at3032

Computer Integrated
Manufacturing and Employment:
Methodological Problems of
Estimating the Employment Effects
of CIM Application on the
Macroeconomic Level
Ayres, R.U., Brautzsch, H.-U. and Mori, S.
IIASA Working Paper
WP-87-019
February 1987
Ayres, R.U., Brautzsch, H.-U. and Mori, S. (1987) Computer Integrated Manufacturing and Employment: Methodological
Problems of Estimating the Employment Effects of CIM Application on the Macroeconomic Level. IIASA Working Paper.
IIASA, Laxenburg, Austria, WP-87-019 Copyright © 1987 by the author(s). http://pure.iiasa.ac.at/3032/
Working Papers on work of the International Institute for Applied Systems Analysis receive only limited review. Views or
opinions expressed herein do not necessarily represent those of the Institute, its National Member Organizations, or other
organizations supporting the work. All rights reserved. Permission to make digital or hard copies of all or part of this work
for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial
advantage. All copies must bear this notice and the full citation on the first page. For other purposes, to republish, to post on
servers or to redistribute to lists, permission must be sought by contacting repository@iiasa.ac.at
WORKING PAPER
C O - W U T E R I N T E G R A T E D MANU-E'ACTURING AND
EMPLOYMENT :
-YETHODOLOGICAL PROBLEMS O F E S T I M A T I N G
T H E EMPLOY-XENT E F F E C T S O F C I X
A P P L I C A T I O N ON T H E MACROECONOMIC L E V E L
R.U. A y r e s
H.-U.
Brautzsch
S. Mori
February 1 9 8 7
WP-87-19
International Institute
for Applied Systems Analysis
NOT FOR QUOTATION
WITHOUT PERMISSION
OF THE AUTHOR
CO-WUTER INTEGMTED MANUFACTURING AND
EMPLOYMENT:
-Yl3THODOLOGICAL PROBLEMS OF ESTIMATING
THE EMPLOY-WNT EFFECTS OF CIM
APPLICATION ON THE MACROECONOMIC LEVEL
R.U. Ayres
H.-U. Brautzsch
S. Mori
February 1987
WP-87-19
Working Papers are interim reports on work of the
International Institute for Applied Systems Analysis
and have received only limited review. Views or
opinions expressed herein do not necessarily represent those of the Institute or of its National Member
Organizations.
INTERNATIONAL INSTITUTE FOR APPLIED SYSTEMS ANALYSIS
A-2361 Laxenburg, Austria
Foreword
This
newly
paper
is
established
Project,
of
one
of
the first research products of the
Computer
which
Prof.
Integrated
Ayres
issues of occupation-by-sectar
comparability, and
Manufacturing
is the leader.
data
suitability for
with
robotics on employment.
session
of
the
some
The paper
American
Orleans, December 30, 1986.
available to
early
Economic
It addresses
availability, international
use with
Methods of estimating labor substitutability
discussed, along
(CIM)
formal 1-0 models.
by
estimates
CIM
are also
of the impact of
was formally
presented at a
Association
meeting in New
As an I IASA working paper it will be
collaborating researchers and institutions in other
countries.
V. Kaftanov
Deputy Director
R.U.
Ayres
H . -U.
Brautzsch
S. Mori
Computer Integrated Manufacturing and Employment:
Methodological Problems
of Estimating
the Employment Effects of
CIM Application on the Macroeconomic Level
1. Problem Statement
Unquestionably,
a d v e n t of
some
of t h e
computer i n t e g r a t e d
labor force impacts.
Relevant
labor
structural
displacement,
changes i n work-content,
The
importance
of
and
most i m p o r t a n t
manufacturing o r
CIM c o n c e r n t h e
aspects
problem i n c l u d e
of
changes
changes
these
problems
the
in
in
the
the
Maier,
Hunt,
1983;
Ayres &
1981, 1985; Leontief
& Duchin,
T h e r e is g e n e r a l a g r e e m e n t
level
the
application
of
CIM
Engelberger,
1983; Ota,
fact that
1986).
on t h e micro-
accompanied by d i r e c t l a b o r
is
displacement and changes i n s k i l l requirements.
number o f s e m i - s k i l l e d
A
f a i r l y large
o p e r a t i v e j o b s is b e i n g e l i m i n a t e d , t h o u g h
A f a r s m a l l e r number o f h i g h l y s k i l l e d j o b s
gradually.
1988;
1984; Haustein &
1 9 8 6 ; a n d Kaya,
on t h e
force,
r e f l e c t e d i n a n immense
is
Miller,
labor
work e n v i r o n m e n t .
number o f p a p e r s a n d b o o k s ( s e e , f o r e x a m p l e ,
Hunt 8
i m p a c t s of t h e
created at the same t i m e .
is b e i n g
T h e r e is a l s o a n a g r e e m e n t w i t h r e g a r d
t o t h e f a c t t h a t t h e c u r r e n t s t a t e o f CIM a p p l i c a t i o n h a s n o t y e t
led
to
any
significant
changes
employment s t r u c t u r e ( e . g .
l e v e l of
is
the
a n a t i o n a l economy.
typical
of
major
s i g n i f i c a n t impact
significant
the
employment
qualification
T.
level or
structure)
Vasko (1983) e m p h a s i z e d :
innovations:
They
begin
to
a t the
"This
have
a
on c e r t a i n b r a n c h e s e v e n b e f o r e macroeconomic
i n d i c a t o r s become r e s p o n s i v e .
any
in
impact
on
T h e r e f o r e i t is d i f f i c u l t t o p r o v e
(always
aggregated)
macroeconomic
data. "
O p i n i o n s a b o u t t h e medium
o f CIM
d i f f e r considerably.
of j o b replacement accompanied
authors are
of t h e
application
will
and long-term
employment i m p a c t s
Some a u t h o r s e m p h a s i z e t h e p r o b l e m
by
higher
unemployment.
opinion t h a t t h e labor-saving
lead
to
higher
productivity
Other
e f f e c t s of CIM
and
income,
resulting
in
higher
domestic
competitiveness) ultimately
demand
creating
a
(.and improved
net
increase
sxport
in total
employment.
These different assessments of the employment effects of CIX
applications are supported--among
different
forecasting
other
methods,
diffusion rates and the
things--by
different
application
the
assumptions
potential
of
CIM
use of
on
the
ana the
expected productivity effects of this technology.
The subject
of this
disadvantages of the
employment
effects
paper is to analyze the advantages and
input-output
of
CIM
approach
for
estimting the
application and to discuss the main
directions of investigations to these problems at IIASA.
2.
Input-Output
Analysis:
An
Approach
for
Estimating the
Employment Impacts of CIM
Advantages
and
disadvantages
estimating the employment effects
as CIM
of
different
methods
of technological
for
changes such
are discussed repeatedly in the literature (Brooks, 1985;
Friedrich & Roenning, 1985;
Informationstechnologie, 1988).
In
this connection Brooks characterized the Input-Output analysis as
the
approach,
which
"provides
projecting employment
the
effects of
most
rigorous
method for
new technolo~iesbecause it is
capable of accommodating economy-wide effects arising
out of the
linkage among sectors and thus of tracing through the system-wide
impacts of introduction of a particular technology."
The first attempt to use an
input-output model
in order to
estimate economic impacts of microelectronic application was made
by Fleissner et al. (1981, in
161,163,164) commented
in
W. Leontief
Austria.
regard
to
this
study:
(1982, pp.
"Although
current business publications, trade papers and the popular press
abound
with
articles
speculation on
about
the economic
"automation"
impact of
and
"robotics"
and
these developments, only
the governmental and scientific agencies of Austria have produced
a systematic
present
assessment of
revolution
in
the prospective
labor
industrial economy and society
been
completed
for
the
presents the Lest model
U.S.
saving
...
consequences of the
technology
No comparable
economy
available for
...
in
a
modern
study has yet
The Austrian study
projections of conditions
in the U. S. of 1998."
Leontief and Duchin (1986) subsequently publisheci a study in
which the
impact of
analyzed using
computer-based automation
an input-output
on employment i s
the U. S.
model for
This model
differs from that of Fleissner in three important ways:
a)
In the Leontief model
the
vector
of
non-investment Tina1
demand is provided from outside the model.
estimated the final demand with the
Fleissner et al.
help of
an econometric
model which i s linked with a demographic and an input-output
mode 1.
b)
While Fleissner et al.
used
a
static
input-output model,
Leontief and Duchin developed a dynamic lnput-output model,.
c>
In
the
Fleissner
subdivided
by
model
sex
the
and
whereas Leontief/Duchin
four
sectoral
formal
labor
forces
educational
are
levels,
used a more detailed occupation-by-
sector matrix ( 5 3 occupations).
One drawback of both models is that they do not
feedback of
reflect the
reduction achieved by CIM application to a
the cost
possible demand increase resulting from lower prices of goods.
A study, in which the approach of Fleissner et al.
was made
Curdy t 19Y5a,b>.
by Mc
output model,
calculate
which
the
is
relative
Howell
similar
industry
to
(1986) used an input-
the
and
was used
Leontief
model, to
occupational
effects of
alternative levels of the use and production of industrial robots
in the U.S.A.
As these
examples suggest, the main advantage of the input-
output approach consists in the
employment effects
also of the
of CIM
effects
consideration
application in
which
are
caused
not
only
of the
a certain sector, but
by
CIM
production and
application in other sectors of the economy.
But one
models not
has to
all
estimated. Some
consider that with the help of input-output
important
of the
effects
2IM
of
application
methodolo~ical limitations
can be
which can be
observed in the above mentioned studies should be mentioned:
a)
In
input-output
models
only
attributes
of
flexible
automation equipment can be considered that can be reflected
in the parameters or
the variables
technological coefficients,
or in
t3e
arises:
in
final
what
demand
way
model, e.g . the
in the labor input coefficierlts
sector.
can
of the
the
However,
effects
of
the quest ion
the increased
'Strictly speaking, it is quasi-static, since
time variation is introduced exogenously.
most of the
L I M be reflected in the model'?
flexibility of
p.5) has noted:
Vasko ! l i j S 3 ,
"There is no established way to measure the
flexibility of the flexible systems."
b,
In
input-output
models
"average" technology
aggregated)
causes
of
production
exceptional
each
the
technology
corresponding
process.
effects
represents
An
which
311
(.more or less
innovation like C i M
can
not
be adequately
reflected in "average" technologies.
c>
In what
way can
such effects as changed work-content, work
C I M application,
environment, etc, which are conditioned by
be reflected in the model?
In
the
literature
these (hardly quantifiable) effects are
especially emphasized.
important to
A s we
have
said
elsewhere,
it is
emphasize--more than once if need be--that the
societal importance of various issues may well be in inverse
ratio to their quantif lability. (Ayres, 1986j.
d,
How
can
the
employment
effects
of
CIM
application
" isolated"'? The current industrial revolution
a
"cluster"
influence
of
on
basic
the
innovations
employment
be
is forced by
which commonly have an
development.
Besides, the
evolutionary development of labor skills and demand patterns
is conditioned by structural
and organizationa'l factors as
much as by technological ones.
e)
Input-output tables
out of date.
intensive
(at least
This i s
work
in the
conditioned by
required
to
process
,
U.S. are many years
the time
the
and labor-
necessary data.
Fleissner et al. used in their study, which was published in
1981, the
input-output tables from the years 1970 and 1976.
Leontief/Duchin used tables from 1967, 1972, and 1977.
This
led
(see
to
severe
problems
in
parameter
forecasting
Friedrich, Roennig, 1985).
f )
On the microeconomic level the effects of application of CiM
are likely
technology.
to be very high in comparison to the traaitional
But these enormous effects on the microezonomic
level will not immediately be "transferable" to the sectol-a1
and the macroeconomic level to
the
same
extent
(see also
Ayres & Killer, 1983,-:. The effects on the sectoral and the
...,
--Haustein& Maier
(1985, called this the "transformation
problem" of the projected dynamical efficiency into a real push
of the average efficiency.
macroeconomic ievel can be
would
lie
within
the
so low
error
in soma
margin
cases that they
of
the
parameter
estimation of the input-output model.
With regard to these
be aware
methodologica? limitations
that any model reflects only a "facet" of reality.
impact of CIM has so many
aspects
completely
any
other
one should
reflected
complementary
by
that
it
is
input-output
approaches
and
unlikely
model.
models
The
to be
A number of
will
be
needed.
Nevertheless, with regard to the estimation of the impacts of (:IN
on
the
level
of
employment,
occupational structure
educational
qualification
arid
of the labor forlze, input-output analysis
i s a powerful approach.
3.
Computer Inte~rated
Manuiacturing and Bmployment :
i~irections
of Ressarch in the CIM Project at IIASA
Bsaring
in
mind
the
backgroun~d outlined
investigation this far has
been
concentrated
above, our own
on
the foilowing
three problems:
a)
The development
of an
estimate tne impacts of
approach to
CIM on employment by occupation;
b)
The computation
of detailed
and internationally comparable
labor matrices (occupation-by-sector matrices);
c>
The
linkage
of
the
labor
output models which are
matrices to the related input-
included
in
the
existing INFORUM
system (Almon, 1977).
Estimation of
the Impacts
of CIM Application on Labor Forces by
Occupation
A preliminary remark is appropriate:
the
influence
structure
of
heterogenous
of
technological
employment,
diversity
of
comparatively influenced by
it
In order
progress
is
working
on
necessary
to
to estimate
the occupational
summarize
places into groups that are
technological
progress.
For this
purpose it is helpful to define similar tasks or occupations
Tasks are
generaliy more
the
.
descriptive of the actual work-content
'The difference between the terms ta.k
and occu~ationcan be
exampie:
The task
simply
explained
by
the
following
"programming" may be done by people with different o~zcupations.
The occupation "programmer" is characterized by doing the task
"programming" in the majority of one's worxlng time.
of a J o b (Warnken, 1986).
the
labor
force
by
On the
,ion
the subcilvi-
other hand
occupation
has
the
advantage
that
of
it
establishes a direct connection with educationai planning.
Hence,
in
technological
order
to
progress
estimate
on
the
the
level
influence
and
the
of
structure
employment and to infer the consequences for education,
be very
useful to
have data
sectors
and
Such a
detailed data
the
of
it would
on the occupational composition by
by tasks as well as the task composition
by sectors.
basis is--to our knowledge--available oniy
for the FRG (Figure 1).
However relationships
between tasks and
occupations are likely to be reasonably similar in countries of a
comparable level of economic development.
Task-by-sector
countries,
but
matrices
the
are
available
only
for
a
few
occupation-by-sector matrices are available
for many more countries.
The following indicators have
to be
considered to estimate
the impact of CIM on employment-by-occupations.
The fractional share of the workers in a certain occupation
potentially affected by the
application
of
certain CIN
a
-
technology (e.g. robotics or CAD) ;
The fractional share of affected workers actually displaced;
-
The resulting increase of labor productivity attributable to
this technology.
Data
about
technologies
the
by
replacement
different
potential
occupations
of
certain
and sectors can best be
determined on the basis of engineering
analysis.
this
engineering
approach
different
follows.
countries
are
information permitting
Detailed
currently
estimates of
CIM
not
An example of
studies
available,
but
for
some
this kind will be sought in
the IIASA project.
Data on the number of machine tools in use, by
by type
of control,
Machinist (Mc
38).
is collected
Graw-Hill) for
The 13th
survey was
category and
every 5 years by the American
each metalworking
published in
sector (SIC 33-
1983 and the 14th will
a2pear in 1588.
In his PhD thesis
tools into
S. Miller
(1983) classified
all machine
4 categories, a s shown in Table 1, below.
allocation is
$:ven
in
percentage
of
all
principle,
be
operated
Appendix
machine
by
tools
level
He
1.
in
I
the
also
A detailed
estimated the
U . S . that could, in
robots
(roughly,
1982
Sectors (99)
Figure 1.
Tasks (96)
System of labor matrices in the FRG.
Table 1.
Low and High Estimates of the Distribution of
Metalcutting Machine Tools By Category
Category
Percent of
Percent of
Machines
Machines
(Low Estimate)
High Estimate
.........................................................................
Category 1
39.4
68.2
9.4
46.7
(Machines designed for
low volume production)
Category 2
(Machines designed for
fully automatic operation)
Category 3
(Machines designed for very large
and/or heavy workpieces)
Category 4
(Machines designed for
medium to large batch production)
............................................................................................
technology) and
Table 2 .
chart
by l e v e l
Combining t h e r e s u l t s
(Figure
the
This
2).
numerical c o n t r o l
of
I1 robots
(roughly, 1 Y G 8 ' s technology),
i n graphic
suggests
that
(and r o b o t i c o p e r a t i o n )
existing
machine
a b o u t t h e u p p e r l i m i t of
form y i e l d s
the pie
upGer l i m i t f a r
the
p r o b a b l y a r o u n d 48%
is
tool
population,
machine
operator
w h i c h would a l s o be
displacement.
compares w e l l w i t h a n e a r l i e r i n d u s t r y survey--admittedly
i n scope--carried
at
out
Carnegie-Mellon
This
limited
University
(Ayres B
1983) which s u g g e s t e d t h a t r e s p o n d e n t s t h o u g h t t h a z 3 9 . 5 %
Miller,
o f o p e r a t i v e s c o u l d be
r e p l a c e d by
1 3 . 6 2 c o u l d b e r e p l a c e d by a l e v e l
The
above
results
can
validation f o r t h e survey
s u r v e y (of
be
a level
(but only
I robot).
regarded
methodology.
as
A far
a
c r u d e s o r t of
more f a r - r a n g i n g
474 r e s p o n d e n t s ) was c a r r i e d o u t i n 1 9 8 4 by t h e J a p a n
I n d u s t r i a l Robot A s s o c i a t i o n J IRA ( J IRA,
focussed
I1 robot
on
the
number
of
workers
r o b o t s by t a s k s a n d
by s e c t o r s .
labor
matrix
displacement
i n d u s t r y c a n be e s t i m a t e d .
The
1985).
JI2A s t u d y
r e p l a c e a b l e by i n d u s t r i a l
Eased
on t h i s ,
the potential
f o r t h e whoie J a p a n e s e m a n u f a c t u r i n g
I t must be n o t e d t h a t t h e J IRA s u r v e y
c o v e r s o n l y a s m a l l p a r t o f t h e J a p a n e s e i n d u s t r y , a l t h o u g h i t is
much more
results
comprehensive
for
are
Japan
than
the
Ayres/Miller
summarized
in
Table
Assuming J IRA'S s u b s t i t u t a b i l i t y d a t a t o b e
manufacturing industry,
of t h e U . S .
3 (columns 1 , 2 ) .
similar t o
t h e U . S.
(Table 3, columns 3 , 4 ) .
We finally
r e s u l t s w i t h t h e e s t i m a t e s by A y r e s & M i l l e r
D e t a i l s of t h e p r o c e d u r e
i n columns 5 , 6 .
J IRA
the potential for labor substitutability
is a l s o e s t i m a t e d
compare t h e
survey.
a r e given
!1983),
i n Appendix
2.
Although
the
classification
d i f f e r e n t b e t w e e n t h e s u r v e y s of
(1983),
it
consistent.
can
be
I t is
r a t i o estimated
concluded
noteworthy
from JIRA's
of
occupations
JIRA
(1985) and
that
the
that
the
s u r v e y is
disc:ussion.
in
the
rather
Ayres & M i l l e r
results are basically
p o t e n t i a l displacement
roughly within t h e range
f o r l e v e l I a n d l e v e l I 1 r o b o t s g i v e n by Ayres B M i l l e r
Another approach t a k e n
is
JIRA
survey
(1983).
a i s o deserves
The 4 7 4 r e s p o n d e n t s w e r e a s k e d ( . i n e f f e c t ) how much
t h e y would b e w i l l i n g t o p a y i n c a p i t a l c o s t s t o r e d u c e t h e t o t a l
number
of
workers
by
one.
m a r g i n a l c a p i t a l v a l u e of a
D a t a is
presented i n
This
robot
Figures 3 , 4
can
system
be i n t e r p r e t e d a s t h e
per
worker r e p l a c e d .
f o r v a r i o u s t a s k s i n t e r m s of
Table 2.
Estimates of the Percent of Metalcutting Machine
Tools That Could be Operated by Level I and Level II robots
MACHINE TYPES ASSIGNED TO
Percent of
All Machines
in Metalworking Industries
------------..-----..--..-..-----.---..-----.-..--..----..-.-.~-.--.-..-.~-....Category 4 only
9.4
Categories 4 and 2
6.3
Subtotal
15.7
--
Max. f o r l e v e l I
robot
46.7
--
Max. f o r l e v e l I1
robot
Machines Which Could be O p e r a t e d
by a L e v e l I r o b o t
9.4
-
Machines Which Could be O p e r a t e d
by a L e v e l I1 r o b o t
46.7%
Categories 4 and 1
and Categories 4 and 3
Total, Category 4
(exclusively and jointly)
15.7%
Specialized
LargeIHeavy
Workpiece
1.1 1.7%
Fully
Automatic
Cumulative %
Automatic or
LOWVolume,
40.5%
v
Low Volume or
Batch
30%
Min P o t e n t i a l f o r KC
Max P o t e n t i a l f o r NC
Min P o t e n t i a l f o r A u t o m a t i c
Max P o t e n t i a l f o r A u t o m a t i c
125
20%
Min P o t e n t i a l f o r M a n u a l
N a x P o t e n t i a l f o r Manual
F i g u r e 2.
C l a s s i f i c a t i o n o f machine t o o l s by u s e and c o n t r o l
( b a s e d on K i l l e r , 1973)
Table 3.
C o m ~ a r i s o no f L a b o r D i s p l a c e m e n t E s t i m a t i o n
i n M e t a l Working I n d u s t r y ( i n 1000 w o r k e r s )
u p p e r : p o t e n t i a l d i s ~ l a c e m e n tw o r k e r
midd1e:Ctotal em~loymentl
lower : ( p o t e n t i a l displacement r a t i o )
[I]; L e v e l I r o b o t ( n o n - i n t e l l i g e n t r o b o t )
CII1;Level I1 robot ( i n t e l l i g e n t robot)
SIC33-38;primary metal, f a b r i c a t e d metal products, general machinery. e l e c t r i c machinery.
t r a n s p o r t a t i o n machinery and p r e c i s i o n machinery
SIC34-37;fabricated metaL products, general machinery. e l e c t r i c machinery
and t r a n s p o r t a t i o n machinery
Japan ( J I R A )
SIC34-37
U.S.
(a)
SIC33-38 I SIC34-37 SIC33-38
U.S.
(b)
SIC34-37
CII
c a s t ing
14.9
C41.81
(35.6%)
d i e c a s t i ns
18.0
C28.51
(63.2%)
~ C a s t i cfanning
22.4
C63 - 71
(35.2%)
heat treatment
23.0
C113.01
(20.4%)
f o r g i ng
press & shearing
54.3
C215.61
( 25.2%)
CIII
SIC33-38
CI1
CIIl
Table 3 .
(continued)
U.S.
Japan ( J I R A )
(a)
SIC34-37 SIC33-38, SIC34-37 SICS-36 1
I
64.9
C180.91
(35.9%)
77.2
C211.21
(36.6%)
i
18.4
22.2
C60.21 C66.51
(30.6%) (33.4%)
(b)
SIC34-37
CII
I
painting
U.S.
CIII
SIC33-38
CIII
CII
32.7 49.1
34.6 51.8
C74.41
C78.51
(44.0%)(66.0%)(44.1%)(66.0%)
grinding &
machining etc.
assembly
Loadi ng &
~ a c k a gng
i
subtotal
others
-
( a ) Based o n t h e J I R A S u b s t i t u t a b i l i t y d a t a
( b ) Based on A y r e s - M i l l e r
-.
..-
--
(1983)
-
.-
10. P l a t i n g
9. P a i n t i n g
8. S p o t
welding
4.
----
0.97
'--'-----*'
1 .O1
4
1 2 . Assembly
0.87
*------
Heat
.-
5. Forging
1 . C a s t i n g ,,,,,
1 3. L o a d i n g &.,
Packaging
7. A r c
welding
15. O t h e r s
. - - - - - - --
..------- - - - -
forming
(mean v a l u e )
- I - - - '
---
'".
1.62
v
F i g u r e 3.
E n t r e p r e n e u r ' s w i l l i n g n e s s t o i n v e s t t o r e n l a c e one
worker
(averaqe robot p r i c e excl. system c o s t )
.
0
Cn
d
2
h,
P
Cn
0
.
10. P l a t i n g
5.
Forging
9. Painting
12. Assembly
.--.I-----
8. Spot
welding
1. Casting
4.
heat
treatment
6. Press &
shearing
13. Loading &
Packaging
15. O t h e r s
- - ------
7. Arc
welding
.--I---
1 4 . Inspection
2. D i e c a s t i n g
11 . IIachininq
.,--,- - -- - - - -- -.-- -- - - - -- - - - - 2.17
3. P l a s t i c s
forming
(mean v a l u e )
Figure 4.
.I----
-- - - - - -
r-0
.71
E n r e p r e n e u r ' s w i l l i n q n e s s t o i n v e s t t o r e ~ l a c eo n e
worker
(average robot p r i c e i n c l . syster. c o s t )
.
ratios between the average marginal
replaced worker
the average
tasks
the
ratio
a robot.
displace only
of
a single
that for most
than unity, implying that ceteria
paribus robots were economically
they could
is noteworthy
It
greater
is
value
by managers or entrepreneurs) and
(.asperceived
cost of
capital
justified
a single
in
Japan
worker.
(i484) if
In most cases, the
observed displacement ratio i s closer to one worker per shift, or
nearer to 2 workers per robot.
already clear that not all workers are substitutable,
It is
even for the most routine tasks.
to-pay data
presented in
Thus, the marginal willingness-
Figure 4
might be
measure of distance from equilibrium.
regarded also as a
If all
justifiable robots
were actually in place, the theoretical ratio should be 8 . 5 "
+ 8.1.
A high ratio suggests that the potential for substitution is much
higher than
the current level of penetration.
Conversely, a low
ratio suggests a very low potential for substitution.
This procedure allows one
range of
to
labor substitutability
get
an
due to
"impression"
CIM.
of the
This procedure is
not necessary if detailed engineering surveys about the potential
labor
substitutability
by
sectors
and
occupations
beccmes
available and more careful computations can be made.
The
Elaboration
of
Detailed
Internationaliy
Comparable Labor
Matrices
Application of
the input-output approach for estimating the
employment effcts of CIM application requires
of
detailed
occupation-by-sector
the reconciliation
matrices
for
different
countries.
Only two prior studies
matrices are known to us.
on internationally
comparable labor
In 1Y69i78 the OECD published a set of
highly aggregated labor matrices
for
53
countries.
The most
sophisticated study was carried out at the World Bank by Zymelman
(1988)
which
sectors
for
work
has
analyzed
26
not
internationally
matrices
countries
been
128
occupations
around the year 1978/71.
updated.
comparable
with
The
problems
matrlces
labor
of
are
and 58
Zymelman's
constructing
discussed
in
Appendix 3.
The main objectives of this task are the following:
a)
The creation of a
direct
employment
data
base
for
displacement
the
computation
effects
of zhe
of CIM by methods
discussed above.
(.Thesubstitutional potential
Zor a given
CIM technology must be referenced to a standard occupational
and sector classification. )
b)
The investigation of
matrices which
possibilities
for
synthesizing labor
are not available from primary sources te.g.
census or micro census).
It must be recalled that labor
matrices are
available only
for a limited number of countries.
If
one
can
find
recognizable
similarities
industry/occupat ion patterns between different
is possible
in
countries then it
to extrapoate countries for which labor matrices are
not available.
Zymelman
(1985) emphasized
that
is a
there
plausible relation between the labor productivities of industries
tsectors) and their occupational structures.
be used
data:
Two approaches can
to synthesize occupational structures from international
judgmental tcomparative)
method,
relationships
assumed and used to
patterns that
between
infer the
are known.
In
and statistical.
occupation
patter for
In
the firs.t
and productivity are
an unkown
case from
the statistical approach, average
coefficients for occupation by sector can be determined by crosssectoral analysis.
much
more
The first method is preferable, but requires
analysis.
represents
intention
the
is
occupations as
Unquestionably,
current
to
use
the
state-of-the-art
the
same
work
in
of
this
nomenclature
for
Zymelman
field.
Our
sectors and
Zymelman to obtain a consistent series covering 3
decades.
The Incorporation of the Labor Matrices in the INFORUM System
In the literature one
application
of
CIM
can find
will
lead
international division of labor
could be
conjectures that
to
important
(see e.g.
the broad
shifts
in
Sadler, 1981).
the
This
caused, for instance, by the increasing competitiveness
of CIM users.
It i s
negation of
the lost
because
the
in
widely
assumed
that
this
could
lead to
advantages of so-called low-wage countries
developed
countries
workers can be largely replaced by CIM.
in important cost reductions.
highly-paid
semi-skilled
If so, this could result
In consequence, one
might foresee
an increasing gap between developed and developing countries.
Confirmation
of
such
hypotheses requires the extension of
economic anaiysis
to include
international trade.
Perhaps the
only suitable instrument available today is the so-called IXFGHUM
system which was designed at the University of Maryland under the
leadership
of
Almon, 1988).
a
number
Clopper
Professor
Almon
(Almon,
1979; Nyhus
An important part of this system is the linkage of
of
national
input-output
models
for
key
trading
countries using special trade models.
Our objective of the present CIM activity at
partly in
the linkage of the occupation-by-sector mmatrices with
the corresponding national input-output
the INFORUM system.
only developed
the INFORUM
countries.
application
developing and
models
now
included in
Unfortunately, not all national input-output
models are included in
CIM
IIASA consists
could
system, and
among these are
Hence, the hypothesis of whether the
lead
to
an
developed countries
increasing
cannot be
gap
between
verified with the
help of this model alone.
Conclusions
The
investigations
application are
of
the
employment
still in the initial stage at IIASA.
approach, namely the incorporation of the
INFORUM model,
might be
impacts
of
CIM
One
this subject
CIM
The chosen
labor matrices
precondition for
application
is
synthesis of detailed labor matrices.
studies on
of
in the
a new departure in the investigation of
employment impacts of CIM.
labor
impacts
a
estimating the
reconciliation
The paucity
and
of available
is an indication of the severity of the
problems of data collection and interpretation.
With
regard
to
potential
labor
substitutability
by CIM
applications in the different sectors and occupations there exits
a deficit in established knowledge.
described
above
allows
one
to
While the
estimate
simple procedure
the
range
of labor
substitution potential, a truly satisfactory computation requires
detailed data from engineering studies.
The importance of the elaboration of detailed labor matrices
is not limited to the estimation of the employment impacts of CIM
or
other
high
technologies.
Rather,
we
expect
investigation about the occupational structure by
contribute
to
answering
further
that
industries can
questions in labor economics,
e.g.
a)
the
What are the determinants of the occupational structure?
b)
How can these determinants be quantified?
c)
Can functional relations be
given
between
the explanatory
factors and the occupational structure?
d)
What possibilities
exist to
prove the estimated functional
relations?
References
Almon, C . , 1979.
The
INFORUK-IIASA
Miller, S.,
Implications.
Ayres,
R.U. ,
System of
I IASA, WP-79-22.
Input-Output Models.
Ayres, R.U. ,
International
1983.
Robotics.
Application & Social
Ballinger Publishing Company, Cambridge.
1986.
Socio-economic
Impacts
of
Robotics.
Carnegie-Mellon University, Pittsburgh, PA.
Brooks, H. 1985.
Automation Technology and Employment.
In: ATAS
Bulletin, UNO, New York, 1985, November.
J . E. ,
Engelberger,
1588.
Robotics
in
Practice.
American
Mangement Association, New York.
Fleissner,
pa,
Dell'mour,
Makrooekonomische
Mikroelektronik -
Aspekte
R.,
Sint,
P.P.,
1981.
der
Kikorelektronik.
In:
Anwendungen, Verbreitung und Auswirkungen
am Beispiel Oesterreichs.
Springer-Veriag Wien-New York.
Friedrich, W. , Roennig, G . , 1985.
Arbeitsmarktwirkungen moderner
I nst itut
Technologien.
fuer
Sozialforschung
und
Gesellschaftspolit ik, Koeln, 1985.
Haustein, H.-D, Maier, H . ,
1981.
H.-D.,
Maier,
Discussion
of Flexible
I IASA, WP-81-152.
Automation and Robotics.
Haustein,
The
H.
1985.
Flexible Automatisierung.
Akademie-Verlag, Berlin.
Howell, D.R. , 1985.
Robots
-
An
The Future Employment Impacts
Input-Output-Approach.
In:
of Industrial
Technological
Foreasting and Social Change, Vol. 28.
Hunt, H . A . , Hunt, T.L . , 1983.
Robotics.
W. E.
Human
Resource
Implications of
Upjohn Institute for Employinent Research,
Kalamaaoo, MI.
Informationstechnologie,
Eeschaeft igung.
JIRA,
1985.
Long
1988.
Informtionstechnologie
und
Econ-Ver lag, Duesseldorf , Wien, 1988.
Range
Forecasting of Demand for Industriai
Robots in Manufacturing Sector.
Kaya, Y. , 1986.
86-8.
Economic Impacts of High Technology.
I IASA, CP-
Leontlef , W. ,
1982.
The Distribution
of Work and Income.
In:
Scientific American, September 1982.
Leontief, W . , Duchln, F. , 1986.
on Workers.
Mc
1'3S5a.
Microelectronic-based
Study for Canada.
Mc
T.H. ,
Curdy,
Technical
1985b.
and
Income
Change:
Effects
of
A Multisectoral
Occupational
Technical
Change:
Implications
of
Multisectoral
A
Discussion Paper 619, Queen's University.
Study for Canada.
S.M. ,
Employment
Discussion Paper 618, Queen's University.
Microelectonic-based
Miller,
imoact of Automation
Oxford University Press, New York, 0:tf ord.
T.H. ,
Curdy,
The Future
1983.
Potential
Manufacturing Costs within the
Impacts
of
Robotics
on
Metalworking Industries. PhD
Thesis, Pittsburgh, PA.
The I NFORUM- I I ASA
Nyhus, D . E . , 1980.
Trade Mode 1:
An Inter1m
Model. I IASA, WP-80-24.
Nyhus, D.E. , Almon, C. ,
1'380.
Linked
I IASA, WP-80-112.
France,- Germany and Belgium.
OECD,
1969.
Statistics
of
Input-Output-Models for
the
Occupational and Educational
Structure of the Labor Force in 53 Lountries, Paris 1969.
OECD, 1970.
Occupational and Educational Structures of the Labor
Force and Levels of Ecanamic Development, Paris 1970.
OTA,
1984.
Office
Manufacturing:
of
Technology
Assessment.
Computerized
Employment, Education and the Workplace.
US
Cangress, Washington, D.C.
Sadler, P. ,
T.
1981.
Welcome back to the 'automation debate' .
Forester
(Ed.> :
The
Microelectronics
In:
Revolution.
Cambridge, MA, the MIT Press.
Vasko,
T . , 1983.
Policy
CAD/CAM
Issues).
Applications in
In:
in an Industrial Environment (Some
Warman,
E.A.
(Ed.> :
Computer
Production and Engineerinq, CAPE 83.
North
Holland, 1983, p.4.
Warnken,
J. ,
19136.
Zur
Anpassungsf aehigkeit
der
Entwicklung
Berufe
bis
der
' internen'
zum Jahre 28(0(0.
In:
MittAB 1986, Heft 1, p. 120.
Zymelman,
M.,
1980.
Occupational
Structures
af Indus~ries.
World Bank, Washington, D .C.
Zymelman, M. ,
1985.
Washington, D.C.
Forecasting Manpower
Demand.
World Bank,
Appendix 1
Categories of Metalcutting Machine Tools In the
American Machinist 1 2 t h Inventory
TYPES OF METALCUTTING MACHINES IN
AMERICAN MACHINIST 12th INVENTORY
TURNING MACHINES
Bench
Engine and toolroom < 8 in swing
Engine and toolroom 9 to 16 in swing
Engine and toolroom 17 to 23 in swing
Engine and toolroom 24 in swing and over
Tracer lathe
Turret lathe; ram type
Turret lathe; saddle type
Auto chuckg vert d horiz; sgl spindl
Auto chuckg vert d horiz; multi-spindl
Automatic between centen chucking
Automatic bar (screw) mach; sgl-spndl
Automatic bar mach; mult-spndl
Vert turn 8 boring mills (VTL. WM)
Other, incl, forin, axle, spin, shell
BORING
Hor. bore,drl,mil (bar mach); tabldplnr type
Hor bore,drl, mile (bar mach); floor type
Precision, horiz and vert
Jig bore, horiz and vert
other (not boring lithes)
DRILLING
Sensitive (hand feed),bench
Sensitive (hand feed), floor d pedestal
Upright:single-spindle
Upright: gang
Upright: turret, not NC
Radial
Multi-spdl cluster (adj and fxd ctr)
deep hole (incl gun drill)
other (not unit head 8 way)
MILLING
Bench type (hand or power feed)
Hand
Ver ram type (swivel head a turret)
Gen prpse, knee or bed:hor (pin, univ a ram)
Gen prpse, knee or bed: vert
Manufacturing, knee or bed
Planer type
Profiling 8 duplct (incl die,skin,spar)
CATEGORY
NOT NC CONTROLLED
NC CONTROLLED
Thread millers
2.4
Others (incl spline.router,engraving)
1
TAPPING MACHINES
4
THREADING MACHINES
24
MULTI-FUNCTIONNC MACHINES (MACHINING CENTERS)
drill-mill-bore.manualtool chg.vert8hor
4
drill-mill-borejndexingturret
4
drill-mill-bore,auto tool chg;vert
4
drill-mill-bore,auto tool chg;horiz
4
SPECIAL WAY TYPE 8 TRANSFER MACHINES
Sgl-statn(several operations on one part)
Multi-station:rotarytransfer
Multi-station:in line transfer
BROACHING MACHINES
Internal
Surface 8 other
PLANING MACHINES
Double column
Openside and other
SHAPING MACHINES (not gear)
Horizontal
Vert (slotters 8 keyseaters)
CUTOFF 8 SAWING MACHINES
Hacksaw
Circular saw (cold)
Abrasive wheel
Bandsaw
Contour sawing 8 filing
Other (incl friction)
GRINDING MACHINES
Externahplain centertype'
Externa1;univ centertype
External; centerless (incl shoe type)
External; chucking
Internal; (chucking, ctrless shoe type)
Surface; rotary table, vert 8 horiz
Surface; reciprocating, horiz, manual
Surface; recipr. vert, horiz. power
Disk grinders. not hand held
Abrasive belt (exclu polishing)
Contour (profile)
Thread grinders
Tool 8 cutter
Bsnch, floor 8 snag
Other (incl jig)
HONING MACHINES
Internal (incl combn bore-hone)
External
LAPPING MACHINES
Flat surface
Cylindrical
Other (incl combn hone-lap)
POLISHING AND BUFFING MACHINES
Polishing s t a d s (bench (L floor)
Abrasive-belt, disk, drum (not grind)
Other (incl spd lathes (L multi-stn type)
GEAR CUTTING (L FINISHING MACHINES
Gear hobbers
Gcar shapers
Bevel-gear cutters (incl planer type)
Gear-tooth finish (grind. lab. shave, etc)
Other Gear Cutting and Finishing
ELECTRICAL MACIiINING UNITS
Electrical-dischargemachines (EDM)
.
Electro-chemical machines (ECM)
Electrolytic grinders (ECG or ELG)
----..---....--.---------.....--..----.-.-.---..-Automatic assembly machines and "other" metalcutting machines are omitted.
Procedure f o r the estimation of
i n JaDan a n d U.S
U.S
23tentiaL
labor s ~ ~ b s t i t u t i o n
The
objectivz
i s t o s s t i m a t e t h e ~ o t e n t i a Ll a b o r s d a s t i t u t a o i l i t y i n
a n d J a c a n a t t r i S u t a b L e t 3 C I M . I n c a s e o f J a p a n , JIRAii985) h a s s u r v e y e d
474
companies
and
reported
the
ratio
between p o t e n t i a l
substitutabie
korkers
by
industrial
robots
and e x i s t i r s p r o c e s s w o r k e r s b y t a s k and DY
i n d u s t r y s e c t o r s . B a s e d o n t h i s , t h e p o t e n t i a l l a b o r r e ~ ! a c e r n e ~ tm a t r i x f o r
Jaoanese m a n u f a c t u r i n g
industry which
csntains
t a s k s i n t h e columns and
i n a u s t r y s e c t o r s i n t h e r o w s ;an b e e s t i m a t e d , a s s h o w n b e l o w .
IJncortunateLj',
a Labor matnix which c o n t a i n s b o t h i q d u s t r y s e c t o r and
tasks
i s
not
avaialble
for
the
U.S.A.
We
can
compare
only
the
occupation-by-sector
matrix
f g r t h e U.S.A
w i t h t h a t f o r J a ~ a n .To c o m ~ o u n d
the
difficulty.
conversion
tables
between t h e
nationa!
occu~ational
classification
systems
for
U.S
andJapan
t o
ISCO a r e
not
currently
a v a i l a b l e . T h i s makes d i f f i c u l t t o a c h i e v e c o r n p z r a b i 1 : t y .
I n the
foll~wing, a
first
tentative
s u b s t i t u t a b i l i t y i n the U.S i s described.
estimate
o f p o t e n t i a l Labor
occupations!
Labor m a t r i x f o r
t h e U.S i n t o t h e n e a r e s t
fl).Aggregate
the
classification
to
that
of
Japan and then
equate
i t t o the task Labor
m a t r i x . H e r e , t h i s D r o c e d u r e i s e n P l o y e d . T h e r e s u l t i s S ~ O ~ Ji ?n F i g u r e A - 1 .
B).Aggregate
J ~ D Z C ~ oSc c~u p a t i o n - b y - i n d u s t r y
labor matrix
i n t o t h e same
Hereafter,
classification
a s J I R f l ' s t a s k - b y - i n d u s t r y l a b o r m a t r i x , s a y A,.
t h i s aggregated occupation-by-industry
L a b o r m a t r i x i s d e n o t e d b y E,.
Let
X, d e n o t e
the
d i s t r i b ~ t i o n o f c c c u ~ a t i o na m o n g ~ ~ S K St h, a t i s ,
c o n v e r s i o n m 3 t r i x f y o n 0, t o A,.
Namely.
.
CIJ=B,X
J
Xj=BJ
(-
> A~
(2)
where i t i s needed t h a t s e n e r a i i z e d i n v a r s e m a t r i x of
B,,
nanreLy
exists.
C>Asgregate U.S
occupation-by-industry
Labor m a t r i x t o a l e v e l s i m i l a r t o
t h a t o f J a ~ a n . T h i s a g g r e g a t e d m a t r i x i s d e c a t e d b y Bu.-.
U n d e r t h e a s s a m ~ t i o nt h a t t h e c o n v e r s i o n m a t r i c e o f J a p a n a n d U . S a r e
same,
we c a n c a l c t i l a t e t a s k - b y - i n d u s t r y
hext.
Let
us
describe the
labor m a t r i x of
contents
of
Q.S., S a y Au;.
applicable
report
and t h e
~rocedure i n order
tc~ e s t i m a t e
disoLacement o f whole manufacturing i n d u s t r y .
the
data i n JIRO's
potential
labor
L S s c t ( i > = p o t e n t i a ? Labor s u b s t i t u t a b i l i t y by s e c t o r
i S j o S ( j ! = p o t e n t i a L l a b o r s t i b s t i t u t a b i i i t y by j o b t y p e
I?jb(:i,j)=respondence
whether t h e f a c t o r y has j c b s t e p i o r n o t ( J I R O ) by
s e c t o r a n d j o b t y p e , w h e r e i i i n d ~ s t r ys e c t o r a n d j ; j g b t y p e r e s p e c t i v e l y .
DSahk(i,k,?)=distribution
of
fu!L
tine
o r o d u c t i o n ~ a r k e r s ,p a r t t i m e
procucticn workers
an2
non-production
workers
by
sector
(JIRA), where
iiindustry
sector,
k ; t y ~ e of
worker,
L;job
type(l;totaL,
2;production
worker,
3:ratio
(2/1!
iBM(i>:number
of
workers
by
'ndustr:;
(MITI
;
whole
manufacturing
industry)
The e s t i m a t i o n p r o c e d u r e i s as f o L l o b ~ s :
C A ] . e s t i m a t i o n o f t o t ~ pl r o d u c t i o n w o r k e r , s a y P R w k ( i > , b y i n d u s t r y
prcduction workers
b y i n d u s t r y S e c t o r and j o b t y p e
[ B ] . d i s t r i b u t i ~ n gf
W R K R ( i . j ) ( w h i c h c o r r e s p o n d s t o A, d e s c r i b e d a b o v e . )
M
where M denotes t o t a i j o b t y p e
.One p r o b l e m o f t h e a b o v e e s t i j n a t i o n i s
M
P R w i < ( t o t a L > . R j b ( t o t a L , j>,/ Z R j b ( t o t a l , j )
j=I
td
Z L.IRKR( i ,j).
i =l
;
(7)
H e r e . t h e r i g h t h a ~ ds i d e v a i u e i s ernDLoyed a s W R K R C t c t a ? , j \ .
- I f 3ppropriate
task-by-industry
Labor m a t r i x data i s a v a i l a b l e . t h j s s t e p
i s n c t needed.
-WRKR!i,j)
g i v e s an
upper L i m i t o f s u b s t i t u t a b l e worker. ( F o r example, t h e
number o f f o r g i n g w g r k e r s i n t h e f o o d i n d u s t r y i s 0.)
:Cl.est?m~tion of
s u b s t i t u t a b l e w o r k e r b y s e c t o r , say S W s c t ( i ) . and b y j o b
[ D l . e s t i m a t i o n o f s u b s t i t u t a b l e w o r k e r b y s e c t o r anb j o b t y p e .
SBST(i,j? should s a t i s f y the folkowing constraints.
say SBST(i,j)
B e c a u s e o f c o n s t r a i n t (11). u s u a l e s t i m a t e d v a l u e
----
r\l
SBST( i j > = S B s c t ! i > - S B j g b ( j 1.1 C S B s c t ( i >
(12)
i= l
on
the
assumptian
o f independent d i s t r i b u t i o n between S B s c t ( i > an2
:
based
S B j o b ( j > s h o u l d b e m o d i f i e d . H e r e t h e f o l l o w i n g p r o c e d u r e was e m p l o y e d .
i n i t i a l value
----
SBSTO(i,j!=SBST(i,j>
T h e n e x t s t e ~i s t o m o d i f y
i f SBSTy(i.j)>WRKR(i.j>
b!ext.
ERRNext,
c a l c u l a t e row-wise
(13)
i n f e a s i b l e terms on j o b t y p e .
then
s e t SBSTk+l(i,l>=WRK4(i,j>
else
RCOM=ROOM+S3STk(i,j>
aqd
set
SBSTk+l(i
e r r o r o f SBSTktl(i.j>
Z SBSTk+l(i,j)
j=1
d i s t r i b u t e e r r o r t e r m ERR o n SSST
,S)=SBSTk(i,;>
(14:
.
SBsct(i?-
i f SBSTk+l(i.j><WRKR(i.j?
kt1
(i,j)<WRKR(i,j>
.
the^ SBST k + l ( ? . j > = S B S T K ~ i , j > ~ ( l + E R R / R C C(163
M>
Next m o d i f r t h e i n f e a s i b l e terms on i n d u s t r y s e c t o r
( s i m i l a r t o t h e P r o c e d u r e (14) t o ( 1 6 ) ?
I'dext, i f m a x i a u m v a l u e o f I ERR'ROOM I i s L e s s t h a n
F
t h e n end.
F i n a l i r , s e t k = k + l and go t o e q u a t i o n ( 1 4 ) .
I n P r a c t i c e , t h e above procedure converges s f t e r f i v e i t e r a t i o n s .