed
what she refers to as a ‘what’s the problem?’ approach to policy analysis, where the
deconstructive analytical emphasis is on how the policy problem is discursively
represented because this contains an explicit or implicit diagnosis about how the
problem should be addressed. Both these political theorists draw on the work of
Foucault and Derrida in making sense of discourse and policy change.
Despite the valuable work of these policy theorists, interrogating and deconstructing
the language of policy has not received sufficient attention in social policy research.
In fact, there is limited evidence about the influence of discourse analysis on the
social policy community (Jacobs, 1999: 210). While the term ‘discourse’ itself has
entered social policy writing, it is also the case that there is often little theoretical
acknowledgement or engagement with the social theories that have placed discourse
on the social policy agenda. The lack of explicit interrogation of policy language may
seem somewhat surprising, given that policy-making is heavily constituted by
discursive practices. As Majone (cited by Fischer and Forrester, 1993: 2) states: ’As
politicians know only too well, but social scientists often forget, public policy is made
of language. Whether in written or oral form, argument is central in all stages of the
policy process‘. Yet, policy research has largely ignored the possibilities offered by
discourse analyses in favour of super-structural accounts of social change pursued
within the social science of variables (Jacobs and Manzi, 1996; Hastings, 1998).
306
The Social Construction of Policy Problems
Positivist conceptions of language as an objective mirror of an external reality are
destabilised by poststructuralism. ’Poststructuralism emphasises the instability and
contingency of the structural context of social interaction‘ (Torfing, 1999: 55). The
plurality of language and the impossibility of fixing meaning once and for all are
basic principles of poststructuralism, informed by a social constructionist
epistemology (Weedon, 1997: 82). The implication of using a social constructionist
epistemology is not to question the existence of the ‘object’ per se; it is the
universalism of knowledge that is questioned in these accounts. Lyotard (1984), for
example, believes that knowledge is stable, and even predictable within specific
situations – but it is not necessarily universal.
There is of course nothing particularly enlightening about suggesting that objects are
constructed in thought. The research innovation comes from showing how policies are
constructed and contested. For showing that policy problems are socially constructed
implies – in theory at least – that they can be constructed differently and that there are
alternatives to the dominant discourses that currently define the terrain of social
policy. The aim then is to encourage reflexivity and inspire the moral imagination of
policy makers, or at the very least to make policy-makers consider the consequences
of talking, acting and thinking in certain ways. To take a current example in the
Australian political context, the Commonwealth Government constructs asylum
seekers as ‘illegal refugees’, or simply ‘illegals’. This construction immediately
positions this group of people as lawbreaking ‘others’, who should be treated with
contempt, rather than compassion.
Moreover, constructing the policy problem in terms of ‘illegality’ immediately
legitimises the current policy prescription of enforced detention. This example
highlights how policy language can become its own form of ‘cultural injustice’
(Fraser, 1997). A basic assumption in this approach to policy research, is that
discourses, which are regularised and ritualised (that is racist, sexist), have as much
capacity to injure, to oppress and to exclude as the material barriers that people face in
their everyday lives (Corker, 2000: 447). Yet, this does not mean that discourses are
fixed; they are also contestable and contingent.
Policies as Socially, Culturally and Historically Contingent Realities
In accepting the poststructural claim that social policies are not pre-existing givens,
reflecting an underlying reality, it follows that what constitutes a policy problem or
solution is historically, socially and culturally contingent (Yanow, 1996; Torfing
1999). Contingent meanings are inherently unstable discursive structures. Discourses
that underpin the welfare state, for example, are not simply imposed from above but
involve the way people view themselves; hence dominant discourses are potentially
fragile and malleable. Policy meanings are only ever partially fixed and different
participants in the policy process are, at various times, engaged in a struggle to
advance their particular version of the policy problem or policy solution. ’In this sense
the problem and the participants are ”mutually constitutive”: the one reinforces the
other‘ (Colebatch, 1998: 27). It follows that from a post-structural perspective the
objects of investigation are both the policy products (documents, legislation) and the
individual and organisational activities that develop and implement these products.
307
GREG MARSTON
In capturing this process, the aim is to find an approach that makes it possible to
consider the relative autonomy of symbolic and cultural systems without giving up the
traditional political-economic focus of critical social policy. According to Nancy
Fraser (1997: 7) both cultural politics (recognition/identity) and social politics
(redistribution) must be re-coupled, both practically and intellectually, if we are to
conceive of provisional alternatives to the present political order:
A critical approach must be ’bivalent‘, in contrast, integrating the
social and the cultural, the economic and the discursive. This
means exposing the limitations of fashionable neostructuralist
models of discourse analysis that dissociate ’the symbolic order‘
from the political economy.
It is important to interpret critical social theory and poststructural accounts in terms of
their possibilities of building a comprehensive conceptual framework for the study of
social policy change which is able to accommodate structure and agency, and able to
investigate the relationship between discursive and material change at multiple levels
of analysis. The next section will briefly articulate a specific form of discourse
analysis that holds onto the critical tradition while at the same time utilising the
insights of poststructuralism and postmodernism discussed above.
3
A Possibility for Dialogue: Overcoming an Either/Or Position
One site in which poststructuralism and critical social policy analysis can meet is in
an adequate theory and method of discourse analysis that recognises both material and
discursive practices. The starting point is to keep a critical focus on questions of
justice, rights and material disadvantage, and to avoid the trap of unqualified
relativism which, at its most extreme, postmodernist thinking embraces (Ellison,
1999: 70). There is a risk in wholeheartedly abandoning conventional paradigms,
embracing the new or adopting an either/or position in relation to social theory.
Poverty, for example, is an ideological formation, it is a truth produced by particular
discursive strategies, but it is also a social construction – and people die from it
(Clarke, 1998: 183). It is clear from this example that where social constructionism
becomes problematic is where it disregards the relative solidity and permanence of
some social entities.
In working through these epistemological issues and complexities, policy researchers
must not be bound by simplistic dichotomies that force them to choose between
critical social theory and postmodernism (Agger, 1992: 75). Each body of theory
needs to be seen as a set of intellectual tools that can be used together, but at the same
time held in constant tension. Moreover, in suggesting that a dialogue needs to be
promoted, it is not simply a case of social policy listening to social theorists, it is also
a case of social theorists being a little more reflexive and a little more attentive to
debates within social policy (Mann, 1998). There is sometimes a tendency among
postmodern writers to represent the welfare state singularly in terms of its public
welfare and disciplinary practices (Rose, 1996, 1999), while remaining silent on the
function and features of occupational and fiscal welfare. Similarly, in the work of
Fairclough (1992) there is a tendency to discuss the forces of marketisation and neoliberalism as if they were a colonising force that has swept into previously protected
spaces. Yet, the Australian welfare state, like many others in the western world, has a
long history of market forms of human service provision through a mixed economy of
308
welfare. Arguably, what has changed in recent years is the intensification of market
forms of welfare provision.
Moreover, the theoretical tools developed in the ‘post’ literature – discourse,
language, power/knowledge, subjectivity – need to be coupled with everyday
struggles and practical concerns. As Mann (1998: 101) argues, it is the struggles of
those who have to rely on public welfare that have done much to generate a more
reflexive but distinctly critical social policy, and it is to these critical voices, these
subject identities that we should pay much more attention. In other words, in social
policy analysis, it is not enough to simply engage in academic discourse analyses of
written policy texts. As Jacobs (1999: 204) states: ’A commitment to both text and
practice is necessary to understand the policy process‘. In the interests of producing a
richer and more politically informed account of policymaking we need to consider the
interpretative accounts of policy actors, paying particular attention to which groups
are able to access the production of policy discourses and the reasons others are
marginalised. In a policy context, this involves empirically exploring the discursive
and non-discursive components of policy implementation and the way meanings are
made, used, conveyed, disseminated and translated (Healey, 1993: 28).
Meanings are produced in the interpretation of a text and therefore they are open to
diverse readings, which may differ in their ideological import, hence the importance
of capturing the voices of those that negotiate the policy field (Fairclough, 1992: 36).
Indeed, the intersection of discourses can be regarded as moments of interactions
between actors, not just of the relationship between discourses, which are moments of
such events. In a research sense, this means bringing together the material focus of
traditional social policy analysis with the poststructural insights about the
constitutional power of language and representation. Paying attention to these voices
demands a theory of discourse that explicitly incorporates the perspectives of those
that find themselves excluded from policy processes. One approach that meets these
demands is critical discourse analysis.
Using Critical Discourse Analysis to Study Social Policy Change
Critical discourse analysis (CDA) is basically a sociological account of language use.
Critical discourse analysis attempts to bring together a political economic framework
appropriated from critical theory, with its focus on material inequalities, together with
the insights of cultural studies and poststructuralism. Critical discourse analysis is one
approach that is able to provide evidence of ongoing processes, such as the
reconstitution of knowledge and the reconstitution of ‘self’ through detailed textual
analysis, which acts as a counter balance to overly rigid and schematising social
analyses that have plagued social policy research (Fairclough, 1995: 209). In
countering an overly rigid structural analysis, however, CDA does not reduce all
social practices to a textual reality. Importantly, CDA differs from some versions of
discourse analysis that end up seeing the social as nothing but discourse, which in
itself represents a form of ‘discourse idealism’ (Chouliaraki and Fairclough, 1999:
117).
CDA is interested in investigating the category of discourse as an instrument of power
and discourse as an instrument of the social construction of ‘reality’ (Sotillo and
Starace-Nastasi, 1991: 412). Whereas much of the earlier work on language use and
discourse has a linguistic bias, critical discourse analysis provides a more socio309
GREG MARSTON
political emphasis; which is what makes it suitable for social policy analysis (vanDijk, 1990: 8). In terms of theoretical orientation, critical discourse analysis shares
critical theory’s concern with ideology, hegemonic politics and the postmodern
critique of the value-free assumptions and cause-effect relationships underlying
positivist social science (Agger, 1991: 109). In this way, critical discourse analysis
makes it possible to forge links among critical theory, poststructuralism and
postmodernism and the links between micro, meso and macro accounts of discourse.
A Multi-Conception of Discourse Place
Discourses have micro features, in terms of their linguistic structure; however,
discourses also operate at more global, societal and cultural levels. At this macrolevel, systems of discourse are closely associated with ideology, hegemony and with
the enactment and legitimation of power (van Dijk, 1998: 9). Hence, the term
‘critical’ implies making visible the interconnectedness between concepts and
practices (Fairclough, 1995: 36). The relationship between macro, meso and micro
levels of discourse is of primary concern to Norman Fairclough (1989, 1992, 1995),
one of the founders of critical discourse analysis, who conceptualises discourse in
terms of three main dimensions. Textual analysis (micro) is concerned with
description about the form and meaning of the text2, discourse practice (meso)
focuses on the discursive production and interpretation of the text and socio-cultural
practice (macro) operates at the level of broader social analysis. In a policy setting,
this model can be used to study how policies are produced and interpreted.
A way of researching discourse at these multiple levels is by focusing on the network
of institutions that produce and interpret policy texts within established policy
communities (such as the housing sector, education policy, health networks, or the
community sector). For Fairclough (1995: 37) the institutional site, such as the
government, the school, or the family emerges as the pivotal site where social change
functions both upwards towards the social formation and downwards to concrete
social actions and material practices. Social institutions are (amongst other things) an
apparatus of verbal interaction, a site of discursive practices (Fairclough, 1995: 38).
Institutions, such as the government, government departments and community
agencies are critically important policy-making sites.
There are a number of recent studies that illustrate how critical discourse analysis can
be made operational in the context of social policy research (see Jacobs and Manzi,
1996; Wodak, 1996; Hastings, 2000; Marston, 2000). What has been missing from a
lot of the empirical research to date has been an account of how policy actors actually
interpret and negotiate contemporary discourses that define the contested discourses
of the welfare state (Hastings, 2000), such as neo-liberalism, communitarianism, and
the ‘new moral economy’ of welfare (Rodger, 2000). To address this gap, there is a
need to do more work on identifying resistant practices in the face of discursive
change. A poststructural reading of hegemonic practices is a useful way of theorising
these forms of contestation.
2
A text is taken to be any instance of written or spoken discourse.
310
Critical Discourse Analysis, Hegemony And Resistance
Hegemony can be defined as the discursive face of power, and as such is a central
concept in critical discourse analysis, and is particularly useful when thinking about
contestation over social policy issues. Hegemony is the attempt to provide
authoritative definitions of social needs, and the power to shape the political agenda
(Fraser, 1991: 100) In the social policy arena, hegemony is therefore concerned with
how social policies are culturally framed. Framing involves the attempt to fix policy
meanings and welfare subjectivities, which are used in political discourse to
legitimatise policy directions (for instance ‘dole bludger’, ‘welfare dependant’).
Hegemonic struggles in social policy discourse are in a sense a struggle about what it
means to be a welfare recipient.
This conception of subjectivity as a site of struggle helps us to focus on the way that
different social identities are discursively formed through power relationships within
welfare state structures. There is a pressing need to understand the way differences
and identities and the attendant ideologies of legitimacy and entitlement are
constructed in welfare discourses and how these relate to social relations of power
(Brah, 1996). This is particularly important given that the ‘regulation of the poorest
groups’ is not only material but also involves a discursive regulation, which is
inscribed with categories of contested identities (Taylor, 1998: 348). Hence, the
welfare subject is understood in critical discourse analysis as decentred and
heterogenous and is neither unified nor fixed, but fragmented (Brah, 1996: 122).
However, the capacity for voluntary action is retained within critical discourse
analysis. To get beyond the philosophical stalemate between subjectivity as a mere
ideological effect on the one hand, and the idea of an autonomous subject on the
other, it is necessary to look at micro-relations of power, which is where the concept
of resistance comes to the fore in critical discourse analyses of social policy issues.
Here, one is trying to render an account of the diverse everyday experiences of
heterogenous subjects as they struggle with the relationships between determining
structures, that which is internalised from these structures, and what remains in their
own intentions, albeit mediated in culture (Leonard, 1997: 47).
In research terms, it is important to focus not only on how hegemonic domination is
secured and reproduced at the expense of transformation, but also how subjects may
contest and progressively restructure domination through everyday practice
(Fairclough, 1992: 34-35). In terms of the broader objectives for policy-making
studies, ‘Discourse analysis can assist policy makers in reconceptualising their
approach to problems and to understand why certain issues come to be seen as
problems’ (Jacobs, 1999: 211). In other words, critical discourse analysis can be used
to deconstruct the policy texts to reveal assumptions, subject positions and social
relations between and within institutional contexts.
The recent Commonwealth Government welfare reform agenda provides an example
of the possible value of this form of analysis. The starting point is to problematise,
rather than accept, the discursive construction of the policy problem. In the case of
welfare reform, the policy problem was predominantly framed by the Commonwealth
Government in terms of ‘welfare dependency’. Structural explanations, such as labour
market changes, were jettisoned in favour of a moralistic, individualist account of the
311
GREG MARSTON
policy problem where welfare claimants were positioned as holding back the
‘economic competitiveness’ of the nation (Newman, 1999).
Policy academics and policy activists using this approach would interrogate the
discourse of welfare reform, rather than accepting the ‘policy problem’ as
representing some form of objective reality. In analysing this issue, critical discourse
analysis is interested in the connections between the socio-cultural context and the
text itself. A critical discourse analyst might ask what made this discursive policy
frame possible and more importantly, what are its multiple effects? Or we could ask:
how has the very notion of welfare become so derisory and why has ‘being
dependent’ become so pejorative? And what does this individualistic discourse mean
for governance and the emergent role of the state? In terms of the policy actors it
would be valuable to get an interpretative account of the policy process that was used
to develop the welfare reform package, to get a sense of the discursive contestation
that defined this particular issue – to capture the ‘messiness’ of policy-making that
gets lost in official discourse and government publications.
The political nature of these questions, and the power of this form of analytics, is to
disrupt the ‘naturalness’ and the inevitability of current rationales at the level of
policy practice. The assumption in this form of analytics is that criticism itself can be
a real power for change, depriving some practices of their ‘self-evidence’ or
‘naturalness’, while extending the bounds of the thinkable to permit the invention of
others (Burchell and Miller, 1991). In contrast to some forms of critical theory, the
intention of critical discourse analysis is not to put forward and impose a ‘blueprint’
for how people should live in the future, but to offer resources for transformation by
identifying resistance practices where they occur and giving a voice to those policy
actors that struggle to be heard in their attempts to access policy practices.
Highlighting positive forms of resistance, and micro-rebellions against dominating
discourses is an important social and political process. This does not mean that we do
not need our grand narratives to make some sense of global social change, however,
we should not make assumptions about how these grand narratives are experienced
and transformed by actors at the local level of policy practice.
4
Conclusion
In the first part of this paper, it was argued that in order to understand the changing
nature of social services, social policy research must go beyond static explanation and
normative policy analysis. Creating an intellectual and practical space for this to occur
will continue to be a struggle, given the weight of history behind social policy
education and the personal and professional investment that maintain particular
paradigms within research institutions. In negotiating the sometimes extreme
positions between positivism and poststructuralism and in seeking to avoid ‘trench
warfare’, this paper has argued for a common ground between the traditional material
concerns of social policy with patterned disadvantage and the radical constructivist
claims that reduce all objects of study to mere artefacts of knowledge represented in
ever shifting textual relations.
In taking this position, I recognise that there is a conservative and relativist strand
with postmodern writing, however, the ‘anything goes’ approach can be avoided as
long as a political focus on contemporary social struggles and ethical practices
remains at the centre of the research focus. It is also imperative that a capacity for
312
voluntary agency is retained, given the difficulty of conceiving of social policy
without its intentionally acting subjects. In addition, ‘critical postmodernism’ must
not only be interested in the sources of discourses, but also the lived effects of
discursive practices, particularly the dialectical relationship between discursive and
material change. These theoretical insights will also need to be worked out
methodologically to suit particular research interests within the social policy
discipline.
Critical discourse analysis is one approach that potentially meets this challenge and
provides a useful dialogue between critical theory and postmodern theorising.
In many ways, critical discourse analysis speaks to the core goals of social policy.
Social policy research has a long and significant history of investigating and
addressing material disadvantage, yet it is still in its infancy in terms of sufficiently
understanding the culture injustices that play a part in maintaining and obscuring
continuing inequalities. Critical discourse analysis is able to show how these material
practices are constituted by discourses. When used in conjunction with other
qualitative approaches, critical discourse analysis has the potential to be a powerful
and revealing form of social inquiry, particularly when used to investigate the nature
of changes and forces that are shaping welfare state programs and forms of service
delivery at the local, institutional and socio-cultural level. In its significance, the
discursive dimension of social change and its role in reproducing material inequalities
and practices of domination and resistance is unlikely to diminish in the future. As
Poynton (2000: 38) concludes:
Possibilities for an oppositional stance – for any kind of resistance
to the operations of new forms of power- are giving way to very
different kinds of engagement with issues of power, government
and their legitimation. Discourse remains at the heart of these
issues, however, rendering critical forms of discourse analysis –
current and future – of ongoing importance.
References
Agger, B. (1991), ‘Critical theory, postructuralism, postmodernism, their sociological
relevance’ Annual Review of Sociology, 17, 105-131.
Agger, B. (1992), Cultural Studies as Critical Theory, The Falmer Press, London.
Bacchi, C. L. (1999), Women, Policy and Politics, Sage Publications, London.
Bosworth, M. (1999), Resisting Identities: Agency and Power in Women's Prisons,
Ashgate Dartmouth Publishing, Aldershot.
Bottomore, T. (1984), The Frankfurt School, Tavistock Publications, London.
Brah, A. (1996), Cartographies of Diaspora: Contesting Identities, Routledge,
London.
Burchell, G., C. and P. Miller (1991), eds, The Foucault Effect-Studies in
Governmentality: with Two Lectures by and an Interview with Michel Foucault,
The University of Chicago Press, Chicago.
Carter, J. (1998), ‘Postmodernity and welfare: when worlds collide’, Social Policy
and Administration, 32, (2), 101-115.
Chouliaraki, L. and N. Fairclough (1999), Discourse in Late Modernity: Rethinking
Critical Discourse Analysis, Edinburgh University Press, Edinburgh.
313
GREG MARSTON
Clarke, J. (1998), 'Thriving on chaos? Managerialism and social welfare', in J. Carter,
ed, Postmodernity and the Fragmentation of Welfare, Routledge, London.
Colebatch, H. K. (1998), Policy, Open University Press, Buckingham.
Corker, M. (2000), 'Disability politics, language planning and inclusive social policy',
Disability and Society, 15. (3), 445-461.
Culpitt, I. (1999), Social Policy and Risk, Sage Publications, London.
Dalton, T., W. Weeks, and J. Wiseman. (1996) eds, Making Social Policy in
Australia: An Introduction, Allen and Unwin, Sydney.
Deacon, A. and K.Mann. (1999), 'Agency, modernity and social policy', Journal of
Social Policy, 28 (3), 413-435.
Dean, M. (1999), Governmentality: Power and Rule in Modern Society, Sage
Publications, London.
Edelman, M. (1988), Constructing the Political Spectacle, The University of Chicago
Press, Chicago.
Ellison, N. (1999), 'Beyond universalism and particularism: rethinking contemporary
welfare theory', Critical Social Policy, 19 (1), 57-87.
Fairclough, N. (1989), Language and Power, Longman, London.
Fairclough, N. (1992), Discourse and Social Change, Polity Press,Cambridge.
Fairclough, N. (1995), Critical Discourse Analysis: The Critical Study of Language,
Longman, London.
Ferge, Z. (1997), 'The changed welfare paradigm: the individualisation of the social',
Social Policy and Administration, 31 (1), 20-44.
Fischer, F. and Forester, J. eds (1993), The Argumentative Turn in Policy Analysis
and Planning, Duke University Press, Durham.
Fitzpatrick, T. (1996), 'Postmodernism, welfare and radical politics', Journal of Social
Policy, 25 (3), 303-320.
Forester, J. (1993), Critical Theory, Public Policy, and Planning Practice: Toward a
Critical Pragmatism, State University of New York Press, Albany.
Fraser, N. (1989), Unruly Practices: Power Discourse And Gender in Contemporary
Social Theory, Polity Press, Cambridge.
Fraser, N. (1991), 'The uses and abuses of French discourse theories for feminist
politics', in P. Wexler (ed) Critical Theory Now, Falmer Press, London, 98-117.
Fraser, N. (1997), Justice Interruptus: Critical Reflections On The ’Postsocialist‘
Condition, Routledge, New York.
Fuery, P. and Mansfield, N. (2000), Cultural Studies and Critical Theory, Oxford
University Press, Oxford.
Gibbins, J. (1999), 'Postmodernism, poststructuralism', in J. Carter (ed) Postmodernity
and the Fragmentation of Welfare, Routledge, London.
Ham, C. and M. Hill, (1993), The Policy Process in the Modernist Capitalist State,
Harvester Wheatsheaf, New York.
Hastings, A. (1998), 'Connecting linguistic structures and social practices: a
discursive approach to social policy analysis', Journal of Social Policy, 27 (2),
191-211.
Hastings, A. (2000), 'Discourse analysis: what does it offer housing studies?', Housing
Theory and Society, 17, 131-139.
Healey, P. (1993), 'The communicative turn in planning theory', in F. Fischer and F.
Forester, eds, The Argumentative Turn in Policy Analysis and Planning, Duke
University Press, Durham.
Healy, K. (2000), Social work practices: contemporary perspectives on change, Sage
Publications, London.
314
Hughes, J. (1990), The Philosophy of Social Research, Longman House, London.
Jackson, N. (1999), 'The council tenants' forum: a liminal public space between
lifeworld and system', Urban Studies, 36 (1), 43-58.
Jacobs, K. (1999), 'Key themes and future prospects: conclusion to the special issue',
Urban Studies, 36 (1), 203-213.
Jacobs, K. and Manzi, T. (2000), 'Evaluating the social constructionist paradigm in
housing research', Housing, Theory and Society, 17, 35-42.
Jacobs, K. and Manzi, T. (1996), 'Discourse and policy change: the significance of
language for housing research', Housing Studies, 11 (4), 543-560.
Leonard, P. (1997), Postmodern Welfare: Reconstructing and Emancipatory Project,
Sage Publications, London.
Lyotard, J-F. (1984), The Postmodern Condition: A Report on Knowledge, University
of Minnessota Press, Minneapolis.
Mann, K. (1998), 'Lamppost modernism: traditional and critical social policy', Critical
Social Policy, 18, (1), 77-102.
Marston, G. (2000) ‘Metaphor, morality and myth: a critical discourse analysis of
public housing reform in Queensland’, Critical Social Policy, 20 (3), 349-373.
Newman, J. (1999), 'The Future of Welfare in the 21st Century', National Press Club,
Canberra.
O'Brien, M. and Penna, S. (1999), 'Oppositional postmodern theory and welfare
analysis: anti-oppressive practice in a postmodern frame', in J. Carter, ed,
Postmodernity and the Fragmentation of Welfare, Routledge, London.
O’Neill, J. (1995) The Poverty of Postmodernism, Routledge, London.
Parsons, W. (1997), Public Policy: An Introduction to the Theory and Practice of
Policy Analysis, Edward Elgar, Cheltenham.
Penna, S. and M.O’Brien. (1996), 'Postmodernism and Social Policy: A Small Step
Forwards', Journal of Social Policy, 25 (1), 39-61.
Petersen, A., J. Dudley and P. Harris (1999), Poststructuralism, Citizenship and
Social Policy, Routledge, New York.
Poynton, C. (2000), 'Linguistics and discourse analysis', in Poynton, C. and A. Lee
(eds) Culture and Text: Discourse and Methodology in Social Research and
Cultural Studies, Allen and Unwin, Sydney.
Rodger, J. (2000), From a Welfare State to a Welfare Society: The Changing Context
of social policy in a Postmodern Era, Macmillan, London.
Rose, N. (1996), 'The death of the social? Re-figuring the territory of government',
Economy and Society, 25 (3), 327-356.
Rose, N. (1999) Governing the Soul: The Shaping of the Private Self, Free
Association Books, London.
Smith, J. (1998), Social Science in Question, Sage Publications, London.
Sotillo, S. and D. Starace-Nastasi (1999), 'Political discourse of a working class
town', Discourse and Society, 10 (3), 411-438.
Spivey, N. (1997), The constructivist metaphor : reading, writing, and the making of
meaning, Academic Press, San Diego.
Steele, M. (1997), Theorising Textual Subjects, Cambridge University Press,
Cambridge.
Taylor-Gooby, P. (1994), ‘Postmodernism and social policy: a great leap backwards’,
in Journal of Social Policy 23 (3), 385-404.
Taylor, D. (1998), 'Social identity and social policy: engagements with postmodern
theory', Journal of Social Policy, 27 (3), 329-350.
315
GREG MARSTON
Thompson, S.and P. Hoggett, (1996), 'Universalism, selectivism and particularism',
Critical Social Policy, 16 (1), 22-43.
Torfing, J. (1999), New Theories of Discourse, Blackwell, Oxford.
van-Dijk, T. (1990), 'Discourse and Society: a new journal for a new research focus',
Discourse and Society, 1 (1), 5-16.
van Dijk, T. (1998), Ideology, Sage Publications, London.
Watts, R. (1993, 1994) ‘Government and modernity: an essay in thinking
governmentality', Arena, 2,103-157.
Weedon, C. (1997), Feminist Practice and Poststructuralist Theory, 2nd edition,
Blackwell, Oxford.
Wodak, R. (1996), Disorders of Discourse, Longman, London.
Wolman, H. (1981), ‘The determinants of program success and failure’, Journal of
Public Policy, 1 (1), 433-464.
Yanow, D. (1996), How Does A Policy Mean? Interpreting Policy and
Organisational Actions, Georgetown University Press, Washington.
Yeatman, A. (1994), Postmodern Revisionings of the Political, Routledge, New York.
316
Reforming the Welfare System in Remote
Aboriginal Communities: an Assessment of
Noel Pearson's Proposals
David Martin1
Centre for Aboriginal Economic Policy Research
Australian National University
1
Introduction
Aboriginal lawyer, activist and social commentator Noel Pearson has recently argued
that the current mode of delivery of welfare services to Aboriginal people is deeply
antithetical to their interests and well-being. Central to his scheme for policy change
and improved welfare outcomes are certain core propositions.
The first of relevance to this discussion is that the welfare policies instituted over the
past three decades have produced what Pearson calls a ‘gammon’ (or artificial)
welfare economy within Aboriginal societies which contrasts with the ‘real’
economies of both traditional Aboriginal subsistence and the market economy. It is
the absence of meaningful reciprocity between recipients of services and those
providing them which renders the welfare economy ‘gammon’, in Pearson’s view
(Pearson, 2000a).
In common with Mark Latham (Latham, 2001), Pearson terms this ‘passive welfare’,
and argues that it promotes detrimental relations of passivity and dependence which
have now become deeply embedded within Aboriginal society and culture. In fact,
Pearson argues that ‘passive welfare’ is the poisoned flour of the modern settler state,
and that it is ‘passive welfare’ which is primarily responsible for the social
devastation in the remote Cape York Aboriginal communities with which he is in the
first instance at least concerned (Pearson, 2000a: 26-39).
Pearson’s second key proposition is that addressing the dysfunctional consequences of
passive welfare for Aboriginal people will require profound structural change, in
particular the development of new institutions of Aboriginal governance, both formal
and informal. It will also require the reform of existing institutional arrangements for
dealing with the resources provided through the welfare system, including those
arrangements that come under the aegis of the State.
1
This paper draws upon research conducted through the Centre for Aboriginal Economic
Policy Research (CAEPR) at the Australian National University, and published as a
CAEPR Discussion Paper (Martin, 2001).
Martin, D. (2002), ‘Reforming the welfare system in remote Aboriginal communities: an assessment of
Noel Pearson’s proposals’, in T. Eardley and B. Bradbury, eds, Competing Visions, Refereed
proceedings of the 2001 National Social Policy Conference, SPRC Report 1/02, Social Policy Research
Centre, University of New South Wales, Sydney, 317-325.
DAVID MARTIN
Pearson argues that genuine partnerships between government and Aboriginal people,
mediated through these institutions, must replace the current mechanisms controlled
by government. Negotiation must replace ‘consultation’. It is through reform of the
existing institutional arrangements, and by means of these formal and informal
Aboriginal institutions, that the reciprocity and individual responsibility necessary to
transform the ‘gammon’ welfare economy to a ‘real’ economy can be implemented
(Pearson, 2000a: 67-82).
2
Positioning Pearson’s Arguments
Noel Pearson’s arguments should be seen as a welcome and politically innovative
contribution to a policy debate of fundamental importance to Aboriginal people, and
indeed to us all. Maintaining the status quo, at least for remote Aboriginal Australia, is
clearly not sustainable.
Pearson has responded to the cry of ‘something must be done’ with a passionately
argued manifesto which sets out his views of the causes of the deep social malaise
confronting remote Aboriginal communities, and a set of proposals for addressing this
malaise which links the requirement for changes at the policy and service delivery
levels with that for socio-cultural change within Aboriginal communities themselves.
While Pearson focuses on a particular region of remote Aboriginal Australia, his
arguments have been carefully crafted to resonate with current concerns in the wider
political arena regarding reform of the welfare sector. It is in this crafting that one
important dimension of his innovation lies. Many of Pearson’s themes would appear
to be drawn from so called ‘third way’ political philosophy, particularly that relating
to welfare policy.
As outlined by Mark Latham, the ‘third way’ is predicated on the belief that a strong
economy and a strong society are mutually reinforcing. It emphasises a particular set
of political values, including the responsibility of people for their actions and for
contributing to society in return for their rights and benefits, incentives for people to
save more, study harder and work more intelligently, and devolution by government
of the powers of democracy and of public provision closer to civil society (Latham,
2001: 25-6).
Pearson’s arguments that the social is the economic, for the need to rebuild the
foundations of social life within Aboriginal communities on mutuality - as he terms it,
‘reciprocity’ - and on the basis of meaningful economic activity, on the importance of
social entrepreneurs rather than welfare bureaucrats as facilitators of change, and on
government adopting the role of enabler rather than the sole instigator of policy and
dispenser of benefits would all appear to be consistent with the ‘third way’ agenda.
Of course, the notion that people should demonstrate some form of reciprocal
obligation in return for benefits provided by the state is not the exclusive province of
‘third way’ politics. It has also been adopted by conservative governments, including
those in Australia, where the current Federal Government has sought to address longterm welfare dependency by progressively instituting a range of policies and programs
under the rubric of ‘mutual obligation’. These various programs are aimed at reducing
318
REFORMING THE WELFARE SYSTEM IN REMOTE ABORIGINAL COMMUNITIES
the economic and social costs of long-term dependency on state benefits, by seeking
to prepare people for entry to the wider labour market, by ensuring that their capacity
to be ‘job ready’ is not lost, and by developing training as a specific component of
job-seeking activity. The relationship between ‘mutual obligation’ and Pearson’s core
concept of ‘reciprocity’ will be examined later in this paper.
3
Innovation
However, Pearson has done more than simply transfer the terms of a wider policy
debate into the Aboriginal affairs arena. His innovation lies in the particular manner in
which he has, almost unremarked, integrated them into an argument for fundamental
reform of the political landscape in Cape York, and for significant power to be
devolved from government to the Aboriginal domain.
Pearson is not averse to government involvement in itself. In fact, he argues that it is
imperative for government to provide the resources on which, in the absence of
economic self-reliance through the market economy, Aboriginal people must depend.
However, he argues for a fundamental restructuring of the means through which these
resources are provided. He proposes a form of self-determination which would reallocate responsibility for policy formulation and service delivery from government
alone to government and community partnerships. This shift would redefine the role
of the state ‘from a disabler to an enabler’ (Pearson, 1999: 33, see also 2000a: 53-4).
Another innovation has been in the way Pearson has subtly transformed the terms of
‘mutual obligation’ in his proposals to institute ‘reciprocity’. Pearson’s scheme gives
considerable weight to mechanisms for instituting what he terms the principle of
‘reciprocity’ as a core means by which the ‘gammon’ welfare economy is to be
transformed to a ‘real’ economy. It is this suggestion, above many others in his
proposals, that has been taken up approvingly by a range of social and political
commentators (e.g. Koch, 1999).
However, there are important differences between Pearson’s concept of ‘reciprocity’
and the political and policy principle of ‘mutual obligation’ as it is currently
articulated by government. As Pearson argues it, ‘reciprocity’ does not lie between the
Aboriginal individual and an undifferentiated, taxpaying Australian ‘community’
mediated by the state. He argues in part that the state is too remote from its citizens,
and furthermore does not have the moral authority with Aboriginal people to
appropriately undertake this role. Rather, ‘mutual obligation’ must be demanded and
implemented between the individual and his or her particular Aboriginal community,
family and local group (Pearson, 2000a: 85-7).
A further, and highly significant, innovation is that Pearson explicitly rejects
arguments that the current parlous state of Cape York’s Aboriginal people can be
solely attributed to the cumulative effects of the racism, dispossession, and trauma to
which they have been exposed. In Pearson’s view they cannot explain the rapid social
breakdown in these communities over the last three decades of the twentieth century.
319
DAVID MARTIN
This Pearson attributes to the artificial economies of these remote communities and
the corrupting nature of passive welfare (Pearson, 2000a: 29-39). Changing the
situation will require partnerships between Aboriginal people, who must take
responsibility for change, and government. Pearson thus directly, and courageously,
confronts a pervasive theme in contemporary Aboriginal political discourse, which
places the causes of current problems firmly within the history of colonisation, and
establishes Aboriginal people as its powerless victims.
4
The Impetus for Social Change
Pearson is clearly arguing for both structural and attitudinal changes in Cape York’s
Aboriginal societies, and recognises that these two dimensions are linked.
Transforming the current economy to a ‘real’ economy is a necessary precursor, in his
view, to transforming the current corrupted values into real reciprocity and mutual
obligation. I want to focus briefly on three inter-related questions which go to the
heart of his scheme. Briefly stated, these are:
•
•
•
5
how is change to be effected when the evidence suggests that certain
widespread Aboriginal values and practices may be inimical to such change;
where might the necessary moral and political authority to effect such change
lie; and
what are the implications of these factors for the new institutional
arrangements Pearson argues are required?
A New Moral Order
Underpinning Pearson’s scheme is the institution of a new moral order. Structural
change is required but the state, he argues, does not have the moral authority with
Aboriginal people to undertake this. Rather, moral authority, exercised through the
requirement for ‘reciprocity’, resides with units within Aboriginal society: variously
the ‘community’, local groups and the ‘family’.
Pearson’s arguments for a new moral and institutional order reflect his concern about
the often quite desperate situation in much of remote and rural Aboriginal Australia,
including Cape York. Maintaining the status quo is clearly an inadequate response,
both ethically and politically. In Pearson’s view, the necessary changes must be
demanded and implemented from within Aboriginal societies themselves, although
the state can assist by creating the necessary structural changes (e.g. at the policy
level). Pearson thus inverts the conventional rhetoric of Aboriginal self-determination
- which is typically focused on maintaining the uniqueness of Aboriginal social and
political and economic forms - and instead harnesses it to a dynamic for social
change.
Pearson’s scheme however is ultimately predicated upon a degree of coercion, albeit
arising from within the Aboriginal polity itself. His implicit reliance upon the capacity
of Aboriginal mechanisms operating within various levels of social grouping ‘family’, ‘local group’, and ‘community’ - to demand reciprocity and responsibility
320
assumes that certain deeply sedimented values and practices in these troubled and
fractured societies can be changed from within. Herein lies a paradox.
It is certainly arguable that a new moral order is required, as part of the necessary
structural change. But the realities of social process within contemporary Aboriginal
societies suggest that there will be considerable difficulty in locating centres of clear
moral authority. Attempts to institute the ‘community’, the ‘family,’ or other such
social units of Aboriginal society as sources of authority or suasion for the purposes
of implementing mutuality are likely to be ineffectual, or even actively resisted. There
is a parallel difficulty in locating clear centres of political authority in these essentially
acephalous societies.
I would suggest that Pearson is wrong for two reasons in attributing the contemporary
dysfunctional nature of these remote Aboriginal communities essentially to the impact
of the welfare system. Firstly, the past three decades have seen not just the
introduction of the welfare system, but a set of other changes which have had
profound impacts on Aboriginal societies of the region: for example, the collapse of
the pastoral industry, the move from the authoritarian regimes in Aboriginal
settlements to a new emphasis on self-management, and the introduction of alcohol, to
name but a few.
Secondly, contemporary Aboriginal values and practices (including those underlying
the dysfunction of which Pearson writes) have not simply arisen as the consequence
of colonialism or, as Pearson argues, the introduction of ‘passive welfare’. Aboriginal
people are not empty vessels into which new ideas have been poured. They have
responded to the dramatic changes in the structural circumstances of their lives in
accordance with distinctive values concerning such matters as personal identity, the
principles by which familial and personal relationships are ordered, and the
relationship between ‘economy’ and society. In turn, of course in an ongoing dialectic
these values have been profoundly impacted by these wider structural changes.
Pearson himself indirectly raises the difficult question as to whether certain core
Aboriginal values and practices may inhibit the capacity of the Cape York
communities to institute the changes that are, in his view, necessary. To put it another
way, Aboriginal people may actively resist attempts to change certain of their core
values and practices which are incompatible with the demands of the dominant
society.
6
Authorising Change
Given the diffuse and contested nature of authority structures in Cape York
Aboriginal societies (see e.g. Martin, 1993a, 1993b; Wyvill, 1990a, 1990b), it is far
from clear how the reciprocity and responsibilities of the kind Pearson envisages
might actually be instituted without active intervention, either by external individuals
or institutions, or by internal institutions (formal or informal) with the necessary
moral and political authority.
External interventions have a long and problematic history, in Cape York and
elsewhere, for example in the attempts in missions and government settlements to
321
DAVID MARTIN
reformulate Aboriginal beliefs and practices to accord with those of the external
agents of change (e.g. Kidd, 1997; for accounts relevant to north Queensland see also
Anderson, 1988; Chase, 1980; Finlayson, 1991; Martin, 1993b).
Pearson rejects such forms of external intervention, and is especially critical of the
‘white dictator’ leadership model of which missionaries and government
superintendents were exemplars (Pearson, 2000a: 49). Yet, it will prove very difficult
to locate appropriate and legitimate sources of moral and political authority within the
existing Aboriginal Cape York polity.
This brings us to the fundamental question: if (as Pearson suggests) the state does not
have the necessary moral authority to institute change, and if (as I would argue)
groupings within the Aboriginal polity such as ‘families’ and other local groups or the
residential ‘communities’ and their organisations also do not currently have the
relevant capacity or authority, then how is change to be authorised and implemented?
7
A New Institutional Order
As an intrinsic component of his proposal for social change in Cape York’s
Aboriginal communities, Pearson has argued for new institutions at the interface
between the Aboriginal polity and the state, and for power and decision-making to be
devolved to both formal and informal institutions at the regional, community and local
levels. Such arrangements should build on existing local and regional organisations
and capacities, Pearson argues, rather than supplanting or competing with them
(2000a: 65-73). It is through these institutions that reciprocity and responsibility are to
be demanded and implemented.
However, Pearson’s focus on economic factors (welfare) as the primary cause of the
contemporary malaise and dysfunction within Cape York’s Aboriginal communities
has led him to give undue emphasis to the role of these proposed institutions in
controlling and distributing government resources. In particular, Pearson has
concentrated on the projected role of these institutions in the distribution of welfarebased resources in return for socially valuable work or other activities, and (more
problematically) in monitoring the utilisation of those resources by individuals.
Pearson has called for a new form of Aboriginal leadership, which he suggests should
be a ‘pervasive’ concept throughout the layers of governance (Pearson, 2000a: 51-2),
a form of what Wolfe (1989) calls ‘dispersed governance’ (see also Rowse, 1992: 8890).
However, Pearson’s discussion elides the essentially political nature of the new
Aboriginal institutions that he advocates; political not just in the shift in power
relations between Aboriginal groups and the state that the new partnerships would
require, but also in the necessity to establish and sustain new dispersed sources of
authority and power within the Aboriginal polity itself (Rowse, 1992: 90).
Fundamental questions are thus raised, not only for Pearson himself but for those who
are in agreement with his call for a radical rethinking of the current policy framework.
What institutional forms might be drawn from, or developed within, the existing Cape
York Aboriginal polity in order to authorise change? What would be the sources of
322
their moral and political authority? How might decisions based on assessments of the
collective good be implemented in political cultures with a characteristically strong
emphasis on personal autonomy and individual rights? These are, it is suggested,
significant issues, for the kinds of structural and attitudinal changes which Pearson
seeks will not be achieved merely by persuasion and negotiation. More broadly, the
issue of moral suasion, or perhaps even coercion, in instituting cultural and social
change, even if arguably for the common good, also raises the matter of basic
citizenship entitlements in a modern democratic state, an issue to which Cape York
Aboriginal people are highly attuned.
While government may not have the moral authority with Aboriginal people to effect
change, as Pearson suggests, it is arguable that it does have a moral responsibility to
ensure that principles of social justice, equity, and accountability are adhered to in the
utilisation of the resources it provides to address Aboriginal socioeconomic
disadvantage. This, and the fractured nature of the contemporary Aboriginal polity,
suggest that government may need to be involved as ‘partners’ at a far more intimate
and hands–on level than Pearson envisages, including assisting with the development
of new Aboriginal governance institutions and facilitating capacity-building within
those institutional arrangements (e.g. through supporting ‘social entrepreneurs’
(Pearson, 2000a)).
This would appear to be the direction taken in the interim report of the recent
Fitzgerald Cape York justice study (Fitzgerald, 2001). This recommends inter alia the
establishment of a Cape York Coordination Unit, responsible for the coordination and
efficient performance of all government activities in the communities, organisation of
consultation between government and community representatives and organisations,
the development of action plans around core issues in each community, and the
implementation of those plans. The Unit’s role is envisaged as being more than just
ensuring effective coordination of government services however. Significantly, the
report explicitly recognises that the current capacity of these communities to exercise
self–management is severely compromised, and recommends a very proactive role for
the Unit with communities which combines both detailed support and, in certain areas
(for example, reducing alcohol consumption and improving school attendance),
intervention and the use of a range of incentives and ultimately sanctions.
Yet, it must be said that such support and intimate involvement also bring their own
inherent risks, since whatever their protestations regarding support for selfdetermination, the state and its agents are ultimately incapable of divesting themselves
of their own political and cultural baggage (Cowlishaw, 1998: 145; Macdonald, 2000:
10-12).
8
Conclusion: Leaching the ‘Welfare Poison’ from Welfare
Dependency
Pearson has argued strongly that profound social and political and economic change
are necessary in Cape York Aboriginal communities, and has been willing to confront
difficult and contentious issues. I support his contention that the dysfunction in these
communities is of such a magnitude that maintaining the policy status quo is simply
323
DAVID MARTIN
not an option. I also support his argument that structural and attitudinal changes are
necessarily interlinked, and that policy should be directed at both levels.
However, Pearson’s argument for an essentially mono-causal connection between the
introduction of the welfare system and increasing social dysfunction ignores the
impact of a whole range of factors since colonisation began. The issues raised by
Pearson go to the heart of the relationship between contemporary Aboriginal groups
and the modern state; they do not just concern the impact of the welfare-based cash
economy. The problems faced by Cape York Aboriginal people therefore cannot be
seen as arising just from ‘welfare poison’. Rather, they derive from a ‘toxic cocktail’
of ingredients (Pearson, 2000b), including some that may ultimately originate within
the Aboriginal realm itself.
Pearson is right to call for a reform of Aboriginal governance as a fundamental
component of the wider reforms he seeks, but his focus on welfare as the primary
cause of the problems within Cape York’s Aboriginal communities has led him to
underestimate the significant internal political dimensions of the necessary
institutional changes. It has also led him to underestimate the necessary, although
risky, role of ongoing active partnerships between government and Aboriginal people
in the implementation of social change.
References
Anderson, J.C. (1988), ‘A case study in failure: Kuku-Yalanji and the Lutherans at
Bloomfield River, 1887–1902’, in T. Swain, and D. Bird Rose, eds, Aboriginal
Australians and Christian Missions, Australian Association for the Study of
Religions, Bedford Park, SA.
Chase, A.K. (1980), Which Way Now? Tradition, Continuity and Change in a North
Queensland Aboriginal Community, PhD thesis, University of Queensland,
Brisbane.
Cowlishaw, G. (1998), ‘Erasing culture and race: practising “self-determination”’,
Oceania, 68 (3) 145-69.
Finlayson, J.D. (1991), Don’t Depend on Me: Autonomy and Dependence in an
Aboriginal Community in Northern Queensland, PhD thesis, Australian National
Fitzgerald, T. (2001), Cape York Justice Study: Volume 1, Summary, Conclusions and
Recommendations, Report of a Commission of Inquiry, Brisbane, November,
2001. Published at http:/www.premiers.qld.gov.au/about/community/capeyorkre
port.htm.
Kidd, R. (1997), The Way We Civilise: Aboriginal Affairs—The Untold Story,
University of Queensland Press, St Lucia, Qld.
Koch, T. (1999), ‘Pearson hits welfare poison’, The Courier Mail, Friday 30 April,
1999.
Latham, M. (2001), ‘The third way: an outline’, in A. Giddens, ed., The Global Third
Way Debate, Polity Press, Cambridge, 25-35.
Macdonald, G. (2000), Assimilation by the Front Door? Rethinking the Ideology of
Self-Management policy in Practice, Paper presented at the ‘Assimilation: Then
and Now’ conference, University of Sydney, November.
324
Martin, D.F. (1993a), ‘Aboriginal and non-Aboriginal homicide: “Same but
different”’, in H. Strang and S. Gerull, eds, Homicide: Patterns, Prevention and
Control, Proceedings of conference held 12–14 May 1992, Australian Institute of
Criminology, Canberra.
Martin, D.F. (1993b), Autonomy and Relatedness: An Ethnography of Wik People of
Aurukun, Western Cape York Peninsula, PhD thesis, Australian National
Martin, D.F. (2001), Is Welfare Dependency ‘Welfare Poison’? An Assessment of
Noel Pearson’s Proposals for Aboriginal Welfare Reform, CAEPR Discussion
Paper No. 213, Centre for Aboriginal Economic Policy Research, Australian
National University, Canberra.
Pearson, N. (1999), ‘Positive and negative welfare and Australia’s Indigenous
communities’, Family Matters, 54, 30–5.
Pearson, N. (2000a), Our Right to Take Responsibility, Noel Pearson and Associates,
Cairns.
Pearson, N. (2000b), ‘Misguided policies a toxic cocktail’, The Australian, Tuesday
24 October.
Rowse, T. (1992), Remote Possibilities: The Aboriginal Domain and the
Administrative Imagination, Northern Australia Research Unit, Australian
National University, Darwin.
Wolfe, J. (1989), ‘“That Community Government Mob”: Local Government in Small
Northern Territory Communities’, Northern Australia Research Unit, Australian
National University, Darwin.
Wyvill, L.F. (1990a), Report of the Inquiry into the Death of the Young Man Who
Died at Wujal Wujal on 29 March 1987, Royal Commission into Aboriginal
Deaths in Custody, AGPS, Canberra.
Wyvill, L.F. (1990b), Report of the Inquiry into the Death of the Young Man Who
Died at Aurukun on 11 April 1987, Royal Commission into Aboriginal Deaths in
Custody, AGPS, Canberra.
325
Volunteers and Community Legal Centres:
A Partnership Under Threat
Rose Melville 1
Sociology Program, University of Wollongong
1
Introduction
This paper examines the notion of partnership, including definitions and some of the
rhetoric surrounding the term. Western governments, such as Australia, Britain and
Canada have enthusiastically embraced the notion of partnership in recent times. It
became quite fashionable to bandy around the term partnership as a means of
describing a new era in relationships between governments, business and the
community sector. Of course, these groups have very different ideas about what this
term means and their practical experiences of it. They range from cynicism about the
extent to which funding bodies will engage in a real or mature partnership to an
attitude of genuine engagement with business or government, where all parties feel
they have an equal stake and input into the process. There are challenges, opportunities
and problems associated with the way in which partnership can operate in the
community sector, especially when there is a dependent ‘funding relationship’
involved in the equation.
The second part of the paper outlines a case study of a successful partnership between
the legal profession, Law Schools, professional legal bodies and community legal
centres. The main impetus for the establishment of this partnership came from within
the legal profession, training institutions and professional bodies. One of the main
outcomes has been the continued supply of legal and para-legal volunteers to help staff
small community based legal centres, which are under-funded and short-staffed. The
unique partnership has facilitated the provision of legal assistance to low-income
Australians for over thirty years. There are benefits and costs involved in this
partnership for all parties, but on the whole it has been able to survive because of the
highly motivated and skilled volunteers who provide their services to community legal
centres (hereafter CLCs). However, this unique partnership is perceived to be under
serious threat because of the foreshadowed introduction of competitive tendering and
contracting.
1
Author’s note: This paper is derived from a report entitled `My Time is not a Gift for
Government’: A Study of NSW Community Legal Centre Volunteers (Melville,
forthcoming 2001). The author wishes to acknowledge and thank the following people.
The volunteers who participated in the study, Myfanwy McDonald who redesigned the
questionnaire, Roberta Perkins who entered and analysed the quantitative data, and Ros
Batten whom assisted with the literature review.
Melville, R. (2002), ‘Volunteers and community legal centres: a partnership under
threat’, T. Eardley and B. Bradbury, eds, Competing Visions: Refereed Proceedings of
the National Social Policy Conference 2001, SPRC Report 1/02, Social Policy
ROSE MELVILLE
2
Partner – Partnership – What Do We Mean?
If we look at the Oxford dictionary, we find that there are a number of meanings
attached to this term, but they all have several things in common. They include an
assumed equality or implicit value in what each brings to the partnership or
relationship, and also that both parties are perceived to share the hazards and benefits
in the relationship.
Sharer (with person, in or of thing); person associated with
others in business of which he [sic] shares risks and profits; a
player associated with another in a game and scoring jointly
with him [sic]; state of being a partner, joint business, associate
[persons, one with another] as partners, be partners of (Sykes,
1976: 804).
It probably goes without saying that values of trust and reciprocity are also implicit in
the notion of partnership, although not necessarily in the way partners or partnerships
operate in practice.
If we use a thesaurus, and look up the word partnership we come up with two key
words: company (noun) and affiliation (noun). If we examine these further we get the
following terms.
Company
Affiliation
Business
Corporation
Enterprise
Joint venture
Organisation
Association
Collaboration
Companionship
Alliance
Relationship
On the left hand are words often associated with the market, whereas on the righthand side we have words associated with the social sphere or civil society. Whilst I
acknowledge that the latter qualities are not the exclusive province of the social being
or social realm, they do allude to various ways and value bases in which partnership
[s] can be conceptualised and as a consequence, different ways in which partnerships
are acted upon.
3
The Literature on Partnerships and its Relevance to the
Community Sector
A brief review of some of the literature on community sector– business/government
partnerships indicates a wide range of theoretical and values bases, which informs the
discussion. Literature about partnerships comes from the field of public policy,
academic and pedagogical theory, organisational and management theory, non-profit,
community work/community sector literature, as well as numerous case studies of
partnerships between government, the community and business. Before canvassing
some of the major ideas in this literature, let us examine some definitions of the term.
327
VOLUNTEERS AND COMMUNTIY LEGAL CENTRES: A PARTNERSHIP UNDER THREAT
The Voluntary Sector Round Table in Canada (hereafter VSR) provides a good
starting point (1998). This paper cites three recent definitions. (See, for example,
Rodal and Mulder, 1993: 28; Treasury Board, Canada, 1995b; New Economy
Development Group, 1996: 16, cited in VSR, 1998: 1). 2
The VSR document identifies four common elements, which differentiate and describe
partnerships including:
•
common objectives and goals among partners (objectives may be the impetus
of the partner or may evolve over time);
•
shared risks and mutual benefits (risks and benefits may be different for each
partner and may accrue with different time frames;
•
contributions from both partners (including monetary and non-monetary);
•
shared authority; responsibility and accountability (VSR, 1998: 1).
According to this document, the term partnership also implies that a lot more can be
achieved by working together – thus there is an added value to the relationship. They
argue that the term is not new – but in the new policy environment its meaning now
describes the collaborative work of multi-agencies with the general public,
government departments, the private and business sector. The rhetoric of partnership
has been extended to try to entice the voluntary (community) sector into relationships
with government and/or business (VSR, 1998:1). 3
A Typology of Partnerships
One common way to analyse or describe the way partnerships operate is through
classificatory systems (VSR, 1998: 1), such as that of Torjiman (1998). This system is
based on notions of social change practices and activities.
• Public education partnerships or 'strategic alliances'
• Social marketing partnerships
• Community investment partnerships
• Social change partnerships (Torjiman, 1998: 6-8, cited in VSR, 1998: 1)
2
3
An arrangement between two or more parties who have agreed to work collaboratively
towards shared and/or compatible objectives and in which there is shared authority
and responsibility; joint investment of resources, shared liability or risk taking; and
ideally, mutual benefits’ (Rodal & Mulder, 1993:28 cited in VSR, 1998:1). ‘ a
relationship in which government and other agents work co-operatively to achieve a
… goal at the community level. It requires the sharing of resources, responsibilities,
decision-making, risks and benefits, according to a mutually agreed –upon formal or
informal arrangement’ (New Economy Development Group, 1996:16 cited in VSR,
1998: 1).
Many community sector organisations initiate partnerships – they are not merely
passive recipients or unequal partners in any joint enterprise.
328
ROSE MELVILLE
Power, and the way in which it is shared, is a central dynamic in one model (VSR,
1998: 1).
They may range from consultative partnerships where the
primary purpose is for government to seek advice or obtain
input to collaborative partnerships or real partnerships, where
there is joint decision-making, pooling of resources and
sharing of ownership and risk (Rodal and Mulder, 1993, cited
in VSR, 1998: 1).
The VSR (1998: 1) document lists three other models:
•
the kinds of partners involved in the relationship - for example government, nonprofit organisations, educational institutions and private sector (Klein, 1992);
•
the aim or focus of the partnership (Long and Arnold, 1995); and
•
the configuration, presence or absence of formality, whether it be contractual,
representational or transactional (Environment Canada, 1992).
There are numerous case studies of successful as well as unsuccessful partnership
ventures, especially between academic institutions and community groups. There is a
very interesting Canadian case study of the successful establishment in 1963-64 of an
Institute of Criminology outside a Faculty structure at the University of Toronto. This
decision was a major risk – but it has meant that strategies had to be developed so that
the different and possibly competing internal and external partners could be catered for
in a unique and productive way. It has not been an easy process. The institute has gone
through a long struggle to obtain ongoing and secure funding as well as to maintain an
independent research agenda while undertaking contract work. Through a great deal of
hard work, they have been successful at securing funding from a number of sources
including private foundations and government bodies, such as the Attorney General's
Department of Ottawa (Edwards, 1985).
In a recently published US anthology, Chibucos and Lerner (1999) have compiled a
series of success stories about university-community partnerships, which were
established to serve children and their families. They note that much of the success is
due to the commitment and passion of the participants, especially those working
within universities. Often faculty staff has to fight a culture that works against the
development of successful external partnerships. To quote Vortuba (1999: 27):
In the process of building partnerships they have had to
challenge traditional views of scholarship and the role of the
university in society. At times, they have even placed their
careers at risk because of a lack of both understanding and
support for their work among department chairs, deans and
promotions and tenure committees. The result is that, far too
often, the building of community-university partnerships takes
place at the periphery rather than at the centre of academic life.
Vortuba (1999: 28-29) identifies six factors, which he argues help to both create and
sustain successful university-community partnerships. The main ingredient is what he
329
calls the scholarship of engagement, a term originally used by Boyd (1996), between
universities and communities.
Before going any further, I should clarify that I am going to focus on community
sector (or nonprofit) and government partnerships or relationships. This is a narrower
and more specific focus. There are three reasons for this stance. First, these
relationships have been in existence for a long time. Second, we desperately need a
better understanding of how to make them work in a more constructive and mutually
interdependent manner than they do at present. Third, the state and the community
sector are inextricably linked in the delivery of services and policy development. If
anything, the role of the community sector will continue to increase in the foreseeable
future.
Community Sector Partnership – Relationship
To those familiar with the community sector, the term partnership is not new, and
unfortunately it has been both abused and used over time, which makes some rather
sceptical about the re-working of it at this point in time. However, not everyone feels
cynical about the possibilities this poses for the community and voluntary sector. The
VSR document provides a very balanced viewpoint about the opportunities and
potential costs involved in forming closer partnerships with government and business.
These include the following:
•
the general public and the community sector want to be involved in policy
making as well as service delivery;
•
the nature of the relationship between funded agencies and government
funding bodies makes an equal partnership impossible because of the power
imbalance between the agencies;
•
there is enormous potential for the community sector to seize the moment and
show it has skills and knowledge that make them an indispensable asset to
government in the new millennium (VSR, 1998: 2).
The Canadian Voluntary Sector claims that the skills which the sector can bring to the
table include 'expertise, links to the community, speed and flexibility and
commitment‘ (VSR, 1998:3).4 Similar claims about the strengths of the sector were
made prior to and during the formation of the Blair Government’s Compact with the
British Voluntary Sector (1998).
Opinions vary considerably about the ability of government to seriously take up the
challenge to involve the community] sector in a truly valued partnership with
government. According to Pascal (1996) the term government partnership, is an
oxymoron in light of the difficulty government frequently has in sharing power and
decision-making. More frequently voluntary sector partnerships and government
bodies undertake relationships involving joint programs and constituents rather than
partnerships.
4
Whilst sympathetic to these claims I would say that there is very little empirical
evidence to support these assertions.
330
ROSE MELVILLE
The evidence for this opinion lies in the problematic and thorny nature of the funding
relationship between governments and agencies. These problems are well rehearsed
and tend to echo around the sector regardless of the country or policy context. They
include the loss of autonomy and ability to undertake advocacy, increasing
bureaucratisation and accountability requirements, disillusionment of volunteers,
mainstreaming of services and narrowing of client groups. (See Dow, 1997 cited in
VSR, 1998: 2; Leat, 1996; Baldock, 1990; Foreman-Nowland, 1998; Kloosterboer,
2001 and Melville, 2001).
A British writer, Marilyn Taylor (1997,1998) has made an attempt to test empirically
how both parties understand the real working of the rhetoric of partnership. She
provides some astute observations about what makes government with community
partnerships successful.
Effective partnership is not easy. It requires clear allocation of
responsibility within partner organisations, with resources,
time and incentive structures for partnerships working.
Partners need to be prepared to change their cultures and ways
of operating to accommodate voluntary sector, community and
user participation (Taylor, 1997: 1).
Taylor goes on to say that:
…the most effective partnerships have been those where there
has been a long tradition of local organising. This gives people
the skills, experience, confidence and infrastructure to engage
on their own terms and to gear up to new opportunities, from
community care planning and the contract culture to the Single
Regeneration Budget and the Lottery (Taylor, 1997: 3).
According to Taylor’s research successful partnership with the community sector
involves seven elements:
• precise objectives, which involve voluntary and community organisations in policy
and service delivery planning. Outcomes which include measurement of the
quality of involvement and not just counting numbers;
• management and organisational commitment: for instance, front-line staff must
know they will be supported if they are involved in risk taking or when adopting
new ways of working;
• acceptance of organisational responsibility to develop and service partnerships;
this includes dedicated time and resources;
• evaluation, benchmarking and review mechanisms must provide incentives for,
and reward, partnership;
• voluntary sector partners must be adequately resourced and trained so they can
effectively participate in making key decisions;
331
• mechanisms for involvement, which accommodate the heavy demands made on
organisations with limited resources and time use; and
• the ability to work with and respect the different cultures, values and resource
capacities of voluntary agencies (Taylor, 1997: 7).
Despite this recipe for successful partnerships with the voluntary sector and local
government, Taylor (1997, 1998) among others, remains somewhat pessimistic about
the ability of government to work in a meaningful partnership with the sector. One of
the major findings of her research is that governments, and especially local authorities,
find it exceptionally difficult to work in an equal and consultative manner with
individuals, community members and voluntary sector agencies. They will not share
power, resources, knowledge and skills in a way that enables the outsider groups to
make a significant impact on policy or decision-making. Many people were very
sceptical about the whole idea and process of consultation (Taylor, 1998: 5-10). They
considered it a useless process (1998). According to Taylor (1997: 8) a great deal more
work has to go in to developing real meaningful partnerships before the sector really
feels involved as equal partners. Of course, Taylor (1998) is not alone in this
assessment. Many others have noted the same problems in other countries, including
Australia (Melville, 2001).
The next section of this paper present a case study of a successful partnership between
public and private sector law professionals working with community based legal
centres. The partnership has been mutually beneficial to both parties for a long period
of time. But the partnership or relationship is also very delicate and finely balanced
and is in danger of being disrupted by particular policy decisions, namely the
introduction of competitive tendering and contracting.
4
CLCs - A Partnership with the Legal Profession and Volunteers
Apart from the work of Chesterman (1996), there is little formal documentation about
the establishment of community legal centres, hereafter CLCs. The most extensive
account documents the history of the Fiztroy Legal Service (Chesterman, 1996).
Noone (1992, 1997) has written about the rationale underlying the unique
characteristics of legal centres. These characteristics make CLCs different to the legal
services previously provided, or not provided, to low- income Australians. From these
and other potted histories, it would appear that CLCs came into existence because of a
strong alliance and commitment from individual legal professionals, state professional
bodies, and various academics in law schools. A motivating factor was the desire of
law professionals and students to gain relevant work experience in poverty and family
law, in the context of social justice. This kind of expertise is not obtained through the
normal course of work in many private legal firms.
Volunteers have played a central role in CLCs since their inception. Like many
organisations of this era, community legal centres were established and operated by
volunteers until granted government funding (Noone, 1992: 121). Volunteers have
continued to play an important role in supplementing and supporting the work of paid
staff in community legal centres. Volunteers in CLCs work in three main areas. They
provide legal advice and information/referral assistance to clients. They undertake
332
ROSE MELVILLE
policy-related and law-reform work. They are members of management committees,
which oversee the administrative and governance functions of centres.
Over the past three decades, these innovative, community-based organisations have
grown into an integral part of the Australian legal system. They fill an important gap
between private legal professionals and the state Legal Aid Commissions. This is
partly because of the clientele they serve. Most clients seen by community legal
centres are neither eligible nor able to access other legal services because of financial
or language and cultural barriers. According to Jukes and Spencer (1998: 5), the legal
profession has come to accept this hybrid model of service provision. There is no
doubt that CLCs should be seen as partners in the overall provision of legal services.
Such a partnership remains precarious, however, because of the continued reliance on
external government funding.
In most Australian states, funding of CLCs is a joint responsibility between state and
federal governments (Noone, 1997: 27). Community legal centres were often
established on minimal amounts of funding. Consequently, funding has always been
precarious and inadequate for their operation (Noone, 1997; Williams; 1999). The
most recent threat to this funding is a Commonwealth Government proposal to move
to competitive tendering and contracting. CLCs are concerned that such a move will
place severe strain on the partnership between volunteers and community legal centres
because of the opposition of these policies amongst volunteers.
5
The Research Project
This paper reports on some of the important findings of a study of NSW CLC
volunteers conducted during 2000 (n=208). The study was partly instigated in
response to a proposed Commonwealth Government review of NSW CLCs, which
was due to occur in 2000. It provided an opportunity for CLCs and interested parties,
such as the NSW Law and Justice Foundation and the NSW Law Society to find out
about the role and motivations of CLCs volunteers, as well as to explore their attitudes
about the proposal to introduce competitive tendering and contracting into CLCs.
One of the main aims of the research was to try to gauge the impact of such policies on
the retention rate of CLC volunteers. Overseas evidence indicates that the introduction
of these policies has a detrimental impact on both the recruitment and retention rates,
as well as the types of tasks undertaken by volunteers (Scott and Russell, 1997; Hedley
and Davis-Smith, 1994).
The study was conducted in two phases. The first phase consisted of an extensive
literature on volunteers in general, and more specifically on volunteers and
competitive tendering. The second phase involved the distribution of a questionnaire
to volunteers in 36 CLCs in NSW (from unknown clusters of volunteers) during
February 2000. The survey instrument contains open and closed-ended questions about
the activities and motivations of volunteers, as well as their attitudes to competitive
tendering. Volunteers at 24 of the 36 centres completed and returned questionnaires,
representing a response rate of 66 per cent.5 For the purposes of this study, voluntary
work was defined as something done 'freely, without remuneration and of benefit to
5
The response rate of volunteers is a little bit more problematic – see full report for
discussion of these methodological issues.
333
the wider community’ (Noble, 2000: 156)6. The quantitative data was coded and
entered into File maker Pro V 3. The qualitative data was coded thematically. This
paper discusses data relating only to the major issues raised by volunteers about
contracting and tendering.
6
Discussion of Findings
This discussion focuses on issues that concern possible changes to the accountability
requirements and the introduction of competitive tendering policies on volunteers in
CLCs. The quantitative and qualitative data provide us with information about the
impact of significant changes to the location, activities, structure, funding and
philosophy of CLCs on volunteer retention rates.
Move to Another Centre
We asked volunteers if they would be prepared to go and volunteer at another legal
centre if their legal centre closed down or moved elsewhere. Eighty-one (40.7 per
cent) people said yes, they would volunteer at another centre, thirty-nine (19.6 per
cent) said no, they would not volunteer at another centre. A surprisingly large number
of volunteers (39.7 per cent) had not decided what they would do if their centre closed
or moved elsewhere (see Table 1).
Table 1 Volunteer at Another Centre?
Yes
No
Undecided
Total
Number
81
39
79
206
Percentage
40.7
19.6
39.7
100
Of the thirty-nine (19.6 per cent) of the respondents who said they would not volunteer
at another centre, thirty-one of these people provided reasons for their negative
response. Twenty-four people said the main reasons were distance and time
constraints. A small group of respondents (seven) said that their skills were too
specialised, that the current centres pursued social justice goals of interest to them for
example, women’s issues, or they liked particular qualities of the centre.
Distance Volunteers Will Travel to Another Centre
The majority of people (86 per cent) were prepared to travel between 15-45 minutes to
volunteer at a centre, while only 12 per cent were prepared to travel over 45 to 75
minutes to do voluntary work in a CLC. Seventy-seven percent of people preferred to
travel 30-45 minute to do volunteer work at a legal centre. This is consistent with the
findings of earlier questions about distance. Distance is clearly a major factor
influencing people's decisions to volunteer. One could hypothesise that the number of
current volunteers would drop off substantially if they had to travel greater distances
than they currently do to volunteer at a legal centre.
6
The term `volunteer’ and not pro-bono work was used to describe an unpaid worker at
CLCs. Pro-bono work is also done by legal firms of behalf of CLC clients but the
legal staff are doing it as part of their paid work, even though the firm may not receive
payment for the case.
334
ROSE MELVILLE
Volunteer at Centre if Nature and Function Changes
Nearly two-thirds of the volunteers (61.2 per cent) indicated that they were undecided
about whether they would continue to do voluntary work if the nature and/or function
of their centre changed. Nearly 23 per cent said they would continue to volunteer
despite any changes and 15.9 per cent said they would no longer volunteer at the
centre. When one considers the size of the sample of this study, this represents a
significant number of people who are signalling they would no longer volunteer at
CLCs.
Surprisingly, 52 respondents provided additional explanations to their answer to this
question. The largest group (38 people) said their decision depended on the types of
changes to legal centres, such as changes in philosophy, issues of access and the nature
of services provided by CLCs. Some provided quite specific reasons, which included:
the loss of focus on consumer credit; lack of skills and expertise; loss of special
qualities of legal centres; for instance the focus on disadvantaged in the community
distance; and time.
Table 2: Volunteer at Centre if Nature/Function Changes
Yes
No
Undecided
Total
Number
46
32
123
201
Percentage
22.9
15.9
61.2
100%
Major Policy Changes - Fees, Merit Testing and Centrally Determined Casework
Guidelines
One of the major changes mooted to current community legal centres is the
introduction of fees, means/merit testing and centrally determined casework guidelines
(Attorney-General's Department, 2000). Many community legal centres have argued
that the introduction of these policies would make them similar in character to state
legal aid services. The respondents were asked whether any of these foreshadowed
changes to CLCs would deter them from volunteering. A majority of people indicated
that the introduction of fees for service would deter them from volunteering and a
significant number were opposed to means/merit testing and centralised casework
guidelines, as indicated in Table 3.
It should be noted that 20-30 per cent of respondents had not yet decided whether the
introduction of the above policies would deter them from volunteering. However, if a
significant number of the undecided group were to oppose the introduction of these
policies, there could be a bigger drop-off in volunteers than the above figures suggest.
These policy proposals prompted a large number of people (71) to make additional
comments on the questionnaire. The themes are outlined below.
335
Table 3: Introduction of Fees Means/Merit Testing and Centralised Casework
Yes
%
No
%
Undecided
%
Total
%
Fees for service
100
49.5
62
30.5
40
20
202
100
Means/merit testing
63
31
89
43.8
51
25.2
203
100
Casework
guidelines
77
38.7
62
31.1
60
30.2
199
100
Objections to Introduction of Fees and Means/Testing
As can be seen from the comments below, there were strong objections expressed to
the introduction of fees and mean/merit testing.
I would object very strongly to fees, means test and casework
guidelines (Respondent 41).
If CLCs are to fulfill their aspirations to provide access to
justice, then they should not charge fees and apply means tests
(Respondent 70).
Effect on Autonomy and Unique Character of CLCs.
Many volunteers were concerned about the potential impact of such policies on the
loss of autonomy, the unique characteristics and local community focus of legal
centres. The comments below highlight these concerns.
I would be extremely hesitant about continuing my
involvement with … Centre if its autonomy was to be
undermined and or it fell prey to increasing bureaucratic
intervention (Respondent 79).
These sorts of polices are incompatible with the notion of
community legal centres is all about. ‘Would like fries with
that?’ – corporatising community services is a ridiculous
notion (Respondent 124).
I would not want to see the autonomy and scope of legal
services limited, nor should access be limited for the general
public who need these services. I believe access to justice is an
important right for all, not just the rich (Respondent 128).
I volunteer for a CLC because it is not a legal aid office. If the
CLC is run as a legal aid office I would not be able to
contribute in the same way as I contribute to the CLC
(Respondent 55).
336
ROSE MELVILLE
Increase in Bureaucratic Control
Several people mentioned that it would increase bureaucratic control of centres, and
one person said that neither profit-motivated nor bureaucratic organisations have a role
for volunteers. One person argued that government had an obligation to provide
services to disadvantaged. Several people mentioned that the changes would seriously
diminish access and equity of certain groups to quality legal services.
It would be a shame to see central bureaucratic decision
making as presently exist in LAC brought to bear on
community legal centres (Respondent 15).
If centres make a profit why should they need volunteers?
(Respondent 189).
The charging of feees would be detrimental to the services of
CLCs, and if we were charging fees and charging money, then
why not pay some one to do my job? (Respondent 131).
The data provides important information about the particular qualities of CLCs that
motivate volunteers. Some of the main qualities that draw volunteers to CLCs include
their non-bureaucratic nature, community-based focus, free service and specialised
services, all delivered within social justice goals. The quantitative and qualitative
results of this current study indicate that fewer people will be inspired to continue
volunteering in an organisation, which resembles the kind of institution they work in
on a daily basis. In other words, if there were no major difference between the
different kinds of legal services, then what would be the point?
Influence on Volunteer Retention Rates if CLCs Become Like a Semi-government
Organisation
We asked respondents if they would continue to volunteer if the centre was changed
and run as a semi-government office. About one-third of people (35 per cent) indicated
that they would not stay, another third (32 per cent) said they would continue to
volunteer and about one-third (33 per cent) said they had not made up their minds.
Fifty (50) people provided additional comments to this question. The main themes are
outlined below.
Opposition to Structural Changes to CLCs
Some people said that they would edeterred from volunteering, depending on the kinds
of structural changes that could take place in CLCs.
It would strongly depend on the extent of government intrusion
- the greater the intrusion the less likely I would continue to
volunteer (Respondent 79).
… It would depend on the culture in a semi-government CLC
(Respondent 80).
It would depend on the nature of the changes involved. Would
not volunteer if place became bureaucratic (overtly)
(Respondent 96).
337
Depends on how the changes would affect the nature of its
services. I want to help people rather than subject them to
further bureaucracy (Respondent 128).
Exploitative Labour and Lack of Appreciation of Volunteers
Some volunteers expressed concerns about exploitative labour and lack of appreciation
by semi-government organisations of voluntary labour.
Government must not rely on unpaid work - as it is, people
work in excess of wage value so NO exploitation of willing
committed community members (Respondent 122).
Firstly, I would then not be a volunteer. What I do with my
time, I decide, it is not a gift to government. (How many times
has government not understood this - let me count the ways?)
(Respondent 81).
As a semi-government organisation it is more likely that
funding will not be reduced as it is in community based
programs because funding should cover wages [and then you
wouldn't need my voluntary labour] (Respondent 108).
Centrality of Community and Community Control
A recurring theme in the empirical findings of this study is the centrality of
community and community participation to the functioning of community legal
centres. Many respondents see this as a defining characteristic of CLCs.
I would object strongly to such a change. A semi-government
organisation is not a community organisation (Respondent 41).
The centre currently has a high level of community ownership
- this would be lost if the centre changed into a semigovernment agency (Respondent 197).
Compromises to Independence and Criticism of Government Policy
Several volunteers mentioned that increased bureaucratic control would compromise
the ability of CLCs to be both critical and independent of government policies.
The value of the centre is its independence and commitment to
social justice. There is no such thing as a ’semi‘ government
organisation. It is either government affiliated or it is free and
independent (Respondent 124).
Government meddles and destroys worthwhile community
programs (Respondent 53).
I volunteer as a person interested in access and equity
principles and actually having the opportunity to do something
about it, even if it means criticising, for example, the police
(Respondent 107).
338
ROSE MELVILLE
Which government? - if Federal then very hard to convince
clients of confidentiality (Respondent 169).
The independence of a community organisation promotes the
empowerment of the community volunteer (Respondent 195).
Loss of Autonomy and Identity
A number of people were concerned about the loss of autonomy and identity of CLCs
and the affect on the quality of services provided to clients and on certain types of
activities undertaken by CLCs.
This would completely change the ability to give independent
advice (Respondent 165).
… centre needs to remain an autonomous centre run on CLC
principles in order to maintain its integrity and the quality of
its services (Respondent 83).
CLCs need to be independent as possible from government they wouldn't be able to offer an appropriate service if they
were government affiliated in that way (Respondent 84).
CLCs must be independent of government. To do otherwise
would be to compromise the integrity of CLCs both in the
impact on their work on law reform and in its perception
within the wider community (Respondent 70).
Semi-government organisations will compromise the
philosophy of CLCs, in particular law reform (Respondent
106).
Loss of Appeal to Volunteers
It was clear that the proposed changes to CLCs would act as a significant deterrent to a
number of people. Some respondents would no longer find it appealing to do voluntary
work if it became too similar to their current place of employment. People need to be
able to do voluntary work somewhere that is different to their waged employment and
in a place that appeals to them (Pusey, 2000).
What's the point if we can't decide policies and activities
(Respondent 59).
If the standard of service provided and/or number of clients
changed - would not volunteer (Respondent 101).
In my day job I already work for the government, so if my
centre became a semi-government agency there would be little
point [in volunteering] (Respondent 152).
I am an employed solicitor and if the centre became a
government agency. I would see no need to provide a free
service (Respondent 155).
339
Volunteer in a Private or Profit Organisation
We wanted to find out if volunteers would continue to volunteer if community legal
centres were transformed into a privately run or contracted service provider. An
overwhelming majority (118) of volunteers said yes, that it would definitely affect
their motivation to volunteer. Forty-four volunteers said that it would not affect their
motivation to volunteer and 44 were undecided. Fifty-six people provided additional
comments to this question. The themes are discussed below.
Table 4 Volunteer in a Private For-Profit Organisation
Yes – will affect motivation to volunteer
No - will not affect motivation to volunteers
Undecided
Total
Number
118
44
44
206
Percentage
57.4
21.3
21.3
100
Opposition to Privatisation and Profit CLCs
A number of respondents expressed concern and criticism of policies to privatise or
introduce profit run organisations into community legal centres.
Privatisation has destroyed enough (Respondent 94).
Private organisations are more than likely operated on a market
driven philosophy and goals and maximising profit rather than
human/community needs (Respondent 108).
You cannot provide these services for a profit (Respondent
113).
Very bad idea, privatisation of this legal service is not the
answer (Respondent 127).
I have already seen the commitment of service providers
become skewed where private service providers have been
introduced in other areas to the great detriment of consumers
(Respondent 143).
Withdraw Labour from For-Profit CLC
A number of people stated that they would withdraw their labour from a for-profit
organisation. One volunteer said that they would want to be paid if CLC services were
privatised, as this would be contradictory to the philosophy and ideals motivating them
to volunteer. It was also clear that volunteers felt that government should not abuse
volunteer labour.
I will not do voluntary work for anything other than a not-forprofit organisation (Respondent 138).
I cannot think of anything that would put me off more. If I
wanted to be involved in a private situation I would work for a
private law firm (Respondent 124).
340
ROSE MELVILLE
If I wanted to volunteer at a privately run organisation I would
have applied a long time ago, but I prefer CLCs (Respondent
135).
Threat to Unique Character and Quality of CLCs.
Privatisation and profit run organisations are seen by many as a significant threat to
the unique character and quality of community legal centres. Respondents pointed out
that this would make CLCs much more like existing private and state run legal aid
services. In addition, there would be no place for volunteers in such services.
A contracted service provider is not a community legal centre.
It is a law firm and we have plenty of these!! (Respondent 41).
Absolutely, the privatisation of legal centres would be
devastating and absurd. The point of being a public service
[community-based] organisation surely, is that such centres
follow procedural fairness and accountability requirements
(Respondent 56).
It would no longer be a community legal centre would it?
(Respondent 111).
No place for volunteers in such a place - it would lose its
access and equity focus (Respondent 168).
Effect on Financial and Administrative Operation of CLCs
Several people mentioned that such a change would affect the internal financial and
administrative operation of CLCs, as well as the kinds of services they provide.
Centres would change to focus on budgetary performance and run on numbers/results
lines so that casework would dominate centre activities.
Yes, if it meant that organisations looked only at casework and
not broader issues (Respondent 43).
Would be concerned that centres will run on a results/numbers
line rather than focusing on client needs (Respondent 104).
Probably withdraw my services - it may as well be legal aid at
best and a private firm at worst - budgetary performance would
become the essential determinants to action taken (Respondent
17).
Community Focus and Involvement under Threat
Community legal centres were seen as public interest focused, in danger of losing their
autonomy and having to compromise in the sort of environment being foreshadowed.
Community involvement is seen as an essential ingredient of the continued success
and legitimacy of community legal centres.
Community legal centres should belong to the community
(Respondent 128).
341
I do not know whether I would volunteer. It is supposed to be a
community centre (Respondent 178).
Again, the independence from Government or from the process
of competitive tendering ensures community involvement is
regarded as valuable for its own sake (Respondent 195).
Privatisation - A Significant Deterrent to Volunteering
The issue of privatisation and the introduction of profit-operating legal centres came
across as one of the most important for volunteers in the questionnaire. A number of
people emphasised how this would change their willingness to volunteer.
I would never volunteer to work for a private organisation that
is profit-oriented (Respondent 58).
Would question a possible profit motive in comparison to a
service run only to meet community needs (not to make
dollars) (Respondent 73).
I assume it would change the culture for the worse - I'm not
interested in volunteering my time so that someone else can
make a big profit. That's why I don't work for a private law
firm in the first place (Respondent 90).
Privately run equals profit. The two don't fit with objects of
CLCs …. (Respondent 120).
Profit motive might become paramount (Respondent 121).
For-Profit Change - Focus of CLCs.
Some other respondents thought that profit-oriented organisations had enormous
potential to change the focus of CLCs. In addition, some saw the services as being
compromised, in a way that would affect the conduct and activities, making it
pointless to participate in such an organisation.
This would definitely affect my attitude, as private firms are
generally influenced by economic rationalism and are
committed to profit rather than moral issues/values
(Respondent 130).
In private organisations the emphasis becomes profit making
as opposed to provision of service. Ability to work on social
justice issues would be compromised (Respondent 205).
If it is privately run, then it might affect the conduct and
activities provided as opposed to being a community
organisation (Respondent 75).
Volunteering in an Externally Controlled Organisation
Respondents were asked if they would continue to volunteer if decisions influencing
centre policies and guidelines were outside local committees. Seventy-six (37.2 per
342
ROSE MELVILLE
cent) said they would not continue to volunteer if decisions were made externally to
local committees, 28 per cent said they would continue to volunteer and 34.8 per cent
said they were undecided.
Table 5: Volunteering in an Externally Controlled Organisation
Yes
No
Undecided
Total
Number
57
76
71
204
Percentage
28
37.2
34.8
100
Fifty-eight people provided additional comments. The themes identified were similar
to those expressed above. They included the need to maintain local control and
autonomy, that government was out of touch with the local community and already
had sufficient control over the way legal centres operated. Volunteers expressed strong
opinions about the need for community legal centres to be able to respond to local
needs and be accountable to the local community. It is clear that the loss of autonomy
would obviously influence people's decisions to remain as volunteers.
Law Reform and CLC Volunteers
Law reform, social action and test case work have long been a core part of CLCs’
activities. Ninety-two people (46.3 per cent) said they would not continue to volunteer
if centres were not involved in these activities. Sixty-two people (30.5 per cent) said
they would continue to volunteer and forty-seven (23.2 per cent) said they were
undecided. Many volunteers expressed the need to know a lot more about the
proposed changes before they would be in a position to make up their minds.
Forty people provided additional comments about this issue. The proposition evoked
some strong reactions from some people as demonstrated by the following comments.
Some of the themes are outlined below.
Why ever would they do that, it would make them nothing
other than another law office? (Respondent 32).
Law reform, social action is intrinsic in the characteristic of
CLCs. To divorce law reform from CLCs is like taking the
`sausage out of the hotdog;' there is no hot dog at all!
(Respondent 70).
Yes, I would continue to volunteer. But I would be most
disappointed - centre would become a `toothless tiger'
(Respondent 24).
An essential feature of CLCs is the law reform work, etc
(Respondent 28).
It is critical that the people at the point of contact with the
public are involved in this kind of work [law reform].
Otherwise, grassroots experience is only about processing
343
people and not attempting to change the law according to the
needs of the public (Respondent 56).
Without law reform work and social action, CLC work would
just be band-aiding (Respondent 64).
One respondent commented that they did very little of this work in their legal centre,
so it would mean little or no change. However, the majority of respondent's (39)
emphasised the importance of law reform to CLCs. It is seen as an integral part of the
function of legal centres.
Summary of Main Effect of Competitive Tendering and Contracting on
Volunteer Retention Rates
One of the biggest concerns to volunteers was that centres would undergo fundamental
changes in their client focus, accessibility, community and social justice orientation.
They also feared that they would lack skills and expertise in a new voluntary work
environment. Although 23 per cent of people said they would continue to volunteer if
the functions of their legal centre changed, 15.9 per cent said they would not
volunteer. In addition, over 60 per cent said they were still undecided.
When it came to the introduction of fees for service, merit testing and centralised
casework guidelines, the results are mixed. Nearly half of the volunteers (49.5 per
cent) said they would not volunteer if fees were introduced, the majority were less
concerned with means testing (43.8 per cent) and volunteers were evenly split about
the issue of the introduction of centralised casework guidelines. A significant number
of people indicated that they had not formed any definite opinions about the proposed
changes. Yet it became clear in analysing the qualitative data that many people were
opposed to the introduction of these policies. Many of the volunteers in the 'undecided'
category had specific beliefs and expectations of CLCs. This in turn influenced their
motivations in working with them. It was clear from the data that any major
transgression of these expectations and beliefs would deter them from volunteering.
Volunteers were very concerned about possible changes to the auspice and structure of
CLCs. Respondents are evenly split on the question of whether they would continue
to volunteer if the agency became a semi-government organisation. About one-third
said they would stop volunteering, a third said they would continue and a third said
they were undecided. However, when one examines the qualitative data, it is clear that
any attempt to alter the fundamental nature of CLCs would evoke a strong reaction
amongst volunteers. Volunteers are committed to the philosophy and current structure
of CLCs, their community focus, independence from government and their ability to
be critical and to advocate on behalf of people.
Respondents expressed their strongest reactions to the suggestion that CLCs be
privatised. A majority of respondents (57.4 per cent) said they this would deter them
from volunteering. Twenty-one per cent were undecided and 21.3 per cent said they
would continue to volunteer. The finding that 21 per cent would continue to volunteer
in a private for-profit organisation is not surprising given that a lot of voluntary work
occurs in private corporations.
For many volunteers, the opportunity to be involved in law reform, social action and
test casework is a crucial component of CLC work. Nearly half the respondents (46.3
344
ROSE MELVILLE
per cent) said they would not volunteer if these activities were no longer part of CLCs.
Surprisingly, when one examines the activity volunteers are currently involved in, only
a small number of people are actually engaged in this work. However, this finding is
extremely important because it means that one of the strongest attractions for a current
volunteer is the possibility that this work exists. If this opportunity does not exist then
the motivation for some volunteers falters, if not fades completely.
A significant number of people indicated that they would discontinue doing voluntary
work if the fundamental services and nature of CLCs change. Some volunteers
indicated they are adopting a wait and see approach prior to making any firm
decisions, as no policy changes have been put into practice as yet. At the same time,
many in the undecided categories stated that there was a point at which once passed
they would withdraw their labour. If a high proportion of the undecided category
become disillusioned then the attrition rate of volunteers in CLCs could be quite high.
However, it is the group of current volunteers who are indicating that the kinds of
changes that often accompany competitive tendering are unacceptable, and who would
withdraw their labour. This is a major concern.
7
Conclusion
This paper has examined two major issues. The first concerns the notion of partnership
in the literature and in Australian policy formation. For many in the community sector,
the relationship they have with funded agencies is barely a relationship let alone a
partnership of any consequence (Nyland and Melville, 1997; Melville, 1999a). One of
the main reasons for this is the dependent nature of the relationship and the impact of
this power imbalance on the weaker party.
In the second part of this paper, the author has outlined a case study of a successful
partnership between the legal profession and community legal volunteers. The main
beneficiaries of this partnership have been low-income Australians gaining access to
legal services. The arrangement has also benefited legal volunteers who feel they are
contributing to society and helping low-income people, as well as gaining valuable
skills in a specialist area of law – namely poverty and family law. The arrangement
has also benefited the community legal centres that are chronically short of funds and
have few paid staff. This partnership is under threat from proposed changes to the
funding regime of CLC.
Overseas experience indicates that the introduction of competitive and privatisation
policies is done with little thought of the consequences for volunteers. One of the main
outcomes is an increased workload and responsibility for voluntary workers and
management committee members. Overseas evidence demonstrates that this leads to a
significant attrition rate amongst volunteers. It also becomes harder to recruit and
retain a pool of volunteers in a contracting regime (Hedley and Davis-Smith, 1994;
Russell and Scott, 1997). Volunteers in this study expressed some of the same
concerns about the introduction of such policies as those surveyed in the overseas
literature. Given the importance of volunteers in community-based organisations it is
up to policy makers to take heed. They need to ensure that the constructive and
mutually beneficial partnerships that already exist between community service
providers continue. To turn ones back on these concerns will no doubt have serious
consequences for the continued provision of legal aid to low- income Australians.
345
References
Baldock, C. (1990), Volunteers in Welfare, Allen and Unwin, Sydney.
Chesterman, J. (1996), Poverty Law and Social Change - The Story of the Fitzroy
Legal Service, Melbourne University Press.
Chibucos, T. R. and R. M. Lerner (1999), eds, Serving Children and Families through
Community-University Partnerships: Success Stories, Kluwer Academic
Publications, Boston.
Dow, W. (1997), ‘The voluntary sector – trends, challenges and opportunities for the
new millennium’, Volunteer Vancouver, September.
Edwards, J. and J. Li (1985), ‘Directing the development of a university Centre of
Criminology’ in Doob, A. N. and E.L. Greenspan, eds, Perspectives in
Criminal Law: Essays in Honour of John LI Edwards, Canada Law Book
Company, Ontario, 1-35.
Environment Canada (1992), Environment Canada in Transition: Consultations and
Partnerships: Working Together With Canadians, paper prepared for the
Transition Team Steering Committee on Consultations and Partnerships.
Hedley, R. and J. Davis-Smith (1994), Volunteers and the Contract Culture, Voluntary
Action Research, Third Series Paper 1, Volunteer Centre, London.
Jukes, J. and P. Spencer (1998), ‘The future for community justice: buying and selling
justice: the future of CLCs’, Reform, 73, 5-10.
Klein, J.L. (1992), ‘Le Parteneriat: vers une planification flexible du development
local?’, Canadian Journal of Regional Sciences, 15(3), 491-505.
Kloosterboer, D (2001), ‘Can advocacy organisations get government funding yet
retain their independence? A strategic analysis of local government subsidy
arrangements’, www.dirkkloosterboer.demon.nl/teksten/text.htm.
Leat, D. (1996), ‘Are voluntary organisations accountable?’ in D. Billis and M. Harris
(Eds) Voluntary Agencies: Challenges of Organisation and Management,
Macmillan, London.
Long, F. and M. Arnold (1995), The Power of Environmental Partnerships,
Management Institute for Environment and Business, Harcourt Brace, Orlando,
Florida.
Melville, R. (1999) ‘Umbrella organisations in a contracting regime: a comparative
review of Australian, British and American literature and experience’,
International Journal of Not-for-Profit Law, 1(4), (http://www.icnl.org).
Melville, R. (2001), ‘Voice and community-sector peak bodies’, Third Sector Review,
Special Issue, Voice and the Third Sector, 7(2).
Melville, R. (forthcoming), My Time is not a Gift for Government: A Study of NSW
Community Legal Centres, Institute of Social Change and Critical Inquiry, Arts
Faculty, University of Wollongong, Australia.
New Economy Group Inc (1996), A Study of Innovative Examples of Local
Partnerships, Human Resources Development, Ottawa, Canada.
Noone, M. (1992), `Imperatives for community legal centres', Alternative Law
Journal, 17(3), June, 120-124.
Noone, M. (1997), `Mid-life crisis: Australian community legal centres', Alternate
Law Journal, 22(1), February 25-29.
Nowland-Foreman, G. (1998), `Purchase-of-service contracting, voluntary
organisations, and civil society', American Behavioural Scientist, 42(1), 108116.
346
ROSE MELVILLE
Pascal, C. (1996), Speech to the Canadian Centre for Philanthropy, Second Annual
Symposium
Rodal, A. and N. Mulder (1993), `Partnerships, devolution and power-sharing: issues
and implications for management’, Optimum, The Journal of Public Sector
Management, 24(3), Winter.
Scott, D. and L. Russell. (1997), Very Active Citizens? The Impact of Contracts on
Volunteers, University of Manchester, Manchester.
Sykes, J. B. (1976), The Concise Oxford Dictionary, 6th edition, Clarendon Press,
Oxford.
Taylor, M. (1997), The Best of Both Worlds: The Voluntary Sector and Local
Government, Joseph Rowntree Foundation, York, http:www.jrf.org.uk/jrf.html.
Taylor, M. (1998), `Dangerous liaisons: policy influence through partnership’, paper
presented at the Centre for Voluntary Organisations Symposium, Third Sector
Organisations in a Changing Policy Context, September .
Torjman, S. (1998), `Partnerships, the Good and the Bad and the Uncertain’, The
Caledon Institute of Social Policy, June, Ottawa.
Treasury Board (1995), Stretching the Tax Dollar, The Federal Government as
`Partner’: Six Steps to Successful Collaboration, Ministry of Supply and
Services, October, Ottawa.
Voluntary Sector Roundtable (1998), Lessons Learned on Partnerships Final Report,
October, Canada, http://www.vsr-trsb.net/publications/ekosoc98/toc.html
Vortuba, J. C. (1999), Foreword’ in Chibucos, T. R. and R.M. Lerner, eds, Serving
Children and Families Through Community-University Partnerships: Success
Stories, Kluwer Academic Publications, Boston.
347
Australian Unemployment Protection:
Challenges and New Directions
Wayne Vroman1
The Urban Institute, Washington, D.C.
Vera Brusentsev
Gettysburg College, Gettysburg, Pennsylvania.
1
Introduction
This paper examines Australia’s scheme of unemployment protection and makes some
comparisons with unemployment protection in the United States of America. Because
unemployment protection arrangements in the two countries are very distinct, the paper
initially describes two broad systems: unemployment insurance (UI) and unemployment
assistance (UA) as alternative ways to protect workers against the effects of
unemployment. Australia operates a system of unemployment protection that limits
eligibility to unemployed persons and families with low income. Its system of
unemployment assistance (UA) has existed for more than 50 years. The United States has
operated unemployment insurance (UI) since the late 1930s.
The paper focuses on two topics: costs of assistance to unemployed people and labour
market disincentives. Section 2 describes the eligibility conditions and the administration
of the systems. A framework for assessing the costs of unemployment protection is
presented in Section 3 while comparative cost data for selected countries are examined in
Section 4. Section 5 presents some empirical examples. The cost of unemployment
assistance in Australia is reviewed in Section 6. The availability of benefits in Australia
and the US is discussed in Section 7. Section 8 describes disincentives and other
Australian policy changes. Section 9 is concerned specifically with potential interventions
to shorten benefit duration in Australia. Sections 10 and 11 summarise the arguments and
draw conclusions.
Four conclusions are reached. First, even though UA systems base eligibility on family
income, the costs of such systems (per percentage point of the unemployment rate) are not
necessarily lower than the costs of UI systems. Examples of high cost and low cost UI
systems are identified as are examples of high cost and low cost UA systems. For the cost
comparison of main interest here, UA in Australia is considerably more expensive than UI
in the US. Secondly, in comparing Australian and US labour markets over the past four
1
This paper was prepared for the National Social Policy Conference, University of New
South Wales, July 4-6, 2001. Opinions expressed do not necessarily reflect official views of
the Urban Institute or Gettysburg College. Financial support was provided by the Urban
Institute.
Vroman, W. and V. Brusentsev (2002), ‘Australian unemployment protection: challenges and new
directions’, in T. Eardley and B. Bradbury, eds, Competing Visions: Refereed Proceedings of the National
Social Policy Conference 2001, SPRC Report 1/02, Social Policy Research Centre, University of New South
Wales, Sydney 348-386.
WAYNE VROMAN AND VERA BRUSENTSEV
decades, major contrasts are found in the growing disparities in the duration of
unemployment and the duration of unemployment benefits. All duration measures in
Australia are much longer than their US counterparts. Thirdly, while selected policy
interventions could be considered to shorten UA benefit duration in Australia, many
approaches have already been tried with only limited success. Finally, two initiatives might
be worth considering in Australia: undertaking a new system for verifying the
measurement of family income and placing a limit on maximum potential UA benefit
duration.
2
UI and UA: Eligibility and Administration
Unemployment insurance (UI) and unemployment assistance (UA) as systems of
unemployment protection have fundamentally different primary objectives. Payments of
UI benefits are intended to smooth income by replacing a portion of an eligible worker’s
lost wages attributable to unemployment. Payments of UA benefits are intended to
eliminate or reduce poverty among low-income families where unemployment occurs.
Thus while both make payments occasioned by unemployment, UI is paid to eligible
individuals regardless of income while UA is paid only to families with unemployment
whose income and assets fall below designated thresholds.
Contrasts between beneficiaries of UI and UA are sharpest in situations where
unemployment is of short duration. Recipients of UI can have high income since payments
are made to partially offset the earnings losses experienced by the individual regardless of
total family income. Because payments of UA, in contrast, are limited just to families
whose income and assets satisfy a means test, benefit payments are received mainly by
those towards the bottom of the income distribution. Thus the share of UA benefits that
goes towards poverty reduction is generally higher than for UI payments.
This contrast (UI paid to the individual workers, UA paid to low-income families)
becomes less clearcut in situations of long-term unemployment. Because earnings
typically constitute the bulk of family income, long-term unemployment often causes a
large reduction in family income. Thus many of the long-term unemployed who receive
payments under a UI program would also be compensated under a UA program where a
means test is used to determine eligibility. Payments from both programs reduce poverty
in such situations.
From a macro perspective, UI and UA both make cash payments that respond strongly to
cyclical developments. Both undertake activities of payments administration, e.g.
decisions about eligibility and payment levels, and these activities are often similar.
Internationally, UI is far more common than UA. To our knowledge UA systems are
present in just four countries (Australia, New Zealand, Hong Kong and Estonia) while UI
programs exist in more than sixty countries.2
A brief comment about the classification of unemployment protection systems may be
appropriate. In several counties UI and UA benefit payments are both present with UI
2
Counts are based on US Social Security Administration (1999).
349
AUSTRALIAN UNEMPLOYMENT PROTECTION: CHALLENGES AND NEW DIRECTIONS
available first and UA then available for UI exhaustees and/or UA is available for those
who do not qualify for UI. When both are present, periodic UA payments are typically
lower than UI payments. Some countries also have a third tier of protection for the
unemployed, an income-conditioned social assistance (or general assistance) program with
benefits payable after UA entitlements have been exhausted. Gornick (1999) describes the
types of unemployment protections offered in OECD countries. Section 1 of Schmid and
Reissert (1996) also provides a summary of UI and UA classification issues.
For present purposes the various ‘mixed’ systems where UI is the initial port of entry for
claimants are treated as UI systems. The cost comparisons in Section 4 are restricted to a
comparison of UI (including mixed systems) with stand-alone UA as an alternative
program for the unemployed.
Table 1 summarises UI and UA activities in two broad areas of benefits administration:
initial entry and continuing eligibility. For both areas, the table lists the requirements the
claimant must satisfy and the decisions (determinations) made by program administrators.
The table compares stylised UI and UA programs. If actual countries were identified, a
more varied picture would be observed. Specific eligibility requirements and
administrative activities are identified with Xs. Several rows have Xs for both
unemployment protections. Key differences between UI and UA are identified in the rows
where only a single X is present.
Both protections make payments for partial unemployment as well as total
unemployment,3 and both specify an explicit waiting period between filing for benefits
and receiving an initial payment. However, UI typically requires the claimant to have
substantial previous work experience (signaled by a required threshold level of previous
earnings [USA], weeks worked [Germany] or hours worked [Canada]) whereas UA can
compensate those with little or no previous work experience. The reason for the job
separation is important for UI eligibility while UA eligibility determinations focus heavily
on family income and assets. Both programs make yes-no decisions about initial eligibility
and the level of the periodic payment, but only UI specifies potential benefit duration at
the time of initial entry into benefit status. A UA program may or may not limit potential
duration. Potential duration is unlimited in Australia, but in Germany UA duration for
many (all but UI exhaustees) is limited.
To continue in benefit status, the claimant must be able to work and be available for work.
Increasingly, countries are requiring claimants to provide evidence of active work search
and/or other socially beneficial activities. The latter requirement has various names, e.g.
activation, reciprocal obligation or mutual obligation. While country practices regarding
activation vary widely, merely waiting until a ‘suitable’ job is offered is generally
becoming less acceptable for maintaining continuing benefit eligibility. Enforcing work
search requirements, judging the suitability of job offers and monitoring job refusals are
administrative tasks common to UI and UA.
3
Partial unemployment is usually more prevalent in UA programs but some paid work while
in receipt of benefits is permitted by both UI and UA.
350
Table 1: Eligibility for UI and UA Benefits
UI
Programs
UA
Programs
X
X
X
X
X
X
X
X
X
X
Initial Entry Eligibility
Requirements on the claimant
Total or substantial unemployment
Substantial prior work experience
Acceptable reason for separation
Serve a waiting period
Low family income (and assets)
Administrative agency determinations
Initial Entitlement, Yes-No
Level of periodic payment
Maximum potential duration
Family Income Assessment
X
X
X
Continuing
Eligibility
Requirements on the claimant
Able to work
Available to work
Active work search
Low family income (and assets)
Administrative agency oversight
Work search
Suitable job offers
Disqualifying and/or deductible
labor income, e.g. pensions
Family income monitoring
X
X
X
X
X
X
X
X
X
X
X
X
X
Both programs also monitor the receipt of other income that may reduce entitlements. In
UI, the other income is typically linked to previous work, e.g. severance pay and pension
benefits. In contrast, all of family income is considered in administering the means test for
UA. Monitoring family income is not simple, especially if a spouse works. If income
monitoring is effective, changes in the spouse’s earnings will alter the UA payment.
As noted, many countries offer both UI and UA with the latter available either after the
claimant exhausts UI or for those not eligible for UI at the onset of unemployment.4 In
4.
See, for example, Charts 4.1-4.5 in Chapter 4 of OECD (1998). Tables 1 and 2 in Gornick
(1999) show that of the 18 OECD countries with UI programs as of the mid-1990s, nine also
351
these situations, the contrasts between UI and UA are smaller than suggested by Table 1.
In effect, UI acts as a screen for some workers who later move into UA benefit status, but
the transition to UA is made by only some of the long-term unemployed, i.e. those with
low income. Typically those who move from UI to UA are paid a lower periodic benefit
while on UA.
A priori, UI and UA would be expected to have contrasting patterns of administrative
costs. Of the two, UI pays more attention to the claimant’s work history and to the
circumstances of the job separation, since entitlement presumes a lengthy period of prior
employment and an acceptable reason for the job separation.5 UA, on the other hand,
focuses more on the current fact of unemployment and whether or not the claimant
satisfies the means test. While UI will review certain types of income for possible offsets
against UI payments (severance pay and pension benefits), UA has to make a complete
assessment of family income. Also, changes in income, e.g. the earnings of a spouse, need
to be monitored to verify continuing UA eligibility. Both have to monitor job search and
work availability as conditions for continuing eligibility.
Of the two systems, administrative costs would usually be higher under UA because of the
costs of monitoring income (initial income assessments for new claims and income
monitoring for ongoing claims). These costs would typically exceed the costs of UI initial
eligibility determinations which are one-time costs per claim. The costs of monitoring
availability and work search are likely to be similar in the two systems. While the
administrative costs of UA are likely to be higher, we have not tried to assemble
comparative cost data to provide empirical support for this inference.
3
Disincentive Issues
The two forms of unemployment protection generate problems of labor market
disincentives. However, the disincentive problems in the two systems are different.
Unemployment Insurance
For unemployment insurance three disincentives can be identified. First, there are entry
incentive effects. When the work histories of recipients are studied, a bunching of
claimants who satisfy minimum eligibility requirement is often found. Prior to 1997,
Canada based eligibility on previous weeks of employment. Each year a consistent
bunching at the minimum weeks threshold was observed. Now that Canada uses hours
worked in determining eligibility (from 420 to 700 depending on provincial
unemployment), a bunching at the minimum hours threshold has been observed.
Second, there may be high replacement rates, i.e. high ratios of weekly benefits to weekly
earnings. High replacement rates encourage longer spells in benefit status. Estimates of the
had UA. The nine were Austria, Finland, France, Germany, Ireland, the Netherlands, Spain,
Sweden, and the United Kingdom.
5
If a claimant quits a job there will usually be both an investigation of the reason for the quit
and assessment of a quit penalty that precludes receipt of benefits for a number of weeks.
352
size of replacement rate effects differ, but the direction of the effect is clear. As claimants
suffer a smaller income loss from unemployment (higher replacement rate), they prolong
periods of recipiency.
At least three factors that contribute to high replacement rates can be singled out.
•
Progressive benefit formulas provide for higher replacement among workers paid
low wages.6
•
Workers subject to high marginal income tax rates often experience high net wage
loss replacement.
•
Paying dependants’ allowances increases replacement rates. A compounding of
these factors occurs among secondary workers with children who are members of
high-income families.
Third, long potential benefit duration can contribute to increases in actual benefit duration.
While empirical estimates vary, each added week of potential duration adds from 0.1 to
0.2 weeks to actual duration (studies from the United States).7
Unemployment duration has many determinants besides UI potential benefit duration. An
increased pace of dislocation and permanent job loss associated with globalisation has
probably played a role in increased unemployment duration in the US. Several measures
show that average unemployment duration increased in the 1980s and 1990s relative to
earlier decades. Increases are observed in both household labour force survey data and in
data from the UI program.8
The increase in unemployment duration in the US during the 1980s and 1990s has
occurred in a period when UI benefit generosity has, if anything, declined. Average
replacement rates are now somewhat lower than in the late 1970s (details vary from state
to state) while potential benefit duration has not increased. These time series patterns
suggest that developments in unemployment duration in the US have not been driven by
developments in UI statutory provisions.
An important component of increased unemployment in Western Europe since the early
1970s has been a lengthening of unemployment duration. Several studies have examined
the linkage between increased duration and the provisions of the UI systems (and other
aspects of employment security) in these countries. Since long UI potential duration
6
7
8
In California, the weekly benefit for beneficiaries at the minimum is 65 per cent of the
average weekly wage while for high wage workers the replacement rate is 39 per cent.
Summaries of the US empirical literature are given in Woodbury and Rubin (1997) and
Vroman and Woodbury (1996). Katz and Meyer (1990) provide estimates of the effects of
potential on actual duration. Atkinson and Micklewright (1991) and Ham, Svejnar and
Terrell (1998) summarise international evidence. See also the analysis of determining social
assistance support levels in Chapter 3 in OECD (1998).
See Chapter IV of Vroman (2001a) for a summary of trends in unemployment duration in
the US from 1950 through 1999. Decade averages of US duration data also appear in Table
3 of the present paper.
353
usually predated the increase in unemployment of the mid-1970s, the linkage between UI
provisions and increased duration is not transparent. A recent investigation by Blanchard
and Wolfers (2000) argues that an interaction between institutions, e.g. unemployment
protection provisions such as the replacement rate and the maximum potential benefit
duration, and macroeconomic shocks have combined to produce the higher unemployment
and lengthened duration observed in many countries since the mid-1970s. Many
researchers have found an effect of UI provisions on unemployment duration and
additional research on the linkage can be anticipated.
Unemployment Assistance
Because UA conditions eligibility on the family income of the unemployed individual, the
static labour supply-income framework provides a useful point of departure for a
discussion of disincentives. For a given family member, family income is given by:
(II-1) Y = X + W*H where,
Y = family income,
X = income from assets plus the earnings of all other family members,
W = the person’s wage rate and
H = hours of work.
When family income falls below Y*, the income guarantee, UA is paid to the family.9
Five aspects of the payment are noteworthy.
9
•
The guarantee (Y*) may depend (positively) on family size. Thus
for a given level of Y, larger families receive larger payments.
•
Ceteris paribus, a low wage rate will cause the UA payment to be
larger. Low-wage workers with both hours of unemployment and
hours of employment, could receive a UA payment even with
substantial hours worked.
•
Ceteris paribus, low hours worked will cause UA payments to be
larger. The second and third points taken together imply that a
large share of UA recipients could be working and receiving
payments simultaneously.
•
Ceteris paribus, the largest payments are received by persons
with zero earnings. If the guarantee level is set too high in an
environment where people have substantial control over their
unemployment, UA could encourage a lengthening of
unemployment duration.
The asset test for UA eligibility is not explicitly treated in the present discussion.
354
•
Payment of UA benefits to an unemployed family member can
influence labour supply decisions of other family members.
Wives with unemployed husbands, for example, may be less
likely to work since their earnings could either make the family
totally ineligible for UA or reduce the size of the payment.
Concern over this possible effect on family labour supply
motivated changes in Australia’s UA system in the mid-1990s.
Part of the reason for Australia changing to a more individualised
UA system in 1995 was to encourage work among other persons
in families (often wives) where one member is unemployed.
These changes are discussed in Section 7.
Empirical evidence supports the presumption of a labour supply effect on other family
members. Terrell, Lubyova and Strapec (1996) found that the presence of an unemployed
spouse lowered the hazard rate of exit to employment by 72 per cent for women and 82
per cent for men in an analysis of data from the Czech Republic. Boeri (1997) reports
similar findings in data from Poland.
Brief consideration of these five points suggests that serious disincentive issues could arise
within UA programs. To minimise artificial prolongation of unemployment, the work
search activity of UA recipients needs to be actively monitored.
Another disincentive issue could arise from worker-initiated job turnover, i.e. quits. Quitto-unemployment flows cause family income to decline. Thus the reason for
unemployment may have to be monitored by a UA program and entitlement limited to
‘acceptable’ reasons for unemployment.
Youth unemployment may also present problems for a UA program. If new workers can
collect UA without demonstrating a substantial job history, some youth might appear as
unemployed for purposes of collecting UA when they are not seriously searching for work
or engaged in training. Again, this would present a monitoring problem for UA program
administrators.
Because UA programs occur with much less frequency than UI programs, there has been
less research on disincentive problems in UA. However, another body of literature is
relevant, analyses of the work disincentives of welfare programs. That research has
emphasised high effective marginal tax rates10 and poverty traps as impediments to work
by the welfare population. Recent policy initiatives in the US have made mandatory work
requirements a prominent feature of a ‘reformed’ welfare system. At a minimum,
advocates of UA as a less costly program than UI would have to present cogent responses
to questions about disincentive effects in a program that conditions eligibility upon family
10 High effective marginal tax rates arise from three sources: 1) the phase-out rate of UA
benefits when family income exceeds the maximum allowed for the receipt of full UA
benefits, 2) the marginal payroll tax rate on earnings and 3) the marginal income tax rate for
the recipient’s family income. A review of the US literature, prior to welfare reform of 1996,
is given by Moffitt (1992).
355
income. Section 7 revisits UA disincentive questions in the context of the Australian
system of unemployment protection.
4
The Cost of Unemployment Protection
Payments of unemployment protection benefits can be compared across countries using a
common metric. This paper examines benefit payments measured as a percentage of total
wages. It first derives a framework and then examines costs for selected countries in
Sections 4, 5, and 6
Benefit payments to the unemployed can be expressed as:
(III-1) TBen = AWBen*NBen*52 where
TBen = total annual benefit payments,
AWBen = average weekly benefits,
NBen = the average weekly number of beneficiaries and 52 converts weekly benefit
payments to an annual benefit flow.
The right hand terms in (III-1) can be rewritten equivalently as:
(III-1a) TBen = (RRate*AWW)*((NBen/Unemp)*(LF*URate))*52 where
AWW = the average weekly wage,
RRate = the replacement rate (average weekly benefits as a ratio to AWW),
Unemp = average weekly number unemployed,
LF = the labour force, and
URate = the unemployment rate (unemployment as a proportion of the labour force, also
commonly termed the TUR, shorthand for the total unemployment rate).
Note that the replacement rate in (III-1a) measures benefit payments relative to the
economy-wide average weekly wage. Since the incidence of unemployment is higher
among low-skilled workers, the average weekly wage of beneficiaries will be lower than
the overall average weekly wage. Thus RRate in (III-1a) could be expressed as the
replacement rate for beneficiaries times the ratio of their weekly wage to the overall
weekly wage. In US data, the weekly wage of UI beneficiaries ranges from 80 to 90 per
cent of the overall weekly wage. This alternative representation would have the advantage
of showing an average replacement rate more directly relevant to labour supply decisions
of beneficiaries.
A convenient metric for examining the costs of unemployment benefit protections is
annual wage and salary payments. This can be expressed as:
(III-2) Wages = Emp*AWW*52 where
356
Wages = total annual wages or the wage bill,
Emp = annual average employment, and
AWW = the average weekly wage.
This expression for the annual wage bill can be rewritten as:
(III-2a) Wages = LF*(1 – URate)*AWW*52 where the terms in (III-2a) have already been
introduced.
Dividing (III-1a) by (III-2a) yields an expression for unemployment benefit costs
measured as a fraction of the wage bill:
(III-3) TBen/Wages = RRate*(NBen/Unemp)*URate/(1 – URate).
This benefit cost rate can be expressed as a fraction or as a percentage. In the graphical
exposition of Section 5 below, B (= TBen/Wages) is shown as a percentage.
The left hand side of expression (III-3) is the cost of unemployment benefits expressed as
a fraction (or percentage) of the wage bill. This cost rate has three determinants.
•
The replacement rate.
•
The share of the unemployed who are compensated.
•
The unemployment rate.
The latter is largely a macro-phenomenon that reflects the overall functioning of the
economy. The replacement rate and the share who receive benefits, in contrast, are
influenced by policy choices made by a country. Statutory provisions and administrative
procedures influence both payment levels and the share of the unemployed who receive
benefits.
Up to this point, the discussion of unemployment benefit costs has not distinguished UI
from UA systems. Regardless of the kind of unemployment protection offered by a
country, total payments can be represented as in expression (III-3). Because the expression
is generic, it can be helpful in making comparisons between UI and UA and showing the
cost of each relative to the total wage bill.
One other feature of expression (III-3) should also be pointed out. The ratio
(NBen/Unemp) is a summary measure of benefit availability, but NBen is not nested
within Unemp. Since both UI and UA can make payments to persons with earnings, NBen
is not a subset of Unemp. In the United States, for example, almost 10 per cent of weeks
compensated by the UI program goes to persons with earnings who receive a so-called
partial unemployment benefit. In Australia, nearly one-fifth of UA recipients have
earnings in the same period when benefits are being received, and, as will be seen, NBen
357
has exceeded Unemp in some years. Thus the (NBen/Unemp) ratio is best thought of as a
macro-indicator of benefit availability where some recipients may be employed.11
In providing unemployment protections, a country may make explicit or implicit decisions
regarding the replacement rate and the share of the unemployed to be compensated. The
product of RRate and (NBen/Unemp) determines how costly unemployment protection is
per percentage point in the unemployment rate. This product can be termed a generosity
index (G). i.e. G = RRate*(NBen/Unemp).
Several combinations of RRate and (NBen/Unemp) can combine to yield a given G. For
example, a G of 0.25 can arise when both RRate and (NBen/Unemp) equal 0.50 or when
RRate equals 0.25 while (NBen/Unemp) equals 1.00. Countries have wide choice in
setting the two components that combine to determine G. Thus the UK and the US have
similar levels of G (See Figure 1 in Section 5) but RRate is much higher in the US while
(NBen/Unemp) is much higher in the UK. If a country wanted to make a cost-neutral
change in its unemployment program, this could be accomplished by changing RRate but
modifying (NBen/Unemp) in the opposite direction.
Regardless of the system used to provide unemployment protection, UI or UA, the costs of
benefit payments per percentage point of unemployment can be characterised with G, the
generosity index. Empirical examples from UI and UA systems are explored in Section 5
while the cost of Australia’s UA system is examined in Section 6.
The coefficient G also has macroeconomic significance. It is a gradient that shows how
much the cost of unemployment protections increases when the unemployment rate
changes. Individual countries may select a smaller or larger G depending upon factors
such as affordability and the size of perceived disincentive effects. As will be seen in
Section 5, a wide variety of choices have, in fact, been made.
5
Some Empirical Examples
To make the preceding discussion more concrete, this section displays graphs that
illustrate the costs of unemployment protections. All figures plot benefits as a percentage
of wages against the unemployment rate.
Twelve Countries in 1992
Figure 1 displays data from 12 countries in 1992. The data are derived from a study by
Schmid and Reissert (1996).12 Their analysis combined UI and UA payments in countries
11
The usual convention in labour force surveys is to count people as unemployed only if they
have been looking for work but had no hours worked during the reference week. In other
words, people with both hours worked and hours of unemployment during the reference
week are counted as employed. Persons in these situations are described as underemployed.
Note that underemployment is a broader concept that can also be applied to persons working
full time but at a skill level below that for which they were trained.
358
like France and Germany where both protections are present. The constituent elements of
G, i.e. RRate and (NBen/Unemp), were also examined in their analysis, but are not
emphasised here.
Three factors stand out in Figure 1. First, costs and unemployment rates vary widely
across the twelve countries. Three countries had 1992 cost rates (B%) that exceeded 3.5
per cent of wages while two had cost rates below 0.5 per cent of wages. Second, of the
high-income countries, the US and the UK rank near the bottom in terms of absolute cost
levels (B per cent) and both are low in terms of G, the generosity index. The average
gradient linking B per cent to the unemployment rate across the 12 countries is roughly
0.20-0.25 while the gradients in the US and the UK are closer to 0.10. Third, and probably
most surprising, there is no statistically significant association between the unemployment
rate and B per cent for these 12 countries in 1992. The adjusted R2 in a homogeneous
regression across the 12 is -0.02.13 In 1992 variation in replacement rates and the share of
the unemployed who were compensated were so wide that they overwhelmed the
association that would be expected between the unemployment rate and the cost of
providing unemployment protections.
The variation across countries is startling. Arrays from the origin for Sweden and
Denmark suggest values of G of 0.697 and 0.506 respectively. The corresponding estimate
for Greece is 0.032. The slope of the highest gradient (Sweden) exceeds that of the lowest
gradient (Greece) by a factor of about 22.
The Costs of UI in the United States
The UI program in the US is administered by individual states that determine the key
benefit provisions. While there is federal (national) oversight of state activities, the federal
performance standards relate primarily to the timeliness of administrative determinations
and financial transactions. There are no federal standards affecting benefit provisions such
as the minimum benefit, maximum benefit or the statutory replacement rate. As a
consequence, benefit generosity varies widely across individual states.14
Figure 2 summarises the evolution of UI costs in the US between 1957 and 1999. This
figure vividly illustrates the low level of UI costs in the US compared to most of the
economies depicted in Figure 1. Only three years had a cost rate of 1.2 per cent or higher.
For the full 43 years, the gradient linking benefits as a percentage of wages (B%) to the
unemployment rate as determined by a regression was 0.110. For the two sub-periods
1957-1980 and 1981-1999, however, the slopes were 0.123 and 0.098 respectively.
12
13
14
Their paper expresses unemployment benefit costs as a percentage of GDP. These have been
converted to an estimated percentage of wages by assuming wages represent 70 per cent of
GDP.
The regressions shown at the bottom of all charts follow equation III-3 of the text. TUR is
shorthand for the total unemployment rate. Similar results obtain when the unemployment
rate enters linearly.
Interstate variation in UI costs for the year 1997 is examined in Section IV of Vroman
(2001b).
359
Figure 1: Benefit Costs, Benefit Generosity and Unemployment Rates in Selected
OECD Countries: 1992
B%
5.0
X - Denmark (G = 0.506)
4.5
4.0
X - Sweden (G = 0.697)
X - Belgium (G = 0.344)
3.5
X - Spain (G = 0.185)
3.0
X - Ireland (G = 0.188)
X - France (G = 0.245)
2.5
X - Netherlands (G = 0.350)
X - 12 Country Average (G = 0.263)
2.0
X - Germany (G = 0.292)
1.5
X - U.K. (G = 0.119)
1.0
X - U.S. (G = 0.102)
0.5
X - Portugal (G = 0.091)
X - Greece (G = 0.032)
4
8
12
B% = (RRATE*(NBen/Unemp))*(TUR/(1-TUR))%
Regression
B% = G*(TUR/(1-TUR))%
B% = 0.197*(TUR/(1-TUR))%
Adj. R2 = -0.02
16
TUR%
Max - Sweden B% = 0.697*(TUR/(1-TUR))%
Min - Greece B% = 0.032*(TUR/(1-TUR))%
Source: Schnid and Reissent, 1996.
Per percentage point of unemployment, UI costs in the US are less than half the average
cost shown for the12 countries in Figure 1.
360
Figure 2: Benefit Costs, Benefit Generosity and Unemployment Rates in the US: 1957
to 1999
B%
1.8
1.6
1.4
X
1.2
R
X
1.0
X
R
X
X
0.8
XX
XX
X XR
X RR
X
0.6
X
X
X
R
R
RR
R
R
R
X
XR
R
X
X
R
RR
XR R
0.4
X X
X
0.2
2.0
B% = (RRATE*(NBen/Unemp))*(TUR/(1-TUR))%
B% = G*(TUR/(1-TUR))%
4.0
6.0
1957-1980 Xs
B% = 0.123*(TUR/(1-TUR))%
Adj. R2 = 0.82
8.0
10.0
TUR%
1981-1999 Rs B% = 0.098*(TUR/(1-TUR))%
Adj. R2 = 0.87
Figure 2 distinguishes the data points from the two sub-periods: Xs for the earlier years
(through 1980) and Rs for the later years. The predominance of Rs towards the bottom of
the envelope of data points is apparent. A formal test for equality of coefficients for the
two periods was rejected at the 0.01 level.
The literature on recipiency in the US, e.g. Burtless and Saks (1985), Corson and
Nicholson (1988) and Vroman (1991), has consistently shown a decrease in UI recipiency
in the early 1980s. Policies enacted by several states experiencing financing problems as
361
The Cost of Unemployment Assistance
Four countries compensate the unemployed with an unemployment assistance program.
Figure 3 summarises the cost experiences of Australia, New Zealand, Hong Kong and
Estonia using five year averages of data spanning the period 1975 to 1999. For a given
unemployment rate, these countries display a wide range of benefit costs. Australia and
New Zealand operate reasonably expensive systems while Hong Kong and Estonia operate
very low cost systems. The data in Figure 3 show a clear rank ordering of costs with New
Zealand having the most expensive UA program and Estonia having the least expensive
program.
The regression that summarises UA costs explains about forty per cent of the variation in
the benefit cost rate. Note that the slope in Figure 3 (0.201) is similar to the slope from the
earlier Figure 1 showing UI costs for 12 countries in 1992 (0.197). The slope in Figure 3 is
strongly influenced by one Estonian data point (shown just above the 10 per cent
unemployment rate). Removing this data point causes the regression slope to increase to
0.24, and the adjusted R2 increases to 0.69.
Based on these data, a most interesting finding emerges. Even though UA programs
condition payments with a means test, there is no assurance that UA is less expensive than
UI in providing unemployment protection. Per percentage point of unemployment rate,
Australia and New Zealand experience benefit costs similar to those of European countries
such as France, Germany and the Netherlands.
6
The Cost of Unemployment Assistance in Australia
Australia has administered a program of unemployment assistance (UA) for over 50 years.
Of the countries where UA is the primary program for unemployment protection, Australia
has the largest population. Its UA program has undergone several changes and continues
to be subject to periodic modifications.
Australia provides a full set of social protections through pensions, allowances and other
kinds of support.15 As a rule, pension payments are larger than allowances. Over the past
two decades, age pensions and UA allowances respectively have averaged about 25 per
cent and 20 per cent of the average male wage. Age pensions, the largest of the individual
programs, are received by over 80 per cent of those aged 65 and older. Traditionally,
pensions have been provided as federally supported payments. In the future, public
pensions are to be supplemented by superannuation payments from individual accounts
financed through payroll-based mandatory employer contributions and voluntary
15
Table 1 in Whiteford (2000) lists five kinds of pensions (age, disability support, wife of
pensioner, carer and parenting payments), eight kinds of allowances (newstart [for
unemployment], partner, parenting payment, youth allowance widow, newstart [short term
sickness], mature age and special benefits), two kinds of ‘other programs’ (child care
assistance and public housing) and eleven kinds of family payments, allowances and
supplements.
362
employee contributions. For persons of working age, there are invalidity (permanent
disability) payments, payments for short term sickness and work injuries, mature age (preretirement) allowances, parenting payments, support payments for training and higher
education and rental subsidies.
Figure 3: Benefit Costs, Benefit Generosity and Unemployment Rates in Countries
with UA Programs (Five year averages, 1975-1979 through 1995-1999)
B%
3.0
N
A
2.5
N
A
A
2.0
N
A
1.5
A
1.0
N
0.5
N
H
H
A = Australia
E = Estonia
H
2.0
H = Hong Kong
N = New Zealand
E
4.0
6.0
E
8.0
10.0
Regression. B% = 0.201*(TUR/(1-TUR))%
Adj R2 = 0.39
The philosophy behind the social protection programs is to provide means-tested benefits.
Except for the superannuation scheme (a comparatively recent innovation) and workers’
compensation, the other programs condition payments on the levels of family income and
family assets (exempting family residences that are owned). Because of its heavy reliance
on means testing, Australia is unique within OECD countries in targeting payments to the
363
TUR%
low-income families and individuals. Roughly 60 per cent of cash benefits are paid to
those in the bottom three deciles of the income distribution.16 Also unusual is Australia’s
reliance on general revenues to finance social payments. Typically OECD countries rely
mainly on payroll taxes. Because most payments are income-conditioned in Australia,
issues arise in structuring payments so that work incentives are appropriate. Effective
marginal tax rates are often high for individuals who contemplate working more hours to
increase their earnings and income.17
During the past 40 years certain evolutionary changes have occurred in Australia leading
to heightened concerns about family income disparities, labour market outcomes and the
structure of the social protection system. Five developments have been particularly
noteworthy.
• Among two-parent families, there has been sizeable reduction in
the share with one working adult and simultaneous growth in the
share with two adult earners and the share with zero adult
earners. Growth of the latter group has prompted a public
discussion about ‘work-poor’ families and the exclusion of some
from the economic mainstream.
• Much of the growth in employment has been in part-time jobs.
Roughly one job in four is part time with about 40 per cent of
women working part time.
• Economic recoveries have been characterised by stickiness in
unemployment. Unemployment rates have declined during
recoveries, but never to the levels experienced prior to 1975.
• There has been a noticeable growth in the share of the workingage adult population (ages 15 to 64) who receive income support
payments. The percentage was about five per cent in the late
1960s, but has varied between 20 to 24 per cent since 1991.
• The average duration of unemployment, of UA payments and of
some other social protection payments have all increased
substantially. In recent years, the median duration of
unemployment as measured both in the labour force survey and
in UA beneficiary data has hovered around one year.
16
17
Comparative data for 13 OECD countries in 1995 show the overall share of transfers going
to the bottom three deciles ranged from 20.8 per cent in Italy to 58.0 per cent in Australia
with the second highest percentage being 53.5 per cent in France. Conversely the top three
deciles in Australia received 7.4 per cent of transfers, the lowest percentage across the same
13 countries. See Figures 7 and 8 in Whiteford (2000).
Clear discussions of effective marginal tax rates in Australia are provided in Appendix 4 of
Interim Report of the Reference Group on Welfare Reform (2000) and in Section 2.4 in
Whiteford (2000).
364
These developments should be kept in mind as the experiences of Australia’s
UA program are reviewed.
Table 2 displays annual financial year data on unemployment and UA benefits. Between
1960 and 1974 the estimates of total unemployment ranged between 68 000 and 153 600
representing from 1.6 per cent to 2.7 per cent of the labour force. Since 1991, in contrast,
the annual averages have ranged between 656 300 and 938 000, and unemployment rates
have ranged from 6.9 per cent to 11.0 per cent.
During the years covered by Table 2, the number of UA beneficiaries has grown even
more rapidly than unemployment. Consequently the (NumBen/Unemp) ratio, which had
ranged from 0.15 to 0.38 between 1960 and 1974, has exceeded 0.60 in every year since
1976 and has exceeded 1.00 in the years since 1995. In a typical week during the most
recent five years there have been as many UA recipients as the number unemployed
reported in the labour force survey.
Data on UA recipients for recent years indicate that 15-20 per cent are working and also
receiving payments. This suggests that about 80-85 per cent of the unemployed as counted
in the labour force survey receive UA payments. What seems to make this possible is the
strong negative effect of long-term unemployment on family income. Among those with
long-term unemployment, it seems that family income is typically low enough to satisfy
the means test for UA eligibility. What appears to be a paradox, i.e. most of the
unemployed collect UA benefits even though eligibility is means tested, is at least partially
resolved by the fact that so many of the unemployed are long term.
Estimates of average weekly benefits and average weekly wages also appear in Table 2.
UA benefit levels were raised substantially in the early 1970s. Average payments tripled
between 1972 and 1976 and replacement rates increased. Since the mid-1980s, the
maximum payment has been indexed to the CPI with semi-annual adjustments. Prior to
1995, weekly benefits also included an allowance for dependent partners. Typically these
allowances were included in from one-third to one-half of payments to unemployed male
UA beneficiaries.
Note in Table 2 that between 1986 and 1994 the estimated replacement rates fall into the
0.30-0.34 range. As part of a reform package effective in 1995, payments to dependent
partners of unemployed individuals were discontinued. Observe that replacement rates
decline after 1994 and hover around 0.25 from 1996 onward. This decrease reflects the
discontinuation of payments to dependent partners. Between June 30, 1994 and June 30,
1995 the number of unemployed recipients decreased by about 75 000, but payments to
218 000 dependent partners also ceased. Thus the total recipients paid by the UA program
decreased by nearly 300 000.18 Total payments of UA benefits declined by about $1.25
billion between financial years 1995 and 1996.
18
These statements about beneficiary counts are based on data not shown in Table 2. The
beneficiary counts in Table 2 include just unemployed individuals, not dependent partners.
365
Table 2: The Cost of Unemployment Protection in Australia: 1959 to
1999
Financial
Year
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
Labor
Force
4109.5
4198.5
4282.0
4381.0
4524.5
4715.0
4851.5
4962.2
5084.1
5204.6
5379.0
5563.3
5666.5
5834.1
5990.3
6103.1
6230.9
6290.6
6400.9
6464.9
6600.7
6757.1
6863.4
6953.8
7067.6
7198.7
7451.4
7679.5
7866.8
8076.3
8346.3
8498.8
8526.0
8539.0
8672.0
8847.0
9055.5
9151.8
9228.0
9363.8
9542.5
Unemp
68.0
88.5
116.0
104.0
85.0
73.5
77.0
90.8
95.4
92.8
93.7
97.1
126.6
153.6
127.3
247.1
301.6
325.1
389.6
408.7
407.6
395.9
423.5
624.9
680.1
619.4
591.5
635.1
610.5
536.0
515.3
710.3
882.0
938.0
916.8
799.5
766.0
797.5
766.8
725.3
656.3
U Rate
TUR%
1.7
2.1
2.7
2.4
1.9
1.6
1.6
1.8
1.9
1.8
1.7
1.7
2.2
2.6
2.1
4.0
4.8
5.2
6.1
6.3
6.2
5.9
6.2
9.0
9.6
8.6
7.9
8.3
7.8
6.6
6.2
8.4
10.3
11.0
10.6
9.0
8.5
8.7
8.3
7.7
6.9
NumBen NumBen/ Total UA
Ann Avg
Unemp
Benefits
39.7
25.9
13.7
14.9
20.7
21.5
17.8
13.2
15.0
29.1
39.6
34.1
116.6
191.7
216.9
265.8
306.2
306.3
310.0
332.0
540.2
619.6
581.7
559.2
574.4
502.5
429.4
385.0
535.9
771.4
883.0
905.7
847.0
812.8
811.4
794.0
767.0
0.382
0.305
0.187
0.194
0.227
0.226
0.192
0.141
0.154
0.230
0.258
0.268
0.472
0.636
0.667
0.682
0.749
0.752
0.783
0.784
0.864
0.911
0.939
0.946
0.904
0.823
0.801
0.747
0.755
0.875
0.941
0.988
1.058
1.058
1.015
1.036
1.058
9
9
30
21
13
7
8
11
11
9
9
11
26
47
58
252
514
618
794
910
925
996
1224
2249
2912
2984
3122
3454
3375
3136
3068
4561
6736
7492
7598
7061
5812
6207
5916
5882
Weekly
Benefits
10
10
10
10
10
10
10
13
14
17
23
33
42
52
55
57
57
58
62
71
80
90
99
107
116
129
140
153
164
168
163
161
160
137
147
143
147
Weekly
Wage
41
43
46
48
51
54
58
63
69
76
83
96
120
138
155
170
183
200
227
263
292
317
338
359
381
404
431
460
487
501
511
526
544
557
574
592
606
621
Rep.
Rate
0.254
0.234
0.208
0.210
0.204
0.186
0.173
0.206
0.200
0.225
0.272
0.341
0.345
0.375
0.354
0.338
0.313
0.290
0.272
0.270
0.274
0.285
0.291
0.299
0.304
0.320
0.326
0.333
0.336
0.335
0.319
0.307
0.295
0.247
0.256
0.242
0.243
G
0.097
0.071
0.039
0.041
0.046
0.042
0.033
0.029
0.031
0.052
0.070
0.091
0.163
0.238
0.236
0.231
0.234
0.218
0.213
0.212
0.237
0.260
0.274
0.283
0.275
0.263
0.261
0.249
0.254
0.293
0.301
0.303
0.312
0.261
0.260
0.251
0.258
B%
0.230
0.134
0.061
0.065
0.085
0.079
0.059
0.051
0.054
0.115
0.185
0.194
0.659
1.152
1.221
1.405
1.481
1.345
1.247
1.306
2.130
2.502
2.355
2.248
2.272
2.043
1.732
1.538
2.121
3.034
3.302
3.205
2.819
2.208
2.267
2.082
1.995
Source: Department of Family and Community Services (FaCS), Australia Bureau of
Statistics and OECD. Labour force, unemployment and beneficiaries in thousands.
Total benefits in millions. Benefit data exclude mature age allowances from 1994
and payments to dependent partners after 1995. Estimates of the labour force and
unemployment for 1960-1964 based on OECD data. Beneficiaries in 1994-1999
estimated from a regression of the annual average on the simple average of June
data for current and past financial year.
366
Much of this decline in payments was merely a relabelling phenomenon. Dependent
spouses, mainly women, often were eligible to collect a partner allowance where
eligibility depended on the unemployed spouse receiving UA benefits. The revised
treatment of dependent partners was part of a reform package that emphasised increased
‘individualisation’ of payments to the unemployed. This will be discussed in Section 7.
The final columns of Table 2 display estimates of G and B per cent. The generosity index,
G, fell below 0.10 in all years before 1975, but has equalled or exceeded 0.25 in all years
since 1984. Note that the index declined noticeably after 1995, a change that parallels the
decrease in the replacement rate.19 Because the (NumBen/Unemp) ratio has hovered
around 1.0 in recent years, however, this high ratio has prevented G from declining to
lower levels despite the reduction in the replacement rate. Australia’s generosity index has
been consistently higher than the average for the 12 countries depicted in Figure 1 even
though it conditions payments on income and assets.
The combination of high unemployment rates and a reasonably high generosity index have
yielded a high cost of unemployment protection in Australia in recent years.
Unemployment benefits as a percentage of wages (B%) averaged 2.35 per cent between
1983 and 1999 and fell below two per cent of wages in just three years (1989, 1990 and
1999). The Australian cost rate for these 16 years averaged more than three times the US
rate (0.65 per cent). As noted above, having a means-tested UA system for the
unemployed is not necessarily less expensive than a UI system.
Figure 4 summarises Australian experiences since 1963 by plotting benefit costs (B%)
against the unemployment rate. The historical record falls into two periods, the 12 years
from 1963 to 1974 and the years after 1974.20 The association between benefit costs and
the unemployment rate was highly significant in regressions fitted for both sub-periods.
Note that the slope during 1975-1999 was twice the slope during 1963-1974 (0.309 versus
0.155). The larger slope reflects both the higher replacement rate and the higher share of
the unemployed receiving benefits during 1975-1999. The regression for these later years
also indicates that Australian costs exceeded the average for the 12 countries displayed in
Figure 1.
7
Unemployment and Benefit Availability in Australia and the US
A direct comparison of unemployment and unemployment benefit availability in Australia
and the US reveals several interesting contrasts. Table 3 displays summary data by decade
for the unemployment rate, unemployment duration, unemployment benefit recipiency,
unemployment benefit duration and (for the US) unemployment benefit exhaustions. The
bottom panel shows Australia/United States ratios for the various series extending from
the 1960s through the 1990s.
19
20
Again, much of the change was more apparent than real as spouse benefits often became
partner allowances with little or no change in the amount paid to the family.
The earlier of the two periods could be extended backward prior to 1963 with outcomes
similar to those depicted for 1963-1974. For earlier years, however, data limitations become
more serious. For example, the labour force survey was started in 1960 but initially covered
only urban areas.
367
Over these years the Australian unemployment rate increased sharply relative to the US
rate. During the 1960s the Australian/US ratio averaged less than 0.50 whereas during the
1990s the ratio exceeded 1.50. A sustained increase in this ratio occurred across the past
four decades.
Figure 4: Benefit Costs, Benefit Generosity and Unemployment Rates in Australia:
1963 to 1999
B%
3.5
X
X
3.0
X
X
2.5
X
X
2.0
X
X
X
XX
X X
X
X
X
1.5
X
XX
X
X
X X
X
1.0
X
0.5
X X
X X X-b
X X X X-a
1.0
2.0
3.0
a - Six years, not four as shown
b Four years, not three as shown
4.0
5.0
6.0
7.0
8.0
9.0
Regression 1963-1974. B% = -0.198 + 0.155*(TUR/(1-TUR))%
10.0
Adj R2 = 0.681
Regression 1975-1999. B% = -0.585 + 0.309*(TUR/(1-TUR))% Adj R2 = 0.955
368
11.0
TUR%
12.0
Table 3: Comparative Unemployment Data, Australia and the US
Mean
UnemMean
Median Unemp. Unemp.
ployment Unemp. Unemp. Benefit Benefit
Rate
Duration Duration RecipiencyDuration
(percent) (weeks) (weeks) Rate
(weeks)-a
Median Unemp.
Unemp. Benefit
Benefit Exhaustion
Duration Rate
(weeks)-a
Australia
1960s
1970s
1980s
1990s
2.0
3.9
7.6
8.9
4.3-b
14.2
42.9
52.2
9.9-d
19.9
23.4
0.24-c
0.43
0.85
0.95
11.9-b
18.6
56.0
71.3
7.3-b
11.2
27.1
41.3
United
1950s
1960s
1970s
1980s
1990s
4.5
4.8
6.2
7.3
5.8
11.3
11.8
11.9
15.0
15.7
0.42
0.36
0.34
0.29
0.31
6.3
7.1
7.6
12.0
12.5
13.8
14.7
14.9
0.25
0.23
0.29
0.33
0.36
Australia/United States
1960s
1970s
1980s
1990s
0.41
0.63
1.05
1.54
0.36
1.19
2.86
3.32
1.57
2.80
3.08
0.68
1.26
2.92
3.07
0.96
1.35
3.81
4.80
Sources:
Australian data from the Australian Bureau of Statistics and Department of
Family and Community Services. US data from the Bureau of Labour
Statistics and Office of Workforce Security of the US Department of Labour.
Note:
a) Means and medians for Australia computed by the authors.
b) 1966-1969
c) 1963- present
The unemployment rate can be expressed as the product of the inflow rate (new
occurrences relative to the labour force) and average duration. Table 3 displays mean and
median duration estimates based on the same labour force surveys that generate the
unemployment rate estimates. Note that the change in the relative duration ratios are even
larger than the changes in the relative unemployment rates. Australian mean and median
369
unemployment duration were both more than three times their US counterparts during the
1990s. The suggestion is that unemployment occurrences are less frequent (relative to the
labour force) in Australia than in the US, but the product of the occurrence rate times
duration has yielded a higher unemployment rate in recent years. Note also that average
duration in the US was higher in the 1980s and 1990s than in earlier decades. Several
duration series in the US show this pattern. (See Chapter IV in Vroman, 2001a)
The likelihood of receiving unemployment benefits has also undergone a sharp change
over the past four decades. During the 1960s about one in four among the unemployed
received benefits in Australia compared to roughly four in 10 in the US. In subsequent
decades the benefit recipiency rate increased in Australia, but decreased in the US. By the
1990s the Australian/US ratio of recipiency rates averaged more than three. Fewer than
one in three received unemployment benefits in the US compared to more than nine of 10
in Australia.
Associated with higher benefit recipiency in Australia has been a major increase in
unemployment benefit duration. Mean benefit durations were similar during the 1960s but
during the 1990s Australian duration averaged more than four times US duration. The
increased ratio occurred despite increased benefit duration in the US. However, whereas
the increase between the 1960s and 1990s in the US was 2.4 weeks (19 per cent), the
change in Australia was much larger. The increase in the Australian mean was 59.4 weeks
(499 per cent) while the median increased by 34.0 weeks (466 per cent).
The final column in Table 3 shows one consequence of increased benefit duration in the
US, namely increased benefit exhaustions. The regular UI program in the US most
typically allows 26 weeks of benefits. Nationwide, average potential duration has averaged
about 24 weeks over the past four decades. As a consequence of increased actual benefit
duration but unchanging potential duration, an increasing fraction of UI recipients exhaust
benefits. The exhaustion rate which averaged 0.23-0.25 in the 1950s and 1960s was 0.36
in the 1990s. More than one of three UI recipients in the US collect their full entitlement
and are terminated from benefit status. A low fraction of such persons (fewer than one in
10) move on to some other form of public income support.21 For both economies the rate
of unemployment occurrences has decreased while unemployment duration has increased.
However the scale of the increases in duration in Australia swamp the much smaller
increases that have taken place in the US.
Figure 5 provides a visual representation of the evolution of unemployment duration in the
two economies. It displays centred five-year averages of the means and medians from the
labour force surveys. Since the mid-1980s the Australian means have consistently
exceeded 45 weeks. Because of the large scale of the Australian changes, the changes for
the US seem modest. However, the US means from the mid-1980s and mid-1990s
21
The two most common income support programs for UI exhaustees are welfare (now termed
Temporary Assistance for Needy Families or TANF) and food stamps. Following welfare reform
of 1996, TANF caseloads have decreased by about one-half and food stamps caseloads by about
one-third.
370
exceeded 15 weeks for several consecutive years. Increased unemployment duration has
occurred in both economies, but on much different scales.
Figure 5: Unemployment Duration in Australia and the US: Five Year Averages
60.00
50.00
Weeks
40.00
Aus. - Mean
Aus. - Median
US - Mean
US - Median
30.00
20.00
10.00
0.00
1969
8
1989
Disincentives and Policy Changes in Australia
Disincentives and the phenomenon of long duration in benefit status have been recognised
as problems in Australia for many years. Several changes in the terms of benefit eligibility
have been implemented but problems persist. Our purpose here is to briefly describe some
salient developments.
Since the mid-1990s increased emphasis has been placed on the principal of mutual
obligation. Those who receive financial (and other) support from public resources are
encouraged to become financially independent and to pursue activities that contribute to
the Australian community. Mutual obligation is an Australian variant of a policy direction
now pursued in several OECD countries and described variously as ‘activation’ and
‘reciprocal obligation.’
Structure of the Benefit Phase-Out
The earliest form of the means test on family income featured a dollar for dollar reduction
in benefits when family income exceeded the guarantee threshold. For age pensions, this
was modified in 1969 with the introduction of a tapered means test coupled with a free
371
area.22 Some earnings were allowed with no reduction in pensions (the free area) followed
by a 50 per cent reduction rate when earnings exceeded the amount allowed by the free
area. Subsequent modifications over the next decade widened the free area, eliminated the
test for those 75 and older, then for those 70-74 and then eliminated the asset element of
the means test.23 Most of the latter changes were reversed between 1978 and 1985, but key
features were retained, e.g. a free area and a phase-out with an effective tax (benefit
reduction) rate of 50 per cent.
Means testing of UA benefits followed a similar history but with changes occurring later.
The dollar-for-dollar benefit reduction was in place through 1979. The free area and the 50
per cent reduction followed by a 100 per cent reduction were introduced in 1980 and
several more modifications occurred between 1982 and 1994.24 Throughout this period the
UA income test was based on family income.
Major changes were then instituted in 1995. First, the basis of entitlement to UA benefits
was changed from family income to individual income (each person’s earnings plus their
share of other countable income). Second, payments to dependent partners of the
unemployed were ended, but replaced by partner allowances in most situations. Third, the
100 per cent phase-out range was replaced with a 70 per cent phase-out range (while the
free zone and the 50 per cent phase-out were retained, although the size of the free zone
was reduced ). Thus as benefits were being reduced due to increased earnings, the phaseout was restructured so that income (earnings plus UA benefits) would always be higher as
a result of higher earnings.
The change to ‘individualisation’ was made in order to improve incentives for combining
work with receipt of UA benefits, particularly among women who often work part time.
Previously, working women were frequently precluded from UA benefits while
unemployed due to earnings and other income of their spouse. While it became more
likely that an unemployed wife could now collect UA benefits, individualisation did not
mean that husbands’ earnings became irrelevant in eligibility determinations. Husbands’
earnings equal to or exceeding 60-62 per cent of national average earnings would preclude
an unemployed wife from eligibility.25 The net effect of the change in many situations was
to allow receipt of UA benefits and to encourage part-time work (due to the lower, 70 per
cent, phase-out), particularly at higher levels of earnings. One analysis using data for the
two and a half years following the changeover suggested this effect did take place.
(Chapters 4, 5 and 6 of Warburton, Vuong, and Evert, 1999).
While the analysis of the effects of individualisation by Warburton, Vuong and Evert,
(1999) suggest positive effects on earnings and labor force participation, the size of the
22
23
24
25
While there is an assets test, it is applied to restrict payments relatively infrequently, in
situations where there are high assets but low income. See footnote 4 in Whiteford (2000).
One description of these developments is given in Section 2.4 in Whiteford (2000).
A concise description of these changes is given in Section 2.1 and Table 2.1in Warburton,
Vuong, and Evert, (1999).
The threshold on spouses’ allowable weekly earnings in August 1999 was roughly $500
compared to a national average for full-time workers of about $800.
372
impacts appears to have been modest. Figure 6 traces aggregate labour force participation
rates and part-time work proportions for women by financial year from 1966-1967 through
20
00
0.60
0.55
0.50
0.45
0.40
0.35
0.30
0.25
0.20
19
67
Proportions
Figure 6: Australian Women’s Labour Force Participation and Part-time Work
Proportions
LFPR
PPT
1999-2000. Developments in participation rates during the 1990s were dominated by
cyclical factors and long-term trends while women’s part-time work proportions increased
at quite a regular pace.26
Job Search and Mutual Obligation
Australia has undertaken a variety of initiatives to promote activation among recipients of
allowances and specifically among the unemployed. Mutual obligation is the term used to
describe situations where the recipient’s right to cash payments (or other support) is
acknowledged, but the receipt of payments is conditioned on the discharge of a reciprocal
obligation. Registration with the employment service and engaging in active job search are
two traditional obligations that have been placed on the unemployed. More recently the
scope of mutual obligation has been expanded with additional changes slated for
implementation in 2002.
26
Regressions conducted within a trend cycle framework using annual data showed a strong
effect of unemployment on labour force participation and part-time work by women but no
breaks at financial year 1995-1996.
373
The scope of activities falling under mutual obligation is now quite wide, and it differs
according to the type of benefit or allowance being received. Persons in receipt of benefits
may (depending on the type of payment): search for work, undertake training, do unpaid
community work, care for the young, engage in physical rehabilitation or undertake life
skills training. These activity requirements show that the recipient is either looking for
work, improving skills or contributing something to the community.
For the unemployed, several changes in the administration of the work search requirement
have been implemented. Throughout, there has been a continuing requirement that the UA
recipient must be unemployed (able to work, available for work and actively seeking
work) and registered as a job seeker.
Until 1982 the agency with primary day-to-day responsibility for administering the work
test was the Commonwealth Employment Service (CES). The CES, part of the
Employment Department, acted as an agent for the Department of Social Security (DSS)
which administered the payment of UA cash benefits. In 1982 DSS, assumed direct
responsibility for work test administration. Responsibility was returned to CES in 1991
though DSS had a partial role in 1993-1994.
Reforms of 1995 altered the administrative structure for the delivery of benefits and
services to the unemployed. A new administrative entity, Centrelink, was established in
December 1996 to deliver social security entitlements. Centrelink was also to perform a
gateway function for the unemployed, i.e. registering people as unemployed and assessing
their degree of labour market disadvantage. Other labour market services previously
discharged by CES such as job matching and case management became the responsibility
of Job Network, a semi-privatised ‘market’ with government, non-profit and for-profit
organisations competing to provide employment services.27 Registration and job search
continue to be required within this revised service delivery structure.
The approach to ‘activate’ the unemployed has also undergone several modifications.
Prior to the large increase in unemployment of the mid-1970s, emphasis was placed
mainly on the acceptance of suitable work. During these years, the number of vacancies
listed with the CES represented some one-third to one-half of UA registrants. Thus CES
could offer jobs to a meaningful share of registrants. This changed when unemploymentvacancy ratios moved from roughly 2.0 upwards to 20.0 and higher in the mid-1970s. In
line with this development, there were changes in the definition of suitable work which
could be refused while retaining an entitlement to benefits. Guidelines were broadened in
1976 to require (after 12 weeks) acceptance of work in line with local job availability even
if it meant a reduction in wages and/or status. By 1989, this definition had been further
modified to require acceptance of casual, part-time or temporary work.28
27
28
In October 1998 the Department of Family and Community Services (FaCS, the entity that
subsumed functions previously discharged by the Department of Health and Family
Services) was created. Its responsibilities included child care, disability and housing
programs, i.e. benefits and services to families.
This evolution prior to 1995 is described in Chapter 7 of Department of Social Security
(1995).
374
Work search requirements also have become more formal with increased emphasis on
evidence of active search. Changes effective in 1991 required both the short-term and the
long-term unemployed to satisfy an activity test. For those unemployed less than 12
months (receiving a payment termed a Job Search Allowance) the activity test included
active work search or participation in labour market or vocational training approved by the
Department of Employment, Education and Training (DEET) as likely to improve job
prospects or the effectiveness of job search. For the long-term unemployed (12 or more
months in benefit status, payments termed Newstart Allowance)29 there was a requirement
to participate in an activity agreement intended to secure reemployment but tailored to
individual circumstances. Several possible activities were to be considered, e.g. work
search activity acceptable for the Job Search recipients plus other activities including paid
work experience and activities proposed by the recipient, e.g. unpaid volunteer work. One
intention of this change was to reallocate CES administrative resources to target the longterm segment of the unemployment pool.
Further changes in the activity test became effective in 1995. Increased emphasis was
placed on early identification of likely long-term UA recipients. The attempt to identify
long-term recipients was perhaps influenced by new administrative practices in the US
that ‘profile’ likely UI exhaustees.30 Also, a wider range of acceptable search activities
could be considered. These changes have not had a noticeable effect on measured duration
which, if anything, has been higher since their implementation.31
New initiatives including enhanced mutual obligation requirements intended to address
long-term unemployment were implemented in 1999. Phased in initially with younger
workers, recipients must undertake one of a set of approved activities besides job search.
Approved activities include paid part-time work, voluntary work, education or training,
relocation, work for the dole (WfD, mandatory participation in temporary public
employment) and several other activities. A Welfare Reform Reference Group was
established in late 1999, and issued reports in March and August 2000. It made
recommendations both to provide added services to the unemployed and to require
enhanced application of the principal of mutual obligation. Many of the Reference Group's
recommendations have been endorsed and will be implemented in phases through 2004.
Increased budget support for the associated reemployment activities has also been
committed.
Older unemployed workers have generally not faced the activity test requirements applied
to younger workers. Those aged at least 50, but younger than retirement age frequently
have been exempted from active search and have been allowed a wider range of
acceptable alternatives to searching for paid work. The Mature Age Allowance, paid since
1995, goes to dislocated workers aged 60 and older and effectively functions as an early
29
30
31
Since September 1996 the term Newstart Allowance has applied to both long-term and
short-term UA recipients.
Section 9 discusses profiling as practiced in the US.
Medians computed by the authors based on data for June 30th averaged from 7.9 to 8.9
months between 1992 and 1996, but increased to 11.9 months in 1997, 14.0 months in 1998
and 16.6 months in 1999.
375
retirement benefit. Older workers consistently exhibit the longest average durations of all
age groups. In recent years, percentages as high as 40 per cent of older workers who leave
UA benefit status have exited to become recipients of invalidity (disability) pensions.
Relative to UA benefits, these payments have higher guarantees and carry no obligation
for active job search. In 1995, disability pensions were received by 15 per cent of men
aged 55-59 and by 25 per cent of men aged 60-64.
Looking back over the past thirty years, it seems that activation has not been very
successful in Australia, at least as reflected in macro-labour market indicators.
Unemployment duration (in both the labour force survey and in UA beneficiary data)
remained stubbornly high during the economic recoveries of the 1980s and the 1990s.
While the ultimate goal of the changes in the activity test and the principal of mutual
obligation has been to speed the movement of people from unemployment to employment,
achievement of the goal has proved elusive.
9
Initiatives to Affect Unemployment Benefit Duration in Australia
The duration of unemployment benefits in Australia seems unusually lengthy, especially
in comparison to benefit duration in the US. In the following paragraphs we discuss policy
interventions that might significantly alter benefit duration. The discussion is organised
into three areas: entry eligibility, ongoing benefit eligibility and limiting potential benefit
duration. Since one of the authors is from the US, much of the discussion will draw upon
US experiences.
Entry Eligibility
Two ways to restrict entry into UA would involve changes in means testing. First, the
income and asset thresholds used in the means test could be reduced. Second, procedures
to verify the income of applicant households could be modified. This second area will be
examined in later paragraphs in a discussion of the Benefit Accuracy Measurement
program currently utilised in the US.
The present analysis will focus on profiling. Is it possible to identify within the pool of
eligibles those especially likely to become long-term UA beneficiaries? Our discussion
briefly summarises US experiences with profiling.
The central idea behind profiling is straightforward. The pool of eligibles is heterogeneous
and those most likely to collect unemployment benefits for long periods can be identified
on the basis of observable characteristics. A profiling algorithm is developed using data
from an earlier period where claimant characteristics and the actual duration of benefits
are observed. The algorithm estimates parameters in a multivariate statistical model, e.g. a
logit regression, where the dependent variable is a 0-1 variable (= 1 when the person
exhausts benefits or when benefit duration exceeds a predetermined threshold). The same
algorithm is then used to assign scores among current applicants. Applicants are ranked
and those with high scores, indicating a high likelihood of exhaustion (or long benefit
duration), are subjected to different administrative treatments than applicants with low
scores. The treatments are intended to shorten benefit duration. Persons refusing to
participate in the treatments may be denied benefits.
376
Profiling is required in all states but enjoys mixed support by UI administrators. Among
the questions raised by profiling, four seem especially important:
•
How well can the potentially long-term unemployed be identified?
•
Does profiling intervene at the right time, i.e., at the start of the unemployment
spell?
•
Can effective treatments be fashioned?
•
Can effective treatments be fashioned for the long-term unemployed?
The third and fourth questions may have quite different answers. It is conceivable that the
most cost effective interventions operate on the short-term end of the duration
distribution. Thus job search assistance can help to shorten duration among those likely to
become reemployed relatively sooner. Under this hypothetical scenario, job search
assistance could be effective, reduce total weeks claimed, but result in increased average
benefit duration because it helps (removes from the unemployment pool) those whose
normal experiences would be characterised by below-average benefit duration.
Evidence on the first of the four questions suggests that the long term unemployed can be
effectively identified. However, Berger, Black and Smith (2000) argue that the most
effective identification algorithms require the use of more variables than typically utilised
by states. Particularly important are: longitudinal variables that incorporate past
experiences with the UI program and variables reflecting local labour market demand
conditions. Using data from Kentucky, they present persuasive evidence on the improved
ability to discriminate among claimants according to potential duration.
The second question also has interesting ramifications. The presumption of profiling is
that early interventions work better than later interventions. This assumption is reasonable
if the previous job has truly disappeared and will not be available to the claimant in the
near future. However, in many situations there is a chance that the previous job will
become available at a later time. If the claimant secures a new job, this may permanently
sever the former job match and preclude return to the former job. A recent paper by
Woodbury and Anderson (2001) finds that some claimants who wait do return to former
jobs and at high wages. Their results also indicate, however, that among those who wait,
overall reemployment rates are lower than for those who start to search for new jobs
sooner. These two findings make the timing of job search a statistical decision problem.
Some should wait (those who eventually return to former jobs) while others should start to
search as soon as possible. Within the claimant pool, some experience net gains while
others experience net costs from waiting.
The presumption of profiling is that claimants are either job attached (and exempt from
profiling treatments) or not job attached (and placed in the profiling pool). In actual labour
market situations what appear to be permanent separations sometimes turn out to be just
temporary. Among some claimants with high profiling scores, not waiting has net costs
because the previous job match is permanently severed.
377
What are the profiling interventions? Data from 1999 were examined with the following
summary results. The total number profiled was 6.5 million or about 70 per cent of all new
initial claims for UI benefits.32 Almost 2.2 million claimants were identified as likely
exhaustees while nearly 0.8 million were actually referred for services. The reporting
system identifies seven service categories. Total counts in 1999 were as follows:
orientation - 0.44 million, assessment - 0.40 million, counselling - 0.20 million, job search
workshops - 0.25 million, placement services - 0.67 million, education and training - 0.14
million and the self-employment program (present in just a few states) - 0.01 million. The
total number of services offered totaled 2.1 million, but the bulk (1.29 million) were very
short term interventions (orientation, assessment, counselling and job search workshops).
Education and training measures, the most significant long-term intervention, totalled less
than 0.15 million. Placement services and education and training measures had completion
rates of about 70 per cent compared to nearly 100 per cent for the quartet of very shortterm (typically one-shot) interventions.
From the preceding, three summary comments seem appropriate.
•
More than half of those identified as likely exhaustees through profiling were not
even referred to a labor market service. Constraints on the availability of services
explain why so many do not receive services. The low ratio of referrals among
likely exhaustees (0.8 million of 2.2 million or 0.37) occurred in 1999, a year of
strong labor markets. Lower ratios would be expected in a period of high
unemployment.
•
The services obtained were typically very short term, usually involving a single
meeting. Long-term services were received by a comparatively modest proportion
(0.13-0.15) of those who were referred and completed at least one service.
•
For some claimants, profiling may cause productive job matches to be severed
because it is applied at the start of the spell of unemployment when it is not yet
certain that the job separation is permanent.
It seems there is a mismatch between services offered to potential exhaustees and the
services needed if duration is to be shortened significantly. A rule of thumb for the effect
of profiling on duration is a reduction of 0.50-0.75 weeks among those referred to services
(Corson et al., 1989). Since less than 10 per cent of initial claimants are referred to
services, the effect of profiling on duration in the US is probably less than 0.1 week.
Ongoing Benefit Eligibility
Two areas of activities related to ongoing eligibility in the US have potential for reducing
average duration in benefit status. The first involves periodic face to face contact between
the beneficiary and local office employees. The second involves measuring the accuracy
of eligibility determinations with information gained from small samples that repeat all
administrative aspects of the original eligibility determinations.
32
These statistics refer to 1999 calendar year summaries for all 53 ‘state’ UI programs.
378
Individual states differ widely in their reliance on eligibility review interviews (ERIs) and
other reporting requirements. These meetings convey information to claimants, assess
their reemployment strategies and demonstrate active administrative oversight of search
and other reemployment activities. The meetings represent a proactive approach to
program administration that goes beyond the typical declaration of job search made by
claimants when they file for a continuation of benefit payments. Failure to participate in
such meetings is a basis for denying a claim for UI benefits.
In a typical continued claim, the beneficiary answers five or six questions about receipt or
earnings, other deductable income, e.g. severance pay, and availability for work for the
upcoming two week period on the front side of a claim form. On the back side of the same
form the names and addresses of employers contacted for possible jobs during the most
recent two week period are supplied. 33
ERIs and other reporting requirements may occur at set intervals during a claim or
following complaints (from, say, a neighbor or a former employer) or when requested by
staff of the administering agency (at random or in light of new case-specific information).
Certain states such as Georgia, New Hampshire and North Carolina add state monies to
federal monies for UI administration to ensure adequate staffing and that the meetings
occur with high frequency. In states that monitor ongoing eligibility most actively, the first
meetings will typically occur four or six weeks after the start of benefits.
These meetings entail some inconvenience for the claimant, but they also help many to
develop an improved understanding of their current unemployment situation. The
meetings have elements of both help to the claimant and a threat to continuing eligibility.
One consequence in the three states just mentioned is that benefit duration is much shorter
than the national average, e.g. 9-11 weeks compared to a national average of 13-15 weeks.
This approach to case administration has ample precedent in Australia, but with a clear
difference in the timing of the initial meeting. Prior to 1999, duration of one year would
trigger increased administrative oversight. More recently, interventions may occur after
three or six months in UA benefit status. It may be appropriate for Australia to consider
requiring even earlier face to face meetings.
For more than a decade the accuracy of UI administrative determinations have been
assessed in each state through a program of Benefit Accuracy Measurement (or BAM).34
The measurements initially focused on benefit overpayments (for both initial eligibility
determinations and determinations of continuing eligibility). Since the mid-1990s, BAM
has also examined the accuracy of employer UI tax payments. At present, UI program
administration is extending BAM to include measurement of the accuracy of denied
benefit claims.
33
34
The same questions can be posed and answered by telephone within an interactive voice
response (IVR) framework. Increasingly new claims and continued claims are transacted by
telephone in the US.
This measurement system was formerly known as Quality Control (or QC).
379
The BAM system selects small samples in each state and undertakes an intensive review
of all elements that contributed to the administrative decisions. It detects errors, estimates
the size of errors (computing both case error rates and dollar error rates) and pinpoints the
source(s) of errors. In benefit payment activities, errors arise from three sources: the
claimant, the employer and the administrative agency. BAM helps agencies identify
problem areas, both staffing limitations and specific administrative procedures that need
attention. The data can help in program management within a state and provide
information on interstate differences in the accuracy of administrative determinations.
The concept of benefit accuracy measurement may be worth considering in Australia.
Because UA considers both labour market activities and family income in decisions about
initial eligibility and continuing eligibility, a BAM system would encompass more
measurement elements than in the US. However, it might provide insights into long
unemployment duration not available from current administrative records. BAM
investigations review all elements that entered original decisions with reference to a
specific time period, e.g. the original decision to award (or deny) benefits or a specific
(two week) period of benefit receipt.
In Australia, a full investigation of continuing eligibility would encompass not only the
customary work activity trio (able, available and actively seeking work), but also other
requirements of mutual obligation and all components of family income. Family income
will often vary in the short run depending on the work activity of individual family
members.
In investigations of income eligibility, it is important to secure information from electronic
sources. This pertains to wage earnings, provisional self employment income, various cash
transfers, interest income and subsidies for housing services, transportation and health
services. Since several types of information are reported using tax identification numbers,
matching electronic information would be straightforward in most situations. Information
from these sources could then be compared with the information used by Centrelink in its
eligibility determinations. These investigations could yield important insights in to
claimant fraud and administrative errors.
It is likely that changes in family income occur with high frequency, especially among
families with two or more persons of working age. BAM investigations could identify the
frequency of income changes and the types of income especially prone to reporting errors
and/or fraud.
Since UA allowances are financed by general revenues, the potential scope of BAM in
Australia would not include program revenues. However, benefit accuracy activities
should extend to erroneous denials as well as possible overpayment errors. As US
measurement of wrongful denials moves to nationwide implementation in the last half of
2001, there may be problems that can be avoided by tracking US experiences.
With improved income measurement it probably would be possible to address a seeming
paradox in Australia’s provision of unemployment protection. The program uses a means
test to determine eligibility, but yet the ratio of UA beneficiaries to labour force survey
unemployment has averaged close to unity in recent years (Table 2). If roughly 20 per cent
380
of UA recipients are employed under labour force survey definitions, that still leaves a
ratio of 0.8 in comparing unemployed UA recipients with labour force survey
unemployment. There are at least two possible explanations for such a high ratio.
•
Long duration unemployment causes such a large reduction in family income that
families in such situations satisfy the means test.
•
The measurement of family income and assets is deficient so that the means test
does not make accurate eligibility determinations.
A BAM measurement system could provide key insights into these two possible
explanations.
Limiting Potential Benefit Duration
One of the strongest contrasts between the Australian and US approaches to
unemployment compensation relates to potential benefit duration. Recipients of UA in
Australia can remain in benefit status for as long as they satisfy the requirements of job
search, other requirements of mutual obligation and the means test. Recipients of UI in the
US, in contrast, typically face a maximum potential duration of 26 weeks during a benefit
year of 52 weeks duration.35 Studies with micro data on benefit duration, e.g. Katz and
Meyer (1990), have consistently found that exit rates among UI beneficiaries increase
even before the maximum entitlement limit is reached. At the point of exhaustion, some
UI recipients exit the labour force while others secure jobs.
Since Australia’s UA program is means tested, the experiences of US means-tested
programs is also germane to this discussion. Temporary Assistance for Needy Families
(TANF), the most important of these programs, has been operating with time limits since
1996 and caseloads have dropped sharply. The number of recipient families which
averaged 5.05 million in 1994 averaged just 2.64 million in 1999, a decrease of 48 per
cent. While some of this decrease can be attributed to the strong economic performance of
the late 1990s, the sustained economic recovery of 1983-1989 was not accompanied by
large caseload reductions in the earlier AFDC (Aid to Families with Dependent Children)
program. Recipient families averaged 3.65 million in 1983 and 3.77 million in 1989. Since
AFDC did not have time limits, this provides a further indication that time limits have a
strong effect on caseloads in means-tested programs.
Two distinct approaches to time limits are followed in the US
•
35
For UI, the time limit of 26 weeks applies during a 12 month benefit year, but is
renewable. A person could collect UI for 26 weeks during the current benefit year
and do so during the next year as well. Recidivism is common, especially in
seasonal industries such as food processing and construction. However, to
In Massachusetts and Washington, the maximum potential duration is 30 weeks. In most
states many claimants are entitled to fewer than 26 weeks due to variable duration formulas
which yield shorter durations for claimants with low wages and/or irregular work patterns.
Nationwide, potential benefit duration has averaged close to 24 weeks since the 1960s.
381
establish eligibility for the next year requires having a threshold level of earnings
and/or work in the current year.36
•
For TANF, the time limit is a lifetime limit. Entitlements used during each year
count towards the lifetime limit. Once the lifetime limit is reached, one is
precluded from further eligibility. The national welfare reform legislation of 1996
specified the lifetime limit at five years, but allowed states to impose shorter
limits. Several states now have shorter limits. Because the TANF authorising
legislation was so recent, the real consequences of lifetime limits are not yet
obvious. As noted, caseloads have decreased sharply, but there has not yet been a
recession to test the immutability of lifetime time limits when large numbers
actually exhaust their TANF entitlements.
The UI program in the US has devised procedures to alter (increase) the 26 week time
limit during recessions. Two programs can extend benefits when unemployment is high.
Both use the 26 week program (so called regular UI) as the port of entry for lengthened
entitlements. The first, the Extended Benefits (EB) program is automatically triggered ‘on’
when an unemployment rate exceeds a predetermined threshold. When EB is ‘on’ the
claimant can collect up to 13 additional weeks of benefits at the same weekly rate as for
regular UI. Setting the appropriate trigger threshold for EB presents a serious operational
problem in the US. Present thresholds are so high that EB would be very difficult to
trigger ‘on’ in most states.37 During the 1990-1992 downturn, only nine states paid any EB
benefits. The second method is through national legislation which can extend UI benefits
on a temporary emergency basis during recessions. The most recent program, Emergency
Unemployment Compensation (EUC), paid benefits during the 30 months between
November 1991 and April 1994. These emergency programs are fully federally financed
and EUC provided half as much in payouts during 1992 and 1993 as the regular UI
program. Emergency national programs have been enacted in every recession since 1958.
While details have varied from one recession to the next, they all have had automatic
sunset provisions and several have had one or more extension of the original benefit
period. All have extended potential benefits for finite periods such as 13 or 26 weeks.
To summarise, three comments about time limits can be offered.
36
37
•
Time limits can be renewable so that the 26 week maximum in regular UI
programs has reference to a single 52 week year, not a longer time period.
•
Recidivism occurs, especially among workers in seasonal industries.
The UI terminology is that the person must have sufficient earnings in the base period to
meet monetary eligibility requirements. Like the benefit year, the base period is also a
twelve month period. The administrative principle for successive base years is that earnings
in one base year cannot be used again in the subsequent base year.
Thresholds are based on UI claims data in nearly all states. The share of claimants that
collect benefits is highly varied from one state to the next. Also, threshold formulas vary by
state.
382
•
Potential duration can be extended during recessions either by some automatic
mechanism (EB) or through ad hoc emergency legislation (most recently EUC).
Under both EB and EUC programs, however, the extensions eventually sunset. Thus, to
secure a renewed entitlement to UI, the individual has to return to work and generate
earnings sufficient to satisfy the monetary eligibility requirement for a later benefit claim.
In most (non-recessionary) situations the claimant has a finite, but renewable entitlement
to UI benefits.
10
Summary
The preceding pages examined three aspects of UA benefit duration in Australia mainly
from a US prism. Our summary assessment of possible applicability in Australia is as
follows. First profiling is a comparatively cheap intervention, but, at least as implemented
in the US, one that offers faint hope of significantly affecting UA benefit duration.
Second, implementation of a BAM system on a pilot basis might yield very useful insights
into the accuracy of means testing among UA recipients. Third, limiting potential benefit
duration could have strong effects on actual duration.
There seems to be a logical sequencing of the latter two interventions. Test a BAM system
first and assess its potential impact on UA caseloads and benefit duration. Placing a time
limit on benefits would represent a more radical departure from current and past practice
in Australia. This might be considered later, after the efficacy of BAM is better
understood.
11
Conclusions
This paper has compared two types of unemployment benefit systems, UI and UA. The
analysis of Sections 1 to 5 was undertaken at a rather general level. The latter sections of
the paper focused specifically on a single UA system (Australia) and a single UI system
(the US). Key differences between Australia and the US are found in the duration of
unemployment and the duration of unemployment benefits. Sections 8 and 9 examined
historical developments and specific aspects of program administration with particular
attention to unemployment benefit duration in Australia.
A most interesting finding of the paper relates to UA program costs. Section 4 examined
cost data from twelve countries with UI programs and four countries with UA programs
for the unemployed. Two of the latter four countries, Australia and New Zealand, have had
costs in recent years that can be described as somewhat above average when compared to
the costs of UI programs for a sample of twelve OECD countries. The other two UA
systems, in Hong Kong and Estonia, have exhibited costs similar to the costs of the
lowest-cost UI system examined here (Greece).
Thus the conclusion about comparative costs of UA versus UI is ‘it depends.’ One would
need to specify exact statutory and administrative provisions of the UA and UI system
being compared before making inferences about their comparative costs. Australia’s UA
program is roughly three times more expensive per percentage point of unemployment
when compared to the US which operates a low cost UI system.
383
The finding that UA has high costs in Australia points to a seeming paradox. Australia
conditions eligibility on income but still most of the unemployed, especially the long-term
unemployed, receive UA benefits. While the level of support payments is modest (about
one-fourth of the average weekly wage), many recipients experience long-term periods in
benefit status. A lower income guarantee probably would result in shorter spells of
unemployment. Two factors could be contributing to this outcome. A number of
Australians could be prolonging their spells of unemployment to satisfy the income and
asset eligibility conditions for UA. This could suggest there is a degree of control over
unemployment duration for at least some recipients (moral hazard). Second, some
claimants may be able to misrepresent their income and/or assets in order to satisfy the UA
means test. Both factors could be contributing to a situation where the number of fully
unemployed UA recipients represents about 80 per cent of all persons measured as
unemployed in the Australian labour force survey.
This could suggest that the Department of Family and Community Services (FaCS) needs
to invest more resources in income and asset verification activities. Section 9 discussed a
specific type of measurement system to improve the measurement of family income.
Improved measurement could provide valuable insights into the long UA benefit duration
characteristic of Australia.
Section 9 also explored some details of placing time limits on benefits. This too has
potential for shortening average UA benefit duration in Australia. Most of the discussion
of Section 9 was based on US experiences. The discussion yielded a more pessimistic
assessment of profiling, at least as currently practiced in the US.
References
Atkinson, Anthony and John Micklewright (1991), ‘Unemployment compensation and
labor market transitions: a critical review’, Journal of Economic Literature, 29,
December, 1679-1727.
Berger, Mark, Dan Black and Jeffrey Smith (2000), Evaluating Profiling as a Means of
Allocating Government Services, University of Western Ontario.
Blanchard, Olivier and Justin Wolfers (2000), ‘The role of shocks and institutions in the
rise of European unemployment: the aggregate evidence’, The Economic Journal,
110, March, C1-C33.
Boeri, Tito (1997), ‘Labor market reforms in transition economies’, Oxford Review of
Economic Policy, 13(2), June, 126-35.
Burtless, Gary and Daniel Saks (1985), The Decline in Insured Unemployment During the
1980s, The Brookings Institution, Washington, DC.
Corson, Walter and Walter Nicholson (1988), An Examination of Declining UI Claims
During the 1980s, Unemployment Insurance Occasional Paper 88-3, US
Department of Labor, Washington, DC.
Corson, Walter, Shari Dunstan, Paul Decker and Anne Gordon (1989), New Jersey
Unemployment Insurance Reemployment Demonstration Project, Unemployment
Insurance Occasional Paper 89-3, US Department of Labor, Washington, DC.
384
Department of Social Security (1995), Social Security in Australia, Department of Social
Security, Canberra, Australia.
Gornick, Janet (1999), ‘Income maintenance and employment supports for former welfare
recipients: the United States in cross-national perspective,’ in Evelyn Ganzglass
and Karen Glass, eds, Rethinking Income Support for the Working Poor, National
Governors' Association, Washington, DC, 49-90.
Ham, John, Jan Svejnar and Katherine Terrell (1998), ‘Unemployment and the social
safety net during transitions to a market economy: evidence from the Czech and
Slovak Republics’, American Economic Review, December, 1117-42.
Katz, Lawrence and Bruce Meyer (1990), ‘The impact of the potential duration of
unemployment benefits on the duration of unemployment’, Journal of Public
Economics, 41, February, 757-82.
Moffitt, Robert (1992), ‘Incentive effects of the US welfare system: a review’, Journal of
Economic Literature, 30(1), March, 1-61.
Organisation for Economic Cooperation and Development (OECD) (1998), The Battle
Against Exclusion: Social Assistance in Australia, Finland, Sweden and the United
Kingdom, OECD, Paris, France.
Reference Group on Welfare Reform (2000), ‘Participation support for a more equitable
society’, and ‘Appendices’, in the Interim Report, Department of Family and
Community Services, Canberra, Australia.
Schmid, Gunther and Bernd Reissert (1996), ‘Unemployment compensation and labor
market transitions’, in G. Schmid, J. O’Reilly and K. Schomann, eds, International
Handbook of Labor Market Policy and Evaluation, Edward Elgar, Cheltenham,
UK, 235-76.
Terrell, K., M. Lubyova and M. Strapec (1996), ‘Evidence on the implementation and
effectiveness of passive and active labor market policies in the Slovak Republic’,
in OECD, Lessons from Labor Market Policies in the Transition Economies,
OECD Proceedings, Paris.
US Social Security Administration (1999), Social Security Programs Throughout the
World, 1999, Social Security Administration, Washington, DC.
Vroman, Wayne (1991), The Decline in Unemployment Insurance Claims Activity in the
1980s, Unemployment Insurance Occasional Paper 91-2, US Department of Labor,
Washington, DC.
Vroman, Wayne (2001a), Low Benefit Recipiency in State Unemployment Insurance
Programs, Report to the US Department of Labor, The Urban Institute,
Washington, DC.
Vroman, Wayne (2001b), Unemployment Insurance and Unemployment Assistance: A
Comparison, The Urban Institute, Washington, DC.
Vroman, Wayne and Stephen Woodbury (1996), Increasing Unemployment Insurance
Benefit Recipiency, prepared for the US Department of Labor, Washington, DC.
Warburton, Mark, Luat Vuong and Heather Evert (1999), An Evaluation of the Working
Nation Income Test Changes for Unemployed People, Department of Family and
Community Services, Canberra, Australia.
Whiteford, Peter (2000), The Australian System of Social Protection – An Overview,
Policy Research Paper No 1, Department of Family and Community Services,
Canberra, Australia.
385
Woodbury, Stephen and Murray Rubin (1997), ‘The duration of benefits’, in C. O’Leary
and S. Wandner, eds, Unemployment Insurance in the United States, W.E. Upjohn
Institute, Kalamazoo, MI, 211-283.
Woodbury, Stephen and Rod Anderson (2001), Job-Search, Unemployment Duration and
Subsequent Earnings: Evidence from Randomized Trials, W.E. Upjohn Institute,
Kalamazoo, MI.
386
Community Activism and Change: the Cases
of Sydney and Toronto
Ariadne Vromen
Government and International Relations, University of Sydney
1
Introduction
This paper uses a notion of participatory citizenship to assist in understanding the
political processes apparent in community development work. It engages, particularly,
with recent uses of both civic republican and communitarian citizenship discourses in
depictions of the non-government, or third, sector. It will be argued that we can only
understand the everyday practice of citizenship through an examination of political
actors’ conceptualisations of the utility of, and agency apparent in, community
development.
2
Conceptualising Citizenship and Participation
Until the late 1980s, discussing citizenship was seen as antiquated, and irrelevant to an
understanding of political rights and participation. Yet since that time there has been an
explosion of theoretical and empirical work on citizenship. Bussemaker and Voet (1998)
suggest that the concept of citizenship has been useful as both a theoretical and practical
tool:
In politics [citizenship] is used to reformulate the relations
between citizens and the state as well as the relations among
citizens. In social movements and activist groups, it refers to
questions of inequality, social cohesion and community life
(Bussemaker and Voet, 1998: 279).
This paper engages with this dual political and activist focus of citizenship theorising by
examining how participatory acts of citizenship are understood conceptually and
practically. In order to argue that participation is a separate consideration within theories
of citizenship, I accept the proposition that a discussion of citizen’s rights based on
Marshall’s liberal model leads to an acceptance of ’passive or private citizenship, because
of the emphasis on passive entitlements and the absence of any obligation to participate
in public life‘ (Kymlicka and Norman, 1994: 354). Whilst it is accepted that participation
is based on securing the status and rights of citizenship, the liberal model of citizenship
(see Marshall, 1963 and 1998) does not centre on participation as a primary component in
the construction of individuals as citizens. Other approaches to citizenship, such as civic
republicanism and communitarianism, do place participation at the centre of analysis, and
they are the conceptual focus of this paper.
Vromen, A. (2002), ‘Community activism and change: the cases of Sydney and Toronto’, in T. Eardley and
B. Bradbury, eds, Competing Visions: Refereed Proceedings of the National Social Policy Conference
2001, SPRC Report 1/02, Social Policy Research Centre, University of New South Wales, Sydney, 387410.
ARIADNE VROMEN
Participation has been the underlying focus for civic republicans who discuss
conceptualisations of active citizenship, and the value which participating in political life
holds for individuals (Barber, 1984; Mouffe, 1992; Kymlicka and Norman, 1994;
Bussemaker and Voet, 1998: 284; Oldfield, 1998). Civic republicans see that
participation in decision-making processes by as many citizens as possible is the ideal
form of democracy. The fundamental characteristic of the civic republican approach to
citizenship is that citizens ought to be directly involved in politics, and they have a civic
duty toward the common good to do so. It ’holds that political life – the life of a citizen –
is not only the most inclusive, but also the highest form of human living together that
most individuals can aspire to' (Oldfield, 1998: 79). Benjamin Barber describes
participatory practice thus: ’literally it is self government by citizens rather than
representative government in the name of citizens‘ (Barber, 1984: 151). Barber suggests
that actions such as neighbourhood assemblies, televised town meetings, civics education
and the use of national referendums will facilitate a realisation of civic republicanism’s
version of active citizenship (1984: 261-312).
The language of active citizenship has also been articulated in Australian public policy
debate. There are two main areas where these normative prescriptions of the worthy
Australian public sphere actor have been used: in debates over the provision of welfare
services and/or in discussions of young people and civics education. In the first area,
active citizenship was articulated as a goal of social policy, and was a precursor to the
argument for ‘work for the dole’ programs for unemployed Australians (see Higgins and
Ramia, 2000). In the second area, active citizenship has been discussed as an ideal for
Australian young people, in that they need to be both more involved in, and more
knowledgeable about, political processes. Several federal government reports have
targeted this issue (for example Civic Expert Group, 1993). Both approaches tend to see
public sphere participation in and of itself as a virtue, yet little recognition is given for
participation that is not shaped by existing institutional and structural forms.
Added to this, civic republicans clearly distinguish between the public and private sphere
in a hierarchical way. That is, real participation and subsequent fulfilment can only be
found in the public world of politics; the private life of individuals is merely a hindrance,
not a positive shaping force on public life. These normative dimensions of civic
republicanism are flawed because participation can only be an ideal since very few
people, and particularly very few women, are able to participate in political processes
because of lack of access, resources, interest or recognition of their ability to participate.
These tensions and contradictions are not resolved by civic republican theorists.
Alternatively, communitarian approaches to citizenship argue that citizens are not
atomised individuals but part of a community whereby their status and their participation
is constructed through their ongoing dependency on other community members.
Participation is therefore determined by the relations citizens have with others and with
the communities of which they are members. Communitarians are intrinsically opposed
to the liberal notion of citizenship because of its basis in conceptualising the autonomous
individual as self sufficient and deriving rights from this status. They argue that
liberalism with its individualist view ’overlooks the differences and material inequalities
between members of society and pays too little attention to the social dimension of
388
COMMUNITY ACTIVISM AND CHANGE: CASES OF SYDNEY AND TORONTO
citizenship‘ (Bussemaker and Voet, 1998: 289). Communitarianism, like civic
republicanism, focuses on the politics of the common good (see Winter, 2000: 30-31).
Yet the difference is that, for communitarians, citizens are constituted as members of
communities, rather than as individual public sphere actors as they are by civic
republicans. Communitarianism focuses on common values and norms, and on people’s
’duties‘ towards one another as citizens (Bussemaker and Voet, 1998: 290). There is
often an emphasis on family life, traditional moral values and traditional ways of caring
(Kane, 2000: 220-1). Communitarianism thus highlights the transferring of private sphere
virtues and practices into the public decision-making sphere, which again highlights a
fundamental difference with civic republicanism.
Civil society is seen as crucial to facilitating day to day life, and civil society
organisations are seen to replace the role and responsibilities of government in ensuring
the rights associated with social citizenship. The communities being discussed in this
approach are likely to be geographical neighbourhoods, working environments, school
communities and families; the list could also include organisations in civil society such as
women’s organisations and sports associations, as well as volunteer work in community
service organisations (see Putnam, 1995). Government’s tasks are mainly to create the
conditions to stimulate citizens to take responsibility for their social surroundings
(Latham, 2001: 240-242). Neither paid work through the market nor political
participation, as constructed through, and facilitated by, the state, is able to impart these
particular moral positions. Instead it is the imperative of a separate sphere of civil society,
constituted by voluntary, self-governing associations of citizens (Kane, 2000: 219;
Kymlicka and Norman, 1994: 363). However, for a more ambivalent position on the role
of the state, see Cox and Caldwell, 2000: 65-7. On the role of non-profits, see Lyons,
2001: 189.
The language of communitarianism has been articulated in recent discussions around the
importance of creating ‘social capital’. Social capital refers to the processes between
people that establish networks, norms and social trust and which facilitate coordination
and cooperation for mutual benefit. Social capital is seen as being generated in
communities and it has been argued that ’social capital should be the pre-eminent and
most valued form of capital as it provides the basis on which we can build a truly civil
society‘ (Cox 1995: 17). Robert Putnam, the academic who has popularised the notion of
social capital, differentiates between political participation and social capital. He defines
political participation narrowly as ’relations with political institutions (1995: 665), and
sees that social capital and its complementary act of ’civic engagement‘ are more
appropriate normative descriptions for acts undertaken by members of communities
(Putnam, 1995: 665-7). Debates have emerged out of a fear that society has lost trust, has
lost the development of social capital and a sense of community. (For Australian
research, see Davis, 2001: 220-224; and Hogan and Owen, 2000).
Discussions of social capital lead to a re-configuring of the debate about citizen’s rights
and responsibilities, and thus move away from the liberal and civic republican
approaches. In communitarianism it becomes community responsibility to create and
sustain participation and citizenship practice. It is no longer a role taken up mainly by the
state as facilitator, or seen as citizen action undertaken in the political, public sphere
389
ARIADNE VROMEN
generally (see Turnbull and Fattore, 1999: 235; Everingham, 1999). Rather it is trusting
and sharing communities that are able to provide for citizenship by simultaneously
supplying and fostering the social rights of individuals, because the impetus to participate
is predicated on undertaking caring for, and with, other community members. This could
also be seen to include acts paternalistically described as ‘women’s work’ - that is,
traditional forms of caring work may well be valued highly in communitarian
conceptualisations of citizenship. The quandary for feminists is that the morality inherent
in this valuation will not lead to change in women’s roles in the public sphere, and
instead reinforce the inequitable status quo.
The question for this paper thus becomes how can we best understand the discursive
constructions political actors provide of the ongoing relevance of their chosen location
for political activity? Is it a communitarian construction of the limited role of the state
and the primacy of community connections and social capital? Or is it a civic republican
notion of political action undertaken in the public sphere and the facilitation of active
citizenship? Or do we need a re-application of citizenship theory to better encapsulate
community based participatory politics? The remainder of the paper will explore these
questions by looking at case studies of two community development organisations.
3
Background: Community Development as a Site for Third Sector
Political Action
I chose community development politics for the focus of my empirical research because it
provides an organisational framework for the location of political action, while also
emphasising democratic processes of social and political change. Community
development is a type of politics that is not electoral yet is more institutional and
formalised than social movement politics; and, importantly, is emblematic of a period of
history in the late 1960s and early 1970s where radical political upheaval occurred in
many Western democracies. Community development was predicated on notions of
participation, and sought to involve people actively in the construction of their own lives
through the provision of information and human resources. Further, the practice of
community development was underpinned by the need to redress inequality and poverty.
Thus it is a form of community involvement that epitomises the ‘state, community,
individual’ focus in participatory citizenship, whilst also being related conceptually to
recent discussions of social capital (see Kenny, 1996 and 1999).
The empirical work for this paper is based on case studies carried out with two
community development organisations, one based in Sydney, Australia and the other in
Toronto, Canada. The main source of analytical material is in-depth interviews conducted
with 40 activists1: quotes are used extensively in this paper so as to provide space for the
voices of political actors in the re-telling of their experiences. (For reports on other
dimensions of this research, such as histories of the organisational practice see Vromen,
1
The interviewees have all been given Greek mythological pseudonyms; after each quote
their pseudonym, the city and major period of time involved with the particular
organisation, is listed.
390
1999a; for discussion of the gendered nature of this form of participation see Vromen,
1999b; and in detail see Vromen, 2000).
Community development activism can be seen to be a political process which is
constantly re-negotiated and re-defined, principally in terms of the relationship between
organisations and the state. Thus, over time, these relationships shift in focus and change,
and consequently modes of action used by political actors are altered. In this paper a
typology, developed by Canadian academics Brian Wharf and Michael Clague (1997),
has been utilised to locate and understand the shifts in the practice of community
development, as seen through the eyes of political actors. This typology was useful
because it suggests that not all community development organisations necessarily have
the same practice, whilst also starting to theorise the reasons for the different roles that
organisations adopt. The typology assists in the conceptualisation of the role and practice
of community based actors, as opposed to primarily theorising the role of the state. It
gives more recognition to the agency of community development organisations and actors
in their everyday practice whilst still acknowledging the structural constraints that shape
practice in, and by, organisations. In addition, Sue Kenny has provided useful
classifications of third sector organisations which reflect their operating frameworks.
Whilst a full exploration is beyond the scope of this paper, the everyday practice of both
the Sydney and Toronto organisations clearly falls into the activist framework, rather than
the charity, welfare industry or market frameworks, that she describes (Kenny, 2000).
There are five roles that community development organisations can be seen to undertake
(Wharf and Clague, 1997):
1.
2.
3.
4.
5.
Agent of the State
Partner with the State
Campaigning against the State
Challenging State Institutions and Values
Independent of the State
There are modes of community development organisation’s practice that are specific to
particular categories of the state-community relationship. Four modes of practice are
important to this research: Bullet list social planning, movement activism, advocacy, and
service delivery. The roles of Agent and Partner often include involvement with service
delivery planning and advocacy work, but with differing levels of agency for the
organisation; for instance Agent is often in a more directed, less equal and nonautonomous relationship than Partner. The role of Campaigner focuses on social planning
and sometimes advocacy work. Much of community development organisations’ work
has been predicated on undertaking advocacy work for non-mobilised constituencies such
as public housing tenants or welfare recipients; but, importantly, it has also been about
facilitating political possibilities for these groups in getting them to argue for a
redistribution of resources. The Toronto organisation, in particular, undertook social
planning research to serve as benchmarks and link in with community based organising
(see Vromen, 1999a).
The Challenger role primarily involves protest and movement based activism, but will
sometimes also utilise social planning. This role was realised in the 1970s when
community development became more widespread as a political process. Both the
391
ARIADNE VROMEN
Sydney and Toronto organisations grew out of urban, resident based movements that
were agitating for more participatory structures in government planning, development,
service delivery and so on. The role of Independent has not been applicable to the two
organisations throughout the last 30 years. This is for two reasons: first, that both
organisations have always chosen a role that provides at least some level of active
engagement with the state; and second, both organisations have always had at least some
of their funding from either the federal or state/provincial government. However, it is
important to point out that this role of Independence is often applied to groups that have
both no state based funding and do not seek to advocate to, or change the focus of, the
state. Examples would be some self generated funded charity based organisations or third
sector organisations which are entirely socially oriented, such as sport and recreation
clubs. However, at times, the role of Independent and Agent can become blurred when
community organisations are providing services that are still regulated by, and once were
predominantly provided by, the state.
Through writing histories of the two case study organisations I tracked their relationships
with the state and their dominant modes of practice over the last three decades. The
organisational practice started to diverge in the mid 1990s, and this was because of the
conservative political climate in both locations. Mobilisation and re-invigorated, largescale protest activism emerged in Toronto after community and activist groups were
totally locked out from any advocacy or negotiation role with the provincial government
(see Vromen, 1999a). Sydney did not see the same level of activism and protest despite
the federal Howard government blocking most possibilities for social planning, advocacy
and any sort of genuine, autonomous, partnership roles for community organisations.
This paper moves beyond a description of the historical and organisational processes of
social and political change to examine how political actors conceptualise their ongoing
role in creating social and political change. It reveals the agency of activists in
constructing their own participatory citizenship, and thus serves as an engagement with
discourses of communitarianism and social capital, and civic republicanism and active
citizenship, which have become prevalent in Australian social policy writing on the
community or third sector.
4
Conceptualisations of Community Development and Participation
In general, the community development activists included in this study conceptualise
community as a setting for an ongoing, relevant form of political activism. Many of the
activists see that community development is currently shaped by the conservative
agendas of recent federal and state or provincial governments. The activists advocate a
continuing opposition to these new government agendas by locating community as the
site from which they can both Campaign and Challenge. The modes of action that the
activists discuss include undertaking advocacy work, coupled with an increased focus on
information distribution and social planning. There is also a general call to re-invigorate,
or in Toronto’s case, to successfully harness, the momentum of community-based
activism through new or strengthened political coalitions.
The activists believe that community development organisations are the most appropriate
political setting for an oppositional type of politics, more so than acting in the party
392
based, electoral arena. Interestingly, while many of the activists have been members of
political parties (generally the ALP in Australia or the NDP in Canada), they do not see
these parties as the vehicles for creating social and political change. Unlike community
based activism, party structures are considered to be restrictive and do not provide forms
that activists feel can represent their ideological positions and political values. However,
this discussion about community politics versus party politics is beyond the scope of this
paper.
The following section of the paper looks at the empirical information generated from this
study, and it is in four sections; the first two on current community development practice
in the two locations and the second two on future possibilities for the practice of
community development activism.
5
Empirical Information Generated by the Study
Current Community Development in Sydney
The Sydney activists discussed how government policy shifts constrain both the
community service sector and community development activism. Complaints about
recent restrictive government policy-making were levelled at both the federal Coalition
and the NSW ALP governments. Many people suggested that the approach of
governments had changed, and the shift to a focus on efficiency in the third sector has
reduced the possibilities of democratic participatory action. For example, Bacchus, below
,argues that the notion of participation underpinning community development work has
recently been constrained by neo-liberal government policy-making emphases on
consultation and service delivery (also see Everingham, 1998):
For me the role of participation and the whole struggle was to
ensure that what people needed happened. Sure there were all
sorts of expressive and therapeutic benefits in the process of
being part of a movement, but the aim as it were of being
involved in the movement was to create an environment where
you could do other things. And as community action groups
became more and more incorporated into the state as a sort of
cheap way of providing services, I found the participative less
and less interesting. It was more about consultation within the
parameters already determined. That whole sort of rights
notion of participation has been massively attacked by the right
wing neo-liberalism views of consumerism, the market place
and such (Bacchus, Sydney Board Member, 1970s).
393
ARIADNE VROMEN
The activists also explicitly name and critique the prominence of ‘efficiency’ in policy
and service delivery at the expense of equity. Both Vulcan and Janus, below, suggest that
community development has lost the emphasis on participation by community members
and now tends to be about the delivering of services. That is, it is the services that are
quantifiable in monetary terms that are being given priority by funding bodies:
The threat to community development and to really effectively
involving people in their communities is that all we will fund,
and will value, is particular services. Not the intangibles, not the
things that we can’t measure and account for. You know, we’ll
fund bandaging leg ulcers on 85 year-old blind people but we
won’t fund a process of getting those people together in a form
of community action, which would actually be of benefit to
health. Because you’re involving them in things that matter, and
you’re involving them in their life and their community. We’re
losing all that (Vulcan, Sydney Board Member, 1980s).
Janus suggests that governments are no longer interested in developing communities for
community’s sake and that the policy shift, which all of these activists are highlighting,
has been primarily towards accountability and notions of targeting:
No government is prepared to say ‘here’s a bag of money, go
into an area and do whatever needs doing and just work on
projects for the public good and for the benefit for the people
who need it most.’ That just doesn’t happen. I think there was
more of a tendency to let people do that 10-15 years ago. Now
everything’s got to be targeted and have a dollar sign up (Janus,
Sydney Board Member, 1990s).
These three activists articulate a sense of disillusionment with what they perceive to be
the imposition of the Agent role onto community development organisations. As Wharf
and Clague (1997) note, a community development organisation becomes an agent of the
state when communities only implement plans conceived by governments, and
government thus abdicates responsibility for service delivery. This argument is put
forcefully by Aurora below. She suggests that there have been explicit strategies
employed by government that seek to force community organisations to take a larger
responsibility for service delivery, whilst also eroding their role as the developers of
community. Implicitly, Aurora articulates the virtues of social capital when she talks
about the role of government in ’remov[ing] the fabric‘ of community interconnections.
The other implication to be taken from this statement is that service delivery alone does
not generate opportunities for acting out participatory citizenship:
I think it’s a difficult time for community organisations, what
with national competition policy. It’s the absolute antithesis of
what people in the community are trying to achieve.
It’s not fair on the little groups who are increasingly being asked
to do more and more ridiculous things by government. It’s
394
because what they do has never been valued, you look at a
Meals on Wheels organisation and look at all the exchanges, that
is a huge network of people and if you’re going to remove that
fabric what are you going to replace it with? They’ve got to
come to their senses sooner or later (Aurora, Sydney Board and
Staff Member, 1980s and1990s).
Jupiter, echoing Bacchus’ earlier point suggests that consultation has become a part of the
rhetoric of government in their dealings with the community sector, and their attempts at
fostering Partner relationships. He suggests that the term and its application needs to be
questioned because community organisations do not always have the resources to commit
themselves to the process, on an equal footing with government:
The trouble with words, particularly when governments pick
them up, is they mean different things to different people. So
what the government means by consultation, in practice for most
governments, means here is what we’re going to do and you can
talk to us about it. Another problem is that more often than not
the community sector is not well equipped for well-informed
consultation. It almost negates consultation, you almost need a
core of professionals out there, gearing up to consultation all the
time. So for consultation to happen on an equal basis, on a
negotiating basis, a joint decision-making basis, is
extraordinarily difficult to achieve. (Jupiter, Sydney Board
Member, 1980s)
Thus these Sydney activists see that in the current social, political and economic climate
only Agent roles are being made available by federal and state governments. What is now
seen as de-legitimated is the advocacy and social planning roles community development
organisations have previously played as Partners or Campaigners. These activists
generally would like to see a re-invigoration of both these active citizenship roles, rather
than advocating a new shift towards that of Independent, and a concentration on solely
localised roles for community development organisations.
A smaller group of the activists suggested that change in community-based activism has
less to do with shifts occurring in government, or the social and political environment
than with change to the make-up and political motivations of geographically-based local
communities. Gaia believes that communities have changed and do not possess the same
cohesive base they once had. Thus she is speaking the language of communitarianism in
terms of the notions of the loss of a sense of shared community. She suggests that it is
self-interest alone that now motivates people to be involved in community-based protest
activity. Whilst Gaia is referring to a change in the geographical community she lives in,
she explicitly blames ‘capitalism’ and the lack of successful policy-making initiatives
being undertaken by governments to address issues of inequality and poverty. Therefore
Gaia sees that it is larger structural issues that shape the behaviour of, and possibilities
for, community based activism:
395
ARIADNE VROMEN
I find that people are not participating now as much as they
were even five years ago. You could go down the street and
door knock and get people to come to something. There’s been
a big change and people are cynical. I find now that they’ll
come if it’s something that’s really going to interest them and
affect their lives.
I think people just think that it’s not going to get any better.
And you can’t blame them, even people getting a bit of money
each week there’s no value in that money. You know, you can
get a feed somewhere or go to one of the charities somewhere
or get some clothing, but that’s not really what people want.
They want a quality of life, and a decent life where they can
see something, and a future and a roof over their head that’s
going to be permanent. And that they don’t have to put
themselves in hock forever. That’s the dilemma that we’re in,
because this is capitalism and this is the whole crux of the
matter, and this is just bandaid treatment, all you’re actually
doing is propping them up (Gaia, Sydney Board Member,
1990s).
While nearly all the Sydney activists saw that the present was a difficult and constraining
time for community development activism, several were optimistic that the broader
political climate would change and become more conducive to the practice of community
development. Mars, below, explains that although the communities themselves have
changed since the initial periods of community development practice in the 1970s, the
basic principles of community development work are still integral to current practice at a
local level. These principles, or modes of action, underpin the social planning and
advocacy work which is fundamental to the Campaigning Against the State and/or
Partner With the State roles of community development organisations:
The inner city communities have changed. Where you have
communities that have changed you have to continually invest
in community development, because if you’re having new
people coming in and there’s a continual necessity to try and
build links with people so that people relate to the local
business, relate to their neighbours and relate to the structures
and the web of relationships in the community. Where you
have rapid change with new coming people you’ve got to try
new gimmicks to try to integrate those people with the other
people in the community.
In stable communities the webs are there and you build on
them and reinforce them, and maybe create new ones. But
when communities change it’s trickier. I basically think you do
it to life stage techniques, you do it through schools, and
mother groups, and child care centres, basically people at life
stages have different interests. Most of the techniques are
396
really life stage focused (Mars, Sydney Staff and Board
Member, 1970s, 1980s and 1990s).
Mars’ comments are clearly related to notions of social capital. That is, community
development practice is seen to be predicated on social capital generation, and this can be
seen in the concepts of localism that Mars articulates. Mars’ suggestion for a renewed
focus on ’life stages‘ is clearly about targeting women as members of communities as
they are the people most likely to interface with the schools and child care centres that he
mentions. However this quote should not be understood simply, as it also taps into ideas
about the broader roles aimed at altering the existing structures of the state, and this is
incompatible with existing communitarian discourse that prioritises autonomous
community based decision-making processes.
Current Community Development in Toronto
The interviews with the Toronto activists revealed that they had similar misgivings to the
Sydney activists about the current political, social and economic environment in their
city. Many activists reflected on the constraints presented by the Progressive
Conservative government in the province of Ontario. However several people were also
optimistic about new community-based developments, such as the activism undertaken
mainly by two groups: the Citizens for Local Democracy (C4LD) and the Metro Network
for Social Justice (MNSJ). This presents a strategic difference in the conceptualisation of
community between the Sydney and Toronto activists. That is Toronto has recently
witnessed a successful re-invigoration of community development organisations’ role as
Challenger, whilst Sydney activists are only hopeful for a similar groundswell of support
in opposing conservative governments. Whilst recent examples of activism in Toronto
permeate the way activists discuss the future possibilities for community development
activism in their cities, their ideas are not only about engagement with the state. Activists
also reflect on the communitarian concerns of galvanising communities based in shared
norms and values.
Several of the Toronto activists discussed Ontario’s Progressive Conservative
government and the shifts in social policy delivery that ensued under the provincial
government. Aphrodite’s comment below reflects on how the political organisation
facilitated by the Metro Network for Social Justice, which culminated in large proteststhe Days of Action, is still pertinent, and provides the space for community development
organisations to engage as Challengers:
One needs to be there even if you feel like you’re never heard.
So we’re in one of those eras where you’re holding a line to
get through and you don’t get to move ahead or improve, but at
least you hopefully don’t get creamed.
...I’m more inclined now to think that we need to keep
organising, and keep talking, keep that capacity out there, that
voice, we won’t always be galvanising people every day into
wonderful things. But we need to keep our own capacity going
397
ARIADNE VROMEN
and keep the learning going and watch for those political
moments. (Aphrodite, Toronto Board Member, 1990s).
Heracles commented directly on the policy changes that have been enacted by the
provincial government, and how now is a time when it has become imperative to just
hold ground, without being too optimistic about the possibility of structural change,
despite prevailing social justice concerns. To some extent, Heracles articulates an
understanding of community development’s current situation as being caught, in
practice, between Agent and Partner roles. He suggests that it is the responsibility of
community development organisations to pragmatically locate what is ’deliverable‘ in
the current times, despite any conflict this may have with previously held ’left
ideology’:
The Tories entered [government] and they’re just not interested
basically, and so you spend all your time just keeping bad
things from happening, instead of attempting to make good
things happen - which is very demoralising at times. The
difficulty is, well I guess I have more appreciation now than I
probably would have had 10 or 15 years ago, for the notion of
a ‘bottom-line’, in the positive sense of that word.
I think that I have got a much stronger appreciation of how you
create workable solutions that make a difference. In my head
I’ve got a lot of left ideology that isn’t going to change, so
that’s always going to partly frame how I look at the world.
But I’m much more interested in what’s deliverable and adding
value, and I think that’s a major thrust that anyone in an
advocacy world these days has to respond to. The only way
you’re going to break through that is with very practical ideas
and I don’t think ideology is going to cut it any more.
(Heracles, Toronto Staff Member, 1980s)
Heracles articulates a different view from that proposed by several of the Sydney
activists such as Aurora, Vulcan and Jupiter. The Sydney activists see that it is
imperative that community development organisations

full document - Social Policy Research Centre

Transcription

Similar documents

Impact of Piracy on Cultural Diversity of European Music

File - THE BONGO BROTHERS

Labour Market Information

The euro banknotes and coins

INTERNATIONAL STUDENT PROSPECTUS

AIR Summer 2016 Press

acla.com - LexisNexis

newsletter - Classroom Connections

inxs final - Seven West Media

04 Japan - Alliance for Healthy Cities