fulltext - DiVA Portal

Relatedness put in place
On the effects of proximity on firm
performance
Lisa Östbring
Department of Geography and Economic History
Umeå University, Sweden
GERUM 2015:1
GERUM – Kulturgeografi 2015:1
Institutionen för geografi och ekonomisk historia
Department of Geography and Economic History
Umeå University
SE – 901 87 Umeå
Sverige/Sweden
Tel: +46 90 786 5000
Fax: +46 90 786 6359
http://www.geoekhist.umu.se
E-mail: lisa.ostbring@umu.se
This work is protected by the Swedish Copyright Legislation (Act 1960:729)
ISBN: 978-91-7601-243-7
ISSN: 1402-5205
Cover image: © Higyou, Dreamstime.com
(http://www.dreamstime.com/stock-image-cartoon-crowd-jigsaw-fieldimage12232251)
Electronic version available at http://umu.diva-portal.org/
Printed by: Print & Media
Umeå, Sweden 2015
Acknowledgements
In the process of writing a doctoral thesis one navigates through countless
spectra of emotions, which are more or less familiar to oneself in the
beginning of the process. There are great expectations and hope, but also the
occasional feeling of despair. There is great freedom and flexibility, but also
hard work and pressure. There is the joy of having efforts confirmed and
appreciated through acceptance to a journal, but there is also the feeling of
belittling that accompanies rejection. There is the fear of public speaking and
doubt in one’s ability to convey information the first time one enters into a
classroom as a teacher instead of as a student. There is also the wonderful
sensation of being an authority in a subject matter and having the
responsibility and ability to encourage, guide, engage and motivate students
into learning slightly more than they thought possible.
All in all, my time as a PhD-student has been a period of extensive
personal and professional growth. I have met numerous extraordinary
people during this time that have contributed to this thesis in several indirect
and direct ways. I am eternally grateful to all of you and some of you are
further mentioned below.
Three people that have contributed indirectly but substantially to the
completion of this thesis are my children, Sigvard and Majken, and my
husband Erik. You guys give me perspective and you remind me that there is
a wonderful life outside of work and you challenge me to evaluate and
articulate my values. Sigvard and Majken, your curious questions motivate
me to continuously work on my communications skills. Your appreciation for
little things in everyday life inspires me to strive for mental and emotional
presence in everything that I do. Thank you Erik, for being your magnificent
self. Your strong belief in me and my abilities has remained unchanged
throughout this long process. You encourage me to continue when I am in
doubt and you are always the first one to celebrate my achievements. You are
proud when I say “oh, that’s nothing!” and you find ways to motivate me like,
“I’ll grow my moustache back out when you reach that stage” or “We’ll drink
champagne once you have finished that”.
I owe great thanks to my supervisors Urban Lindgren and Rikard
Eriksson. Urban, you have this ability to always know what I need. Be it an
encouraging pep-talk, discussing ideas, or just leaving something to be
resolved by me in due time. You have had confidence in me from the
beginning and you have trusted my ability to come up with my own ideas.
You have this egalitarian perspective that we can both learn from each other,
which was initially terrifying! I did not feel like an authority on any matters
discussed. But I have learnt and now I really think that we can complement
each other’s knowledge. Thank you Rikard, I am in awe of your level of
ambition, your knowledge and thoughtfulness, your efficiency and accuracy,
i
and your willingness to help in any ways imaginable. Your close readings of
my texts and your precise questions have motivated me to constantly do
better. The pair of you make a wonderful, inspiring, and dynamic supervising
duo that I am fortunate to have worked with.
I am grateful to my two Erikas; Erika Knobblock and Erika Sandow
whom, during different periods of my PhD-project, took me in during my
numerous stays in Umeå. Your hospitality and generosity know no
boundaries.
Staying with you guys have enriched my social life
tremendously and you have come to be two of my closest friends. In the later
part of the project, Katarina Haugen and Olle Stjernström took over some of
the responsibility of housing me. You selflessly invited me to your home,
disregarding any inconvenience it may entail. A special thanks to Katarina
for our talks of big and small things. You too, have come to be a close and
valued friend. Thank you also, to Johanna Liljenfeldt for so generously
letting me stay with you, often with short notice! There are more of you that
have offered me a place to stay, thank you to all of you! This project would
not have been possible, had it not been for the generosity and kindness of my
colleagues and friends.
During my time as a PhD student there have been several people that have
helped me with practical issues related to my work. I am grateful to Erik
Bäckström, Lotta Brännlund, Ylva Linghult and Fredrik Gärling for your
invaluable help in; working with the database, administrative concerns, and
dealing with students. Your help is highly appreciated and you deliver it with
such ease and always with a smile. Thank you to Magnus Strömgren for
always finding the time to help out with answering unanticipated questions
from the students related to the GIS- and statistics software in the computer
lab. Great thanks to Martin Henning, Einar Holm and Olle Stjernström for
commenting on an earlier draft of the thesis. Your insightful comments
contributed to significantly improving the quality of the thesis.
Special thanks to Lars Larsson and Kajsa Åberg for always making me
smile and think. You do this by being funny and serious at the same time.
Also, thank you to Olle Olsson for years of teaching together and for your
ability to be your awesome self, and to Mats Borrie for our little talks in the
corridor about this and that. Also, a great thank you to Einar Holm and
Kerstin Westin for thinking, early on, that I could contribute with something
in academia. And thank you Einar, for introducing me to Sture Öberg. Apart
from showing me the width and depth of human geography, Sture managed
to do something that no one else has; give me advice on life and work that I
listened to. I have pondered over things you said in the passing and it has
been a great help in my professional and personal life.
Thank you to my parents-in-law, Agneta and Rolf, for so generously
helping with the logistics of my travels. You have made everything run so
much smoother than it otherwise would have! Thank you also to Linda
ii
Thorell for regularly having lunch with me and by so doing giving me a social
life during long periods of working alone, from home. Finally I want to thank
my parents, Kaj and Lena, for always believing in me and giving me the
impression that you truly think I can accomplish anything I set out to do.
And thank you dad for helping out with the kids during my numerous visits
to Umeå.
Lisa Östbring
Umeå, March 2015
iii
iv
Table of Contents
1. Introduction
1
1.1. Contribution and aim
4
1.2. Outline of the thesis
6
2. Theoretical research context
7
2.1. Setting the scene: From agglomeration externalities to proximity
2.2. An evolutionary framework
7
9
2.3. Knowledge in the learning economy
10
2.4. Sector and firm as structures of knowledge
14
2.5. Proximity and relatedness
16
3. Methodological comments
22
3.1. The data
23
3.2. Measuring knowledge and ability
25
3.3. Estimating proximity in the labour force
26
3.4. Measurement of innovativeness
29
4. Paper summaries
31
4.1. Paper I: Labour mobility and plant performance: On the (dis)similarity
between labour- and capital-intensive sectors for knowledge diffusion and
productivity
31
4.2. Paper II: Labor mobility and organizational proximity: Routines as
supporting mechanisms for variety, skill integration and productivity
32
4.3. Paper III: Relatedness through experience: On the importance of
collected worker experiences for plant performance
5. Concluding discussion: Relatedness put in place
35
37
5.1. Findings
37
5.2. Discussion
42
5.3. Directions for future research
45
6. Sammanfattning (Summary in Swedish)
46
7. References
54
v
Appendices
Paper I:
Östbring, L., and Lindgren, U. (2013): Labour mobility and plant
performance: On the (dis)similarity between labour- and capitalintensive
sectors
for
knowledge
diffusion
and
productivity.
Geografiska Annaler: Series B, Human Geography 95:287-305.
Paper II:
Östbring, L.; Eriksson, R.; and Lindgren, U. (submitted): Labor
mobility and organizational proximity: Routines as supporting
mechanisms for variety, skill integration and productivity.
Paper III:
Östbring, L.; Eriksson, R.; and Lindgren, U. Relatedness through
experience: On the importance of collected worker experiences for
plant performance.
vi
1. Introduction
This thesis considers the effects of learning on firm performance. It
specifically examines how various forms of proximity between agents
(individuals) influence the competitiveness of firms. The emphasis is on the
role played by the individuals, their knowledge and their collective ability to
learn and generate novelty within the firm. The thesis rests on the assumption
that learning, knowledge and innovation are central not only to firm
competitiveness, but also to the discipline of economic geography.
During recent decades, knowledge, learning and innovation have gained
increased attention as the main drivers of the economy. Today, it is commonly
argued that the importance of knowledge for sustained firm competitiveness
supersedes that of capital. Early contributions by economists highlighted the
role of human capital and technological change as promoters of economic
growth (Arrow 1962; Romer 1990; Solow 1956). In their studies of an
increasingly globalized world, economic geographers have developed this idea
further. Knowledge and human capital are now often considered the most
localized and most important of all production factors. The accessibility of
information and goods, for both consumers and producers, has changed
patterns of production and consumption and made place “slippery” in many
respects (Dicken 2007; Markusen 1996). The locational advantage of firms is
no longer mainly determined by proximity to physical input factors or markets
or geographically differentiated wages (von Thünen 1966; Weber 1929).
Instead, it is increasingly access to knowledge that has become crucial to firm
survival and competitiveness. The escalated competition brought about by
globalization has increased the rate and frequency of processes of creative
destruction (Schumpeter 1942), in which only the most innovative firms in an
industry survive. Therefore the success of a firm is highly dependent on its
available knowledge resources. It is through these resources that external
knowledge is absorbed and internal knowledge combined and applied in new
ways to generate novelty.
The prominent role of knowledge in the production of goods and services
has given rise to concepts like the knowledge society (Bell 1973), the learning
economy (Lundvall and Johnson 1994) and the knowledge-based economy
(David and Foray 1995). Two of these concepts, the learning economy and the
knowledge-based economy, are often used synonymously, but when they were
first introduced their meanings differed. David and Foray (1995) described the
knowledge-based economy as relying on science and technological knowledge,
mainly in science-driven high-tech sectors. Lundvall and Johnson (1994)
described a learning economy in which all sectors seemed to rely on learning
more than on any other input factor for their development and
competitiveness. It is the perspective of a learning economy in which all
industries can generate economically valuable knowledge that is adopted in
1
the present thesis. The success of a firm is no longer determined by the
distance to physical input factors, which through globalizing processes have
become increasingly ubiquitous (Maskell and Malmberg 1999b), but rather by
the accessibility of relevant knowledge. Even though information accessibility
is at an all-time high, knowledge is more specialized and localized than ever
(Howells 2012; Maskell and Malmberg 1999a). Every place in the space
economy fosters its own unique combination of formal and informal
institutions and knowledge, which to a large extent is the result of past
economic, political and social structures (Massey 1984). The socialization of
knowledge makes it localized because social depth is not easily transferred
over the great physical distances that globalization has brought about. In
economic geography (Morgan 2004) and in theories of the firm (Nonaka et al.
2000), it is argue that social depth is crucial in transforming information into
knowledge. This has motivated an upsurge of theoretical and empirical
contributions on non-physical forms of proximity between agents that could
alleviate the need for physical proximity in the generation of trust and
knowledge (Almeida et al. 2002; Boschma 2005; Boschma et al. 2009;
Grabher 2002a; Lam 1997; Torre and Gilly 2000; Torre and Rallet 2005). It
is in this body of literature that the thesis is embedded.
Traditionally, economic geographers have studied the relationship
between geographical agglomerations and knowledge production. The
primary focus has been on the role of geography in facilitating external
relations between firms. For example, localised learning refers to interactive
learning processes that are stimulated by localised capabilities and the ability
to interact easily with other actors in one’s physical proximity (Malmberg and
Maskell 2002; Maskell and Malmberg 1999a). Similar perspectives on
knowledge creation through external relations are emphasized in theories of
learning regions (Morgan 1997), innovative milieus (Maillat and Lecoq 1992),
industrial districts (Piore and Sabel 1984) and clusters (Porter 1990). All these
different perspectives emphasise the economically desirable characteristics of
co-location, such as competitiveness and high growth. Yet another strand of
literature focused on the generation of knowledge is that concerned with
national and regional innovation systems (Asheim and Gertler 2005; Asheim
et al. 2011; Cooke 2001; Lundvall 1992). These perspectives include
institutions and relationships between various forms of actors within a
geographically delimited territory in their analyses of innovation processes. A
commonality between all these theories is the focus on knowledge
externalities as important sources of novelty. The role of external knowledge
for firms is neither denied nor disregarded, but it is not the focus of the thesis.
Firms cannot rely exclusively on external knowledge for their innovativeness,
nor can they absorb all available knowledge. It is the in-house composition of
knowledge that determines the absorptive capacity of the firm (Cohen and
Levinthal 1990) and the combinative capabilities (Kogut and Zander 1992).
2
Whereas absorptive capacity represents the ability to recognize and absorb
external knowledge, the combinative capabilities of the firm are determined
by employees’ ability to learn from each other and combine their knowledge
in new ways. It is these internal abilities of the firm to learn and innovate that
are under scrutiny in the thesis. This is because the economy is ultimately
made up of individuals in different networks and structures and the firm is the
primary organizing unit of interactive learning processes. Penrose (1959)
acknowledged that the distinction between firm and market is administration,
which occurs in firms but not in markets. Furthermore, Boschma (2004)
discusses and questions whether and how regions can be competitive, as
learning and innovation occur in or between firms that are constituted by
individuals. Brenner and Broekel (2011) examine to what extent the
innovation performance of regions can be measured. They present a number
of indicators of how the region, given that it is a coherent unit, may stimulate
or contribute to firm innovativeness, thus acknowledging that innovation
processes take place within and between firms by combining individuals’
knowledge in new ways.
In the current economic geography literature, and in economics and
innovation studies, the beneficial role of relatedness for economic growth is
highlighted for regions, firms and plants (Boschma et al. 2014; Eriksson and
Lindgren 2009; Neffke et al. 2008). The relatedness, whether it be co-located
firms in related industries (Boschma and Iammarino 2009; Neffke et al.
2012), firms diversifying into activities that are related to their main activity
(Tanriverdi and Venkatraman 2005) or a labour pool of related competences
(Boschma et al. 2009), presumably offers the greatest potential for successful
innovation. The reason is that added knowledge and technology that are
related to existing knowledge and technology in a plant are similar enough to
be absorbed easily, while at the same time different enough to generate ideas
for novel applications. A firm may have all the physical capital to produce a
related good that renders the cost of diversification low. Similarly, a firm in a
locality with a high number of related industries and firms reaps efficiency
gains (externalities) both from knowledge spill overs and from supporting
structures aimed at promoting the regional knowledge in a specific field (see,
e.g., the regional innovation systems literature; Cooke 2001).
In theories of the role of relatedness in the labour pool, whether it concerns
the in-house competences in a firm or labour mobility studies, the link
between relatedness and competitiveness is less clear cut than in theories of
the benefits of related agglomeration or firms expanding into related
activities. Cognitive relatedness in the labour pool facilitates effective
individual and group learning, i.e., knowledge growth. In the case of labour
mobility, it presumably also ensures effective integration of new employees in
firm activities. The idea that relatedness represents a cognitive distance,
which is neither too long nor too short, stems from the literature on proximity
3
dimensions (Boschma 2005; Kirat and Lung 1999; Torre and Gilly 2000;
Torre and Rallet 2005). The literature on (mainly) non-physical proximity
dimensions states that there should be cognitive relatedness between agents
for effective learning to occur. But there are also other dimensions (social,
organizational, institutional and geographical) that potentially can reduce the
demand for cognitive relatedness through their socializing functions. Short
distances in these dimensions reduce uncertainty and promote trust and
understanding between individuals. It is suggested in the present thesis that
individuals can come to be associated in such a manner that they can learn
from each other, and advance firm knowledge, despite a long distance between
(heterogeneity of) their formal knowledge.
1.1
Contribution and aim
As discussed in the introductory section, knowledge is considered the most
important resource in the production of goods and services in the learning
economy. There is nothing controversial in claiming that processes of learning
and innovation are essential to firm competitiveness in all sectors. For
example, Hipp et al. (2000), Miles (2005) and Tether et al. (2001) have
studied learning and innovation processes in service industries, whereas Acha
and Brusoni (2003) and Tunzelmann and Acha (2005) have studied
innovation processes in low- and medium-tech industries. It is through
learning that knowledge is obtained and applied to potentially generate
novelty. But whereas the role of knowledge production and spill over effects
has been thoroughly discussed in economic geography (Heimeriks and
Boschma 2013; Neffke et al. 2011), few studies have dealt with the ability of
firms to absorb existing knowledge (Howells 2002). In the present thesis, the
primary study object is the plant and its employees, i.e., the labour force. In
the micro-perspective, the focus is on the interactive learning processes and
absorptive capacities of firms and plants. It is through the employees that the
firm consumes and absorbs knowledge, and by examining the knowledge
variety in plants, the present thesis hopes to contribute to our understanding
of the role of the labour force in firm competitiveness.
Apart from absorbing external knowledge through monitoring and cooperations with other actors, firms can also acquire knowledge by hiring new
individuals. In such instances, one or a few individuals are brought into the
firm, which constitutes a specific socio-technical context. In this context, the
new employees are expected to communicate their knowledge, whereas
existing employees are expected to absorb, learn and apply the new
knowledge. The effect that labour force knowledge has on competitiveness and
firm performance is primarily determined by the collective ability of
employees to learn and apply the new knowledge. Therefore, analyses of the
relational character of knowledge and learning should take multiple proximity
dimensions and thus types of relations into account. Torre and Gilly (2000)
4
suggest that despite the frequent discussions of the beneficial role of proximity
in economic geography, the exact workings of proximity remain a black box.
This is a black box in urgent need of opening; if the concept is to be more than
a theoretical notion, it needs to be analytically relevant and empirically
applicable. Both Boschma (2005) and Torre and Gilly (2000) discuss the
probable interconnectedness and overlap effects between different forms of
physical and non-physical proximity. Nevertheless, most existing empirical
work on proximity dimensions and knowledge flows (i.e., labour flows) within
and between firms analyse one form of non-physical proximity (see, e.g.,
Boschma et al. 2009; Timmermans and Boschma 2013; Eriksson 2011). These
studies have shown slightly diverging results, which may be attributed to the
interconnectedness of multiple forms of proximity. The potentially variable
impact of formal knowledge variety on plant performance, due to differences
in other types of proximity, is the primary focus of the present thesis.
Furthermore, the thesis also considers the in-house composition of
knowledge, not only the relationship between inflowing knowledge (labour)
and existing knowledge. By so doing, the relative importance of high levels of
human capital can be contrasted with the influence of knowledge relatedness
(variety). Boschma et al. (2009) found high levels of knowledge relatedness to
be more important to firm performance than having high levels of human
capital. But without the integration of human capital and knowledge variety
into the same analytical model, the relationship between the two forms of
knowledge and their unique and combined influence on firm performance
cannot be properly addressed.
Our understanding of the shifting role of the labour force, from physical
labour to knowledge resources, is still limited. Therefore, starting from the
literature on proximity dimensions and relatedness, the aim of the present
thesis is to examine how different forms of proximity in the labour force
influence the competitiveness of firms in the learning economy. The point of
departure is the understanding that the proximity dimensions afford trust and
reduce uncertainty between individuals. These are aspects that promote
interactive learning processes.
There are multiple claims that processes of knowledge production and
learning differ in different sectors. Pavitt (1984) introduced a taxonomy of
drivers of technical change in different sectors that has been extensively used
ever since. Furthermore, Strambach (2013) claims that knowledge and
innovative behaviour accumulate at the level of the industry, the firm, and the
region, whereas Malerba (2006) points to the differentiated innovation
processes in different sectors. But most empirical studies of labour force
knowledge as a promoter of growth have either studied the entire economy
(e.g., Boschma et al. 2009) or just one sector (e.g., Bienkowska 2007). This
leaves either very broad generalizations about the workings of knowledge
throughout the economy, or thorough descriptions of the particular. In the
5
thesis, the different conditions under which the labour force interacts and
applies its knowledge are examined. Each plant in the economy comprises a
unique context that is reproduced by the participants and their knowledge and
abilities, and by cognitive and physical structures within the plant
(Nooteboom 2009). By scrutinizing the interplay between local microdiversity (i.e., the skills and knowledge of the local labour force) and
overarching structures (e.g., firm routines or technological regimes), an
integration of micro- and meso-level mechanisms is achieved, which allows
for a rich analysis of the role of knowledge in promoting competitiveness for
firms and plants. Through this integration of the role of the labour force and
its knowledge, on the one hand, and structures in the form of firm routines
and production technologies, on the other, it is acknowledged that both parts
affect the potential for innovation. It is shown that labour force composition
and knowledge are essential parts of firm competitiveness, but only if the
surrounding structures offer ways of applying knowledge that is circulating.
In the thesis, knowledge is analysed at the level of the plant and combined
with factors that are believed to affect the innovative power of learning in
firms. These factors are embedded in the local context and industry setup.
Moreover, the labour force is formed by real-life individuals that react
differently to the same situations due to their different abilities, knowledge
and experiences. Therefore, the efficacy of labour force knowledge in
enhancing firm performance is not only contingent on the sum of the
employees’ formal knowledge. Learning processes are social processes and
similarities in past work experiences and institutional environment influence
the ease and success of interactive learning processes. Factors that are
assumed to influence the innovativeness of firms are explored in three
empirical studies. Each of them pursues a different aspect of learning and
knowledge application in firms. The three papers have different foci; the
effects of cognitive and geographical proximity on learning in plants in
different sectors, the effects of cognitive, organizational and geographical
proximity on learning at the plant level and the effects of various forms of
cognitive proximity on learning in plants. At the same time, they share the
umbrella theme of the labour force as a (mainly localized) knowledge resource
and potential promoter of economic growth.
1.2
Outline of the thesis
The thesis comprises three empirical papers, a theoretical framing and
discussion and remarks on methodological considerations and delimitations.
In this introductory section, Section 1, the overarching aim of the thesis and
the three papers is introduced. The following section, Section 2, places the
empirical papers (referred to as Paper I to III) in the theoretical context to
which the thesis aims to contribute. By taking an evolutionary perspective on
knowledge production and reproduction, the context-dependent nature of
6
learning and competitiveness is highlighted. The role of the individuals and
the potential benefits of being embedded (Granovetter 1985) in different
contexts for learning processes are discussed. There is also a short discussion
of the presumed hindering or enabling meso-level structures associated with
learning and knowledge application in specific sectors. Furthermore, the
underlying theoretical implications of the proximity dimensions framework
(Boschma 2005; Torre and Gilly 2000) are considered. It is argues that the
relational (non-physical) aspects of the proximity dimensions are of greater
importance to learning processes on the micro-level of interpersonal
knowledge exchange. In Section 3, methodological considerations are
discussed, mainly those concerned with operationalization of theoretical
concepts. Section 4 offers summaries of the empirical papers and in Section 5
the main findings and conclusions of the papers are discussed. In this section
there is also an introduction of potential future research based on the findings
in the thesis. A short Swedish summary is provided in Section 6.
2. Theoretical research context
This section presents a general theoretical background for the thesis. It does
not reflect the theoretical frameworks for the different papers, which can be
found in the initial part of each paper. Instead it provides a description of the
theoretical context to which it aims to contribute. It starts with a short
overview of the changing role of proximity in theories in economic geography.
It then moves on to specifically present the evolutionary perspective on
growth, and the role of physical and non-physical proximity in interactive
learning and knowledge production for firms. The point of departure is the
understanding that contemporary capitalist society is substantially shaped by
knowledge and skills. These properties of society reinforce the role of
individuals in theories of economic growth. It also points to the importance of
micro-level analysis (the level of the firm or plant) as complements to mesoor macro-level analyses of sectors, networks, regions and nations in
understanding the uneven spatial development of successful economic
activities.
2.1. Setting the scene: From agglomeration externalities to
proximity
Already in 1890, Marshall (1890) acknowledged that there was something
“in the air” in agglomerations that seemed to stimulate the competitiveness of
firms. Proximity is thus not a new idea in economic geography, but for a long
time it was mainly physical proximity to numerous external auxiliary
functions that was thought to be important to firm competitiveness. The
benefits of high concentrations of agents and firms have been repeatedly
highlighted in the agglomerations literature. Proximity to competitors, sub7
contractors, clients, large markets and a specialized labour pool have all been
considered locational advantages for firms, dating back to von Thünen (1966
(English version of the 1826 publication Der Isolierte Staat)) and Marshall
(1890) and their predecessors. After the Second World War, Fordist mass
production was gradually replaced with leaner forms of production. This was
mainly a result of technological advancements occurring at an increasingly
rapid pace, which allowed for consumer influence and custom-made products.
On a global scale, the technological advancements meant substantial
improvements of, initially, the physical infrastructure and, later on, the digital
infrastructure (Dicken 2007), which gave the impression that ‘distance was
dead’ (Cairncross 1997). The role of science and technology in the
advancements of the global networks that emerged steered the foci of
economic geography research toward specific geographical agglomerations
and clusters, namely those of highly competitive and successful high-tech
industries (Cooke 2004; Frenkel 2001; Klepper 2010; Saxenian 1994;
Sternberg 1996).
With the focus on high-tech regions came an inflated interest in knowledge
and people as resources, rather than just physical labour. In the literature,
there was increased mention of the embodiment of knowledge. With such a
socialized competence view came reflections on how human capital,
knowledge workers and talents formed the local environment (Florida 2002;
Lucas 1988; Romer 1990; Scott 2004). This was described as occurring not
only indirectly, through their place of work, but also through their lifestyle
preferences. The role of individuals in shaping the local environment was put
forward by economic geographers as being twofold: partly as being part of the
labour force and as such –, small units of cognitive abilities – collectively
shaping the innovativeness of firms, and partly as being cultural and social
beings that were part of communities and networks that reproduced not only
knowledge, but also values and norms (Bathelt and Glückler 2005; 2013;
Farole et al. 2010; Storper and Scott 2009; Storper and Venables 2004).
Silicon Valley is probably one of the most researched regions with regard to
the relationship between technological innovativeness and social and
institutional structures (see, e.g., Saxenian 1994; Kenney and von Burg 1999).
The recognition of workers as embodying knowledge and knowledge
production as occurring in a cumulative fashion in a specific socio-technical
context (Strambach and Klement 2012) offered further insights into the
variable outcomes of seemingly similar knowledge inputs in economic
processes in different spatial contexts (Gertler 2003; Leiponen and Drejer
2007).
The benefits of geographical proximity, as discussed in the agglomerations
literature, came to concern the structures and networks, in a region, that
supported or sustained high levels of technical innovation rates. Institutional
and social perspectives on differentiated regional growth introduced non8
physical forms of proximity between actors. As mentioned in the introduction,
geographical proximity – or co-location – facilitates and stimulates
interaction between actors (Knoben 2009). In theories of the learning region,
geographical proximity facilitates collaborations based on trust between
actors (Morgan 1997), and in the industrial districts literature the role of
collaborations between small- and medium-sized firms is stressed (Piore and
Sabel 1984). The importance of geographical proximity in the innovative
milieu literature is based on the facilitated relations between knowledge
producing actors (Maillat and Lecoq 1992). The role of geographical proximity
in cluster theory highlights proximity to competitors and partners (Porter
1990). Eriksson (2011) found an inverse relationship between the positive
effects of related variety and distance of origin of labour inflow, indicating the
importance of place-specific institutions in understanding variety in
production technology. Similarly, Rigby and Essletzbichler (2006) and
Haltiwanger et al. (1999) found great spatial differences in production
techniques within the same manufacturing industries. The difficulty in
imitating and learning from other firms in the same sector but from different
regions preserves the heterogeneity between firms and demonstrates a place
dependence of knowledge that is due to specific social and institutional
characteristics.
There is a voluminous body of literature on place-specific institutions and
firms as structures that affect, and are affected by, the behaviour of microagents. The reciprocal relationship allows a collective past to influence the
perceived probability of success associated with different actions (see, e.g.,
Amin and Cohendet 2005; Bathelt and Glückler 2013; Farole et al. 2010;
Maskell 2001). These structures are integrated into individuals’ cognitive
abilities, partly as past experiences and partly as current practices.
2.2. An evolutionary framework
The seminal work of Nelson and Winter (1982), on the evolutionary
perspective, has had a major impact on the promotion of evolutionary theory
in economics. According to Nelson (1995), evolutionary theory in any
discipline discusses the same three key principles: they explain the temporal
development of something, they consider random elements that generate
variation, and they emphasize the existence of inertial forces that provide
continuity among entities. In evolutionary economics, the three main
mechanisms that influence the outcome of economic activities are: the
routines, the search for new and better routines and the selection
environment. Evolutionary economics is quite different from neoclassical
economics in its rejection of the economy as a closed system. Furthermore, it
argues that technological change should be the primary focus of economic
analysis. There is great emphasis on novelty and change in evolutionary
economics, which is in contrast to traditional equilibrium theory. In
9
evolutionary economics, decreasing returns due to price competition is a
secondary phenomenon in comparison to innovation processes and the search
for novelty and Schumpeterian rents (Boschma and Frenken 2006).
Following theories of evolutionary economics, an evolutionary economic
geography framework has recently been developed (Boschma and Frenken
2006; 2009; Boschma and Martin 2007; 2010). The concept of evolution was
not altogether absent in geography and economic geography prior to this work
(see, e.g., Martin 2000; Massey 1984; Rigby and Essletzbichler 1997), but the
intention of the framework is a clearly defined theory of economic
development in economic geography. The ambition with the framework is a
theory complemented with research methodology and a research agenda. In
evolutionary economic geography, firm routines and their development over
time are core attributes in the competitiveness of firms. The overarching aim
of evolutionary economic geography is thus to analyse the changing spatial
distribution of routines and the resulting impact on the distribution of
economic activity.
Within an evolutionary framework, analyses of economic processes can be
carried out on multiple levels (Boschma and Frenken 2006). At the macrolevel, the development of sectors and networks can be analysed within spatial
systems, such as cities, regions or nations, and their structural and dynamic
prerequisites compared in a global system. At the meso-level, evolutionary
economic geography highlights the spatial development of sectors and
networks and the analysis of relations of competition and complementarity.
Finally, at the micro-level, the firm and its routines are under scrutiny. The
development of the firm and the effects of variety on the competitiveness of
the firm are highlighted as important research topics. At all levels of analysis,
the fundamental assumption is that structures (institutions, routines, etc.)
lead to the accumulation of knowledge and capabilities, in firms, sectors,
networks and regions, which can be steered in different directions through a
combinatorial dynamic driven by the interactions of agents (Strambach and
Klement 2012; Strambach 2013). The present thesis focuses exclusively on the
micro-level of analysis and learning processes that occur as the result of
internal variety, at the level of the plant. The reproduction and adaption of
firm routines occur within the firm and its plants as responses to market
changes. The accumulated knowledge in a plant determines the efficacy of
knowledge absorption and adaption. Therefore, studying the in-house
composition of knowledge and abilities is a crucial step in understanding the
variable ability of firms to adapt and consequently remain competitive.
2.3. Knowledge in the learning economy
The territorial accumulation of knowledge occurs on multiple scales and
has many names, for example varieties of capitalism (Hall and Soskice 2001),
regional innovation systems (Cooke 2001) and localized learning (Maskell and
10
Malmberg 1999). The spatial accumulation of knowledge in structures and
people suggests that there is a place-specific characteristic of knowledge and
how it is produced, reproduced and ultimately applied (Lam 1997; Rigby and
Essletzbichler 2006; Storper 1997). Knowledge accumulates in regions due to
the interactions of social and technological networks that have developed over
time (Leiponen and Drejer 2007; Rigby and Essletzbichler 1997). Knowledge
can thus be seen as a structuring component throughout society. On the
individual level, people make increasingly reflexive decisions about their
everyday life and practices based on their knowledge (Giddens 1990). On the
systems level, knowledge has come to be an essential part in explaining how
the economy works (Bell 1973; Lundvall and Johnson 1994). But firm
routines, the industry and its technological regime and the local labour market
all influence the choices and outcomes of knowledge applications. The pathdependency of knowledge generates non-physical proximity between
individuals belonging to the same firm, industry and labour market. Thus,
knowledge does not only have to be technologically relevant to be conversed
and applied within the firm. It also has to be understood and communicated
throughout the specific firm in a specific region. This thesis scrutinizes the
micro-diversity of labour force knowledge and links it to macro-structures in
an attempt to better understand the contingent nature of learning and
innovation.
The concept of the learning economy (Lundvall and Johnson 1994) did not
only introduce knowledge and processes of learning as key factors in
innovation and firm competitiveness. It also pointed to the role of knowledge
in all sectors, not only in the traditionally R&D-intensive sectors, but also for
firms in, for example, the service industry. Barras (1986; 1990) argues that the
rise of ICT (Information and Communication Technology) led to an industrial
revolution in the service sector. The use of ICT allowed firms in the service
sector to improve their efficiency and quality and develop new services at a
faster pace than had previously been possible. In the learning economy, the
acquisition, creation and implementation of knowledge are the key factors in
explaining competitive advantage, as firms must continuously innovate to not
lose market shares (Nonaka 1994; Lawson and Lorenz 1999). Under such
circumstances, it is not only the learning that is important, but also the ability
to apply knowledge to productive practice.
The embodiment of knowledge in people makes it a unique resource.
Because people are embedded in local contexts, their knowledge is not only
personal, but also localized and contextual. Therefore, the effects of sharing
bits of knowledge vary between interactions due to different interpretations
(Nonaka et al. 2000). These interpretations are based on personal experiences
(Polanyi 1962; March 2010), understanding of the firm routines (Nooteboom
2009) and technological knowledge of the specific industry (Pavitt 1998). But
innovations within firms are not just the result of the combined abilities of all
11
employees to absorb, converse and apply knowledge. There are also external
factors that always, but variably, affect the formation of a context and
influence the propensity for knowledge application and innovation. These
factors are the industry in which an activity is carried out (see, e.g., Pavitt
[1984] on the determinants for sectoral technological trajectories) and the
firm in which the activity takes place (see, e.g., Dosi et al. [2008] on firm
routines as a selection mechanism for knowledge absorption). Whereas placespecific institutions (i.e., Gertler 2003; Maskell and Malmberg 1999a)
contribute to the formation and production of knowledge, there are factors
associated with firm and industry that do not only steer the knowledge view
within the firm, but also affect the firm’s capacity to implement the ideas of
individual employees. Individuals, firms, sectors and regions are all shaped by
past events and experiences. The configuration and development of economic
activities in specific places are evolutionary and path-dependent on multiple
levels.
In contemporary economic geography, descriptions of knowledge have
come to include relational and social aspects. For instance, Howells (2002)
describes knowledge as a socially constructed process, whereas Amin and
Cohendet (2005) define it as a process and practice rather than a possession.
Lam (2000) considers knowledge to exist between individuals rather than
within them. Thus, knowledge is a resource in itself, but it also impacts how
the firm uses other resources. Knowledge is therefore a major differentiating
factor for firm competitiveness, because the ability of firms to procure,
converse and apply knowledge varies. In addition to that, knowledge is also a
volatile resource due to its embodiment in people. If a person decides to
change jobs, the firm risks losing crucial bits of knowledge (Howells 2012).
Due to the dependence on knowledge, and implicitly on the labour force or
even single individuals, throughout the capitalist economy, the importance of
micro-level analysis of labour force characteristics and collective learning
processes cannot be stressed enough. The present thesis aspires to uncover
some of the complexity associated with learning, innovation and economic
growth within firms.
Although knowledge and learning have only recently begun to attract the
attention of economic geographers, the economy has always depended on
knowledge (Hudson 2009). The complexity of knowledge and learning, and
the later association with innovation, has generated a vast body of literature,
ranging from psychology and cognitive science (Vygotsky 1962) to sociology,
organizational studies (Cohen and Levinthal 1990; Lam 2000) and economics
(Harrison et al. 2001; Penrose 1959). Whereas philosophy largely views
knowledge and learning as internal processes within the individual, sociology
sees learning as an interactive process between two or more individuals. In
economics, Penrose (1959) regarded the varying ability of firms to
administrate similar production factors to be the main differentiating factor
12
in firm growth. The notions introduced by Penrose have been further
developed in economics, and knowledge has come to be regarded as the main
resource or capability of the firm in the knowledge-based theory of the firm
(Grant 1996a).
Howells (2012) goes back to reinvestigate Michael Polanyi’s work on tacit
knowledge and knowledge flows to determine the appropriateness of the
rather flexible use of the concept in economic geography. He theorizes that
knowledge cannot flow freely, only information can. Instead, the embodiment
of knowledge in people requires human interaction for learning to occur. As
mentioned in the introduction, information becomes knowledge only when it
is meaningful to the receiver (Nonaka et al. 2000). Meaning is created from
associating information with what one already knows and has experienced
(March 2010; Polanyi 1962). If past experiences help individuals interpret
information and situations, the implication is that processes of understanding
are tacit. Therefore, processes of knowledge creation and learning are always,
but varyingly, tacit and thus socially embedded (Morgan 2004) and
cumulative (Strambach 2013). In his early work, Polanyi (1962; 1966) put
forth the idea that knowledge exists in a wide spectrum (not a dichotomy)
ranging from wholly tacit to wholly explicit knowledge. But because the
understanding and application of knowledge are based on personal and
collective experience, which affects perception and anticipation, there is no
such thing as completely explicit knowledge – it is always subjective and
contextual.
In economic geography, Bathelt and Glückler (2003) deal with this
subjectivity in terms of a relational perspective on knowledge, in particular,
and resources, in general: Knowledge is in the eye of the beholder and the
socio-technical context determines how knowledge advancements are best
achieved. March (2010) also acknowledges the subjectivity of knowledge by
discussing employees’ experiences as useful but imperfect teachers in dealing
with situations at the current place of work. The success of communication
between two individuals, thus, depends as much on past experiences and
awareness of practices associated with a specific industry and firm as it does
on their formal knowledge. The contextual and relational aspect of knowledge
was introduced in economics and the resource-based view of the firm in the
late fifties by Penrose (1959). She claimed that the main differentiating factor
for firms with apparent similarities in input factors was their capability to
combine existing resources in different ways. A few decades later, the same
phenomenon but on a different scale – the local environment or the region –
was picked up by economic geographers. Theories of localised capabilities
(Maskell and Malmberg 1999a) and learning regions (Morgan 1997) are a few
contributions that consider the socially embedded and relational character on
knowledge economic geography.
The socializing and structuring of
knowledge occurs within firms as well as within regions. It is therefore
13
reasonable to assume that knowledge and the effects of learning in firms and
regions cannot be understood in isolation, but rather in combination with
socializing structures, such as technological trajectories and formal and
informal networks. Individuals in a certain place produce and reproduce
capabilities through their formal competences and interpretations of values,
norms and goals.
2.4. Sector and firm as structures of knowledge
One form of macro-structure is the technology used within a sector or
specific industry. The possibility that different production technologies offer
structurally different opportunities for learning and innovation has been
somewhat overlooked in studies on knowledge dissemination within and
between firms. Pavitt (1998) stated that the specificity of input factors in
various sectors renders them differently equipped to integrate inflows of skills
and knowledge into plant routines. The reason is differences in technological
regimes and routines (Dosi 1982; Nelson and Winter 1982), which generate
unequal absorptive capacities (Cohen and Levinthal 1990; Nooteboom et al.
2007) and partly explain asymmetric industrial development (Patel and Pavitt
1997). Because the technological trajectories of firms vary, their propensity to
act on and respond to new knowledge differs with regard to how well the
information is understood and can be used within the plant. According to
Pavitt (1984), learning processes are sector-specific and dependent on the
firm’s principal activity and the knowledge inputs.
Knowledge has a cumulative aspect to it, which can be explained by strong
firm routines and rigid structures, but cumulativeness may also arise from a
dependence on machinery and physical capital in other forms (Dosi 1982;
Patel and Pavitt 1997; Pavitt 1998). Thus, firms that rely heavily on their
physical capital may exhibit high levels of path-dependence partly because of
knowledge accumulation in physical capital. This kind of knowledge
accumulation lowers the probability of innovation through labour (i.e.,
knowledge) inflow. It also impedes the ability to absorb and utilize external
knowledge. Lagerholm (2007) showed that within mature industries (i.e.,
manufacturing in the Swedish context), process innovation is mainly the
result of supplier innovation rather than in-house innovation. Thus,
substantial knowledge advancements are brought to the plant via investments
in new machinery rather than through knowledge contributions by in-house
or in-flowing staff. In labour-intensive sectors, on the other hand, both
process and product innovations are mainly the result of in-house knowledge
applications, and the sheer number of innovative opportunities is thus greater
than in capital-intensive sectors. Therefore, the variety (relatedness) of
knowledge in industries that are dependent on their labour rather than their
14
physical capital will be characterised by a larger number of potential
applications of that knowledge.
Within firms, knowledge is administrated and structured in accordance
with the knowledge view and firm goals (Nonaka et al. 2000). The firm can,
thus, be perceived as an administrative unit (Penrose 1959) of individuals that
creates cohesion through its routines. It also influences the knowledge view of
its employees through these structures. When Nelson and Winter (1982)
originally presented their evolutionary theory of the firm, they perceived
routines as the capabilities of the firm and as the genes of the firm. This
implied that routines could both have gene-like generative qualities and be
manifested as firm behaviour. The ambiguity of the concept has persisted and
the interpretations and use of the concept are manifold, but there is consensus
on the context-dependence and regularity aspects of routines (Cohen et al.
1996; Dosi et al. 2008). Context-dependence and regularity suggest that
routines create a form of proximity (intra-organizational proximity) between
agents within the same firm and subsequently steer their view on knowledge
usability and applicability. This will be further discussed in the section on
proximity.
Within the conceptual evolutionary economics literature, routines are
often conceived of as the latent genes of the firm, synonymous with macrolevel social and cognitive maps of how to interpret problems and context,
manifested through firm behaviour. Whereas Pentland (1995) and Pentland
and Reuter (1994) liken routines to grammar, where a routine is not one
pattern but a set of possible patterns, Miner et al. (2008) consider
organizational routines to be the foundation of organizational learning
because routines function as organizational memory. These perspectives
agree on the role of routines as the rules, social and physical structures and
cognitive scripts and abilities of organizational members. These
organizational restraints create path dependence on the firm level, in both the
knowledge view and attitudes toward external knowledge. Routines at the
micro-level are often concerned with the individual’s explicit task and position
within a routine process and the performance of that task (Becker 2004). Most
empirical research has been carried out as case studies at the micro-level (e.g.,
Becker and Zirpoli 2008; Feldman 2000, 2003; Weick and Roberts 1993),
where specified routines have constituted tangible and measurable processes
from which inference has been drawn about their interconnectedness, their
adaptable and/or stable nature and the role of individual agency.
The dichotomy between micro- and macro-views of routines is misleading
due to the interdependent relationship between the two. Routines take place
in a context that makes explicit routines partly tacit through individuals’
interpretation of “how things are done” within in firm. In the everyday
workings of routines, individual members use their understanding of the
macro-level routine to adjust their performance at the micro-level in response
15
to circumstances and outcomes (Feldman 2000, 2003). In order for a routine
to be efficient, there need to be strong and continuous connections and
interactions between the different agents partaking in the process (Feldman
and Rafaeli 2002). The connections should preferably generate extensive
knowledge of the links between distributed activities within the organization
(Weick and Roberts 1993). This can be expressed as a high level of contextual
understanding, which presumably leads to extensive knowledge about the
individual’s and others’ role in the firm. Furthermore, organizational
members are aware of latent structures and “how things are done” in the firm,
which potentially allows rapid application of internal ideas.
2.5. Proximity and relatedness
In the present thesis, the analysis of learning and innovation processes
emphasizes the relations between individuals rather than using a full systems
approach. There is no doubt that the relations between individuals are
embedded in greater systems and that external knowledge input is necessary
for firms to sustain competitiveness (see, e.g., Bathelt et al. [2004], Grabher
[2002b; 2004] and Maskell et al. [2006] on the importance of interorganizational projects for firm innovativeness). However, by only focusing
on relational aspects of individual attributes in the labour force and thus
abstracting from a number of influences, an increased analytical clarity is
achieved. MacKinnon et al. (2002) argue that in order for learning to be a
relevant factor in explaining the spatial distribution of activities, the role of
knowledge needs to be carefully grounded empirically. By focusing on
individual knowledge and interactive learning within plants, the thesis
attempts to isolate the effect that learning has on plant performance. This
section presents the proximity dimensions framework. It also discusses the
role the proximity dimensions for interpersonal learning play on the microlevel of the plant.
The Russian developmental psychologist Vygotsky (1962) claimed that
learning occurs through interaction between internal cognitive processes and
the context. He introduced the idea that a person can gain a certain amount of
knowledge on her own through internal processes, but that through
interaction with others (helpers or instructors) there is potential for further
knowledge advancements. He called this potential further learning that could
be achieved through interactions the Zone of Proximal Development (ZPD).
In order to promote someone else’s learning, there must be pieces of shared
interests or mutual focal points from which to depart, otherwise
communication will fail. In reference to recent contributions within economic
geography, Vygotsky’s theory relates to the growing body of literature on
cognitive proximity and other forms of non-physical (relational) proximity.
When the distance between agents, in one or several of these non-physical
proximity dimensions, is just right, related –, learning is facilitated and
16
innovation more likely to occur than if there are long distances between the
dimensions (Boschma 2005; Harrison et al. 2001; Song et al. 2003; Torre and
Rallet 2005). Referring back to Vygotsky (1962), long distances in all
proximity dimensions cause a void between individuals that complicates
communication and learning. On the regional level, a short distance in
multiple proximity dimensions between firms may lead to a regional lock-in.
This is a scenario in which understanding and agreement between agents are
high, but equally high is the risk of myopia (Maskell and Malmberg 2007) and
an inability to identify new relevant technological knowledge (Boschma 2005;
Visser and Boschma 2004). The present thesis considers that the same
phenomenon is likely to occur on the interpersonal level of the plant.
Theories of non-physical types of proximity between firms were originally
introduced by the French Proximity Dynamics Group (Kirat and Lung 1999;
Torre and Gilly 2000; Torre and Rallet 2005). They distinguish between two
dimensions of proximity: the physical dimension, which is geographical
proximity, and the non-physical dimension, which is organizational
proximity. Their theory illustrates how interactive knowledge creation and
innovation processes are shaped by a relationship between relational factors
and physical distance. It is further argued that the relevance of the different
proximities changes in different stages of the innovation process. In their
framework, geographical proximity refers to physical distance in a twofold
manner: first: firstly, as perceived by individuals and thus partly socially
constructed, and second, as a separation of activities and relations in space.
The organizational dimension, on the other hand, refers to a relational
distance that affects the ability to coordinate actions. This dimension is
constructed in accordance with two related logics. First, similarity in firm
routines, which are partly built on shared beliefs and high degrees of tacit
knowledge, offer the same reference space and shared understanding of
structures and knowledge view. This is a perspective also taken by Blanc and
Sierra (1999). Second, firms that are close in organizational terms belong to
the same networks and have interactions of various forms. In the present
thesis, where labour force knowledge and relatedness between individuals are
under scrutiny, the former definition of organizational proximity is adopted.
The logic here is that individuals will benefit from having a shared
understanding of structures and knowledge view in their knowledge
interactions. A short organizational distance between individuals ensures
overlapping, or ideally identical, reference spaces. This proximity, or
familiarity, represents an institutional micro-context that will reduce
uncertainty and promote trust between individuals. Trust is a key factor in the
transfer of highly tacit knowledge and thus in interactive learning.
Multiple proximity models have been developed (e.g., Blanc and Sierra
1999; Kirat and Lung 1999; Zeller 2004) since the initial work of the Proximity
Dynamics Group. The most detailed account of the analytical relationship
17
between proximity and innovation is provided by Boschma (2005). In his
model, Boschma presents five proximity dimensions: geographical,
organizational, cognitive, institutional and social proximity. Geographical
proximity is defined as both absolute and relative distance, and the impact on,
and relationship with, the other proximity dimensions is underlined.
Boschma’s (2005) definition of organizational proximity differs from that of
Torre and Gilly (2000) in its focus on control of relations and hierarchy. In his
model, organizational proximity refers to the coordination and control of
intra- and inter-organizational relations. Organizational proximity is achieved
when firms are connected through coordinated channels and activities, and
this proximity reduces uncertainty and the risk of opportunism. This
description of organizational proximity is better suited to examination of
inter-firm learning processes and innovation. This is because it is associated
with the accessibility and power relations between firms that determine how
organizational networks are shaped.
The cognitive proximity dimension, as presented by Boschma (2005),
deals with differences and similarities in the capabilities of economic agents.
Antonelli (2000) and Howells (2012) have argued that knowledge is
distributed among different firms and industries. Firms learn and create
knowledge by combining existing pieces of complementary knowledge in new
ways (Nooteboom 2000). In order for firms to absorb external knowledge
(Cohen and Levinthal 1990) and combine it with internal knowledge, the
external knowledge must be similar enough – related – to the existing
knowledge to be recognized as valuable (Boschma and Lambooy 1999). The
cognitive distance should not be too long, nor should it be too short, because
short distances do not offer any learning potential. The institutional
proximity dimension refers to the degree of similarity in values, norms, rules
and laws between economic agents. It is, thus, regarded as being dependent
on both informal and formal structures. Boschma (2005) discusses the close
and intertwined relationship between institutional proximity and
geographical proximity. In the present thesis, they are regarded as
representations of the same phenomenon, namely similarity in norms, values
and understanding of formal and informal rules. Because the labour force and
labour flows are examined, the physical distance between firms influences the
probability that a person will change employer from one firm to another. But
the distance also represents an institutional similarity, or difference, that
affects the understanding of norms and practices among individuals and may
therefore influence the level of trust between individuals. As mentioned
previously, trust is thought to be necessary for the transfer of highly tacit
knowledge. Lastly, the social proximity dimension, as presented by Boschma
(2005), is associated with social ties between individuals resulting from
friendship and kinship. Thus, the social dimension is the only dimension that,
in its original form, specifically focuses on the micro-level of inter-individual
18
relationships. This proximity dimension is not further examined in the thesis,
but the other dimensions are considered on the micro-level of interpersonal
distances within the plant. The rationale for the micro-perspective is that, in
order to understand how firms learn, we need to examine the knowledge that
constitutes the firm and its absorptive capacity. Following Torre and Gilly
(2000) and MacKinnon et al. (2002), the thesis attempts to empirically
ground the concept of knowledge by further opening the black box of
proximity relations in relation to knowledge transfer and learning.
Each proximity dimension offers a way to construct shared values between
individuals and potentially also a shared basis for communication and
learning. In the proximity dimensions framework suggested by Blanc and
Sierra (1999), they identify a dimension that they call relational proximity:
“Relational proximity accounts for the existence of non-economic
relationships among individuals in the form of a common working ethos, a
common language and culture, good mutual knowledge, mutual trust,
mutually respected norms of behaviour” (Blanc and Sierra 1999;197). In the
present thesis, all of the above-discussed proximity dimensions fit under the
description of relational proximity, suggested by Blanc and Sierra (1999). This
relational characteristic implies that what two individuals lack in mutual
formal knowledge may be compensated for by mutual norms and codes of
conduct or a background in similar organizational contexts. Whereas
Boschma (2005) argues that cognitive relatedness is a necessary but not
necessarily sufficient prerequisite for learning to occur between firms, it is
argued here that, on the interpersonal level, relatedness in any one of the
proximity dimensions may be sufficient for learning to occur. The reason for
this goes back to Vygotsky’s ZPD and the need for a connection between
individuals in order for them to communicate efficiently, begin to trust one
another and subsequently learn from each other. The temporal aspect is also
crucial to the distinction between intra- and inter-firm learning. Within the
firm or plant, individuals remain for extended periods of time and inflows of
labour are regarded as, more or less, permanent additions to the labour force.
In inter-firm learning, numerous relations are temporary. Therefore, there is
a greater need for cognitive relatedness, which allows for efficient
communication of the unique knowledge that each project participant can
contribute. In the case of mergers and acquisitions, it is unlikely that two firms
with completely different core competences will merge. In such cases,
similarity or relatedness in the organizational or institutional dimension can
facilitate or impede learning processes (Boschma 2005) initiated by
relatedness in the cognitive dimension.
Regardless of the level of analysis, there is still great uncertainty about
whether or not the proximity dimensions are of equal importance to learning
and innovation. Moreover, few previous empirical studies have examined to
what extent proximity in one dimension can replace, or compensate for, a long
19
distance in another dimension. The theoretical literature on proximity
dimensions and relatedness between agents offers plausible interpretations of
the relational characteristics of knowledge. But many of the suggestions made
in the literature have yet to be empirically tested and grounded. The effects of
cognitive and geographical proximity on learning and innovation have been
the most theorized and empirically tested on both the interpersonal and
interfere level. On the interpersonal level, Boschma et al. (2009) found that
cognitively related in-house competences as well as cognitively related labour
inflow positively impacted plant performance. They further identified
interdependence between the cognitive and the geographical/institutional
dimension. Cognitively unrelated inflow that originated from the same local
labour market positively influenced plant performance. This is interpreted as
indicating that cognitive relatedness is not always a necessary condition for
interpersonal learning to occur. It is rather the combined influences of
multiple types of proximity/distance between individuals that determine the
learning potential. On the other hand, Neffke and Henning (2013) created a
measure of relatedness between industries, based on labour flows between
them. Their results indicate that firms are more likely to diversify into
activities to which they have pre-existing ties through their labour force than
into activities that diverge substantially from their core activities.
Furthermore, Neffke et al. (2011) showed that evolution of the regional
industrial setup is strongly path-dependent. They demonstrated that
industries that are related to pre-existing industries are more likely to enter a
region than those that are not.
In studies on labour mobility, the cognitive dimension is commonly
measured as relational differences in formal education, sector belonging,
occupation similarity or intensity of previous knowledge exchange between
actors (Boschma et al. 2009; Neffke and Henning 2013; Timmermans and
Boschma 2013). There is coherence in the view that, for fruitful
communication to occur, knowledge and skill should be neither too similar
nor too different. Therefore, bringing together people with related skills and
knowledge is the most efficient way to promote learning. In such instances,
people are able to understand one another while simultaneously drawing on
each other’s knowledge to advance their own knowledge, as suggested by
Vygotsky (1962). But the indicators for cognitive relatedness do not consider
the plethora of factors that influence individuals’ cognitive abilities. Merely a
fraction of an individual’s skills and abilities are measured with each of the
different techniques of estimating cognitive relatedness. This is mainly due to
methodological and empirical restraints and limitations and leads to subdivisions of cognitive relatedness. This is fine, as there is no way to include all
factors that influence individual ability, mainly because of the subjectivity of
knowledge and learning and partly because there are no registers and microdata that include such levels of detail. But a greater awareness of the
20
multifaceted ways of achieving cognitive relatedness, in particular, and
relatedness in other dimensions, in general, is important.
As in most empirical work, Paper I and II in the present thesis included
and analysed, analyse one aspect of cognitive relatedness. Because different
studies use different definitions of cognitive relatedness, there is a risk of overor under-estimating the effects of relatedness on firm/plant performance.
This is due to the exclusion of other factors that also influence the abilities and
skills of the agents under study. One such other factor is organizational
proximity that arises from similarity in firm routines. With regard to the
notion that firm routines are fundamental in creating organizational
proximity, there are several researchers who claim that knowledge within a
firm is distributed rather than dispersed, meaning that everyone knows some
information, but no one knows all of it (Becker 2004; Cohen and Bacdayan
1994; Feldman and Rafaeli 2002; Nonaka et al 2000; Lam 2000; Weick and
Roberts 1993). This unique context creates isolating mechanisms that hinder
not only external imitation of firm capabilities, but also internal articulation
of what these capabilities are exactly. The tacit nature of firm routines
generates intra-organizational proximity and possibly also a social proximity
that together create trust between individuals. This trust potentially allows
them to discuss ideas over wider cognitive spectra than would otherwise have
been possible.
Due to the relational and contextual nature of knowledge and the varying
degree of tacitness, knowledge is difficult to transfer and develop. Labour
mobility is thus the most efficient channel for dissemination of highly tacit
knowledge, but presumably also for dissemination of less tacit knowledge
(Szulanski 1996). Many have studied the transfer of embodied knowledge by
examining the effects of labour mobility on performance and innovation (see,
e.g., Agrawal et al. 2006; Almeida and Kogut 1999; Boschma et al. 2009;
Breschi and Lissoni 2009; Maliranta et al. 2009; Song et al. 2003;
Timmermans and Boschma 2013). Whereas Almeida and Kogut (1999)
showed that the mobility of engineers influenced local transfers of knowledge,
Power and Lundmark (2004) found labour mobility among key persons to be
the driving force behind the upgrading of skills within ICT clusters. Breschi
and Lissoni (2009) built on the results of Almeida and Kogut (1999) and put
forward the idea that labour mobility is the main conductor of local knowledge
transfer, superior to formal networks and social ties. Lacetera et al. (2004)
were able to quantify the effect that inflowing high-ability individuals can have
on their colleagues, thus showing that single individuals can impact the
knowledge of the entire workforce.
When knowledge is put in a new context, there is potential for completely
new applications of that knowledge, due to the new employee’s interpretation
of what is needed as well as the “old” employees’ understanding of the new
knowledge. Strambach and Klement (2011) argue that there is a combinatorial
21
knowledge dynamic at the micro-level of the firm. This works, in combination
with the cumulative knowledge dynamics inherent in the socio-technical
context of the region, to generate novelty (innovation). By bringing a specific
skill-set from one context to another, the individual potentially contributes to
the generation of new combinations of existing knowledge as well as to the
creation of entirely new knowledge (Asheim and Coenen 2005; Nonaka et al.
2000; Pinch and Henry 1999). But there is still uncertainty, not only about the
learning capability of the employees, but also about the external
circumstances (i.e., factors external to the individuals partaking in the
knowledge exchange) under which knowledge is most likely to be successfully
applied within a plant. The present thesis incorporates theories of sector- and
firm-specific structures into the description of the relationship between
physical and non-physical proximity and interactive learning in firms.
3. Methodological comments
The thesis is based on quantitative modelling of socio-economic processes
within and between plants. Methodology is closely related to the philosophy
of science, as different methods are used to capture different aspects of the
social world. The scientific methods employed in the thesis are shaped by
critical realism (Bhaskar 1978). The collective learning processes associated
with the knowledge composition within a plant are considered to be
contextual and cumulative and manifested in changes in plant productivity.
The view on learning as a process and practice driven by the variety of
accessible knowledge fits well into an evolutionary perspective on the
accumulation of knowledge and the growth of the economy, as discussed
earlier in the thesis. The evolutionary perspective has many common traits
with a critical realist ontology. Castellacci (2006) puts forth a view on reality
as complex, differentiated and structured, as a system that is open and
continuously changing and transforming, and as a world characterized by
great uncertainty.
In critical realism, the world exists independent of human awareness
(Sayer 1981). Furthermore, the world exists in three domains: the real, the
actual and the empirical (Collier 1994). The real domain contains all physical
objects, but also mechanisms and structures associated with the objects and
combinations of objects. Thus, it acknowledges the existence of non-physical
objects and processes in society. In the actual domain, events occur
irrespective of whether humans register them or not. In much social science
production, mechanisms in the real world are used to explain the development
of events in the actual world. Finally, the empirical domain represents the
events that are experienced by humans (Collier 1994).
Social science, in general, and evolutionary economic geography, in
particular, study phenomena in the empirical domain (the spatial distribution
22
of economic activities), the ambition being to understand the mechanisms
that underlie these phenomena (institutional settings, social networks,
embodied knowledge, etc.). Similar to the evolutionary perspective on
economic growth, critical realism regards science as constantly evolving.
Theories of the deeper domains are produced and reproduced based on their
plausibility in the empirical world. There is likely not one absolute truth in
social science, but there are explanations that are more plausible than others,
based on observations in the empirical world. Thus, constantly checking for
sources of error in the empirical setup and building on previous research are
essential elements in validating the theoretical claims made about structures
and mechanisms in the real domain (Collier 1994). Scientific explanations of
events offer theoretical descriptions of the mechanisms that caused them. By
empirically testing knowledge relations and innovations in firms, fragments
of the real world can be captured that, in turn, can inform us about probable
causal mechanisms. These causal mechanisms, found in both humans and
structures, can improve our understanding of pieces of the real world. By
combining insights from numerous theoretical descriptions, a synthesized
understanding of the functioning of the real world may be achieved (Lawson
1997).
3.1. The data
The empirical analyses in the three papers are based on data from the
ASTRID database at the Department of Geography and Economic History at
Umeå University. The database integrates several administrative registers
maintained by Statistics Sweden (SCB) and offers detailed, geo-referenced,
longitudinally linked information about all individuals, plants and firms in
Sweden. By integrating registers, it is possible to connect attributes to people,
and people to plants and firms with attributes of their own. The longitudinal
aspect of the database enables analyses of changes in work force composition
within individual plants, as well as recording of individuals’ past work
experiences with regard to plant, industry, and firm. Because it includes
recorded coordinates for all buildings, residences and plants, the database has
a high spatial resolution. This enables analyses of mobility patterns and
agglomeration activities. The data in ASTRID are measured and encoded on a
yearly basis, and the values of the variables in the database reflect the situation
at a specific point in time. Thus, it is not possible to see whether a person has
had multiple employments in a year. But by comparing personal wage and
social benefits, it is possible to determine whether a person has been employed
and unemployed during the same year. The analytical unit in all the empirical
papers is the plant. A plant is traditionally associated with manufacturing, but
in the database and in this thesis it a workplace. This workplace can be of any
size and belong to any industry. The plant represents a physical building or
part of a building that is administrated by a firm. A firm can have multiple
23
plants, which all have their unique plant identification number (CFAR), but
share the same organizational identification number (OrgID). Individuals are
associated with plants, not with divisions or subunits within a plant.
Therefore, it is not possible to determine whether a person has changed jobs
or positions within a plant, but intra-firm mobility between plants can be
detected. Mobility is captured as a change in plant identification number and
change in plant coordinates for individuals in the database, and consequently
internal mobility (intra-plant mobility) is concealed in the data, not just in the
analyses conducted in the thesis but in the structure of the database in general.
The intra-firm mobility between plants is recorded as a change in plant
identification number within the same firm identification number.
The Swedish labour market is characterized by fairly high job mobility
rates, which correspond to those in several other north European countries,
e.g., Denmark, Finland and the Netherlands (EUROFOUND 2006). Between
1988 and 2012, 68 percent of all employees who ended their employment did
so to change jobs, not due to retirement or layoffs triggering unemployment
(Andersson et al. 2014). Thus, job change is the dominant type of labour flow
on the Swedish labour market. Furthermore, job mobility is more prevalent in
some segments of the labour market than in others. Highly educated
individuals show higher job change rates than do the less educated. Men
change jobs more often than women do, younger individuals more often than
older people and highly paid individuals more often than those in the lower
wage segments (Jansson 1997; Widerstedt 1998). In the analyses of job
mobility (Paper I and II), only mobility related to job change is considered.
Thus, transition from unemployment to work is excluded from the analyses,
whereas change of workplace is included.
The empirical analyses are based on data covering the period between 2000
and 2010. Paper I uses data from the period 2002-2005, whereas the analysis
in Paper II is based on data from 2003-2008. Paper III examines data from
the period 2000-2010. Thus, the results in Paper II and III could,
theoretically, be affected by the financial crisis that struck the labour market
in 2008 and 2009 (Öhman 2010). But the Swedish labour markets and
financial markets were relatively mildly impacted by the global crisis. Between
the middle of 2008 and the end of 2009, 120,000 people lost their jobs. This
can be compared to the crisis in the early 1990s during which 600,000 people
lost their jobs (Öhman 2010). During troughs, job mobility decreases
(Andersson and Tegsjö 2006), but in Paper II all job changes were recorded
between 2003 and 2006, and only plant performance was recorded up until
2008. In Paper III the empirical analysis is based on the in-house composition
of knowledge in knowledge-intensive business services (KIBS). Thus the flows
are implicitly considered in the yearly estimations of in-house knowledge
variety. The performance of the plants may have been negatively impacted by
the recession, which may have led to an underestimation of the impact of
24
different forms of knowledge variety on performance. But the empirical setups
in the two papers have mitigated the risk of any severe underestimation. In
Paper II a longitudinal panel was estimated, and in Paper III pooled OLS
regressions with observations over a ten-year period and yearly dummies to
control for time-specific heterogeneity were calculated.
In Papers I and II the focus is on labour mobility and its impact on plant
performance. The models used in these papers do not control for the exit of
labour from the plants, only the entry. Thus, in some cases the inflow of labour
replaces exiting employees, whereas in other cases the inflowing labour is a
new addition to an expanding plant. There are a few central concepts in the
papers that are discussed in greater detail below in an attempt to visualize the
transformation of theoretical concepts into quantitative variables.
3.1 Measuring knowledge and ability
This section deals with the concept of knowledge and ability and to what
extent knowledge can be assessed through quantitative methods. The central
role of knowledge in the learning economy has been emphasized throughout
the thesis. The European Commission states that:”…the key aspect of human
capital has to do with the knowledge and skills embodied in people and
accumulated through schooling, training and experience that are useful in
the production of goods, services and further knowledge” (De la Fuente and
Ciccone, European Commission 2002, p7). Formal measures, such as
academic degrees and occupation titles, offer guiding information about the
theoretical and practical scope of parts of an individual’s knowledge. Relating
back to the critical realist ontology, an individual’s formal knowledge
(education, occupation or industry experience) is a signal (Spence 1973) in the
empirical domain, which allows the researcher to make informed assumptions
about a portion of the individual’s ability. Arrow (1973) introduced the idea of
the educational system as a filter through which the most able students pass,
whereas the less capable individuals are caught in the filter and prevented
from further knowledge advancements. In Spence’s (1973) article on job
market signalling, academic degrees are signals to employers about the ability
of the individual to perform. Over the years, the individuals change and adapt
their signals with increased experience and additional work titles. Thus,
formal measures of knowledge are conscious efforts on the part of the labour
force to signal not only ability, but also acquired knowledge to employers.
Factors that are believed to socialize knowledge and affect learning are, to
some extent, accounted for through the integration of proximity dimensions
other than the cognitive dimension into the analytical models. But the main
focus of the papers is on the relationship between formal knowledge variety
and plant performance. Educational background is used to estimate
relatedness in knowledge between individuals in Paper III. Educational
background and industry experience are used as measures of knowledge in
25
Paper I and occupational title is used in Paper II. Whereas educational
background represents formal theoretical knowledge, industry experience
informs us of how the individual’s knowledge has been practically applied.
Occupation title considers both the applied knowledge through the work tasks
and the individual’s qualification for those tasks (educational background).
These three formal measures describe parts of the individual’s total knowledge
that is codified and actively acquired. They accurately represent knowledge
that the individuals have obtained, but they do not represent all the knowledge
that the individuals possess. The formal measures of knowledge tell us little
about the social and tacit components of knowledge (Grabher 1993; Polanyi
1962). In Paper III aspects of the social character of knowledge are touched
upon through the integration of individuals’ past experiences. These past
experiences have influenced and shaped employee knowledge beyond the
information in the strictly formal records. By tracing employees’ former coworkers an image of the context in which they have been active emerges,
which allows for an analysis of the relative similarity of previous work
experiences in the current labour force within a plant. It is suggested here that
this is a way to approach the tacit dimension of knowledge and ability in a
quantitative manner. Past work experience informs us about how formal
knowledge has been applied. It also offers insights into the knowledge
networks of individuals. The networks indicate what parts of the individual’s
own formal knowledge may have been accentuated and developed further as
a result of these interactions.
In contrast to Becker’s (1964) human capital theory, in which he claims that
individuals weigh the costs and benefits of investments in higher education,
the Swedish context offers few costs associated with higher education because
education is free. The possibility for student loans and social and economic
benefits for students are the same for the entire Swedish population. The
general educational level in the population is high, and the individual reward
(income) of additional years of schooling is below average in the European
Union (Björklund et al. 2000; Eliasson 2006). The European Commission
found the average return on each year of schooling to be a 6.5 percent increase
in personal income in the European Union (De la Fuente and Ciccone 2002).
3.2 Estimating proximity in the labour force
In this section the characteristics of the labour force and the various forms
of proximity that are estimated in the papers are presented and discussed.
Throughout the empirical papers (Papers I-III) various forms of proximity
dimensions are referred to and their effects on interactive learning are
discussed. A short distance in any of the dimensions implies similarity
between agents. Boschma (2005) argues that similarity reduces uncertainty
between actors. Actors can be firms, organizations and governmental
institutions, but also individuals. Whereas Boschma (2005) mainly discusses
26
the role of the various forms of proximity on the inter-firm level, the present
thesis examines the role of proximity in interactive learning processes on the
inter-individual level within plants. The plant is an administrative unit of
inclusion in the sense that it generates cohesion between its members
(employees). This is achieved through a common understanding of firm
routines and the technological trajectory of the specific industry in a given
region. From this rationale it follows that individuals who enter the plant, as
newly employed labour, may diverge from the existing staff. Apart from
differing in formal knowledge, new employees may also have different
experiences of firm routines and institutional settings that originate from a
different firm in a different region. The efficiency of absorption of individual
knowledge into plant knowledge is affected by the degree of similarity between
the inflowing labour and the existing in-house labour. In the remainder of this
section, the operationalization of geographical, cognitive and organizational
proximity is elaborated upon and discussed.
Geographical proximity is estimated between inflowing labour and the
existing in-house labour in Paper I and Paper II. The inflowing labour is
categorized as being either intra-regional or inter-regional with regard to local
labour markets. Thus, the inflows are considered proximate (similar) or
distant (different or unrelated) in a dichotomous manner. The rationale for
this division is that the geographical proximity dimension is closely related to,
not to say intertwined with, the institutional proximity dimension (Boschma
2005; Farole et al. 2010; Rallet and Torre 1999). In the regional innovation
systems literature, the varieties in regional (or local) institutional settings are
emphasized in the competitiveness of firms and industries (Asheim and
Gertler 2005). This aspect of institutional proximity may be of little
importance to interpersonal learning processes, but belonging to a community
and enjoying insider advantages within one’s own labour market are aspects
of the geographical proximity dimension that are considered in the thesis.
Inter-regional labour mobility represents a scenario in which the individual
enters into a new community from a context that the new co-workers know
nothing or little about. Using the same logic, it is argued that intra-regional
labour flows preserve the insider advantages and remain within the same
institutional environment or community. The familiarity associated with
intra-regional labour flows presumably reduces uncertainty and generates
trust between the in-house labour and the inflowing labour.
The non-physical proximity dimension of cognitive proximity is estimated
in various ways in the papers. The reason for this is to show that different types
of knowledge can generate cognitive relatedness between individuals, through
which they can learn from each other. In Paper I the in-house knowledge
composition is calculated as an entropy measure of variety (as suggested by
Frenken et al. 2007). The estimations of relatedness in the in-house labour
force are based on educational background, whereas the relationship between
27
the inflowing labour and the existing in-house labour is based on industry
codes (as done by Boschma et al. 2009). It is assumed that a knowledge inflow
from related industries is efficiently absorbed while simultaneously
contributing new pieces of knowledge that can be combined with the existing
knowledge in new ways. In Paper II, entropy measures based on occupation
are used to determine the degree of relatedness in the in-house labour force.
The same entropy calculations as in Paper I are used to determine the degree
of similarity, relatedness and unrelatedness in the in-house composition of
knowledge. Occupation is a slightly more refined measure of knowledge than
educational background, as it reflects both work tasks and individual
qualifications. The relatedness between the inflowing knowledge and the
existing knowledge in Paper II is estimated using the same relational entropy
measures that Boschma and Iammarino (2009) used when determining the
relative similarity of trade linkages and the regional industrial setup. The
inflowing labour and its occupation titles are compared to the occupational
distribution within the plant. By so doing, the levels of similarity, relatedness
and unrelatedness are determined between the inflowing labour and the
existing labour.
In Paper III, in-house knowledge variety is estimated in three different
ways. This is done to demonstrate that cognitive relatedness can be achieved
in more than one way. Individuals’ experiences shape their cognitive abilities
(March 2010), and it is argued in Paper III that, in an evolutionary framework,
the knowledge of the labour force should be regarded as evolving over time
and, thus, as being influenced by past events. Just as firms, industries and
regions are affected by their past history, so are the micro-agents, the
individuals, which constitute the labour force. The first measure of cognitive
relatedness is an entropy measure of formal education variety, exactly like the
in-house estimations in Paper I. The degree of similarity, relatedness and
unrelatedness in formal education in the labour force are followed by an
experience measure of variety. The similarity in industry experience is
estimated in the in-house labour force. This type of knowledge represents
applied knowledge. For each employee a fifteen-year backlog is retrieved from
the database that allows us to see what industries the individual has worked
in, during the past fifteen years. The industry experiences of all employees at
the plant are combined in an “experience pool”. From this pool of industry
experiences, the degree of similarity, relatedness and unrelatedness in labour
force industry experience are determined. The third measure of cognitive
proximity is a measure of similarity in knowledge networks in the in-house
labour force. This type of knowledge represents variations in experiences in
practices and epistemic communities within the current place of work. Similar
to the industry experience measure, the knowledge network is calculated from
a pool of knowledge exposures that each employee has been subject to during
the past fifteen years. The educational background of every person who has
28
worked with the individual under study is recorded and added to the pool of
knowledge exposures at the plant. The same entropy measures, as for the
other two types of cognitive proximity, are used to estimate the degree of
similarity, relatedness and unrelatedness in knowledge exposure. The
rationale for estimating cognitive proximity in three separate ways is to
demonstrate that knowledge relatedness can be achieved in multiple ways.
Thus, a labour force that is characterized by high levels of unrelatedness in
one of these measures may be related through another type of cognitive
relatedness.
Organizational proximity is estimated in Paper III in a similar fashion to
geographical proximity, as explained above. Individuals are regarded as being
either organizationally proximate or organizationally distant in a
dichotomous manner. Intra-organizational labour inflows are individuals
changing plants within the same firm. These flows represent proximity in the
organizational dimension. The inter-organizational flows are individuals
originating from another plant and firm. These flows are organizationally
distant. Organizational proximity is considered to generate a mutual
understanding of “how things are done” and a coherent knowledge view and
awareness of firm norms and values. These aspects create combined
institutional and social cohesion within the micro-context of the firm. On the
interpersonal level, this is presumed to reduce uncertainty and contribute to
efficient communication. Inter-organizational labour flows originate from
other micro-institutional environments that may differ significantly from the
new context. Therefore, these flows need to familiarize themselves with the
new structures and internal networks and their initial interactions with the
existing in-house labour will entail more uncertainty.
3.3 Measurement of innovativeness
Before proceeding to the measurement of successful learning processes at
the plant, the relationship between knowledge and innovation is considered.
According to Pavitt (2005), innovation processes contain three overlapping
sub-processes: production of knowledge, conversion of knowledge into
physical and non-physical artefacts and the matching of these artefacts to
market needs and demands. Thus, all innovation processes depend on the
acquisition, creation and application of knowledge. This knowledge is
disseminated on the market and conversed within the firm or plant (Nonaka
1994). It is important to keep in mind that knowledge creation and acquisition
do not occur at one specific point in time during the innovation process, but
rather throughout the entire process. This has been both empirically
demonstrated (see Butzin and Widmaier [2012] on innovation biographies)
and theoretically argued (e.g., Kline and Rosenberg [1986] critique the linear
model of innovation [Bush 1945]).
29
On a related note, the reverse product cycle (RPC) introduced by Barras
(1986; 1990) in the service industry also enters into traditional
manufacturing. RPC constitutes three successive phases: improve efficiency,
improve quality and, lastly, new services. The order of these stages is opposite
to that of the product cycle model for manufacturing introduced by Abernathy
and Utterback (1978). But approximately 35 years after the introduction of the
product cycle model, the reach of capitalism and the increasing pace of
Schumpeterian creative destruction have impacted the product cycle in
manufacturing. The cost associated with improving the quality of existing
products and the efficiency of production processes are substantially lower
than they are in relation to developing new products. Therefore, improving
quality, improving efficiency, finding new markets, creating services around
existing products and making incremental changes to existing products are
viable options that generate noticeable results within a foreseeable future.
This development, one of increased global competition, indicates that
collective learning processes, which generate incremental innovations within
plants, not only can but do occur in all sectors. But to pinpoint exactly how
and when essential bits of knowledge are exchanged or recombined is
impossible using quantitative methodology, and also beyond the scope of the
present thesis. Instead, the thesis rests on the assumption that learning occurs
in all sectors, and that it can be detected in the performance of plants due to
the predominance of innovations that increase efficiency.
To estimate and evaluate the potential increase in efficiency due to
collective learning processes, plant performance is specified as labour
productivity at the level of the plant in Paper I, II and III. The database does
not contain information on productivity, so performance is instead measured
as value added per employee. But the database only contains information
about value added for firms, not their respective plants. For firms with only
one plant this is unproblematic, but for multiplant firms the database ascribes
the value added of the entire firm to all of its plants. This is dealt with in Paper
I, Paper II and Paper III by distributing value added to plants in accordance
with their wage quota. By ascribing value added to plants based on their wage
quota, both the education and experiences of the employees are considered
(Wictorin 2007). As discussed in a previous section, individuals signal their
knowledge and ability (Spence 1973) and the employer is ensured of their
ability through the filtering function of higher education (Arrow 1973). Highability individuals are efficient (Becker 1962), which is recognized by the
employer, who awards them with higher salaries. High efficiency is a mixture
of innate ability, specific knowledge related to educational degree and generic
knowledge in communicating, learning and synthesizing knowledge inputs.
Following this logic, it is reasonable to allocate value added in proportion to
wage quota. In Papers II and III, value added was also proportionally
allocated to plants with regard to the share of firm employment (as done by
30
Martin et al. 2011). This approach is perceived as being more egalitarian,
because every employee is valued as being equally important to the firm. This
was done as a robustness check, and the allocated value added remained the
same in Paper III. The reason for this may be found in the sample of firms in
Paper III, which only consists of KIBS. Unlike a manufacturing firm with
multiple plants with different functions, firms in KIBS will most likely have
similar functions and human capital levels at all their plants. In Paper II the
models were estimated with the alternative productivity measure, and the
trends in the models did not change. The procedure for allocating value added
is therefore deemed fit for the purpose of the papers.
4. Paper summaries
4.1. Paper I: Labour mobility and plant performance: On
the (dis)similarity between labour- and capital-intensive
sectors for knowledge diffusion and productivity
The first paper, Paper I, analyses the differences between capital- and
labour-intensive sectors with regard to the impact of workforce knowledge
composition and labour mobility on plant performance. As discussed in the
theoretical background, knowledge relatedness between individuals is
assumed to be beneficial to knowledge absorption and learning. But the
conditions under which new knowledge can be applied and generate novelty
presumably vary between different industries and firms. In the literature on
proximity and relatedness, especially papers focusing on individuals and
labour flows, knowledge transfer and learning are often treated as if they occur
in isolation, detached from any surrounding context. This disregard of the
time-place fusion in which every action is carried leaves the analyses
incomplete and at risk of over- or underestimating the effects of knowledge
and relatedness on plant performance. The aim of the paper is to bridge a
perceived gap between micro- and macro-level theories of knowledge
generation and innovation in the space economy and thereby to nuance the
debate on non-physical proximity and the labour force as a knowledge
resource. Whereas there is one strand of literature advocating the benefits of
complementarities and relatedness for plant performance (Boschma et al.
2009), there is another strand discussing the different prerequisites for
innovation in different technological regimes (Castellacci 2008; Pavitt 1984).
These strands are neither contradictory nor opposing, but they rather exist in
parallel and should be combined if one wishes to model the context in which
individuals interact and learn to promote innovation and productivity.
Through the integration of these theories in conjunction with theories of the
effects of labour mobility, the paper shows that learning and innovation do
not occur in isolation. Everything takes place in a context determined partly
by the individuals involved, but also by the main technology used by the firm.
For the empirical part of the paper, geo-referenced longitudinal employeremployee micro-data are used. All firms in Sweden are categorized as being
either capital-intensive or labour-intensive. The effects of labour flows
31
between plants, within and between labour markets, are then analysed for
capital-intensive and labour-intensive sectors, respectively. The sample
contains all plants in Sweden with 10 employees or more that have had a
labour inflow in 2003 and value-added for both 2003 and 2005. The analysis
is carried out using weighted least square (WLS) regression analysis. An
additional variance analysis (ANOVA) is performed to determine the relative
effect of each of the co-explaining factors on plant performance.
The aim of the paper is achieved by testing the effects of in-house
constellation of knowledge on productivity, of type of knowledge inflow on
productivity and of geographic origin of the inflow on productivity. By so
doing, the effects of cognitive and geographical proximity in the labour force
on plant performance are tested in the labour-intensive sectors and the
capital-intensive sectors, respectively. The assumption is that the labourintensive sectors will show overall higher effects of labour inflow on plant
performance than will the capital-intensive sectors, which rely heavily on the
efficiency of their physical capital (machinery) and supplier ingenuity for their
productivity.
The results presented in Paper I show that knowledge type (similar, related
or unrelated) and knowledge origin (local or non-local) have different effects
on plant performance in the two sectors. A high degree of related knowledge
in the in-house workforce positively impacts plant performance in both
sectors, but the relative effect (when compared to other explanatory factors)
is larger in the labour-intensive sectors. The analysis of labour inflow indicates
that knowledge in the capital-intensive sectors is localized – only intraregional labour flows give rise to increased plant productivity. In the labourintensive sectors, the geographic and cognitive dimensions complement one
another; similar knowledge needs to be non-local in order to be beneficial to
plant performance, and unrelated knowledge mainly contributes to plant
productivity growth when it originates from inside the region. The findings
presented in the paper indicate that there are substantial differences between
the capital-intensive sectors and the labour-intensive sectors. With regard to
the analyses of in-house skill portfolio, related variety has considerably more
influence on firm performance within the labour-intensive sectors.
The overall weak effects of knowledge on performance in the capitalintensive sectors may be attributed to high levels of cumulativeness of
knowledge stored in the physical capital.
4.2. Paper II: Labor mobility and organizational
proximity: Routines as supporting mechanisms for variety,
skill integration and productivity
In Paper II the context-dependent nature of knowledge is further
scrutinized through examination of the role of firm routines (proxied as intraorganizational proximity) in learning and innovation. The aim of the paper is
to contextualize labour force knowledge by focusing on the socializing and
structuring nature of firm routines, the goal being to offer new insights into
theories of knowledge as a resource, in general, and theories of knowledge
32
transfer via labour mobility, in particular. The impact of knowledge inflow,
which is assumed to have an innovative effect, is measured as changes in per
capita productivity at the plant. By operationalizing familiarity with firm
routines as intra-firm labour flows, and contrasting them to unfamiliarity with
firm routines, operationalized as extra-firm labour flows, knowledge transfer
is studied in light of not only cognitive and geographic proximity, but also
organizational proximity.
Paper I showed that there are differences in the effects of related knowledge
inflow in capital- and labour-intensive sectors, respectively. This supports the
hypothesis that learning and innovation depend on more than just the
relatedness of formal knowledge between the partaking individuals. The
probability of successful application of new knowledge is also contingent on
the firm’s routines, which steer the knowledge view and shape patterns and
processes of communication within the firm. Whereas studies in economic
geography have shown the importance of having a related knowledge base
within a plant, as well as related knowledge entering the plant (Boschma et al.
2009; Eriksson 2011; Timmermans and Boschma 2013), the literature in
organization studies indicates that firm routines constrain and/or enable
learning through the rules, social and physical structures and cognitive scripts
and abilities of organizational members (Dosi et al. 2008; Miner et al. 2008;
Nelson and Winter 1982; Pentland and Reuter 1994). The firm can be
perceived as a social community specialized in the generation and transfer of
knowledge. In the present paper, familiarity with firm routines is perceived to
generate intra-organizational proximity between organizational members.
This gives internal labour flows the advantage of a shared context with the inhouse staff. Internal labour flows thus have the benefit of a shared social and
cognitive map and knowledge of firm goals with the in-house staff. This
advantage presumably offers a shared basis of understanding and eases
communication, which alleviates the need for related formal knowledge in
order for learning to occur.
The analysis employs longitudinal micro-data from the ASTRID database
at the Department of Geography and Economic History at Umeå University.
For the empirical estimations, the Arellano-Bond dynamic panel generalized
method of moments (GMM) is applied. The analysis is based on 8,633 internal
job moves and 9,418 external job moves to 33 plants and four firms, offering
a total of 99 observations, in Sweden between the years 2003 and 2006.
Together with a group of co-explaining factors, the effects of cognitively
differentiated intra- and inter-firm labour flows on plant performance are
determined. The effects of the inflow variables are estimated at three points in
time: the same year as the inflow occurs, the year after and two years after;
this is done because internal flows are believed to be integrated more rapidly
and produce, more or less, instant productivity gains.
The results indicate that the theoretically positive relationship between
cognitively related labour flows and plant performance is not unaffected by
the inclusion of an organizational proximity dimension in the analysis.
Cognitively related knowledge inflow positively impacts productivity when it
33
originates in another firm and thus is organizationally unrelated. The positive
effect on productivity is delayed, theoretically due to the longer time needed
for absorption and conversion when the cognitive dimension is the only
shared dimension. Cognitively related intra-firm labour flows (individuals
changing plants within the same firm) are familiar with firm routines and thus
organizationally similar (a short distance in the organizational dimension).
This combination of similarity in the organizational dimension and
relatedness in the cognitive dimension generates too much similarity for any
innovative learning to occur. The findings suggest that cognitively related
labour flows are primarily economically beneficial between firms. In such
cases, cognitive relatedness is enriched by sets of firm-specific norms and
routines associated with the previous employer that allow for rich
combinations of knowledge at the new plant.
When a third proximity dimension is added to the models, namely
geographical proximity, the effects of cognitively related inflows are positively
reinforced in the extra-regional model. This indicates that relatedness in the
cognitive dimension is mainly necessary when there are no other socializing
proximities present, i.e., organizational or geographical proximity. In the
absence of organizational and geographical proximity, cognitive relatedness is
necessary and sufficient to generate effective learning and application of
knowledge. These results imply that a cognitively related labour inflow
originating from afar (firm and place) contains a greater variation, associated
with the past place and firm, than does intra-firm or intra-regional inflows of
cognitively related knowledge. Throughout most of the analysis there is a
positive relationship between cognitively unrelated flows and plant
performance in the intra-firm models. These flows are organizationally
proximate and absorbed instantly and efficiently, generating productivity
growth for the plant over consecutive time periods. The proximity in the
organizational dimension allows for unrelated knowledge to be combined and
applied in a familiar setting. When the inflow is still organizationally
proximate but geographically distant, the positive effects of cognitive
unrelatedness disappear. This indicates that distance (unrelatedness) in two
dimensions, of which one is the cognitive dimension, cannot be overcome by
proximity (similarity) in the organizational dimension.
Another outcome of Paper II is the framing of the intrinsic nature of the
proximity dimensions with regard to learning and knowledge transfer in the
labour force. Without any hierarchical ranking of their relative importance, it
has been shown that multiple proximities exist that can promote learning and
affect plant performance. The effect of the cognitive dimension can be reduced
or amplified by the distance in other dimensions. In the paper, a combination
of organizational and geographic proximity was shown to generate a state in
which cognitively similar and related knowledge inflows either were
detrimental or insignificant for productivity. It is concluded that cognitive
relatedness is not a necessary condition for learning to occur in the labour
force.
34
4.3. Paper III: Relatedness through experience: On the
importance of collected worker experiences for plant
performance
In Paper III the role of labour force experience in learning and innovation
in firms is examined. In contrast to the two previous papers, Paper III does
not explicitly study labour mobility, instead in-house competences are
observed over time. The formal cognitive abilities of employees, as expressed
by their education, are supplemented with cognitive abilities in the form of
employees’ previous work experiences. These experiences are divided into
industry experience and past knowledge exposure associated with previous
workplaces. Theory clearly states that experience is an elemental part of
individuals’ interpretative abilities (Polanyi 1962; March 2010) and should
thus be considered when analysing the in-house innovative potential. Past
experiences are essential in forming individuals’ abilities to seize ideas,
communicate them and transform them into productive practice. Therefore,
in the present paper, the cognitive proximity dimension is divided into three
forms of knowledge, each with its own variety and distance. The experience
variables represent socialized cognitive abilities that have been used in
processes and practices at plants. Furthermore, interaction terms are
constructed and their impact estimated in order to determine the potential
overlap effects of different forms of cognitive proximity. By analysing the
interaction effects, Paper III not only shows that the combined effects of
different forms of proximity differ from the independent effects, but it also
shows that the impact of one form of knowledge on plant performance is
conditional on the level of another knowledge variable.
The reason for Paper III is the vast interest in knowledge as a valuable
resource for firm performance. This interest has gradually shifted focus from
the accumulation of human capital to the inflow and accumulation of related
and complementary knowledge, as the most important determinants of
knowledge transfer, learning and innovation (Boschma 2005; Harrison et al.
2001; Nooteboom et al. 2007; Tanriverdi and Venkatraman 2005). Paper III
builds on the theory that a shared practice can be achieved through related
cognitive abilities between individuals (Boschma 2005), represented as past
work experiences. Shared practice is manifested in an understanding of how
things are done, familiarity with firm goals, coherence in knowledge view and
a good ability to communicate. It is further suggested, in the paper, that
relatedness in formal knowledge between employees facilitates
communication, whereas relatedness in industry experience and knowledge
exposure offers a plethora of possible applications of formal knowledge and
understandings of multiple practices. Therefore it is assumed that cognitive
relatedness within the firm can be achieved in different ways by highlighting
different aspects of employees’ total cognitive abilities.
The empirical analysis, a weighted pooled OLS regression with year-,
industry- and region-fixed effects, is based on geo-referenced longitudinal
matched employer-employee data. In the study, the abilities and experiences
of employees in knowledge-intensive business services (KIBS) are analysed
35
for firms in Sweden between 2000 and 2010. Firms in KIBS depend on their
employees’ ability to learn and apply knowledge for their performance, as they
mainly engage in consultancy activities and research in various knowledge
fields. Thus, the micro-diversity within plants is more thoroughly scrutinized
in the present study than in the previous ones because the employees are,
more or less, the only resource for many of these firms. In addition to the
examination of staffs’ formal abilities, a 15-year backlog of industry experience
and past knowledge exposure for every current employee is created and
analysed to determine the degree of relatedness in industry experience and
knowledge exposure.
The results show that high human capital ratios positively impact plant
performance. Considering that these firms mainly produce and sell
knowledge, the results are not surprising. The resource of knowledge
presumably promotes not only technological development, but also economic
growth in both firms and regions in all industries. Thus, high human capital
ratios represent the knowledge that has accumulated in a place over time. In
this instance, the place is a firm, or rather a plant. But when the level of human
capital is put in the same model as the various forms of knowledge variety, a
substantially more nuanced picture of the workings of knowledge on plant
performance emerges. It is shown that the effects of the different forms of
knowledge composition variables on plant performance are conditional on the
level of human capital at the plant. That is to say that the effect of, e.g., high
levels of similarity in formal knowledge on plant performance varies as a
function of level of human capital. This is something that previous theories of
proximity dynamics have put forward, but been unable to empirically test. The
results further indicate that the effects of the negative relationship associated
with high levels of similar or unrelated formal knowledge on learning are
abated as the level of human capital increases at the plant. High levels of
similar formal knowledge and human capital even generate productive
learning processes in plants in KIBS. In previous studies on the impact of
labour force knowledge on plant performance, the different forms of
knowledge have been tested separately. By combining them and creating
interaction terms, Paper III is able to more accurately represent the actual
conditions for learning and innovation at a plant. Not only are the different
forms of knowledge combined and tested simultaneously, but the multiple
ways of accomplishing and measuring cognitive relatedness are also
recognized. In most existing empirical work, high levels of similar knowledge
have negatively impacted plant performance. But by combining multiple
forms of knowledge and estimating their relationship with performance
simultaneously, this study has been able to show that the impact of similar
knowledge is conditional on the human capital ratio at the plant. The
conditional relationship between human capital levels and the cognitive
relatedness of the labour force have not been empirically tested previously.
36
Instead most studies have chosen to include either human capital or cognitive
relatedness in the labour force when estimating the effects of interactive
learning on plant performance.
5. Concluding discussion:
Relatedness put in place
The present thesis set out to examine the relationship between different
types of proximity in the labour force and the competitiveness of firms. The
rationale for the micro-level focus is the assumption that interactive learning
processes within plants occur under different conditions than inter-firm
learning processes. Therefore, the relationship between interactive learning
and the proximity dimensions is assumed to differ from that for inter-firm
learning. Moreover, it is suggested that the functioning of the proximity
dimensions varies between technological and organizational contexts. The
workings of knowledge dynamics at the micro-level of the plant have been
scrutinized in three empirical papers. The empirical findings indicate that the
proximity dimensions are interconnected and that cognitive distance in the
labour force is not the only factor that determines the efficiency of interactive
learning processes. The geographical and organizational distances between
inflowing labour and in-house labour also influence the potential for
interactive learning. A third factor that impacts the effect of learning on
performance is the degree of dependence on physical capital associated with
sector belonging. It is demonstrated that capital-intensive sectors benefit
substantially more from having a local labour inflow than a cognitively related
inflow. Furthermore, it is established that relatedness is not necessary for
interactive learning to occur, neither in the cognitive dimension nor in any of
the other dimensions examined. Instead, it is the combined distances of
multiple proximity dimensions that determine the degree of interpersonal
learning.
The next section is structured around a discussion of the key findings in the
papers and their implications for interactive learning in firms. The subsequent
section discusses the findings in relation to the analytical framework of the
proximity dimensions and evolutionary economic geography. It focuses on
how we can understand firm competitiveness through analysis of the labour
force. Moreover, the relationship between the findings and the spatial
development of economic activities is outlined. The last section offers
suggestions for future research.
5.1 Findings
When the three papers are considered together, five main findings emerge.
First, as theoretically suggested by Boschma (2005) and Torre and Gilly
(2000), the proximity dimensions are interrelated and can influence one
another. It is evident that the geographical dimension influences the impact
37
of the cognitive dimension on plant performance. The observed relationship
is captured in Paper I and Paper II and is supported by previous empirical
findings by Boschma et al. (2009) and Eriksson (2011). It is also shown that
the organizational dimension influences the cognitive dimension. The
interdependence is manifested through the variable impact on plant
performance that a given distance in one dimension has, depending on what
other type of proximity is accounted for in the model. Put differently, a short
distance in one dimension may require a long distance in another dimension
in order for interactive learning to take place.
In Paper I and Paper II it is demonstrated that inflows of unrelated formal
knowledge positively impact plant performance if they are also geographically
proximate. This can be explained by the interrelationship of cumulative and
combinatorial knowledge dynamics (Strambach 2013; Strambach and
Klement 2012), which allows for heterogeneous knowledge to be combined in
a context where there is accumulation (similarity) in another dimension.
Proximity in the geographical dimension indicates a mutual understanding of
norms and values associated with a community (Farole et al. 2010) and also a
shared appreciation of, and involvement in, the localized capabilities (Maskell
and Malmberg 1999a). This finding demonstrates nicely how local labour
market flows reproduce and develop the localized capabilities by allowing for
combinations of unrelated pieces of knowledge through the accumulation of
place-specific values and norms
Proximity in the organizational dimension, which here is estimated as
intra-organizational proximity, is assumed to generate a coherent knowledge
view in the labour force. Moreover, intra-organizational proximity ensures a
shared understanding of the structures and procedures within the firm that
reduces uncertainty about the role of others (Weick and Roberts 1993). The
interdependence of the different proximity dimensions is found in all three
papers1 (Paper I, II and III). These findings suggest a twofold relational and
social aspect of knowledge interactions. Interactive learning occurs between
individuals who possess knowledge and work in specific socio-technical
contexts. Their knowledge can be quite similar, but it can also be completely
unrelated. The degree or relatedness determines how well knowledge is
understood and appreciated in knowledge interactions. This relational
distance determines the efficiency of communication and learning in groups,
and varies between any two individuals. This is the first relational aspect of
knowledge interactions. Concurrently, employees at a plant are also connected
through dimensions other than the cognitive2 dimension. Depending on the
1 Paper III examines different types of cognitive proximity, rather than different proximity dimensions.
But the same pattern is detected, where the influence of one knowledge type on plant performance depends
on what other type of knowledge is accounted for in the model.
2 E.g., the organizational, the geographical/institutional or the social dimension. The social dimension is
not examined in the thesis.
38
distance in these other dimensions, learning can be either facilitated or
complicated. This is the other aspect of the relational character of the
proximity dimensions.
The second finding demonstrates that the proximity dimensions have
conditional effects on learning and innovation in firms. The empirical findings
of the interdependence of the proximity dimensions in Paper I and Paper II
are further established and elaborated upon in Paper III, where it is
established that the impact of a given distance in one dimension varies with
the distance in another dimension. Whereas such relations have been
theoretically discussed in the literature (e.g., Boschma 2005), they have not
been demonstrated empirically before. The conditional effects are called
overlap effects. These overlap effects can either enhance the initial impact of
either variable – a positive overlap effect – or they can mitigate the initial
impact – a negative overlap effect. The empirical findings show that the
impact of high levels of in-house similarity in formal knowledge is conditional
on the level of in-house similarity in past knowledge exposures. High levels of
similarity in knowledge exposure substantially reduce the negative impact of
similarity in formal knowledge to the extent that high levels in both variables
positively impact plant performance. These findings could help explain the
slightly diverging results of previous studies on the effects of knowledge
variety on plant and firm performance.
Third, it is not only the proximity dimensions that are interrelated, but the
impact of the cognitive dimension on interactive learning is contingent on the
human capital ratio at the plant. The findings in Paper III demonstrate that it
is the combination of knowledge variety and human capital ratio at the plant
that determines the competitiveness of the plant. Previous studies have
argued that it is not human capital per se that influences the performance of
plants, but rather the degree of relatedness in the in-house labour force
(Boschma et al. 2009). However, the present findings suggest that the impact
of knowledge variety on plant performance is conditional on the level of
human capital at the plant. Other empirical studies have refrained from
estimating the impact of knowledge variety and human capital ratio in the
same model. But in Paper III the parallel impacts of knowledge variety and
human capital ratio are estimated together with their conditional impacts on
plant performance. It is argued that human capital represents ability, whereas
knowledge variety symbolizes specialized expert knowledge.
It is theoretically suggested that high levels of human capital represent high
generic abilities to communicate, learn and synthesize. It also signals high
individual abilities (Spence 1973) that are guaranteed through the successful
completion of higher education resulting in a diploma (Arrow 1973).
According to Becker (1964), the high abilities of highly educated individuals
render them more efficient in their execution of different tasks compared to
those without higher education. The degree of knowledge variety within a
39
plant, on the other hand, informs us of the potential for combining different
pieces of specialized knowledge into new knowledge. Paper III shows that the
impact of similar, related and unrelated formal knowledge (educational
background of the employees) is conditional on the level of human capital. The
negative impact on plant performance associated with high levels of similar or
unrelated knowledge is mitigated by high levels of human capital. This
suggests that interactive learning processes cannot be properly understood
and estimated if attention is not paid to both the generic abilities of the labour
force and the combined variety of their specialized knowledge. Furthermore,
as discussed above human capital represents an ability, presumably an innate
ability, whereas formal knowledge expresses the individual’s area of expertise.
Productive interactive learning processes within plants are influenced by both
the abilities of the labour force and the knowledge variety present in the plant.
The fourth finding shows that there are differences between sectors that
generate variable outcomes of learning processes. This implies that the
workings of the proximity dynamics also vary between sectors, not only
between different levels of analysis. In Paper I the role of knowledge variety
in interactive learning in capital-intensive sectors and labour-intensive
sectors, is examined. In line with previous empirical studies (Haltiwanger et
al. 1999; Leiponen and Drejer 2007; Rigby and Essletzbichler 1997), the
results illustrate a strong dependence on localized capabilities in the capitalintensive sectors, where only local labour inflows positively influence plant
performance. The regional variations in production technology reinforce and
develop the localized capabilities, by allowing local inflows of unrelated
knowledge to be absorbed into the plant. Furthermore, it is established that
the overall impact of both in-house knowledge variety and inflowing
knowledge variety is smaller in the capital-intensive sectors than in the
labour-intensive sectors. The implications are that there are structural
differences between capital-intensive and labour-intensive sectors. This may
be explained, at least partially, by the dependence on physical capital
(machinery), which is produced externally, in the capital-intensive sectors.
This dependence reduces the number of innovative opportunities, because
process innovations are mainly the result of supplier innovations (Lagerholm
2007). Moreover, the physical capital used in the production of goods may
restrict the potential to apply new knowledge, generated from interactive
learning processes. The possibility to apply knowledge is presumed to
completely depend on firm routines and the technologies used within the firm.
This is a step that is often disregarded in economic geography, where
interactive learning processes, in and between firms, are treated as generating
equally applicable knowledge regardless of the sectoral and organizational
context. Both the technology used in an industry and the firm routines within
a specific firm can be viewed as structures that either hinder or promote
40
effective absorption of inflowing knowledge and the possibility to apply new
knowledge.
Lastly, the findings suggest that cognitive relatedness is not necessary for
interactive learning to occur on the micro-level of the plant. Paper I
demonstrated that the combination of either geographical proximity and
cognitive distance or geographical distance and cognitive proximity generated
significantly stronger positive impacts on plant performance than cognitive
relatedness did. In Paper II organizational and geographical proximity
allowed for unrelated knowledge to be efficiently absorbed and applied in the
plant. Within plants, interactions take place between individuals on a daily
basis. Under such circumstances, it is more important to share a short
distance in one dimension, than it is to be cognitively related. This implies that
trust and social structure are more important to learning than is sharing
formal pieces of knowledge. This is in contrast to the agglomerations literature
and the RIS literature. With regard to externalities and knowledge spill overs
between co-located firms, Neffke et al. (2012) demonstrated that a related
industry setup in the region substantially increases the survival rate of plants.
Similarly, Boschma and Iammarino (2009) show that technological
relatedness in the regional industry base positively impacts export rates. The
presumed reasons for the positive effects of a related local industry setup or
RIS are explained by supporting structures and knowledge accessibility
(Trippl 2011).
In cases of mergers and acquisitions or temporary projects between firms,
relatedness is a necessary condition for the joint ventures to be initiated and
for innovation to occur. The temporal character of such case and the potential
lack of any other proximity necessitate cognitive relatedness (Boschma 2005).
This is in contrast to the inter-individual level and labour mobility. In such
scenarios, one or a few individuals enter a context in which they are to stay for
an indefinite period of time. They need to be “accepted” by the receiving group
– the in-house labour force – and any type of proximity relation will speed up,
and ease, the integration of the inflow. Once accepted, the individual will be
able to communicate more distant pieces of knowledge, with regard to the
existing in-house composition. Thus, on the interpersonal level all proximity
dimensions seem to have mainly a socializing function, including the cognitive
dimension, whereas for cooperation between firms, relatedness in the
cognitive dimension is probably the reason why the joint project came to be in
the first place, and the other proximity dimensions can hinder or promote the
learning and innovation processes. The present results indicate that the
workings and effects of proximity dynamics are different at different levels of
analysis.
41
5.2.
Discussion
The above findings show that in order to explain the uneven spatial
development of economic activities it is crucial to understand how knowledge
is produced and reproduced in firms. Since learning and innovation occur in
plants and firms rather than regions we need to understand the interplay
between internal processes and the external environment. In the present
thesis it has been demonstrated that sector belonging, firm routines, and the
knowledge resources in the firm interact to generate unique conditions for
interactive learning and innovation. Firms and regions reproduce increasingly
specialized knowledge in order to sustain a competitive advantage (Howells
2012), while utilizing the accumulated knowledge resources in the firm and
region. A micro-perspective on knowledge offers significant empirical
evidence that distributed knowledge production is an essential element in the
learning economy. This means that knowledge is becoming increasingly
spatially fragmented, due to the local or regional reproduction of knowledge.
One major reason for this is the labour force and its relative fixity in space that
makes it a mainly local resource. Labour force knowledge is fostered in a sociotechnical context in which formal knowledge and skills are integrated with
social and institutional structures that affect the view on knowledge and
knowledge application (e.g., Maskell and Malmberg 1999a; Morgan 2004;
Storper and Venables 2004; Strambach and Klement 2012).
The findings further suggest that for learning to occur on the interpersonal
level, it is more important to share a bond than to be cognitively related. The
interrelationship between proximity and distance that has been suggested
here is in contrast to much of the existing empirical literature on proximity
dynamics (e.g., Boschma et al. 2009). It is found that relatedness in the
cognitive dimension is not unambiguously positive for interactive learning
and innovativeness. Having similarity in one dimension and unrelatedness in
the cognitive dimension has a significantly stronger impact on interactive
learning than simply having relatedness in the cognitive dimension. It
therefore seems as if the combined distance of several proximity dimensions
should be taken into account when estimating the innovative power of a firm
or industry. The empirically demonstrated interdependence of different types
of proximities may help to explain the somewhat diverging empirical results
of the relationship between firm performance and knowledge variety in the
labour force. The findings also offer support for the notion of a relational
economy (Bathelt and Glückler 2003; Torre and Gilly 2000) in which
economic actors are unequally able to combine their knowledge and skills with
the knowledge and skills of other actors, depending on their level of
association.
The micro-foundation of innovation and economic performance is
knowledge transferred between individuals. The findings indicate that
innovation emerges as a result of cumulative and combinatorial knowledge
dynamics. Whereas Crevoisier and Jeannerat (2009) argue that the
cumulative aspect of knowledge in firms and regions is losing importance, it
42
is argued here, and in line with Strambach (2013), that the cumulative and
combinatorial knowledge dynamics interact in the generation of novelty. It is
the cumulative aspect of knowledge that has driven the theoretically
motivated need for cognitive relatedness in all empirical work on learning so
far. There have been two parallel logics to explain the need for cognitive
relatedness in inter-firm and interpersonal learning. First, on the inter-firm
level, firms are unlikely to diversify into activities that are completely
unrelated to the existing competences (Harrison et al. 2001). The efficiency of
such mergers, acquisitions, alliances and projects is too low to be economically
justified (Rumelt 1974). Second, knowledge in a firm has accumulated over
time and any additional contribution to the existing stock needs to be
cognitively related in order to be recognized as relevant (Cohen and Levinthal
1990; Nooteboom 2009). But, instead of just focussing on cognitive
relatedness as a bridge between accumulated knowledge and new knowledge,
the present thesis has demonstrated that similarity in the organizational or
the geographical/institutional dimension allows unrelated knowledge to be
integrated into production processes. It has been shown, for instance, that
organizationally proximate individuals successfully generate novelty by
combining their unrelated formal knowledge (Paper II). Moreover, unrelated
formal knowledge originating from the same institutional context (i.e., being
intra-regional) contributed to the upgrading of skills in the capital-intensive
sectors in Paper I. Thus, it is argued that the accumulation of social capital
and firm-specific knowledge occurs on the level of the individual. This
accumulation represents a certain degree of familiarity with firm routines and
norms and values in a region that are shared with other individuals in the
same context. This similarity represents a short relational distance between
individuals that generates cohesion and reduces uncertainty on the
interpersonal level. If there is similarity in any one of the dimensions,
unrelated knowledge can be integrated into firm practices.
Combining unrelated pieces of knowledge is only possible between
individuals who share some kind of common ground that enables trust and
reduces uncertainty. In line with recent contributions in economic geography
(Boschma 2005; Strambach and Klement 2012) and the reasoning underlying
the ZPD (Vygotsky 1962), it is argued here that proximity and distance
represent accumulation and combination and that interactive learning in the
labour force is facilitated by close proximity in any one dimension.
Furthermore, the findings in this thesis support the view that micro-agents
(i.e., individuals in the present thesis) are embedded in (Granovetter 1985)
and inseparable from the structures that surround them. Physical and
cognitive structures influence and are influenced by the actions of microagents. Whereas structures (institutions, technology and firm routines) offer
possibilities and impose restraints on agents, they are also interpreted every
time an agent attempts to act accordingly. As time passes, the structures are
incrementally changed and adapted by the individuals that live and work
under them. The local environment generates relational proximity between
agents through formal and informal networks (Storper 1997). This proximity
43
reproduces and rejuvenates localized capabilities by enabling the combination
of heterogeneous pieces of knowledge in firms through local unrelated labour
inflow. Thus, time and place can be perceived as the paramount dimensions
that shape the micro-dynamics of knowledge generation and innovation.
The notion presented by Feldman and Kogler (2010), that places are not
equal and are defined by evolutionary processes, is empirically corroborated
in the present thesis. An important element in the evolution of economic
activities is the accumulated knowledge embodied in the labour force in a
region (Maskell and Malmberg 1999a). The impact that labour force
knowledge has on plant performance is substantial, multifaceted,
differentiated, and demonstrated through multiple relational measures of
knowledge and ability throughout the empirical papers. The findings in the
thesis support the claim by Storper and Scott (2009) that the productivity of
resources and skills depend on selective geographical matching processes. It
has been shown that the proximity dimensions are interdependent and it has
also been demonstrated that physical and cognitive structures are
interconnected with the proximity dimensions. These interdependencies lead
to path-dependent trajectories for firms and regions. These trajectories will
offer different absorptive capacities and innovative behaviours throughout the
economic landscape (Storper and Scott 2009).
The contextual and relational aspects of learning has been demonstrated in
the papers and emphasised throughout the thesis. The findings of the
relational character of knowledge offer significant support for the notion
presented by Bathelt and Glückler (2003) that economic actors are situated in
contexts of social and institutional relations. According to the knowledge and
competence-based theory of the firm (Kogut and Zander 1992; Teece et al.
1997), knowledge accumulation in firms is mainly driven by firm routines and
organizational practices. These are based on localized learning processes that
have emerged over time in specific institutional and technological contexts, of
which the labour force is a vital part. The specificity of these localized learning
processes helps explain why the behaviour and knowledge of successful firms
cannot easily be replicated or transferred to other contexts (Szulanski 1996;
Teece 2010). The existence of localized capabilities is also demonstrated
through the variable impact of knowledge inflow depending on the
geographical (institutional) origin of the labour inflow. It would seem that
geography is not dead (Cairncross 1997) and space is not neutral with regard
to the development and competitiveness of economic activities.
44
5.3. Directions for future research
The present findings suggest that the conditions for learning and
knowledge application vary throughout the economic landscape. Therefore,
interactive learning processes and innovation should be examined within the
context of the firm and its technological trajectory and institutional
environment. Future studies on interactive learning and knowledge dynamics
should therefore consider the structures that surround the labour force and
the knowledge variety that the labour force constitutes. A further division of
sectors based on the structures that their technological trajectories dictate
could advance our understanding of how knowledge generation is impeded
and promoted in different contexts. The reciprocal relationship between
structures and agents suggests that the role of knowledge variety in firm
competitiveness should be regarded as one part of a greater system. Thus,
future research aimed at understanding local conditions for competitiveness
and growth could benefit from the integration of meso- and micro-level
analyses.
Another interesting direction for future research is to further detangle the
interdependencies and conditional influences of the proximity dimensions. In
Paper III it was shown that different types of cognitive proximity have
conditional impacts on plant performance. These findings indicate that
conditional effects may also be found between other types of proximity. This
would require the development of continuously integrated frameworks where
multiple types of proximity are analysed simultaneously. Furthermore, the
trend away from appreciating “just” high levels of human capital may be
premature due to the conditional effects of human capital and knowledge
variety. Thus, future studies on the influence of labour force knowledge on
competitiveness should integrate ability (human capital) and knowledge into
the analysis. This effect is also likely to vary between different sectors.
Moreover, the present findings indicate that the role of different proximity
dimensions may vary at different levels of analysis. For example, cognitive
relatedness is not found to be necessary for knowledge generation on the
interpersonal level of intra-firm or intra-plant learning. On the other hand, in
discussions of inter-firm relations, Boschma (2005) and others argue that
cognitive relatedness is a necessary but not sufficient requirement. Further
analytical clarity about the workings of the proximity dynamics on different
levels of analysis is therefore needed. In relation to that issue, there is a need
to continue to develop empirical tools for estimating proximity in the different
dimensions on the different analytical levels.
Our understanding of the external mechanisms that impede and/or
facilitate learning (i.e., the firm and its routines, the sector and the region)
should be further looked into, preferably in an increasingly integrated
framework that considers multiple factors that affect learning and the
innovative capabilities of firms or regions. The evolutionary framework offers
45
great opportunities for integrating various levels of study as well as different
mechanisms that influence the potential for learning and innovation.
Regardless of level of analysis, the passage of time leaves the observed entity
or place with events and experiences that impact the ability to adapt to new
circumstances in the future. Therefore, differences in the economic
performance of firms, sectors and regions should be perceived as varieties of
accumulated paths on several interconnected levels. Through more extensive
research on each of the above-mentioned mechanisms thought to obstruct
and/or promote knowledge transfer, we will eventually come closer to a
complete synthesis of how and why differences in innovative capabilities exist.
6. Sammanfattning (Summary in
Swedish)
I den här avhandlingen undersöks förutsättningar för lärande och
tillämpning av kunskap (innovation) i företag. I ekonomisk geografi och
besläktade ämnen har kunskap kommit att betraktas som den viktigaste
faktorn för företags och regioners konkurrenskraft. Det beror på att kunskap
är lokalt förankrad till skillnad från materiella produktionsfaktorer som kan
tillgängliggöras genom transporter över hela världen. Genom
socialiseringsprocesser kopplade till såväl platsers bransch- och
företagsstruktur som deras historia tolkas, används och reproduceras
kunskap på olika sätt på olika platser. Det handlar inte huvudsakligen om att
varor och tjänster produceras på olika platser utan snarare om att likartade
produkter produceras på olika sätt på olika platser. Den lokala kunskapen styr
hur användningen av andra resurser (produktionsfaktorer) utformas och det,
i sin tur, påverkar framtida förutsättningar för förändring och nytänkande i
såväl enskilda företag som i regioner.
Innovationskraften - förmågan att förnya produkter, processer, styrning
och marknadsföring m.m. - anses avgörande för bibehållen eller utökad
produktivitet för företag och regioner. Således blir den tillgängliga kunskapen
och förmågan att förnya den av yttersta vikt. Individer representerar olika
kunskap, vars kvalitet och kvantitet bedöms och avgörs av sammanhanget de
befinner sig i, dvs. av de andra deltagande individerna. Att skaffa sig en
konkurrensfördel genom arbetskraften handlar därför inte om att ackumulera
expertkunskap inom ett område utan snarare om att sammanföra individer
med varierad kunskap som kan kommunicera med, och lära av, varandra. På
så vis skapas en dynamisk miljö med förutsättningar för fortsatt lärande och
förnyelse som i förlängningen kan ge stärkt konkurrenskraft. Individers och
gruppers lärande är förvisso en nödvändighet, men är i sig självt otillräckligt
för att initiera en innovationskedja. För att kunskapen ska kunna omsättas i
praktik krävs en närvaro av fysiska och kognitiva strukturer som tillåter detta.
Dessa strukturer varierar mellan såväl branscher som företag och påverkar i
46
hög grad möjligheten för enskilda individer att tillämpa ny kunskap i
produktionen. Det är således inte enbart kunskapssammansättningen som är
avgörande för företagens produktivitet utan även regional bransch- och
företagsspecifika strukturer som påverkar möjligheten att omsätta
kunskapen.
Utgångspunkten i den här avhandlingen är arbetskraftens kunskap och den
variation i kunskap som finns på arbetsplatser och i deras närmiljö.
Arbetskraftens individuella och samlade förmågor utgör grunden för lärande.
Således är även de potentiella tillämpningarna (omsättningen av teori i
ekonomiskt betydelsefull praktik) av ny kunskap i företag ytterst beroende av
arbetskraftens kunskap och förmågor. Men för att implementera kunskap och
ge den ett ekonomiskt värde räcker det inte med att kunskapen existerar, utan
det måste även i företaget och branschen finnas inkluderande och flexibla
strukturer som tillåter enskildas eller gruppers teorier att bli praxis. Mycket
av den befintliga forskningen fokuserar antingen på kunskapens betydelse för
företagens produktivitet eller på de övergripande strukturernas hindrande
alternativt befrämjande egenskaper för innovationskraften. Få studier
integrerar de två perspektiven, vilka kan sägas vara olika analysnivåer av
samma fenomen, innovationskraft och konkurrenskraft. Att försöka fånga och
kartlägga interaktionen, mellan kunskapssammansättning och strukturer är
denna avhandlings huvudsakliga syfte.
Syftet uppnås genom tre artiklar som använder sig av kvantitativa
analysmetoder för att belysa olika aspekter av arbetskraftens kunskap i
relation till företagens produktivitet och konkurrenskraft. Samtliga studier
utgår
från
formella
mått
på
arbetskraftens
kunskap
(t.ex.
utbildningsinriktning eller yrkestitel) och den variation av kunskap som finns
på arbetsplatsen. Dessa mått relateras sedan till andra strukturer som varierar
mellan arbetsplats, så som fysiska strukturer kopplade till sektoriella
skillnader, kognitiva strukturer i form av företags skilda rutiner och
professionella nätverks effekt på arbetskraftens faktiska förmågor.
Analyserna bygger på detaljerad, longitudinell socioekonomisk data ifrån
statistikdatabasen ASTRID som finns på institutionen för geografi och
ekonomisk historia vid Umeå universitet. ASTRID innehåller data från
Statistiska Centralbyrån och information finns på årsbasis om samtliga
företag, arbetsplatser och individer i Sverige.
Syftet med den första artikeln, Paper I, är att testa huruvida det finns
systematiska skillnader i möjligheten att omsätta kunskap till praktik mellan
två stora sektorer i ekonomin, nämligen den kapitalintensiva- och den
arbetskraftsintensiva sektorn. Genom att analysera effekten av såväl
sammansättningen av kunskap på arbetsplatsen som inflödet av ny
arbetskraft till arbetsstället, studeras både kunskapssammansättningens
betydelse för arbetsplatsens produktivitet och de skilda möjligheterna att
integrera inflödande kunskap med befintlig kunskap inom sektorerna.
47
Artikeln bygger dels på litteratur som diskuterar hur produktionsmetoder och
produktionsfaktorer
genererar
strukturella
skillnader
i
innovationsmöjligheter mellan branscher, och dels på litteratur om
arbetskraftsrörlighet och värdet av relaterad kunskap för företags
produktivitet. Resultaten i den första artikeln visar att det är gynnsamt för alla
arbetsplatsarbetsplatser, oavsett sektor, att ha anställda vars formella
kunskaper är besläktade (related) men inte identiska (similar). Den positiva
effekten är dock större i arbetskraftsintensiva sektorer än i kapitalintensiva
sektorerna. Den svaga positiva effekten av inflödande besläktad kunskap i de
kapitalintensiva sektorerna antas bero på att kunskap i stor utsträckning
ackumuleras i det fysiska kapitalet, vilket är kostsamt att förändra och
utveckla. I de arbetskraftsintensiva sektorerna lagras majoriteten av
kunskapen i arbetskraften och utbyte och utveckling av idéer och kunskap kan
ske genom den dagliga interaktionen.
Kunskapsinflödenas effekt på arbetsplatsens produktivitet skiljer sig åt
mellan sektorerna vad gäller såväl kunskapens sammansättning (identisk,
besläktad eller orelaterad) som geografiska ursprung (lokalt eller icke-lokalt).
Effekten av arbetskraftsinflöden, tillika kunskapsinflöden, är genomgående
större för den arbetskraftsintensiva sektorn än för den kapitalintensiva. Det
förklaras med ett beroende av fysiskt kapital (maskiner och utrustning) i de
kapitalintensiva sektorerna. Det innebär att potentiella effektiviseringar och
processinnovationer delvis kommer till stånd hos underleverantörer och
tillverkare. Således är antalet innovationsmöjligheter färre i de
kapitalintensiva sektorerna än i de arbetskraftsintensiva sektorerna där såväl
produkt- som processinnovationer sker inom företaget.
Inom de
kapitalintensiva sektorerna är det endast de lokala inflödena som har en
positiv effekt på arbetsplatsens produktivitet. Det är i enlighet med tidigare
forskning som funnit att produktionstekniken inom en och samma bransch i
tillverkningsindustrin varierar kraftigt mellan olika platser. Det visar på en
socialiseringsprocess av kunskapstillämpning som innebär att i de
kapitalintensiva sektorerna är kunskap om hur saker görs och kommuniceras
på en plats av större betydelse än den formella kunskapen en individ besitter
och kan bidra med i produktionen. I de arbetskraftsintensiva sektorerna finns
det kombinerade effekter av inflödets kunskapssammansättning och
geografiska ursprung som gör att såväl lokal orelaterad kunskap som ickelokal identisk kunskap har en positiv effekt på produktiviteten. Resultaten för
de arbetskraftsintensiva branscherna visar att det finns andra aspekter av
arbetskraftens kunskap, än den formellt kognitiva förmågan, som påverkar
förmågan att kommunicera och implementera kunskap på arbetsplatsen. I
branscher där människor, tillika sociala kunskapsbehållare, utgör den
huvudsakliga produktionsfaktorn förekommer en större känslighet för
kunskapens geografiska ursprung. Det lokala- respektive icke-lokala inflödet
av arbetskraft representerar institutionellt lika och olika inflöden. När hänsyn
48
tas till såväl formell kunskap som det geografiska (institutionella) ursprunget
framträder en bild där kombinationen av likhet och olikhet i dessa
dimensioner genererar ett sammantaget intellektuellt avstånd som gynnar
lärande och innovationsförmågan på arbetsplatsen. De samlade resultaten i
den första artikeln indikerar att förutsättningarna för absorption och
tillämpning av kunskap skiljer sig åt mellan de två stora sektorsgrupperna.
Den andra artikeln, Paper II, handlar om hur företag socialiserar och
strukturerar kunskap genom sina rutiner och på det viset skapar en praxis som
är unik för företaget. Denna praxis skapar en organisatorisk närhet och
samhörighet mellan de anställda som tillåter effektivare kommunikation
mellan individer inom företaget än mellan individer på olika företag.
Rutinerna skapar således olika förutsättningar för de anställda att tillvarata
intern och extern kunskap. Syftet med artikeln är att jämföra effekten på
produktiviteten av interna kunskapsflöden (individer som byter arbetsplats
inom ett och samma företag) med externa kunskapsflöden (individer som
byter arbetsplats och företag). Interna och externa flöden ställs mot varandra,
samtidigt
som
hänsyn
tas
till
arbetskraftens
formella
kunskapssammansättning och dess geografiska ursprung. Liksom företagen
strukturerar och socialiserar kunskap gör även platser (regioner) det genom
den ackumulation av tekniskt kunnande och sociala nätverk som byggts upp i
regionen över tid. Denna samsyn kring teknisk utveckling och kunskap som
finns i regioner, sägs utgöra en del av regionens institutionella grund. Denna
formar såväl branschstrukturen i regionen som tillämpningen av kunskap
inom en specifik bransch och representerar i denna avhandling normer och
värderingar. Dessa är huvudsakligen grundade i regionens historiska
utveckling av branscher och företag. Bilden av vad som är relevant kunskap
förstärks alternativt försvagas av samarbeten mellan branscher eller mellan
företag som historiskt sett varit starka och framgångsrika. Genom att
inkludera den geografiska dimensionen i analysen framträder en bild av hur
den formella kunskapssammansättningen samspelar med såväl rumsligt som
organisatoriskt ursprung för att skapa förutsättningar för lärande och
innovationskraft.
Resultaten i denna artikel visar att inflöden av kunskap som liknar den som
finns på arbetsplatsen endast ökar produktiviteten om den samtidigt
härstammar från att annat företag med annan praxis. Den formella
kunskapen har då tillämpats inom en annan fysisk och social struktur och
eventuellt även på ett annat sätt. Den är således differentierad från den
befintliga kunskapen på den nya arbetsplatsen i ett organisatoriskt
hänseende. Detta ger den likartade kunskapen ett ytterligare djup och bredd,
vilket innebär potentiellt fler nya kombinationer och tillämpningar av
kunskap. Den positiva effekten på produktiviteten är som snabbast och mest
omfattande när det likartade kunskapsflödet härstammar från en annan
region. Det tolkas som att ett större avstånd i såväl den geografiska- som den
49
organisatoriska dimensionen skapar en stor bredd av möjliga nya
tillämpningar av den relaterade formella kunskapen. Resultaten visar också
att bekantskap med praxis (intra-organisatoriska flöden) skapar en
sammanhållen kunskapssyn. Även om individens formella kunskap är
likartad snarare än identisk med den befintliga kunskapen på den nya
arbetsplatsen så sker inget nyskapande. Det krävs att den inflödande formella
kunskapen är olik den befintliga samt att den är lokal (intraregional), och
således har samma socio-tekniska ursprung, för att ny kunskap och nya
tillämpningar ska genereras. Detta visar på en starkt konvergerade effekt av
platsspecifika värderingar på kunskap. Det indikerar även att intraorganisatorisk närhet skapar förtroende mellan de deltagande individerna på
ett sådant sätt att mer avlägsna idéer kan mötas och diskuteras och potentiellt
implementeras. I Paper II framgår det att avstånd kan mätas på fler sätt än
som fysiskt (geografiskt) avstånd och att dessa icke-rumsliga relationella
avstånd påverkar lärande inom företag. Genom närhet i någon eller fler av
dessa dimensioner kan förtroende skapas ur vilket ny kunskap och nya
tillämpningar kan växa fram. Likartad eller identisk formell kunskap är inte
alltid nödvändig för att lärande och innovationskraft ska uppstå. Resultaten
visar att lärande och innovationsprocesser är sociala processer och som
sådana grundas de på tillit och trovärdighet, något som kan uppnås genom
närhet i den organisatoriska och den geografiska dimensionen. Huruvida det
finns någon hierarkisk struktur mellan olika former av närhet, för lärande, går
inte att säga utifrån resultaten i denna artikel. Men, det har framkommit att
korta avstånd i en socialiserande och strukturerande dimension kan
kompensera för stora avstånd i den formella kunskapsdimensionen.
I den tredje artikeln, Paper III, testas effekten av arbetskraftens kunskap i
företag inom kunskapsintensiv företagsservice (KIBS). Medan de två första
artiklarna fokuserar på effekten av inflöden av arbetskraft på arbetsplatserna
så fokuserar denna artikel på den befintliga arbetskraften på en arbetsplats.
Till kunskapsintensiv företagsservice räknas bl.a. juristfirmor, arkitektkontor,
olika IT-tjänster och forskning m.m. Dessa branscher genomsyras av lärande
och föränderlighet då de huvudsakligen säljer expertkunskap och arbetar på
uppdragsbasis mot kunder vars behov varierar. Kunskapens och lärandets roll
förväntas således vara betydande för företagens konkurrenskraft. Dessa
branscher lämpar sig därför extra bra för studier av kunskapsstrukturer,
lärande och innovationskraft. Utöver att undersöka vilken effekt den formella
kunskapen och dess sammansättning har på innovationskraften i dessa
företag så presenteras och testas en teori om att kunskap hos individer kan
mätas på fler sätt än som formell utbildning. Genom sina
arbetslivserfarenheter kommer människor i kontakt med såväl olika
branscher
som
arbetsplatser
vars
medarbetare
har
olika
utbildningsbakgrund. Dessa möten och erfarenheter formar individens
kunskap och förståelse av hur kunskap kan användas. Både branscherfarenhet
50
och tidigare kunskapsexponering betraktas som egna kunskapsvariabler.
Effekten på lärande och innovationskraft på arbetsplatsen beräknas för de
olika typerna av kunskap dels var för sig, dels i olika kombinationer. Detta för
att testa huruvida förekomsten av en viss typ av kunskap påverkar effekten
som en annan kunskap har på lärandet. Grundantagandet i artikeln är att
kunskap kan ta sig olika uttryck och således även bör mätas på olika sätt. Det
anses även rimligt att olika kunskapsstrukturer påverkar varandra. Detta kan
ske på ett sådant sätt att den initiala effekten av en form av kunskap på
företagets produktivitet förändras i kombination med en annan kunskapstyp.
Tidigare branscherfarenhet representerar tillämpad kunskap. Därför är det
rimligt att liknande (relaterade) branscherfarenheter hos medarbetarna på en
arbetsplats skapar förtrogenhetförtrogenhet med multipla tillämpningar av
en viss typ av formell kunskap. Likaså vittnar tidigare kunskapsexponering
om förmågan att absorbera och kommunicera kunskap. Att medarbetarna har
liknandeliknande erfarenheter av kunskapsexponering skapar förtroende för
kompetensen genom igenkänning.
Avsikten med artikeln är dels att nyansera måttet på kunskap, och dels att
testa hur kunskapssammansättningen på en arbetsplats påverkar dess
produktivitet. I teoretiska diskussioner lyfts kunskap fram som relationell och
socialt formad. Detta till trots mäts den oftast som utbildningsbakgrund i
studier av kunskapens effekter på företagens produktivitet. Artikeln
undersöker hypotesen att individer kan skaffa sig likartad kunskap genom
arbetslivserfarenheter likväl som genom utbildningsbakgrund. Resultaten
visar att företags produktivitet påverkas positivt av att ha en hög andel
anställda med högre utbildning.
Med tanke på att företag inom
kunskapsintensiv företagsservice huvudsakligen säljer och producerar
kunskap är detta inte särskilt förvånande. Kunskap anses driva såväl teknisk
utveckling som ekonomiskt tillväxt i företag och regioner i alla branscher.
Andelen högutbildade representerar det humankapital som ackumulerats på
en plats, i detta fall en arbetsplats. Men när humankapitalnivån på
arbetsplatsen ställs mot olika sammansättningar av kunskap träder en mer
nyanserad bild fram. Det framgår att effekten av humankapital och de olika
formerna av kunskapssammansättning påverkar varandravarandra, dvs. de är
villkorade av varandra. Detta är något som tidigare forskning fört fram som
troligt men det har hittills inte testats. Resultaten visar att den negativa effekt
som höga nivåer av identisk och olik formell kunskap (avseende
utbildningsinriktning bland de anställda) har på lärande avtar när
humankapitalnivån ökar. Höga nivåer av såväl identisk formell kunskap som
humankapitalhumankapital genererar, till och med, produktivitetsfrämjande
lärprocesser inom KIBS. I tidigare studier av arbetskraftens kunskap och dess
effekt på företags innovationskraft har kunskapstyperna testats var för sig.
Identisk kunskap har då alltid varit skadlig för innovationskraften. Genom att
i denna artikel testa fler typer av kunskap samtidigt framgår det att effekten
51
av identisk kunskap är villkorad av utbildningsnivån på arbetsplatsen. Detta
är något som tidigare forskning inte lyckats visa.
När resultaten av de tre artiklarna beaktas i sin helhet, växer det fram en
bild av att lärande i företag är socialt snarare än kognitivt betingat. Lärande
drivs och styrs av såväl fysiska och kognitiva strukturer kopplade till specifika
branscher, som av förmågan hos de individer som deltar i aktiviteten. I denna
avhandling fokuseras på individernas förutsättningar till kollektivt lärande
och innovationskraft på arbetsplatsen genom analys av rumsliga och ickerumsliga avstånd mellan individerna. Avhandlingen bidrar på tre sätt till att
utveckla befintliga teorier om kunskap som resurs och effekter av rumslig och
icke-rumslig närhet på lärande. Det första bidraget är att avhandlingens
samtliga studier visar att de olika formerna av närhet samspelar för att skapa
förutsättningar för lärande och innovationskraft. Inget universellt fungerande
optimalt avstånd har framträtt i någon dimension utan studierna har snarare
visat att lärande på arbetsplatser är starkt kontextberoende. Ett kort avstånd
(närhet) i en dimension kan kompensera för ett långt avstånd i en annan
dimension. Effekten av avstånden skiljer sig åt mellan sektorsgrupper, vilket
påvisats i Paper I, liksom mellan inflöden av arbetskraft med olika kännedom
om, och förtrolighet med, företagets rutiner (Paper II). En fördjupning och ett
ytterligare bidrag, nås i Paper III där villkorade effekter (interaktionseffekter)
påvisas mellan olika former av kunskap. Dessa effekter benämns
överlappande lärandeeffekter (overlap effects). Detta betyder att den
påverkan som ett visst avstånd i en variabel har på lärande på arbetsplatsen
förändras med olika värden i en annan kunskapsvariabel. I artikeln används
fyra olika mått på kunskap varav tre är relationella avståndsmått, dvs.
varianter av kognitivt avstånd/närhet. De kunskapsavstånd som är
opåverkade av förekomsten av andra kunskaper och avstånd benämns
oberoende lärandeeffekter. Dessa resultat, de villkorade effekterna, är
sannolikt applicerbara även på övriga närhetsdimensioner (social,
organisatorisk, geografisk och institutionell närhet). Att korta avstånd i en
dimension inte bara kompenserar för långa avstånd i en annan dimension
utan också att effekten av en kunskapsform på lärande varierar med nivån på
andra kunskapsvariabler har såvitt jag vet aldrig testats och påvisats tidigare.
Dock har sannolikheten att sådana samband finns påtalats, liksom behovet av
att empiriskt testa interaktionen mellan de olika dimensionerna.
Det tredje bidraget, som har en mer implicit karaktär bygger på slutsatsen
att likhet i formell kunskap inte är nödvändig för att lärande ska äga rum
mellan individer. Det tycks viktigare, för att lärande ska uppnås, att det finns
närhet mellan individerna i någon dimension. Denna närhet skapar
förtroende mellan individerna, vilket skapar förutsättningar för att koppla
samman olika kunskap. Således är det sammantagna avståndet i flera
dimensioner av större betydelse för lärande på en arbetsplats, än vad det är
att enbart det kognitiva (kunskaps) avståndet är det rätta. Resultaten i
52
avhandlingen visar att innovationskraft kan uppstå på arbetsplatser med höga
nivåer av olik (unrelated) kunskap, dvs. stora avstånd i den kognitiva
dimensionen. Det kan vid en första anblick tyckas kontraintuitivt och
framförallt som att det står i kontrast till den befintliga litteraturen om ickerumsliga avstånd (t.ex. Boschma 2005). Det framhålls i litteraturen att det
teoretiska ramverket med rumsliga och icke-rumsliga avståndsdimensioner
kan användas för analyser på flera olika nivåer. Detta är än så länge främst ett
teoretiskt och generellt antagande. Men med allt fler empiriska studier av
kunskap, lärande och icke-rumsligt avstånd utkristalliseras skillnader i
effekter av närhet mellan individer och mellan företag. På mikronivån, den
nivå som studerats i denna avhandling, bidrar korta avstånd i samtliga
närhetsdimensioner med att skapa förtroende och förtrogenhet mellan
individer. En arbetsplats utgör en specifik kontext som medarbetarna är väl
bekanta med. Ibland kommer det nya medarbetare, vilkas kunskap bör
tillvaratas, införlivas i rådande rutiner och omsättas praktiskt. För att den nya
kunskapen ska kunna absorberas och tillämpas krävs att de befintliga
medarbetarna kan bedöma trovärdigheten hos den nya medarbetaren. Den
eller de nya medarbetarna kommer till en etablerad arbetsplats med
existerande rutiner och kunskapsstrukturer. Absorptionen och assimileringen
av den nya kunskapen underlättas av att nya individerna delar erfarenheter i
någon närhetsdimension med de befintliga medarbetarna. Då inflöden av
arbetskraft till en arbetsplats inte är tillfälliga samarbeten i vare sig tid eller
rum utan ofta permanenta lösningar, så är villkoren som omger lärandet i
dessa situationer annorlunda från dem som gäller vid samarbeten mellan
företag kring en produkt eller en process. De skiljer sig även från de processer
som genererar lärande i form av agglomerationsfördelar. Den kognitiva
dimensionen utgör i dessa situationer en dimension inom vilken ett kort
avstånd kan bidra till att skapa en gemensam nämnare med andra
medarbetare. Finner man en gemensam nämnare i någon dimension ökar
förtroendet för individen och dennes kunskap kan troligtvis absorberas även
om den är olik den befintliga kunskapen på arbetsplatsen.
53
7. References
Abernathy, W., and Utterback, J. (1978) Patterns of innovation in
technology. Technology Review 80:40-47.
Acha, V., and Brusoni, S. (2003) Complexity and industry evolution: New
insights from an old industry, in Best Papers Proceedings of 2002. European
Association for Evolutionary Political Economy Conference, 2003.
Agrawal, A.; Cockburn, I.; and Mchale, J. (2006) Gone but not forgotten:
knowledge flows, labor mobility, and enduring social relationships. Journal
of Economic Geography 6:571–591.
Almeida, P., and Kogut, B. (1999) Localization of knowledge and the
mobility of engineers in regional networks. Management Science 45:905–917.
Almeida, P.; Song, J.; and Grant, R. M. (2002) Are firms superior to
alliances and markets? An empirical test of cross-border knowledge building.
Organization Science 13:147-161.
Amin, A., and Cohendet, P. (2005) Geographies of knowledge formations
in firms. Industry and Innovation 12:465-486.
Anderson, L., and Tegsjö, B. (2006) En rörlig arbetsmarknad – dynamiken
bland job, individer och företag [Labour market mobility – dynamics among
jobs, individuals and business]. In Statistics Sweden: Fokus på närlingsliv och
arbetsmarknad våren 2006 [Focus on economy and labor market, spring
2006]. Statistics Sweden (SCB), Stockholm.
Antonelli, C. (2000) Collective knowledge communication and innovation:
the evidence of technological districts. Regional Studies 34:535-547.
Arrow, K. (1973) Higher education as a filter. Journal of public economics
2:193-216.
Asheim, B., and Coenen, L. (2005) Knowledge bases and regional
innovation systems: Comparing Nordic clusters. Research Policy 34:11731190.
Asheim, B., and Gertler, M. (2005) The geography of innovation: Regional
Innovation Systems, in: J. Fagerberg, D.C. Mowery, and R.R. Nelson (eds):
The Oxford handbook of innovation. Oxford: Oxford University Press, pp.
291-317.
Asheim, B.; Smith, H.; and Oughton, C. (2011) Regional Innovation
Systems: Theory, empirics and policy. Regional Studies 45:875-891.
Barras, R. (1986) Towards a theory of innovation in services. Research
Policy 15:161-173.
Barras, R. (1990) Interactive innovation in financial and business services:
The vanguard of the service revolution. Research Policy 19:215-237.
Bathelt, H., and Glückler, J. (2003) Toward a relational economic
geography. Journal of Economic Geography 3:117-144.
54
Bathelt, H., and Glückler, J. (2005) Resources in economic geography:
from substantive concepts towards a relational perspective. Environment and
Planning A 37:1545-1563.
Bathelt, H., and Glückler, J. (2013) Institutional change in economic
geography. Progress in Human Geography Online First Version of Record Oct 18, 2013:1-24
Bathelt, H.; Malmberg, A.; and Maskell, P. (2004) Clusters and knowledge:
local buzz, global pipelines and the process of knowledge creation. Progress
in Human Geography 28:31-56.
Bathelt, H., and Taylor, M. (2002) Clusters, power and place: inequality
and local growth in time-space. Geografiska Annaler B 84:93-109.
Becker, G. (1962) Investment in Human Capital: A Theoretical Analysis.
Journal of Political Economy 70:9-49.
Becker, G. (1964) Human capital. New York: Columbia University Press.
Becker, M. C. (2004) Organizational routines: A review of the literature.
Induatrial and Corporate Change 13:643-677.
Becker, M. C., and Zirpoli, F. (2008) Applying organizational routines in
analysing the behaviour of organizations. Journal of Economic Behavior and
Organization 66:128-48.
Bell, D. (1973) The coming of post-industrial society. New York: Basic
Books.
Bhaskar, R. (1978) A realist theory of science. Brighton: Harvester Press.
Bienkowska, D. (2007) Arbetskraftens rörlighet och klusterdynamik: En
studie av IT- och telekomklustren i Kista och Mjärdevi [Labour mobility and
cluster dynamics. A study of ICT clusters in Kista and Mjärdevi, Sweden].
PhD-thesis, Geografiska regionstudier 75, Depertment of Social and
Economic Geography, Uppsala University, Uppsala.
Björklund, A.; Edin, P-A.; Holmlund, B.; and Wadensjö, E. (2000)
Arbetsmarknaden, (The labor market), Second edition. Kristianstad:
Kristianstads boktryckeri.
Blanc, H., and Sierra, C. (1999) The internationalization of R&D by
multinationals: a trade-off between external and internal proximity.
Cambridge Journal of Economics 23:187-206.
Boschma, R. (1999) Learning and regional development. GeoJournal
49:339-343.
Boschma, R. A. (2004) Competitiveness of regions from an evolutionary
perspective. Regional Studies 38:1001-1014.
Boschma, R. A. (2005) Proximity and Innovation: A Critical Assessment.
Regional Studies 39:61-74.
Boschma, R.; Eriksson, R.; and Lindgren, U. (2009) How does labour
mobility affect the performance of plants? The importance of relatedness and
geographical proximity. Journal of Economic Geography 9:169-90.
55
Boschma, R.; Eriksson, R.; and Lindgren, U. (2014) Labour market
externalities and regional growth in Sweden: The importance of labour
mobility between skill-related industries. Regional Studies DOI:
10.1080/00343404.2013.867429.
Boschma, R., and Frenken, K. (2006) Why is economic geography not an
evolutionary science? Towards an evolutionary economic geography. Journal
of Economic Geography 6:273-302.
Boschma, R., and Labooy, J.G. (1999) Evolutionary economics and
economic geography. Journal of Evolutionary Economics 9:411-429.
Boschma, R., and Martin, R. (2007) Editorial: Constructing an evolutionary
economic geography. Journal of Economic Geography 7:537-548.
Boschma, R., and Martin, R. (2010) (eds) The Handbook of Evolutionary
Economic Geography. Cheltenham: Edward Elgar.
Boschma, R., and Iammarino, S. (2009) Related variety, trade linkages and
regional growth in Italy. Economic geography 85:289-311.
Brenner, T., and Broekel, T. (2011) Methodological Issues in Measuring
Innovation Performance of Spatial Units. Industry and Innovation 18:7-37.
Breschi, S., and Lissoni, F. (2009) Mobility of skilled workers and coinvention networks: an anatomy of localized knowledge flows. Journal of
Economic Geography 9:439-68.
Bush, V. (1945) Science: The endless frontier. Washington: United States
Government Printing Office.
Butzin, A., and Widmaier, B. (2012) The study of time-space dynamics of
knowledge with innovation biographies. Working Papers on Innovation and
Space 07.12. Philipps University, Marburg.
Cairncross, F. (1997) The death of distance. How the communication
revolution will change our lives. Boston: Harvard Business School Press.
Castellacci, F. (2006) A critical realist enterpretation of evolutionary
growth theorising. Cambridge Journal of Economics 30:861-880.
Castellacci, F. (2008) Technological paradigms, regimes and trajectories:
manufacturing and service industries in a new taxonomy of sectoral patterns
of innovation. Research Policy 37: 978–994.
Cohen, M. D.; Burkhart, R.; Dosi, G.; Egidi, M.; Marengo, L.; Warglien, M.;
and Winter, S. (1996) Routines and other recurring action patterns of
organizations: Contemporary research issues. Industrial and Corporate
Change 5:653-98.
Cohen, M. D., and Levinthal, D. A. (1990) Absorptive capacity: A new
perspective on learning and innovation. Administrative Science Quarterly
35:128-52.
Collier, A. (1994) Critical Realism – An Introduction to Roy Bhaskar’s
Philosophy. Oxfordshire: Verso.
Cooke, P. (2001) Regional innovation systems, clusters and the knowledge
economy. Industrial and Corporate Change 10:945-974.
56
Cooke, P. (2004) Regional knowledge capabilities, embeddedness of firms
and industry organization: Bioscience megacentres and economic geography.
European Planning Studies 12:625-641.
Crevoisier, O. (1999) Two ways to look at learning regions in the context of
globalization: The homogenizing and particularizing approaches. GeoJournal
49:353-361.
Crevoisier, O., and Jeannerat, H. (2009) Territorial Knowledge Dynamics:
From the Proximity Paradigm to Multi-location Milieus. European Planning
Studies 17:1223-1241.
David, P.A., and Foray, D. (1995) Accessing and expanding the science and
technology knowledge base. STI review 16:13-68.
De la Fuente, A., and Ciccone, A. (2002) Human Capital in a Global and
Knowledge-based Economy. Final Report for the DG Employment and Social
Affairs, European commission.
Dicken, P. (2007) Global shift, mapping the changing contours of the
world economy, 5th edition. London: Sage publications.
Dosi, G. (1982) Technological paradigms and technological trajectories: a
suggested interpretation of the determinants and directions of technical
change. Research Policy 11:147–162.
Dosi, G.; Faillo, M.; and Marengo, L. (2008) Problem solving and
governance in the capability-based view of the firm: the roles and theoretical
representations of organizational routines. In Handbook of Organizational
Routines, ed. M. C. Becker, 107-21. Cheltenham, U.K.: Edward Elgar
Publishing Limited.
Duranton, G., and Puga, D. (2004) Micro-foundations of urban
agglomeration economies, in Henderson J.V. and Thisse J.V. (Eds) Handbook
of Regional and Urban Economics, Vol. 4, pp. 2063–2117. North-Holland,
Amsterdam.
Eliasson, K. (2006) College choices and earnings among university
graduates in Sweden. Nationalekonomiska institutionen (Department of
Economics), Umeå University, Umeå.
Eriksson, R. (2011) Localized spill overs and knowledge flows: How does
proximity influence the performance of plants? Economic Geography 87:12752.
Eriksson, R.; and Lindgren, U. (2009) Localized mobility clusters: Impacts
of labour market externalities on firm performance. Journal of Economic
Geography 9:33-53.
EUROFOUND (2006) Mobility in Europe. Analysis of the 2005
Eurobarometer survey on geographical and labour market mobility. Office
for Official Publications of the European Communities, Luxembourg.
Farole, T.; Rodriguez-Pose, A.; and Storper, M. (2010) Human geography
and the institutions that underlie economic growth. Progress in Human
Geography 35:58-80.
57
Feldman, M. P., and Kogler, D. (2010) Stylized facts in the geography of
innovation, in: B. H. Hall and N. Rosenberg (Eds) Economics of Innovation.
Amsterdam: Elsevier, pp. 411–427.
Feldman, M. S. (2000) Organizational routines as a source of continuous
change. Organization Science 11:611-629.
Feldman, M. S. (2003) A performative perspective on stability and change
in organizational routines. Industrial and Corporate Change 12:727-52
Feldman, M. S., and Rafaeli, A. (2002) Organizational routines as sources
of connections and understandings. Journal of Management Studies 39:30931.
Florida, R. (2002) The rise of the creative class. New York: Basic Books.
Frenkel, A. (2001) Why high-technology firms choose to locate in or near
metropolitan areas. Urban Studies 38:1083-1101.
Gertler, M. (2001) Best practice? Geography, learning and the institutional
limits to strong convergence. Journal of Economic Geography 1:5-25.
Gertler, M. S. (2003) Tacit knowledge and the economic geography of
context, or The undefinable tacitness of being (there). Journal of Economic
Geography 3:75–99.
Giddens, A. (1990) The consequences of modernity. Stanford: Stanford
University Press.
Grabher, G. (2002a) Cool projects, boring institutions: Temporary
collaboration in social context. Regional Studies 36:205-214.
Grabher, G. (2002b) The project ecology of advertising: Tasks, talents and
teams. Regional Studies 36:245-262.
Grabher, G. (2004) Temporary architectures of learning: Knowledge
governance in project ecologies. Organization Studies 25:1491-1514.
Granovetter, M. (1985) Economic action and social structure: the problem
of embeddedness. American Journal of Sociology 91:481–510.
Grant, R. M. (1996a) Toward a knowledge-based theory of the firm.
Strategic Management Journal 17:109-122.
Grant, R. M. (1996b) Prospering in dynamically-competitive environments:
Organizational capability as knowledge integration. Organization Science
7:375-387.
Hall, P., and Soskice, D. (2001) Varieties of Capitalism: The Institutional
Foundations of Comparative Advantage. Oxford University Press.
Haltiwanger, J. C.; Lane, J. I.; and Spletzer, J. R. (1999) Productivity
differences across employers: the roles of employer size, age, and human
capital. American Economic Review 89:94–98.
Hamilton, B. H.; Nickerson, J. A.; and Owan, H. (2003) Team incentives
and worker heterogeneity: An empirical analysis of the impact of teams on
productivity and participation. Journal of Political Economy 111:465-497.
58
Harrison, J.; Hitt, M.; Hoskisson, R.; and Ireland, R. (2001) Resource
complementarity in business combinations: Extending the logic to
organizational alliances. Journal of Management 27:679-690.
Heimeriks, G., and Boschma, R. (2013) The path- and place-dependent
nature of scientific knowledge production in biotech 1986-2008. Journal of
Economic Geography, Advance Access; doi:10.1093/jeg/lbs052, 1-26.
Hine, D.; Parker, R.; Pregelj, L.; and Verreynne, M-L. (2013)
Deconstructing and reconstructing the capability hierarchy. Industrial and
Corporate Change (Advance Access doi:10.1093/icc/dtt046).
Hipp, C.; Tether, B.; and Miles, I. (2000) The incidence and effects of
innovation in services: Evidence from Germany. International Journal of
Innovation Management 4:417-454.
Howells, J. (2002) Tacit knowledge, innovation and economic geography.
Urban Studies 39:871-884.
Howells, J. (2012) The geography of knowledge: never so close but never so
far apart. Journal of Economic Geography 12:1003-1020.
Huckman, R. S., and Pisano, G. P. (2003) The effects of organizational
context on individual performance. NBER Working Paper No. 10027.
Hudson, R. (2009) From knowledge-based economy to … knowledge-based
economy? Reflections on change in the economy and development policies in
the north east of England. Regional Studies 45:997-1012.
Jaffe, A. B.; Trajtenberg, M.; and Henderson, R. (1993) Geographic
localization of knowledge spill overs as evidenced by patent citations.
Quarterly Journal of Economics 108:577–598.
Jansson, F. (1997) Åldersstrukturens betydelse för arbetskraftens
rörlighet. Arbetsmarknad och Arbetsliv 3: 295-306.
Kenney, M., and von Burg, U. (1999) Technology, entrepreneurship and
path dependence: Industrial clustering in Silicon Valley and Route 128.
Industrial and Corporate Change 8:67-103.
Kirat, T., and Lung, Y. (1999) Innovation and proximity: Territories as loci
of collective learning processes. European Urban and Regional Studies 6:2738.
Klepper, S. (2010) The origin and growth of industry clusters: The making
of Silicon Valley and Detroit. Journal of Urban Economics 67:15-32.
Kline, S.J., and Rosenberg, N. (1986) An overview of innovation, in R.
Landau, and N. Rosenberg, (eds) The positive sum strategy: Harnessing
technology for economic growth. Washington DC: National Academy Press,
pp. 275-304.
Knoben, J. (2009) Localized inter-organizational linkages, agglomeration
effects, and the innovative performance of firms. Annals of Regional Science
43:757-779.
59
Kogut, B., and Zander, U. (1992) Knowledge of the Firm, Combinative
Capabilities, and the Replications of Technology. Organization Science 3:383397.
Lacetera, N; Cockburn, I. M.; and Henderson R. (2004) Do firms change
capabilities by hiring new people? A study of the adaption of science-based
drug discovery. Advances in Strategic Management 21:133-159.
Lagerholm, M. (2007) Kunskap och innovation i ett moget kluster. En
ekonomisk-geografisk studie av aluminiumindustrin i Småland-Blekinge.
Geografiska regionstudier 74, Uppsala Universitet, Uppsala.
Lam, A. (1997) Embedded firms, embedded knowledge: Problems of
collaboration and knowledge transfer in global cooperative ventures.
Organization Studies 18:973-996.
Lam, A. (2000) Tacit knowledge, organizational learning and societal
institutions: An integrated framework. Organization Studies 21:487-513.
Lawson, T. (1997) Economics and reality. London: Routledge.
Lawson, C. and Lorenz, E. (1999) Collective learning, tacit knowledge and
regional innovative capacity. Regional Studies 33:305-317.
Leiponen, A., and Drejer, I. (2007) What exactly are technological regimes?
Intra-industry heterogeneity in the organization of innovation activities.
Research Policy 36:1221-1238.
Lucas, R. (1988) On the mechanics of economic development. Journal of
Monetary Economics 22:3-42.
Lundvall, B-Å. (1992) (ed) National systems of innovation. London: Pinter.
Lundvall, B-Å., and Johnson, B. (1994) The learning economy. Journal of
Industry Studies 1:23-42.
MacKinnon, D.; Cumbers, A.; and Chapman, K. (2002) Learning,
innovation and regional development: a critical appraisal of recent debates.
Progress in Human Geography 26:293-311.
Maillat, D., and Lecoq, B. (1992) New technologies and transformation of
regional structures in Europe: The role of the milieu. Entrepreneurship and
Regional Development: An International Journal 4:1-20.
Malerba, F. (2006) Sectoral systems: How and why innovation differs
across sectors, in J. Fagerberg, D. Mowery and R. Nelson, (eds): The Oxford
Handbook of Innovation. New York: Oxford University Press, pp 380-406.
Maliranta, M.; Mohnen, P.; and Rouvinen, P. (2009) Is inter-firm labor
mobility a channel of knowledge spillovers? Evidence from a linked
employer–employee panel. Industrial and Corporate Change 18:1161–1191.
Malmberg, A., and Maskell, P. (2002) The elusive concept of localization
economies: towards a knowledge-based theory of spatial clustering.
Environment and Planning A 34:429-449.
March, J. (2010) The ambiguities of experience. Ithaca: Cornell University
Press.
60
Markusen A. (1996) Sticky places in slippery space: A typology of industrial
districts. Economic geography 72:293-313.
Marshall, A. (1890) Principles of Economics: An Introductory Volume.
London: MacMillan.
Martin, R. (2000) Institutional approaches in economic geography, in E.
Sheppard, and T.J. Barnes, (eds): A companion to Economic Geography.
Malden: Blackwell Publishing, pp 77-94.
Martin, P.; Mayer, T.; and Mayneris, F. (2011) Spatial concentration and
plant-level productivity in France. Journal of Urban Economics 69:182-195.
Maskell, P. (2001) The firm in economic geography. Economic Geography
77:329-344.
Maskell, P.; Bathelt, H.; and Malmberg, A. (2006) Building global
knowledge pipelines: The role of temporary clusters. European Planning
Studies 14:997-1013.
Maskell, P., and Malmberg, A. (1999a) Localised learning and industrial
competitiveness. Cambridge Journal of Economics 23:167-185.
Maskell, P., and Malmberg, A. (1999b) The competitiveness of firms and
regions – ‘Ubiquitification’ and the importance of localized learning.
European Urban and Regional Studies 6:9-25.
Massey, D. (1984) The spatial division of labour. Social structures and the
geography of production. Basingstoke: Macmillan.
Miles, I. (2006) Innovation in services, in J. Fagerberg, D. Mowery and R.
Nelson, (eds): The Oxford Handbook of Innovation. New York: Oxford
University Press, pp 433-458.
Miner, A. S.; Ciuchta, M. P.; and Gong, Y. (2008) Organizational routines
and organizational learning, in: M. C. Becker (ed) Handbook of
organizational routines. Cheltenham, U.K.: Edward Elgar Publishing
Limited, pp. 152-186.
Morgan, K. (1997) The learning region: Institutions, innovation and
regional renewal. Regional Studies 31:491-503.
Morgan, K. (2004) The exaggerated death of geography: learning,
proximity and territorial innovation systems. Journal of Economic
Geography 4:3-21.
Neffke, F., and Henning, M. (2013) Skill relatedness and firm
diversification. Strategic Management Journal 34:297-316.
Neffke, F.; Henning, M.; and Boschma, R. (2011) How Do Regions Diversify
over Time? Industry Relatedness and the Development of New Growth Paths
in Regions. Economic Geography 87:237-265.
Neffke, F.; Henning, M.; and Boschma, R. (2012) The impact of aging and
technological relatedness on agglomeration externalities: a survival analysis.
Journal of Economic Geography 12:485-517.
Neffke, F.; Svensson Henning, M.; and Boschma, R. (2008) Surviving in
agglomerations: plant evolution and the changing benefits of the local
61
environment. Papers in Evolutionary Economic Geography 08.20, Utrecht
University, Utrecht.
Nelson, R. (1995) Recent evolutionary theorizing about economic change.
Journal of Economic Literature XXXIII:48-90.
Nelson, R., and Winter, S. (1982) An evolutionary theory of economic
change. Cambridge, MA: Harvard university Press.
Nonaka, I. (1994) A dynamic theory of organizational knowledge creation.
Organization Science 5:14-37.
Nonaka, I.; Ryoko, T.; and Akiya, N. (2000) A firm as a knowledge-creating
entity: A new perspective on the theory of the firm. Industrial and Corporate
Change 9:1-20.
Nooteboom, B. (2000) Learning and innovation in organizations and
economies. Oxford: Oxford University Press.
Nooteboom, B. (2009) A cognitive theory of the firm, learning,
governance and dynamic capabilities. Cheltenham: Edward Elgar.
Nooteboom, B.; Van Haverbeke, W.; Duysters, G.; Gilsing, D.; and van den
Oord, A. (2007) Optimal cognitive distance and absorptive capacity. Research
Policy 36:1016–1034.
Ottaviano, G. I. P. (2011) ‘New’ new economic geography: firm
heterogeneity and agglomeration economies. Journal of Economic
Geography 11:231-240.
Paci, R., and Usai, S. (1999) Externalities, knowledge spill overs and the
spatial distribution of innovation. GeoJournal 49:381-390.
Patel, P., and Pavitt, K. (1997) The technological competencies of the
world’s largest firms: complex and path-dependent, but not much variety.
Research Policy 26:141–156.
Pavitt, K. (1984) Sectoral patterns of technical change: towards a taxonomy
and a theory. Research Policy 13:343–373.
Pavitt, K. (1998) Technologies, products and organization in the innovating
firm: what Adam Smith tells us and Joseph Schumpeter doesn’t. Industrial
and Corporate Change 7:433–452.
Pavitt, K. (2005) Innovation Processes, in J. Fagerberg, D.C. Mowery, and
R.R. Nelson, (eds) The Oxford handbook of innovation. Oxford: Oxford
University Press, pp. 86-114.
Penrose, E. (1959) The theory of the growth of the firm. Fourth edition.
Introduction by C. N. Pitelis (Reprint 2013) Oxford: Oxford University Press.
Pentland, B. T. (1995) Grammatical models of organizational processes.
Organization Science 6:541-56.
Pentland, B. T., and Reuter, H. H. (1994) Organizational routines as
grammars of action. Administrative Sciences Quarterly 39:484-510.
Pinch, S., and Henry, N. (1999) Paul Krugman’s geographical economics,
industrial clustering and the British motor sport industry. Regional Studies
33:815–827.
62
Piore, M. J., and Sabel, C. F. (1984) The second industrial divide:
Possibilities for prosperity. New York: Basic Books.
Polanyi, M. (1962) Tacit knowing: Its bearing on some problems of
philosophy. Reviews of Modern Physics 34:601-616.
Polanyi, M. (1966) The logic of tacit inference. Philosophy 41:1-18.
Porter, M. (1990) The competitive advantage of nations. New York: The
Free Press.
Puga, D. (2010) The magnitude and causes of agglomeration economies.
Journal of Regional Science 50:203-219.
Rallet, A., and Torret, A. (1999) Is geographical proximity necessary in the
innovation networks in the era of global economy?. GeoJournal 49:373-380.
Rigby, D., and Essletzbichler, J. (1997) Evolution, process variety and
regional trajectories of technological change in U.S. manufacturing. Economic
Geography 73:269-284.
Romer, P. M. (1990) Endogenous technological change. The Journal of
Political Economy 98:S71-S102.
Rumelt, R. (1974) Strategy, structure and economic performance. Boston:
Harvard Business School Press.
Saxenian, A. (1994) Regional advantage: Culture and competition in
Silicon Valley and Route 128. Cambridge, MA: Harvard University Press.
Sayer, A. (1981) Abstraction: a realist interpretation. Radical Philosophy
28:6-15.
Schumpeter, J. (1942) Capitalism, socialism and democracy. New York:
Harper & Row.
Scott, A. (2004) A perspective of economic geography. Journal of Economic
Geography 4:479-499.
Solow, R.M. (1956) A contribution to the theory of economic growth.
Quarterly Journal of Economics 70:27-50.
Song, J.; Almeida, P.; and Wu, G. (2003) Learning-by-Hiring: When is
mobility more likely to facilitate interfirm knowledge transfer? Management
Science 49:351-65.
Spence, M. (1973) Job market signaling. The Quarterly Journal of
Economics 87:355-374.
Statistics Sweden (2014) Fokus på närlingsliv och arbetsmarknad hösten
2014 [Focus on business and labor market]. Information om utbildning och
arbetsmarknad 2014:3. Örebro: SCB-Tryck.
Sternberg, R. (1996) Reasons for the genesis of high-tech regions –
Theoretical explanations and empirical evidence. Geoforum 27:205-223.
Storper, M. (1997) The regional world: Territorial development in a global
economy. New York, NY: The Guilford Press.
Storper, M., and Scott, A.J. (2009) Rethinking human capital, creativity
and urban growth. Journal of Economic Geography 9:147-167.
63
Storper, M., and Venables, A.J. (2004) Buzz: face-to-face contact and the
urban economy. Journal of Economic Geography 4:351-370.
Strambach, S. (2013) Micro-dynamics of knowledge: actors, processes and
territorial organization. Working Papers on Innovation and Space 01.13.
Philipps University, Marburg.
Strambach, S., and Klement, B. (2012) Cumulative and combinatorial
microdynamics of knowledge: The role of space and place in knowledge
integration. European Planning Studies 20:1843-1866.
Szulanski, G. (1996) Exploring internal stickiness: Impediments to the
transfer of best practice within the firm. Strategic Management Journal
17:27-43.
Tanriverdi, H., and Venkatraman, N. (2005) Knowledge relatedness and
the performance of multibusiness firms. Strategic Management Journal
26:97-119.
Timmermans, B., and Boschma, R. (2013) The effect of intra- and interregional labour mobility on plant performance in Denmark: the significance
of related labour inflows. Journal of Economic Geography (Advance Access
doi:10.1093/jeg/lbs059).
Teece, D. J.; PISANO, G. P.; and SHUEN, A. (1997) Dynamic capabilities
and strategic management. Strategic Management Journal, 18:509-533.
Teece, D. J. (2010) Technological Innovation and the Theory of the Firm:
The Role of Enterprise-Level Knowledge, Complementarities, and (Dynamic)
Capabilities, in: B. H. Hall, and N. ROSENBERG, (Eds) Economics of
Innovation. Amsterdam: Elsevier, pp. 679-730.
Tether, B.; Hipp, C: and Hull, F. (2001) Standardization and
particularization in service: Evidence from Germany. Research Policy 30:115138.
Torre, A., and Gilly, J.P. (2000) On the analytical dimension of proximity
dynamics. Regional Studies 34:169-180.
Torre, A., and Rallet, A. (2005) Proximity and localization. Regional
Studies 39:47-59.
Trippl, M. (2011) Regional innovation systems and knowledge-sourcing
activities in traditional industries – evidence from the Vienna food sector.
Environment and Planning A 43:1599-1616.
Von Thünen, J. H. (1966) Von Thünen’s The Isolated State: An English
Edition of Der Isolierte Staat. Oxford: Pergamon.
Von Tunzelmann, N., and Acha, V. (2006) Innovation in “low-tech”
industries, in J. Fagerberg, D. Mowery and R. Nelson, (eds): The Oxford
Handbook of Innovation. New York: Oxford University Press, pp 407-432.
Vygotsky, L. (1962) Thought and language. Cambridge, MA: The MIT
Press.
Weber, A. (1929) Theory of the location of industries. Chicago: University
of Chicago Press.
64
Weick, K. E., and Roberts, K. H. (1993) Collective mind in organizations:
Heedful interrelating on flight decks. Administrative Sciences Quarterly
38:357-81.
Wernerfelt, B. (1984) A resource-based view of the firm. Strategic
Management Journal 5:171-180.
Wictorin, B. (2007) Är kluster lönsamma? EN undersökning om platses
betydelse för företags produktivitet [Are clusters profitable? The role of place
for firm productivity]. Doctoral Thesis No. 124, Department of Business
Studies, Uppsala University, Sweden.
Widerstedt, B (1998) Job Mobility, Wage Growth and Match Quality in
Sweden. Doctoral thesis: Moving or Staying? Job Mobility as a Sorting
Process. Umeå Economic Studies No. 464.
Yeung, H. W-C. (1998) The socio-spatial constitution of business
organizations: a geographical perspective. Organizations 5:101-128.
Zeller, C. (2004) North Atlantic innovative relations of Swiss
pharmaceuticals and the proximities with regional biotech arenas. Economic
Geography 80:83-111.
65
GERUM kulturgeografi
(1997-20xx)
(ISSN 1402-5205)
1997:1
Holm, E. (ed.): Modelling Space and Networks. Progress in Theoretical and
Quantitative Geography.
1997:2
Lindgren, U.: Local Impacts of Large Investments. (Akad. avh.)
1998:1
Stjernström, O.: Flytta nära, långt bort. De sociala nätverkens betydelse för val
av bostadsort. (Akad. avh.)
1998:2
Andersson, E.; L.-E. Borgegård och S. Hjort: Boendesegregation i de nordiska
huvudstadsregionerna.
1999:1
Stjernström, O. & I. Svensson: Kommunerna och avfallet. Planering och hantering av farligt avfall, exemplen Umeå och Gotland.
1999:2
Müller, D.: German Second Home Owners in the Swedish Countryside. (Akad.
avh.)
2000:1
Håkansson, J.: Changing Population Distribution in Sweden – Long Term Trends
and Contemporary Tendencies. (Akad. avh.)
2000:2
Helgesson, L.: Högutbildad, men diskvalificerad. Några invandrares röster om
den svenska arbetsmarknaden och vägen dit.
2000:3
Tollefsen Altamirano, A.: Seasons of Migrations to the North. A Study of
Biographies and Narrative Identities in US-Mexican and Swedish-Chilean
Return Movements. (Akad. avh.)
2001:1
Alatalo, M.: Sportfisketurism i Västerbottens läns inlands- och fjällområde. Om
naturresursanvändning i förändring. (Lic. avh.)
2001:2
Nilsson, K.: Migration among young men and women in Sweden. (Akad. avh.)
2001:3
Appelblad, H.: The Spawning Salmon as a Resource by Recreational Use. The case
of the wild Baltic salmon and conditions for angling in north Swedish rivers.
(Akad. avh.)
2001:4
Pettersson, R.: Sami Tourism - Supply and Demand. Two Essays on Indigenous
Peoples and Tourism in Sweden. (Lic. avh.)
2002:1
Alfredsson, E.: Green Consumption Energy Use and Carbon Dioxide Emissions.
(Akad. avh.)
2002:2
Pettersson, Ö.: Socio-Economic Dynamics in Sparse Regional Structures. (Akad.
avh.)
2002:3
Malmberg, G. (red.): Befolkningen spelar roll.
2002:4
Holm, E.; Holme, K.; Mäkilä, K.; Mattsson-Kauppi, M. & G. Mörtvik: The SVERIGE
spatial microsimulation model. Content validation, and example applications.
2003:1
Abramsson, M.: Housing Careers in a Changing Welfare State. (Akad. avh.)
2003:2
Strömgren, M.: Spatial Diffusion of Telemedicine in Sweden. (Akad. avh.)
2003:3
Jonsson, G.: Rotad, rotlös, rastlös – Ung mobilitet i tid och rum. (Akad. avh.)
2004:1
Pettersson, R.: Sami Tourism in Northern Sweden – Supply, Demand and
Interaction. (Akad. avh.)
2005:1
Hjort, S.: Rural migration in Sweden: a new green wave or a blue ripple? (Lic.
avh.)
2005:2
Malmberg, G.: Sandberg, L. & Westin, K.: Den goda platsen. Platsanknytning och
flyttningsbeslut bland unga vuxna i Sverige.
2005:3
Lundgren, A.: Microsimulation and tourism forecasts. (Lic. avh.)
2005:4
Brandt, B. F.: Botniabanan – Förväntningar i tid och rum på regional utveckling
och resande. (Akad. avh.)
2006:1
Helgesson, L.: Getting Ready for Life – Life Strategies of Town Youth in Mozambique and Tanzania. (Akad. avh.)
2006:2
Lundmark, L.: Restructuring and Employment Change in Sparsely Populated
Areas. Examples from Northern Sweden and Finland. (Akad. avh.)
2007:1
Li, W.: Firms and People in Place. Driving Forces for Regional Growth. (Akad.
avh.)
2007:2
Lundholm, E.: New Motives for Migration? On Interregional Mobility in the
Nordic Context. (Akad. avh.)
2007:3
Zillinger, M.: Guided Tourism - the Role of Guidebooks in German Tourist
Behaviour in Sweden. (Akad. avh.)
2008:1
Marjavaara, R.: Second Home Tourism. The Root to Displacement in Sweden?
(Akad. avh.)
2008:2
Sörensson, E.: Making a Living in the World of Tourism. Livelihoods in
Backpacker Tourism in Urban Indonesia. (Akad. avh.)
2009:1
Hjort, S.: Socio-economic differentiation and selective migration in rural and
urban Sweden. (Akad. avh.)
2009:2
Eriksson, R.: Labour Mobility and Plant Performance. The influence of proximity,
relatedness and agglomeration. (Akad. avh.)
2010:1
Khouangvichit, D.: Socio-Economic Transformation and Gender Relations in Lao
PDR. (Akad. avh.)
2010:2
Eriksson, M.: (Re)producing a periphery - popular representations of the Swedish
North. (Akad. avh.)
2010:3
Phouxay, K.: Patterns of Migration and Socio-Economic Change in Lao PDR.
(Akad. avh.)
2011:1
Hjälm, A.: A family landscape. On the geographical distances between elderly
parents and adult children in Sweden. (Akad. avh.)
2011:2
Sandow, E.: On the road. Social aspects of commuting long distances to work.
(Akad. avh.)
2011:3
Sandberg, L.: Fear of violence and gendered power relations. Responses to threat
in public space in Sweden. (Akad. avh.)
2011:4
Phommavong, S.: International Tourism Development and Poverty Reduction in
Lao PDR. (Akad. avh.)
2012:1
Haugen, K.: The accessibility paradox – everyday geographies of proximity,
distance and mobility. (Akad. avh.)
2012:2
Olofsson, J.: Go West – East European Migrants in Sweden. (Akad. avh.)
2013:1
Ouma, A.: From Rural Gift to Urban Commodity – Traditional Medicinal
Knowledge and Socio-spatial Transformation in the Eastern Lake Victoria
Region. (Akad. avh.)
2013:2
Åkerlund, U.: The Best of Both Worlds – Aspirations, Drivers and Practices of
Swedish Lifestyle Movers in Malta. (Akad. avh.)
2014:1
Olsson, O.: Out of the Wild: Studies on the forest as a recreational resource for
urban residents. (Akad. avh.)
2014:2
Gustafsson, C.: ”For a better life…” A study on migration and health in Nicaragua.
(Akad. avh.)
2015:1
Östbring, L.: Relatedness put in place. On the effects of proximity on plant
performance. (Akad. avh.)