University of Southern California Dissertations and Theses

The impact of social capital on economic performance: lessons from small and medium size enterprises (SMEs)

(USC Thesis Other)

The impact of social capital on economic performance: lessons from small and medium size enterprises (SMEs)

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content THE IMPACT OF SOCIAL CAPITAL ON ECONOMIC PERFORMANCE:
LESSONS FROM SMALL AND MEDIUM SIZE ENTERPRISES (SMES)
by
Iva Božović
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(POLITICAL ECONOMY AND PUBLIC POLICY)
December 2007
Copyright 2007 Iva Božović
Dedication
To my parents.
ii
Acknowledgments
This work has benefited immeasurably from Professor Timur Kuran’s guidance,
insights, and critique. I have learned immensely from his teaching and his
valuable comments about the process of research and academia in general. I
thank him for unbounded encouragement, patience and inspiration. I am grateful
to Professors Gillian Hadfield and Robert English for indispensable comments,
brain-storming sessions, and encouragement. I also thank Professors Jeffrey
Nugent, Carol Wise, John Odell, and Todd Sandler for many helpful comments
and advice throughout this research. I have received generous financial support
from the University of Southern California, USC School of International
Relations, and the USC Center for International Studies during the course of my
studies and research, and I greatly appreciate their trust and support.
The greatest source of unwavering support have been my parents, Nebojša and
Vlasta, which is why I dedicate this, and future work to them. Their direct help
with data collection was critical for this research but, more importantly, the
project would not have been completed without their inspiration and
encouragement. I thank them for believing in me unconditionally, for their endless
love, sacrifice, and patience. My sister Rajka took care of my sweet tooth over the
years and also helped with data entry. From near and far, she always cheered me
on. I want to also express profound gratitude to Kris Bruvold for sustaining me
with love, friendship and laughter. His help in editing and formatting the
presentation of this work has been crucial to its completion.
iii
Sincere gratitude is due to Erik and Mr. and Mrs. Bruvold, for their friendship,
valuable advice, and for offering a home away from home. I am grateful to Jelena
Tešić and Nick Angelov who encouraged me to begin this journey in the first
place. They have helped me overcome hurdles along the way while their friendship
has served as a vital source of motivation. Numerous other friends have each in
their own way enriched my life and chased away the stress over the past few years.
Among them are Iva Jakšić, Dollie Davis, Kris Campanale, Yuliya Minkova,
Marko Davidović, Steve Tavoni, Jovana and Dragiša Trifković and Ivana
Bošković. I am grateful to Ksenija Vidulić for always looking out for me and to
Una Matko, and Aleksandra Mandić for unforgettable adventures.
Thank you all.
iv
Table of Contents
Dedication ii
Acknowledgments iii
List of Tables viii
List of Figures ix
Abstract x
Chapter 1 Introduction 1
Chapter 2 SmallandMediumSizeEnterprises(SMEs)inTransition 9
2.1 Innovation and Competition . . . . . . . . . . . . . . . . . . . . . . 11
2.2 Employment Growth . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.3 Flexible Specialization . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.4 The Restructuring of Eastern Europe . . . . . . . . . . . . . . . . . 19
2.5 SMEs in South-Eastern Europe . . . . . . . . . . . . . . . . . . . . 22
2.6 Limits to SME Growth . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.6.1 Access to Finance . . . . . . . . . . . . . . . . . . . . . . . . 29
2.6.2 Rule of Law . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
2.6.3 Private vs. Public Order . . . . . . . . . . . . . . . . . . . . 34
Chapter 3 Social Capital: Concepts and Measures 39
3.1 The Social in Social Capital . . . . . . . . . . . . . . . . . . . . . . 40
3.2 The Capital in Social Capital . . . . . . . . . . . . . . . . . . . . . 42
3.3 Operationalizing Social Capital . . . . . . . . . . . . . . . . . . . . 44
3.3.1 Attitudes: Norms and Trust . . . . . . . . . . . . . . . . . . 44
3.3.2 Structures: Networks . . . . . . . . . . . . . . . . . . . . . . 51
3.4 The Future of Social Capital . . . . . . . . . . . . . . . . . . . . . . 55
Chapter 4 Social Networks in Transition: Better a Hundred Friends
than a Hundred Rubles 58
4.1 Networking Through Transition . . . . . . . . . . . . . . . . . . . . 60
4.2 Networks, Transition, and Economic Outcomes . . . . . . . . . . . . 61
4.3 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
v
Chapter 5 Methods and Data Collection 68
5.1 Network Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
5.1.1 Positional Approach . . . . . . . . . . . . . . . . . . . . . . 69
5.1.2 Relational Approach . . . . . . . . . . . . . . . . . . . . . . 72
5.2 Nodes and Ties: Who is in the Network. . . . . . . . . . . . . . . . 73
5.3 Research Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
5.3.1 Qualitative Data Collection . . . . . . . . . . . . . . . . . . 78
5.3.2 Quantitative Data Collection . . . . . . . . . . . . . . . . . 82
5.3.3 Scope of Analysis . . . . . . . . . . . . . . . . . . . . . . . . 88
Chapter 6 Typology of Social Networks 92
6.1 Reform Progress. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
6.2 Business Environment . . . . . . . . . . . . . . . . . . . . . . . . . 101
6.3 Content of Exchange . . . . . . . . . . . . . . . . . . . . . . . . . . 103
6.3.1 Personal Relations . . . . . . . . . . . . . . . . . . . . . . . 104
6.3.2 Connections . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
6.4 Network Boundaries . . . . . . . . . . . . . . . . . . . . . . . . . . 108
6.5 The Cost of Networking . . . . . . . . . . . . . . . . . . . . . . . . 111
Chapter 7 Networks and Investment: Exploring the Determinants
of Profit Reinvestment 115
7.1 Determinants of Profit Reinvestment . . . . . . . . . . . . . . . . . 116
7.1.1 Modeling Reinvestment . . . . . . . . . . . . . . . . . . . . . 121
7.2 Data and Measures . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
7.2.1 Profit Reinvestment. . . . . . . . . . . . . . . . . . . . . . . 127
7.2.2 Property Rights . . . . . . . . . . . . . . . . . . . . . . . . . 129
7.2.3 Use of Networks . . . . . . . . . . . . . . . . . . . . . . . . . 132
7.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
7.4 Robustness Checks . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
7.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
Chapter 8 Network Roles and Business Performance 149
8.1 Networks as Determinants of Economic Outcomes . . . . . . . . . . 150
8.2 Hypotheses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
8.2.1 Overall Impact . . . . . . . . . . . . . . . . . . . . . . . . . 153
8.2.2 Network Structure . . . . . . . . . . . . . . . . . . . . . . . 154
8.2.3 Relationship Characteristics . . . . . . . . . . . . . . . . . . 157
8.2.4 Embedded Resources . . . . . . . . . . . . . . . . . . . . . . 159
8.3 Data and Network Measures . . . . . . . . . . . . . . . . . . . . . . 160
8.3.1 The Dependent Variable . . . . . . . . . . . . . . . . . . . . 161
8.3.2 Network Measures . . . . . . . . . . . . . . . . . . . . . . . 162
8.3.2.1 Structure . . . . . . . . . . . . . . . . . . . . . . . 164
vi
8.3.2.2 Types of Relationships . . . . . . . . . . . . . . . . 167
8.3.2.3 Volume of Resources . . . . . . . . . . . . . . . . . 168
8.3.3 Additional Control Variables . . . . . . . . . . . . . . . . . . 169
8.4 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
8.5 Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
Chapter 9 Economizing on Transaction Costs: The Role of Social
Networks 182
9.1 Transaction Cost Framework . . . . . . . . . . . . . . . . . . . . . . 183
9.2 Choice of Transaction Governance . . . . . . . . . . . . . . . . . . . 185
9.2.1 Transaction Cost Elements . . . . . . . . . . . . . . . . . . . 186
9.2.2 Determinants of Transaction Costs . . . . . . . . . . . . . . 189
9.3 Illustrating the Choice of Transaction Governance . . . . . . . . . . 191
9.4 Transaction Characteristics and the Use of Networks . . . . . . . . 198
9.5 Efficiency Implications . . . . . . . . . . . . . . . . . . . . . . . . . 206
9.6 Networking Structures: Implications for Policy . . . . . . . . . . . . 212
9.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
Chapter 10 Economic Sociology and Economic Transitions 222
10.1 Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
10.2 Notes on Methodology . . . . . . . . . . . . . . . . . . . . . . . . . 233
10.3 Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . 237
Bibliography 240
Appendix A: Interview Participants 256
Appendix B: Quantitative Dataset 259
B.1 Sampling Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259
B.2 Sampling Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . 260
vii
List of Tables
2.1 Employment change . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.2 SME sector in SEE . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.3 Structure of SMEs by type of activity . . . . . . . . . . . . . . . . . 27
2.4 Short-term interest rates . . . . . . . . . . . . . . . . . . . . . . . . 30
2.5 Confidence in courts . . . . . . . . . . . . . . . . . . . . . . . . . . 35
5.1 Select respondent characteristics . . . . . . . . . . . . . . . . . . . . 87
7.1 Profits and reinvestment . . . . . . . . . . . . . . . . . . . . . . . . 128
7.2 Perceived security of property rights . . . . . . . . . . . . . . . . . 129
7.3 Perceived security of property rights, with JMW indicators . . . . . 130
7.4 Networks in dispute resolution . . . . . . . . . . . . . . . . . . . . . 132
7.5 Time devoted to maintaining contacts . . . . . . . . . . . . . . . . . 134
7.6 Determinants of profit reinvestment . . . . . . . . . . . . . . . . . . 137
7.7 Predicted profit reinvestment rates 1 . . . . . . . . . . . . . . . . . 141
7.8 Predicted profit reinvestment rates 2 . . . . . . . . . . . . . . . . . 143
8.1 Sales growth over 3 years . . . . . . . . . . . . . . . . . . . . . . . . 161
8.2 Descriptive statistics for two network types by network size . . . . . 166
8.3 Regressions of network characteristics on sales growth . . . . . . . . 171
8.4 Regressions for sales growth as a function of network roles . . . . . 174
8.5 Predicted sales growth rates (trust network) . . . . . . . . . . . . . 176
8.6 Predicted sales growth rates (resource access network) . . . . . . . . 178
9.1 Summary of transaction costs . . . . . . . . . . . . . . . . . . . . . 188
9.2 Transaction characteristics and illustrative examples . . . . . . . . . 192
A.1 SME interview participants . . . . . . . . . . . . . . . . . . . . . . 256
viii
List of Figures
2.1 Minimum efficient scale of production . . . . . . . . . . . . . . . . . 12
5.1 Plan of analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
6.1 Industrial output and GMP, FR Yugoslavia (1990–2001) . . . . . . 96
6.2 Types of networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
7.1 Property rights components . . . . . . . . . . . . . . . . . . . . . . 124
9.1 Transaction cost framework . . . . . . . . . . . . . . . . . . . . . . 190
ix
Abstract
The study bridges transition scholarship with social network studies to examine
whether the social embeddedness of economic actors in transition is a source of
advantage or an impediment for private sector growth. Transition scholars have
recognized the omnipresent role of informal personal exchange, reciprocal ties,
social networks, and exchange of favors as socialist legacies and path-dependent
cultural aspects in the post-communist context. However, they have been less
successful in recognizing the role of these cultural factors as agents of institutional
change. Adopting a structural approach to social capital, this analysis centers on
social networks and the resources embedded in the relationships that form them.
Original empirical evidence from small and medium size enterprises (SMEs) in
Serbia and Montenegro is utilized to examine whether differences in business
success can be explained through the differences in micro-level social capital and
the nature of informal exchange relationships.
Empirical analysis shows that, much like confidence in courts, informal
networks are an important component of the security of property rights that
encourage private sector investment. Variation in the extent to which networks
are used in dispute resolution and the degree to which SME managers and owners
invest in their business ties help explain differences in profit reinvestment.
Probing further into what network aspects drive small business success, the study
finds that network structure positively impacts growth in sales. However,
structural impact is dependent on the content of exchange facilitated by network
ties. When it comes to networks’ role in serving as an informal mechanism for
x
contract enforcement, dense and homogenous networks are preferred. For
accessing valuable resources, capital, and favors, heterogeneous and less dense
networks perform better.
Lastly, this study employs the transaction cost framework to show that costly
network ties represent an efficient investment of limited resources. However,
individual decisions to optimize on transaction costs through the use of networks
can coordinate to socially inefficient outcomes. Suboptimal outcomes can persist
when personalized exchange prevents changes to formal institutions designed to
lower incentives for the use of social networks.
xi
Chapter 1
Introduction
Almost 15 years after the beginning of transition reforms in most of socialist
Eastern Europe, Dani Rodrik boldly proclaimed that the “Washington Consensus”
is dead (2006). Formulated in the early 1990’s this approach became the norm for
analyzing and executing the policies necessary to successfully guide post-socialist
countries to open-market economics and economic growth. Central to the
prescription was the assumption that the relevant economic technocrats would
implement, through a top-down approach, macroeconomic stabilization,
liberalization, and privatization, which would open up the economies to free trade,
eliminate sources of macroeconomic instabilities, and remove any traces of state
involvement in the markets.
The alternative, or the “Gradualist Approach,” consists of a rather
heterogeneous body of scholarship unified in its criticism of the policies prescribed
by the “Washington Consensus.” Gradualists center their arguments on the notion
that the process of transition demands a radical restructuring of the economic
system which also alters actors’ expectations (Arrow 2000; Murrell 1995; Stiglitz
2000). Thus, institutions providing the bedrock of rule of law and security of
1
property rights, shown to be key prerequisites for markets, must be established
before reforms can begin. Unlike the dominant approach, criticized for its
championing of neo-liberal policies, the latter approach focuses predominantly on
“getting the institutions right” (Rodrik 2006). This complex approach contains
fewer concise policy prescriptions compared to the clear policy recommendations
originating from the “Washington Consensus.” Recognizing that ultimately all
institutions are deeply embedded in societies, cultures and histories, the
gradualist approach offers little in terms of quick policy solutions.
Though the experience over the last two decades has provided evidence to
suggest that the heyday of the “Washington Consensus” has past, the debate over
the appropriate approaches to development and economic transitions is hardly
over. A key problems is that few studies suggest concrete prescriptions for
developing the optimal institutional infrastructure necessary to initiate or sustain
economic growth. Even Rodrik’s approach to formulating growth strategies,
which requires diagnostic analysis and targeted policy design, recognizes that the
reforms needed to build institutions for safeguarding the economy from external
shocks are difficult to implement. Correctly determining the constraints to
economic growth, and formulating the correct set of policy reforms to ameliorate
those constraints, may still not produce economic growth if the necessary
institutional reforms cannot be implemented. Essentially, the debate over the
most appropriate approach to institutional change will continue because we know
little about the actual process of institutional change.
Institutions consist of formal and informal rules, as well as their enforcement
mechanism, that structure the economic and political environment by setting the
"rules of the game" and by altering actor’s incentives (North 2005). Together,
2
institutions reflect the accumulated beliefs of a society over time. Formal
institutions include formally codified and legally enforced political, economic and
judicial rules, such as written laws and property rights. Informal institutions
include conventions, norms and customs: “Arising to coordinate repeated human
interaction, [informal institutions] are 1) extensions, elaborations and
modifications of formal rules, 2) socially sanctioned norms of behavior, and 3)
internally enforced standards of conduct” (North 1990, 40). Informal institutions
are not legally enforced and are not necessarily codified but together with formal
constraints, they form a complex set of rules that constrain individuals’ behavior
and reduce uncertainty. Because informal institutions are embedded in the
societal beliefs and experiences accumulated over time, they are less amenable to
change and are highly path-dependent. Thus, changes in formal institutions
through direct policy measures have unpredictable outcomes because neither the
interaction between formal and informal institutions, nor the impact of policy
instruments on informal institutions can be understood or predicted a priori.
Transition scholarship did not fully recognize the importance of informal
institutions to the process of democratic and economic transformation. Offered
explanations for the differences in transition outcomes stopped short of grasping
the multifaceted nature of economic transition and failed to point out that it is as
much a bottom-up as it is a top-down process. Consequently, scholars overlooked
economies’ social and cultural characteristics that, as informal institutions, could
bring economic reforms to fruition or hinder their implementation. In the end,
what the policy-oriented transition scholarship missed is that the economic
principles of Western developed economies, and the inviolability of the rule of law
along with the requisite institutions, would not necessarily become ingrained in
3
the social and cultural fabric of formerly communist societies, and their impact
would not inevitably correspond to the intended objectives. Overall, transition
literature adopted the “undersocialized” approach to the transition process,
denying “any impact of social structure and social relations on production,
distribution, or consumption” (Granovetter 1985, 483).
Key to economic growth and improvements in economic performance is “some
combination of formal rules and informal constraints and the task we face today is
to achieve an understanding of exactly what combination will produce the desired
results both at a moment of time and over time” (North 2005, 79). Formal and
informal institutions, along with the state of technology, determine transaction
costs for engaging in political and economic exchange. Through transaction costs,
the complex web of institutions determines which markets will function and how
effectively. Implicit in this last statement is that institutions ultimately determine
the extent of specialization and division of labor and, consequently, the level of
productivity that affects economic performance (Yeager 1999). New Institutional
Analysis, and more specifically Comparative Institutional Analysis, provide tools
for the examination of formal institutions and their implications in terms of
efficiency and social welfare. Similar tools for examining informal institutions are
far less developed or consistently employed.
1
To begin to understand the
interaction between the two systems of constraints on social and economic
activities, we must systematically and empirically approach the study of informal
institutions, especially their sources of path-dependence and impact on economic
incentives.
1
Nevertheless, groundbreaking examples of scholarly work that can speak to the issues at
hand include among others Platteau (2000), Greif (1994), Kuran (2003), Kuran (2004), Landa
(1994), and legal scholarship examining the relationship between private and public forms of
legal order.
4
Many transition scholars have turned to the concept of social capital to
explain the impact of norms and informal networks that dominated socialist
society and continue to exist as coping mechanisms during transition. Social
capital is a metaphor that has been evoked to suggest the potential benefits of
beliefs, norms, relationships, networks, and trust for economic and political
advancement. In the mid-1990’s the concept of social capital came into
prominence as “the missing link” that, together with natural, physical and human
capital, can explain the causes of economic development (Grootaert 1998). As the
process of transition began to approach the end of its first decade, with less than
spectacular results, the concept of social capital received increased attention in
policy-oriented scholarship struggling to better specify the relationship between
markets and states, markets and culture, and formal and informal institutions.
Social capital studies offered inconsistent conclusions. Some blamed
institutional change and the breakdown in the rule of law for the decline in trust
and norms of reciprocity. Others stressed the importance of socialist legacies for
encouraging the use of informal social exchange to compensate for the weaknesses
in the institutional setting. Yet others saw informal social exchange as an attempt
to exploit the system and undermine the establishment of lasting market and
democratic institutions. One aspect of transition consistently mentioned in social
capital literature is the prevalence of exchange networks that operated in the
socialist system and that continue to permeate market structures since the onset
of transition. The primary objective of this study is to investigate to what extent
embeddedness of economic activity in social exchange relations or social networks
facilitates or impedes private sector growth. This objective is important for
understanding the growth of small and medium size enterprises (SMEs) which
5
represent a key resource on which a broader economic growth strategy can
capitalize. The study undertakes an empirical examination of how social networks
impact entrepreneurial incentives for reinvestment and the success of business
performance in a transition setting.
Chapter 2 reviews the importance of SMEs to the success of economic
transition in South-Eastern Europe. SMEs play an important role in innovation
and employment growth, and are an integral part of liberalization, privatization
and restructuring reforms. The chapter argues that slower transitioning economies
of South-Eastern Europe have poorly developed SME sectors. Growth of
individual businesses is found to be critically important to the success of the SME
sectors. Two central barriers for the growth of SMEs are introduced, namely
access to finance and the rule of law, of which the latter is more critical and
partially explains the presence of the former. The chapter also discusses the
importance of informal ways for doing business that compensate for the critical
obstacles to SME performance.
The following chapter (Chapter 3) offers a critical examination of the field of
social capital which has grown out of an interest to explain the role of norms,
values, trust and social networks in economic and political institutional change.
The chapter draws attention to the problems in conceptualizing and
operationalizing the term social capital and argues that the most valuable
direction for the future development of the field rests upon a greater focus on
social capital’s structural elements such as social networks. Chapter 4 reviews the
transition scholarship examining the role of informal social exchange in the
process of transition. The chapter concludes that social networks compensate for
the deficiencies in the socialist economic system at the same time that they enable
6
the exploitation of the public sector for personal gain. The implications for
private sector growth, however, are not clear.
To understand how networks and resources embedded in them translate into
costs and benefits for economic actors, and whether they positively or negatively
impact business performance in the private sector, the study adopts a relational
approach. This approach to the study of social networks is contrasted with the
alternative, positional approach in Chapter 5. Within the relational approach,
both structural and individual elements are emphasized and the chapter offers a
detailed definition and conceptualization of the social network. Two elements of
data collection are also introduced in the chapter. The study is based on
qualitative data, collected through in-person, open-ended interviews with SME
managers and owners in Serbia and Montenegro, and an original quantitative
dataset constructed using questionnaires mailed to a random sample of SME
managers and owners in the same region. Some of the characteristics of the data
collected and the analyses that follow are also reviewed.
Chapter 6 offers a typology of networks whose use was revealed in the
interviews with SME managers and owners in Serbia and Montenegro. The
following chapter (Chapter 7) employs the surveys from Serbia and Montenegro to
demonstrate the important role that networks play in providing an informal
mechanism for insuring that parties to a business transaction respect the terms of
the contract—a mechanism that does not rely on the courts or the formal legal
system for contract enforcement. The chapter uses the model from Johnson,
McMillan, and Woodruff (2002b) as a starting point but argues that as a
determinant of reinvestment, security of property rights must be disaggregated
into measures of the threat of expropriation, and confidence in formal and
7
informal means for contract enforcement. The chapter demonstrates that despite
costly investments, networks are positively and directly impacting the rate of
profit reinvestment.
Chapter 8 explores what aspects of social networks are critically important for
improving business performance. In doing so, it considers the relative impact of
network relationships, network structure and network resources. Most
importantly, by considering two different types of networks—a trust network and
a resource access network—the analysis also addresses the impact of the content
of network exchange on the success of individual businesses. The self-reported
growth in sales over the course of the previous 3 years of operations is used as a
dependent variable. The chapter shows that to understand the impact of social
networks we must separate them into trust and resource access networks and then
consider the impact of structure, relationships and resources in reference to each
of the two network roles.
Chapter 9 uses the transaction cost approach to clarify the elements of an
individual manager’s decision whether to use personal social networks or
impersonal markets to conduct any given transaction. Contrasting the total costs
of transacting in each of the exchange settings shows that social networks
sometimes offer a cost-saving alternative to the impersonal market but that their
predominant use is associated with significant social costs. Individually efficient
strategies thus result is less than efficient social outcomes. The remainder of the
chapter is devoted to summarizing these social costs and explaining why their
presence raises policy concerns. The final chapter (Chapter 10) summarizes the
findings of the overall study, its main implications, and a few methodological
lessons.
8
Chapter 2
Small and Medium Size Enterprises
(SMEs) in Transition
A vibrant small and medium size enterprise (SME) sector has been identified by
transition literature as a key factor for successful economic transition in Eastern
Europe. SMEs are roughly defined as enterprises with 250 or fewer employees.
Small enterprises have fewer than 50 employees and medium size companies have
between 50 and 250 employees. A more nuanced definition also includes micro
businesses which have less than 10 employees (OECD-EBRD 2003; Hull 1999;
OECD 1997). An alternative definition, often used by U.S. scholars and
practitioners, classifies all businesses with up to 500 employees as SMEs. With
respect to transition economies, the former definition is more commonly used,
especially by the World Bank and its partner organizations in Europe, and is
adopted in this work.
Recent attention to the small and medium size enterprise sector (SME) in
transition economies has resulted from the positive assessment of the sector’s
contribution to sustained economic growth. Several trends have elevated SMEs to
9
the top of the economic agenda, among them, the unprecedented growth of
employment associated with this sector in the 1980’s and 90’s in the United
States, and its role in technological progress of localized innovation centers such
as Silicon Valley, which have challenged the traditional conceptualization of
national competitive advantage. Reflecting upon SME performance in the
Western developed economies, we learn that the primary contributions of SMEs
are a positive impact on innovation and the job creation rate.
In the context of transition economies, SMEs play additional roles. At the
onset of transition, the SME sector emerges to fulfill the demand for goods and
services undersupplied by the socialist, large-enterprise sector. Newly emerging
start-up businesses in post-communist economies, which by definition are small
and medium enterprises, are relied upon to absorb the excess labor shed by the
large state-owned enterprises undergoing restructuring and privatization reforms.
The entire SME sector, which includes both start-up and spin-off enterprises,
plays a central role in generating employment opportunities. In the EU, close to
75% of all employment belongs to the SME sector, which also contributes 30–60%
of the Gross Domestic Product (GDP) (OECD 1997, 17–18).
1
In successful
transition economies, the central role of the SME sector is reflected in similar
contributions to growth and employment (UNECE 2006, 106). SMEs are also
lauded for their ability to mobilize local resources and facilitate innovation.
Overall, in transition economies, SMEs help ameliorate the dislocations resulting
from transition processes and are collectively conceptualized as the primary force
behind realized and potential economic growth of transition economies.
1
OECD data does not include businesses in agriculture. Moreover, OECD considers SMEs
to have 500 employees or less and therefore its data is not directly comparable with the data for
Eastern European SMEs that include 250 employees or less.
10
This chapter reviews the critical contribution of SMEs to economic growth and
the success of economic, post-communist transition. This overview is based on the
SME sector’s role in developed market economies and the additional roles the
sector undertakes during the process of transition. The chapter then reviews the
performance of the SME sector in South-Eastern Europe and argues that the poor
performance of SMEs in the region can help explain the poor results of transition
overall. The chapter also reviews critical obstacles to SME growth and shows that
the ability to enforce contracts and the formal rule of law play a dominant role.
Lastly, the chapter reviews private forms of contract enforcement as an alternative
to the formal rule of law and argues that we must further investigate the ability of
these private order mechanisms to encourage SME growth.
2.1 Innovation and Competition
Up to the end of 1970’s, industrial organization theory strongly emphasized the
notion of scale economies and the need for businesses to work at a capacity that
would produce at least the level of output at which the minimum average cost
(AC) is attained—Q
o
in Fig. 2.1 (Audretsch 2002a).
2
Below this level of output,
for example Q
b
, marginal cost (MC) is below the average total cost and the
average cost is decreasing with each additional unit of output produced up to Q
o
.
If that is the case, economies of scale theory dictates that producing additional
units above Q
b
increases efficiency because the average cost of the units produced
is lowered despite the rising cost of the last unit produced (rising MC). In other
words, all businesses operating below the Q
o
level of output are considered
economically inefficient because economies of scale remain underutilized. SMEs,
2
At this level of output, marginal cost is equal to the minimum level of average cost.
11
MC
AC
$
Q o Q b
Q
p
Figure 2.1: Minimum efficient scale of production
most studies found, are suboptimal since they operate at an inefficient level of
output. This “static” view of industrial organization theory suggested that any
policies in favor of enlarging the SME sector negatively impact economic welfare
because they favor the establishment of low-productivity and thus low-wage firms.
An alternative viewpoint slowly emerges beginning with the late 1970’s (Acs,
Carlsson, and Karlsson 1999). Placing at the center of analysis the inventor, as
opposed to the individual firm, this alternative approach focused on the role of
knowledge and innovation in producing change. Due to the asymmetric nature of
knowledge, an innovator and his current employer will have diverging estimates of
the potential value of an invention (Audretsch 2002a). At times, the inventor may
find it more profitable to establish a start-up firm dedicated to researching and
developing a particular innovation as opposed to working with the current
employer. Hence, small firms are established not to emulate the performance of
large firms but to take advantage of innovative activity, and introduce ideas and
products that are not taken up by large firms operating at capacity. The notion of
scale economies has not been abandoned in this dynamic viewpoint of industrial
12
organization. Of the small firms that enter the market, only those that offer a
successful product and are able to achieve the minimum efficient scale of
production will be able to survive and grow (Audretsch 2002a). Others will
stagnate and eventually exit the particular industry. Hence, in an economy, a
multitude of SMEs operating below the capacity that guarantees minimum
average cost is not necessarily representative of underutilized economies of scale
but may serve as an indicator of dynamic economic efficiency and the potential for
long-term economy-wide benefits.
The dynamic view of the role of SMEs emphasizes two specific contributions:
one to innovation and the other to the process of “creative destruction.”
3
With
respect to the former, recent evidence of the overwhelming role of SMEs in
developing new technologies and industries has initiated an exploration into
whether SMEs enjoy certain advantages over large firms when it comes to
innovation. In Canada for example, in 1990, 68.9% of companies carrying out
formal research and development (R&D), had fewer than 50 employees (OECD
1997, 18). However, this group of enterprises comprised only 13.2% of the total
formal R&D expenditures (19). This indicates that although formal R&D
expenditures are concentrated in large firms, SMEs tend to undertake innovative
activities in greater numbers. Large firms also have more employees and if we take
into account the rate of introduction of new products or processes per employee,
3
Schumpeter (1942) introduced the concept of creative destruction to explain the distinctive
capacity of the capitalist system to renew itself from within. In his interpretation, the
evolutionary forces of the capitalist system are unleashed when new commodities or new
technologies are introduced, awarding companies with significant cost or quality advantages that
fundamentally transform the production and supply of that product, forcing other businesses in
the industry to exit. This form of competition goes beyond the more simplistic notion of price
competition (84).
13
small firms dominate with the rate 0.309 compared to 0.202 for large firms
(Audretsch 2002b).
Often, small enterprises compensate for small R&D budgets by developing
close ties with other firms so that their interactions facilitate the spread of
information and spillover of knowledge, which reinforces innovation (Almeida
1999). This strategy is especially successful when inter-firm networks are
geographically concentrated. These “clusters” arise when firms locate in the same
area to take advantage of the proximity of competitors, research universities,
suppliers, and highly specialized labor markets. The best known example of the
SMEs’ innovation success in this form is Silicon Valley in California.
SMEs also differ from large firms in the types of innovative activities they
undertake. Overall, SMEs play a primary innovative role through projects that
large firms find too risky to undertake. In the case of high uncertainty,
independent small firms invest in discovery-oriented research whereas large firms
dominate the field of design-improvement research and innovation (Carlsson
1999). Because small firms tend to populate the fields that have not yet attracted
a lot of research attention they add to the diversity of new discoveries. Large
firms, on the other hand, often acquire innovations of small enterprises to develop
them further and market them for final consumption. Spreading the cost of design
innovation over a large output lies at the root of large firms’ efficiency in design
research and development. At the same time, SMEs expedite the diffusion of
innovations by finding alternative usages for products or developing and adapting
key innovations to meet specific client needs (Pfirrmann 2002). For example,
many small firms buy second-hand equipment and improve it by changing or
adding functions, or by transforming it beyond the intentions of the manufacturer.
14
The research and development (R&D) expenditures for this form of technological
innovation are informal and do not enter into official statistics, which
underestimates the total level of R&D activities performed by small firms (OECD
1997). In sum, small firms have a greater rate of innovation per employee, carry
out far more informal R&D activities, and tend to dominate technological fields
ignored by large firms, even though in terms of dollar value, they cannot outdo
the innovation expenditures of large enterprises.
The second element at the heart of the dynamic conception of the role of
SMEs in an economy is the notion of “creative destruction” (see footnote (3) on
page 13). This notion represents the process whereby firms that can no longer
compete exit the market making room for more creative and efficient competitors.
Central to this renewal are small new firms built around innovative ideas that, if
successful, enable the firms to grow into large competitive enterprises, or to be
bought up by large firms that have the resources and ability to develop their ideas
further. Small firms thus reinforce market competition and assist in pushing it to
the edge of technological advancement and innovative research. The entry of small
and innovative firms also ensures that heterogeneity in the economy is maintained.
Heterogeneity and volatility, reinforced by competition and exit, provide the most
essential elements of dynamic economic growth (Carlsson 1999). Accordingly, a
distribution of firms that is skewed heavily towards the small firm size is not a
sign of unused capacity and inefficiency, but evidence of a vibrant economy based
on the entry of small firms that respond to new opportunities and introduce new
ideas that ultimately fuel economic competition and productivity growth.
15
2.2 Employment Growth
SMEs also contribute to economic growth through their role in creating jobs.
Since the 1970’s, non-farm employees in Western developed economies have been
working in increasingly smaller firms. The SME share in total employment has
been increasing continuously, particularly in the goods producing sectors such as
construction and manufacturing (Acs 2003, 26). In major European economies,
small firms employ between 55 and 85% of the labor force (OECD 1997, Table
1.1). Table 2.1 illustrates that the rise in the SME share of employment, however,
cannot be explained by the entry of new small firms. This is because SMEs also
have high death rates compared to large enterprises.
4
The net increase in
employment due to small and medium size businesses results because firms of this
size, once established, grow more quickly than large firms (Hart 2000). These
figures do not take into account the differences in the sectors within which
business operate. Thus for some sectors, large businesses may be a greater source
of job growth than small or medium size businesses. Considering the economy as
a whole, however, shows that, net of job losses, small firms are better job creators
than large firms. Table 2.1 shows that the net change in employment (for 1992) in
small and medium size businesses exceeds the change in employment within large
enterprises, even though the latter category employs more individuals overall.
Expansion of existing SMEs is thus the most significant source of job creation in
the E.U. and the U.S. (Hull 1999; Audretsch 2002a).
4
Line 5 in Table 2.1 shows the percentage change in employment due to job destruction per
size of the firm. The job destruction rate for large enterprises is anywhere between 12.7 and
14.5% and for small and medium enterprises this rate is anywhere between 14.5 and 20.7% of
the total change in employment.
16
Table 2.1: Employment change due to birth and death of enterprises (U.S.A., 1992–1993)
Firm Size (Number of Employees)
Data Type Total 1 – 4 5 – 9 10 – 19 20–99 100 – 499 500 +
Employment in 1992 92,791,532 5,171,122 6,195,603 7,383,80 17,111,967 13,304,504 43,624,956
Net Change 1,948,253 1,081,375 288,193 106,260 68,760 242,648 161,017
Percent Change, Total 2.1 20.9 4.7 1.4 0.4 1.8 0.4
Percent Change, Job Creation 16.8 41.5 23.1 18.6 16 16.4 13.1
Percent Change, Job Destruction 14.7 20.7 18.5 17.2 15.6 14.5 12.7
Percent Change, Job Reallocation 31.35 62.2 41.6 35.38 31.6 30.9 25.8
Source: Acs, Carlsson, and Karlsson (1999).
17
2.3 Flexible Specialization
Increased appreciation for the contribution of SMEs to economic welfare through
innovation and job creation also coincides with the change in the determinants of
success in the globalized market. Since the oil shocks of the 1970s, survival and
success in a dynamic global economy has been ensured by flexibility in responding
to significant market changes. Certain markets now cater to specialized products
that meet changing consumer needs and result in shorter product life-cycles.
Small firms can meet these conditions more easily than large firms focused on
capital-intensive, mass production of standardized products. While large firms
can better cope with short term shifts in demand and supply, small firms respond
better to persistent shocks that encourage the formation of new markets, or entail
changes in the workforce, equipment and techniques (Casson 1999). The more
large firms suffer from managerial inertia, such that management and firm
bureaucracy are resistant to change and prefer to continue with existing
techniques and projects, the greater the advantage of small firms in terms of
flexibility and adaptability to changing market conditions.
The trend toward “flexible specialization” has encouraged the downsizing of
large companies and the formation of subcontracting and outsourcing
relationships with SMEs (OECD 1997, 18). Instead of inefficient small firms
trying to emulate the success of large firms, today we see small and large firms,
specializing in different segments of the same industry, forming inter-firm network
linkages to lower transaction costs and facilitate cooperation, while maintaining
flexibility in production. Hence, emphasizing the importance of SMEs in an
economy does not stipulate that we negate the importance of large firms for
productivity and economic growth. It does, however, entail noting the increased
18
participation of SMEs in the overall volume of economic activity and job creation,
their significant role in supporting innovation and renewal though creative
destruction, and their advantages in responding to changing market conditions.
2.4 The Restructuring of Eastern Europe
Economic theory based on the experience of developed economies indicates that a
large number of SMEs and the expansion of individual SMEs are both desirable
for economic growth. The former contributes to long-term growth and dynamic
efficiency through the role of SMEs in innovation and the latter is important
because of the role of SMEs in creating jobs. Economic reforms that have been
undertaken in SE Europe suggest that SMEs could play additional roles
specifically in the context of transition. Economic transition refers to the process
of moving from an economy based on state-owned property and centrally planned
governance to an economy based on the principles of free market economics,
which include the dominance of private property and the use of the market as a
coordination mechanism.
5
Three crucial components of successful reform are
liberalization, privatization, and restructuring.
6
Each of these transition reforms
envision an important role for the SME sector.
Liberalizing the markets and opening them up to domestic and foreign
competition directly promotes the entry of new SMEs. Though foreign companies
may establish large-scale production, new domestic entrants are almost by
definition small firms. As a result, the largest share of the SME sector is
5
Such reforms are naturally preceded or accompanied by a set of complementary political
reforms but they are not addressed in this work in order to maintain the focus on reforms that
have consequences in the economic sphere.
6
An overview of essential transition reforms can be found in Svejnar (2002), Murrell (1996),
Estrin (2002), and Lipton and Sachs (1990).
19
comprised of newly established firms. At the same time, large state-owned
enterprises can privatize and restructure by downsizing and converting viable
units into smaller self-sustained enterprises. These spin-offs continue to operate
independently in the SME sector as private entities or entities in mixed
ownership.
7
In addition, depending on the country in question, the SME sector
can also include a number of SMEs that were established prior to the onset of
transition reforms. Surviving pre-transition SMEs are most common in transition
economies where prior socialist governments tolerated a sector of small enterprises
employing a few workers and operating in light industry and traditional services
(Kolodko 2003). This is particularly true of Hungary, Poland, and the countries of
former Yugoslavia. The SME sector in transition economies is thus comprised of
newly established enterprises, spin-offs of large enterprises, and enterprises
established before the beginning of transition.
In accordance with the goals of privatization and restructuring, SMEs promote
the establishment of private ownership and are likely to diversify the structure of
economic activity by encouraging entrepreneurship. For example, SMEs are relied
upon to supply goods and services that were not previously supplied, or were
undersupplied by state-owned monopolies. Also, SMEs can facilitate integration
into international trade by encouraging imports and exports, and in some cases
serving as export agents for large state-owned enterprises and industries.
8
7
State retains a controlling portion of equity.
8
For example, before the transition and the break-up of former Yugoslavia, frozen fruit and
vegetable producers worked as providers for large food and beverage conglomerates. With the
break-up of Yugoslavia, such links were broken as the fruits and vegetables are now grown
mainly in Serbia, while many of the major food and beverage producers are in Slovenia and
Croatia. The remaining companies in Serbia and Montenegro represent only a fraction of the
market previously supplied by the Serbian frozen fruit and vegetable industry. As a result, many
small enterprises emerged to help domestic producers establish links with food and beverage
industries in what are now foreign neighboring countries, and to help them compete in foreign
20
Compared to present or formerly state-owned firms, new SMEs are superior in
introducing new technologies and production techniques as discussed above.
Hence, SMEs can improve the competitiveness of large state-owned enterprises.
SMEs can lower large-enterprises’ operation costs by providing them with
competitively priced inputs and by enabling them to take advantage of
outsourcing and subcontracting relations (Scase 2003). SMEs can also play a role
in encouraging state-owned enterprises to restructure (McMillan and Woodruff
2002). Without political backing, in most cases, state enterprises cannot compete
successfully with private SMEs unless they restructure. After reforming, those
enterprises that are still unable to perform efficiently and to compete have to shut
down. SMEs can thus improve allocative efficiency by actively competing with
state-owned enterprises and forcing the unsustainable ones to exit.
When state-owned enterprises restructure or shut down, they may release the
excess capital and workers that they were employing as a result of inefficient,
politically determined objectives. One of the most important roles of SMEs is to
alleviate the problems of rising unemployment by creating new employment
opportunities. Thus, the role of SMEs in job creation is expected to play an
integral role in economic restructuring and in providing the impetus for growth
after transition is completed. This role also has significant political ramifications.
Reducing unemployment relieves political pressure from the disgruntled
state-sector employees, and it lowers government expenditures in terms of
unemployment assistance, early retirement payments, and other social programs
(McIntyre 2003). Tax revenues from the SME sector also significantly augment
the state and local budgets of transition economies.
markets with other suppliers of frozen fruits and vegetables. These SMEs serve as agents that
represent the domestic producers in foreign markets (Božović 2006a).
21
In general, the SME sector is expected to ameliorate negative consequences of
adjustments that must occur during transition. Also, they are expected to
accelerate transition reforms. These transition-specific roles are combined with
the roles that SMEs play in developed economies such as job creation and
innovation. As a result, a growing SME sector is at the heart of the economic
transformation that began in the post-communist world and is expected to lay the
foundation for self-sustained growth in the future.
2.5 SMEs in South-Eastern Europe
Given the importance of SMEs to an economy in transition, it is necessary to
review their performance since the beginning of transition in the early 1990’s.
Table 2.2 provides a snapshot of the size of the SME sector in South-Eastern
Europe (hereafter SEE), which is the focus of this study. The table also offers a
comparison with the more successful reformers of Central-Eastern Europe and the
European Union (EU). The share of SMEs in the total number of enterprises in
SEE is comparable to that of other regions and Western Europe. However, Balkan
countries, including Croatia, have a significantly smaller number of SMEs
compared to the economies of Central Europe and the EU. This is also reflected
in the statistics measuring the number of SMEs per 1000 inhabitants, which show
that all SEE economies have less than nearly a third of the SMEs than the
countries of Central Europe (Table 2.2).
The above figures are even more alarming if we take into account that
officially available statistics in some transition economies do not offer the most
accurate account of the number of SMEs operating in a country. Studies have
reported significant discrepancies between the number of firms registered in
22
Table 2.2: SME
1
sector in South-Eastern Europe (SEE)
Number of
SMEs
2
Share of SMEs
in total number
of enterprises
(%)
2
Share of micro
enterprises in
total number of
SMEs (%)
2
SMEs per
1000
inhabitants
2
Share of SME
employment in
total employment
(%)
3
Index of SME
Development
3
SEE
Albania 56,237 99.55 96.28 18.1 58.4 274.2
Bosnia and Herzegovina 30,000 99.34 85.33 7.0 53.0 118.2
Bulgaria 224,211 99.67 92.61 27.6 42.2 188.4
Croatia 63,135 99.33 66.51 13.7 65.0 1820.9
FYR Macedonia 27,938 99.31 93.01 14.0 65.1 306.1
Moldova 22,138 89.68 80.22 6.1 21.6 11.2
Romania 612,862 99.68 50.79 27.4 20.8 155.5
Serbia and Montenegro 66,968 98.90 95.57 7.8 32.4 156.0
Select CEB countries
Czech Republic 876,990 99.81 94.71 85.1 56.7 1000.0
Hungary 858,981 99.88 96.37 85.9 56.8 1103.5
Poland 3,368,367 99.80 95.19 87.0 68.1 1222.3
Slovak Republic 365,783 99.96 96.88 67.7 66.0 1053.1
Slovenia 26,915 98.87 82.80 13.5
5
62.6 2534.9
EUplus = EU19
4
20,415,000 99.80 93.26 52.8 — —
1
SME refers to non-farm small and medium size enterprises with 250 or fewer employees.
2
Source: Sanfey et al. (2004). Data for Albania, Croatia, Czech Republic and Hungary are from 2002. Data for Bosnia and Herzegovina, Moldova, Poland,
Serbia and Montenegro, Slovak Republic and Slovenia are from 2001. Data for Bulgaria, FYR Macedonia, Romania and EUplus are for 2000.
3
Source: UNECE (2006). Data is from 2002.
4
EU member countries plus Iceland, Liechtenstein, Norway, and Switzerland.
5
This number has been listed much higher in other sources. For example, Kozak (2007) lists it at 45.6% for 2003.
23
official commercial court registers and the number of firms actually operating.
Sometimes, to attract foreign investors, statistics are presented to inflate the
vitality of the SME sector by exaggerating the rate of new firm formation while
not taking into account SME death rates (McIntyre 2003). Further, many of the
registered firms are not actually operating because they were set up to collect
state subsidies, transfer funds out of state budgets, or facilitate other forms of
money laundering. Others have simply shut down. Thus, their existence is
reflected in court registers but they are not operating, at least not in any way that
contributes to economic growth. Some studies estimate that the number of these
dormant firms can be as high as one third of the total number of registered SMEs
(Glinkina 2003). The data for the period 1990–1999 in Serbia indicate that only
somewhere between 31–40% of firms actually recorded some activity in their
accounts and filed annual financial statements with the Bureau for Budgets and
Payments (Bolčić 2002, 113, Table 5).
9
Similarly, in Montenegro in 2000, 14,350
small businesses were registered, of which only 38% were actually operating
(Vukčević 2005). Therefore, the already low number of SEE SMEs should be
interpreted with caution as it is most likely an overestimate.
The United Nations Economic Commission for Europe (UNECE) publishes an
index of SME development, that is a combination of indicators measuring the
share of private ownership in the economy, the share of the SME sector in the
nation’s GDP, and the share of the labor force in the SME sector as a percentage
of the total labor force. This index is reported for select countries in column 6 of
Table 2.2 where a higher value of the index indicates better performance. Other
than Croatia, all SEE countries have an index at around 300 or below while the
9
This bureau (Zavod za Obračun i Plaćanje) was abolished both in Serbia and in Montenegro
in 2004.
24
Central European economies are all above 1000. For comparison, the value of the
index for advanced emerging market economies is between 500 and 2200 (UNECE
2006).
We noted above that in developed economies SMEs play a key role in creating
new jobs. The size of the SME sector as a whole does not necessarily impact total
employment due to the high entry and exit rate of SMEs in the market.
Moreover, the share of employment associated with the SME sector can be
deceiving if we do not simultaneously take into account changes in large sector
employment and the overall trend in unemployment. For example, in Serbia
between 1991 and 1998, total employment in the large sector fell more than it rose
in the SME sector. As a result, the rise in the SME share of total employment is a
result of the rising unemployment (Božović 2006a). Thus, unless the SME sector
is growing enough to absorb the workforce released from large enterprises
undergoing restructuring, the share of SME employment in the total employment
is not an instructive measure of the SME sector’s contribution to generating jobs.
Earlier we noted that there is an inverse relationship between the size of the
firm as measured by the number of employees, and the rate of firm growth. This
counterintuitive relationship is due to the fact that total employment grows when
established SMEs expand their activities and as a result hire more workers. Thus,
to understand the impact of the SME sector on job creation we have to look at
the growth of individual firms. Unfortunately, data indicating expansion of SMEs
in SEE and the increase in their employment is scarce. Also, due to the
incompatibility of size definitions and the inaccuracy of the data, we do not have
a good picture of the within-country distribution of SMEs in terms of their size
(micro, small, and medium). Some general assessments of the overall composition
25
of the sector and the proclivity of SMEs to grow can, however, be found in the
literature. First, studies have concluded that the number of micro enterprises is
significantly larger in transition economies compared to the EU, which could
explain the sector’s lackluster demand for additional labor (Hull 1999). Also, the
average number of employees per firm has in some cases declined. Such a decline
can be interpreted either as a statistical outcome of a large entry of micro
businesses, or as evidence of the inability of small firms to grow (Franičević and
Bartlett 2002). Moreover, transition economies have observed a large increase in
self-employment. Compared to the OECD average of 11%, the self-employment
rate of 20% recorded in transition economies is considered high. This is an
indication that many businesses are formed by unemployed individuals in order to
take advantage of short-term arbitrage opportunities, or to sell a few assets and
services as a means for survival. In other words, individuals are looking for
alternative sources of income and not opportunities with long-term investment
potential. This is in line with the “Birmingham” model of firm formation that sees
the rise in self-employment as a response to rising unemployment (Smallbone and
Welter 2001; Bartlett and Hoggett 1998). The resulting distribution of SMEs in
SEE is concentrated at the micro-level, whereas a much smoother distribution
within the EU indicates a growth over time of micro businesses into small and
medium SMEs (Kolodko 2003). Regardless of the willingness of entrepreneurs to
expand their business, a high percentage of micro enterprises in the SME sector is
an indication that individual businesses are not successfully growing and suggests
that SMEs in transition face obstacles to growth.
26
Table 2.3: Structure of SMEs by type of activity (percent of all
SMEs)
1
Construction and
manufacturing Trade Services Other activities
SEE
Albania 50.0 22.0 17.0 11.0
Bosnia and Herzegovina 50.0 5.0 30.0 15.0
Bulgaria
2
50.8 28.9 18.0 2.3
Croatia 43.7 39.1 23.9 2.1
Moldova
2
48.6 28.5 19.0 3.9
Romania
3
50.2 16.9 19.4 8.6
Serbia and Montenegro
2
47.5 5.3 25.8 21.4
Select CEB countries
Czech Republic 33.7 21.5 24.2 20.7
Hungary 24.7 52.1 19.8 3.4
Poland
3
36.0 37.2 22.6 4.1
Slovak Republic 39.5 32.0 21.9 5.5
Slovenia 20.9 42.0 29.8 6.5
1
Source: UNECE (2006). The data is from 2003. Some of the rows add up to values
not equal to 100. The source of these errors is not known, but they have probably
resulted from the multiplicity of sources, including years and definitions, used to
generate the dataset on which UNECE (2006) reports.
2
ThedatawerenotavailableinwholeorinpartwithinUNECE(2006), sodatafrom
Sanfey et al. (2004) is reported instead. This data is for 2001.
3
The data is from 2002.
The size of the SME sector in SEE is considerably smaller than in successful
Central European transition economies and SMEs that do emerge are unable to
grow. The type of the SMEs that have been established reveals a poor
composition of the sector as well. As in most transition economies, the majority
of SEE SMEs sprung up in the trade and services sectors (Kolodko 2003;
Pfirrmann 2002). These enterprises were established to exploit the demand for
previously unavailable goods and especially the imports from foreign companies
that were made available as a result of trade liberalization. They also filled the
demand for professional services such as accounting, marketing, and advertising
that were not previously needed and were now employed by all sectors, including
the emerging SME sector for its own needs. The slowest rate of SME entry in
transition economies has been observed in the manufacturing sector (Kolodko
27
2003; Pfirrmann 2002). Economies of scale are not uniform across sectors and
some sectors are naturally going to attract more firm entry than others. However,
in SEE, the size of the trade sector significantly exceeds the size of the
manufacturing, construction and service sectors combined (Table 2.3). In the
advanced transition economies, such as the Central European and Baltic
countries, the situation is reversed: combined, manufacturing, construction and
service SMEs combined far outdo the number of SMEs in the trade sector. The
initial distribution of SMEs across sectors could be explained by SMEs’ incentives
to operate in areas with low investment and labor requirements, high turnover,
and low levels of risk and uncertainty. However, the persistence of a SME
distribution which is heavily skewed toward the trade sector indicates that these
underlying conditions have not improved significantly and are affecting either the
incentives of SMEs to emerge or opportunities for them to succeed in higher
value-added sectors.
Analyzing the evidence from SEE, it is clear that economic reality has not met
the expectations of economic theory. The idea that small businesses will harness
entrepreneurial ambitions pent up by a centrally planned system, and that they
will mushroom in numbers sufficient to overtake the post-transition recession in
output and employment, has been, at minimum, optimistic. In SEE, the SME
sector has sprung up but the number of SMEs is still lagging compared to other
successful transition economies and the EU. The relatively low number of SEE
SMEs overall and their slow expansion produced a sector which cannot
adequately contribute to lowering unemployment. Numerous micro enterprises
have concentrated in retail trade, which offers limited potential for self-sustained
growth compared to successful reformers in Central Europe. Many of these micro
28
enterprises are established to provide survival income and have unclear potential
for vertical and horizontal integration into productive “systems” of enterprises. To
unearth some of the reasons for the slow development of the SME sector in SEE
we must examine the factors that have constrained the entry of new firms or the
expansion of already established SMEs.
2.6 Limits to SME Growth
The literature addressing the SME sector performance has consistently identified
two critical elements of the business environment that are necessary for SME
growth in transition economies.
10
The first is the need for businesses to have
access to sources of finance. The second element singled out in transition
literature is the rule of law.
2.6.1 Access to Finance
Problems with access to credit and high credit costs are often cited as the main
culprits for the poor performance record of SMEs in SEE. The credit markets
have improved as many foreign banks have been allowed to set up branches in
transition economies. Nevertheless, the costs of obtaining loans have fallen only
marginally. Most of the credit is available in the form of short-term loans which
are not optimal for long-term projects that businesses need to finance. Moreover,
these loans come with high collateral requirements and high interest rates.
Table 2.4 shows that interest rates in most of SEE exceed 10%, going as high as
10
The following provide an overview of this literature: Hull (1999), Pissarides, Singer, and
Svejnar (2000), Bartlett, Bateman, and Vehovec (2002), Glinkina (2003), McIntyre (2003),
Radaev (2003), and Cengic (2003).
29
26% in Romania. Foreign and domestic banks face a high risk of default in these
economies and to ensure themselves against risk they require high collateral (200
to 300%) and charge high processing fees (Glinkina 2003; Radaev 2003). The
survey of SMEs in Serbia and Montenegro used later in this work indicates that
on average, 78.29% of those who took out a loan had to provide collateral. The
average rate of the collateral reported for the sample is 177.60% and it ranged
from 30 to 300%.
11
Table2.4: Short-terminterestratesinselectedSouth-EasternEuropeaneconomies,
2002–2004
1
Nominal interest rates on
short-term credits (per cent)
Total credit
2
(per
cent of GDP)
2002 2003 2004
3
2002 2003 2004
3
Albania 15.3 14.3 12.1 6.3 6.6 7.8
Bosnia and Herzegovina 12.7 10.9 10.4 32.7 39.2 43.0
Bulgaria 9.8 9.2 9.1 16.5 22.8 28.6
Croatia 13.0 11.6 11.8 47.1 54.6 56.1
Romania 35.4 25.7 26.3 10.7 12.5 13.8
Serbia and Montenegro 24.3 16.5 15.9 12.0 15.9 14.8
The former Yugoslav
Republic of Macedonia
18.3 16.0 12.6 16.2 16.4 17.9
Turkey — — — 13.5 13.8 15.3
1
Source: UNECE (2005).
2
Total outstanding claims of commercial banks on the non-government sector. GDP data for 2004 are prelim-
inary estimates.
3
January–September.
The type of collateral accepted by banks is most often made up of real estate.
These requirements are especially difficult to meet for businesses that are looking
to expand their production as they may not have sufficient real estate to provide
as collateral. The high costs of credit and stringent collateral requirements
explain why commercial loans cover a relatively small percentage of investment
needs for SMEs in SEE (this figure ranges from the 21% in Croatia to the 4% in
11
The dataset is described in more detail in Chapter 5.
30
Albania) (Sanfey et al. 2004). As a result, internal funds have been the biggest
source of finance in transition (McMillan and Woodruff 2002). In SEE, internal
funds cover between 50–86% of financing needs for SMEs, which is a large
percentage even if we take into account that SMEs everywhere often find it
difficult to secure external sources of finance (Sanfey et al. 2004). Alternatively,
SMEs can borrow from friends, family and even informal and illegal markets,
which does not necessarily imply superior terms of credit. The inability to access
credit under favorable conditions, and the resulting limitations placed on the
ability of businesses to plan certain investments, directly impacts the sector’s
overall performance.
Alternative sources of capital, namely equity financing, are not well developed
in transition economies, and especially not in SEE. In 2002, the share of equity
capital in financing the investment needs of SMEs in SEE was recorded at 1%
(Sanfey et al. 2004). While all transition economies have functioning stock
markets, they have been slow in generating significant volumes of activity and a
noticeable supply of capital (Sanfey et al. 2004). An important body of work
examining the determinants of financial market development has indicated that
the extent and effectiveness of financial markets is determined by the ability of
legal rules to protect investors from corporate insiders (La Porta et al. 1997;
Demirgüç-Kunt and Maksimovic 1998). Similar results have been found with
reference to countries in transition (Pistor, Raiser, and Gelfer 2000). Moreover,
this work has stressed that the quality of legal enforcement mechanisms is
sometimes divorced from the extensiveness and efficiency of the law on the books.
In fact, legality, or the extent to which the rule of law is implemented in practice,
was found to be a better determinant of the financial sectors’ development than
31
the quality of the law protecting insiders against informational asymmetries and
conflict of interest. This suggests that the first critical obstacle to SME sector
development in SEE, or the inability of businesses to rely on external sources of
finance, is largely a result of the lack of legal enforcement necessary to support
the development of the financial sector, especially the debt and equity markets.
As a result, the first obstacle is closely related to the second obstacle for the SME
sector’s growth introduced above, namely the rule of law.
2.6.2 Rule of Law
Inadequacy of the rule of law in SEE stems from two separate factors: the
frequently changing legal code and the disconnect between the law on the books
and its implementation in practice. The second aspect of this problem is more
critical and is rooted in the ineffectiveness of the judiciary systems across Eastern
Europe. The Law in Transition Report, published biannually by the European
Bank for Reconstruction and Development (EBRD), shows that in terms of 1)
quality, education and diversity of judicial qualifications, 2) judicial powers, 3)
financial resources, 4) structural safeguards, 5) accountability and transparency,
and 6) efficiency, judiciaries and judicial training structures across SEE are in
need of serious reform (EBRD 2005). The EBRD’s Legal Indicators Survey has
shown that in terms of effectiveness of insolvency laws, local courts in SEE
perform worse than those in Central Europe, where Estonia, Hungary and
Lithuania scored the highest (EBRD 2005). SEE economies such as Albania,
Bosnia and Herzegovina, Moldova, Bulgaria, and Serbia and Montenegro showed
the worst performance in terms of court efficiency.
32
Individual assessments of the courts’ efficiency in enforcing the rule of law can
be observed from the Business Environment and Enterprise Performance Survey
(BEEPS) developed by the World Bank and the EBRD. Table 2.5 offers a glimpse
into the results of this survey and shows that the popular belief in the
effectiveness of courts in SEE is rather low. For questions about the courts’
fairness and impartiality, honesty, and ability to enforce decisions, answers in the
categories “never” and “seldom” provide the worst level of assessment. The
percentage of respondents who answer in either of these two categories is higher in
SEE than in countries such as Slovenia, Poland and Hungary, for example.
Although there is variation in terms of how badly SEE countries score in each of
the three dimensions, the table shows that in SEE, courts are not believed to be
very impartial, fair, honest, uncorrupt, and able to enforce decisions. Interviews
with SME managers and owners in Serbia and Montenegro have shown that
courts are poorly evaluated both in terms of their effectiveness and the overall
cost of dispute resolution.
12
Courts are too slow in processing and deciding cases,
which increases the cost of administrative and legal assistance. In addition, the
general impression that judges and other court officials are vulnerable to influence
from the parties involved in disputes, amplifies the public’s distrust of the courts.
The distrust in the ability of courts to enforce the rule of law generates
disincentives for the use of courts in contract enforcement. The ability to enforce
contracts efficiently is the key prerequisite for business growth since without
effective forms of contract enforcement, businesses cannot insure transactions and
investments against those who renege on their contracts. In addition, the absence
of efficient forms of contract enforcement can encourage actors to cheat because
12
Interviews with SME managers and owners are discussed in Chapter 5 and Appendix A.
33
they expect others not to pursue legal action. Hence, the inability to enforce
contracts can have a paralyzing effect on economic transactions.
2.6.3 Private vs. Public Order
Surely, transactions take place in transition economies, including SEE, despite the
poor ability of the courts to enforce insolvency, bankruptcy, and other commercial
laws. Mutually beneficial exchanges take place because going to court is not the
only option businesses have for resolving breaches of contract. Even in countries
with well-developed legal systems and highly efficient courts, businesses often rely
on informal means of enforcing contracts in order to expedite conflict resolution
and produce transactional assurance (Macaulay 1963; Richman 2004). Private
order literature has argued that when formal procedures and rule of law fail to
provide effective forms of contract enforcement, other self-enforcement
mechanisms can lower the incentives for cheating on contracts by altering the
consequences of one’s actions (Hadfield 2004). Others have empirically
demonstrated that a simple reputation mechanism can sustain an alternative form
of contract enforcement (Ellickson 1991; Landa 1994). Actors are aware that if
they cheat in a transaction their actions will become known to other members of
the community. Those who wish to benefit from future transactions have an
incentive to abide by the terms of their agreements in order to protect their
reputation. Robert Ellickson (1991) vividly illustrates this mechanism at work in
a ranching community of Shasta County, California, where residents are shown to
prefer self-enforcing, informal rules for settling disputes to written and publicly
enforced law. Individuals use gossip as a social sanctioning mechanism to spread
the word about unfairly behaving actors. They also keep mental accounts of
34
Table 2.5: Confidence in courts
Courts are fair and
impartial Never Seldom Sometimes Frequently Usually Always
Albania 15.85 27.44 41.46 7.32 5.49 2.44
Bosnia and Herzegovina 11.92 29.80 33.77 10.60 11.92 1.99
Bulgaria 20.43 27.83 25.65 8.26 12.17 5.65
Croatia 13.45 27.49 33.92 11.70 9.94 3.51
Czech Republic 11.30 25.10 36.82 12.13 11.72 2.93
Macedonia 19.50 25.79 25.79 8.18 10.69 10.06
Hungary 14.03 18.55 12.22 22.17 23.08 9.95
Moldova 18.82 27.06 38.24 8.24 7.06 0.59
Poland 8.21 24.21 35.79 14.11 14.95 2.74
Romania 13.20 24.80 32.40 12.40 13.60 3.60
Slovakia 6.67 28.00 39.33 9.33 12.67 4.00
Slovenia 6.82 25.00 25.57 26.70 11.36 4.55
Serbia and Montenegro 12.14 26.21 30.10 11.17 14.56 5.83
Courts are honest and
uncorrupted Never Seldom Sometimes Frequently Usually Always
Albania 23.46 30.86 32.72 6.79 4.32 1.85
Bosnia and Herzegovina 17.81 28.77 28.08 13.01 9.59 2.74
Bulgaria 23.45 25.22 28.32 7.52 10.62 4.87
Croatia 12.05 29.52 30.72 12.65 10.84 4.22
Czech Republic 16.18 27.39 30.71 11.20 10.79 3.73
Macedonia 24.05 20.89 28.48 9.49 7.59 9.49
Hungary 12.27 10.45 21.36 25.00 21.36 9.55
Moldova 15.20 40.35 25.73 12.87 4.68 1.17
Poland 7.44 28.45 33.92 14.22 13.57 2.41
Romania 19.58 27.92 23.33 10.00 13.75 5.42
Slovakia 10.00 27.33 37.33 12.00 10.00 3.33
Slovenia 7.98 23.31 22.09 30.67 11.04 4.91
Serbia and Montenegro 12.31 27.81 28.21 10.26 13.33 8.72
Courts are able to enforce
decisions Never Seldom Sometimes Frequently Usually Always
Albania 11.11 21.57 30.07 22.22 9.80 5.23
Bosnia and Herzegovina 14.19 30.41 29.05 10.14 13.51 2.70
Bulgaria 10.26 14.96 23.93 11.97 18.38 20.51
Croatia 9.25 23.13 28.90 17.34 11.56 9.83
Czech Republic 10.97 21.94 32.49 12.66 15.61 6.33
Macedonia 18.12 23.12 30.00 8.12 8.75 11.88
Hungary 16.20 11.57 21.30 21.76 24.07 5.09
Moldova 10.78 28.14 32.93 17.37 10.18 0.60
Poland 10.04 29.49 31.84 13.25 10.68 4.70
Romania 10.40 22.80 26.00 14.40 17.20 9.20
Slovakia 6.00 27.33 35.33 8.67 12.00 10.67
Slovenia 8.28 23.08 21.89 22.49 12.43 11.83
Serbia and Montenegro 9.41 23.76 29.21 11.39 12.38 13.86
Source: Business Environment and Enterprise Performance Survey 2002 (World Bank). This data was
collected from small and large businesses.
35
violations, which can be offset by future accidents or exchanges in goods and
services. Ultimately, the effectiveness of self-enforcing rules results in wealth
maximizing benefits, argues Ellickson, because residents have a vested interest in
maintaining relationships with their neighbors. Similar studies have shown that
private forms of contract enforcement can reliably sustain exchange when effective
third-party enforcement provided by state and the courts is lacking (McMillan
and Woodruff 2000).
However, some have warned that private forms of contract enforcement have
only limited relevance as a “general social technology” for governing transactions
(Eggertsson 2005). For the reputation mechanism to work, community boundaries
have to be clearly identified so that the sanctioning mechanisms, such as
expulsion form the group, can have credible consequence for cheaters. As a result,
it is argued that the reputation mechanism is restricted to groups based on
geographic location, ethnic orientation, religious affiliation, or other membership
criteria. As the economy develops and the needs of market actors become more
specialized, trade can no longer be supported by the reputation mechanism
operating within groups. Avinash Dixit (2004) offers a formal model to illustrate
the above limitation by showing that as the distance between actors increases
(socially, economically, and geographically), the probability that instances of
cheating become known decreases, causing the informal system of exchange
governance to break down. This model, however, offers little explanation of the
parameters that determines the rate at which actors’ cheating becomes known to
others (Rauch 2005). Without illustrating the system for knowledge and
information distribution, Dixit’s model seems to reproduce the elemental
assumption that information distribution is in some way restricted to the
36
members of the group. To illustrate the weakness of Dixit’s model, James Rauch
(2005) brings up China as a counter-example which shows that significant
presence of informal governance does exist even in the world’s second largest
economy. This raises the question, as Rauch (2005) observes, of “how can the
decay of information be so attenuated that effective informal governance can be
sustained in an economy as vast as China’s” (484)?
Rauch’s counterexample raises the possibility that informal social networks,
which as a form of social capital play an important role in post-communist
transition, are in fact providing a system of information distribution that can
sustain the self-enforced, private forms of contract enforcement more broadly. In
other words, a reputation mechanism does not have to be restricted to the groups
defined by religious affiliation, ethnic identity or geographic region. The
prevalence of informal exchange relations in facilitating commercial exchanges
prior to and during post-communist transition suggests that such a reputation
mechanism can sustain transactions on a much broader scale by using personal
relationships for information transfer across various group memberships. In other
words, it is possible for personal crosscutting relationships of various origins to
allow information to be transmitted beyond a single group. Personal relations
forming the structure of informal social networks can in this manner provide a
legitimate alternative form of contract enforcement applicable beyond a single
geographic location or ethnic community. In this case, the degree to which private
order can be sustained in a manner that positively impacts economic performance
is determined by the nature, structure, and role of informal exchange through
social networks. This demands that we turn our attention to social networks as
the root of social capital described by transition scholars and take a closer look
37
into their structure and impact on economic performance, specifically with respect
to the needs of the SME sector.
This chapter has argued that success in economic transition depends on a large
SME sector populated by individual businesses that are successfully expanding.
Such an SME sector represents a dynamic component of the economy which
contributes to innovation and job growth. The absence of a vibrant and growing
SME sector signals significant obstacles to SME development and economic
growth more broadly. Economic literature has already identified two critical
obstacles to the growth of individual enterprises, namely the problems with access
to and cost of finance, and the rule of law. The two obstacles, however, are closely
interrelated since the rule of law determines the development of and the ability of
businesses to access external sources of finance for their investments. Hence, the
rule of law, and specifically contract enforcement, appears to present a key
obstacle to SME growth. The alternative private order mechanisms for contract
enforcement are greatly supported by the availability of social networks for
distribution of critical business-related information. Therein, the infrastructure of
informal social relations is used to compensate for the lack of effective forms of
formal contract enforcement. Thus, it is necessary to explore whether the informal
forms of contract enforcement facilitated by social networks, as key elements of
social capital, positively impact business performance and growth. In other words,
we need to examine to what extent social capital and social networks can
substitute for the lack of formal rule of law and thus contribute to SME growth.
38
Chapter 3
Social Capital: Concepts and
Measures
Just as scholars in the 1980s struggled to peek into the “black box” of economic
growth by examining human knowledge and investment in human capital, so is
the goal today, as Mark points out, to make “networks, norms, and institutions,
history and culture fully endogenous to economic models” (2005, 47). The field of
social capital has partially contributed to these efforts. The objective of the
scholarship produced under the banner of social capital is to explain how the
informal elements of political and economic institutions, such as culture, values,
norms, trust, and social relationships translate into distinct advantages in terms
of political and economic well-being.
This chapter reviews the broad field of social capital and its utility in studying
the impact of informal institutions that govern social exchange on the
performance of the SME sector. Further, the chapter voices a sense of
disappointment with the ability of the attitudinal approaches to social capital to
provide explanations of how informal elements of social exchange impact economic
39
outcomes. Greater potential for linking social capital and economic outcomes with
causal explanations is offered by the alternative structural approaches to social
capital. The chapter offers a review of these two approaches and shows that the
latter is associated with better conceptualization, operationalization and
measurement techniques. Moreover, the structural approach to social capital
based on the study of social networks can benefit from the methods of social
network analysis already largely developed in sociology.
3.1 The Social in Social Capital
Inevitably, all scholarship in this field can be traced to one of the three definitions
of the central concept. The first is attributed to the sociologist Pierre Bourdieu
who is also sometimes credited with coining the term social capital. Bourdieu
wanted to understand how the interaction of physical, cultural and human capital,
as well as their different distributions, perpetuate social inequalities. He defined
social capital as “the sum of resources, actual or virtual, that accrue to an
individual or a group by virtue of possessing a durable network of more or less
institutionalized relationships of mutual acquaintance and recognition” (Bordieu
1983, 249). The second elemental definition of social capital is found in the work
of James Coleman. In examining the role of social capital in the production of
human capital, Coleman defined social capital “not as a single entity, but a variety
of different entities having two characteristics in common: they all consist of some
aspect of social structure, and they facilitate certain actions of actors—whether
persons or corporate actors—within the structure” (1988, S98). Coleman’s view of
social capital was less symbolic and more normative than Bourdieu’s, which is
fitting given his effort toward producing a rational choice theory of sociology.
40
However, Coleman shared with Bourdieu the idea that social capital results from
interactions among individuals. This sets them both apart from Robert Putnam,
the source of the third pivotal definition in the field. In his study of the
divergence in the quality of governance between Southern and Northern Italy,
Putnam (1993) defined social capital as the “features of social organization, such
as trust, norms and networks, that can improve the efficiency of society by
facilitating coordinated action” (167). Putnam’s work on the decline of social
capital in the United States probably did the most to publicize the concept across
the academic disciplines and beyond, but he also contributed the broadest and
most vague definition of the term, which has invited others to use terms such as
social trust, norms of reciprocity, and civic engagement interchangeably.
Whichever definition is adopted, at the center of the concept of social capital
is the emphasis on its roots in the social environment which individuals inhabit.
The emphasis is always on the positive outcomes that individuals can derive from
their interactions with other members of the community, which is not necessarily
geographically defined. Not surprisingly, since it was introduced by sociologists,
the concept is infused with the sociological perspective that individuals are not
autonomous units under analysis but that they are both a product of their social
or cultural environment and are influenced by that structure in the way they
perceive the world, process information, and make their decisions. To put it
succintly, the social in social capital suggests a complex and interdependent form
of capital whose value is derived from interactions between individuals and is not
necessarily a result of deliberate accumulation.
1
1
Even Coleman, who wrote from the rational choice perspective, was not able to explain the
self-interest motive in accumulating or generating social capital. He explained its accumulation
as a by-product from pursuance of other activities, according to Field (2003).
41
3.2 The Capital in Social Capital
While there is very little disagreement about the social roots of social capital, a
lot less is agreed upon when it comes to its conception as capital. Some see it as
any other form of capital, though less tangible, that individuals can use to their
own advantage. Physical capital is most tangible because it includes products and
resources. Human capital constituted by indirectly observable skills and
knowledge possessed by actors themselves is much less tangible. Social capital
“inheres in the structure of relations between actors” and is therefore least
tangible (Coleman 1988, S98). Nevertheless, like all other forms of capital, it
facilitates productive activity. For example, when in a bind, an individual can
receive assistance from close family members, friends, and coworkers. The fewer
and more limited relations one possesses, the more limited his or her social
capital, and therefore the lower the chances that they will resolve a problem
successfully. Social capital literature, however, is not clear upon whether some
individuals are economically advantageous because they are more trustworthy or
trusting of others, have more resourceful relationships, belong to somehow better
networks, or join various civic organizations. Social capital scholars do not
identify clearly the sources of social capital nor the forms and functions in which
social capital can be observed.
The root of this problem lies in the multitude of definitions adopted to
examine social capital. For one, it is not clear whether the value of this capital
somehow resides under individual control, and for individual benefit, or if its value
is an attribute of the group and thus generated through civic engagement, trust,
or other outcomes of collective action. The distinction between the private and
public good attributes of social capital remains unclear. Some see it as a value of
42
the resources that one can access through social ties (Bordieu and Wacquant
1992; Burt 2000; Lin 2001). In that case, the formation of social ties or links can
be seen as any other form of investment based on the revenue-generating principle
(Westlund and Nilsson 2005). Coleman’s perspective is close to this definition
since he observed an instrumental element of social capital in its ability to
generate goods and services for individuals and groups. However, his definition
could not reconcile the rational choice aspects of the individual investment
decision and the collective aspects of social capital (Field 2003). Since social
capital is a resource that resides in social structure, it is not clear why
self-interested individuals would make investments that increase socially available
capital. As a result, its accumulation was explained as a byproduct and not as a
result of the deliberate actions of individuals. For others still, social capital is
clearly a public good since it produces positive externalities such as reciprocity,
civic engagement and norms of cooperation that result in collective action
(Putnam 1993).
In addition to the distinction between the public and private nature of capital,
an added difficulty in conceptualizing social capital as a form of capital is the
contention that unlike other forms of capital, social capital grows with use and
diminishes with disuse. For example, the value of machinery depreciates during
its productive period, and materials are used up, but trust and norms of
reciprocity that enable successful exchange to take place today can facilitate even
more mutually beneficial transactions in the future (Woolcock 1998, n.11). A
corollary is offered by Putnam (2000) who suggests that with the reduction in the
use of social ties and increased isolation, social capital diminishes and so do civic
engagement and associational activities that result in collective action.
43
Whether taken to represent the benefits that individuals derive from their
social ties for individual use or public goods that produce collective action, at the
aggregate level, social capital is thought to translate into specific advantages.
These advantages translate into better economic performance manifested in a
better standard of living and higher economic growth. Alternatively, social capital
is associated with a more vibrant civil society, better political institutions,
improved welfare, and overall a happier society. This is, unfortunately, where the
metaphor of capital reaches its theoretical limits. Good explanations of the
process whereby the elements of social capital result in the advantages described
have been largely absent. Overall, the metaphor has simply remained a metaphor
and has not generated a robust explanatory variable. The reviewed problems with
conceptualizing social capital are partly responsible. However, problems with
operationalization and measurement have also contributed to this deficiency.
3.3 Operationalizing Social Capital
3.3.1 Attitudes: Norms and Trust
Social capital literature has adopted two broad approaches when it comes to
operationalizing and measuring social capital. The first approach centers on
attitudinal elements of social capital such as norms and trust. Norms are socially
imposed constraints on one’s behavior. The individual will follow social norms
even when they are not particularly useful because he or she wishes to be accepted
by the group or fears that the violation of the norm will initiate sanctions from
the group. Despite the fact that the individual alone does not decide which norms
are adopted by the society in general, he or she nevertheless makes the choice to
44
follow or disobey those norms. In that sense, the degree to which a specific norm
is followed, such as the norm of reciprocity, is an individual choice.
The focus on norms as social capital is supported by numerous empirical
studies that use survey instruments to measure the acceptance of specific norms of
reciprocity, cooperation, tolerance and overall participation in civic life. To
measure these elements, following the example of Putnam’s empirical work,
scholars use proxies, of which the most popular is membership in voluntary groups
and associations.
2
Such indicators of social capital are then related to a country’s
political and economic performance. The most significant shortfall of this
approach is the tendency to take the measures of individuals’ attitudes and to
aggregate them into national averages in order to derive conclusions about
national political and economic performance. However, differences in the
acceptance of a specific norm within the society are lost when individual attitudes
are aggregated. This practice detaches the measures of individual attitudes from
the underlying social context that is primarily responsible for generating them in
the first place (Foley and Edwards 1999). Also, the use of measures such as
membership in organizations conflates the causes and outcomes of social capital
because the relationship between norm generation and membership in
organizations is not explained (Levi 1996). Organizations and clubs are
responsible for fostering networks of civic engagement which generate norms of
reciprocity that in turn support future acts of collaboration and collective actions.
However, the proposed causal mechanism cannot explain how membership in civic
groups fosters the spread of norms of reciprocity and trust to nonmembers.
2
For indicators of “civic-ness” Putnam (1993) used the vibrancy of associational life,
newspaper readership, electoral turnout, and incidence of preference voting.
45
Neither micro-level theory nor empirical evidence show how and why associational
membership translates into norms of reciprocity and generalized trust.
The concept of trust takes center stage in the literature that examines social
capital in the transition economies of Eastern Europe. On the one hand, due to
socialist legacies of authoritarian government and repressed civil society, the
literature has argued that these economies are cursed with high levels of distrust
that has resulted in a low level of social capital, which in turn negatively impacts
the process of transition (Raiser et al. 2001). Others see the presence of informal
exchange networks as contributing to high levels of trust which is a productive
asset that economies in transition can utilize to cope with economic and political
transformation (Kolankiewicz 1996; Sik 1994). The contradictory assessments of
trust present after the collapse of communism result from differing conceptions of
trust employed in the studies.
The concept of generalized or “extended trust” has been introduced as the key
aspect of social capital facilitating economic interactions among individuals who
lack information about one another and encouraging collective action that results
in better governance. Putnam (1993) differentiated between the concepts of
“personal” and “social” (or generalized) trust. In small communities, trust is based
on the information about individuals and the familiarity with their disposition
that one obtains through interpersonal ties. In a more complex setting, a more
indirect form of trust is preferred. This generalized, extended, or “social” trust is
impersonal and, unlike interpersonal trust formed within networks, it is not based
on information about actors but on an inherent and socially conditioned
predisposition to trust anonymous individuals. Generalized trust, according to
social capital theory, allows one to move beyond ties with family members,
46
friends, and acquaintances to forge relationships with complete strangers, which
encourages collective action outside the membership of a specific group.
Fukuyama (1995) went the furthest in terms of the definition of trust and he
simply equated generalized trust with social capital. In his work, he argued that
social capital is a capability that results from the prevalence of trust which
through its impact on industrial structure, and economic institutions more
broadly, impacts economic growth. What these various conceptions of trust
indicate is that social capital literature does not have a solid ground in terms of
the relationship between social capital and trust. While work grounded in
Putnam and Coleman’s conceptualization of social capital views trust as a
component of social capital, Fukuyama simply equates the two.
The focus on trust has been promulgated by the World Values Survey (WVS),
which systematically traces values and beliefs for about 80% of the world’s
population. Numerous studies looking into the relationship between the levels of
trust, or civic participation, and economic development have used WVS for
operationalizing and measuring social capital through the concept of trust (Knack
and Keefer 1997; Raiser 1997; Raiser 1999; Paldam 2000). To measure the extent
of trust in outsiders, the survey uses the following question: “Generally speaking,
would you say that most people can be trusted or that you need to be very careful
in dealing with people?”
The fundamental problem with the concept of extended trust is that the
survey questions used to measure it are misleading. Unfortunately, the question
does not specify who the “people” are that the respondent is asked to refer to in
answering the question. One may consider his or her people to include the closest
group of friends and associates while others may interpret “people” to include
47
randomly encountered strangers.
3
Secondly, the question fails to discount the
influence of third-party contract enforcement institutions. The survey question
does not ensure that the individual, when devising his or her answer, deliberately
ignores the ability to seek recourse in the formal court of law and therefore insure
oneself against potentially damaging commitments to individuals of questionable
reputation. In other words, the response to the survey question does not
disentangle the importance of extended trust from trust in institutions.
The link between personal and generalized trust has also not been established.
Similarly to the problem of specifying how norms of reciprocity extend beyond
group membership, proponents of the concept of generalized trust cannot explain
how personal trust fosters the formation of extended trust. The explanations
offered are incomplete because they do not specify how trust formed among
individuals in a group, who interact with one another repeatedly and are familiar
with each others’ reputations, can be used to form relationships with individuals
outside the group. Also, if extended trust evolves from or cannot exist without
personal trust, then we should observe a significant positive correlation between
indicators for each type of trust. However, data measuring interpersonal trust
show a negative but insignificant correlation between strong reliance on family
and friends and extended trust (Raiser 1999). Other micro-level empirical
analyses have been unable to demonstrate any relationship between group
membership and trust (Levi 1996; Stolle 2003; Rose-Ackerman 2001) Therefore,
3
Similar criticism was raised by Glaeser et al. (2000) and Badescu and Uslaner (2003). To
test the validity of this question, Krishna and Shrader (2002) have included in their surveys
both the general trust question and questions asking respondents to identify the people to whom
they thought the trust question was referring. Responses to these questions varied to such an
extent that in a group of only 42 respondents, three distinct categories of people were identified.
48
the data presented in favor of extended trust demonstrate inadequate
conceptualization and measurement tools for the concept itself.
The unclear relationship between the various types of trust and social capital
has worsened any efforts to explain the relationship between social capital and
trust on the one hand and aggregate economic outcomes or indicators of political
performance on the other. Knack and Keefer (1997) find a positive relationship
between trust and economic growth but find no relationship between economic
performance and associational membership. However, Raiser et al. (2001)
determine that there is no relationship between trust and growth but they do find
that civic participation and trust in public institutions are independently and
positively related to economic growth. This shows that while trust may be
important in facilitating various types of exchanges, there is not enough evidence
to show that larger proportions of people expressing trust has any definitive
implications for the economic growth or health of a democratic society (Foley and
Edwards 1999). The use of social capital measures such as trust and membership
in associations to understand differences in economic and political outcomes at
the macro-level has thus been merely “suggestive rather than conclusive” (Field
2003). These aggregate measures of social capital oversimplify the relationship
with economic outcomes as they discount the large within-country variation
observed empirically.
Studies of social capital that have focused on the concept of trust have not
clearly identified the relationship between between generalized trust and personal
and/or institutional trust; between types of trust and social capital; and between
trust as an indicator or component of social capital and aggregate indicators of
political and economic performance. Extended trust as a concept offers little
49
benefit to understanding the impact of social capital because it provides us with
no clear explanation of its determinants. The concept is vague and at times
subsumes the concept of institutional trust. Most importantly, the concept lacks
theoretical utility because it cannot sufficiently link the identifiable benefits at the
micro-level with indicators of successful macro-level performance. A solution may
lie in operationalizing social capital as a construct that is analytically independent
from trust. The concept of trust is itself a complex phenomenon and incorporating
it into social capital may never result in less problematic operationalizations of the
term, or motivate better measurement techniques. Trust as an analytical concept
has already produced a distinct body of scholarship and attitudinal approaches to
social capital have not been tremendously successful in absorbing this literature to
advance the concept of social capital.
4
Rather than employing the concept of
trust to operationalize social capital, the field must employ concepts that will
consider trust as an independent factor, often itself an outcome of social capital.
The complex and ambiguous operationalization of trust and norms as elements
of social capital prevents us from aggregating their measurements and
incorporating them into economic models as endogenous variables or a shift factor
in the production function. Further, we lack causal explanations of how these
elements affect individual actions and their aggregate outcomes. More
importantly, rather than identifying the correlation between social capital
elements and economic success at the aggregate level, we need social mechanisms
that outline causal relationships and specify the conditions for which they hold
(Hedström and Swedberg 1998). This analytic approach is essential for opening
up the black box that connects norms, relations, trust and culture with economic
4
Examples include works such as Gambetta (1988), Cook (2001), Ensminger (2001), Hardin
(1993), Levi (2000), Sztompka (1999), and Yamagishi (2001) among others.
50
development and growth. It is also an approach through which social capital
literature can move beyond the metaphor and contribute studies that specify clear
and concise determinants, distinct relationships with particular outcomes, and
differentiate between macro-to-micro and micro-to-macro mechanisms. The
second approach to operationalizing social capital is better positioned to advance
social capital scholarship to meet these expectations.
3.3.2 Structures: Networks
The second approach to operationalizing social capital has largely focused on its
role as an enabling structure. Rather than viewing social capital in terms of trust
or norms, this approach focuses on the role of social networks. The central
premise of this approach is rather simple: it is not what you know that matters,
but who you know. This statement summarizes the view that relationships
provide individuals with resources to reach ends they could not reach on their own
or do so more efficiently. The structure formed by a set of relationships that bring
actors together constitutes a network. The network is then a structure within
which actors interact. It influences their beliefs, preferences, and constraints.
Clearly, in this approach an individual actor does not make decisions isolated
from others. Instead, actors are embedded in a system of social relations that
they build in every day encounters.
5
The structural features of the network and
the resources embedded in them are the primary points of study in the network
approach to social capital.
The most important contributions in this tradition are those of Ronald Burt
and Nan Lin. Both of them work with conceptualizations of social capital rooted
5
The concept of embeddedness originated with Granovetter (1985)
51
in Bourdieu’s emphasis on network size and capital possessed by network
members. Burt sees social capital as the “contextual complement to human
capital” (2000, 3). In his study of managerial performance in a U.S. company, he
identifies social capital in terms of “structural holes.” On either side of a
structural hole, groups operate with a distinct set of resources. Individuals who
form relationships across structural holes take advantage of diverse sources of
information and capital and the opportunity to broker relationships between
diverse groups of individuals. The manager thus adds value by spanning the hole,
and the greater the number of structural holes that an individual can span, the
greater the value of his or her social capital. Networks with a higher number of
structural holes result in greater innovation and creativity on the part of the
manager and better job performance evaluations. In Burt’s approach, structure
produces social capital: “Holding a certain position in the structure of these
exchanges can be an asset in its own right” since it can provide better access to
resources (2000, 3).
For his part, Nan Lin emphasizes resources embedded in relationships within
various types of networks. To fully operationalize social capital, Lin (2001) offers
a multi-stage model that illustrates how individuals access and use resources
embedded in social networks to gain returns in instrumental activities. In the first
stage of the model, structural and positional factors affect investments in social
capital. At the second stage, various distributions of social resources are
mobilized. This is where the structural contributions to social capital interact
with individual choice for action. At the last stage of the model, mobilized
resources generate returns. Returns to instrumental or purposive action can be
economic, social or political. In the same vein, Flap (1994) and Portes (1998) also
52
operationalize social capital as a function of size, relationship strength, and
resources found in the network.
Proponents of the attitudinal approach to social capital are critical of the
structuralist perspective because it emphasizes structural aspects of social capital
at the expense of attitudinal components and other factors that can impact the
structure of the networks. This prevents network approaches from recognizing
that not all network structures are capable of producing positive collective action
(Hooghe and Stolle 2003). As a result, proponents of attitudinal approaches argue
against examining social networks as an end in themselves. This criticism,
however, breaks down under closer scrutiny. Lin’s model of social capital, for
example, shows that access to resources in the networks is conditioned by not only
by the structural and positional embeddedness but also collective assets such as
the economy and technology, as well as social, cultural and political participation
(2001, Figure 1). The ability to show that not all network structures or social
capital result in beneficial collective action is in fact a strength of the structural
or network approach to social capital. Understanding which structures result in
positive or negative social capital is especially helpful for advancing the field and
it avoids the tautological tendency of attitudinal approaches to equate all positive
outcomes with causes of social capital. A particularly useful and necessary
research agenda at this stage is to specify the conditions under which specific
network structures generate superior returns (Lin 2001).
The conceptualization of social networks adopted in this work shows that the
structural and attitudinal approaches are not incompatible. To adequately
understand the concept of social capital, we must first recognize that norms held
by individual network members serve as elements upon which social relations are
53
founded or which dictate when relationships will be established. Norms serve to
coordinate expectations about others’ behavior in exchange. Networks are
informally structured social relations that result from repeated interactions among
individuals, not necessarily confined to the political or economic sphere. Social
networks can include extended family, social circles, mafia, and business and
industry networks. Within networks, repeated social interactions foster the
establishment and strengthening of specific norms that influence the relations in
that specific network. Members belonging to a network who want to gain the
respect of network members will follow the conventions and norms lest they face
sanctions and the loss of benefits associated with expulsion from the network.
While social and moral norms are not tied to a specific network, different
networks espouse different sets of norms. Both mafia and business networks can
value the norm of reciprocity. However, business network members place far less
emphasis on the norm of revenge that is highly revered in mafia organizations.
Social networks and their embedded resources also foster the development of
process-based or interpersonal trust. Trust is essential to social networks because
it sustains them at the same time that it is cultivated by them. Members of
networks trust one another to abide by the terms of the agreements into which
they voluntarily enter. This trust is based upon repeated exchange and shared
expectations of behavior, strengthened by accepted norms. This community of
individuals who trust one another in social relations is what constitutes a social
network.
This conceptualization shows that norms can be found at the basis of
exchange relations that form networks that are central to social capital in the
structuralist perspective. Trust, on the other hand, can be seen as an outcome
54
and thus conditioned by network structure. In this operationalization, the
structuralist approach that centers on networks and resources still allows us to
consider attitudinal elements as closely related concepts, but provides a clear
explanation of how they relate to social capital. This perspective on social capital
also enables us to operationalize the term independently of trust and its
outcomes. The discussion of problems that arise when trust alone is identified as
an indicator of social capital has concluded that defining social capital
independent of trust is in fact the best strategy. Thus, it is possible to approach
the study of social capital by recognizing the symbiotic relationship between
attitudinal and structural elements. Unlike attitudinal elements alone, the
network approach enables us to also consider the negative consequences of social
capital, which we will find particularly useful for our discussion of the impact of
social networks in transition economies.
3.4 The Future of Social Capital
The field of social capital has reached a point in its development that calls for
critical self-reflection. It is time to evaluate strengths and weaknesses of the field’s
contributions in order to set the path ahead and steer scholarship in the most
productive direction.
6
Currently, the field is flooded with too many definitions
that often lead to functionalist or tautological operationalizations of the key
concept and provide weak or improper measurement techniques. Despite the
significant volume of scholarship associated with this field, social capital has not
6
Some scholars such as Adam and Rončević (2003), Edwards, Franklin, and Holland (2006),
Foley and Edwards (1999), and Lin, Burt, and Cook (2001) are already contributing to this call.
55
moved significantly beyond the metaphor of capital and has not contributed clear
causal linkages to political and economic outcomes.
The structural or network approach to social capital possesses certain
advantages over other approaches and offers the basis upon which future social
capital scholarship should develop. This approach is not antithetical to social
capital approaches which have focused on trust and norms. Instead, it focuses its
analysis on social relationships and network resources which are central to all
social capital definitions, especially those introduced by Bourdieu, Coleman and
Putnam that now represent the pillars of the field (Field 2003). The premise that
relationships embed resources is not new but enables better operationalization of
social capital than we have observed with approaches focusing on norms and
trust. Moreover, the social network approach can benefit from existing methods
employed by sociologists who have studied community relationships and their
impact on values and norms, well before the term social capital came into
prominence. To avoid the risk of becoming just another “catch-all” term or a
social sciences fad, the field of social capital has to move beyond the discussion of
what it is and how it matters. Where the field stands to make the greatest
contributions is in showing how social capital matters. For examining how and
when social relationships produce distinct payoffs, structural approaches seem to
provide distinct advantages. “With the ever-sharpening definitions and
measurements, social networks scholarship may have much to contribute to the
sustained development of social capital as an intellectual enterprise” (Lin 2001,
24). The field has already generated extensive research exploring the impact of
society and culture on development. In doing so, it has admirably bridged
56
disciplines, bringing together scholarship from sociology, political science,
economics, and management, as well as anthropology.
57
Chapter 4
Social Networks in Transition:
Better a Hundred Friends than a
Hundred Rubles
This old Russian proverb reflects the important role that the informal system of
economic exchange played as a complement to the formal socialist system.
Operating silently in its background, the primary role of this system was to
address the systemic problems of shortages in goods and services resulting from
the inefficiencies of planned economic systems (Kornai 1992). The essential
infrastructure of this informal system was provided by fluid structures of personal
relationships formed as networks between friends, relatives, coworkers and
neighbors. These instrumental networks of contacts became known as sviazy in
Russia, veze in the Balkans and guanxi in China. In addressing scarcity in their
every day lives, social networks enabled consumers to obtain goods in short
supply or at least receive information about where to acquire specific goods and
services. In her anthropological study, Ledeneva (1998) describes the networks of
58
blat, meaning pull or influence, as the use of informal contacts or relationships
rooted in trust and sympathy to procure goods and services through a form of
nonmonetary exchange. Networks were needed in all aspects of life: for obtaining
meat, shoes, train tickets, medical services, apartments, entry into colleges and
jobs among others. Due to limited resources, producers would use networks of
personal relations to trade in resources and finished goods, in order to ensure that
they could reach their predetermined production targets. Enterprises in the Soviet
Union often designated an employee (tolkach) whose primary responsibility was to
procure resources through long-term personal relationships with other businesses
(Litwack 1991). In the realm of socialist production, networks facilitated the
exchange of goods and services across economic sectors to eliminate shortages in
primary inputs and to assist producers in meeting inaccurately determined
production targets. The norm of reciprocity was central to the functioning of
informal networks as reliable complements to the ineffective legal framework and
poorly managed system of distribution. Informal networks, and the unwritten
rules of exchange they cultivated, enabled the rigid institutional framework to
coexist with the disparate needs of the society (Ledeneva 2001).
The informal system not only complemented the formal system from within,
but continued to exist alongside it as a means to evade it (Ledeneva 2001;
Lomnitz and Sheinbaum 2004). Networks of relations allowed individuals with
privileged access to public resources to redistribute state property and services for
personal gain and exercise influence over individuals seeking assistance. The mark
of a good position was not the size of the salary one was awarded but the number
of favors one could obtain on account of his or her position. For example, a
position in the party elite allowed one to trade on political appointments or
59
permissions for foreign travel while a position in a military hospital allowed one to
capitalize on access to medical staff or supplies. In such a way, the informal
system of networks penetrated the formal system in order to abuse it. Informal
practices often crossed into the realm of illegal activities such as extortion of
payments and favors, favoritism, nepotism and clientelism, all of which constitute
primary examples of corruption (Karklins 2005). Often, the role of informal
exchange in facilitating appropriation of public resources for individual gain and
the role of networks in addressing shortages for survival were blurred.
4.1 Networking Through Transition
The dismantling of the socialist system of government and the introduction of
liberalized markets was expected to dissipate the role of networks and informal
(nonmarket) exchange. Since goods and services become freely available and in
abundant supply, the need to redistribute goods informally should dissipate.
Similarly, if businesses are allowed to set their goals according to market signals,
discrepancies between objectives and resources in economically sound companies
should cease to exist, eliminating the need to trade in resources with other
businesses. However, the process of transition created its own opportunities for
exploiting state resources through informal exchange. The weakness of the rule of
law and the unprecedented change in state–society relations created an ambiguous
landscape of economic policy that opened up opportunities for turning state
property into private property. Privatization and restructuring, two critical
reforms of post-communist transition, enabled those occupying positions of access
to state-owned property to “grab” resources in order to trade with them
informally (Sik 1994). Social networks played a role both in the provision of
60
access to such individuals and in facilitating the informal private exchange of
state assets. The logic of deregulation and privatization, namely the return of
state property to the public, and the impression that the communist state
betrayed its citizens, provided the justification for the “grabbing” of state
resources. Thus informal exchange relations used to exploit the communist system
from within continued to play a similar role in the transition process and created
additional opportunities for exploiting state resources and services.
4.2 Networks, Transition, and Economic
Outcomes
Social networks also continued to provide the means for compensating for the
deficiencies in the newly reformed state structures and markets, or simply, for
getting by. During the period of adjustment and social learning, individuals
resorted to established informal systems of exchange as a coping mechanism to
use in crises and under uncertainty (Sik 1994; Ragaru 2003; Lonkila 1997). Even
though shortages in goods and services became rarer, networks were used to
overcome shortages in money and knowledge and in this manner represented an
important contributor to individual economic status and well-being. Networks of
social relationships continued to serve as a familiar mechanism for insuring oneself
against negative adjustments in economic opportunities. The unprecedented pace
of change in state and local regulations, for example, created information and
knowledge shortages which were ameliorated with exchange through personal
relationships in social networks (Caldwell 2004). Ledeneva (2001) argues that the
incomplete institutions and the changing market conditions have generated
61
circumstances in which the use of networks has mutated to the extent that it is
even more prevalent than during the previous economic system. According to Sik
(1994), both aspects of the “networking culture” are more prevalent today than
before the transition. Empirical studies from Russia, Kyrgyzstan, Hungary and
Bulgaria confirm this finding (Woodruff 2004; Radaev 2004; Kuehnast and
Dudwick 2004; Lonkila 1997; Ragaru 2003). However, the relationship between
the exploitative role of social networks and their role as a coping mechanism
remains ambiguous.
Studies of transition in Eastern Europe and the former Soviet Union that focus
on social capital in terms of networks and their resources are interested in the role
that networks play in the building up of democratic institutions and the market
economy. Ragaru (2003) and Rose-Ackerman (2001) see informal networks of
social relations as barriers to the development of lasting democratic institutions.
Informal exchange allows for the continuation of the blurred boundaries between
public and private spheres that allow for the institutionalization of clientelistic
and corrupt relationships with political elites. Together, these factors can
undermine the legitimacy and support of public institutions (Ragaru 2003).
Karklins (2005) notes that the reliance on networks of relationships is “in tension”
with the new system of political and market institutions since the exclusivity of
network memberships inhibits the development of an inclusive civil society.
Similarly, Rose-Ackerman (2001) questions whether the legacy of past
relationships is antithetical to the establishment of an institutional framework
based on democratic principles of accountability and impartiality.
Scholars focusing on the role of networks in the establishment of market
institutions offer a far less unanimous opinion about the use of networks and their
62
impact. For some, the legacy of informal exchange has negative implications for
economic transition. Kolankiewicz (1996) notes that networks inherited from the
socialist period are dense and characterized by ”amoral familism” and clientelism
because they were formed to cope with the specific problems not addressed by the
state. As a result, network relationships impose constraints on economic units,
which must be self-interested in order to advance the goals of privatization, and
cannot be encumbered by obligations of reciprocity (Kolankiewicz 1996). A
particularly dangerous effect of reciprocity for economic performance can occur
when businesses are forced to accept compromises in quality of products, services,
or delivery because they are unable to sever business ties rooted in friendships
and other personal relationships (Mihaylova and Harriss 2003). Similarly, Sedaitis
(1997) argues that businesses formed based on “tradition, loyalty and the memory
of past accomplishments” enjoy limited benefits in terms of organizational
flexibility since they lack access to information outside their limited number of
contacts. On the other hand, some were concerned that to survive in the rapidly
transforming environment, only those individuals with ties to existing business
circles would find it profitable to join in private production and start private
businesses, shutting the rest of the society out.
An even broader concern was that personal exchange would breed corruption
that would essentially dampen the incentives to enter into productive economic
activities and slow down the emergence of the private sector. Rose (2000)
observes a difference between the use of networks in an “antimodern” and
“modern” society. In the former, such as Russia, networks are used for getting
favors from public officials, or using connections to bend or avoid rules, all of
which contributes to a system lacking transparency. Such a system breeds
63
uncertainty that disrupts the formation of stable expectations associated with
more efficient systems. Paldam (2000) argues that informal networks, which
supported the grey or the black economy, were necessary during the socialist
system but that during transition they become detrimental allowing harmful
social capital to grow and prevent economic development.
The use of networks for corruption is not sufficient for a successful argument
that informal networks and the social capital they represent has a negative impact
on institutional development and economic growth. Ethnographic studies of social
networks in transition have argued that networks and corruption, while
reinforcing each other in some instances, are not synonymous (Lonkila 1999;
Ledeneva 1998; Kuehnast and Dudwick 2004; Caldwell 2004). In the case of
exchange of favors personal relationship are established prior to the exchange of
help, money or resources. On the contrary, corruption and bribery often involve
exchange of monetary payments and are not necessarily based on a personal
relationship. In this sense, network ties are less instrumental as they are not
established solely to facilitate bribery. Moreover, corruption, if and when
facilitated by personal relations, implies a short term transaction. Network
relationships, however, continue to exist even after the material exchange involved
in the favor has already taken place (Wang 2000).
1
Hence, equating network
exchange with corruption and black market outcomes is not a sufficient
explanation of how social networks can negatively impact economic outcomes.
Also, numerous scholars have identified a positive role of social networks in
economic transformation of Eastern Europe.
1
Wang (2000) studies guanxi networks in China, but her discussion of the difference between
guanxi and bribery is relevant for the transition of economies of Eastern Europe.
64
Scholars who emphasize the role of informal exchange as a coping mechanism
in a time of uncertainty often see a positive impact on economic activity. Not only
are personal relationships needed in the state-owned industry where the “old
ways” of thinking and conducting transactions are still prevalent, but they also
prove useful in establishing trust and reducing risks in business transactions
within the privatized sector (Lonkila 1997; Manolova and Yan 2002). A study of
small scale, cross-border traders in Central and Eastern Europe revealed that
traders rely on the use of friendships and personal relationships as a form of
assurance against cheating in established trade relations (Wallace, Shmulyar, and
Bedzir 1999). They build patron–client relationships to pay lower fines, avoid
unexpected payments, obtain licenses, and pay lower fess. Similarly, Todeva
(1998) emphasizes that in order to deal with structural uncertainties, business
actors rely on informal exchange as a vehicle for the delivery of information.
Social networks thus serve as a mechanism for managing risks due to changes in
market conditions arising from discrepancies between supply and demand,
changes in government regulation, taxation, and the availability of resources and
labor supply (Todeva 1998). Peng and Heath (1996) argue that informal exchange
relationships are part of a growth strategy for businesses in transition because, in
addition to protection, such relationships can foster trust among business actors
enabling resource pooling and a reduction in transaction costs. In reference to the
use of informal relationships in present-day China, Xin and Pearce (1996) find
that network relationships are often used to defend against the threats of
expropriation and extortion that dominate an unstable business environment in
the process of transition. In such an uncertain environment a network of relations
65
can assist individuals in taking advantage of previously unavailable opportunities
for personal enrichment (Sik 1994).
4.3 Summary
Preview of the scholarship emphasizing the role of social networks in the context
of transition shows that among those who have noted the predominance of
informal exchange in post-socialist economies, little consensus exists in regard to
the implications for the private sector and economic growth. It is not clear to
what extent the networks compensating for the institutional inefficiencies have
assisted the development of the private sector, or to what extent the exploitative
role of networks is subversive of the business environment necessary for private
sector growth.
Social network studies dedicated to examining the lackluster transition
performance of economies of SEE are exceptionally rare. The scholarship on social
networks in Eastern Europe in general is based on qualitative studies that are
mostly anthropological and ethnographic in nature.
2
The focus has largely been
on explaining the nature and role of social networks in relation to their socialist
legacies. This means that we lack studies that examine the relationship between
the use of networks and economic performance, or investigate whether networks
are impeding or facilitating economic transition and the growth of the private
sector. Also, if there is indeed a significant impact on the business environment
and the incentives for economic activity, we ought to explore further to what
extent the differences in the characteristics and use of informal exchange
2
Woodruff (2004) and Batjargal (2003) are some of the exceptions to this generalization and
these works will be reviewed in Chapters 7 and 8 respectively.
66
relationships can explain the differences in the success of individual enterprises.
Ideally, these explanations can further be linked to differences in outcomes of
private enterprises across Eastern European post-socialist economies.
67
Chapter 5
Methods and Data Collection
Outside the context of transition, scholars exploring social networks have
extensively relied on methods of network analysis grounded in sociology. The
selection of the most appropriate method is in large part determined by the
research questions and the goals of the study. Here, the primary objective is to
empirically examine the nature of social networks and explore their impact on the
economic performance of the SME sector. Hence, some methods, such as the
positional approach to social networks, will be passed over in favor of other more
appropriate methods and techniques, such as the relational approach. It is
important to note at this point that the rejection of the positional approach
should by no means indicate its inferiority as compared to other available
methods. In fact this approach has achieved great recognition by introducing
social network analysis developed by sociologists to the analysis of networks in
economics. Next we review the strengths of these two approaches and the
associated measurement techniques as the two options which can be utilized in
the present study.
68
5.1 Network Analysis
Studies of social capital that focus on networks and social ties, and especially
their impact on the success of economic transition reforms, can benefit greatly
from the methods of network analysis developed in sociology. Social network
analysis is a diverse body of methodologically oriented scholarship that is
interested in examining social structure. It can be considered as a research
program whose goal is to design measures of social structure that can be
employed in other theories (Cook and Whitmeyer 1992). Though a diverse body
of scholarship, social network analysis has at its core one identifiable premise: it
rejects explanations of human action based on the individuals’ possession of
certain attributes or norms, and explains actions as a result of their involvement
in social relations (Emirbayer and Goodwin 1994).
5.1.1 Positional Approach
Network analysis usually adopts one of two conceptualizations of social structure.
The positional approach is concerned with identifying and describing patterns in
social networks which can explain how they impact behavior and social change
over time (Cook and Whitmeyer 1992). The relevant factor to observe is the
position or role that a set of actors occupies within a system as a whole. The
patterns of relations that define an actor’s position relative to other actors in the
system are used to explain behaviors and processes within the network
(Emirbayer and Goodwin 1994). The positional approach views networks as
patterns of social ties and ignores the characteristics of the individual occupying
specific positions (or nodes) in that structure.
69
The positional approach is closely related to the methods of network analysis
adopted in economics. The treatment of networks in economics is based on the
rational choice assumption that individual actors will choose whether to form or
sever network ties based on the rational calculation of benefits (net of costs
involved in forming the link) obtainable through such ties. The costs and benefits
are specified as a function of the network structure (Jackson forthcoming). The
payoffs then shape the decision of whether to form or sever the links and provide
the basis for evaluations of social welfare. In studying social networks, economists
recognize the importance of network externalities which can cause a tension
between individual choices and the objective of maximizing social welfare.
Network externalities are clearly evident in the examples of product or standard
adoption. The more actors adopt a specific type of computer operating system, or
a keyboard layout, the greater the value of those goods to the marginal actors and
the society as a whole. The increase results from the complementarity between
individual decisions to form or sever links with others in the network (Economides
1996, Farrell and Klemperer forthcoming). The objective of network analysis is
economics is to examine network formation and specify network architectures or
structures which maximize social welfare. The relative positions of actors within
the structure determine network configurations and the resulting welfare
implications. The main concern of game-theoretic studies of network formation is
the tradeoff between stability of the network and the efficiency in terms of social
welfare (Jackson 2005).
An omission in this treatment of social networks is the recognition that
sometimes even costly links have to be maintained and cannot be severed. Various
affect-based relationships such as family relationships and close personal ties,
70
where one actor does not necessarily gain economic benefits, are maintained
because of the sense of obligation. Depending on the underlying culture of the
society under observation, individuals in familial and close personal relationships
will feel differing degrees of responsibility to one another. As a result, individual
actors are sometimes obliged to provide resources to one another, or connect the
same individuals with other actors, without obtaining any benefit in return for
their efforts. This notion does not necessarily violate the assumption of rational
choice that underlies economics’ treatment of social networks since obligations
and reciprocity can be modeled as reductions in opportunity costs or equivalently,
as increases in benefits. However, this omission points to the broader problem
with the assumption that all ties are formed based on the same underlying
motivation. In other words, the presence of the link is treated as the opposite of
an absence but there is no differentiation in terms of the nature of the links
forming the network. All links are assumed to be the same.
Moreover, the positional approach is not concerned with investigating how the
presence or absence of links affects individual behavior. If we are concerned with
the influence of network ties on individual outcomes, it may not be appropriate to
examine positions of actors in the network and the resulting network structures.
In other words, the positional approach may not be successful in examining the
volume and types of resources accessed through network ties or whether these
resources translate into economic payoffs. For this specific objective the relational
approach to network analysis is more appropriate.
71
5.1.2 Relational Approach
The relational approach to network analysis explains behavior and network
characteristics by focusing on the direct and indirect ties among actors. Density,
strength, symmetry, range and other elements of ties describe and explain
network processes. From this perspective, the nature of relations among actors is
more important than the location of individual actors in the network. However,
within the relational approach, we can still differentiate between the studies that
focus on the properties of the network as whole and those that focus on the
capital made available to the individuals within the network. The first group
studies the characteristics of the networks as determined by, for example, the
strength and density of the ties among actors. These studies refuse to use the
number of ties and diversity of ties to which an actor has access in a network as
measures of social capital and predictors of economic behavior (Foley and
Edwards 1999). Scholars working in this perspective believe that such relational
measures should never be divorced from the network structure captured by
measures of network properties. The sorts of ties, as well as the resources they
give access to, are a function of the network structure, and focusing on the capital
available to individual actors ignores the importance of the overall structure.
Therefore, according to this perspective, measures of network properties, and not
individuals’ capital within the network, should be used to assess the levels and
nature of social capital.
While an individual’s number of network ties certainly does not capture the
value of social capital that one can employ in an economic activity, focusing on
the properties of the network alone fails to explore the role of relationship
content. The overall strength of ties and the density of the network are variables
72
that do not adequately describe the types of relationships that individuals forge,
how they benefit network members, and what their effect is on actors’
decision-making. Focusing on network properties alone means that actors are
neglected and therefore we cannot observe the resources and advantages which
individuals enjoy as a result of being in a network.
To fully understand the use of networks and the advantages they afford their
members it is important to work with a perspective that recognizes that
individuals comprise the networks and that not all relationships have the same
value to individuals. Relationships with some individuals in the network may be
more important than others, while simultaneously the structure that is created
enables and constrains the individuals who comprise it. A complete analysis needs
to also consider the ways in which individuals’ choices are impacted by the
specific structure. Examining both the structure and the relationships is
necessary to understand how different networks affect resources and opportunities
available to their members. Therefore, an analysis has to consider the properties
of the network, but it also must pay attention to how different properties
translate into individuals’ resources that in essence determines actors’ economic
decisions. We must therefore operationalize networks in a way that considers
network structure (in terms of strength or density of ties, for example) and also
the content and nature of ties (in terms of who is in the network and what
resources they provide to network members).
5.2 Nodes and Ties: Who is in the Network
A network under analysis can be envisioned as a structure formed by a set of
relationships around an individual actor. Both the actors linked and the ties
73
among them constitute the network. A group of actors in close physical proximity,
or belonging to a similar organization, do not constitute a network unless every
pair of actors is connected to other actors via a direct or indirect tie. Direct ties
refer to dyadic relationships between two actors who interact with each other.
Indirect ties refer to relationships where actors may not interact with each other
directly but they each have a direct relationship with another common actor in
the network. In general, actors form and maintain relationships in order to take
advantage of resources embedded in network ties.
The actor at the center of the network, depending on the analysis, can range
from an individual, or a firm, to various groups of individuals within larger
organizations, and even nation-states. The point of interest in this study is the
manager and/or owner of an individual SME. Organization and management
theory has long established that the success of a business venture and a firm’s
subsequent growth depend on the owner-manager’s personal characteristics as well
as the strategies he or she develops for the management of available resources and
constraints. Recent additions from entrepreneurship research have emphasized the
role of the owner-manager’s social network as an important tool for gathering
information and mobilizing resources (Ostgaard and Birley 1994). Given the
broad objective of understanding the role of personal relations as a business
strategy and the impact of social networks on business success, this study focuses
on the networks of firm managers and owners as the relevant networks that
facilitate or constrain the decision making of the firm.
A strong definition of a network recognizes only the social ties, which are
backed by enduring, reciprocated exchange relations (Rauch 2001). A weaker
definition includes short and long-term relationships without regard to
74
reciprocation. Unidirectional linkages, where one actor is only at the receiving end
of the exchange, or affect-based linkages, where actors are linked through familiar
or emotional relationships, can also represent social ties at the basis of a network.
The weaker definition of a network essentially includes all linkages based on
relationships (realized or potential) which can provide access to economic
resources. This definition then subsumes the common distinction between social
and economic ties. These two types of ties are frequently blurred in exchange to
make such a distinction practical. In reality, it is difficult to disassociate the
business owners’ trade agreement from any underlying social ties that may exist
between the two actors based on friendship, familial connection, or other
connection outside the scope of the purely business relationship. All relationships
that can be used to take advantage of resources, no matter how established, are
considered equally relevant to the definition of a network.
This conception of the network and its focus on the individuals’ social
relations sets this study apart from the research that analyzes the formation and
benefits of inter-organizational networks.
1
While social ties may be precursors for
the formation of relationships between firms, inter-organizational network studies
are focused on formal linkages among firms and the structures that the linkages
produce. Neither the individuals themselves, nor the social ties among them, are
the subject of their analysis. A specific example of networks studied includes
community-based networks, such as clusters and industrial districts, which are
loosely linked firms whose specialization and competitive advantage is based on
ties to a specific locality and the geographic proximity of other network members.
Other examples of inter-organizational networks include business networks held
1
These include contributions from industrial economics, industrial marketing, game theory,
organizational sociology, resource dependence theory, and institutional theory (Ebers 1997).
75
together through “relations of ownership, investment or shared membership,” and
they include business groups such as Japanese keiretsu or Korean chaebol,
strategic alliances, and joint ventures (Perry 1999). Some inter-organizational
studies tend to view networks as a mechanism for coordinating exchange
relations—an alternative to markets and firms (Casson and Cox 1997; Perry
1999). Such a conceptualization is highly restrictive for the purposes of this study
because it renders relationships among individual units pertinent only to the
proposed network form of governance when in fact relationships are ubiquitous
both within firms and across markets (Granovetter 1985). More importantly, the
focus here is on the relationships formed by individual decision makers, without
regard to the degree of their formalization.
The broad conceptualization of networks used in this study differs from
inter-organizational network studies in considering networks of individuals who
are connected to others through both formal and informal relationships, and who
interact with each other across firms and markets. Although the operating
definition is different from that adopted in the literature dealing with
inter-organizational networks, relationships formed between business owners and
managers are certainly not excluded in this operationalization of the network.
This broad definition has clear limitations from the positional approach to
network analysis, which attempts to develop precise measures of social networks,
by quantitatively measuring the attributes of patterned exchange among
individual members. From this perspective, a clear specification of the criteria by
which network members are identified as well as a narrow specification of the
relationships which qualify as network linkages are preferred. The focus in this
study, however, is not to empirically describe the structure of a network and to
76
identify all of its members, but rather, to discuss why and how specific network
relationships are utilized. In this case, a broad definition of networks is
appropriate since it allows us to consider a great variety of applicable
relationships under the single criterion that they provide individuals with a
resource that can impact their economic decisions.
5.3 Research Design
For this study, the empirical analysis of networks and their impact on business
performance in SEE could not be performed using an existing dataset. As we
have already noted, empirical studies and data sources addressing the nature and
role of social capital in SEE are scarce. Most of the preexisting evidence has been
collected using the World Values Survey and the New Democracies Barometer,
and only recently in SEE countries. These surveys provide attitudinal measures of
social capital which, we have seen, are inferior to the measures based on structural
elements such as networks. An extensive report on social capital in Bosnia and
Herzegovina was produced by the World Bank (2002) using qualitative interviews
and focus groups. It offers a rich contextual analysis of the relationship between
social capital and the interaction between ethnic groups, but its findings are not
conducive to the analysis of firm-level performance.
Firm-level survey data is provided by the BEEPS survey administered jointly
by the World Bank and the EBRD. This database offers extensive information
about firms’ perception of their business environment, focusing specifically on
barriers to business entry and business growth. While this database provides us
with general assessments about firm-level experience with institutions in SEE, it
does not include any measures of social capital. Therefore none of the currently
77
available sources of data for the region under consideration are pertinent for this
study. In order to overcome the significant drawback of currently available social
capital measurements (such as those derived from the World Values Survey) and
the lack of firm-level data related to social capital, this study has relied on
original data collection.
2
5.3.1 Qualitative Data Collection
The data collection process was constructed to produce both quantitative and
qualitative data dealing with social networks and firm-level performance. The
collection of both quantitative and qualitative data was conducted concurrently
but independently. The review of each collection process will explain the
differences between the two data sources.
Qualitative evidence was collected during field research in Serbia and
Montenegro in the summers of 2005 and 2006. Face-to-face interviews were
conducted with managers and owners of SMEs in Belgrade, the capital of Serbia,
and three cities in Montenegro: the capital, Podgorica, Kotor, and Herceg-Novi.
The interviews were extended to two cities other than the capital because, unlike
Belgrade, the capital of Montenegro is not the single most important economic
center in the region. Two cities in the booming coastal region were selected to
complement the evidence obtained from the SMEs Podgorica.
The sample of SME managers and owners that could potentially be included
in the study was identified from the registries obtained from the Chamber of
2
An alternative source of data not described above is the Phare-Ace Project 97 8089-R which
apparently examines barriers to SME development in select SEE countries (Macedonia, Bosnia
and Herzegovina, and Slovenia) and includes questions that could offer reliable indicators of
social capital (Bukvic et al. 2001). Further information on this database could not be obtained
and one of the principal investigators has confirmed that access to this database is not readily
available.
78
Commerce of Serbia and the Business Center associated with the Agency for the
Development of Small and Medium Size Enterprises in Montenegro. SMEs were
randomly selected from the obtained databases and were contacted by phone to
inquire about their willingness to participate in the study. Approximately 80% of
the contacted persons accepted to do an interview. The interview acceptance rate
was higher in Montenegro. In each case, the person interviewed was the top
manager, or owner of the SME, or both. Approximately 10% of the SME
managers and owners who accepted to do an interview were identified through the
principal investigator’s social interactions while in the field. For this subset of
SMEs, careful attention was paid to the selection process so as to avoid the
snowball sampling strategy. Including businesses whose owners or managers were
mentioned by already interviewed businesses would have inadvertently restricted
this data to a single network of business contacts. To avoid this outcome, contacts
suggested by interviewed businesses were not pursued and 10% of the SMEs were
identified through independent interactions in the field.
A total of 41 interviews were conducted, 16 in Serbia and 25 in Montenegro.
Among them 13 were micro enterprises (1–9 employees), 20 small enterprises
(10–50 employees), and 5 medium enterprises (50–250 employees). The sample
also includes 3 large enterprises since the information on the number of employees
obtained prior to the interview was inaccurate or incomplete. Interviews with the
managers of large enterprises were not discarded because they offered information
pertinent to their growth from small and medium businesses into large businesses.
An overview of the businesses that accepted to participate in the study is
provided in Appendix A. SMEs from various sectors of the economy accepted to
participate in the study. The distribution of interviewed SMEs by their sector of
79
operations is fairly consistent with their overall distribution within the economy.
As a result, the largest number of interviewed SME managers and owners is
involved in trade. Other sectors represented among the interviewees include food
production, services, construction and manufacturing. Other parameters with
respect to which the sample captures significant variety include the age of the
respondent, previous experience of the respondent, the year in which the
enterprise was established, and the general assessments of the enterprise success.
Data was collected using open-ended interviews conducted in the local
language (Serbo-Croatian). The author conducted all of the interviews in person.
The interviews, which lasted between 45 minutes and 2.5 hours investigated issues
such as the ways that businesses obtained information about prospective
employees and business opportunities, sources of help and the circumstances in
which they were used, procedures for ensuring that business transactions are
completed, and the willingness to sever business relationships to pursue more
desirable contracts. The interviewer had a list of questions designed to encourage
discussion but she adhered to the questions loosely, changing them to suit the
discussion at hand, adding questions to clarify interviewees’ responses, or
encouraging them to provide additional information on a specific topic. This
proved to be a rewarding strategy because the interviewees had the opportunity
to offer any information they wished without waiting to be prompted by the
interviewer. They also offered information that otherwise might not have been
invited by the scripted interview questions. This aspect of the interview process
has indirectly increased the validity of the responses because any similarities in
the use of network resources identified by the managers cannot be attributed to
the design of specific questions or the particularities of the interview process.
80
Overall, the open-ended nature of the interview process produced a setting in
which the interviewee felt relaxed and free to offer any information about their
business, even the description of semi-legal strategies they have adopted to
overcome barriers to their business growth.
Interviews with SME managers and owners were conducted to gather
information about the ways in which network relations are used to address
obstacles to everyday business activities in Serbia and Montenegro. The group of
interviewed businesses does not include individuals who are contemplating starting
a new business or have had a business that has since shut down. The interview
responses thus apply to existing businesses only and provide us with only a partial
representation of the role of networks, or lack of network ties, with respect to
entry and growth of SMEs. To ameliorate this problem, interviewees were
encouraged to speak about the strategies they would pursue to avoid obstacles in
pursuing a new line of business. Another limitation of the interview responses is
that they cannot be easily coded into standardized measures of firm performance.
While businesses can be loosely classified as low, medium, or high performers,
these levels do not carry enough information to be included in any statistical
assessment of the relationship between the use of networks and firm performance.
Neither is the number of respondents sufficient to produce statistically
representative results. The poor quality of the databases used for generating the
contact information for businesses in Serbia and Montenegro presented a
significant problem for the selection of interview respondents. Thus, substantial
time was devoted to contacting businesses in order to persuade them to
participate in the study. Moreover, interviews had to be conducted concurrently
with the collection of quantitative data. Both of these collection techniques had
81
to be implemented during the summer period during which the researcher could
visit the area, but had to be accomplished before August when most of the
businesses in the area shut down and most managers go on vacation. This
provided a limited window within which to complete both qualitative and
quantitative data collection using limited financial resources.
5.3.2 Quantitative Data Collection
To a large degree, the decision to use the relational approach in operationalizing
the network as a unit of analysis dictates the most appropriate approach for data
collection. Positional approaches to the study of social networks most often use
socio-centric methods for data collection and analysis. The main objective of this
method is to quantify relationships among actors (nodes) in a particular network
such as a university department, organization, team, or a community tied to a
specific locale. All of the actors within the network are asked to provide
information about their relationships with everyone else.
3
Information such as the
extent to which individuals interact or how well they know one another is used to
quantify relationships and construct measures of network structure. Graph
analysis is often used to “map” the positions of network members in order to
describe network attributes (such as stability and efficiency in economics).
Alternatively, statistical methods, such as cluster analysis, are applied using
quantitative measures to uncover differences in relationship structures among
subgroups. Critical empirical measures include the position of nodes with respect
to all others and the level of connectivity with other nodes in terms of the type
3
For large networks whose size exceeds 100 members, for example, it is not feasible to ask
individuals to comment about their relationships with everyone else. In this case, individuals are
asked to comment about their immediate relationships instead. Measures are combined to try to
construct the entire network as closely as possible.
82
and density of individual ties. In this approach, networks have to be strictly
defined to enable the specification of network boundaries and to identify who does
or does not belong to a network.
4
Using the relational approach in the study of social networks does not require
the mapping of the entire network. The emphasis here is on the individuals and
their relationships for which the ego-centric methods for data collection and
analysis are suitable. In this approach, the information is collected about
individuals’ relationships with the most immediate members of their network that
may cut across group memberships. In this manner, the researcher does not
indirectly select network boundaries but allows the respondent to identify
relationships which contribute the most to their behavior. The researcher asks the
respondent—ego—to name the individuals with whom he or she
interacts—alters—and to characterize the nature of the relationship with each one
of the named individuals. To get a sense of network characteristics, the ego is also
asked to describe the relationships among the alters. Depending on the study,
various criteria can be provided to the ego to use in identifying the alters, and in
differentiating the relationships with alters. Instead of interviewing every one of
the stated alters, the researcher relies on the information provided by the ego to
describe the features of the network to which this individual has access. Measures
of the nature and strength of relationships provide the measures of network
composition but they are nevertheless considered attributes of the individual
respondent. Given the objective to relate network measures to economic
outcomes, this measurement and analysis technique is more resource efficient.
Firm-level data on performance must be obtained from every business but it
4
Marsden (1990) and Cook and Whitmeyer (1992) provide a review of network analysis
methods in sociology.
83
would be prohibitively costly to map the entire network of every single business
included in the study. Thus, to obtain an individual level measure of the social
capital accessible through one’s social network, this study uses standard survey
sampling techniques.
The requirements of the ego-centric approach to measuring network attributes
have influenced the construction of the survey instrument. First, the survey
includes elements which ask the respondent to name the members of the network
(alters). These are followed by a series of questions designed to collect specific
information about the alters. Next, the survey asks the respondents to describe
the nature of the relationship between the respondent and the identified alters, as
well as the nature of the relationship between the alters. All of the elements
discussed thus far are included for two different types of networks in order to
capture the information about the content of exchange transmitted through the
networks. This feature of the survey is discussed in Chapter 8. The survey also
includes questions that form measures of firm performance such as the rate of
profit reinvestment and self-reported growth in sales. Basic information about the
business and demographic information about the respondent are also collected.
The survey was constructed in the local language (Serbo-Croatian). The
accompanying recruitment letters and survey information sheets were also written
in Serbo-Croatian and mailed to the businesses in the sample.
The sample targeted with the described survey instrument was identified using
the database of registered SMEs obtained separately from the Serbian and
Montenegrian Chamber of Commerce. Using various delimiting criteria described
in Appendix B, the total population of businesses registered with each Chamber
of Commerce was reduced to a sampling frame that included business from the
84
county limits of Belgrade, Serbia and counties of Podgorica, Kotor, and
Herceg-Novi in Montenegro. From each sampling frame, every twentieth business
was selected for each of the Serbian and Montenegrian samples. Businesses in the
sample were mailed a questionnaire between June and August of 2005 and 2006.
Only the businesses’ managers or owners were asked to fill out the questionnaire
and they were instructed to return it in a postage-paid envelope. A total of 1820
questionnaires were mailed. 337 questionnaires were returned by the post office
because they were undeliverable and 12 questionnaires were mailed back but were
unusable. That means that 1197 questionnaires were either 1) delivered to the
incorrect location, 2) the respondent forgot to respond or mail the questionnaire,
or 3) the business at that address declined to participate in the survey. A total of
274 questionnaires were filled out correctly and used in this study, implying a
response rate of 22.89%.
5
Of the total number of surveys, 57.6% of surveys were
mailed to Serbia and 42.4% to Montenegro. The response rate from Serbia was
higher than that of Montenegro—20.71% and 15.17% respectively.
6
Response rates for mailed questionnaires tended to be lower than for in-person
or phone interviews. In addition, the study did not include any secondary
mailings to encourage those who had received the questionnaire to respond.
7
As a
point of comparison, Cromie and Birley (1992) obtained a 24% response rate for
questionnaires mailed to businesses in Norway. The same response rate was
recorded by Westlund and Nilsson (2005) for private enterprises in Sweden. The
authors cite a comparable study in the United Kingdom that received responses
5
The response rate is 15.05% if we do not exclude the number of questionnaires which were
undelivered by the post office and were returned to sender.
6
Including the number of undelivered questionnaires, these rates are 17.74% and 11.40%.
7
Given the problems with the contact information in the databases obtained from the
Chambers of Commerce, it was decided that a second mailing, such as postcard reminders,
would have been an unwise expense of limited resources.
85
to only 14% of mailed questionnaires. Aldrich, Rosen, and Woodward (1987)
received responses to less than 41% of questionnaires mailed to businesses in
North Carolina. It is not clear whether these response rates were calculated using
only the total number of delivered questionnaires or the total number mailed.
None of these studies, however, used an ego-centric questionnaire that included
not only information about the direct (ego–alter) relationships, but also
information about the indirect ties among network members (alter–alter pairs).
For the current study, these questions were laid out in tables asking respondents
to fill out the corresponding columns and rows. In retrospect, these questions
complicated and lengthened the survey because the respondents had difficulty
filling out the provided tables. Questions about relationships forming alter–alter
pairs would have probably been handled better with in-person interviews but the
nature of the study and the available resources did not make this form of data
collection possible.
Table 5.1 provides select summary statistics for the survey’s respondents.
Among the respondents, 54.01% are the businesses’ managers, 12.41 are owners,
27.74% are both, and the remaining 6% are founders, section managers, or both.
The majority of respondents were male (87.59%). Individuals who returned the
survey ranged from 37–67 years of age. Among them, 64.59% had completed a
college or graduate degree, indicating a highly educated group of respondents. Of
the respondents, 7.03% were unemployed before starting the business for which
they are reporting, while 35.40% held positions in the private sector, 55.83% held
a position in the public sector and only 2.19% worked in the bureaucracy or the
government. Of all the respondents, 41.24% previously worked for businesses that
offered the same product or services as the ones they are currently managing.
86
Table 5.1: Select respondent characteristics
Respondent’s Total Percent
Position
Manager 148 54.01
Owner 34 12.41
Manager & owner 76 27.74
Founder 3 1.09
Section manager 6 2.19
Founder and manager 7 2.55
Gender
Male 240 87.59
Female 34 12.41
Education
High school or tech school 27 9.86
Some college 70 25.55
College 140 51.09
Masters & above 37 13.50
Previous experience
Unemployed 18 6.57
Private sector 97 35.40
Public sector 153 35.84
Government 6 2.19
Age
Min 31
Max 67
Mean 49.56
St. Dev. 9.23
Total 274 100
87
5.3.3 Scope of Analysis
The biggest obstacle encountered in the process of data collection involved the
SME database itself. Identifying the database for the entire population of SMEs
from which to draw a representative sample was the most difficult task of the data
collection process. Court registries in Serbia and Montenegro, which are
responsible for registering and documenting all businesses, did not offer access to
searchable databases. In each case, a business’s record could only be pulled up by
its name, and the entire record would have to be printed. This record did not
indicate the size of the business, which prevented the researcher from restricting
the sample to only those businesses with 250 or fewer employees.
8
As a result, the
Chamber of Commerce database obtained directly from the head offices in Serbia
and Montenegro provided the best alternative.
Over the course of the data collection process, the quality of the database
revealed itself to be rather poor. While it included an extensive selection of
businesses with abundant information for each individual business, the database
contained a surprisingly large number of records for closed or non-existent
businesses. Contacting the individual owners listed in the database revealed that
a high percentage of businesses were registered but never operated, or had shut
down. In an attempt to improve the response rate, a select number of businesses
were located by neighborhoods in the city of Belgrade and the researcher made an
attempt to visit the businesses to personally persuade their managers to fill out
the questionnaire. It was thought that such a visit would make the businessmen
less inclined to refuse participation. To honor the researcher’s visit and the
8
Similar problems were found with the online version of the database managed by the
Chamber of Commerce in Serbia. The Chamber of Commerce of Montenegro did not offer an
online database.
88
attention given to the business, the prevailing business culture suggests that the
owners and managers would feel compelled to reciprocate by filling out the survey.
Unfortunately, when these businesses were visited, the researcher witnessed the
poor quality of the information in the SME database. Out of the first 20
businesses, none were found at the specified locations. Some addresses did not
exist at all. The rest pointed to residential apartments, sheds, or storage places
and nothing resembling a legitimate place of business. The failed attempt to
increase the response rate revealed that the information included in the Chamber
of Commerce databases was poor, which explains why many of the questionnaires
were returned undelivered by the post office. Evidently, only some of the delivered
questionnaires went to existing and operating SMEs. By this point in the data
collection process, over 80% of all questionnaires had been mailed and despite its
problems, the database was kept as an integral part of the sample selection
process.
9
The poor quality of the initial database is critically responsible for the
low number of interviews conducted and the less than desired response rate for
the mailed questionnaires.
Although the true quality of the database obtained for the purpose of sample
selection was eventually revealed, this data was nevertheless difficult to locate and
obtain. Data collection was also extended for a year since the mailings had to be
repeated in order to increase the response rate. These unexpected difficulties in
data collection, although not uncommon in any field and survey research, forced
the principal investigator to narrow the scope of the project. The original research
design included data collection plans for both Croatia and Serbia and
Montenegro. Unfortunately, the limited time and financial resources and the
9
See Appendix B for the discussion of supplements to this database.
89
resulting data collection delays forced the researcher to sacrifice the comparative
elements of the research design in order to ultimately ensure the project’s
feasibility. As a result, all of the data come from Serbia and Montenegro and all
of the analysis is based on this region.
During data collection, Montenegro initiated and passed a referendum
declaring independence from Serbia. The two republics split their union on June
3, 2006 and have been recognized as independent states by the United Nations.
The data for this project was collected both in Serbia and Montenegro but the
project was never designed to treat these cases as elements of a comparative
study. The pool of responses obtained is too small to warrant any meaningful
comparative analysis. Moreover, a comparative research design would be
weakened by the inclusion of Serbia and Montenegro as two independent cases.
They have similarly poor rates of SME performance and selecting similar cases on
both the independent and dependent variable side would not add any value to a
comparative research project.
10
Hence, the selection of Serbia and Montenegro
was intended as a single case while the study of Croatia, as the second,
independent case was unavoidably dropped from the research design.
The acquired database and interview responses provide us, nevertheless, with
a rich source of information about the case of Serbia and Montenegro and such a
single case study can be of immense value for a hypothesis-generating project
(Odell 2001). The goal of the study described here is in fact to point to a new line
of inquiry based on individual-level measures of social capital as networks in an
analysis of the differences in economic outcomes. We will demonstrate the
usefulness of such an empirical approach to the study of social capital and
10
This idea is based on the comparative case study method discussed in Mill (1970) and Odell
(2001).
90
Transition
Outcomes
SME
Growth
SME Sector
Growth
Profit
Reinvestment
Property
Rights
Courts
SC/
Networks
A
B
Figure 5.1: Plan of analysis
generate hypotheses that should be examined in future research. To this end, the
study of a single case is strengthened by the use of multiple methods of analysis
and types of data. In the following chapters, the interview data are first used to
illustrate the role of social networks in a transition setting by focusing specifically
on the nature of the ties forming the networks and the resources exchanged
through these network ties. The insights from this descriptive project were used
to formulate two parts of the quantitative analysis. The first part explores the
role of social networks as an informal source of contract enforcement and their
impact on the perceived security of property rights (relationship A in Fig. 5.1).
The second part of the quantitative analysis examines whether social networks
impact the performance of businesses directly (relationship B in Fig. 5.1). Lastly,
to complement the analysis of the quantitative dataset, qualitative data is used to
illuminate any inconsistencies in the analysis and to resolve empirical puzzles not
already explained by the quantitative dataset.
91
Chapter 6
Typology of Social Networks
An owner of a publishing company in Belgrade, Serbia recently agreed to comment
on the nature of doing business in Serbia and Montenegro and the strategies he
employs to ensure the continued success of his enterprise. He explained that the
personal relationships he maintains with managers and owners of the businesses
who purchase his publishing services are critically important for conducting
day-to-day operations. The owner is expected to vacation with his customers and
their significant others, and he must encourage his wife and children to join him
on such trips and other social gatherings. Relationships with other businessmen
ensure that he is “high on their list” and that they will honor their agreements
toward him when it comes to paying for the services rendered. If he refrains from
extending the relationships outside the strictly commercial setting, he would have
no leverage to persuade the other managers to consider making payments, and the
latter would have no incentives to make such payments. The costs associated with
maintaining relationships with other managers are necessary for avoiding losses
from contract infringement. This portrayal of the business environment and the
strategies employed to facilitate business transactions raises an important
92
question. Why must businesses rely on personal relationships that demand
extensive sacrifices of time and effort in their everyday transactions?
This chapter uses qualitative data to describe basic characteristics of networks
accessed by SME managers and owners.
1
To understand the role of networks, it
examines the types of relationships formed, with whom they are formed, and with
what intensity. The discussion illustrates the content of exchange facilitated in
networks, which is an essential first step in the empirical analysis of the link
between that content of exchange and businesses’ performance. Additionally, the
chapter investigates the costs that managers and owners must face when accessing
network ties.
6.1 Reform Progress
At the beginning of post-communist transition in the late 1980’s, the Socialist
Federative Republic of Yugoslavia was poised to successfully and quickly
implement key reforms of privatization and liberalization.
2
Prior to transition, the
country enjoyed a higher standard of living than most countries of the former
Soviet block and maintained a relationship with the Western world. One of the
key consequences of President Josip Broz Tito’s skillful balancing between the
Eastern and Western blocks was overexposure toward foreign creditors.
Eventually, in the early 1980’s Yugoslavia had to draw upon IMF reserves to
1
We will consider qualitative evidence in much greater detail in Chapter 9. In this chapter,
qualitative evidence is used to illustrate the nature of the overall business environment and the
challenges business overcome through the use of social networks.
2
Prior to 1991 the official name of the country was the Socialist Federative Republic of
Yugoslavia. Following the breakup, only Serbia and Montenegro remained from the original
constitutive republics and they formed the Federal Republic of Yugoslavia. Later, the country
changed its official name to Serbia and Montenegro. In 2006, Montenegro declared independence
and Serbia and Montenegro have been recognized by the United Nations as two independent
countries.
93
finance its ever-expanding foreign debt. The macroeconomic adjustments, which
included the cuts in federal expenditures, were directly related to disagreements
over budgetary resources among federal units. This internal struggle over
contributions and entitlements in the federal budget is one of the key factors
behind the disintegration of Yugoslavia and its eventual breakup into 5, and later
6, independent countries.
3
Yugoslavia had experimented with economic principles that were far from the
planned economic system of the Soviet Union but also different from the
principles of free market economics dominant in the capitalist societies. The most
well known characteristic of Yugoslav socialism was the principle and practice of
workers’ self-management. The basic premise of this system was that factories in
public ownership are best managed by their workers. The authority over the
resources and direction of the company was placed in the hands of the companies’
workers’ councils. Broader production goals were still loosely imposed by the
upper party echelons through the company’s director, but companies were free to
manage their own resources and investments according to consensus within the
workers’ councils. Some of the issues decided by the councils included the division
of revenues between salaries and investment resources, and the assignment of
vacation. Questions dealing with the technological aspects of necessary
investments were left for the technical staff which was consulted by the worker’s
council. The council had the least authority over the personnel issues. This aspect
of management was mostly controlled by local units of the political party, which
maintained a watchful eye through the assigning of the councils’ members.
3
Kosovo, a United Nations protectorate, is still part of Serbia. It’s eventual status will likely
include autonomy from Serbia but the terms of the agreement have not been finalized and the
future of Serbia as a candidate for the potential entry into the European Union is heavily
dependent on the successful resolution of the status of Kosovo.
94
The principle of self-management represented a unique contrast between, on
one hand, political control over economically productive units and, on the other
hand, proto-democratic principles within companies’ management structures. The
idea that workers had democratic control over the fruits of their own labor
instilled the concept of private ownership in the country. These foundations of
entrepreneurial spirit contributed to the formation of SMEs during the late 1980’s
and early 1990’s. However, the currently poor record of SME sector’s performance
in Serbia and Montenegro is troubling when juxtaposed with the history of worker
management and organizational elements conducive to the formation of private
ownership. The slow progress in SME growth is in large part a consequence of the
breakdown in privatization reforms that began with the disintegration of former
Yugoslavia in the early 1990’s . The decade that followed the onset of transition
in Serbia and Montenegro was strewn with violent ethnic conflicts, influx of
refugees from Croatia, Bosnia and Kosovo, economic sanctions, hyperinflation,
domestic protests, and unsustainable economic policies.
Figure 6.1 shows the drop in industrial output in Yugoslavia (Serbia and
Montenegro) since the beginning of transition. The early 1990’s drop in industrial
output is sometimes associated with the onset of transitional recession but the
more significant cause of this drop is the loss of industrial capacity from Slovenia
and Croatia who declared independence in 1991. The breakdown of economic
relationships with these two republics severed the flow of goods and capital that
supplied the intricate production systems also operating across the internal
borders. The shrinkage of internal markets presented the most immediate costs,
but long-term consequences arose when companies abandoned their investment
and development strategies. Violent wars in Croatia and Bosnia and Herzegovina
95
110
100
90
80
70
60
50
40
30
110
100
90
80
70
60
50
40
30
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001
Industrial output
Gross material product
110
100
90
80
70
60
50
40
30
110
100
90
80
70
60
50
40
30
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001
Industrial output
Gross material product
110
100
90
80
70
60
50
40
30
110
100
90
80
70
60
50
40
30
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001
Industrial output
Gross material product
Figure 6.1: Industrial output and GMP, FR Yugoslavia (1990–2001)
a
a
Source: Federal Statistical Office, Statistical Yearbook of Yugoslavia, 2001 in OECD (2003).
The calculation of gross material product (GMP) excludes services that do not directly
contribute to the production, distribution and repair of material goods. As a result, GMP is an
underestimate of gross domestic product (GDP). Index: 1990=100%.
between 1992 and 1995 exhausted government resources and hindered production
in Serbia and Montenegro. Economic embargo and sanctions authorized by the
U.N. in 1992 had an even more detrimental impact on Yugoslavia’s industrial
output. Over the course of the next 4 years, Serbia and Montenegro saw its
economy crumble as the country faced complete international isolation and the
freezing of its international assets. Domestically, the sanctions resulted in extreme
shortages, an increase in unemployment and poverty, and the near collapse of the
healthcare and social support systems. The economic and political isolation of
Yugoslavia was accompanied by poor macroeconomic policies that contributed to
the hyperinflation that ravaged the republics from May, 1992 to January, 1994.
According to some estimates, the inflation resulted in prices that were 2 billion
times higher at the end of 1993 than at the beginning of the same year (Sokić
2003). The inflation rates surpassed those recorded in the Weimar Republic in
96
1923. Price inflation reached its highest level between December, 1993, and
January, 1994, when prices increased at the rate of 60% per day. Economic
sanctions and hyperinflation provided the breeding ground for the grey economy,
which emerged as a response to shortages in goods and services, and especially
the demand and supply of foreign currency. Illegal activities such as the sale of
humanitarian aid, cigarette smuggling, counterfeiting of money, and drug and
human trafficking flourished.
The estimates of the cost to the economy of Serbia and Montenegro as a result
of the breakup of the former Yugoslavia vary widely with some coming in as high
as US$96.5 trillion (Sokić 2003). The estimates of the direct damage caused by
the imposition of U.N. sanctions are more precise and are said to equal US$36
billion (Sokić 2003; UNDP 2000). Further, costs of US$29.6 billion resulted from
the NATO bombings in Serbia during 1999 in response to Serbian actions in
Kosovo (OECD 2003). An approximately 18% drop in the GDP can be ascribed
to direct losses in production and investment resulting from the conflict in Kosovo
(OECD 2003).
In an attempt to distance itself from the disastrous political and economic
policies in Serbia, the leadership in Montenegro pushed to restructure the
constitutional relationship of Serbia and Montenegro. To curb the effects of
inflation, Montenegro adopted the German mark in 2000 and eventually
transitioned to the Euro for its official currency in 2002. These policies gave
Montenegro a slight head start in terms of returning to the reform agenda and
undertaking necessary transition reforms. In 2001, following the end of the violent
conflicts in neighboring states and the overthrowing of Slobodan Milosevic’s
dictatorial regime, Serbia finally returned to transition reforms. Also, following
97
the lifting of U.N. sanctions and the restructuring of its foreign debt, Serbia
finally received much needed financial assistance. Today, the GDP for Serbia and
Montenegro together is still around 50% of its 1990 level.
4
Further, the World
Bank estimates that the poverty level in 2000 was twice as high as that in 1990
(OECD 2003). Unemployment also remains high: both Serbia and Montenegro
have unemployment rates of around 27%. In addition to unemployment, the now
independent countries face problems with current account deficits and foreign
debt. On average, Serbia faces US$2 billion of current account deficits annually
(OECD 2003; UNDP 2000). The same figure is US$275 million for Montenegro,
which amounts to, on average, 25% of the GDP in each country. To cover the
current account deficits in the balance of payments, Serbia and Montenegro both
rely on foreign direct investment. This is accomplished through the sale of
state-owned enterprises to cover budgetary expenditures or current account
deficits, which is not sustainable in the long run and misrepresents current
problems in both economies. Privatization reforms that are the source of finance
for the states, although progressing in both countries, have had meager success,
especially with respect to large enterprises.
Forms of private ownership existed even before the breakup of Yugoslavia, but
they were limited to specific activities, number of workers, or size of profits.
However, experience with private ownership and self-management provided an
impetus for the entry of large numbers of new private firms and a significant
boost to the privatization of SMEs. At the start of transition, the republics of
Serbia and Montenegro each decided to follow a different strategy for the transfer
of state-owned enterprises to private ownership. Among its privatization methods,
4
Fig. 6.1 illustrates the level of GMP as opposed to GDP. See the figure’s notes for an
explanation of the difference between the two.
98
Montenegro adopted the issuance of mass vouchers to all legal residents over the
age of 18, the sale of shares though public tenders, and the issuance of free shares
to employees. Mass vouchers have already been issued and they were typically
either transferred to family members, invested in privatization funds, or exchanged
for shares in privatizing companies. The sale of companies through public tenders
has begun but so far has not attracted many interested investors. Serbia adopted
a model of privatization that focused less on restitution and more on a transparent
process of the sale of capital to domestic and foreign owners. Its privatization
model relies on the sale of up to 70% of firm capital through public tenders or
auctions. Only after 50% of the capital is sold, up to 30%, can be transferred to
company employees free of charge based on their rank and tenure. The problem
with the Serbian model of privatization is that it relies on complicated procedures
for determining the market value of companies and their future earnings potential.
This process not only opened up the possibilities for creative illegal manipulations
of capital valuations, but also completely crumbled in the 1990’s economic
downturn caused by U.N. sanctions and record hyperinflation. As a result, the
process of privatization had to be virtually restarted in 1994 and again in 2001 for
the companies that remained to be privatized.
In terms of the extent to which privatization has been accomplished, Serbia
and Montenegro are some the worst performers in SEE (Sanfey et al. 2004).
Together they have the lowest private sector share in GDP from any SEE or
Central European economy (Sanfey et al. 2004, Table 2.1). The pace at which the
privatization of state enterprises was undertaken has contributed largely to such
meager performance. Out of 1506 Serbian enterprises that were to be privatized,
901 were large enterprises, and only 181 had initiated the process by January,
99
1999 (Sokić 2003). Montenegro especially lags behind. First, only selected
companies were eligible to participate in mass voucher privatization. Second, the
state still retains controlling shares in 40% of companies and is planning to
auction these in international public auctions. Up until the sale of the state’s
telecommunications company “Telekom” in 2005, of all the enterprises with 250
employees or more, none were privately controlled (CEED 2002, 69).
5
The process
of privatization was undertaken in both countries with very little transparency,
which contributed to the workings of the grey economy. Some of the abuses that
occurred as a result of privatization schemes included management purposely
destroying company’s profitability in order to lower the sale price for a known
buyer, using credits from inflationary monetary policy (freshly printed money) to
purchase shares, establishing spin-offs for draining the value of the parent
company through the cost of inputs, making various forms of money transfers to
private accounts and private firms, and management purchasing companies for
private ownership with funds from public enterprises slated for privatization.
Since 2001, the eagerness to complete the transfer of ownership to private
parties has been revived and several large state-owned companies have been sold
to private investors. However, in both Serbia and in Montenegro the governments
have retained the oversight role in the transfer of ownership. Thus, they approve
which companies will be privatized and in what order. Also, they are involved in
debt restructuring, as well as the issuing of credits to assist the restructuring
efforts of companies that could potentially be sold to foreign investors. The
government has exuberantly touted the interest of foreign investors in the
companies of Serbia and Montenegro but such proclivities have not materialized
5
This means that the state is either the owner or it controls at least 51% of the shares
through state-owned social security funds.
100
to the extent expected. The result of privatization reforms is described today as a
“sea of dead enterprises, unemployed persons, and ghosts of promised investments”
(Sokić 2003, 461). Due to the delayed privatization reforms and the poor results
achieved with respect to the transfer of ownership most large enterprises have not
been restructured and privatized. Some are managing to continue to produce but
many sit idle as they are awaiting financial appraisals and offerings from potential
investors. As a result, attempts by those enterprises to create spin-offs, or to
subcontract and outsource have been minimal. Consequently, relationships
between the large, mostly state-owned, enterprise sector and the SME sector have
not taken root to any noticeable extent. Thus, the SME sector has not been able
to take advantage of vertical integration into supply and production chains which,
as described in Chapter 2, is highly beneficial for SME growth.
6.2 Business Environment
Qualitative evidence collected from SMEs in Serbia and Montenegro offers a
picture of the overall conditions in which businesses operate. From the
perspective of owners and managers, the most problematic barriers to doing
business are the inability to enforce contracts and the frequently changing and
inadequately administered business regulations. In terms of enforcement, courts
in Serbia and Montenegro are poorly evaluated in both their effectiveness and the
overall cost of dispute resolution. They are slow in processing and deciding cases,
which increases the cost of administrative and legal assistance. In addition, judges
and other court officials are generally considered vulnerable to influence from the
parties involved in the case, amplifying the public distrust in the courts. As a
result, businesses lack incentives to rely on courts to enforce agreements with
101
other parties. The inadequacy of laws stems from the incomplete specification of
their scope and frequent changes, including occasional retroactive application of
implemented legislation. In both Serbia and Montenegro managers expressed
concern over the number of procedures that businesses have to follow and note
that obtaining licenses and permissions drains considerable efforts and finances.
An illustrative example of the burdensome regulation comes from the World
Bank’s “Doing Business Database.”
6
Among the 175 countries for which the
database tracks the cost of obtaining licenses necessary to build a single
warehouse, Serbia and Montenegro have the eighth and second highest figures
respectively.
7
The issue of numerous and onerous administrative procedures has
other closely related implications for the day-to-day business environment in
Serbia and Montenegro. For example, no efforts have been made to streamline the
procedures for obtaining common documents and to eliminate superfluous
bureaucratic layers. There is a significant absence of automation because
computer- and internet-based access to information or application instructions is
absent. Managers and owners are thus forced to interact frequently with
bureaucrats whom they perceive as opportunistic, inefficient and unprofessional.
The lack of specificity and automation in formal procedures allows bureaucrats to
interpret and enforce procedures in whatever way they find most appropriate.
This ability to supersede formal policy and decide unilaterally how its elements
will be implemented empowers individual officials to apply the law selectively and
in accordance with personal interests.
6
Excerpts from the database are available at:
http://www.doingbusiness.org/ExploreTopics/DealingLicenses/
7
The average cost of obtaining licenses in Europe and Central Asia is 564.9% of per capita
income, compared to 5869.20% in Montenegro and 1946.70% in Serbia.
102
The lack of specificity and automation in formal procedures and the selective
application of law have resulted in personalization of institutions. From the
perspective of market actors, government bodies cease to represent impersonal
embodiments of the written rule of law, and become a domain of the individual
bureaucrat who is empowered to selectively enforce the rules. This blurs the
boundary between public and private spheres since it allows public officials to
extort payments from business managers and owners while the latter can bargain
with officials on the extent, nature, and the cost of implementation. To
compensate for the structural weaknesses in the business environment, managers
and owners therefore often turn to personal relations.
6.3 Content of Exchange
The interviews conducted with managers and owners of SMEs in Serbia and
Montenegro reveal that many aspects of doing business necessitate the use of
relationships. Businessmen access two qualitatively different types of networks,
namely networks built by personal relations (lični odnosi) and networks of
connections (veze). At any one time, an economic actor may have access to
personal relations, connections, or both. The two types of networks can be
differentiated using two specific network dimensions: 1) the strength of the
relationship in terms of the extent to which individuals feel committed to one
another, and 2) degree of openness to outsiders. In discussing these basic
differences between personal relations and connections, the content of exchange
facilitated through networks will become evident.
103
6.3.1 Personal Relations
Ties comprising a network of personal relations varied in strength. Interviewees
identified family members, friends, previous and current business partners, and
acquaintances as individuals with whom they maintain personal relations. These
are primarily networks of contacts that allow individuals to distribute information
about business actors and their activities. What one does, how well one’s business
is doing, and how one treats the employees are examples of information
transferred via personal relations. This type of information assists individuals in
finding other actors who can participate in mutually beneficial exchange. For
example, an exclusive distributor of bathroom supplies uses his friends to put him
in touch with owners and managers of restaurants and hotels with whom he could
potentially form long-term supply contracts. All of the businesses also reported
the need to use personal relations to obtain suggestions on individuals they should
hire, or information about individuals they are thinking about hiring.
The second role of networks of relations in business activities arises from their
ability to support interpersonal trust. Even when the effectiveness of courts is in
question, economic actors will take part in exchange if they have some reason to
trust that the other party will abide by the terms of the agreement. Essentially
networks spread information about an actor’s reputation that other individuals in
the network can use when deciding with whom to trade or, once a potential
partner is identified, to verify his or her reputation. Information offered by a
friend, relative or business partner about a specific actor’s reputation serves to
reduce the expected probability that he or she will renege on the terms of the
agreement. Since no one would choose to enter into agreements with businesses
that have questionable dealings, loss of reputation serves as an informal and
104
multilateral sanctioning mechanism. For example, the manager of a security
company in Serbia had obtained reliable information from a competitor that a
specific bank had refused to make payments for services it purchased. In order to
avoid the risk of entering into a dispute over nonpayment, the security company
simply refused to provide its services to the bank. The reputation mechanism
built into network transactions and the ability of network members to sanction
businesses that do not abide by their terms of contract are critical elements of a
multilateral enforcement mechanism.
Personal relations can also play the role of a unilateral enforcement mechanism
by assisting managers in resolving conflicts before they are forced to consider legal
action. Managers involved in an established relationship have a mutual interest in
addressing problems in production, supply, or payment in order to prevent
relationship breakdown and the potential compromise of their reputation in the
network. Parties can trust one another to invest efforts into finding alternative
solutions and obtain feedback on proposed strategies in order to meet the original
terms of the agreement. Overall, personal relations serve as an added assurance
that both parties are committed to an agreement.
6.3.2 Connections
Similarly to relations, connections include ties ranging in strength from relatives
and close friends to business partners and acquaintances. Where they differ from
personal relations is that they facilitate the exchange of favors, or usluge.
Connections are said to exist among individuals who are willing to do favors
employing personal resources, and those of their contacts, to help others get
ahead. These resources can include information, wealth, or access to positions of
105
Connections
Established
relationships rooted
in reciprocity,
which facilitate
exchange of favors.
Personal Relations
Relationships based
upon repeated
exchange, which
produce interpersonal
trust as a form of
contract enforcement.
Figure 6.2: Types of networks
authority. The value of the resources made available entitle one to request a return
favor at some future point. Thus, the issue that separates personal relations from
connections is not the nature of the underlying tie but the use of the relationship,
and the extent to which actors are committed to one another through exchange.
Connections can be useful in all of the ways associated with the use of personal
relations, namely for obtaining and distributing information, resolving disputes or
ensuring that the terms of contracts are respected. However, connections can
facilitate exchange of resources and information beyond those accessible through
personal relations. Thus, while not every relationship is considered a connection,
every connection is a personal relation. Fig. 6.2 summarizes this point by showing
the network of connections as a subset of network of relations.
Use of connections identified by SME managers and owners can be classified
into two categories: 1) obtaining privileged access, or 2) securing preferential
treatment. The most notable examples in the first category involve the provision
106
of access to valuable resources, finances, or public contracts. Managers reported
using connections with bank officials to secure favorable loan conditions such as
lower interest rates, higher credit lines or a longer term for the loan. The most
commonly discussed cases of using connections to obtain access involve the
issuance of public contracts. The low availability of private investment and
general lack of liquidity have increased the demand for contracts with state-owned
companies and private companies where the state has retained a controlling
majority. However, jobs and contracts with such companies are exclusively
obtained through the use of connections. In cases where larger contracts have to
be assigned through a public auction system, connections can be used to ensure
that one’s firm is ranked highest among all applicants or to ensure that the terms
of the auction are specified so narrowly as to match only a specific company’s
terms of offer.
Respondents also described how connections can be used to receive preferential
treatment from judges, bureaucrats, and other firms. First, they noted that
having a connection with a judge can enable one to influence the outcome of a
court proceeding. For the same reason, today’s best lawyers are reported to be
former judges who exchange favors with their former colleagues—or current
judges. Second, businessmen noted that connections are useful for breaking
through multiple layers of regulation and licensing that serve as a form of
rent-seeking. Reported cases vary in the nature and extent to which they
circumvent or violate the law, but they all involve exchanging favors with contacts
in local and regional bureaucracies to avoid dealing with outdated procedures and
regulations. Similarly, connections can be useful in helping businessmen avoid
dealing with bureaucrats who demand side-payments in exchange for rapid and/or
107
fair processing. In general, connections with individuals in the bureaucracy or
other positions of power enable the procurement licenses, documents or privileged
information that must otherwise be obtained through complicated, expensive and
burdensome procedures. Thus, connections help businessmen accomplish things
more quickly, cheaply, and with fewer obligations, especially where procedures are
not clearly specified, are not automated, or are selectively enforced.
8
6.4 Network Boundaries
The second distinction between personal relations and connections involves the
degree of openness to outsiders. Personal relations are easily formed through
social interactions, or membership in associations, community groups, and clubs.
All types of ties, regardless of the setting in which they are formed, can carry
valuable information. New market entrants face difficulties in establishing their
reputation when there is no record of past business transactions to speak for their
trustworthiness. However, information obtained about one’s social dealings can
serve as a proxy for his or her trustworthiness in business transaction.
Businessmen claim that this is always a rewarding strategy since “all one has to do
is ask around” and “everyone knows everyone else.” Networks of personal relations
8
Interviewed business managers and owners were more open about their use of personal
relations than connections in their every day business activities. The use of connections
sometimes violates the norm of fairness in terms of equality of opportunity and can be
associated with the stigma of improper use of influence to advance one’s position. The
reluctance of some businessmen to admit to the use of connections was anticipated due to
similar findings in research on Russian informal relationships or blat (Ledeneva 1998). To
overcome this problem, interviewees were not asked directly about their use of connections but
were encouraged to talk about the strategies they employed to overcome business problems and
the type of assistance they received from friends and business associates. They were also
encouraged to comment on the use of connections by other businesses in their sectors of activity.
108
thus have fluid memberships and porous borders where all types of relationships
are used to obtain useful information for conducting commercial transaction.
Networks of connections are comprised of individuals who will risk personal
resources in order to help someone else get ahead. Such commitment to other
actors entails the possibility that the recipients of favors have to reciprocate in
some way for the resources and access they have been granted. The return favor
does not have to be comparable in terms of the resources and effort that have
been invested, but it is considered the norm that the individual must be willing to
return the favor when called upon to help.
This network of individuals who “owe it” to one another resembles Ellickson’s
(1991) community of ranchers and farmers in Shasta County who maintain
“mental accounts” of debts incurred as a result of animal trespass incidents (55).
The distinguishing characteristic of networks of connections among businessmen
in Serbia and Montenegro is that mental accounts of debts and credits towards
others do not arise as a result of violations of community norms but as a result of
deliberate assistance to secure preferential access or treatment. When an
individual is asked to perform a favor for someone else, they can deny the request
unless 1) they have been assisted by that partner in the past, or 2) they could be
assisted by that partner in the future because they expect the relationship to
continue, and the partner has access to contacts and resources otherwise not
available.
9
Another characteristic of networks of connections that separates them from
the provision of private order described by Ellickson is that in the case of
9
In some circumstances, an individual will risk resources to help others without hoping to
receive a favor in return, but such connections typically only exist between family members and
close friends where the sense of commitment is rooted in emotional ties and the culturally
instilled norm to always assist one’s kin.
109
connections, “mental accounts” are highly transitive. To obtain privileged access
to a particular resource an individual is not restricted to individuals he or she
knows directly. Instead, one can find an intermediary who “owes them one” and
who will then request a favor from another individual capable of securing access
to the desired resource. Networks of connections thus represent multi-level and
crosscutting links of indebtedness which can easily extend beyond the boundaries
of a geographic region, or a closed community of a specific ethnic or religious
orientation. However, the more expensive or unusual the favor, the more difficult
it is to find someone who can help procure the desired resources or benefits.
The most lucrative of all the connections are those with individuals who enjoy
a direct or indirect form of political authority. Direct political authority belongs
to individuals holding positions of authority in the various levels of government.
Individuals who hold management positions in state-owned enterprises, and those
who have close personal or political ties to the individuals in government positions
enjoy an indirect form of authority. This latter group also includes managers and
owners who previously held positions in the government but have since moved to
the private sector most likely to take advantage of business opportunities. Ties
with these individuals can secure access to the most valuable and exclusive
resources such as import and export licenses, contracts with state-owned
enterprises, rights to extract national resources and rights to use public property
at no cost. Special treatment that can be secured using these connections
includes, for example, tax exemptions, reductions in requirements that need to be
fulfilled for a specific type of business, and cancelation of work safety inspections.
Connections with individuals who enjoy some form of authority over state and
public resources are the most lucrative but also the most exclusive of all ties.
110
They are reserved for relatives, friends and long-term business partners. In sum,
the extent to which individuals feel committed to one another, due to the
exchange of favors, makes connections more difficult to access and more closed for
outsiders, compared to networks of relations.
6.5 The Cost of Networking
In the case of Serbia and Montenegro, there is an intricate relationship between
the social milieu in which businesses operate and their needs in conducting daily
operations. Businessmen utilize the two previously discussed types of
relationships. The relationships forming both types of networks range from family
and friends, to business colleagues and partners. However they differ in terms of
the content of exchange they facilitate. Personal relations facilitate the spread of
information about other business actors, lower the risk of cheating, and help
prevent relationship breakdown that entails the use of expensive and inefficient
courts to settle potential disputes. In effect, they serve as informal means of
contract enforcement. In addition, connections facilitate the exchange of favors
that can procure preferential treatment or exclusive access to goods, services, and
information. Personal relations and connections also differ in terms of the
intensity of commitment characterizing the ties that form them. In the case of the
latter, exchange of favors requires a sacrifice of personal and business resources
and is thus reserved for much more intense, committed and reciprocal
relationships which are less easily accessed by outsiders.
The use of network relationships for contract enforcement or access to
information and capital is not without costs. Different types of costs have been
identified with respect to the use of relationships in business activities. The most
111
elemental and commonly noted costs involved in using relations and connections
stem from the need to search for contacts. Resources must also be devoted to
building relationships, or if the relationship precedes the commercial contact, to
maintain personal and business ties. Individuals must spend time and financial
resources visiting each other’s factories and offices, or arranging to meet in social
engagements, such as parties and vacations, in an attempt to become more
familiar with each other’s business practices and to gather information about each
other’s trustworthiness. An illustrative example of this cost comes from the
manager of a publishing house in Montenegro who donates to the local water polo
team not only as a gesture of charity but also to expand and maintain his network
of contacts. Building and maintaining connections is even more costly because
individuals must be identified who possess the authority or the resources to
accomplish something and then direct and indirect ties to those individuals must
also be identified. Making connections is more difficult than identifying key
personal relations since the identified individuals must also be willing to risk their
authority and resources to accomplish the necessary task. Much like the
publishing house manager, SME owners and managers are aware that social
encounters with business partners and their contacts, including ski trips and
summer vacations, are the costs of expanding a network of personal relations and
developing them into more lucrative connections. Time and financial resources
devoted to these activities are a necessary cost of using networks.
With respect to connections, costs are also related to the potential need to
return a favor that has resulted in exclusive access to goods and valuable
information. If the individual is called upon to reciprocate a favor, and he or she
refuses, the relationship between the two would be severed and the reputation of
112
the individual refusing to help would be tarnished in the network. The costs or
the timing of the return favor cannot be predicted in advance. Also, a return
favor does not necessarily imply an expense of time or resources in assisting
others. Individual businesses can be asked to abort certain activities, such as an
expansion of service or sale of products to a particular business. In the extent
that reciprocity can limit individual action, it creates opportunity costs. The
benefits that could have been realized through such transactions, had the business
not been asked to reciprocate by aborting them, must also be considered as costs
of accessing favors and resources through connections.
Reciprocity often entails that the individuals trade with specific actors to the
exclusion of other, nonnetwork members. When businesses maintain specific
relationships to the exclusion of others, commitment can produce costs, as better
offers outside the network are disregarded. Excluding nonnetwork members from
exchange is not necessarily a cost associated only with connections; even the use
of personal relations can result in opportunity costs. Individual businesses may be
willing to sever unsuccessful business relations provided they can find alternative,
better terms of trade. When information about the reputation of businesses
making alternative offers cannot be confirmed, businesses will be reluctant to
switch to new partners. Thus, even when reciprocity does not dictate that a
business trade with specific firms alone, the lack of information about other
options and distrust in nonnetwork members results in opportunity costs.
Working with established, overlapping relations could also block individuals from
accessing new products, technologies, and information outside the network, which
can limit opportunities for learning about and responding to changing market
conditions (Sedaitis 1997). Overall, refusal to trade with unfamiliar businesses
113
outside one’s network of relations or connections can result in costly reductions in
productivity and efficiency with negative implications for economic performance
(Cook, Rice, and Gerbasi 2004).
Interview responses suggest that managers and owners in Serbia and
Montenegro take advantage of all ties regardless of whether the other party is
one’s relative, friend, acquaintance or business partner. Also, all ties, whether
formed in a social or business setting, are useful for doing business. Depending on
the content of exchange and the degree of commitment on the one hand, and
openness to outsiders on the other, a qualitative difference exists between
networks of personal relations and networks of connections. This difference
becomes important when we consider the roles that the two types of networks
play. Networks of personal relations are primarily used to transfer business related
information and provide useful contacts for business transactions. In addition to
these roles, networks of connections are used to secure privileged access to
valuable resources and to obtain preferential treatment that can translate into
positive economic outcomes. Connections and personal relations also differ in
terms of direct and indirect costs that businesses must pay to access the resources
embedded in network ties. The costs of maintaining and building contacts are
common to both types of networks. However, the use of connections entails the
reciprocal exchange of favors where both the return favor and the commitment to
specific exchange relationships can present significant costs for business activities.
The implications of network roles and the potential costs they create must be
further investigated to fully understand the contribution of social networks to
economic activity in transition.
114
Chapter 7
Networks and Investment: Exploring
the Determinants of Profit
Reinvestment
The expansion of the SME sector’s impact on the economy through the growth of
individual enterprises is assumed to be a critical component of the success of
economic transition. Studies of the SME sector have used the underlying motives
for firm formation to suggest that some businesses simply do not wish to grow.
Taking into consideration the large number of micro enterprises that have sprung
up in transition economies and the fact that many of them have been established
with the goal of securing an alternative source of income, some authors have
distinguished between enterprises established for purely entrepreneurial motives
and those established for less productive, proprietary motives (Scase 2003;
Dallago and McIntyre 2003).
Entrepreneurial SMEs take advantage of business opportunities with the
potential for realizing profits and sustaining future business growth through
115
reinvestment of gains and expansion of business activity. On the other hand, the
motive of proprietorship encourages the formation of SMEs centered on securing
property ownership that can yield rents. The rents provide owners with a desired
level of income. The proprietor does not intend to use the gains of his or her
activity to invest in long-term growth of the enterprise but only to sustain a
specific level of consumption. This group of SMEs often includes enterprises
established to extract government subsidies, to launder money, or to facilitate
asset-stripping (i.e. the transfer of state assets into private SMEs owned by
former state-enterprise managers and politicians) (Gustafson 1999; Shleifer and
Vishny 1998).
If growth of individual businesses is critically important for transition
performance, then entrepreneurial businesses have a greater role to play than
those established for generating rents alone. What separates entrepreneurial
businesses from proprietorships is the degree to which they reinvest profits. Thus,
to understand the factors that influence the growth of individual SMEs that can
contribute to the growth of the economy, we ought to focus our attention on the
factors that encourage businesses to reinvest. By analyzing SMEs reinvestment,
we are effectively focusing on entrepreneurial SMEs.
7.1 Determinants of Profit Reinvestment
Studies of determinants of firm growth have coalesced around two major strands.
One group of studies has made the case that access to external sources of finance
and financial market development are critical components of firm growth
(La Porta et al. 1997; Demirgüç-Kunt and Maksimovic 1998) Others have argued
that the security of property rights instead is the key determinant of growth
116
(Acemoglu and Johnson 2005; Johnson, McMillan, and Woodruff 2002b). For
example, Brunetti, Kisunko, and Weder (1998) use aggregates of firm-level
measures for “credibility of rules.” They find a strong and stable relationship
between investment levels and the predictability of judicial enforcement, as well as
the security of property rights and persons. In his recent book, De Soto (2000)
identifies property rights as an essential vehicle for allowing physical assets to
generate values that can be employed in productive activities. The ability to
reinvest capital productively is the central contributing factor to economic growth
(De Soto 2000).
The two strands of literature arguing for different sources of firm growth may
not be incompatible. Johnson, McMillan, and Woodruff (2002b) and Cull and Xu
(2005) have suggested that for some countries, such as those in the early stages of
post-communist transition, property rights may in fact be more important than
sources of finance. The latter may come to play a greater role as transition
progresses. The problem for within-country studies, however, has been
distinguishing the effects of property rights and access to finance in examining the
determinants of investment and growth.
Johnson, McMillan, and Woodruff (2002b) used survey data from
post-communist countries to examine whether the perceptions of (in)security of
property rights affect the rate at which businesses reinvest their profits. Statistical
regressions indicated a significant and negative effect of the insecurity of property
rights on the reinvestment of profits. Entrepreneurs with the least secure property
rights reinvest forty percent less than those with the most secure property rights.
In addition, when the security of property rights is held fixed, businesses with
access to credit do not opt for more investments, suggesting that the security of
117
property rights is not only a necessary but also a sufficient condition for private
sector investment in general.
Using a proxy based on several indicators of investment, Frye (2004) took the
next step in the analysis to determine the most important components of secure
property rights.
1
The underlying assumption was one established by Johnson,
McMillan, and Woodruff (2002b), namely that investments take place when
decision makers feel that the state actors and individuals cannot subvert their
rightful ownership of property. Frye (2004) concluded that the effectiveness of
courts, especially the perceived ability to use courts to defend one’s interests
against government actors, had the most significant impact on managers’ decision
to commit to new capital investments in the near-term future. Johnson,
McMillan, and Woodruff (2002b) also found that entrepreneurs’ perception of the
effectiveness of domestic courts positively affects the reinvestment of profits.
Bringing the two sets of results together implies that individuals who are
confident in courts believe that the contracts protecting their property from the
state and other actors can be legally enforced and are therefore willing to invest
more. It can be concluded then that the belief that contracts can be enforced is
what produces the positive relationship between confidence in courts and security
of property rights. Hence, individuals who believe in the effectiveness of courts are
more comfortable accepting contracts that include less insurance against risk.
Several studies have found that individuals with higher confidence in the
effectiveness of courts are in fact more inclined to extend trade credit—allowing a
customer to pay the bill upon delivery or with a delay—or ask for lower
1
The indicators of investment used are: whether the business had constructed a new building
in the last two years, whether in the same period the business had extended trade credit to its
partners, and whether the business is planning to commit to a significant capital investment in
the coming year.
118
prepayment in transactions (Frye 2004; Johnson, McMillan, and Woodruff 2002a).
Therefore greater confidence in courts translates into greater confidence in
contracts and the security of property rights, which ensure higher reinvestment.
The quality of courts and the confidence in their ability to enforce contracts is
rather low in SEE transition economies, as discussed in Chapter 2. Instead,
businesses rely on personal relationships and trust as sources of confidence in the
contracts they form with other business actors. Qualitative reports from SME
managers and owners showed that social networks can play the role of an informal
mechanism for contract enforcement (Chapter 6). Empirical evidence from
Poland, Romania, Russia, Slovakia and Ukraine, where courts are not as
sophisticated as in the developed world, inter-firm and social relationships are the
main form of ensuring contract enforcement available to businesses (Johnson,
McMillan, and Woodruff 2002a). Existing empirical evidence, however, is less
successful in illuminating the relationship between the use of informal and formal
forms of contract enforcement.
Hendley, Murrell, and Ryterman (2000) found that relational contracting and
third-party enforcement based on a reputation mechanism are the dominant
strategies for enforcement in Russia, but that they are complementary to the use
of formal contract enforcement through the courts. However, Hendley and Murrell
(2003) showed, in the case of Romania, that using relationships for contract
enforcement is negatively correlated with the use of courts, suggesting that the
two are not complementary strategies. Using the same survey evidence as
Johnson, McMillan, and Woodruff (2002a), McMillan and Woodruff (2000) found
that social networks and effective courts are substitutes in contract enforcement
while courts and business networks are complementary to each other. Woodruff
119
(2004) offered a model demonstrating that the relationship between interpersonal
or informal mechanisms for contract enforcement, and formal contract
enforcement through the courts, is theoretically ambiguous. However, in his
empirical analysis, he took the social and business networks as defined by
McMillan and Woodruff (2000) and put them together to show that network
relationships are complements to courts as strategies in contract enforcement.
The existing literature successfully documents the use of reputation based
strategies for contract enforcement, but the relative importance of the use of
courts on the one hand, and the relational mechanisms provided within networks
on the other, remains somewhat ambiguous. What has not been addressed,
however, is the bigger question of how the use of both types of enforcement
mechanisms affects reinvestment and business growth. In other words, it is not
clear to what extent business activity is encouraged by the availability of informal
means of contract enforcement and how business growth is affected by the precise
relationships between formal and informal means of contract enforcement.
The question of the effect of informal means of contract enforcement on
business activity is further complicated by the fact that informal means of
contract enforcement can present serious additional costs in terms of productivity
and efficiency. Interviews with SME managers and owners in Serbia and
Montenegro have shown that utilizing close relationships with trusted partners for
contract enforcement has potentially negative consequences for firm performance.
Interpersonal trust formed in relationships can breed distrust in individuals who
are not network members. Given the low confidence in the courts, if no member of
the network can vouch for the reputation of a certain actor, the rest of the
members in the network will be reluctant to enter into contracts with them.
120
Essentially, what makes networks viable is that they connect individuals who,
while distrustful of others in general, rely on established relationships to reduce
the risk of misconduct on the part of a business partner. As a result, businesses
may be willing to pay higher prices in order to continue using existing
relationships (Johnson, McMillan, and Woodruff 2002a). We have also seen that
working with existing relationships limits access to new information and resources,
restricting organizational flexibility and learning. These negative consequences of
utilizing existing relationships must not be forgotten in assessing the role of
networks for business investment. Hence, given the positive and negative aspects
of using networks as informal means of contract enforcement we must examine the
effect of informal enforcement strategies on SME reinvestment of profit.
7.1.1 Modeling Reinvestment
Johnson, McMillan, and Woodruff (2002b), hereafter JMW, aim to disentangle
the effects of property rights from those of access to finance in determining the
demand for reinvestment. In particular, they are interested in examining whether
secure property rights are a necessary and sufficient condition for firms’
reinvestment. They use an additive index of property rights to analyze a model of
profit reinvestment that is derived from the latent variable model of demand for
investment funds (internal and external to the firm). They consistently find
results that the security of property rights and trust in courts are sufficient and
necessary conditions for investment. It is their model of profit reinvestment that
provides a strong and tested starting point for our analysis of the role informal
contract enforcement plays in reinvestment.
121
In the JMW model, a firm makes decisions about investment and borrowing
simultaneously. Firm’s demand for total funds to be invested is given by
I
d
= I(π,s,r
i
,r
l
), (7.1)
where π denotes expected profits, s indicates the value of profits that will be
extracted by corrupt bureaucrats and other criminals, r
i
is the rate of return
businesses can earn by investing their profits internally, and r
l
is the interest rate
that the business must pay for borrowed funds. This total demand for investment
funds can be satisfied from internally generated funds or from funds borrowed
externally. The demand can thus be stated as:
I
d
= R+L
d
, (7.2)
where R refers to reinvested earnings and L
d
refers to the business’s demand for
loans. The central assumption in this model is that the cost of internal funds, r
i
,
is lower than the rate at which funds can be borrowed externally, r
l
. The
difference between r
i
and r
l
exists due to information asymmetry. External
lenders demand higher interest rates to insure themselves against risk because the
true quality of the investments is not known. The difference between the two
rates has been confirmed empirically in developed economies and is expected to
be even higher in transition economies where investment uncertainty is much
higher. The difference creates a pecking order in which internal funds are
exhausted before external funds are brought in for investment. Hence, the model
treats the demand for investment of internal funds independently from the
decision whether to invest external funds. The pecking order hypothesis allows us
to investigate the determinants of reinvestment using econometric models that do
not take into account the demand for external finance.
122
Let the amount that a business is willing to invest from its internal funds,
which can be less than or equal to total profits, be given by E
i
. The independence
between the decisions to invest internal and external funds implies that
I
d
=







R if I
d
≤ E
i
,
E
i
+L
d
if I
d
> E
i
.
(7.3)
And finally a restatement of the demand for investment funds gives the following:
R =







I(π,s,r
i
) if I
d
≤ E
i
,
E
i
if I
d
> E
i
.
(7.4)
We investigate the determinants of the firm’s decision to reinvest its profits as
given in the equation (7.4). Industry controls can capture the differences in r
i
,
but the estimates of industry profits and the security of property rights are
provided in econometric analysis.
The model above provides a natural starting point for investigating the
determinants of the decision of SMEs to reinvest profits within a specific country.
Given the strong results obtained by JMW, this model offers a successful strategy
for isolating the impact of access to external finance from the factors impacting
the rate of reinvestment. However, we introduce additional variables into the
model in order to investigate the role of informal contract enforcement. First,
proxies for the security of property rights are analyzed in terms of two
independent constituent factors. Following Acemoglu and Johnson (2005) and
Cull and Xu (2005), the impact of property rights is separated into: 1) elements
indicating how safe ones’ property is from expropriation (both by the government
and third-parties), and 2) elements indicating the quality of contract enforcement.
Using an instrumental variable approach, Acemoglu and Johnson (2005)
concluded that institutions providing protection against the expropriation of
123
Security of property rights
Threat of
expropriation
Contract
enforcement
Informal
contract
enforcement
Formal
contract
enforcement
Figure 7.1: Components of security of property rights
property by the government and powerful elites are a more direct determinant of
growth than the institutions accessed for contract enforcement among
third-parties. They suspected that the impact of contracting institutions is
lessened because individuals can utilize formal contracts and informal contract
enforcement to compensate for weak contracting institutions but they are less able
to use these resources for lowering the risk of expropriation by the government.
To pursue this claim further, in addition to distinguishing between the threat of
expropriation and the quality of formal contract institutions, informal means of
contracting are also considered within the model here. To capture the impact of
informal contracting, as an element of the security of property rights, contract
enforcement needs to be further broken into measures of formal and informal
contract enforcement as illustrated in Fig. 7.1.
7.2 Data and Measures
The determinants of profit reinvestment are investigated here using the dataset
from Serbia and Montenegro introduced in Chapter 5 and described in more
124
detail in Appendix B. The dataset contains 274 responses to a questionnaire
mailed to a sample of SMEs with 250 and fewer employees. In almost 97% of the
cases, the responses were provided by the manager and/or owner; the remaining
3% came from founders or high-level managers. Among the businesses in the
dataset, only 7.66% were formerly a state-owned enterprise and of these, only 19%
reported that the state still has a controlling part in the business (1.46% of the
total sample). Having a sample composed of mostly privately owned and
controlled businesses is preferred in this analysis. First, it reflects the true SME
population since almost all of the remaining state-controlled enterprises have over
250 employees and thus are not of interest here. Second, by focusing our attention
on privately owned and controlled businesses, we effectively avoid the problems
that arise when working with data from state-owned enterprises. Some of the
common problems include the misrepresentation of profits, number of employees,
reinvestment rates and the role of political pressure in reinvestment decisions.
Over 90% of businesses in the sample are registered as limited liability
companies with the remaining 10% spread over joint-stock and joint-venture
businesses, partnerships and sole entrepreneurships. In terms of the year in which
the business began to operate, almost 60% of businesses in the sample were
opened between 1989–1994, 16.06% were opened between 1995–2000, and 16.42%
were started after 2000. The sample also includes 8.03% of businesses that were
started before 1989, the officially accepted beginning of transition. Lastly, in
terms of the main business activity, the sample overrepresents the manufacturing
sector (42.22% of the sample) and underrepresents the trade sector (23.71%). The
remaining categories of primary business activities for the SMEs in the sample
include construction (16.67%), food industry (7.41%), and services such as
125
tourism, telecommunications, financial consulting, publishing, and information
technology (10%).
For the whole SME population, Table 2.3 in Chapter 2 suggests that the
distribution in trade and manufacturing sectors should be almost the reverse of
what is reflected in the current sample. The underrepresenting of the trade sector
and the overrepresenting of the manufacturing sector is a result of the sampling
strategy and the choice of the SME database from which the sampling frame was
constructed. To ensure that the sample captured as many legitimate businesses as
possible, the sampling frame was selected from businesses that had provided a
phone number at the time of registration with the Chamber of Commerce and
had reregistered their businesses in accordance with a new 2002 law.
2
This
sampling strategy eliminated a lot of micro businesses from the sampling frame
and especially those involved in retail trade—the largest of all business categories.
Due to low capital and labor requirements, the retail trade sector has the highest
rate of entry and exit and includes a high number of illegitimate businesses set up
for tax purposes. The category also includes most of the street vendors and small
resellers which are not particularly interesting for the analysis of profit
reinvestment and firm-level growth. Thus, limiting the number of retail trade
businesses, the overall trade sector was undersampled and as a result, the largest
share of businesses was captured from the manufacturing sector. To at least
partially, account for activity-related differences in reinvestment determinants, the
following analysis will make use of industry controls.
2
The law from 2002 required all operating businesses to register their activities with the
Chamber of Commerce and to pay a reregistration fee.
126
7.2.1 Profit Reinvestment
Incorporating the lessons from JMW, the survey did not ask the respondents to
specify the exact percentage of the profits reinvested during the previous year.
Instead, the question “How much did you reinvest out of profits during 2005?”
offered respondents the following choice of ranges: 0%, 1–10%, 11–25%, 26–49%,
50–75% and over 75%. By transforming an open-ended question into preset
categories, the question sacrificed some precision that would have been contained
in a continuous variable. The goal was to reclaim in terms of increased validity
what was lost in precision. Previous studies have concluded that much higher
response rates can be obtained when questions about profits and investments are
not answered in nominal values but in terms of ratios. In this case, survey
respondents were offered categories of ratios from which to choose in describing
their reinvestment rate and the level of after-tax profits as a percentage of annual
sales.
Table 7.1 shows the firms’ average profit after taxes as a percentage of sales,
and profit reinvestment as a percentage of profits after taxes. For comparison,
Table 7.1 also includes figures obtained by JMW for other transition economies.
The average profits of 27.4% reported in Serbia and Montenegro are closer to the
results from Russia than the more advanced transition economies of Central
Eastern Europe. This adds further evidence to the JMW claim that the rate of
profit decreases as transition progresses. During transition, as improvements in
market institutions eliminate the instabilities that can result in disequilibria,
increased competition contributes to the lowering of overall profits. The reported
rate of reinvestment for businesses in Serbia and Montenegro (42.08%) is also in
line with the rates reported in JMW.
127
Table 7.1: Profits and reinvestment
Serbia and
Montenegro
Measure Poland Slovakia Romania Russia Ukraine
Last year’s profit after taxes,
percentage of annual sales
15.5 9.9 5.7 12.9 20.6 18.0
Profit reinvestment, percentage of
profits after taxes
42.08 52.6 42.3 52.8 38.8 29.6
Estimated industry profit rate
after taxes
14.26 11.1 10.0 13.4 17.6 14.3
Tax payments as percentage of
sales for firms in industry
25.89 15.5 16.4 17.2 26.9 28.0
Number of firms 274 303 308 321 269 270
Notes: Profit reinvestment as a percentage of profits excludes firms with zero or negative profits. Respondents
choose categories to indicate their profit after taxes as a percentage of sales and the rate of reinvestment.
Profit after taxes and profit reinvestment as a percentage of profit after taxes were calculated using the
midpoint for each category. The results for Poland, Slovakia, Romania, Russia, and Ukraine are reported in
Johnson, McMillan, and Woodruff (2002b).
In the survey, entrepreneurs were also asked to estimate the profit rate and tax
payments as a percentage of sales for firms in their industry. Both of these
resulting figures were comparable to figures JMW obtained in other transition
economies. As expected, the estimated profit rate in the sector is highly
correlated with the firms’ own profits (p =0.51) and is a useful control for
industry-level determinants of profit.
3
We controlled for the estimated level of
taxes in order to capture the effect to which taxes can be a disincentive to
investment in general.
4
Figures for other transition economies obtained by JMW
suggest that in terms of tax levels, Serbia and Montenegro are closer to the less
advanced transition performers than the more advanced countries of Central
Eastern Europe. Unfortunately, without information about sample distributions,
it is difficult to say, in any meaningful terms, how the estimated profit rates
compare across transition economies captured in the JMW sample.
3
See Section 7.3 for the results of the models that make use of this control variable.
4
This decision follows JMW.
128
Table 7.2: Perceived security of property rights
Indicator Percent
Firms make payments for protection 21.53
Belief in court effectiveness
1 (min) 18.45
2 17.34
3 29.15
4 25.09
5 (max) 9.96
7.2.2 Property Rights
To get information about respondents’ perceptions of the security of property
rights, several questions were used. First, entrepreneurs were asked the following:
“It is thought that some firms may need to make payments for “protection” of their
activities. Do you think this is true for firms in your sector?”
5
Previous research
has shown that sensitive questions about unofficial payments, bribery, corruption
and similar receive better response rates when asked in this indirect format, rather
than with reference to the respondent’s firm.
6
Table 7.2 shows that 21.53% of the
firms in the sample responded affirmatively to this question. The second indicator
for the security of property rights is based on the respondents’ perception of the
effectiveness of courts in enforcing contracts. The question is based on a 5-point
Likert scale and is stated as follows: “How confident are you on a scale 1 to 5 (1
being ‘completely lacking confidence’ and 5 being ‘completely confident’) that a
decision of the court in your favor will be enforced if it goes against the other
party in the dispute?” Table 7.2 reports the percentage of respondents who
responded in each category reflecting confidence in the effectiveness of courts.
5
This question appears in JMW original questionnaire.
6
JMW follow this practice developed in Daniel Kaufmann’s empirical work. See for example
Hellman et al. (2000).
129
Table 7.3: Perceived security of property rights, with JMW indicators
Serbia and
Montenegro
Survey result countries Poland Slovakia Romania Russia Ukraine
Percentage of respondents who
say firms make payments for
protection
21.53 7.9 14.9 0.6 92.9 88.8
(272) (302) (308) (320) (126) (107)
Percentage of respondents who
say firms make extralegal
payments for government services
51.46 20.1 38.2 20.0 91.2 86.9
(271) (298) (306) (315) (114) (84)
Percentage of respondents who
say firms make extralegal
payments for licenses
47.45 19.3 42.2 17.0 91.7 87.5
(272) (300) (303) (317) (120) (88)
Percentage of respondents who
say courts cannot be used to
enforce contracts
44.16 27.1 32.1 13.1 44.2 45.4
(274) (303) (308) (321) (269) (269)
Number of entrepreneurs
surveyed
274 303 308 321 269 270
Notes: The number of observations is given in parentheses below each response level. The results for Poland,
Slovakia, Romania, Russia, and Ukraine are reported in Johnson, McMillan, and Woodruff (2002b).
Perceived court effectiveness may vary within the sample for several reasons.
First, individuals form their opinions based on personal experience and the
information they receive about others’ experiences with the courts. The first of
these can vary due to personal characteristics of the manager, such as education,
personal trustfulness, and the experience with varying quality of courts in
different locations. The changing circumstances and institutions in transition can
cause the quality of institutions as well as the incidence of bribes and corruption
to vary across jurisdictions, all of which can affect individual experiences. Others’
experiences, affected by the same factors, will shape the information individuals
use in forming their evaluations. Moreover, an individual will process information
about others’ experiences to varying degrees, also contributing to within-country
variation in the perceptions of court effectiveness
Generating a new survey for the purpose of data collection offers unique
freedoms in deciding which questions to include in the survey. Capitalizing on this
opportunity, the survey conducted for this study replicated the JMW questions
130
necessary to construct the measures of the security of property rights. In addition
to asking entrepreneurs whether they make payments for protection, the survey
also included the JMW questions on whether businesses make “extra legal”
payments for government services and for obtaining licenses. Using already
existing survey questions, we were also able to construct a measure equivalent to
the JMW indicator for whether firms believe courts can be used to enforce
agreements with trading partners. The responses to these questions are included
in Table 7.3 along with JMW results for the other transition economies from their
study. Although the table illustrates great variation between countries, JMW note
a distinction between two groups: Central European Countries (Poland, Slovakia
and Romania) and the two former members of the Soviet Union, Russia and
Ukraine. In all cases except the last row in Table 7.3, Serbia and Montenegro fall
between the two groups identified by JMW. With respect to confidence in courts,
responses from Serbia and Montenegro are much more in line with those obtained
from Russia and Ukraine than those from the more advanced Central European
economies. The goal of the comparison with other transition economies is not to
quantify any differences or similarities but to place Serbia and Montenegro in a
broader context. A cursory examination of survey results from Serbia and
Montenegro against other representative transition cases allows us to check
whether they include any significant departures from the results found in the rest
of the region. Given that Serbia and Montenegro do not show extremely large or
small results compared to the other cases, the table demonstrates that with
respect to property rights components, the two cases under observation are not
outliers. Thus, any analysis of the role of property rights in profit reinvestment
will generate results that are not specific to some outliers and the unique
131
Table 7.4: Extent to which networks used in dispute resolution
Networks used in
dispute resolution Freq. Percent
1 37 14.68
2 35 13.89
3 47 18.65
4 30 11.90
5 39 15.48
6 28 11.11
7 36 14.29
Total 252 100.00
combination of conditions specific to them, but can translate to other cases in the
region.
7.2.3 Use of Networks
To examine the role of networks as a form of informal contract enforcement,
entrepreneurs were asked the following question (based on a 7-point Likert scale
with one 1 for very little and 7 for very extensive): “Please circle the number
which best describes the extent to which managers in your business use personal
relations and connections to resolve disputes with other businesses.” The
responses for each level are shown in Table 7.4.
An alternative measure of networking activity was also employed in an
attempt to more accurately examine the use of networks and their impact on
business performance through reinvestment. This measure was based on the
following survey question: “In an average week, approximately how much time do
you devote to maintaining contacts with the people you know and with whom you
discuss your business activities?”
7
The responses ranged from 0 to 40 hours per
week with a mean of 14.69 hours and a standard deviation of 10.16 hours.
7
This question has been adapted from a similar question used by Aldrich, Rosen, and
Woodward (1987) and borrowed by Ostgaard and Birley (1994).
132
Surprisingly, almost 70% of the sample reported spending 10 hours or more per
week maintaining contacts. Table 7.5 shows all of the responses given and the
frequency of respondents for each specific response. The question was designed to
capture the use of both social and economic ties since the conceptualization of
social networks used in this study does not wish to discriminate between the two.
This measure is not restricted to the use of networks in dispute resolution alone
but captures the broader benefits of utilizing network ties and their embedded
resources. More importantly, this question captures both the positive and negative
aspects of the use of networks. Those individuals indicating that they spend 30 or
40 hours per week maintaining contacts are not only deriving benefits from such
ties but are also incurring costs in terms of time and resources (including the
opportunity cost of using this time for other productive activities as opposed to
contact maintenance). Studies of networks in transition economies and the
qualitative evidence from Serbia and Montenegro have both shown that networks
provide benefits and generate costs. Determining whether networks positively or
negatively impact reinvestment will be made possible with a measure of network
use that does capture more than just the positive influence of social networks.
None of the network measures considered are significantly correlated with the
measures of court effectiveness and the risk of expropriation. This means that
network variables, contract enforcement measures, and expropriation measures are
capturing different sources of variation and can be safely included together in
regressions of profit reinvestment.
133
Table 7.5: Time devoted to maintaining contacts
Hours per
week Freq. Percent
0 2 0.81
1 3 1.22
2 17 6.91
3 2 0.81
4 13 5.28
5 15 6.10
7 3 1.22
8 13 5.28
10 64 26.02
11 7 2.85
12 9 3.66
15 5 2.03
16 3 1.22
17 3 1.22
20 36 14.63
24 3 1.22
25 3 1.22
30 30 12.20
35 8 3.25
36 2 0.81
40 5 2.03
Total 246 100.00
134
7.3 Results
The goal of the regression analysis is to test whether the use of social networks
impacts businesses’ reinvestment decisions. The hypothesis is explored that
businesses that utilize networks to a greater extent, have greater confidence in
court effectiveness, and perceive a lesser threat of expropriation will reinvest more
of their profits. Moreover, if the confidence in courts and the use of networks for
informal contract enforcement are complements, then including network measures
into the model should make no impact on reinvestment.
In the regression models tested, the dependent variable is the percentage of
profits firms reinvested during the previous calendar year. As described already,
this variable includes categorical data ranging from 1 (lowest reinvestment) to 6
(highest reinvestment category), which dictates that we use ordered probits to
investigate the determinants of profit reinvestment. The independent variables
considered in the model include the measures of the extent to which businesses
trust contract enforcement, the extent to which networks and relations are used in
contract enforcement, a variable measuring the time managers spend maintaining
contacts per week, and a dummy variable indicating whether businesses feel that
others in the industry must pay for protection. Table 7.6 presents the regression
results. Only those businesses with positive profits are included in the analysis.
Since the goal is to examine the factors impacting reinvestment of profits,
businesses with no internal funds available for reinvestment are not interesting for
the analysis as their reinvestment rate cannot be measured. The numbers
reported in the table are coefficients of the ordered probit models: a positive
coefficient indicates that an increase in the particular independent variable
increases the probability that a business falls in the higher reinvestment category.
135
In parentheses just below the coefficients is the z value calculated using robust
standard errors. The size of the dataset did not permit the use of standard errors
clustered per industry and region.
Consider first the columns between 1, 3, and 5. Column 1 reports the results
of the model using only the variables that represent the threat of expropriation
and the confidence in contract enforcement. The dummy variable indicating
whether businesses make payments for protection has the hypothesized sign but is
insignificant. The extent to which businesses believe in the contract enforcement
ability of the courts has the hypothesized sign, meaning that an increase in the
degree to which businesses feel confident in the courts is associated with a higher
level of reinvestment and this relationship is significant at the 5% level. Column 3
reports the results of the base model reported in column 1 with two variables
representing the use of networks added. They are the degree to which networks
are used in contract dispute resolution and the time spent maintaining conacts.
Interestingly, both of these variables are positively associated with reinvestment
and highly significant. The results show that using networks to a greater extent
for dispute resolution and spending more time during the week maintaining
contacts increases the probability of being in a higher reinvestment category.
Next, the estimated industry profit rate reported by firms for their sectors and
tax payments as a percent of sales are included as proxies for the expected
investment opportunities and disincentives for reinvestment, respectively.
Following JMW, the firms’ own profit rates are not used in the models estimating
reinvestment rates in order to avoid the potentially damaging effects of reverse
causation. This practice is less than perfect because there are still some risks of
endogeneity resulting from the fact that expected profit rates for the sector are
136
Table 7.6: Determinants of profit reinvestment
Variable (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)
Firms make payments for
protection (d)
-0.10 -0.28 -0.66*** -0.67*** -0.66*** -0.79*** -0.86*** -0.82***
(-0.63) (-1.69) (-3.73) (-3.82) (-3.79) (-4.39) (-4.21) (-4.50)
Belief in court effectiveness 0.12* 0.15** 0.14* 0.13* 0.13* 0.15* 0.15* 0.15*
(2.27) (2.69) (2.13) (2.05) (2.07) (2.37) (2.19) (2.25)
JMW 4 point index -0.00 -0.10 -0.26*** -0.26***
(-0.00) (-1.73) (-3.43) (-3.58)
Networks used in dispute
resolution
0.10** 0.10** 0.12** 0.12** 0.13** 0.12** 0.13** 0.12* 0.12** 0.13**
(2.72) (2.93) (2.63) (2.71) (2.83) (2.75) (2.90) (2.57) (2.69) (2.94)
Time maintaining contacts 0.03*** 0.02*** 0.04*** 0.03*** 0.03*** 0.05*** 0.04*** 0.05*** 0.04*** 0.04***
(3.78) (3.41) (4.24) (4.22) (4.14) (5.18) (5.01) (5.29) (5.00) (5.10)
Estimated industry profit
rate
-0.02 -0.02 -0.02 -0.01 -0.01 -0.01 -0.02* -0.01
(-1.92) (-1.95) (-1.91) (-1.74) (-1.60) (-1.57) (-2.13) (-1.53)
Tax payments (percent of
sales)
-0.00 -0.00 -0.00 0.00 0.00 -0.00 0.00
(-0.51) (-0.56) (-0.06) (0.08) (0.03) (-0.17) (0.12)
Log age of firm -0.26* -0.37** -0.31* -0.32* -0.34* -0.32*
(-1.99) (-2.90) (-2.45) (-2.45) (-2.46) (-2.37)
Loan before last year (d) -0.21 0.07
(-0.95) (0.35)
Collateral to offer bank (d) 0.10 -0.07
(-0.95) (-0.59)
Industry controls yes yes yes yes yes yes yes yes
Manager characteristics yes yes yes yes yes
Log of number of employees yes yes yes yes yes
Region controls yes
N 271 274 243 246 213 213 213 211 214 211 211 214
Log likelihood -455.29 -462.44 -396.34 -406.06 -334.44 -334.28 -332.65 -321.78 -332.37 -320.33 -321.34 -332.32
chi_square 5.60 0.00 31.90 20.40 73.35 76.90 80.82 129.22 75.39 130.82 136.52 83.95
p 0.06 1.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Notes: All regressions are ordered probits using the last year’s rate of reinvestment as a dependent variable. (d) in the variable name indicates a dummy
variable. Numbers in the rows are coefficients. z values based on robust standard errors are included in the parentheses. Stars indicate variable significance:
* for p<.05, ** for p<.01, and *** for p<.001. When region and industry controls are both included in the model they are interacted.
137
largely based on a firm’s own rate of profit. However, to capture the impact of
expected profits on future reinvestment, an indicator such as the estimated profit
rate for the sector as a whole is less subject to endogeneity problems than the use
of the firm’s own profit rate.
8
Both the estimated profits and taxes in the firm’s
sector are used in the models reported in columns 5 and 6. The models also
include controls for the industry in which the business operates using dummy
variables for all of the industries in the sample. In models 5 and 6 the dummy
variable indicating that firms make extra legal payments for protection is now
negative and highly significant, showing that the perception of the threat of
expropriation is strongly associated with a lower reinvestment category. Variables
showing belief in court effectiveness, use of networks in dispute resolution, and
time spent maintaining contacts continue to show positive and highly significant
coefficients. Although estimated profit rates and tax payments have negative
coefficients, neither value is significant.
9
None of the industry coefficients show
significant results but the use of industry dummies along with the estimated
profits and taxes is an important form of control for industry-level determinants
of profit reinvestment and it increases the overall model validity.
Models reported in columns 7 and 8 include variables measuring the age of the
firms in order to control for the different reinvestment needs of younger and older
businesses. The expected negative relationship between firm age and reinvestment
8
Additional tests performed by JMW have indicated that their model is impervious to
endogeneity problems, suggesting that it is safe to use the estimated profit rate for the sector in
models of profit reinvestment.
9
The insignificant and small coefficients for tax payments (coefficients consistently come out
to +/- 0.00) are similar to results obtained by JMW. They explain this outcome by suggesting
that perhaps tax payments vary little across industries and remain constant within a specific
region/country. Later, in model 10, we will interact industry controls with the dummy variable
for the region in which the business operates to try to test for these regional effects, including
differences in tax codes.
138
is significant, showing that younger firms tend to reinvest more than older firms.
Model 8 also includes additional controls for manager characteristics. This
includes a dummy variable indicating whether the manager had any experience in
the management position in the private sector, dummy indicators for the highest
level of education achieved, and the log of the managers’ age. In addition, model
8 includes the variable measuring the firm’s numbers of employees. Including
these additional controls has improved the overall model significance (chi_square
for the models is higher). In model 8, the coefficient for the log of the age of the
entrepreneur is negative and significant at the 1% level of significance (coefficient
-0.72 and z value of -3.07 are not reported in the table). This indicates that the
reinvestment rates are higher for younger entrepreneurs.
10
Private sector
experience in management did not have a significant effect. However, education
controls indicating the manager’s highest level of education completed were
significant for all categories beyond high school (the omitted category) and 2-year
technical school.
11
This suggests that better educated entrepreneurs reinvest more
regardless of their previous experience (results omitted from the table).
The model reported in column 10 illustrates similar results to model 8 with
only a slight modification in the control variables. In this model, we tested
whether the region in which the business operates (Serbia or Montenegro) has any
impact on reinvestment when combined with other industry controls.
12
This
allows us not only to control for industry determinants of profit rates and
reinvestment, but also for different conditions in the same industry across regions.
10
JMW obtained the same result for their selection of transition economies.
11
These categories are 2-year college completed, education beyond 2 years but not completed,
and a masters degree.
12
Chapter 5 explained that Serbia and Montenegro are treated as a single economy in this
study. In June 2006, these two regions became independent states.
139
The industry and region controls are interacted and the results reported in
column 10, which show that neither of the interacting terms is significant and the
region controls do not contribute significantly to the model.
The results demonstrate that, as hypothesized, a higher threat of
expropriation perceived by businesses lowers the rate of reinvestment. Also, the
more confident managers are about the effectiveness of courts in enforcing
contracts, the more they reinvest. Further, the models illustrate the consistently
significant and positive impact of network measures on reinvestment. As
hypothesized, the more networks are used to enforce agreements without relying
on formal contracting, the more confident managers and owners are to reinvest
their profits back into the business. Coefficients reported by the ordered probit
regression models are not easily converted into measures of increases in
probability and even so, it is difficult to interpret increases in probability in terms
of their implications for the dependent variable. However, we can use the results
of model 8 to calculate predicted probabilities that a manager in the sample will
report a specific level of profit reinvestment, given one of the network measures,
while holding all other independent variables at their mean. Table 7.7 illustrates
the results for predicted profit reinvestment given the degree to which networks
are used in dispute resolution. The last row shows the average level of
reinvestment for the specific level of the network measure. The weighted average
is calculated using the midpoint for each investment category and weighing it by
the percentage of businesses predicted to report that particular level of
reinvestment. In Table 7.7 we can see that as the extent to which networks are
used for dispute resolution increases, so does predicted reinvestment.
140
Table 7.7: Predicted profit reinvestment rates 1
Percentage of firms in reinvestment category
Extent to which networks are used in dispute resolution
1 2 3 4 5 6 7
Profit reinvestment rate (%) (very
little)
(very
much)
0 8.05 6.37 4.98 3.84 2.92 2.20 1.63
1–10 29.45 26.58 23.66 20.78 17.99 15.37 12.95
11–25 29.81 29.82 29.41 28.59 27.40 25.88 24.10
26–49 13.58 14.60 15.46 16.14 16.60 16.82 16.79
50–75 9.20 10.41 11.60 12.74 13.78 14.69 15.43
76–100 9.91 12.22 14.89 17.91 21.30 25.04 29.10
Weighted reinvestment rate 26.40 29.43 32.63 35.96 39.41 42.95 46.53
Note: The extent to which networks are used in dispute resolution ranges from 1 (very little) to 7 (very
much). The percentages reflect the probability of being in each reinvestment category given the extent to
which networks are used in contract enforcement. The rest of the independent variables are held at their
mean. The results of regression model 8 are used to calculate the probabilities. The weighted reinvestment
rate is calculated using the midpoint for each category of profit reinvestment.
The surprising result, however, is the robust positive impact of the variable
measuring the time managers spend during the week maintaining contacts. The
measure is designed to capture the costs that occur to individual businessmen
when they devote time and resources to maintaining network relationships instead
of concentrating their efforts on other productive activities. Earlier we noted that
the reported levels of the variable measuring the extent of networking activity
through the time spent maintaining network contacts ranged from 0 to 40 hours
per week. Among the respondents, almost 70% reported spending 10 hours or
more per week on activities helping to maintain network relationships. These
numbers are surprisingly high when one considers the absolute size of these time
commitments. However, without a direct comparison to similar measures from
other surveys, little more can be said about the absolute levels. A panel study of
SMEs in the Research Triangle Area of North Carolina can serve as a possible
benchmark (Reese and Aldrich 1995). The study found that 29.68% of businesses
belonged in the low category, 36.04% in the medium, and 34.27% in the high
category, where categories are defined by the number of hours per week spent
141
maintaining contacts. Interestingly, the cutoffs for the categories are 2.9 hours for
the low, 7.9 hours for the medium and the highest category includes any business
spending 8 hours or above per week. In the case of the respondents from Serbia
and Montenegro, the last category in terms of the above definitions would include
approximately 80% of the sample. This means that in absolute levels, the costs of
networking, when measured by the time spent maintaining network contracts, are
very high.
The study of SMEs in North Carolina did not find any relationship between
the time spent maintaining networks and the indicators for business survival and
performance. Here, however, we have shown that despite the costs associated with
network use, networking activity positively impacts the level of profit
reinvestment. Table 7.8 shows the predicted probabilities of a manager in the
sample reporting a specific level of profit reinvestment given the degree of his or
her networking activity (measured by the time spent maintaining contacts). The
predicted probabilities are calculated using the results from model 8, and the last
row in the table reports the weighted average level of reinvestment for select
values of the time during the week that managers devote to maintaining contacts.
Table 7.8 shows that managers who devote more time during the week to
activities that help them maintain their contacts are also predicted to reinvest
more. Managers devoting no time to maintaining contacts are expected to
reinvest 19.08% of their profits while those investing 40 hours per week to
maintaining contacts (the sample maximum) will on average reinvest 67.50% of
their profits. In effect, the cost of networking has a positive and robust effect on
reinvestment of profits suggesting that for businesses that wish to grow and
expand their activities, networking activities are a necessary investment. Further
142
Table 7.8: Predicted profit reinvestment rates 2
Percentage of firms in reinvestment category
Hours per week devoted to maintaining contacts
Profit reinvestment rate (%) 0 10 20 30 40
0 14.36 6.37 2.36 0.01 0.00
1–10 36.83 26.58 16.02 7.95 3.25
11–25 27.66 29.82 26.29 18.96 11.17
26–49 10.31 14.60 16.79 15.68 11.89
50–75 6.08 10.41 14.48 16.35 14.99
76–100 5.22 12.22 24.06 40.34 58.52
Weighted reinvestment rate 19.08 29.43 42.05 55.41 67.50
Note: The percentages in the columns reflect the probability of being in each reinvestment category given the
hours per week devoted to maintaining contacts. While the hours per week spent maintaining contacts span a
continuous range from 0–40, five discrete points were chosen for illustration. The weighted reinvestment rate is
calculated using the midpoint for each category of profit reinvestment.
research is needed to confirm this relationship in other samples and more
importantly, to investigate whether the relationship, for a bigger sample, might in
fact illustrate the presence of a threshold level beyond which too much networking
activity crowds out other productive activities.
Network measures, such as the degree to which networks are used for dispute
resolution and the time spent maintaining contacts in any given week, show
significant and positive results for reinvestment beyond the measures of the
effectiveness of courts. At first glance, this outcome suggests that informal and
formal means for contract enforcement are substitutes. However, a model
including an interaction term between the use of networks in dispute resolution
and the effectiveness of courts produces a positive but insignificant coefficient for
the interaction (model not included in the table). At this point we can conclude
that formal and informal contracting produce independent positive effects on the
level of profit reinvestment for SMEs.
143
7.4 Robustness Checks
The regression results in Table 7.6 include additional models used to examine
whether alternative specifications of the property rights indicators would produce
similar results in terms of the effect of network measures. For these alternative
specifications, we use the JMW 4-point additive index of the insecurity of
property rights. Affirmative answers to the questions of whether businesses make
extra legal payments for services, licenses, and protection, as well as a negative
response to the question whether courts can be used to enforce contracts, are
combined to produce the index. The higher the value of the index, the less secure
the property rights, as perceived by the respondent. The index is used in models
2, 4, and 9 instead of the expropriation and court effectiveness measures used in
the analysis up to this point. The results show that both network variables—the
degree to which networks are used in dispute resolution and the time invested in
maintaining contacts—are positive and significant even when the alternative
specification for property rights is used.
Including the JMW index alone (as in model 2) does not produce any results
in terms of the ability of the model to predict profit reinvestment. The index is
also not a significant predictor of reinvestment when included with the measure of
network use (model 4). However, when all of the other control variables are added
into the model (model 9), the property rights index shows the expected negative
and significant impact on reinvestment. Thus, the alternative specification of
measures for the security of property rights does not improve the models overall.
Comparing model 4 with model 3 and model 9 with model 8 shows that our
original measures of property rights produced better specified models (higher
chi_squared).
144
A more important check for the robustness of the models presented and the
results obtained is to test whether the original assumption used to create the
model of reinvestment is confirmed in the empirical results. The JMW model of
reinvestment is based on the assumption that the decision to reinvest internal
funds (which are proxied by reported profits) and the decision to borrow
externally are independent. This assumption is important for allowing us to
consider determinants of property rights independent of the firms’ access to
external finance. However, JMW note that it is possible for the two decisions to
be complementary to one another. When investments are “lumpy,” businesses may
make their decision to reinvest profits and borrow externally simultaneously
(Johnson, McMillan, and Woodruff 2002b, 1352). In other words, if they can
borrow externally, they will also reinvest internally. Or the corollary, if they
cannot borrow externally to fund their planned investments, they will not want to
reinvest internally either. To test whether this critical model assumption holds,
we follow JMW and include measures of access to external finance. The latent
variable model for investment demand determines both the demand for
reinvestment (the model examined in this study) and the demand for external
funds. Thus, including the direct measure of access to finance, such as dummy
indicator for “whether a firm had a loan in the previous year,” is not appropriate
because it can impact the latent variable (demand for investment funds).
However, we can include measures that are correlated with access to finance but
only weakly correlated with investment demand, namely the indicator for whether
a firm has had a loan prior to 2005 and the indicator for whether a firm has assets
to offer as collateral for a loan. In fact, these two measures are significantly and
positively correlated with the dummy for whether the firm has had a loan in the
145
previous year. The correlation coefficients are 0.67 and 0.81 and both are
significant at the 0.01 level. Of firms who had received a loan in 2005, 78.29% had
offered collateral. Similarly, of those who had received a loan in 2005, 87.60% had
also received a loan prior to 2005.
Model 11 tests the impact of two measures of firms’ creditworthiness—loan
before 2005 and an indicator for whether the firm has assets to offer as collateral.
Neither measure has a significant impact on reinvestment, confirming that access
to finance and reinvestment decisions are in fact independent from each other. As
an additional check, the model was rerun with the JMW 4 point index for the
insecurity of property rights (model 12). Again, dummy variables for a loan
accepted before 2005 and the availability of collateral do not have any impact on
reinvestment, offering evidence that investments are not lumpy and that the
model’s specification is valid. Correlation coefficients between the indicator for
access to a loan in 2005 (or the availability of collateral) and the measures for
expropriation, credibility of courts, and use of networks are insignificant. This
means that our measures are not proxying for access to finance, and including
them together with proxies for access to finance in models 11 and 12 is
appropriate. Our earlier conclusions in terms of the threat to expropriation,
credibility of courts, and the use of networks remain robust. The inclusion of the
two finance variables has shown that proxies for access to finance do not impact
the decision to reinvest profits. The critical model assumption—treating demand
for external funds and demand for reinvestment as independent from each
other—has been confirmed.
146
7.5 Conclusion
There is a growing body of literature suggesting that personal relationships are
important factors in market exchange, but little has been done to empirically
examine the roles that networks play and their impact on individual businesses.
This chapter investigated the possible impact of social networks’ roles in contract
enforcement. Using an original dataset from SMEs in Serbia and Montenegro, the
chapter has specifically explored the impact of networking activity on the
individual business decision to reinvest profits.
The use of networks in dispute resolution is strongly and consistently
associated with greater profit reinvestment. As shown in Table 7.7, businesses
that rely minimally on networks when it comes to resolving disputes with other
businesses reinvest 26.40% of their profits while those who rely the most on
networks reinvest 46.53%. Using networks for dispute resolution can produce
reinvestment rates up to 100% higher. This effect was strong even after we
controlled for the formal elements of property rights recognized as important
determinants of profit reinvestment. As expected, greater confidence in courts was
associated with higher reinvestment, as was a lesser risk of expropriation by the
government or third parties. However, the use of networks for contract dispute
resolution impacted profit reinvestment even after we took into account the
business owner or manager’s confidence in the courts and his or her perceived risk
of expropriation.
The results presented in this chapter also document a strong relationship
between the cost of networking activity and profit reinvestment. Devoting time to
maintaining contacts that are related to one’s business activity was associated
with higher reinvestment. Individuals who spent 40 hours per week maintaining
147
contacts reinvested approximately 300% more than those businesses that devoted
zero hours to maintaining contacts (see Table 7.8). The evidence suggests that
although a costly activity, maintaining network relationships is an efficient
strategy for individual businesses. Maintaining business contacts or furthering
mutual relationships was shown to be an important component of a
growth-oriented business strategy. The effects of networking activity appeared to
be independent of the impact of contract enforcement and the risk of
expropriation, indicating that networks play a role that goes beyond the informal
mechanism for contract enforcement. Despite the costs involved in maintaining
ties, networks had a positive impact on reinvestment. Determining the effect of
networks beyond their role in informal contract enforcement, and examining
whether they impact business performance directly, is a task undertaken in the
next chapter.
The sample used to conduct the above analysis is based on the existing and
operating SMEs in Serbia and Montenegro. Businesses who are no longer
operating, or have never been established, are excluded and thus we lack evidence
from this group of entrepreneurs. The inclusion of only operating businesses may
understate the impact of the role of networks. However, the analysis
demonstrated that in encouraging reinvestment networks play an important role
that should translate into the growth of individual businesses—a growth that
contributes to the performance of the SME sector as whole and potentially the
success of the transition overall.
148
Chapter 8
Network Roles and Business
Performance
The interviews with SME managers and owners in Serbia and Montenegro,
introduced in Chapter 5, have demonstrated that the impact of networks extends
beyond their role in contract enforcement when network relationships are used to
secure access to invaluable sources of capital, information, or opportunities to
secure advantages for their businesses. This has raised the question of whether
these other network roles are also positively correlated with factors that encourage
SME growth. Thus, we must investigate further whether the prevalence of
networks in business activities is associated directly with other measures of
business performance. In doing so, we must pay special attention to the role that
social networks play in economic activities. One way to distinguish distinct
network roles is to pay closer attention to the content of exchange facilitated
through networks. Understanding the content of network exchange should provide
greater insight into the roles that networks play and how the roles translate into
improved business performance.
149
8.1 Networks as Determinants of Economic
Outcomes
We saw in Chapter 4 that transition scholarship, while arguing successfully that
social networks play an important role in transition societies, has not offered
concise lessons in terms of the impact of social networks on business performance.
Scholars have noted the usefulness of social networks in exploiting weaknesses of
institutional infrastructures or for openly violating formal rules for individual
advantage. Others have emphasized the positive aspects of social networks and
have noted their utility for compensating for institutional failures. In particular,
networks compensate for the inefficiencies in the third-party contract enforcement
usually provided by the courts. Through the reputation mechanism, networks
assist individuals in enforcing contracts by ensuring that cheaters are identified
and expelled from the network. By the same token, transition scholars have
recognized that networks can result in "lock-in" effects, limiting individuals’
ability to conduct business activities beyond their networks (Ledeneva 2004;
Radaev 2004; Mihaylova and Harriss 2003; Božović 2006b; Johnson, McMillan,
and Woodruff 2002a). What they have not illuminated are the implications of
these aspects of social networks for private sector growth or individual firms’
business performance.
Social network studies developed almost exclusively outside the context of
Eastern Europe or post-communist societies suggest a positive role of social
networks in mobilizing resources for economic activity. A subset of social network
studies has been advanced by management and entrepreneurship scholars who
have been interested in examining whether social networks encourage
150
entrepreneurship and the success of newly founded firms. Borrowing central ideas
from economic sociology, they recognize that actors’ opportunities and resources
are dependent upon the nature of their social environment and relationships with
individuals around them. They have emphasized the role of personal networks in
facilitating exchange and improving opportunities for business success by
providing members with emotional support, information, knowledge, capital and
other resources. Their primary objective has been to determine empirically
whether venture start-up and business survival can be traced to any or all of these
positive aspects.
Empirical studies of networks in entrepreneurship differ in terms of what
aspects of networks they see as critical in delivering key benefits. Some have
suggested that the types of ties connecting individuals in the network are the
potential sources of advantage (Lechner and Dowling 2003; Dubini and Aldrich
1991). Others have argued that the specific characteristics of network structure
are the sources of advantages provided by the networks (Renzulli, Aldrich, and
Oody 2000; Aldrich, Rosen, and Woodward 1987). Resources embedded in social
networks have also been implicitly tied to performance. However, even though
resources provide the best potential for applying the concept of social capital to
the analysis of business performance, studies of the relationship between network
resources and business success are absent.
1
Whether we focus on ties, resources,
or the overall structure, social network studies make the case that networks can
provide members with specific advantages that can results in economic payoffs.
However, the perspective of the network roles adopted in management and
entrepreneurship studies is much more limited than that of transition scholars
1
Batjargal (2003) is a notable exception.
151
who are concerned with social networks as a broader strategy for conducting
business activities. As a result, social networks studies have not considered
multiple network roles, or multiple types of network content, and their impact on
performance. Unfortunately, there has been a significant lack of dialogue between
these two areas of research.
Without notable overlap between social networks and transition studies, it
remains unclear to what extent social networks constrain or improve business
performance in transition. Social network studies give us a method with which to
bridge the gap with transition scholarship. Transition scholars, however, bring to
the table a wider and more eclectic perspective of networks that has developed as
a response to the need to understand the role of networks in numerous aspects of
post-communist transition (Ledeneva 2004). This chapter adopts the approach of
social network studies that have been applied by entrepreneurship and
management scholars examining business founding and growth. This approach is
applied to an investigation of the overall impact of networks on the success of
individual businesses in a transition setting, while taking into account multiple
network roles. This is the first study to consider specifically how distinct network
roles, determined by the content of exchange facilitated through the network,
impact business performance in transition. In doing so, it aims to uncover
whether the impact of network structure, relationship types, and network
resources is dependent on, or mediated by, the content of exchange facilitated
through the networks. In other words, it explores whether the impact of structure,
relationships and resources varies depending on the network role in question.
152
8.2 Hypotheses
8.2.1 Overall Impact
Previous research on networks and their impact on business performance has
produced inconsistent results. Based on insights from social capital and social
network studies, many have implicitly suggested a positive relationship between
the use of networks and business success that has not been confirmed by empirical
evidence (Johannisson 1995). Aldrich, Rosen, and Woodward (1987) concluded in
a longitudinal study that networking activity is important for business founding
but that it does not appear to improve business profitability. However, Ostgaard
and Birley (1996) found that networks are sources of advantages even beyond the
business startup stage. Birley (1985) and Reese and Aldrich (1995) found
contradictory results where the the first work argued that networks are a crucial
source of help in assembling resources for business startup, and the second study
found no evidence in support of the claim that networks impact survival and
performance.
To understand the net effect of networks, it is important to look at business
performance and not just business survival following startup. The full
implications of the use of networks may not be observable at the time of a
business startup since the relationships used to assist someone in setting up a
business can become limiting over the course of a business lifetime. Also,
businesses have different needs beyond the startup and so different network roles
may become relevant later. The interviews with managers and owners in Serbia
and Montenegro have revealed that networks play two distinct yet important
roles. First, they are used as a mechanism for contract enforcement and second,
153
they can secure access to capital and information. By avoiding legal action and
the use of expensive and ineffective courts, social networks can lower the cost of
transacting and improve business competitiveness. Also, to remain competitive
and ensure long-term growth, business can benefit from maintaining ties that
grant them exclusive access to capital and information as these translate into
direct economic payoffs. Thus, we hypothesize:
Hypothesis 1: Use of network relationships for contract enforcement
and securing access to resources should positively impact SME
performance.
8.2.2 Network Structure
Several aspects of network structure have been found to have an impact on a
business’ founding or the likelihood of its success. The most elemental indicator of
a network’s impact on business outcomes is its size. The larger the network that
the individual accesses, the greater the likelihood that the network will provide
the individual business with advantages that can result in economic payoffs.
2
Beyond network size, characteristics of network structure include density and
diversity. Though related, lower density and higher diversity are often considered
the optimal atributes of one’s network. In one of the elemental studies of
economic sociology and network scholarship, Burt (1992, 2000) argues that
networks provide advantages when they are less dense or full of structural holes
such that individuals bridge the diverse pools of information and resources to
their advantage. The greater the diversity among network members and the more
structural holes within a network, the greater the advantages the network can
2
The empirical evidence demonstrating the positive impact of network size is mixed
(Ostgaard and Birley 1996; Podolny and Baron 1997).
154
generate. Similarly, the literature addressing networks’ impact on job market
outcomes has argued that diverse network ties generate the least redundant and
most valuable sets of information about potential opportunities for job market
candidates. Empirical evidence supporting the above claim is inconsistent. With
respect to individual firm-level outcomes, Aldrich, Rosen, and Woodward (1987)
and Jansen and Weber (2004) found no evidence that the diversity or size of a
network have a positive impact on increases in turnover or profitability. Renzulli,
Aldrich, and Oody (2000), however, found that network diversity, or
heterogeneity, increases the odds for business startups.
One way to understand the inconsistencies in empirical evidence is to consider
whether lower density and greater diversity are always the preferred
characteristics of networks regardless of the role that networks play. It is possible
that for some types of networks, greater density and lower diversity are more
likely to produce a positive outcome. Consider, for example, the role of networks
in enabling businesses to enforce their contracts. Since networks lower the costs of
arbitration through informal means, bigger networks should provide greater
opportunities for conducting business without the risk of cheating or problems
with production and delivery. The reputation mechanism at work in the trust
network is rooted in the transfer of information about actors’ reputation in past
dealings. As a result, the more trusted and more overlapping the information
channels among the alters, the better the ability of the trust network to transfer
the necessary information and discourage cheating among network members. The
extent to which alters are connected in the ego’s network is an indicator of the
degree to which there already exists an infrastructure necessary for the transfer of
trusted information. Trust networks are most valuable for business performance
155
when they can provide the ego with the cheapest and most efficient mechanism
for contract enforcement. The number of indirect ties in the network—ties among
the alters—will be greater if these individuals come from a “similar walk” of the
ego’s life. In other words, diversity in terms of the relationship that the ego has
with the alters in the network will negatively impact the potential for the alters to
know one another and to have relationships among themselves. If, however, the
ties between the ego and the alters are similar in nature, then the likelihood that
the alters know one another will be greater. Overall, greater density and lower
diversity should maximize the potential that the trust network provides the ego
with a cheap and efficient mechanism for contract enforcement.
The preferred density and diversity of a network whose members provide
valuable access to resources are entirely different when it comes to the network’s
impact on business performance. Following Burt’s (2000) structural holes
argument, to provide the ego with the largest benefit that can translate into
business success, individuals should maintain ties with a diverse group of
individuals thereby spanning structural holes and accessing diverse source of
capital and information. For the same number of ties, managers with overlapping
ties are at a disadvantage when competing for access against someone with a set
of diverse ties (Dubini and Aldrich 1991; Aldrich, Rosen, and Woodward 1987).
The less network members know each other, and the more diverse the
relationships they have with the ego, the greater the potential for the ego to
access valuable and diverse resources with direct benefits for their business.
Diversity and density, two critical components of network structure, can have
a direct positive impact on business performance. However, depending on the role
156
of the network, different attributes of network structure will have a positive
impact on performance. Thus, two hypotheses proposed here are:
Hypothesis 2: The more dense the network used for contract
enforcement is the better will be the firm’s performance. The
less dense the network providing access to resources and
information is the better will the firm’s performance. Density
of the network impacts performance in a way that is
conditional upon network role.
Hypothesis 3: Diversity of a network assisting in contract
enforcement negatively impacts business performance, but the
diversity of a resource access network has a positive impact on
performance. Diversity affects performance in a way that is
closely dependent on the dominant role of the network.
8.2.3 Relationship Characteristics
The type of relationships forming the direct ties between the ego and the alter are
closely related to the concept of diversity. The central argument, put forth by
Granovetter (1974, 1973), is that individuals with weak ties, in terms of the
emotional connection and frequency of interaction, are likely to gain better
positions in the job market than those with access to predominantly strong or
close familial ties. Similar to the arguments in favor of diverse networks, weak ties
are better at linking individuals to nonredundant sources of information.
Individuals with strong ties are likely to gain access to overlapping sources of
information and are unable to reach out to the sources outside their immediate
familial network. Empirically, as related to any aspect of business performance,
the importance of weak ties has not been established. Batjargal (2003), Jansen
and Weber (2004), Renzulli, Aldrich, and Oody (2000) and Lechner and Dowling
157
(2003) conclude from their evidence that weak-tie networking, with network
members who do not know each other, is important for firm performance, business
startup, or business survival. However, Ostgaard and Birley (1996) present
evidence to the contrary that networks with a higher percentage of strong ties are
associated with more profitable firms. Also, Aldrich, Rosen, and Woodward
(1987) show that more closed networks, in terms of the number of friends as
opposed to strangers among the alters, are associated with a greater likelihood of
business startup.
Strong familial ties are often valuable sources of emotional support and
information based on shared experience that business owners and managers can
utilize during business start-up and to ensure business survival once the business
is established. For already established businesses, the ability to enforce contracts
and access sources of capital, information or favors are more important aspects of
a successful business strategy. Weak ties are more valuable as channels for
transfer of such information as opposed to strong ties between family and
extended relatives.
Whether we speak of the roles of networks in enforcing contracts or in
providing access to resources, weak ties between the ego and the alters will
positively impact business performance. Those individuals whose networks of
trusted individuals extend beyond family members enjoy a wider array of
opportunities for conducting successful and beneficial business transactions.
Nonfamilial ties are also better in securing access to resources and information
since they are less costly in terms of commitment. The absence of affection or
emotional obligation to assist one’s family implies a greater ability of businessmen
to switch to other network members to improve the terms of the exchange or
158
obtain resources and information unavailable to family members. At the same
time, nonfamilial ties are less costly in terms of obligations because individual
businessmen are not expected to trade in favors for relatives. Instead, they are
free to pursue the best course of action for their business. Access to resources is
thus best secured in a network with as few family members as possible to ensure
that the cost of utilizing network ties is the lowest. Further, to increase the
probability that the ego will access nonredundant sources of information and
capital, network alters should include strangers and not individuals who know one
another. Therefore, we hypothesize here that individuals who utilize sources
outside the family and those who trust beyond their immediate family members
should be able to produce greater results for their firm’s performance:
Hypothesis 4: The use of networks for contract enforcement and the
use of networks for access to services benefit from weak ties
which in turn translates into a positive impact on business
performance.
8.2.4 Embedded Resources
The amount of wealth, status, and power enjoyed by the members of one’s
networks can be used as measures of social capital (Lin 2001). In Chapter 6, we
saw that one’s connections can generate economic benefits through the exchange
of favors. There, we also saw that the most lucrative connections are those with
individuals who enjoy some form of political authority or hold management
positions in state-owned or large enterprises. Exchanging favors through ties with
individuals holding these positions can produce access to lucrative contracts,
exclusive licenses, access to public property, tax breaks, and provisions of other
benefits. Similarly, access to bank managers in one’s network can result in
159
favorable terms for loans (Uzzi 1999). A greater volume and diversity of resources
results implicitly in better performance. The volume of the resources embedded in
a network is more closely relevant for the role of the network in providing access
to capital and information, as opposed to enabling contract enforcement outside
the courts. The types of resources to which network members have access is not
very relevant for the use of networks in facilitating the exchange of information
about members’ reputations. As a result, the remaining hypothesis to be
examined in the current study states:
Hypothesis 5: The volume of resources accessed is a critical aspect of
networks that provide access to resources and information
necessary for successful business performance. This is not,
however, an important aspect of a network whose role is to
facilitate contract enforcement and thus will not contribute to
business performance.
8.3 Data and Network Measures
The empirical data for this analysis is derived from the survey instrument
introduced in Chapter 5 and Appendix B. The dataset is based on 274 returned
questionnaires which were mailed to a sample of SME managers and owners in
Serbia and Montenegro. The demographic characteristics of the businesses and
the respondents were introduced in Chapter 5 and Chapter 7, respectively. The
analysis of the evidence from SMEs in Serbia and Montenegro undertaken in the
previous chapter is now expanded to examine the impact of various aspects of
social networks on indicators of business performance.
160
Table 8.1: Sales growth over 3 years: total and by profit reinvestment category.
Reinvestment rate,
last year
Sales growth over past 3 years
Total / Per-
cent total ≤0 1–10 11–25 26–49 50–75 >75
0 2 6 2 4 0 0 14
6.86
1–10 15 9 9 11 0 4 48
23.53
11–25 9 15 14 7 2 2 49
24.02
26–49 2 12 2 8 4 2 30
14.71
50–75 4 4 3 7 0 3 21
10.29
over 75 2 6 11 15 0 8 42
20.59
Total 34 52 41 52 6 19 204
Percent total 16.67 25.49 20.10 25.49 2.94 9.31 100.00
8.3.1 The Dependent Variable
To capture the success of firms’ performance, the study relies on a self-reported
measure. Individual respondents were asked to state what kind of growth in sales
they had observed over the last three years of the firm’s operation.
3
Rather than
specifying an exact percentage, respondents selected the category that best
represented their case (0% or less, 1–10%, 11–25%, 26–49%, 50–75%, and over
75%). Table 8.1 shows the total number of respondents in each category of
reported sales growth.
4
Self-reported measures of business performance have been used previously in
studies of networks and their characteristics (Aldrich, Rosen, and Woodward
1987). The current choice of the survey question and its resulting measure of
performance was influenced by the concern for the privacy of the respondents,
validity of the results, and brevity of the survey instrument. In transition
3
Businesses less than 4 years old at the time of the survey were asked to report on sales
growth since the beginning of the firm’s operations.
4
Table 8.1 also divides each category of sales growth by the rate of profit reinvestment
employed as a dependent variable in the previous analysis.
161
economies, as in many other regions, the practice of misreporting business data
for tax purposes is common. As a result, respondents are less eager to answer
questions that measure sales, taxes, profits, and levels of employment. The
original expectation was that questions about the specific levels of yearly sales
would generate weak data since they would reflect the publicly reported levels of
sales and not the actual performance. Consequently, taking differences between
yearly figures to calculate sales growth over time would produce invalid and
inaccurate trends. Asking the respondents to evaluate the behavior of sales
instead was used to signal to the respondent that the investigators are interested
only in the degree of progress and that they will not focus on yearly performance.
Additional security for the respondents—that the survey would be used only for
the stated purpose and in the agreed manner—was provided by the use of
categories for possible responses. Following the lessons of Johnson, McMillan, and
Woodruff (2002b) the question offered a set of categories in order to increase
response rates and their validity. However, categorical as opposed to interval data
reveals less about the true behavior of sales. As in the case of profit reinvestment
categories, it is hoped that the loss of information has been compensated by the
increase in validity of the obtained dataset.
8.3.2 Network Measures
A successful relational approach to the study of networks builds the analysis upon
network information obtained from one member—the ego—who then provides
information about his or her direct ties with other members—the alters—and any
information about the indirect ties—the relationships among alters. The
advantages of this approach in terms of the resources and objectives of the study
162
were discussed in Chapter 5. The questions used to encourage the ego to speak
about network alters are known as name-generating questions, while questions
exploring further details about the direct ties and alter characteristics are known
as name-interpreter questions.
5
The name-generator/name-interpreter method
forms the core element of the survey that generated the evidence used in this
chapter.
The critical aim of this study is to examine how the impact of networks on
business performance is moderated by the content of exchange that the networks
facilitate. Relying directly on the interviews with the SME managers and owners,
a qualitative examination of network exchange revealed two specific types of
networks. One type of network is formed by the relationships used for contract
enforcement, while the second is composed of relationships utilized for exchange of
favors and access to resources and information. The types of relationships defining
the two network types differ in terms of the degree of commitment, as we saw in
Chapter 6. To explore measures for both of these networks, one name-generating
question was used to identify each type of the network. Name-interpreter
questions, while the same for each network type, are asked separately after each
name-generating question to acquire information about network members specific
to the network type.
6
The name-generating question used to define the trust network was stated as:
“Please list up to 5 individuals with whom you think it would be least risky to
enter into a business deal (Please use first name only, nickname or false name).”
5
Originally introduced by McCallister and Fischer (1978), the use of name-generating and
name-interpreting questions has been adopted in the work of Burt (1992), specifically as a
strategy for the measurement of individual levels of social capital. Burt’s adaptation of the
method can be found at http://faculty.chicagogsb.edu/ronald.burt/research/QUEST.pdf
6
Podolny and Baron (1997) adopt a similar technique in their study of network types and
their effects on inter-organizational mobility.
163
The question used to specify the resource access network was stated as: “Please
list up to 5 individuals who have helped you accomplish something critical for the
success of your business or whom you have contacted (or continue to contact) for
beneficial information or advice (Please use first name only, nickname or false
name).” Each of the two name-generating questions was followed by a series of
questions examining the sex and age of the alter, nature of the relationship
between the ego and the alter, frequency of contact, duration of the relationship,
and the occupation of the alter. The names were used to trace information from
name-interpreting questions to specific individuals.
7
After identifying the basic
characteristics of the direct relationships, respondents were asked to specify
indirect relationships in the networks by indicating which pairs of alters are
strangers, friends, or acquaintances. Responses to these questions were used to
generate indicators of network structure, relationship types, and volume of
network resources.
8.3.2.1 Structure
Name-generating questions specified that up to 5 individuals could be named with
respect to each network type. This limit is normally imposed to improve response
rates and because it would be prohibitively expensive to collect information about
alters, and alter pairs, for a large number of network members. The number of
individuals listed provided an indicator of network size. To ensure that size was
not limited by the structure of the question itself, the respondents were asked to
specify how many individuals they would have listed as members of each network
7
To ensure privacy, the respondents were asked to provide only the first names or nicknames,
but no last names.
164
type had they been given the opportunity to list more than 5 names.
8
Table 8.2
shows that most of the respondents indicated that they had 5 members in their
trust and resource access networks (90.51% and 69.34% respectively). Also, the
bottom portion of Table 8.2 shows that 181, or 70.15%, with non-missing
information would have listed more individuals in the trust network and 132, or
48.70%, would have expanded the list of members of their resource access
network. This means that network size is a rather poor indicator of the network’s
true size and was not used as a measure of network structure. However, other
measures of network structure based on the information about the alters in the
network were closely related to the total number of individuals listed. Hence, size,
or the number of direct ties, was not used as a direct measure of network
structure but merely as a control variable for other size-dependent measures of
network structure.
The number of relationships among alters was used as an indicator of network
density. If the ego identified a specific pair of alters to be friends or acquaintances,
the density measure was increased by 1. If the alters are strangers, there is no
relationship between them and the pair does not affect network density. The total
density for each type of the network for each observation ranges from 0 to
N(N−1)
2
.
Heterogeneity of network ties was used as a measure of network diversity,
which is another distinct characteristic of network structure. It is not based on
indirect ties among alters, but takes into account the nature of the direct ties
between the ego and the alter. The types of relationships identified in the survey
were grouped into 5 categories: family members, friends, current and former
colleagues, business associates, and acquaintances. The more diverse networks are
8
This question and technique for checking the validity of the measures of network size has
been borrowed from Podolny and Baron (1997).
165
Table 8.2: Descriptive statistics for two network types by network size
Trust Network Resource Access Network
Size Number
of firms
Density:
No.
indirect
ties
(mean)
Diversity:
Hetero-
geneity
No. weak
ties
(mean)
No. key
occupa-
tions
(mean)
Number
of firms
Density:
No.
indirect
ties
(mean)
Diversity:
Hetero-
geneity
No. weak
ties
(mean)
No. key
occupa-
tions
(mean)
0 0 0 0 0 0 6 0 0 0 0
(2.19)
1 0 0 0 0 0 23 0 0 0 0
(3.65) (8.39)
2 10 0.6 0 1.4 1 22 0.54 0 1.3 1.3
(3.65) (8.03)
3 10 1.8 1.78 3 0.4 21 1.8 0.2 2 1
(3.65) (7.66)
4 6 1 0.31 3.5 2 12 4.25 0.24 4 2
(2.19) (4.38)
5 248 4.71 0.33 4.64 1.8 190 5.58 0.27 3.93 2
(90.51) (69.34)
Number who would add more than 5 names: 181 Number who would add more than 5 names: 132
Mean number of additional names: 9.45 Mean number of additional names: 9.63
Note: Numbers in parentheses are percentages of the total sample, N = 274.
166
those whose members are connected to the ego through a greater variety of
relationship types. Following Renzulli, Aldrich, and Oody (2000), heterogeneity is
formally defined as the probability of randomly choosing people with two different
attributes from the 5 relationship categories. This probability is given by:
h = 1−
"

#family
i
total

2
+

#friends
i
total

2
+

#colleagues
i
total

2
+

#bus. associates
i
total

2
+

#acquaintances
i
total

2
#
(8.1)
A heterogeneity of h= 1 indicates perfect heterogeneity, and h= 0 indicates
complete homogeneity. A larger number of each particular type of direct
relationship lowers network heterogeneity, while greater variety in relationship
types increases diversity. Thus, the nature of the tie is not relevant on its own
(this will be captured by other measures), but only to the extent that it increases
the diversity of network ties, which is a critical component of network structure.
Heterogeneity in the sample for both types of networks ranged from 0 to 0.72.
The density and diversity of each network type are summarized by network size in
Table 8.2.
8.3.2.2 Types of Relationships
The nature of the direct ties between the ego and the alters gives us an indication
of the types of relationships that form the trust and resource access networks.
These relationships are classified into 5 categories as described above. Among
them, relationships with nuclear and extended family are considered strong ties.
The rest of the categories are classified as weak ties. The number of weak ties in
each network was calculated for each respondent. This measure was then used to
167
examine the role of weak ties in sales growth. Table 8.2 shows the mean number
of weak ties per network type and by network size.
8.3.2.3 Volume of Resources
The name-generator/name-interpreter method described above was extended to
include questions about the alters’ occupations. It is assumed that the alters’
occupation reveals information about the resources that they make available to
the ego. This is a key idea behind the alternative method for network data
collection, also known as the position-generator method (Van der Gaag, Snijders,
and Flap forthcoming). Rather than collecting information about ego and alter
ties, position-generator questions ask respondents to indicate how many
individuals they know who have a specific occupation and they are asked to
describe the nature of the relationship with that individual. This approach uses
occupations to measure the variety, range and volume of resources that are
available for instrumental action. In the survey conducted for this analysis,
occupations for specific alters were collected. This approach follows Batjargal
(2003) where individuals selected from among predetermined occupation
categories in describing their network members, or wrote the information
themselves, for subsequent coding.
A number of critical occupations were identified among the responses. They
include individuals who are employed by government ministries and agencies at
both the national and local levels,
9
bank managers, and managers or owners of
large enterprises. A unifying characteristic among these occupations is control
over sources of rents. Individuals holding these positions control opportunities for
9
The survey asked about the government at the federal level and the level of the republic
since at the time Serbia and Montenegro were republics comprising a federation.
168
obtaining income beyond the profit-generating employment of factors of
production. The number of individuals with these occupations in each network
type was used as an indicator of resource volume. The mean number of key
occupations among alters is summarized in Table 8.2 for each network type by
network size.
8.3.3 Additional Control Variables
Two groups of controls were also included in the analysis to capture the effect of
other variables that are important determinants of sales growth. The first group
of controls included characteristics of the individual manager and owner. The
manager’s highest level of education was entered as a set of dummies. We also
included a dummy indicator to capture whether the manager had experience
managing a private enterprise prior to their current position. Lastly, the
manager’s age was also included in the controls. The second group of variables
controlled for various firm characteristics. They included the log of the SME’s
age, the log of the total number of employees at the end of the previous calendar
year, and various dummy indicators for the main business activity.
8.4 Results
The goal of the regression analysis was to examine how each of the network roles
impacted reported sales growth as the indicator of business performance.
Moreover, regression analysis was used to examine which network characteristics,
depending on the role of the network, impacted reported growth in sales. Thus, in
all of the reported models, the dependent variable was the growth in sales over
169
the preceding 3 years. As we have discussed already, this variable includes
categorical data ranging from 1 (lowest sales growth category) to 6 (highest sales
growth category), which dictated the use of ordered probit regressions to
investigate the impact of network roles on sales growth.
Table 8.3 illustrates the results of regressing network characteristics on sales
growth. The results are separated by network type. All of the models tested
include the control variables. The numbers reported in the table are the
coefficients of the ordered probit model, where a positive coefficient indicates that
an increase in the particular independent variable increases the probability that a
business falls in the higher sales growth category. The z value calculated using
robust standard errors is reported in parentheses below the coefficients. The size
of the dataset did not permit the use of standard errors clustered per industry
and region.
We consider first the results for the trust network. Model 1 tested the effect of
network size and network diversity, measured by the number of indirect ties
specific to the trust network. The results of this baseline model indicate that the
density of the trust network has the hypothesized sign but is significant only at
the 10% level. Model 2 adds network heterogeneity as the second structural
measure. The density of the network again has the hypothesized sign but it is
only significant at the 10% level. Heterogeneity has the hypothesized sign but is
insignificant. Models 1 and 2 show limited evidence, then, that the structural
aspects of the trust network are determinants of sales growth.
Models 3 and 4 separately test the effects of the number of weak ties and the
number of key occupations in the trust network. Neither variable is a significant
determinant of sales growth. Model 5 finally takes the variables for network
170
Table 8.3: Regressions of network characteristics on sales growth
Trust Network Resource Access Network
variable (1) (2) (3) (4) (5) (1) (2) (3) (4) (5)
Net. size 0.20
+
0.21
+
0.20 0.27** 0.05 0.16
+
0.11 0.03 0.02 -0.14
(1.79) (1.93) (1.52) (2.59) (0.28) (1.91) (1.36) (0.23) (0.31) (-0.68)
No. indirect ties 0.05
+
0.06
+
0.06
+
-0.07* -0.08* -0.07*
(1.77) (1.87) (1.92) (-2.14) (-2.29) (-2.04)
Heterogeneity -0.38 -0.17 0.90* 1.25*
(-0.93) (-0.40) (2.28) (2.39)
No. weak ties 0.08 0.12 0.00 0.22
(0.96) (1.15) (0.03) (1.15)
No. of key occupations 0.05 0.04 0.02 0.03
(0.88) (0.59) (0.40) (0.40)
Edu: tech school -0.61* -0.68* -0.47
+
-0.58* -0.78** -0.73* -0.82* -0.62
+
-0.61 -0.85*
(-2.29) (-2.53) (-1.82) (-2.07) (-2.67) (-1.98) (-2.19) (-1.69) (-1.63) (-2.25)
Edu: 2 yr. college -0.90* -0.98* -0.93* -1.15* -1.08* -1.33** -1.57** -0.99* -1.00* -1.56**
(-2.09) (-2.41) (-2.24) (-2.52) (-2.27) (-2.83) (-3.10) (-2.25) (-2.35) (-3.18)
Edu: college incomplete -1.04** -1.12*** -1.01** -1.15** -1.22*** -1.33*** -1.44*** -1.13** -1.13** -1.47***
(-2.92) (-3.34) (-2.84) (-3.16) (-3.44) (-3.55) (-3.74) (-3.16) (-3.18) (-3.82)
Edu: college -0.44
+
-0.53* -0.49* -0.58* -0.56* -0.90** -0.93** -0.62* -0.63* -0.91**
(-1.77) (-2.10) (-1.96) (-2.15) (-2.05) (-3.01) (-3.21) (-2.22) (-2.25) (-3.17)
Edu: masters -0.86** -0.99** -0.84** -0.90** -1.04** -1.21** -1.33** -0.91* -0.91* -1.23**
(-2.69) (-3.09) (-2.60) (-2.89) (-3.26) (-2.89) (-3.05) (-2.22) (-2.21) (-2.86)
Previous experience in
private sector management
-0.05 -0.00 0.02 0.04 0.08 -0.20 -0.11 -0.01 0.01 -0.07
(-0.21) (-0.01) (0.08) (0.16) (0.29) (-0.74) (-0.44) (-0.02) (0.04) (-0.29)
Log manager age -3.46*** -3.45*** -3.18*** -3.29*** -3.49*** -3.15*** -2.97*** -3.17*** -3.19*** -3.11***
(-7.07) (-6.97) (-6.87) (-6.89) (-6.84) (-6.68) (-6.34) (-6.37) (-6.46) (-6.66)
Log SME age -0.06 -0.06 -0.05 -0.06 -0.06 -0.15 -0.15 -0.06 -0.05 -0.11
(-0.24) (-0.27) (-0.23) (-0.28) (-0.29) (-0.61) (-0.63) (-0.27) (-0.23) (-0.49)
Log no. of employees 0.16* 0.15* 0.13
+
0.10 0.13 0.14
+
0.14
+
0.13
+
0.13 0.11
(2.12) (2.08) (1.75) (1.20) (1.51) (1.67) (1.74) (1.77) (1.58) (1.39)
Main bus. activity (dummies) yes yes yes yes yes yes yes yes yes yes
N 204 204 204 204 204 204 204 204 204 204
log likelihood -287.04 -286.53 -288.15 -287.96 -285.78 -288.25 -284.91 -291.27 -291.19 -283.84
chi_square 313.21 309.84 324.78 346.70 351.64 284.87 274.84 284.73 294.66 289.62
p 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Notes: All regressions are ordered probits using sales growth in the past 3 years as a dependent variable. Numbers in the rows are coefficients. z values based
on robust standard errors are included in the parentheses. Star symbols indicate variable significance:
+
for p<.1, * for p<.05, ** for p<.01, and *** for
p<.001.
171
structure, types of relationships and volume of resources and tests them against
sales growth in a single regression. Only network density is significant at the 10%
level. The results thus far indicate that the more dense the trust network is, and
the greater the number of relationships among network alters, the greater is the
probability that the business managed by the ego will be in the higher sales
growth category.
The next 5 models report similar regressions using measures for density,
diversity, number of weak ties, and resource volume specific to the network whose
primary role is the provision of access to valuable resources. Model 1 shows that
the number of ties among individuals in one’s resource access network has the
hypothesized negative sign and the result is significant. It also shows that the
density of the resource access network negatively impacts the probability of being
in a higher sales growth category. Model 2 adds the second structural variable,
heterogeneity. As hypothesized, the heterogeneity of the resource access network
positively impacts sales growth. Model 3 tests the impact of the number of weak
ties. The coefficient is of the hypothesized sign but is insignificant and its value is
close to zero, indicating a negligible effect. The same can be said of the variable
measuring the number of key occupations in the network analyzed in Model 4.
The last model tests the impact of all network characteristics on sales growth. As
expected from the results of individual models, putting all the elements together
shows that only network density and heterogeneity have a significant impact on
sales growth. In the case of the former, this impact is negative. As hypothesized,
a higher density of ties in the resource access network leads to lower sales growth.
Network heterogeneity, however, positively impacts sales growth showing that the
172
more diverse the ties in one’s network are, the higher is the probability that the
business will be in a higher sales growth category.
Table 8.4 reports the results of the models where the relevant variables for the
trust network were included along with the variables for the resource access
network in order to examine whether the hypothesized impacts of structure,
number of weak ties, and resource volume can be confirmed when both network
types are considered in the same model. The results of model 1 show that the
characteristics of the trust and resource access networks jointly impact the growth
of sales. However, model 1 also shows, once again, that the variables for the
number of weak ties in the network and the variables measuring the volume of
resources are insignificant. As a result, they are dropped from the analysis and a
new model is estimates using only the measures of the structural characteristics
for each network type. In model 2, the number of ties among alters in the trust
network is highly significant and has a positive impact on sales growth. On the
other hand, as expected, the density of the resource access network is also
important to sales growth, but has a negative impact on the probability that the
individual business will be in a higher sales growth category. The model also
shows that the heterogeneity of the trust and resource access networks affects
performance in opposite ways. The lower the heterogeneity of the trust network,
the greater the network impact on sales growth. However, the greater the
heterogeneity, the greater the the impact of the resource access network on a
business’ sales growth.
To check the robustness of the results, model 3 also includes a dummy
indicator for the region in which the business operates. The region and activity
indicators are interacted in the model and the implications of model 2 are
173
Table 8.4: Regressions for sales growth as a function of network roles
Variable (1) (2) (3) (4)
Trust network
Net. size 0.38 0.28* 0.30
+
0.23
(1.46) (2.03) (1.93) (1.46)
No. indirect ties 0.08* 0.09** 0.15*** 0.13**
(2.11) (2.77) (3.35) (2.82)
Heterogeneity -0.74 -0.89* -1.18* -0.93
+
(-1.50) (-1.97) (-2.39) (-1.82)
No. weak ties -0.09
(-0.59)
No. key occupations 0.12
(1.64)
Resource access network
Net. size -0.29 -0.02 -0.05 -0.05
(-1.20) (-0.15) (-0.34) (-0.37)
No. indirect ties -0.09** -0.09** -0.15** -0.12*
(-2.75) (-2.75) (-2.90) (-2.12)
Heterogeneity 1.53** 1.07* 1.49* 1.39*
(2.70) (2.16) (2.49) (2.27)
No. weak ties 0.21
(1.15)
No. key occupations -0.05
(-0.75)
Controls
Edu: tech school -0.58 -0.57 0.14 -0.28
(-0.71) (-1.26) (0.24) (-0.48)
Edu: 2 year college -1.46
+
-1.28* -0.59 -1.18
+
(-1.67) (-2.22) (-0.93) (-1.86)
Edu: college incomplete -1.38
+
-1.26** -0.52 -0.89
(-1.77) (-2.86) (-0.91) (-1.55)
Edu: college -0.48 -0.59 0.98 0.50
(-0.61) (-1.31) (1.64) (0.82)
Edu: masters -0.95 -1.17* 0.28 -0.46
(-1.12) (-2.08) (0.41) (-0.66)
Previous experience in
private sector management
0.17 0.04 0.91* 0.54
(0.39) (0.17) (2.47) (1.42)
Log manager age -3.42*** -3.21*** -4.56*** -4.17***
(-6.33) (-6.65) (-7.79) (-6.68)
Log SME age -0.07 -0.09 -0.56
+
-0.69*
(-0.33) (-0.37) (-1.93) (-2.38)
Log no. of employees 0.10 0.17* 0.09 0.11
(1.16) (2.02) (0.83) (1.04)
Main bus. activity (dummies) yes yes yes yes
Activity + Region (dummies) yes yes
Belief in court effectiveness 0.20*
(2.36)
N 204 204 204 201
log likelihood -276.09 -277.80 -253.72 -249.44
chi_square 118.05 268.67 360.70 355.26
p 0.00 0.00 0.00 0.00
Notes: All regressions are ordered probits using sales growth in the past 3 years as a dependent variable.
Numbers in the rows are coefficients. z values based on robust standard errors are included in the parentheses.
Star symbols indicate variable significance:
+
for p<.1, * for p<.05, ** for p<.01, and *** for p<.001.
When region and industry controls are included in the model they are interacted.
174
confirmed. The last model in Table 8.4, in addition to the interaction effects
between the region and business activity indicators, also includes the measure of
the business owners’ and managers’ confidence in courts. This variable was
utilized extensively in the analysis in Chapter 7 and it measures the degree to
which individuals are confident in the courts’ ability to enforce contracts and
resolve disputes. This variable is included to test whether trust network
characteristics continue to play a role in sales growth even after measures of
formal contract enforcement are included in the model. As the results of model 4
indicate, the density and heterogeneity of the trust network are still significant
determinants of sales growth although heterogeneity is now significant only at the
10% level. The density and heterogeneity of the resource access network also
continue to play a role in sales growth, as predicted in earlier models. In addition,
belief in court effectiveness positively and significantly impacts the probability
that a business will have higher sales growth.
The coefficients reported by the ordered probit regression models in Table 8.4
are not easily interpreted. However, we can use the results of the final model in
Table 8.4 to calculate the predicted probabilities that a manager in the sample
will report higher sales growth, given any one of the network variables, while
holding the other independent variables at their mean. The upper section of
Table 8.5 illustrates the predicted sales growth category given the density of the
trust network. The lower section of the table illustrates similar calculations for
the degree of network heterogeneity. The last column shows the weighted-average
sales growth for each specific value of the independent variable. The
weighted-average is calculated using the midpoint for each sales growth category
and weighing it by the percentage of businesses predicted to report that particular
175
Table 8.5: Predicted sales growth rates (trust network)
Density:
number of
alter-alter ties
Percentage of firms in sales growth category
Weighted
sales
growth ≤0 1–10 11–25 26–49 50–75 >75
0 17.02 46.14 26.21 10.00 0.30 0.33 11.25
1 13.93 44.23 28.63 12.32 0.41 0.48 12.66
2 11.24 41.77 30.79 14.94 0.56 0.70 14.20
3 8.95 38.88 32.58 17.85 0.76 1.00 15.86
4 7.02 35.65 33.93 21.00 1.00 1.40 17.62
5 5.43 32.20 34.79 24.34 1.30 1.94 19.52
6 4.14 28.66 35.11 27.79 1.67 2.64 21.54
7 3.11 25.13 34.87 31.25 2.10 3.54 23.68
8 2.30 21.71 34.10 34.62 2.59 4.68 25.94
9 1.68 18.47 32.81 37.79 3.15 6.10 28.34
10 1.20 15.49 31.08 40.63 3.76 7.84 30.85
Diversity:
heterogeneity
of ego-alter ties
0 2.77 23.80 34.64 32.56 2.28 3.95 24.54
0.32 5.28 31.83 34.85 24.70 1.34 2.00 19.72
0.44 6.59 34.80 34.20 21.82 1.07 1.52 18.08
0.56 8.14 37.63 33.17 19.07 0.85 1.14 16.54
0.625 9.09 39.08 32.47 17.65 0.74 0.97 15.73
0.72 10.61 41.07 31.28 15.66 0.61 0.77 14.62
1 16.17 45.70 26.87 10.58 0.32 0.36 11.61
Note: The percentages reflect the probability of being in each sales growth category given the number of
relationships among alter-alter pairs and the degree of network heterogeneity. The results of model 4 are used
to calculate the probabilities. The weighted sales growth rate is calculated using the midpoint of each
category for sales growth.
176
level of sales growth. As postulated in hypotheses 2 and 3, as we go from lower to
higher levels of network density, the weighted sales growth increases from 11.25%
to 30.85%. As the heterogeneity of ties in the trust network increases from 0 to 1,
weighted sales growth decreases from 24.54% to 11.61%.
Table 8.6 shows the calculations using the density and diversity measures for
the resource access network. The last column in the table illustrates most clearly
the hypothesized reversal in the effect of network structure on sales growth due to
the different type of network. In the last column of Table 8.6, the
weighted-average sales growth decreases as the network density increases (30.37%
to 12.10%), and increases as the heterogeneity of ties forming the resource access
network increases (from 14.80% to 39.14%).
8.5 Implications
Table 8.4 shows that characteristics of both the trust network and the resource
access network are significant determinants of sales growth. This confirms the
first hypothesis that both network roles impact business performance. However,
network characteristics do not impact performance in a uniform manner. For the
trust network to boost sales growth, density must be high and heterogeneity low.
In other words, ties forming the trust network should overlap and originate in
similar types of relationships for the network’s role in contract enforcement to
have the greatest impact on sales growth. For the resource access network, it is
important that its density be low and its heterogeneity high. Networks formed by
various types of relationships between the ego and the alters, and whose alters are
strangers to one another, provide the most valuable access to resources and
information that translate into increased sales growth. These findings confirm
177
Table 8.6: Predicted sales growth rates (resource access network)
Density:
number of
alter-alter ties
Percentage of firms in sales growth category
Weighted
sales
growth ≤0 1–10 11–25 26–49 50–75 >75
0 1.28 16.02 31.43 40.13 3.64 7.49 30.37
1 1.73 18.80 32.97 37.47 3.08 5.94 28.08
3 3.06 24.94 34.85 31.44 2.12 3.60 23.80
4 3.99 28.19 35.11 28.25 1.72 2.75 21.82
5 5.13 31.45 34.90 25.07 1.38 2.07 19.94
6 6.52 34.65 34.24 21.96 1.09 1.54 18.17
7 8.19 37.70 33.14 19.00 0.84 1.13 16.49
8 10.15 40.51 31.65 16.22 0.65 0.82 14.93
9 12.45 42.98 29.82 13.67 0.49 0.59 13.47
10 15.08 45.04 27.72 11.38 0.36 0.42 12.10
Diversity:
heterogeneity
of ego-alter ties
0 10.34 40.74 31.50 15.99 0.63 0.80 14.80
0.32 4.40 29.46 35.09 27.01 1.58 2.46 21.07
0.44 3.06 24.95 34.85 31.43 2.12 3.59 23.79
0.56 2.08 20.62 33.73 35.69 2.77 5.12 26.72
0.625 1.67 18.41 32.78 37.85 3.16 6.14 28.39
0.72 1.19 15.40 31.02 40.71 3.78 7.90 30.93
1 0.41 8.31 24.00 46.20 5.80 15.29 39.14
Note: The percentages reflect the probability of being in each sales growth category given the number of
relationships among alter-alter pairs and the degree of network heterogeneity. The results of model 4 are used
to calculate the probabilities. The weighted sales growth rate is calculated using the midpoint of each
category for sales growth.
178
Hypotheses 2 and 3 and illustrate that the impact of network structure is
conditional upon the role played by the network.
None of the models suggested that the number of weak ties and the volume of
resources impact sales growth, regardless of the network type. As a result, we
were unable to confirm Hypotheses 4 and 5 regarding the role of weak ties and
resource volume. However, this conclusion does not imply that network resources
are unimportant to sales growth. To construct the measures of resource volume,
certain occupations were selected for providing the greatest value of resources to
be used in instrumental activities. However, this classification is incorrect if the
rankings of occupations differ across business activities. In that case, the measures
based on the presence or absence of specific occupations among the alters in the
networks do not capture the value of the resources for specific business activities
in the sample. The control variables used in the models indicate that most of the
indicators for the main business activity are significant predictors of sales growth.
A cross-tabulation of sales growth categories with the indicators for business
activity, however, shows that some table cells are empty. Due to its small size, the
sample does not capture the full sales growth variability across each type of
business activity. As a result, it is not possible to include measures of resource
diversity based on the rankings of occupations specific to business activities. An
alternative approach is to use preset occupation rankings as the basis for resource
volume measures but then to restrict the sample to a single business activity. For
samples that include businesses operating in various sectors, a safer approach to
measuring the impact of embedded resources is to move away from any
occupation rankings and to look for more universal measures of resource value.
One alternative examined in the analysis above considered the impact of resource
179
diversity by testing a variable that measured the number of different occupations
present among alters (results not reported here). The variable did not show any
significant impact on the growth of sales. However, alternative measures of
resource variety should be examined before any conclusions can be made about
the impact of embedded resources on firm performance.
To understand the impact of networking activity on business performance, this
chapter focused specifically on the content of exchange facilitated through network
relationships. To do so, we have distinguished between two types of networks: 1)
networks used primarily as an informal mechanism for contract enforcement
through the reputation mechanism and the formation of interpersonal trust; and
2) networks of individuals who provide exclusive access to resources and
information through the exchange of favors. The main contribution of the analysis
is to demonstrate empirically that the effect of network characteristics on business
performance is not uniform across network type. Depending on the network type
considered, network structures will impact business performance differently.
The results presented here show that individuals with trust networks whose
direct ties are similar in nature, and whose members are not strangers to one
another, are more likely to be in charge of companies that have experienced
growth in sales in the previous three years. Also, the results indicate that to access
resources with the greatest benefit for business performance, business managers
and owners have to utilize networks based on a diverse set of relationships and
there should be a low level of interaction between different network members.
Both network types are important for the growth of sales. Among network
characteristics, structure has the strongest impact on performance. However, the
impact of network structure is dependent on the type of the network and its
180
primary role. Here, considerations of content of exchange and network types
appear for the first time in an analysis of business performance to show that not
all network characteristics have a uniform impact on performance. Trust networks
are better for performance when they are more dense and less heterogenous.
Resource access networks are better for business success if they are heterogenous
and less dense. Future socio-metric studies are needed to examine the degree to
which the membership in trust and resource access networks differs. The survey
results analyzed in this chapter indicate that the membership in the two networks
rarely overlaps. Socio-metric studies are needed to examine further the structural
characteristics of different network types and the degree to which network roles
are dependent on their memberships.
181
Chapter 9
Economizing on Transaction Costs:
The Role of Social Networks
In Chapter 6, we encountered the owner of a publishing company who
begrudgingly invests resources in socializing and vacationing with his partners to
ensure that they honor agreements involving his business. The chapters that
followed have used quantitative analysis to examine whether such investments in
network ties generate economic payoffs and how different network roles impact
business performance. This chapter probes further into the incentives that
businesses have for relying on personal relationships that demand extensive
sacrifices of time and effort when conducting their everyday business transactions.
To complement the results of the quantitative analysis, this chapter uses
qualitative interviews with SME managers and owners to examine whether forms
of exchange result in avoidable costs of doing business. One possible conclusion
that can be derived thus far is that greater reliance on social networks ought to
translate into greater economic welfare since social capital accessed through
182
networks of relations and connections is beneficial for business performance. This
chapter explores whether qualitative evidence offers support for this conclusion.
Before we examine the qualitative evidence, we must step back and consider
the basic elements of an economic actor’s decision to use network relationships to
conduct a business transaction or to approach an unknown partner in the
impersonal marketplace. This chapter adopts the transaction cost framework to
examine more closely the determinants of an individual’s decision to utilize social
networks. The chapter then shifts focus to the social costs resulting from the
dominant role of social networks in exchange. Lastly, the chapter explores the
possible role of social networks in preventing beneficial reforms of
market-supporting institutions and comments on why the incidence of personal
exchange and the related social costs demand policy attention.
9.1 Transaction Cost Framework
The framework of transaction cost economics rests on the behavioral assumption
that economic actors do not have unlimited cognitive abilities and that rationality
is bounded by the information actors can process and their ability to anticipate
future exchange conditions (Williamson 1985; Rindfleisch and Heide 1997). The
second essential assumption is that of opportunism, which recognizes that actors
have an incentive to exploit imperfections in rationality (Williamson 1985; Pilling,
Crosby, and Jackson 1994). The two assumptions together imply that
transactions between economic actors cannot be costless because investments
must be made to curb opportunism and ensure that the terms of agreements are
followed even in cases of unpredictable future exigencies. Costs must be borne
when negotiating the agreement ex-ante, measuring the good or service quality
183
ex-post, and if necessary, enforcing the original agreement. The standard claim of
transaction cost economics is that when the costs of transacting in the market
become sufficiently high, goods or services should be provided in-house by
organizing production and exchange within the hierarchical structure of the firm.
Unfortunately, this comparison between transaction governance structures in
the markets and the firm assumes that production benefits remain unchanged
(Dietrich 1994). Once we consider the costs of production, an internal solution is
not necessarily a feasible option for SMEs, especially for those businesses
operating in transition environments. SMEs generally lack the financial and
human capital necessary to expand production in a radically new direction, which
would be required of them for the in-house production of goods otherwise
obtainable in the market. Size limits the ability to absorb administrative costs
associated with increased scope and volume of production while the managers
may not have the skills and experience necessary for new lines of production. In
sum, SMEs may be too small to internalize the production of goods and services
available in the market. Hence, any reduction in transaction costs would possibly
be offset by an increase in the costs of production.
Although internal production is not an attractive option for SMEs, we can
identify another alternative for curbing opportunism and delivering transactional
assurance. The private order literature has shown that when formal procedures
and the rule of law fail to provide effective forms of contract enforcement, other
self-enforced mechanisms tend to emerge to limit cheating. Chiefly focusing on
the reputation mechanism, this literature has demonstrated that the threat of
punishment by community members (i.e., expulsion) is sufficient to sustain
cooperation in the group. Moreover, the reputation mechanism does not have to
184
be restricted to groups defined by religious affiliation, ethnic identity, or
geographic region. The evidence presented in Chapter 6 on the role of social
networks in facilitating commercial exchanges indicates that the reputation
mechanism can sustain transactions on a much broader scale by using network
relationships for information transfers across various group memberships. In other
words, networks provide cross-cutting relationships of various origins, which by
their transitive nature allow information to be transmitted beyond a single group.
In such a manner, networks provide a legitimate alternative form of exchange
governance rooted in the private mechanism for contract enforcement but
applicable outside a single geographic location or ethnic community. Thus, the
relevant choice for SME managers and owners that must be investigated further is
not between firms and markets, also known as the "make-or-buy" decision, but
between impersonal exchange in the market and personal exchange made possible
through social networks.
9.2 Choice of Transaction Governance
Previous chapters have demonstrated that social networks facilitate exchange
through a private enforcement mechanism and offer exclusive access to preferential
treatment, resources and services. However, there is no indication that social
networks would always be the preferred structure for organizing exchange. While
the beneficial roles of social networks can translate into transaction cost savings,
Chapter 6 has shown that networks are not without cost. To understand how
conducting personal exchange through networks can result in benefits despite the
involved costs, we must compare the costs associated with the market and the
costs associated with social networks (Rindfleisch and Heide 1997).
185
9.2.1 Transaction Cost Elements
To carry out a transaction impersonally in the market, an actor has to invest time
and resources to locate sellers and negotiate a price for the desired goods or
services. After a seller has been identified, resources must be devoted to
negotiating and drafting an agreement that will govern the transaction. Next, the
actor must monitor and verify the quality of the goods or services ordered to
ensure that the terms of the agreement are followed. If the original agreement is
violated, the actor must bring the dispute to the court that has the authority to
enforce the original contract. Given some nonzero probability of violation, it will
be necessary to invest resources to insure the original terms of the agreement.
Together, these costs are equal to the direct costs of transaction management.
There are also costs associated with choosing an inferior option for exchange
governance (Rindfleisch and Heide 1997). When a transaction is carried out in the
market, the opportunity cost represents benefits that could have been secured by
carrying out the transaction through one’s social networks. These include the
monopoly rights or price differential that can be secured through contacts in the
networks. Since exclusive access and preferential treatment cannot be obtained
through impersonal exchange in the market, the value of their loss has to be
considered when evaluating the total cost of carrying out a market transaction.
The costs of transacting in the market are summarized in Table 9.1, and they
include the direct costs of managing the transaction as well as the opportunity
costs of choosing an inferior form of exchange governance (Rindfleisch and Heide
1997; Rao 2003).
To carry out a transaction in the network, one must first identify direct or
indirect ties that can put the individual in contact with the supplier of the
186
particular good or service. Once the supplier is identified it is necessary to invest
resources in order to establish a personal relationship with the actors, or if the
relationship precedes the commercial contact, to maintain this relationship.
Individuals must spend time and financial resources visiting each other’s factories
and offices, or arranging to meet in social engagements, such as parties and
vacations, in an attempt to become more familiar with each other’s business
practices and trustworthiness. Resources must also be devoted to negotiating and
drawing up a contract. Interview responses with managers and owners have
demonstrated that standard contracts are drawn even in the case of transactions
within networks because they are legally required for commercial transactions,
and because they represent the ultimate safeguard when relationships break
down. If an individual wishes to monitor and verify the quality of the goods or
services obtained, they will also incur measuring costs. Since networks enable the
reputation mechanism to curb opportunistic behavior among network members
and to encourage the formation of interpersonal trust, contract violations are
expected to be rare. If contracts are violated, enforcing the original agreement
becomes necessary. However, when a transaction is conducted in the network,
parties do not have to resolve their disputes in courts. Instead, actors can use
close relationships to help devise a mutually beneficial solution and to avoid going
to court, which would permanently damage the social and commercial
relationship. As a result, in personal exchange within networks, enforcement costs
are minimal or close to zero.
Opportunity costs also exist when transactions are carried out in the network.
Again, we can think of them as benefits from transacting in the market that are
lost when one conducts exchange through the network, or as costs that might be
187
Table 9.1: Summary of transaction costs
Transaction costs Markets Networks
Search and investment S
i
S
j
Contracting X
i
X
j
Measuring M
i
M
j
Enforcement p∗E
i
p∗E
j
Opportunity O
i
O
j
Total costs TC
M
=S
i
+X
i
+M
i
+p∗E
i
+O
i
TC
N
=S
j
+X
j
+M
j
+p∗E
j
+O
j
incurred in the network but would not arise in a market transaction. These costs
are related to the potential need to return a favor that has resulted in exclusive
access to goods and services or a supply of goods at favorable prices. If the
individual is called upon to reciprocate a favor, and he or she refuses, the
relationship would be severed and the reputation of the individual refusing to help
tarnished in the network. Further, reciprocity often entails that individuals trade
with specific actors to the exclusion of other, nonnetwork members. This
commitment can be costly when better offers outside the network are rejected to
maintain existing ties (Cook, Rice, and Gerbasi 2004). Overall, network
commitment can result in costly reductions in productivity and efficiency. These
potential costs of using relationships to conduct exchange do not arise in
impersonal markets where there is no expectation of return favors among
participants in exchange. In network transactions, reciprocity results in
opportunity costs that can be equated with the present value of the return favor
to be required at a future point. The total costs of transacting in the network are
summarized in Table 9.1 and include direct costs, such as search and investment
costs, contracting costs, measuring costs, and enforcement costs, as well as the
potential opportunity costs of choosing an alternative form of exchange
governance.
188
The choice of transaction governance is motivated by the objective to
economize on transaction costs. The transaction setting which results in the
lowest level of transaction costs will be the chosen governance structure.
Impersonal transactions will be preferred when transaction costs in the network
exceed those in the market (TC
n
> TC
m
), and personal exchange facilitated by
social networks will be used when the transaction costs in the market exceed the
transaction costs in the network (TC
n
< TC
m
). Hence, some transactions will be
carried out through networks rather than impersonal markets despite costly
investments necessary to build and maintain network ties. The transaction cost
framework and examples from interviews with managers and owners in Serbia and
Montenegro can assist us in determining the conditions under which personal
exchange will be preferred to impersonal markets.
9.2.2 Determinants of Transaction Costs
So far we have addressed how the level of transaction costs affects the choice of
the governance structure. Following Williamson’s (1979) framework illustrated in
Fig. 9.1, we can now relate the cost elements described above to the underlying
transaction characteristics in order to determine the appropriate form of
governance. Given the essential behavioral assumptions of bounded rationality
and opportunism, specific transaction characteristics, namely uncertainty,
frequency and asset specificity, determine the level of transaction costs
(Williamson 1979; Williamson 1985). All three transaction characteristics impact
the ability to constrain opportunism in exchange. The uncertainty of the
transaction indicates the degree to which it is possible to predict future
contingencies and developments in the transaction. The degree to which
189
Form of
transaction
governance
Transaction
Costs
Transaction
characteristics:
Uncertainty
Frequency
Specificity
Figure 9.1: Transaction cost framework
investments made by parties to the transaction are specific to the exchange is
known as asset specificity and is dependent on the value such investments would
have in the best alternative use (Pilling, Crosby, and Jackson 1994). Lastly,
frequency refers to the number of transactions per given exchange relationship
such that greater frequencies pose greater risks associated with losing the current
trading partner (Pilling, Crosby, and Jackson 1994). Together, uncertainty,
frequency and asset specificity determine the level of transaction costs—the costs
of managing the transaction and constraining opportunism in exchange.
To simplify the discussion, we limit ourselves to transactions of an
intermediate level of uncertainty. Holding uncertainty constant allows us to vary
transactions only in terms of frequency and specificity, and to look at the
implications of each for transaction cost levels (Williamson 1979). Specificity in
Williamson (1979) refers to the degree to which any parties in exchange (think
suppliers) are investing resources specific to the transaction. In developed
markets, it can be assumed that suppliers are numerous and monopoly rights over
specialized resources are absent (Williamson 1979, 246). This assumption cannot
be upheld in the emerging markets of Serbia and Montenegro. Depending on the
good or service in question, the degree of competition can be low, and monopoly
190
ownership rights abound as a result of the privatization reforms and the manner
in which they were implemented. For example, when an actor wishes to obtain a
specific good that only one firm is able to supply, then the supplier can exert
significant monopoly power over the buyer. Alternatively, if a buyer orders
specific material for a production that is currently unique and in low demand
elsewhere in the market, the supplier has no alternative use for the material if the
buyer backs out of the transaction. Consequently, in the case of a transition
economy, transactions can be specialized due to the degree of market competition
and market demand, or transaction-specific investments.
9.3 Illustrating the Choice of Transaction
Governance
The manner in which the frequency and specificity characteristics of transactions
interact to determine the level of transaction costs is essential to understanding
the choice SMEs make between transacting through social networks and relying
on impersonal markets. In terms of frequency, we will consider two types of
transactions, namely onetime, nonrecurrent purchases and repeated or recurrent
purchases. We will also consider three levels of asset specificity: nonspecific,
mixed, and specific transactions. To determine the governance structure
appropriate for each case, we next illustrate the process of economizing on
transaction costs using examples for each pair of transaction characteristics in
Table 9.2. The choice of transaction governance and the examples illustrating
each transaction type are included in the table.
191
Table 9.2: Transaction characteristics and illustrative examples
SPECIFICITY
Nonspecific Mixed Specific
FREQUENCY
Non-
recurrent
One time purchase of
standard type wire
Business plan
writing/consulting services
Construction contractor
MARKET
MARKET/
NETWORKS (relations)
NETWORKS
(connections)
Recurrent
Recurring purchase of
primary materials for
a bakery production
Line of products from an
exclusive distributor
Purchasing steel tubes of
specific measurements
and durability
NETWORKS
(relations)
NETWORKS
(connections)
NETWORKS
(connections)
To illustrate the transaction costs involved in a onetime purchase of a
nonspecific good we consult the manager of a firm supplying construction
material, primarily wiring, in Belgrade, Serbia. The owner of the firm was once
employed in the publicly owned wire factory. He has used his contacts at that
firm to start his own wholesale business, which he supplies with low-cost materials
directly from the factory. When the business needed to fulfill a special order by a
customer in Belgrade, the manager found that that the existing factory could not
supply the wire according to desired specifications. Instead of reaching out for
assistance to contacts from his previous employment, the owner decided to obtain
the wire from a Chinese factory whose representatives he had met at an
exposition in Germany. Ordered from China via email, the wire was successfully
delivered to Serbia. The manager of the company explained that for every
transaction the risk that the other party will cheat is enormous, especially when
one is unable to enforce the contract in the commercial courts of Serbia or the
courts overseeing international commercial transactions. However, the manager
said, “you have to trust others at least a little bit, otherwise you won’t be in
business.” Hence, in the case of a onetime transaction he was confident to
interpret the size of the Chinese firm and its participation in an international
192
exposition as a proxy for good business reputation. The manager was asked to
explain why he preferred to work with the Chinese firm as opposed to other firms
whose owners and managers he knew personally. He explained that locating the
company over the internet was easy, he did not have to spend considerable time
arranging the order and calling “over-and-over again about the deal,” and he did
not have to “ask anyone to do anything special” for him. In this case, the cheapest
and quickest way to complete a nonspecific and nonrecurrent transaction was to
carry it out in the impersonal market.
A telecommunications firm in Montenegro, which needed to obtain a business
plan in order to apply for the right to lease specific radio frequencies managed by
the state, provides us with an example of a nonrecurrent, mixed transaction. The
business plan in this case is a mixed good because the it has to be tailored to the
applicant, but can also borrow from a general format used for similar applications.
To select a consulting firm that offers professionally written business plans, the
manager of the telecommunication firm made inquires among his contacts in the
agency, which manages the radio spectrum. There he learned that the son of the
agency’s head runs a business consulting firm. The manager decided to do
business with this particular firm, outside the area of his operations no less, and
at an approximate 15% price premium, because he knew that the business plan
written by this firm would provide him with the best chance of obtaining the
approval from the commission. The consulting firm was informed about the
details that must be present in the plan in order to win approval by the
commission, and the head of the commission was unlikely to deny applications
presenting a business plan created by his son’s firm.
1
The most efficient way for
1
In this discussion, we must put aside the issues of nepotism and conflict of interest that
appear when the head of the commission reviewing applications for radio frequencies allows his
193
the manager of the telecommunications firm to purchase the business plan was to
use his personal contacts to obtain services from a specific business able to
guarantee a successful outcome in the application process for a frequency lease.
The examples of nonrecurrent, yet highly specific transactions include a
construction company operating in both Serbia and Montenegro. The company
must regularly hire contractors to work on specific projects for which it has
secured licenses and financing. The particular transaction in this example involves
a trade in services for a single project and any investments made by the
contractors are highly specific. The manager of the construction company expects
contractors to maintain the quality standards mandated by the project and to
complete their assignments on time so that agreements toward financiers and
clients can be fulfilled. In such a line of business, explained the manager,
enforcement of contracts is critical, yet he is unable to rely on the courts since
they are slow or corrupted and the financial guarantees offered by banks are too
expensive or unenforceable. As a result, search and contracting costs associated
with identifying the right contractors are the most significant costs of transacting.
The manager has to find companies that can do the required jobs and he must
create foolproof contracts that specify in detail all the possible contingencies and
damages that will be awarded if the contracts are broken. The manager estimated
that he would be willing to sacrifice 6% of his profits in order to eliminate these
search and contracting costs. He added that he is usually able to do so by
contracting only with businesses he knows or who come recommended from
professional and social contacts. To attract these contractors to work with him,
son to earn a profit by supplying business plans required for such applications. The intention
here is to present the transaction details from the perspective of the telecommunication firm
purchasing consulting services.
194
the manager offers to pass on to them some of the profit margin he retains by not
transacting with contractors identified through impersonal markets.
Another advantage of “working only with people you know” is the ability to
lower monitoring costs. Normally, the manager invests resources and time in order
to keep contractors in a “dependent” position with respect to labor and materials
to limit their opportunities to take advantage of the construction company. Such
monitoring is easier to perform with businesses with whom the manager is already
familiar. He recognized, however, that working with someone in the network of
contacts creates other costs. As already stated, in order to attract their business,
the construction company has to offer attractive conditions and sacrifice profit
margins. Sometimes, the construction company manager is even willing to build
additional time into his projects in order to wait for a specific contractor to finish
with an existing engagement and join his team. These costs, however, are offset by
the savings in search, contracting, and monitoring. As a result, the construction
company manager always prefers to work within his network as opposed to
working with unfamiliar contractors identified in the impersonal market.
A bakery that supplies products across Serbia and Montenegro is frequently
involved in transactions that are nonspecific yet recurrent. The lack of clear and
detailed regulation is a significant problem with respect to food production and
packaging, which raises special concern for the quality of inputs used in the
bakery’s production. Monitoring the production and packaging of primary
resources is difficult because the bakery has no means to inspect the their quality
once they are obtained directly from the suppliers’ factories. One way the bakery
can lower the monitoring costs is to contract only with businesses with which it
already has a successful business relationship. Enforcement, search and
195
contracting costs are not significantly different than those the bakery would incur
by offering the contract to the best bidder in an impersonal market. However, the
concern with input quality and the problems it can cause in the production
process have prompted the manager to save on monitoring costs by contracting
with businesses known in his commercial and social network. The manager did
not feel committed to these relationships in the sense that he would not refuse a
better offer from other companies of an equal or better reputation. The bakery
and its suppliers were not bound by a sense of reciprocity which meant that
better offers could be pursued. However, the alternative offer would only be
entertained if it came from a company that also enjoys a sound reputation in the
manager’s network. This shows that exchange between the bakery and the
suppliers of primary inputs is rooted in social and business relationships, but the
manager prefers to rely on personal relations and not connections which involve a
higher degree of commitment.
Similar reasoning for preferring personal exchange to impersonal markets
applies to the case of a recurrent mixed transaction conducted by the owner and
manager of a wholesale supermarket in Montenegro. The manager used
connections to persuade an exclusive distributor of foreign cosmetics to supply the
supermarket at prices lower than those offered to other firms in the area. The
goods obtained in this case are not specific in terms of the degree to which they
are tied to the needs of the buyer. However, the exclusive distributor of foreign
cosmetics effectively has monopoly rights over the sale of products in the region,
and in the sense of market competition, the goods in this example are mixed. The
use of connections to obtain a favorable deal from the distributor probably did
not result in any savings in terms of search, measurement, contracting or
196
enforcement. If anything, the relative search costs were higher than those that
would have been incurred in impersonal exchange because the supermarket
manager had to search for direct or indirect ties that would allow him to trade
favors with the distributor. However, the opportunity cost of transacting in the
market in this particular case would have been significant because the benefits in
terms of the lower supply price would not have been realized. In this case,
transacting through the network was considered a superior choice.
A final example illustrates the decision of what governance mechanism to
employ for a transaction that is both recurrent and specific. The manager of a
metal furniture manufacturing firm in Montenegro relies on an outside supplier for
steel tubes used in the production of all furniture designs. The tubes must be
produced according to specific measurements and quality standards so that the
manufacturer can guarantee the durability of materials and furniture quality to
the customers. Unfortunately, steel tube producers and suppliers can hide the
true quality of their products behind inadequately specified and enforced
regulation. As a final user, the furniture company has no way to monitor the
production and certification process of the materials purchased. Instead, the
manager relies on personal connections in choosing suppliers. He carefully
maintains connections with them to ensure that they do not take advantage of the
working relationship. Working with someone “they know,” explained the manager,
is the best way to lower monitoring and enforcement costs. Even though other
costs do not differ between the two transaction settings, the manager prefers to
rely on personal exchange in order to enhance the predictability and ensure the
continuity of his production process.
197
9.4 Transaction Characteristics and the Use of
Networks
Table 9.2 summarizes the examples from Serbia and Montenegro which illustrate
how transaction characteristics impact the choice of the appropriate governance
structure. When a transaction is specific either in terms of related investments on
the supplier’s side, or in terms of market competition, buyers feel the need to
protect themselves against the advantageous position of the seller causing
enforcement costs to become the dominant factor in decision-making. Networks
offer lower enforcement costs and thus often appear as an attractive option for
transacting. Moreover, as specificity increases and we move horizontally across
the examples in the table, opportunities to secure exclusive deals and other
non-price benefits grow. To capture these benefits, network ties must be utilized
but the search of the appropriate channels for influence increases the search costs
in the network relative to the market. Thus, when specificity increases,
enforcement costs and opportunity costs in the markets rise, but so too do search
and investment costs in the network. In the case of mixed and nonrecurrent
goods, the relative difference in the transaction costs associated with the two
forms of transaction governance is ambiguous. Hence, we find examples of both
personal and impersonal exchange for mixed and nonrecurrent transactions.
However, for highly specific goods, transacting through social networks tends to
be the preferred choice.
Moving vertically in the table and increasing the frequency of the transactions
places monitoring and enforcement costs at the center of analysis. In most of the
cases, managers were concerned that working frequently with a particular firm
198
challenges their ability to continually ensure the quality of the goods supplied and
enforce the terms of the agreement. As a result, they prefer to work with
businesses they know or can identify through existing network contacts. As the
frequency increases, relative monitoring and enforcement costs in market
transactions rise, encouraging managers to choose personal exchange.
Table 9.2 also illustrates the choice between personal relations and
connections. The interview responses have shown that when the primary objective
of transaction participants is to lower enforcement costs, information supplied by
networks of relations proves to be a sufficient deterrent against opportunistic
action. Connections, on the other hand, are used when increased search and
investment costs that precede an exchange of favors can be justified with benefits
exclusive to the network exchange. When the benefits do not exceed the search
and investment costs, relations rather than connections are used to facilitate
network exchange.
In Table 9.2, we see that network exchange becomes attractive as transactions
increase in frequency and asset specificity. However, no mention has been made of
the uncertainty that was held at an intermediate level in the table. Recall that
Williamson (1979) defines uncertainty as the degree to which market actors are
unable to predict the circumstances surrounding exchange and thus cannot
specify their full decision tree. The implications of the transaction cost approach
are that as it becomes more difficult for actors to prepare detailed contracts that
can guide exchange under unforeseen conditions, transaction costs associated with
monitoring, contracting and enforcement typically rise, making the option of
producing the good in-house more attractive. In the current work, however, we
are not considering the choice of hierarchy as a governance mechanism and are
199
instead interested in examining the impact of uncertainty on the likelihood that
individuals will resort to network relationships for organizing exchange. The
impact of uncertainty in this case is much less clear.
Perhaps considering the characteristics of governance structures more closely
can help us determine the impact of uncertainty. Transaction costs are minimized
when the costs and competence of governance structures are properly aligned with
transaction attributes (Williamson 1991). Governance structures vary in terms of
their effectiveness in controlling incentives and adapting to unforeseen problems in
exchange. Market governance is ideal for nonrecurrent and nonspecific
transactions because parties can easily switch to alternative buyers and sellers if
they are unhappy with their current partners. Availability of market alternative is
what protects each partner from self-interested behavior on the part of the other.
Transactions are dependent on the legal framework only to the extent that
“classical contracts” specify the process whereby the transfer of ownership will
take place and how the contract will be enforced. For nonrecurrent and
nonspecific transactions, formal contracts fully specify the nature of an agreement
and the consequences that each party will face in case of nonperformance.
Because of the availability of market alternatives, it is not important to preserve a
relationship and thus the identity of the parties participating in exchange is not
relevant (Williamson 1979).
For highly specific and recurring transactions, “classical contracts” are less
effective in curbing opportunism. The absence of market alternatives and the
recurring nature of the transaction creates mutual dependencies between the
parties to the transaction such that opportunities for self-interested behavior
abound. Also, the recurring nature of the transaction makes it difficult to create a
200
contract that can foresee and provide solutions for all possible contingencies. At
the furthest end of the specificity and frequency spectrum, vertical integration
becomes the most effective governance structure. Integrating the production of a
good within the firm’s hierarchy effectively places both parties to the exchange
within a single entity, thereby aligning their identities and interests. Future
contingencies are easier and less costly to address as a result of internal
consultation and agreement, and without the need to rely on the legal framework.
For situations in which classical contract law cannot provide for all the
contingencies, Williamson (1979, 1991) notes the option of using a hybrid form of
governance. In particular, “neoclassical contracts” can be used to set the terms of
a transaction while also providing a mechanism for addressing potential
disturbances during exchange. Contracts in this case specify the framework for
detecting disturbances, disclosing the necessary information, and also engaging
alternative and adaptive forms of dispute resolution, such as private arbitration,
prior to litigation (Williamson 1991). The advantage of hybrid forms of
governance is that they provide more flexibility in addressing disturbances that
simple contracts cannot accommodate because they do not recognize mutual
dependencies among exchange partners nor their incentives for continuing the
exchange.
Neoclassical contracts are not complete. Williamson (1991) argues that hybrid
forms of governance become ineffective in cases of frequent and consequential
disturbances to exchange, making them less preferred than markets and hierarchy.
Parties to exchange governed by a hybrid structure are still autonomous units,
which means that disturbances with severe consequences for the relationship will
encourage self-interested behavior among parties. Moreover, hybrid forms of
201
adapting to disturbances can be costly since they are based on negotiations that
take time. In the case of frequent disturbances, hybrid governance ceases to be a
cost efficient and effective form of organizing exchange. Thus, under high levels of
uncertainty, actors will choose either markets or integration for conducting
exchange.
In many respects, hybrid governance exhibits attributes that were discussed in
reference to social networks as a form of governance structure. The attributes are
characteristic of transactions where both parties are committed to exchange due
to the absence of market alternatives and the presence of transaction-specific
investments. As a result, the parties are interested in maintaining the relationship
and prefer to avoid litigation. They employ mechanisms that allow them to settle
disputes and resolve disturbances outside the courts in order to preserve already
committed resources. Both the network and hybrid governance is less dependent
on formal contracts and litigation and more reliant on the parties’ commitment to
exchange as an incentive to curb self-interested behavior. The question then is
whether, as Williamson (1991) predicted for the hybrid forms of governance, we
can expect increased uncertainty to diminish the use of social networks in favor of
integration or markets.
We saw earlier that the reason why hybrid forms of governance lose out to
markets and integration under frequent and consequential disturbances in
exchange is because otherwise flexible adaptation mechanisms become
burdensome and less capable of curbing self-interested behavior. However, certain
characteristics of social networks make network exchange especially conducive to
overcoming the problems associated with hybrid governance structures. When
relying on network relationships to govern the transaction, actors are not
202
necessarily entering into formally structured arrangements such as franchising,
joint-venture partnerships, and other forms of contractual reciprocal trading.
Instead, the trading relationship is supported by informal agreements based on
inter-personal trust, norms of reciprocity, obligation toward network members,
and a commitment to maintaining one’s reputation in the network. These
mechanisms, as we discussed in earlier sections, serve as informal, unwritten
safeguards against self-serving actions in exchange.
To preserve network membership and a positive reputation, actors are
committed to the success of exchange and trust one another to undertake actions
to avoid recourse to the courts following a relationship breakdown. The
interpersonal trust forming the basis of network relationships encourages
information exchange between actors, which Williamson (1991) cites as a critical
component of successful adaptation to transaction disturbances. Also,
interpersonal trust leaves more room for handling disturbances without needing to
rely on consent at every stage of the adaptation process. Informal safeguards,
such as network reputation and interpersonal trust, extend beyond the life of the
specific exchange. The norms of reciprocity embedded in network relationships
enable actors to form expectations about each other’s behavior that stabilize the
expected value of exchange regardless of the levels of specificity and uncertainty
(Zhou and Poppo 2005). In other words, through informal mechanisms for dispute
resolution, relational governance provided by network relationships can
successfully coordinate that adaptation that is the most costly and difficult aspect
of hybrid governance structures (Zhou and Poppo 2005).
In a transition environment with a weak legal framework, such as Serbia and
Montenegro, it is not clear that hybrid governance structures would be passed
203
over for markets and hierarchies under extreme levels of uncertainty. Essentially,
in a weak legal environment, courts cannot be trusted to enforce agreements
among parties. The security provided by classical and neoclassical contracts, and
assumed by Williamson (1979, 1991), is not guaranteed. In such a weak legal
environment, alternative, informal forms of governance structures that provide a
framework for addressing potential disturbances and disputes are deemed more
valuable. Given the inability of the institutional framework to guarantee court
effectiveness, perhaps the best way to maximize competency and minimize the
costs of governance structures for transactions susceptible to high levels of
uncertainty is to use network relationships as the basis for exchange. Thus it is
possible to expect that with the greater risks to exchange posed by uncertainty
and asset specificity, there will be a greater degree of reliance on governance
structures based on network relationships (Zhou and Poppo 2005).
Ultimately, the impact of uncertainty on the propensity to organize exchange
in markets or social networks (or vertical integration) is an empirical question.
The evidence from Serbia and Montenegro suggests that businessmen devise
simple contracts to form the basis of exchange not just because they are legally
obligated to do so, but also because litigation serves as a final resort in the event
of unreconcilable differences in resolving an exchange problem. These simple
contracts, however, are formed after network relationships have been accessed to
identify partners, and to support the informal agreements. This complementarity
between the formal and informal mechanisms for contract enforcement was also
illustrated using survey data from SMEs in the region (Chapter 7). The evidence
so far supports the idea that social networks provide a governance structure that
can support exchange beyond classical contracts. The SME managers and owners
204
in this study suggest that uncertainty pushes them to place an even greater
emphasis on working with familiar parties and to invest in relationships that can
constrain opportunistic behavior. Greater uncertainty can make the use of
networks more appealing for all pairs of transaction characteristics. This means
that under high uncertainty, even nonrecurrent and nonspecific transactions, as
well as nonrecurrent mixed transactions, will tend to be carried out through
network relationships.
The intention here is not to introduce an alternative form of transaction
governance per se, but to show why transacting with individuals in one’s social
network is an efficient decision despite the significant investment costs in setting
up and maintaining network ties. By borrowing the transaction cost framework
from Williamson (1979), we were able to organize qualitative evidence in a way
that tells us in what circumstances it is preferred to have a network to successfully
conduct exchange and to shed light on the appeal of personal exchange to SME
managers and owners. Despite costly investments in searching for contacts in the
network, as well as building and maintaining the relationships, networks can still
result in lower transaction costs by offering tremendous cost savings in terms of
contract enforcement. Also, the increased cost of maintaining network
relationships is offset by the relative savings in opportunity costs that occur when
networks provide preferential access and treatment. Hence, in a transition
environment, costly network ties assist individual SMEs by lowering transaction
costs below those associated with impersonal market exchange and thus improving
efficiency and performance of individual businesses. In other words, network
relationships can be seen as informal institutional arrangements that enable
transactions that would otherwise not take place and thereby assist individual
205
businesses in generating economic surplus. It is easy to see how this translates
into better economic outcomes for individual businesses, but the implications for
economic welfare are a lot less clear. We turn to this discussion next.
9.5 Efficiency Implications
Table 9.2, unfortunately, does not provide any indication of the volume of
transactions that will be handled in markets and networks, respectively.
Nonetheless, we can still see that the contexts (pairs of transaction
characteristics) in which networks are preferred are more numerous than the
contexts in which actors wish to transact in impersonal markets. The SME
owners and managers were unanimous in their criticism of the outcome illustrated
in Table 9.2 by declaring repeatedly that it is hard to do business when a plethora
of mundane activities requires that they rely on personal relations and
connections. For example, a manager of a construction company had to use
personal relations to establish rapport with the manager of another business in
order to pressure him to authorize a payment for services received. Similarly, to
obtain certified documents, one must often socialize with court officials, or search
for contacts that can ensure that the paperwork will be completed within a
reasonable time period. Managers view these activities as generally unproductive
and they would prefer not to rely on such strategies to accomplish necessary
business tasks. They recognize the efficiency benefits that can be secured by using
relations and connections but, at the same time, they are aware that with better
functioning legal systems, they would have access to less costly and more
productive options to “get business done.”
206
Obtaining connections or competing against those who have access to good
connections is an especially problematic consequence of the business environment
in Serbia and Montenegro. The problem lies in the nature of connections that sets
them apart from personal relations. Connections can be useful for transacting in
the private sector, but they are even more beneficial for transactions in the public
sector. We have already noted that having connections in the courthouse or city
hall can help one obtain licenses and permits more quickly by cutting through
multiple layers of regulation, or by avoiding bureaucrats who use public offices for
extortion and financial gains. The higher the individuals are in the government or
political structure, the greater the benefits they can bestow upon relatives, friends
and acquaintances that seek a special favor. “Connections with the establishment”
are the most lucrative of all ties because of the value of the exclusive access and
treatment they can provide.
Since connections provide exclusive benefits beyond those associated with
personal relations, they can significantly tilt the playing field in favor of
individuals who have them. They allow businesses to jump through legal
loopholes and reach goals cheaply, gaining an advantage over rivals who must go
through costly procedures. Those with better connections, in terms of the value of
the opportunities and resources they can secure or the costs they can avoid, with
all else being equal, will enjoy greater economic success. Hence, managers can use
their access to powerful connections as a source of competitive advantage.
Conversly, those without access to the “right connection” are seriously
disadvantaged. When networks of connections are responsible for determining
one’s success, market principles are visibly violated. In the words of a manager of
a publishing company in Montenegro: “It’s pointless for me to be fair and
207
hardworking when someone with connections will run me over.”
2
The result is
that those that follow the rules, have a great product, and offer competitive prices
are not necessarily the businesses that will grow. Connections violate market
principles because they enable inefficient and less profitable businesses to succeed,
often at the expense of more efficient competitors.
3
The second problem associated with connections, especially their ability to
facilitate transactions between the public and private sectors, is that they
constitute valuable tools for rent-seeking and rent-avoidance. Rent-seeking refers
to profit-seeking or utility-seeking activities that entail the expense of resources to
modify the political and economic institutions in order to secure a competitive
advantage.
4
The classic example is that of expending resources to lobby for
monopoly rights that allow a business to capture rents in the market (Tullock
1980). Connections facilitate rent-seeking when, in order to secure rents, parties
exchange favors with those in the position to assign monopoly rights or other
forms of restricted market entry. The efforts involved in searching for and
investing in connections that can generate monopoly rights are efficient from the
perspective of the individual businesses. However, at least in the case of monopoly
rights, these efforts do not contribute to social welfare (Tullock 1980). Also, the
2
The use of connections to tilt the playing field in one’s favor and the interlaced issues of
equality of opportunity and fairness explain why the managers and owners interviewed in the
study were reluctant to admit to using connections. This outcome was expected and
interviewing techniques were adjusted to inquire about the prevalence of the use of networks in
general. As a result, respondents were much more comfortable in discussing the use of
connections with reference to general business practices in their particular sector.
3
This analysis is closely related to Stark’s (1997) claim that connections among businesses
may have held back the number of bankruptcies we have observed in Russia. Inefficient firms are
thus using their relationships and social capital to continue to operate at the expense of the
long-term social welfare.
4
Correspondingly, rent-avoidance refers to efforts and resources expended to avoid paying
rents.
208
efforts and resources involved in investing and maintaining connections can exceed
the value of the rents secured, in which case connections generate social costs.
An illegal form of rent-seeking that is facilitated through connections is
corruption.
5
Not all use of connections results in corrupt activities, but when
favors are exchanged instead of payments and they result in actions that
deliberately break the law in order to advance the interests of one party, then we
say that connections are facilitating corruption. In Serbia and Montenegro,
connections are instrumental in creating monopolies or limiting market
competition, obtaining public contracts, securing outsourcing deals with
state-owned enterprises, and, most commonly, obtaining building licenses on state
property, and altering urbanization plans to allow lucrative projects in otherwise
unapprovable locations. This exchange of favors instead of financial payments,
and the assumption of tacit reciprocity built into the social relations, separate
“social exchange corruption” from “market corruption” (Cartier-Bresson 1997).
Connections that facilitate social exchange corruption are simply an extension of
legal networks of relations and connections. Interviewed managers suggested that
in order to complete a specific transaction in the network, they first seek a
relation, then a connection, and ultimately, if the previous two options fail, they
utilize the “blue envelope” or “money under the table”. In the case of open market
corruption, connections can be instrumental in identifying individuals who would
be willing to accept payment in exchange for a favor. Whether it facilitates
5
Corruption refers to responses to the beneficial rules, such as when favors are exchanged to
avoid monitoring of harmful conduct, and responses to harmful rules, when favors are exchanged
to avoid bad policies and inefficient institutions (Svensson 2005).
209
exchange of favors for corrupt goals or exchange of payments in market
corruption, the use of connections can contribute to the social costs of corruption.
6
The use of connections for corrupt practices encourages corruption since some
actors find it more profitable to devote their resources to obtaining preferential
treatment as opposed to improving productivity. In this manner, corruption
through networks can become a strategy for growth. It can also necessitate the
use of connections in order to avoid corruption in exchange. For example, a
business may need connections to bypass a specific official who is seeking bribes in
return for government services, or to receive compensation after being violated by
corrupt officials. In sum, the use of connections for corruption as an illegal form of
rent-seeking creates social costs, invites additional corruption as an effective
strategy for growth, and encourages the use of connections for corruption- and
rent-avoidance.
The last among the observed negative effects is the increase in the public’s
distrust in formal institutions. The use of connections for rent-seeking and
corruption heightens the personalization of institutions since they become
increasingly identified with individuals who can determine, interpret, or bend the
rules. When connections can be used to secure preferential treatment unavailable
to others, the inviolability of the rule of law is undermined. The detrimental
factor to the public trust in formal institutions is the awareness that the same
rules do not apply to everyone. When the principles of formality, fairness and
accountability built into institutions are openly and visibly violated, public trust
6
The most often noted social costs of corruption arise when projects made possible through
the use of connections are not necessarily the projects that satisfy the interest of society as a
whole and could potentially have a negative impact on efficiency and productivity. Technology,
capital and human knowledge are not employed in the most socially desirable ways when
corruption enables the implementation of inefficient projects (Svensson 2005).
210
is undermined with significant negative implications to the prospects of
strengthening the rule of law.
In this chapter, the analysis of the outcome in which personal exchange
dominates over the use of the impersonal market has revealed that personal
relations or connections do not have the same consequences in terms of costs to
social welfare. When speaking about the business environment and the difficulty
of operating under uncertainty, managers mostly expressed disappointment that
they could not easily form enforceable contracts and have to supplant them with
the time-consuming maintenance of reciprocal relationships. Nevertheless,
managers recognized that relations help them operate more productively by
enabling transactions that otherwise would not take place in impersonal markets.
The role of connections in securing preferential treatment and access, however,
has been linked to several social costs. First, connections are responsible for
tilting the playing field and violating market principles by giving “connected”
businessmen comparative advantage over other equally or more productive
businesses. Second, connections facilitate rent-seeking and corruption with
already familiar negative externalities for social welfare. Third, connections,
especially when used in rent-seeking and corruption, contribute to the general
distrust of formal institutions. From the perspective of social welfare, connections
are far more damaging than personal relations. The incentives to individual
businessmen to rely on connections increase as more market principles are
violated; the more visible that the cases of rent-seeking and corruption become;
and the more distrustful individuals become of their formal institutions. In sum,
individual decisions to economize on transaction costs produce an outcome with
significant costs to social welfare.
211
9.6 Networking Structures: Implications for
Policy
The presence of significant social costs is an indicator that individuals have failed
to coordinate their activities to produce a socially efficient outcome. Negative
spillover effects, such as those resulting from individuals’ decisions to optimize on
transaction costs, can produce equilibria that are socially suboptimal. As Karla
Hoff explains in her review of coordination problems in development economics:
Individual actions that are privately rational when the individual takes
his environment more or less as given, need not be socially rational.
There might be another set of individual actions that would create a
different environment, within which a different set of actions would be
an equilibrium, and that environment might be a better state of
affairs. There could be multiple equilibria in institutions, with no
tendency for market forces to select the one that was best (2001, 5).
That the use of social networks can persist even though they result in a
socially inefficient outcome has been formally demonstrated by Kranton (1996)
who studied the interaction of self-enforcing exchange agreements and impersonal
market exchange. Even though her focus is on reciprocal or gift exchange, the
parallels with the exchange conducted within network relationships are easily
discernible. The model of a dynamic environment where actors choose between
reciprocal or market exchange shows that when the initial market size is small,
reciprocal exchange survives even though markets would be more efficient. This is
because the more individuals rely on social networks, the more it becomes
advantageous for them to use connections, and the more costly it becomes to rely
on anonymous market transactions. Thus, it is possible for a steady state outcome
to emerge where everyone invests in maintaining and accessing network ties even
though everyone would be better off if the use of networks were less important.
212
Insofar as institutional arrangements are developed by rational actors to
increase efficiency in exchange,
7
the informal arrangements developed to address
the absence of a well-functioning legal system become part of a suboptimal
equilibrium. The problem emerges when the institutional arrangements prevent
further improvements in market-supporting institutions (Kranton 1996; Hoff 2001;
Bowles, Durlauf, and Hoff 2006).
A few observations from Serbia and Montenegro can illustrate this point. Say
an individual actor decides to pursue a transaction in the anonymous market; his
or her refusal to take advantage of a good connection can result in losses for the
company. A manager of a printing company in Montenegro revealed that on one
occasion he had declined to accept a favor from a public official responsible for the
issuance of building licenses who would have ensured that his application received
immediate attention. Instead, the company decided to pursue the matter through
formal channels that prolonged the application process by ten years. During this
period the company incurred the costs of obtaining a temporary permit,
extending the temporary permit, and renewing the application, all the while
lacking complete ownership rights over the completed premises and the option to
use them as collateral for additional capital. Accepting the favor from the official
would have shortened the processing period to less than a year and would have
amounted to lower fees and administrative costs during the building process.
Thus, opposing the status quo results in immediate costs. Moreover, since the
costs are immediate but dispersed—not concentrated in a single group or strata of
7
A survey of New Institutional Economics would begin with a similar explanation followed by
a discussion of various disagreements over the process of institutional design and change. See,
for example, Furubotn and Richter (2000).
213
the population—it could be difficult for collective opposition to the status quo to
emerge.
Further improvements in market-supporting institutions, chiefly the rule of
law, may not emerge as a result of opposition from the winners in the status quo
(Hellman 1998). In what Hellman (1998) calls the "partial reform equilibrium,"
benefits of incomplete transition reforms, which have introduced market
institutions but have not secured the rule of law, enable those who have access to
lucrative connections to secure comparative advantage over other market
participants. The examples of nepotism or clientelistic politics, where friends and
relatives of individuals in power receive public contracts or other access to public
resources, illustrate the basic logic of the argument. Similarly, individuals who are
able to reap significant benefits because they are more competitive when they
transact though personal relationships and connections will be reluctant to accept
institutional reforms that require them to compete anonymously in the
marketplace. In other words, actors with a vested interest in the status quo are
unwilling to sacrifice benefits they have secured through connections to see an
improvement in overall economic efficiency.
Furthermore, calls from the political establishment for additional reforms,
necessary to strengthen the rule of law that supports impersonal market
transactions, are not credible. The political establishment is perceived by the
public as the clearest winner in partial reforms as identified by Hellman (1998).
Experience in the field and the interviews with SME managers and owners seem
to indicate that a large segment of the public views the ruling elites and their
powerful associates, as individuals who have used the transition process to reap
individual benefits. They have done this through asset stripping, taking
214
advantage of arbitrage opportunities, using public resources, or setting up private
businesses to profit from the transformations in the public sector. These
individuals rely on personal relationships both to protect their positions and to
reward individuals who have helped them amass wealth and power. Also,
managers of domestically owned banks are perceived to have capitalized on their
personal ties with politicians to secure monopoly rights or to have used
government resources to hedge their start-up risks. Lastly, managers of
state-owned enterprises are suspected of using personal relationships with the
political establishment to support nonviable operations in order to secure inflated
earnings at the time of sale to outside investors. Together, they are viewed as the
greatest winners through the status quo and any calls for institutional reforms
necessary to improve the rule of law, and increase the incentives for the use of
anonymous market exchange, are perceived as directly threatening to their
interests. While calls for reform originating in the political establishment are
plentiful, they are often unable to gather public support because their credibility
is suspect. Also, an environment characterized with high distrust in formal
institutions and the inviolability of the rule of law contributes to the problem of
the lack of commitment to the process of institutional reform.
One would expect the judiciary to be the source of the strongest commitment
to reforms that would advance the rule of law by curbing the practice of selective
enforcement and personalization of bureaucratic institutions, and by improving
the automation of regulatory procedures. Such reforms would reinforce their
independence from the political elites and their associates by removing channels
through which favors can be demanded from the court officials. However,
improving the effectiveness of the courts is a difficult and resource-intensive
215
process, even if we do not take into account individuals’ incentives for refusing to
implement necessary reforms. For some, the use of network relationships in
business related transactions is a stable infrastructure for securing sources of rents.
Additionally, low salaries in the judiciary are a common problem in transition
economies, and the exchange of favors provides an opportunity for increasing
one’s income. Not only can court officials exchange favors with friends, relatives,
and other members of the general public, but they can also engage in trading
favors within the judiciary in order to effectively "deliver" on promised favors.
Lawyers also capitalize on this system of informal trade in favors. The most
successful and wealthiest lawyers are former judges who trade favors with their
former colleagues—current judges—to ensure a desired outcome for their clients.
Successful appeals to the judges to trade cases is one example of this practice. In
addition, judges and court officials are more vulnerable to pressure to participate
in the exchange of favors because, unlike political elites, they can be approached
more easily by the public. A refusal to assist someone can have long-term
negative repercussions for personal relationships within their local communities.
Examples of personal incentives to reject calls for reform, or to prevent and avoid
their implementation in practice, can thus be found even in the judiciary.
Lastly, calls for reforms in market-supporting institutions may clash with the
cultural elements that have accepted personalized exchange through socialization.
Individuals become accustomed to approaching organizations, or obtaining goods
and services, through friends, acquaintances and relatives, because they have
observed this as a socially acceptable and very successful strategy. They resist
adopting any new rules that clash with what they have accepted as a form of
legitimate and conventional behavior. In other words, the informal arrangements
216
that have emerged to address the weaknesses in the legal framework and are
prohibitive of impersonal market exchange, are path-dependent and difficult to
alter.
These are only several examples that illustrate why personalized exchange
could ultimately impede the further development of formal institutions or their
enforcement mechanisms that are necessary to reduce network advantages and
lower the use of connections. The examples also warn against the danger of
partially introducing market institutions without the effective rule of law since
these do not make impersonal markets efficient governance structures for most
types of transactions. A more detailed political analysis of the winners and losers
from the status quo is needed to get a fuller understanding of the reasons behind
the lack of collective action necessary to eliminate the sources of the
suboptimality. To fully establish that the dominance of impersonal exchange is
suboptimal from the perspective of social welfare, individual benefits from
transaction cost optimizing in network exchange would have to be compared with
the spillover costs identified with the dominant use of personal relationships.
Also, to fully establish that this “partial reform equilibrium” is a stable and
suboptimal equilibrium (i.e. equilibrium trap) and not just a snapshot of an
adjustment process to an equilibrium, an empirical analysis, which is still in its
infancy, is necessary (Hoff 2001).
These are only some of the research questions from whose elucidation social
capital and social network studies can benefit in the future but which are beyond
the scope of the present study. The objective here was to show that markets alone
may not be capable of altering incentives in a way that will introduce anonymous
market exchange as the norm. Personalized exchange and the informal
217
institutions that provide its critical infrastructure emerge to compensate for the
weaknesses in the formal insinuations. Together, however, they can persist in an
equilibrium with possible social costs. The analysis here suggests that the
departures from optima can be lasting and costly. As a result, personalized
exchange and the social costs it produces deserve the attention of policymakers.
The objective should be to design institutions that can alter the actors’ incentives
in a way that would enable the coordination of individual activities to a more
optimal equilibrium. Development theorists have dealt with this issue for some
time and the skeptics among them believe that neither markets nor governments
are capable of addressing coordination failures effectively (Hoff 2001).
Kranton’s (1996) model of the choice between reciprocal exchange and
impersonal exchange suggests that only slight modifications in incentives would
eventually lead actors to coordinate to an outcome where market exchange
dominates. The more that individuals reject personalized exchange and approach
unknown actors in the market, the more transactions become costly for those who
still choose to invest in relationship maintenance. Thus policies that would make
it easier for individuals to access impersonal markets and to successfully conduct
their transactions in that setting are needed. While "big-push" policies may be
distrusted by some development economists, if efforts are coordinated at several
levels, results could be possible.
A speculative list of policies that might comprise these efforts can include
prescriptions aimed at training court officials and improving court capacity in
order to lower the costs of accessing the courts. A closely related reform is
increasing salaries awarded to judges and court officials, which can directly lower
the incentive for exchanging favors, accepting bribes and participating in corrupt
218
activities. Training programs and higher salaries in the offices of local and
national governments can also result in lower incidence of the selective application
of the law and the use of connections for rent-seeking. Improving bureaucratic
professionalism by enforcing educational and work qualifications for all applicants,
as well as prescribing performance standards for public service, would help lower
selective applications of the law by bureaucratic officials. Lastly, streamlining
procedures and making them available to the public outside the government
offices, over the internet, or through automated telephone systems can increase
the automation of procedures and can curb the influence of public officials in
interpreting and implementing regulation.
Another important component of efforts to curb the exchange of favors and
other costly uses of personalized exchange is to offer an inexpensive and clear
mechanism for enforcing accountability among bureaucratic officials. With such a
mechanism in place, individuals who experience selective enforcement of rules and
abuse of connections can: 1) draw attention to the violations of market principles
and abuse of official positions through rent-seeking and corruption; and, 2) utilize
the mechanism to demand corrective action by the government. The goal is to
organize efforts that will force individual actors to learn through experience and
socialization that connections cannot provide any exclusive benefits, or conversely,
that the costs of looking for friends, relatives and acquaintances cannot be offset
by advantages obtainable only through network exchange. The list presented here
is not exhaustive but is intended merely to offer suggestions on what specific
current or future policies are especially likely to alter incentives in a way that
would assist actors in coordinating to a more optimal equilibrium from the
perspective of the society.
219
9.7 Conclusion
A complete shift from personal to impersonal exchange is not desirable. Personal
exchange can successfully complement impersonal exchange by providing an
alternative form of contract enforcement that can lower enforcement costs even
when disputes can be resolved in inexpensive and efficient courts. For some
transactions, using personal relations to resolve disputes and avoid the use of
courts altogether will always be a more efficient solution than purely impersonal
market governance. This chapter has argued that such use of personal relations in
exchange is not associated with significant social costs. However, the use of
connections in conducting network exchange can have negative implications for
social welfare. Some of the social costs identified in this study are the violation of
market principles, prevalence of rent-seeking and corruption, and distrust in
institutions. Hence, moving away from the equilibrium where personal exchange
dominates impersonal exchange is highly desirable as it reduces the benefits of
connections and lowers the social costs resulting from their use. Unfortunately,
without an improvement in the courts’ capacity, training and judicial salaries, as
well as the professionalism of public servants, such a shift is unlikely to occur.
Businesses have no incentives to abandon the use of connections, which from their
individual perspectives result in transaction cost savings, unless the role of
connections in securing privileged access and information is deliberately restricted.
This study recognizes that further research is needed to empirically asses the
presence and implications of suboptimal outcomes where personalized exchange
dominates the use of impersonal markets. However, the use of qualitative
evidence from Serbia and Montenegro helped us reach some conclusions about the
efficiency of the equilibrium where networks are the preferred choice of
220
transaction governance for relatively more types of transactions. The chapter has
shown that a move toward impersonal exchange centered upon formal institutions
and the rule of law is difficult because the current suboptimal equilibrium blocks
reform in formal institutions designed to lower the incidence of personal exchange
and benefits of connections. Further research is needed to investigate potential
thresholds, in terms of the degree to which personal exchange is used, and below
which personal exchange blocks the improvement of the conditions that would
increase the incentives for the use of impersonal markets in exchange.
221
Chapter 10
Economic Sociology and Economic
Transitions
Grounded in transition scholarship, the primary concern of the study presented
herein has been to answer whether the social embeddedness of economic actors in
a transition economy is a source of advantages or impediments for entrepreneurial
activity and private sector growth. Transition scholars have documented the
omnipresent nonmarket exchange, and the use of reciprocal ties, social networks,
exchange of favors, and corruption, as remnants of the socialist legacies that have
become path-dependent cultural aspects in the post-communist world. However,
they have been less successful in integrating their concerns over these elements
into broader questions dealing with the process of institutional change during
transition. As a result, they have been unable to determine the impact of social
embeddedness on the success of economic reforms or individual-level economic
outcomes.
Importing concepts from social capital studies did not bear fruitful results as
the this field of study itself lacks compelling explanations of how elements of
222
social capital translate into positive economic results. Part of the problem in
providing these explanations lies in the lack of concepts that can be
operationalized, measured, and employed in the analysis of individual-level
outcomes. Closely related social network studies, developed exclusively outside
the context of the post-communist transition, provide concepts and methods with
which we can finally make sense of the conflicting evidence regarding the impact
of social networks on economic outcomes in transition. This study bridges
transition scholarship with social network studies and demonstrates that through
a network approach, we can generate measures of social capital at the individual
level and examine them in relation to firm-level economic performance. In
adopting this approach the study offers a rare example of quantitative empirical
analysis of social capital in transition.
10.1 Implications
Transition scholarship has shown that social networks formed an integral part of
the socialist systems both as a way to compensate for its weaknesses and to
exploit the system for private gains. The same is true in transition. The
unprecedented change in political and economic institutions opened numerous
opportunities to gain private returns using informal exchange relations. Fearing
that social networks might provide the means by which powerful interests could
“capture” the state, some scholars wanted to understand how social capital might
benefit or inhibit the formation of strong participatory democratic institutions.
Specifically with respect to the success of economic reforms, there was concern
that personalized exchange and corruption could destroy entrepreneurial
incentives and block further entry of SMEs. Yet, there was plenty of evidence to
223
suggest that personalized exchange provided insurance against uncertainty in the
process of structural adjustment and institutional change. Using the alternative
informal set of institutions to guide transactions seemed to compensate for the
voids in the formal institutional infrastructure. However, the role of personalized
exchange in the process of transition and its impact on private sector growth
remained unclear.
A potential toolset for studying the informal exchange of the post-communist
world is offered by social capital literature. Scholars in this field are concerned
with describing the various benefits of norms, trust, and networks to political
participation, strength of democratic institutions, and effects on economic
well-being. Unfortunately, the field relies on diverse approaches with numerous
definitions and concepts that are not consistently operationalized. Attitudinal
elements such as norms and trust are particularly vulnerable when it comes to
contributing precise definitions and demonstrating the relationship between the
elements of social capital and political or economic outcomes. Turning to social
capital for tools with which to examine the context of transition did not result in
explanations of how social capital causes positive outcomes and presented only a
few micro-level empirical studies for reference.
Adopting a structural approach to social capital instead, this study borrowed
the concept of the social network—a structure formed by a set of relationships
that bring actors and their resources together. In this approach, social capital is
envisioned as an enabling structure as it centers on the value of resources to which
individuals gain access as a result of their network membership. With this
definition of social capital, it is possible to clearly operationalize its key elements
and borrow from sociology a set of methods for their measurement. Together,
224
these tools enable us to examine whether differences in micro-level outcomes can
be explained through differences in the nature and characteristics of informal
exchange relationships in a transition setting.
The study was mostly concerned with the performance of SMEs because they
are critical to the success of transition and represent key players in the
endogenous strategy for growth. SMEs play a role in generating employment
opportunities, innovating, and alleviating some of the negative shocks associated
with the transition process. This study has shown that the slow transition
performers of South-Eastern Europe also have poorly developed SME sectors.
The causality in this case can go either way but it raises the concern that SMEs
face obstacles to their growth, which therefore affects future growth possibilities
for the economy as a whole. Thus, it is necessary to examine whether concerns
over impersonal exchange relations in transition economies represent such
obstacles and in what way they may be affecting the performance of SMEs. A
considerable amount of research has been devoted to addressing the relationship
between inter-organizational networks and the success of individual businesses,
but the link between managerial social ties and business performance has till now
not been addressed sufficiently (Peng and Luo 2000).
Interview data from Serbia and Montenegro shows that SME managers and
owners utilize networks for two specific purposes. First, networks serve as an
informal mechanism for contract enforcement. Second, they provide exclusive
access to valuable information and capital. Together, these two roles also
summarize the findings from other transition economies such as Russia, Bulgaria,
and China (Ledeneva 2001; Ledeneva 1998; Peng and Luo 2000; Lonkila 1997;
Manolova and Yan 2002; Radaev 2004; Fan 2002). In general, transition scholars
225
emphasize the role of networks in compensating for weaknesses in the legal
framework while social network studies are mostly concerned with the role of
networks in providing access. Evidence from Serbia and Montenegro presented
above also demonstrates that using social networks is not without cost. Receiving
only cursory attention in previous social capital studies, the costs in searching for
and maintaining networks, returning favors, or staying committed to specific
relationships become evident in interviews with SME managers and owners.
To understand how costly social networks affect business performance, this
study employed an original quantitative dataset from Serbia and Montenegro. The
relationship between formal and informal mechanisms for contract enforcement
and its impact on the level of profit reinvestment was examined first. Following
existing studies that demonstrate the critical connection between the use of courts
and profit reinvestment, we tried to determine whether similar evidence exists for
social networks as an informal mechanism for securing the inviolability of one’s
property rights. Using a model in which property rights measures were broken
into elements of formal and informal means for contract enforcement, the study
found that those who relied more on social networks for dispute resolution and
those who invested more in network maintenance reinvested more of their profits.
The surprising finding was the high level of investment in network ties—on
average, managers spent close to 15 hours per week maintaining current business
relationships. Greater investment in social capital resulted in greater reinvestment
of profit, even after we controlled for the confidence in courts and the threat of
expropriation. This suggests that social networks have a direct and independent
effect on the level of profit reinvestment: the security of property rights is
strengthened by the use of social networks in dispute resolution and additional
226
investments in network ties that serve as an insurance policy against contract
defection. Hence, differences in businesses’ reinvestments can be explained by the
variation in the degree to which networks are used in dispute resolution and the
degree to which SME managers and owners invest time in their business ties.
The finding that confidence in courts and the absence of the threat of
expropriation encourages reinvestment is consistent with results obtained by
Johnson, McMillan, and Woodruff (2002b) for Poland, Slovakia, Romania, Russia
and Ukraine. The use of social networks for facilitating exchange has also been
documented in other transition countries, but their impact on the security of
property rights and profit reinvestment has not been examined prior to this study.
The relationships between social networks as an informal mechanism for contract
enforcement and profit reinvestment was shown here to be robust but more
studies are needed to confirm these results in countries other than Serbia and
Montenegro. In the future studies, it is especially important to examine the
properties of this relationship further and to uncover any possible thresholds
beyond which investments in social network ties experience diminishing returns.
Since evidence emerged for the positive impact of social capital on
reinvestment and presumably the growth of businesses, we needed to examine
what about networks specifically matters for growth. Using the quantitative
dataset from Serbia and Montenegro, this study found that the content of
exchange facilitated by the network and its structure are directly and positively
related to the reported growth in SME sales. To fully understand the impact of
networks, we have to use the content of exchange to distinguish between the two
network types, namely the use of networks for contract enforcement (trust
networks) and the use of networks for access to resources (resource access
227
networks). Both network types positively impact sales performance, but the
impact of structure is conditional on the network role. Highly dense (high number
of alter–alter ties) and homogenous (low diversity in ego–alter ties) trust networks
encourage sales growth. However, it is the low density and heterogenous resource
access networks that directly and positively impact sales growth. Hence, SME
managers and owners should seek to attract into their trust networks individuals
from similar walks of social life and who know one another. Conversely, they
should seek to encourage individuals from diverse groups, who will most likely not
know each other, to join their resource access networks.
Empirical treatments of social networks, found in management studies, ignore
the role of networks in assisting with contract enforcement. They also discount
the impact of the wider institutional environment, or implicitly assume that it is
the same for everyone. The present study demonstrated that this role must be
included when examining the impact of networks on business performance since it
affects the security of property rights. When networks were included in the model
for their role in contract enforcement, the positive impact of network roles and
network structure on sales growth held even when we controlled for confidence in
courts.
Surprisingly, no evidence emerged from this study to confirm that weak ties in
the network promote business growth. One possible explanation is that the
heterogeneity variable measuring the diversity of the ego–alter ties was already
capturing some of this effect. However, the role of network diversity in
encouraging business performance was contingent on the network type.
Additional studies are needed to examine whether the number of weak ties is
helpful for business success. Existing empirical studies show evidence in support
228
of both weak and strong ties (see Section 8.2.3). Future socio-centric studies are
needed to examine the degree to which memberships in the two network types
differ. This would allow the examination of membership both within and across
network types.
Given the cost of using network ties and the evidence demonstrating the
positive role of social networks, it was necessary to examine whether network ties
represent an efficient investment of limited resources. Using the transaction cost
framework, we have shown that an individual’s decision to invest in costly
networks as a business strategy is a result of the motive to economize on
transaction costs. However, individual decisions typically coordinate to a less than
optimal outcome. The predominant use of social networks is associated with
social costs such as the use of connections for rent-seeking and corruption,
distrust in state institutions and distortions in the market structure. The level of
social costs depends on the mix of formal and informal institutions in the local
equilibrium. Suboptimal outcomes characterized by the presence of significant
social costs can persist. This occurs when the characteristics of personalized
exchange supported by social networks prevent changes in formal institutions that
may lower incentives for the use of personal exchange in the first place. However,
it may be possible to design policies that can simultaneously attempt to alter
actors’ incentives in order to encourage impersonal market exchange as the
preferred governance structure for conducting business transactions.
These results shed light on what was described as the “ambiguous” role of
social networks in post-communist contexts (Ledeneva 2004). In her study of
informal exchange in Russia, Ledeneva (2004) claims that networks can work both
in favor and against market institutions and that their benefits cannot be
229
disentangled from their consequences. Through the use of an empirical approach
to social capital at the micro-level, this study clarifies the impact of networks on
economic performance. At the individual level, the use of personalized exchange
through social networks is the result of the decision to economize on transaction
costs. However, the use of social networks is associated with spillovers because the
individually optimizing decisions do not always coordinate to a socially optimal
outcome. These social costs can represent obstacles to further economic reforms
and the society can thus find itself in a partial-reform equilibrium (Hellman 1998).
Unqualified support for increasing the volume of social capital or greater
networking activity is not necessarily going to have a positive impact on economic
outcomes. Social capital studies often conclude that greater participation in clubs
and bowling alleys, or wider webs of contacts, will help one maximize their income
or will contribute positively to broader development prospects (Putnam 2000;
Woolcock and Narayan 2000). Sometimes, however, greater individual benefits
can come at the price of greater social cost. These considerations take us beyond
policy prescriptions and lead into political analysis of the society’s relative
preference between individual growth and social consequences. They also raise the
question whether alternative investments of resources generate better results. For
an answer, we need a better understanding of the conditions that discourage
adjustments to more optimal equilibria and the relationship between formal and
informal institutions, resulting in the predominant use of personalized exchange.
The degree to which personalized exchange dominates impersonal market
exchange shown here is specific to the case of Serbia and Montenegro and its
experience with transition to date. We have identified several contributing
elements that are specific to the business environment in the two newly
230
independent states. They included a weak legal framework, selective application
of the law, unspecified business regulation, expensive and inefficient courts, and
the low level of bureaucratic professionalism. Thus, social costs identified in
relation to the outcome discussed in the case under study are specific to Serbia
and Montenegro.
In a well-known study, Macaulay (1963) argued that relational contracts and
personal exchange relations play a large role in avoiding legal disputes and
contract enforcement costs among business partners in the U.S. Today, the
consensus seems to suggest that U.S. society is highly litigious and that the use of
courts for formal contract enforcement is predominant. The analysis in the
Chapter 9 suggests that the explanation for the possible change in trends must
rely on the analysis of the degree to which personalized exchange has over time
given way to impersonal market exchange and the potential reasons behind this
shift. Although, networking is not absent in the case of U.S. the relative
importance of the two network roles can vary depending on the broader
institutional infrastructure in the economy. In particular, networking in the U.S.
is synonymous with using relationships to hear about employment opportunities,
gain new clients, secure real estate deals, receive recommendations for a doctor,
babysitter, or a school. We hear of relatively few cases where network
relationships were needed to provide assurance that a contractual relationships
would not break down. Also, in conducting day-to-day business transactions such
as obtaining import–export licenses or notarizing documents, one does not
necessarily have to seek help from friends, relatives, or acquaintances. In that
sense, the U.S. does not rely on personalized exchange to the extent that some
transition economies might. However, the relative degree to which personalized
231
exchange is used in a particular country or region requires additional research. We
must develop techniques for evaluating the degree to which transactions can be
conducted without the need to rely on network contacts, but before the costs of
transacting increase relative to other market participants. Thus, the case of Serbia
and Montenegro has served to set up a broader discussion pointing to the need for
further examination of the relative use of personal and impersonal exchange.
In other transition cases where personalized exchange seems to be dominating
the use of impersonal exchange, such as China and Russia, similar social costs to
those discussed in relation to Serbia and Montenegro have been identified
(Ledeneva 2003; Radaev 2004; Fan 2002; Standifird and Scott 2000). It appears
that the discussion of efficiency relating to the use of personal exchange in Serbia
and Montenegro has broader implications. The cause for concern arises from the
finding that the dominance of personal exchange can block improvements in
formal institutions that are capable of lowering the use of the social networks
responsible for the related social costs. Some studies suggest that China is
developing its own brand of capitalism based on personalized exchange and thus
no grounds exist for the expectation that societies such as this one will ever
exhibit the characteristics of Western developed markets (Boisot and Child 1996;
Lovett, Simmons, and Kali 1999). The case of Serbia and Montenegro warn that
such arguments may be premature. The presence of social costs suggests that
individual decisions to optimize on transaction costs can create a less than
optimal outcome in a social sense. It is necessary to further examine the degree of
suboptimality that arises as a result of personalized exchange and how it can
persist over time. Accepting high levels of personalized exchange without
establishing its consequences can amount to choosing less efficient outcomes. The
232
case of Serbia and Montenegro was thus not intended to provide generally
applicable conclusions. The discussion in Chapter 9 was presented merely to
suggest important questions that apply beyond this single case. To answer such
questions, future comparative studies are necessary.
10.2 Notes on Methodology
Relational and ego-centric approaches to social network analysis were utilized to
examine the role of impersonal exchange in transition economies. These tools
enable the collection of individual-level measures of social capital that were
employed as explanatory variables for the differences in business performance.
Micro-level studies of the impact of social capital on economic outcomes are rare,
and so are quantitative empirical studies of social capital in transition. As a
result, borrowing methods from sociology and employing them in the analysis of
business success in transition produced a uniquely valuable quantitative dataset
with powerful methodological lessons for both economic sociology and transition
scholarship.
One of the superior qualities of the relational approach to social network
analysis is that it focuses on the characteristics of network ties to describe
network structure, and does not consider the relative positions of individuals
within networks. While the latter strategy is important for other approaches to
social network analysis (e.g. the positional approaches), it is inefficient and
cumbersome for a study that wants to employ individual level measures of
network characteristics as independent variables in an econometric analysis. In
this study, specific techniques were selected for data collection in accordance with
the relational approach to social network analysis. The study used the
233
name-generating and name-interpreting questions as part of a mailed
questionnaire collecting responses from SME managers and owners. These
questions asked the individual respondents to describe their relationships with
other members of their most immediate networks, to provide select member
characteristics, and to comment on the nature of the relationships possibly
existing between network members. The resulting measures do not focus on ties
alone but describe the actors as well as the resources to which members of the
network have access.
The single most critical aspect and highly original component of the survey is
the use of a technique inspired by Podolny and Baron (1997) to look at the
content of network exchange. Name-generator questions were used to distinguish
between two types of networks—trust and resource access—while
name-interpreting questions were included for each network type. In this manner,
the study was able to incorporate variables for network characteristics that
included the content of exchange facilitated through network ties. This is the first
social network study to employ this technique in examining the impact of
networks on economic outcomes. As a result, it has not only contributed to
transition scholarship by employing novel techniques for measuring and studying
social networks, but also to advancing the measurement techniques of social
network analysis. Thus, we have demonstrated that it is possible to capture the
content of information transmitted via network ties and that through the design
of name-generating questions used to identify network memberships, numerous
network types can be studied.
The techniques employed here to collect the measures of social network
characteristics and the degree to which individuals rely on them are fully
234
compatible with the features of the survey incorporated to collect information
about business performance, business and management characteristics, and other
control variables necessary for an econometric study of firm-level outcomes. Some
of these additional survey questions were borrowed from related studies in order
to ensure validity and reliability of results, while others were constructed anew.
Overall, this survey, designed for the purpose of generating a quantitative dataset,
illustrates that one can integrate sociology’s methods for capturing the
characteristics of social networks into surveys examining micro-level economic
performance. In the end, social network measures can yield interesting and robust
results as micro-level, social capital variables in an econometric analysis.
Following Batjargal’s (2003) successful experience with the position-generator
questions, this study attempted to capture resource diversity in the network by
collecting information about members’ occupations. The position-generator
method requires respondents to list how many individuals they know in each of
the offered categories of occupations. To avoid adding such a complex question to
an already long survey, the name-interpreting questions were adjusted to collect
information about the members’ occupations (see Section 8.3.2.3).
1
The goal was
to include some of the beneficial attributes of the position-generator technique
without increasing the likelihood that the respondent would not complete the
survey due to its length. Unfortunately, this method did not generate the
expected results. The variation in written-in occupations was too large to
comfortably fit into the predetermined rankings of occupations from which we
calculated measures of resource volume. While the rankings of occupations could
1
Several occupation categories with appropriate codes were offered in a table, but
respondents could also write in their own responses. Occupations were coded and ranked
according to predetermined criteria largely following Batjargal’s (2003) analysis.
235
have been adjusted after the completion of data collection, they would have been
highly arbitrary without justification for valuing some occupations more than
others in terms of resources provided to network members.
To overcome the shortfall of the technique where name-interpreting questions
are blended with questions about members’ occupations, the analysis may be
restricted to a specific sector. In this manner, the rankings of occupations can be
customized to the needs of the businesses working in that particular sector and
the same rankings can then be applied to occupations listed by all respondents.
Alternatively, rankings of occupations can be eliminated altogether if other
measures are used to capture resource volume or resource diversity.
Restricting the sample to one or two highly divergent sectors can have other
benefits as well. For example, this would provide additional controls to ensure
that the results are independent of disturbances specific to one’s business activity.
The present study employed business activities as a control variable in the
regression analysis. However, because all sectors were targeted in the survey, some
activities were poorly represented in the final sample. As a result, they did not
provide sufficient variation in terms of other necessary variables and had to be
dropped to successfully process logit and probit regressions. Sampling only a
limited number of activities would avoid this problem. A particularly interesting
type of analysis can be performed by choosing two smaller samples: one restricted
to a service-oriented business activity and one oriented heavily towards
manufacturing. In this manner one could examine how network characteristics
and the relative importance of network roles change from one sector to another.
The present experience with data collection emphasizes the importance of
locating a detailed database from which to select the sample for study. With the
236
benefit of hindsight, it appears that the best strategy is to locate existing studies
focused on the same target population and to utilize their field experience to
identify and locate the “population” database. Not only is it difficult to identify
organizations or institutes that manage a desired database in the chosen case or
cases, but it can be especially difficult to secure the right to access the specific
database. Individuals in Serbia and Montenegro were mostly distrustful of the
researcher’s intentions with the database or were completely disinterested in
assisting the researcher because such activities did not fall under the purview of
their employment.
Testing the database during the pilot stage is crucial because it allows the
researcher to verify the quality of the contact information. Understandably, the
database should be carefully examined prior to the pilot stage to identify any
glaring omissions in the information needed for selecting a sampling frame or a
target sample. The same lessons hold true for those conducting in-person
interviews with a random selection of respondents. Difficulties with the quality of
the chosen database and the complexity of the questionnaire suggest that the
above study could have benefited from employing interviewers in the data
collection process. Resource permitting, this type of data collection is more
conducive for surveys employing the ego-centric approach and including questions
about network members and relationships among them.
10.3 Concluding Remarks
Central to the growth of modern economies, explains Douglass North (2005), has
been a shift from personal exchange, based on the principle of reciprocity rooted
in strong personal ties and collectivist beliefs, to impersonal exchange supported
237
by formal rules and effective forms of contract enforcement. A complex structure
of formal institutions and organizations supportive of impersonal exchange, as
identified in theory and from the experience of developed economies, was
prescribed for the economies of Eastern Europe embarking upon post-communist
transition beginning in the late 1980s. When these countries adopted the
prescribed changes, they achieved diverse results. Some, such as Serbia and
Montenegro, have settled in states where personal exchange still plays an
important role in political and economic markets, and further moves toward
impersonal exchange are not bearing the expected economic gains. Responsible
for this outcome, in part, is our lack of a complete understanding of the proper
combination of formal institutions and historically conditioned informal
constraints, such as beliefs, norms and cultural values that can generate positive
economic outcomes (North 2005). The preeminent task for these economies is to
determine the implications of the relationship between personal and impersonal
exchange for further economic growth and how the existing culturally and
historically influenced institutions might mediate any attempts to move toward
impersonal exchange. Post-communist transition in South-Eastern Europe can
thus be viewed as an attempt to move away from a socially and culturally
influenced equilibrium where personal exchange centered upon networks of
relations and connections is the preferred form of transaction governance.
To understand the relationship between formal and informal institutions, we
need tools with which to study the informal institutions that include beliefs,
norms, and social networks. Some of these tools have been developed by social
network studies and have raised significant interest in interdisciplinary research
for sociologists, economists, and management scholars. Transition scholarship has
238
probably more to gain from economic sociology since the latter can be used to
uncover the dangers of partial reforms and the problems that could arise when
informal and formal institutions produce equilibria associated with significant
social costs. In a reciprocal fashion, economic sociology, benefits by expanding its
testing ground into the transition context where its traditional assumptions may
not hold, and where socially embedded strategies for overcoming institutional
voids flourish. Together, it is possible for the two fields to generate powerful
insights into the process of institutional change.
239
Bibliography
Acemoglu, Daron, and Simon Johnson. 2005. Unbundling Institutions. The
Journal of Political Economy 113 (5): 949–995.
Acs, Zoltan J. 2003. The Historical Role of the SME Sector in Developing and
Developed Capitalist States. In Small and Medium Enterprises in
Transitional Economies, edited by Robert McIntyre and Bruno Dallago,
18–36. New York: Palgrave Macmillan, in association with the United
Nations University/WIDER.
Acs, Zoltan J., Bo Carlsson, and Charlie Karlsson, eds. 1999. Entrepreneurship,
Small and Medium-Sized Enterprises and the Macroeconomy. Cambridge,
U.K.: Cambridge University Press.
Adam, Frane, and Rončević. 2003. Social Capital:Recent Debates and Research
Trends. Social Science Information 42 (2): 155–183.
Aldrich, Howard, Ben Rosen, and William Woodward. 1987. The Impact of
Social Networks on Business Foundings and Profit: A Longitudinal Study. In
Frontiers of Entrepreneurship Research, edited by N. Churchill, J. Hornaday,
O.J. Krasner, and K. Vesper, 154–168. Babson College Center for
Entrepreneurial Studies.
Almeida, Paul. 1999. Semiconductor Startups and the Exploration of New
Technological Territory. In Are Small Firms Important? Their Role and
Impact, edited by Zoltan J. Acs, 39–51. Boston: Kluwer Academic Publishers.
Arrow, Kenneth. 2000. Economic Transitions. Journal of Institutional and
Theoretical Economics 156 (1): 9–18.
Audretsch, David B. 2002a. The Dynamic Role of Small Firms: Evidence from
the U.S. Small Business Economics 18 (1): 13–40.
. 2002b. Small Firms and Entrepreneurship: The Western Experience. In
Small Firms and Entrepreneurship in Central and Eastern Europe: A
Socio-Economic Perspective, edited by Oliver Pfirrmann and Gunter H.
Walter, 15–46. Heidelberg: Physica-Verlag.
Badescu, Gabriel, and Eric M. Uslaner. 2003. Social Capital and the Transition
to Democracy. London: Routlege.
Bartlett, Will, Milford Bateman, and Maja Vehovec, eds. 2002. Small Enterprise
Development in South-East Europe: Policies for Sustainable Growth. Boston:
Kluwer Academic Publishers.
240
Bartlett, Will, and Paul Hoggett. 1998. Small Firms in Southeast Europe: A
Comparative Analysis of Experience in Bulgaria, Hungary and Slovenia. In
How to Support SMEs in Yugoslavia, edited by Jelica Minić and Aleksandar
Denda, 163–186. Belgrade: European Movement in Serbia, Institute of
Economic Sciences, Ekonomska Politika and Konrad Adenauer Stiftung.
Batjargal, Bat. 2003. Social Capital and Entrepreneurial Performance in Russia:
A Longitudinal Study. Organization Studies 24 (4): 535–56.
Birley, Sue. 1985. The Role of Networks in Entrepreneurial Process. Journal of
Business Venturing 1 (1): 107–117.
Boisot, Max, and John Child. 1996. From Fiefs to Clans and Network
Capitalism: Explaining China’s emerging economic order. Administrative
Science Quarterly 41 (4): 600–29.
Bolčić, Silvano. 2002. Rast Privatnog Sektora i Preduzetništva u Srbiji Tokom
1990-tih. In Srbija Krajem Milenijuma: Razaranje Društva, Promene i
Svakodnevni Život, edited by Silvano Bolčić and Anđelka Milić, 107–122.
Beograd: Institut za Sociološka Istraživanja Filozofskog Fakulteta u
Beogradu.
Bordieu, Pierre. 1986 [1983]. The Forms of Capital. In Handbook of Theory and
Research for the Sociology of Education, edited by John G. Richardson,
241–258. New York: Greenwood Press.
Bordieu, Pierre, and Loic Wacquant. 1992. Invitation to Reflexive Sociology.
Chicago: University of Chicago Press.
Božović, Iva. 2006a, May. Economic Constraints in Serbia and Montenegro:
Perceptions of Small and Medium Size Enterprises. M.A. Economics,
University of Southern California, Los Angeles.
. 2006b. The Prospects of Socia Capital: Networks in Bosnia and
Herzegovina. In Assessing Social Capital: Concept, Policy and Practice,
edited by Rosalind Edwards, Jane Franklin, and Janet Holland, 111–128.
Newcastle, UK: Cambridge Scholars Press.
Bowles, Samuel, Steven N. Durlauf, and Karla Hoff. 2006. Poverty Traps.
Princeton, NJ: Princeton University Press.
Brunetti, Ayomo, Gregory Kisunko, and Beatrice Weder. 1998. Credibility of
Rules and Economic Growth: Evidence from a Worldwide Survey of the
Private Sector. The World Bank Economic Review 12 (3): 353–84.
Bukvic, V., W. Bartlett, A. Rus, D. Sehic, and Stojanova V. 2001. Barriers to
SME Development in Bosnia, Macedonia and Slovenia. Phare-ACE Project
P997 8089-R Final Report. Ljubljana, Slovenia: GEA College.
241
Burt, Ronald. 1992. Structural Holes. Cambridge, MA: Harvard University
Press.
. 2000. The Network Structure of Social Capital. In Research in
Organizational Behavior (v.22), edited by Robert I. Sutton and Barry M.
Staw. Greenwich, CT: JAI Press.
Caldwell, Melissa L. 2004. Not By Bread Alone: Social Support in the New
Russia. Berkeley and Los Angeles, CA: University of California Press.
Carlsson, Bo. 1999. Small Business, Entrepreneurship, and Industrial Dynamics.
In Are Small Firms Important? Their Role and Impact, edited by Zoltan J.
Acs, 99–111. Boston: Kluwer Academic Publishers.
Cartier-Bresson, Jean. 1997. Corruption Networks, Transaction Security and
Illegal Social Exchange. Political Studies 45 (3): 463–476.
Casson, Mark. 1999. Entrepreneurship and the Theory of the Firm. In
Entrepreneurship, Small and Medium-Sized Enterprises and the
Macroeconomy, edited by Zoltan J. Acs, Bo Carlsson, and Charlie Karlsson,
45–79. Cambridge and New York: Cambridge University Press.
Casson, Mark, and Howard Cox. 1997. An Economic Model of Inter-Firm
Networks. In The Formation of Inter-Organizational Networks, edited by
Mark Ebers, 174–196. New York: Oxford University Press.
CEED. 2002. Barriers to Doing Business in Montenegro. Business Idea.
Podgorica, Serbia and Montenegro: Center for Entrepreneurship and
Economic Development (CEED).
Cengic [Čengić], Drago. 2003. Small Entrepreneurs in Medjimurje: Between
Tradition and Challenges of New Forms of Cooperation. In Regional
Economic Growth, SMEs and the Wider Europe, edited by Bernard Fingleton,
Ayda Eraydin, and Raffaele Oaci, 246–270. Aldershot, England: Ashgate.
Coleman, James S. 1988. Social Capital in the Creation of Human Capital.
American Journal of Sociology 94 (Supplement): S95–S120.
Cook, K. S., and J. M. Whitmeyer. 1992. Two Approaches to Social Structure:
Exchange Theory and Network Analysis. Annual Review of Sociology
18:109–27.
Cook, Karen S., ed. 2001. Trust in Society. New York: Russell Sage Foundation.
Cook, Karen S., Erik R.W. Rice, and Alexandra Gerbasi. 2004. The Emergence
of Trust Networks under Uncertainty: The Case of Transitional Economies -
Insights from Social Psychological Research. In Creating Social Trust in
Post-Socialist Transition, edited by Janos Kornai, Bo Rothstein, and Susan
Rose-Ackerman, 193–212. New York: Palgrave Macmillan.
242
Cromie, Stan, and Sue Birley. 1992. Networking by Female Busines Owners in
Northern Ireland. Journal of Business Venturing 7 (3): 237–251.
Cull, Robert, and Lixin Colin Xu. 2005. Institutions, Ownership, and Finance:
The Determinants of Profit Reinvestment Among Chinese Firms. Journal of
Financial Economics 77 (1): 117–146.
Dallago, Bruno, and Robert McIntyre. 2003. Conclusions: The Role of Public
Policy in SME Development. In Small and Medium Enterprises in
Transitional Economies, edited by Robert McIntyre and Bruno Dallago,
206–225. New York: Palgrave Macmillan, in association with the United
Nations University/WIDER.
Demirgüç-Kunt, Asli, and Vojislav Maksimovic. 1998. Law, Finance, and Firm
Growth. The Journal of Finance 53 (6): 2107–2137.
De Soto, Hernando. 2000. The Mystery of Capital: Why Capitalism Triumphs in
the West and Fails Everywhere Else. New York, NY: Basic Books.
Dietrich, Michael. 1994. Transaction Cost Economics and Beyond: Towards a
New Economics of the Firm. New York: Routledge.
Dixit, Avinash K. 2004. Lawlessness and Economics: Alternative Models of
Governance. Princeton, NJ: Princeton University Press.
Dubini, Paola, and Howard Aldrich. 1991. Personal and Extended Networks are
Central to the Entrepreneurial Process. Journal of Business Venturing 6 (5):
305–313.
Ebers, Mark. 1997. Explaining Inter-Organizational Network Formation. In The
Formation of Inter-Organizational Networks, edited by Mark Ebers, 3–40.
New York: Oxford University Press.
EBRD. 2005. Law in Transition: Courts and Judges. London, UK: European
Bank for Reconstruction and Development.
Economides, Nicholas. 1996. The Economics of Networks. International Journal
of Industrial Organization 14 (2): 673–699.
Edwards, Rosalind, Jane Franklin, and Janet Holland, eds. 2006. Assessing
Social Capital: Concept, Policy and Practice. Newcastle, UK: Cambridge
Scholars Press.
Eggertsson, Thrainn. 2005. Imperfect Institutions: Possibilities and Limits of
Reform. Ann Arbor: University of Michigan Press.
Ellickson, Robert C. 1991. Order Without Law: How Neighbors Settle Disputes.
Cambridge, MA: Harvard University Press.
Emirbayer, Mustafa, and Jeff Goodwin. 1994. Network Analysis, Culture, and
the Problem of Agency. American Journal of Sociology 99 (6): 1411–54.
243
Ensminger, Jean. 2001. Reputations, Trust and The Principal Agent Problem.
In Trust in Society, edited by Karen S. Cook, 185–201. New York: Russell
Sage Foundation.
Estrin, Saul. 2002. Competition and Corporate Governance in Transition.
Journal of Economic Perspectives 16 (1): 101–124.
Fan, Ying. 2002. Guanxi’s Consequences: Personal Gains at Social Cost.
Journal of Business Ethics 38 (4): 371–380.
Farrell, Joseph, and Paul Klemperer. Forthcoming. Coordination and Lock-In:
Competition with Switching Costs and Network Effects. [Available at
http://www.nuffield.ox.ac.uk/economics/papers/2006/w7/
Farrell_Klempererwp.pdf].
Field, John. 2003. Social Capital. London: Routledge.
Flap, Hendrik Derk. 1994. No Man Is An Island: The Research Program of a
Social Capital Theory. World Congress of Sociology. Bielefeld, Germany.
World Congress of Sociology.
Foley, Michael W., and Bob Edwards. 1999. Is it Time to Disinvest in Social
Capital? Journal of Public Policy 19 (2): 141–173.
Franičević, Vojmir, and Will Bartlett. 2002. Small Business Development in
Croatia. In Small Enterprise Development in South-East Europe: Policies for
Sustainable Growth, edited by Will Bartlett, Milford Bateman, and Maja
Vehovec, 267–294. Boston: Kluwer Academic Publishers.
Frye, Timothy. 2004. Credible Commitment and Property Rights: Evidence
from Russia. American Political Science Review 98 (3): 453–66.
Fukuyama, Francis. 1995. Trust: The Social Virtues and the Creation of
Prosperity. New York, NY: Free Press Paperbacks.
Furubotn, Eirik G., and Rudolf Richter. 2000. Institutions and Economic
Theory: The Contribution of the New Institutional Economics. Ann Arbor:
University of Michigan Press.
Gambetta, Diego, ed. 1988. Trust: Making and Breaking Cooperative Relations.
New York: Basil Blackwell.
Glaeser, Edward L., David I. Laibson, Jose A. Scheinkman, and Christine L.
Soutter. 2000. Measuring Trust. The Quarterly Journal of Economics 115
(3): 811–846.
Glinkina, Svetlana. 2003. Small Business, Survival Strategies and the Shadow
Economy. In Small and Medium Enterprises in Transitional Economies,
edited by Robert McIntyre and Bruno Dallago, 51–63. New York: Palgrave
Macmillan, in association with the United Nations University/WIDER.
244
Granovetter, Mark. 1985. Economic Action and Social Structure: The Problem
of Embeddedness. American Journal of Sociology 91 (3): 481–510.
Granovetter, Mark S. 1973. The Strength of Weak Ties. American Journal of
Sociology 78:1368–80.
. 1974. Getting a Job: A Study of Contacts and Careers. Cambridge, MA:
Harvard University Press.
. 2005. The Impact of Social Structure on Economic Outcomes. Journal
of Economic Perspectives 19 (1): 33–50.
Greif, Avner. 1994. Cultural Beliefs and the Organization of Society: A
Historical and Theoretical Reflection on Collectivist and Individualist
Societies. Journal of Political Economy 102 (5): 912–950.
Grootaert, Christiaan. 1998. Social Capital: The Missing Link? Social Capital
Initiative Working Paper 3. World Bank, Washington DC.
Gustafson, Thane. 1999. Capitalism Russian-Style. Cambridge: Cambridge
University Press.
Hadfield, Gillian. 2004. The Many Legal Institutions that Support Contractual
Committments. In Handbook of Institutional Economics, edited by C. Ménard
and M. Shirley, 175–204. Netherlands: Kluwer Academic Publishers.
Hardin, Russell. 1993. The Street-Level Epistemology of Trust. Politics &
Society 21 (4): 505–529.
Hart, P.E. 2000. Theories of Firms’ Growth and the Generation of Jobs. Review
of Industrial Organization 17 (3): 229–248.
Hedström, Peter, and Richard Swedberg, eds. 1998. Social Mechanisms: An
Analytical Approach to Social Theory. New York: Cambridge University
Press.
Hellman, Joel, Geraint Jones, Daniel Kaufmann, and Mark Schankerman. 2000,
February. Measuring Governance and State Capture: The Role of
Bureaucrats and Firms in Shaping the Business Environment. Workshop on
the Institutional Foundation of a Market Economy, Berlin.
Hellman, Joel S. 1998. Winners Take All: The Politics of Partial Reform in
Postcommunist Transition. World Politics 50 (1): 203–34.
Hendley, Katherine, and Peter Murrell. 2003. Which Mechanisms Support the
Fulfillment of Sales Agreements? Asking Decision-Makers in Firms.
Economic Letters 78:49–54.
Hendley, Katherine, Peter Murrell, and Randi Ryterman. 2000. Law,
Relationships and Private Enforcement: Transactional Strategies of Russian
Enterprises. Europe-Asia Studies 52 (4): 627–656.
245
Hoff, Karla. 2001. Beyond Rosenstein-Rodan: The Modern Theory of
Coordination Problems in Development. Proceedings of The World Bank
Annual Conference on Development Economics 2000, Supplement to The
World Bank Review. Washington, D.C.: The World Bank, 145–188.
Hooghe, Marc, and Dietlind Stolle. 2003. Chapter 1 of Generating Social
Capital: Civil Society and Institutions in Comparative Perspective, edited by
Marc Hooghe and Dietlind Stolle, 1–18. New York: Palgrave Macmillan.
Hull, Galen Spencer. 1999. Small Businesses Trickling Up in Central and
Eastern Europe. New York: Garland Publishing.
Jackson, Matthew O. 2005. A Survey of Models of Network Formation: Stability
and Efficiency. In Groups Formation in Economics; Networks, Clubs and
Coalitions, edited by Gabrielle Demange and Myrna Wooders. New York:
Cambridge University Press.
. Forthcoming. The Study of Social Networks in Economics. In The
Missing Links: Formation and Decay of Economic Networks, edited by Joel
Podolny and James Rauch. New York: Russell Sage Foundation.
Jansen, Dorothea, and Mike Weber. 2004. Helping Hands and Entrepreneurship
- Supporting Newly Founding Firms. In Advances in Interdisciplinary
European Entrepreneurship Research, edited by Michael Dowling, Jürgen
Schmude, and Knyphausen-Aufsess, 57–79. Piscataway, NJ: Transaction
Publishers.
Johannisson, Bengt. 1995. Paradigms and Entrepreneurial Networks - Some
Methodological Challenges. Entrepreneurship and Regional Development 7
(3): 215–232.
Johnson, Simon, John McMillan, and Christopher Woodruff. 2002a. Courts and
Relational Contracts. The Journal of Law, Economics & Organization 18 (1):
221–277.
. 2002b. Property Rights and Finance. American Economic Review 92
(5): 1335–56.
Karklins, Rasma. 2005. The System Made Me Do It: Corruption in
Post-Communist Societies. Armonk, NY: M.E. Sharpe.
Knack, Stephen, and Philip Keefer. 1997. Does Social Capital Have an Economic
Payoff? A Cross-Country Investigation. The Quarterly Journal of Economics
112 (4): 1251–1288.
Kolankiewicz, George. 1996. Social Capital and Social Change. British Journal
of Sociology 47 (3): 427–441.
246
Kolodko, Grzegorz. 2003. Transition to a Market and Entrepreneurship:
Systemic Factors and Policy Options. In Small and Medium Enterprises in
Transitional Economies, edited by Robert McIntyre and Bruno Dallago,
153–170. New York: Palgrave Macmillan, in association with the United
Nations University/WIDER.
Kornai, Janos. 1992. The Socialist System: The Political Economy of
Communism. Princeton, NJ: Princeton University Press.
Kozak, Marta. 2007. Micro, Small, and Medium Enterprises: A Collection of
Published Data. World Bank Online Database. [Available at:
http://rru.worldbank.org/Documents/other/MSMEdatabase/
msme_database.htm].
Kranton, Rachel E. 1996. Reciprocal Exchange: A Self-Sustaining System. The
American Economic Review 86 (4): 830–851.
Krishna, Anirudh, and Elizabeth Shrader. 2002. The Social Capital Assessment
Tool: Design and Implementation. In Understanding and Measuring Social
Capital: A Multidisciplinary Tool for Practitioners, edited by Christiaan
Grootaert and Thierry Van Bastelaer, 17–40. Washington, D.C.: The World
Bank.
Kuehnast, Kathleen, and Nora Dudwick. 2004. Better a Hundred Friends Than
a Hundred Rubles? Social Networks in Transition – the Kyrgyz Republic.
Washington, DC: The World Bank.
Kuran, Timur. 2003. The Islamic Commercial Crisis: Institutional Roots of
Economic Underdevelopment in the Middle East. Journal of Economic
History 63 (2): 417–49.
. 2004. Why the Middle East is Economically Underdeveloped: Historical
Mechanisms of Institutional Stagnation. Journal of Economic Perspectives 18
(3): 71–90.
Landa, Janet. 1994. Trust, Ethnicity, and Identity: Beyond the New
Institutional Economics of Ethnic Trading Networks, Contract Law, and
Gift-Exchange. Ann Arbor: University of Michigan Press.
La Porta, Rafael, Florencio Lopez-De-Silanes, Andrei Shleifer, and Robert W.
Vishny. 1997. Legal Determinants of External Finance. The Journal of
Finance 52 (3): 1131–1150.
Lechner, Christian, and Michael Dowling. 2003. Firm Networks: External
Relationships as Sources for the Growth and Competitiveness of
Entrepreneurial Firms. Entrepreneurship and Regional Development 15 (1):
1–26.
247
Ledeneva, Alena V. 1998. Russia’s Economy of Favours: Blat, Networking and
Informal Exchange. Cambridge: Cambridge University Press.
. 2001. Networks in Russia: Global and Local Implications. In Explaining
Post-Soviet Patchworks, V.2, Pathways from the Past to the Global, edited by
Klaus Segbers, 59–77. Aldershot, England: Ashgate.
. 2003. Informal Practices in Changing Societies: Comparing Chinese
Guanxi and Russian Blat. Working Paper 45. Center for the Study of
Economic & Social Change in Europe, University College London, UK.
. 2004. Ambiguity of Social Networks in Post-Communist Societies.
Working Paper 48. Center for the Study of Economic & Social Change in
Europe, University College London, UK.
Levi, Margaret. 1996. Social and Unsocial Capital: A Review of Robert
Putnam’s Making Democracy Work. Politics & Society 24 (1): 45–55.
. 2000. When Good Defenses Make Good Neighbors. In Institutions,
Contracts, and Organizations: Perspectives from New Institutional
Economics, edited by Claude Menard, 137–157. Chichester, UK: Edward
Elgar.
Lin, Nan. 2001. Building a Network Theory of Social Capital. In Social Capital:
Theory and Research, edited by Nan Lin, Ronald S. Burt, and Karen Cook,
3–30. New Brunswick, NJ: Aldine Transactions.
Lin, Nan, Ronald S. Burt, and Karen Cook, eds. 2001. Social Capital: Theory
and Research. New Brunswick, NJ: Aldine Transactions.
Lipton, David, and Jeffrey Sachs. 1990. Creating a Market Economy in Eastern
Europe: The Case of Poland. Brookings Papers on Economic Activity 1990
(1): 75–1333.
Litwack, John M. 1991. Legality and Market Reform in Soviet-Type Economies.
Journal of Economic Perspectives 5 (4): 77–89.
Lomnitz, Larissa Adler, and Diana Sheinbaum. 2004. Trust, Social Networks and
the Informal Economy: A Comparative Analysis. Review of Sociology of the
Hungarian Sociological Association 10 (1): 5–26.
Lonkila, Markku. 1997. Informal Exchange Relations in Post-Soviet Russia: A
Comparative Perspective. Sociological Research Online 2, no. 2. [Available at
http://www.socresonline.org.uk/socresonline/2/2/9.html].
. 1999, November. Social Networks in Post-Soviet Russia: Continuity and
Change in the Everyday Life of St. Petersburg Teachers. Ph.D. diss.,
Department of Sociology, University of Helsinki, Helsinki.
248
Lovett, Steve, Lee C. Simmons, and Raja Kali. 1999. Guanxi Versus the Market:
Ethics and Efficiency. Journal of International Business Studies 30 (2):
231–248.
Macaulay, Stewart. 1963. Non-Contractual Relations in Business: A Preliminary
Study. American Sociological Review 28 (1): 55–67.
Manolova, Tatiana S., and Aimin Yan. 2002. Institutional Constraints and
Entrepreneurial Responses in a Transforming Economy. International Small
Business Journal 20 (2): 163–184.
Marsden, Peter V. 1990. Network Data and Measurement. Annual Review of
Sociology 16:435–63.
McCallister, L., and C. Fischer. 1978. A Procedure for Surveying Personal
Networks. Sociological Methods and Research 7 (2): 131–148.
McIntyre, Robert. 2003. Small Enterprises in Transition Economies: Causal
Puzzles ad Policy-Relevant Research. In Small and Medium Enterprises in
Transitional Economies, edited by Robert McIntyre and Bruno Dallago,
1–17. New York: Palgrave Macmillan, in association with the United Nations
University/WIDER.
McMillan, John, and Christopher Woodruff. 2000. Private Order Under
Dysfunctional Public Order. Michigan Law Review 98 (8): 2421–58.
. 2002. The Central Role of Entrepreneurship in Transition Economies.
Journal of Economic Perspectives 16 (3): 153–170.
Mihaylova, Dimitrina, and John Harriss. 2003. Breaking Out of the Vicious
Circle? Trust and Economic Development in the Balkans. Second
International Conference ICES 2003 "From Transition to Development:
Globalisation and Political Economy of Development in Transition
Economies", Volume 1. Sarajevo, Bosnia and Herzegovina: Faculty of
Economics of Sarajevo, 551–570.
Mill, John Stuart. 1970. Two Methods of Comparison. In Comparative
Perspective: Theories and Methods, edited by A. Etzioni and F.L. Du Bow,
205–213. Boston, MA: Little, Brown.
Murrell, Peter. 1995. The Transition According to Cambridge Mass. Journal of
Economic Literature 33 (1): 164–178.
. 1996. How Far Has the Transition Progressed? Journal of Economic
Perspectives 10 (2): 25–44.
North, Douglass C. 1990. Institutions, Institutional Change and Economic
Performance. Cambridge: Cambridge University Press.
. 2005. Understanding the Process of Economic Change. Princeton, NJ:
Princeton University Press.
249
Odell, John S. 2001. Case Study Methods in International Political Economy.
International Studies Perspective 2 (2): 161–176.
OECD. 1997. Globalisation and Small and Medium Enterprises (SMEs).
Synthesis Report, Vol.1. Paris: Organisation for Economic Co-operation and
Development (OECD).
. 2003. Federal Republic of Yugoslavia: Economic Assessment. OECD
Economic Surveys, Supplement No.3 (2002). Paris: Organisation for
Economic Co-operation and Development (OECD).
OECD-EBRD. 2003. South East Europe Region: Enterprise Policy Performance
Assessment. Joint Report, Organisation for Economic Co-operation and
Development and European Bank for Reconstruction and Development.
Paris: Organisation for Economic Co-operation and Development (OECD).
Ostgaard, Tone A., and Sue Birley. 1994. Personal Networks and Firm
Competitive Strategy - A Strategic or Coincidental Match. Journal of
Business Venturing 9 (4): 281–305.
. 1996. New venture Growth and Personal Networks. Journal of Business
Research 36 (1): 37–50.
Paldam, Martin. 2000. Social Capital: One or Many? Definition and
Measurement. Journal of Economic Surveys 14 (5): 629–653.
Peng, Mike W., and Peggy Sue Heath. 1996. The Growth of the Firm in
Planned Economies in Transition: Institutions, Organizations, and Strategic
Choice. Academy of Management Review 21 (2): 492–528.
Peng, Mike W., and Yadong Luo. 2000. Managerial Ties and Firm Performance
in a Transition Economy: The Nature of a Micro-Macro Link. Academy of
Management Journal 43 (3): 486–501.
Perry, Martin. 1999. Small Firms and Network Economies. New York:
Routledge.
Pfirrmann, Oliver. 2002. Introduction. In Small Firms and Entrepreneurship in
Central and Eastern Europe, edited by Oliver Pfirrmann and Gunter H.
Walter, 1–14. Heidelberg: Phisica-Verlag.
Pilling, Bruce K., Lawrence A. Crosby, and Donald W. Jr. Jackson. 1994.
Relational Bonds in Industrial Exchange: An Experimental Test of the
Transaction Cost Economic Framework. Journal of Business Research 30 (3):
237–251.
250
Pissarides, Francesca, Miroslav Singer, and Jan Svejnar. 2000. Objectives and
Constraints of Entrepreneurs: Evidence from Small and Medium Size
Enterprises in Russia and Bulgaria. William Davidson Institute Working
Paper Series 346. Stephen M. Ross Business School, University of Michigan,
Ann Arbor.
Pistor, Katharina, Martin Raiser, and Stanislaw Gelfer. 2000. Law and Finance
in Transition Economies. Economics of Transition 8 (2): 325–368.
Platteau, Jean-Philippe. 2000. Institutions, Social Norms, and Economic
Development. Amsterdam: Harwood Academic Publishers.
Podolny, Joel, and James Baron. 1997. Resources and Relationships: Social
Networks and Mobility in the Workplace. American Sociological Review 62
(5): 673–693.
Portes, Alejandro. 1998. Social Capital: It’s Origins and Applications in Modern
Sociology. Annual Review of Sociology 24:1–24.
Putnam, Robert. 1993. Making Democracy Work. Princeton, NJ: Princeton
University Press.
. 2000. Bowling Alone:The Collapse and Revival of American
Community. New York: Simon & Schuster.
Radaev, Vadim. 2003. The Development of Small Entrepreneurship in Russia. In
Small and Medium Enterprises in Transitional Economies, edited by Robert
McIntyre and Bruno Dallago, 114–133. New York: Palgrave Macmillan, in
association with the United Nations University/WIDER.
. 2004. How Trust is Established in Economic Relationships When
Institutions And Individuals Are Not Trustworthy: The Case of Russia. In
Creating Social Trust in Post-Socialist Transition, edited by Janos Kornai
and Susan Rose-Ackerman, 91–110. New York: Palgrave Macmillan.
Ragaru, Nadege. 2003. Uslugi: The Role of Political Favors and Connections in
Post-Communist Bulgaria. In New Approaches to Balkan Studies, edited by
Dimitirs Keridis, Ellen Elias-Bursac, and Nicholas Yatromanolakis. Dulles,
VA: Brassey, Inc.
Raiser, Martin. 1997. Informal Institutions, Social Capital and Economic
Transition: Reflections on a Neglected Dimension. Working Paper 25.
European Bank for Reconstruction and Development (EBRD), London.
. 1999. Trust in Transition. Working Paper 34. European Bank for
Reconstruction and Development (EBRD), London.
Raiser, Martin, Christian Haerpfer, Thomas Nowotny, and Claire Wallace. 2001.
Social Capital in Transition: A First Look at Evidence. Working Paper 61.
European Bank for Reconstruction and Development (EBRD), London.
251
Rao, P.K. 2003. The Economics of Transaction Costs: Theory, Methods and
Applications. New York: Palgrave, MacMillan.
Rauch, James E. 2001. Business and Social Networks in International Trade.
Journal of Economic Literature 39 (4): 1177–1203.
. 2005. Getting the Properties Right to Secure Property Rights: Dixit’s
Lawlessness and Economics. Journal of Economic Literature 43 (2): 480–487.
Reese, Pat Ray, and Howard E. Aldrich. 1995. Entrepreneurial Networks and
Business Performance: A Panel Study of Small and Medium-Sized Firms in
the Research Triangle. In International Entrepreneurship, edited by Sue
Birley and Ian MacMillan, 124–144. London; New York: Routledge.
Renzulli, Linda A., Howard Aldrich, and James Oody. 2000. Family Matters:
Gender, Networks, and Entrepreneurial Outcomes. Social Forces 79 (2):
523–546.
Richman, Barak D. 2004. Firms, Courts, and Reputation Mechanisms: Toward a
Positive Theory of Private Ordering. Columbia Law Review 104 (8):
2328–2367.
Rindfleisch, Aric, and Jan B. Heide. 1997. Transaction Cost Analysis: Past,
Present, and Future Applications. Journal of Marketing 61 (4): 30–54.
Rodrik, Dani. 2006. Goodbye Washington Consensus, Hello Washington
Confusion? Journal of Economic Literature 44 (4): 973–987.
Rose, Richard. 2000. Getting Things Done in an Antimodern Society: Social
Capital Networks in Russia. In Social Capital: A Multifaceted Perspective,
edited by Partha Dasgupta and Ismail Serageldin, 147–171. Washington, DC:
IBRD/The World Bank.
Rose-Ackerman, Susan. 2001. Trust and Honesty in Post-Socialist Societies.
KYKLOS 54 (2-3): 415–44.
Sanfey, Peter, Elisabetta Falcetti, Anita Taci, and Sladjana Tepic. 2004.
Spotlight on South-Eastern Europe. An Overview of Private Sector Activity
and Investment. London: European Bank for Reconstruction and
Development (EBRD).
Scase, Richard. 2003. Entrepreneurship and Proprietorship in Transition: Policy
Implications for the SME Sector. In Small and Medium Enterprises in
Transitional Economies, edited by Robert McIntyre and Bruno Dallago,
64–77. New York: Palgrave Macmillan, in association with the United
Nations University/WIDER.
Schumpeter, Joseph A. 1975 [orig. pub. 1942]. Capitalism, Socialism and
Democracy. New York: Harper.
252
Sedaitis, Judith B. 1997. Network Dynamics of New Firm Formation:
Developing Russian Commodity Markets. In Restructuring Networks in
Post-Socialism: Legacies, Linkages, and Localities, edited by Gernot Grabher
and David Stark, 137–157. Oxford; New York: Oxford University Press.
Shleifer, Andrei, and Robert W. Vishny. 1998. The Grabbing Hand: Government
Pathololgies and Their Cures. Cambridge, MA: Harvard University Press.
Sik, Endre. 1994. Network Capital in Capitalist, Communist and
Post-communist Societies. International Contributions to Labour Studies 4
(1): 73–93.
Smallbone, David, and Friederike Welter. 2001. The Distinctiveness of
Entrepreneurship in Transition Economies. Small Business Economics 16 (4):
249–262.
Sokić, Sreten. 2003. Ekonomija Tranzicije. Beograd, Serbia: Zavet.
Standifird, Stephen S., and Marshall R. Scott. 2000. The Transaction Cost
Advantage of Guanxi-Based Business Practices. Journal of World Business
35 (1): 21–42.
Stark, David. 1997. Recombinant Property in East European Capitalism. In
Restructuring Networks in Post-Socialism, edited by Gernot Grabher and
David Stark, 35–69. New York: Oxford University Press.
Stiglitz, Joseph E. 2000. Whither Reform? Ten Years of Transition. In Annual
World Bank Conference on Economic Development, edited by B. Pleskovic
and J.E. Stiglitz, 27–56. Washington, D.C.: World Bank.
Stolle, Dietlind. 2003. The Sources of Social Capital. In Generating Social
Capital: Civil Society and Institutions in Comparative Perspective, edited by
Marc Hooghe and Dietlind Stolle, 19–42. New York: Palgrave Macmillan.
Svejnar, Jan. 2002. Transition Economies: Performance and Challenges. Journal
of Economic Perspectives 16 (1): 3–28.
Svensson, Jakob. 2005. Eight Questions about Corruption. Journal of Economic
Perspectives 19 (3): 19–42.
Sztompka, Piotr. 1999. Trust: A Sociological Theory. Cambridge: Cambridge
University Press.
Todeva, Emanuela. 1998. East European Business Networks: A Review of
Dependencies and Strategies and Their Influence on Company Success. CIBS
Research Papers in International Business, No 12-98. London South Bank
University, London.
253
Tullock, Gordon. 1980. The Welfare Costs of Tariffs, Monopolies, and Theft. In
Toward a Theory of the Rent-Seeking Society, edited by James M. Buchanan,
Robert D. Tollison, and Gordon Tullock. College Station, TX: Texas A&M
University Press.
UNDP. 2000. Suspended Transition (1990-2000): Vulnerability Trends and
Perception. Early Warning Reports. Belgrade, Serbia and Montenegro:
United Nations Development Programme (UNDP).
UNECE. 2005. Economic Survey of Europe. New York and Geneva: United
Nations Economic Commission for Europe (UNECE).
. 2006. Small and Medium-Sized Enterprises in Countries in Transition.
Entrepreneurship and SMEs. New York and Geneva: United Nations
Economic Commission for Europe (UNECE).
Uzzi, Brian. 1999. Embeddedness in the Making of Financial Capital: How
Social Relations and Networks Benefit Firms Seeking Financing. American
Sociological Review 64 (4): 481–505.
Van der Gaag, Martin, Tom A.B. Snijders, and Henk D. Flap. Forthcoming.
Position Generator Measures and their Relationship to Other Social Sciences.
In Social Capital: Advances in Research, edited by N. Lin and B. Erickson.
New York: Oxford University Press.
Vukčević, Zoran. 2005. Development of SMEs in Montenegro: Strategy,
Institutional Support and Experience. In A Region in Transition. Marking
the EBRD Annual Meeting, 24–27. London: European Bank for
Reconstruction and Development (EBRD).
Wallace, Claire, Oksana Shmulyar, and Vasil Bedzir. 1999. Investing in Social
Capital: The Case of Small-Scale, Cross-Border Traders in Post-Communist
Central Europe. International Journal of Urban and Regional Research 23
(4): 751–70.
Wang, Hongying. 2000. Informal Institutions and Foreign Investment in China.
The Pacific Review 13 (4): 525–556.
Westlund, Hans, and Elin Nilsson. 2005. Measuring Enterprises’ Investments in
Social Capital: A Pilot Study. Regional Studies 39 (8): 1079–1094.
Williamson, Oliver E. 1979. Transaction-Cost Economics: The Governance of
Contractual Relations. Journal of Law and Economics 22 (2): 233–261.
. 1985. The Economic Institutions of Capitalism. New York: The Free
Press.
. 1991. Comparative Economic Organization: The Analysis of Discrete
Structural Alternatives. Administrative Science Quarterly 36 (2): 269–296.
254
Woodruff, Christopher. 2004. Establishing Confidence in Business Partners:
Courts, Networks, and Relationships as Pillars of Support. In Creating Social
Trust in Post-Socialist Transition, edited by Janos Kornai, Bo Rothstein, and
Susan Rose-Ackerman, 111–25. New York: Palgrave Macmillan.
Woolcock, Michael. 1998. Social Capital and Economic Development: Toward a
Theoretical Synthesis and Policy Framework. Theory and Society 27 (2):
151–208.
Woolcock, Michael, and Deepa Narayan. 2000. Social Capital: Implications for
Development Theory, Research and Policy. The World Bank Research
Observer 15 (2): 225–249.
World Bank. 2002. Bosnia and Herzegovina: Local Level Institutions and Social
Capital Study. Washington, DC: World Bank. [Available at
http://lnweb18.worldbank.org/ESSD/sdvext.nsf/09ByDocName/
BosniaandHerzegovinaLocalLevelInstitutionsandSocialCapitalStudyVol1/
$FILE/BosniaMainRpt.pdf].
Xin, Katherine R., and Jone L. Pearce. 1996. Guanxi: Connections as
Substitutes for Formal Institutional Support. Academy of Management
Journal 39 (6): 1641–1658.
Yamagishi, Toship. 2001. Trust as a Form of Social Intelligence. In Trust in
Society, edited by Karen S. Cook, 121–147. Nw York: Russell Sage
Foundation.
Yeager, Timothy J. 1999. Institutions, Transition Economies, and Economic
Development. Boulder, CO: Westview Press.
Zhou, Kevin Zheng, and Laura Poppo. 2005, September. Relational Contracts in
China: Governance and Contractual Assurance. Mimeo.
255
Appendix A
Interview Participants
Table A.1: SMEs whose ownwers and/or managers participated in the interviews.
Company Main Activity
Type of
Ownership
Number of
Employees
Person
Interviewed
1 Import and export of large electrical
machinery and equipment
Mixed 15 Manager
2 Exclusive distributor of tobacco
products from an MNC based in the
UK
Private 31 Manager
3 Publishing and printing Private 78 Manager/Owner
4 Security guards and monitoring sys-
tems
Private 800 Manager
5 Importanddistributionoffoodprod-
ucts. Trade in "second-level" iron;
Import and trade of plastic reservoirs
for agricultural machinery
Private 3 Manager/Owner
6 Publishing, import and export of
magazines, journals and newspapers
Private
(formerly state)
50 Manager
7 Exportoffrozenfruitsandvegetables Private 2 Manager/Owner
8 Exclusive distributor for computer
supplies from an international part-
ner
Private 28 Manager
9 Authorized dealer of bathroom sup-
plies
Private 4 Manager/Owner
10 Authorized dealer for plastic pipes
and reservoirs
Private 3 Manager/Owner
11 Food production, poultry products Private 60 Manager/Owner
12 Security guards Private 30 Manager/Owner
13 Manufacturing, metal furniture Private 14 Manager/Owner
14 Printing Private 15 Manager/Owner
Continued on next page
256
Table A.1: (continued)
Company Main Activity
Type of
Ownership
Number of
Employees
Person
Interviewed
15 Wholesale and retail trade in con-
struction materials
Private 16 Manager/Owner
16 Wholesale and retail trade (super-
markets)
Private 100 Manager/Owner
17 Large-scale baking production for
breads, pastries and cakes
Private 30 Manager/Owner
18 Medical lab Private 3 Manager/Owner
19 Food production, spices and oils Private 6 Manager/Owner
20 Large-scale baking production for
breads, pastries and cakes
Private 800 Manager/Owner
21 Food production, pickled vegetables
and fruit preserves
Private 38 Manager/Owner
22 Construction, equipment renting and
subcontracting
Private 25 Manager/Owner
23 Construction, aluminum and plastic
materials
Private 16 Manager/Owner
24 Setup and maintenance of telecom-
munication systems; Radio station
Private 6 Manager/Owner
25 Production of cleaning supplies and
soap based products; Transport ser-
vices in coastal areas
Private 20 Manager/Owner
26 Wholesaleandretailtrade; Foodpro-
duction, meat and dairy
Private 600 Manager/Owner
27 Authorized dealer for mobile services
and related equipment; Wholesale
and retail trade, computer equip-
ment, books, office and school sup-
plies
Private 43 Manager/Owner
28 TV station and news agency Private 30 Manager/Owner
29 Import and distribution of oil and its
derivatives
Private 80 Manager
30 Production and wholesale trade of
metal wiring
Private 12 Manager/Owner
31 Engineering and construction Private 52 Manager/Owner
32 Import and distribution of chemical
supplies
Private 8 Manager
33 Boutique wine production and sale Private 6 Manager/Owner
34 Information technology and software
design
Private 3 Manager
35 Chemical testing for food certifica-
tion
Private 32 Operations
Manager
36 Accounting and financial consulting Private 5 Manager
37 Information technology services and
repairs
Private 3 Manager
38 Import and distribution of watches
and small leather goods
Private 30 Manager
39 Retail trade in air conditioning units Private 3 Manager/Owner
Continued on next page
257
Table A.1: (continued)
Company Main Activity
Type of
Ownership
Number of
Employees
Person
Interviewed
40 Monitoring systems, installation and
security services
Private 25 Owner
41 Engineering and construction Private 23 Owner
258
Appendix B
Quantitative Dataset
B.1 Sampling Design
The total population of businesses registered with the Chambers of Commerce in
Serbia and Montenegro was first reduced to a sampling frame from which
businesses were randomly selected to be included in the final sample targeted for
the mail-in survey. The criteria used to reduce the total population of registered
businesses included: has less than 250 employees, does not operate in agriculture,
education, culture or entertainment, and has a registered telephone number. This
last criterion was included as an attempt to weed out non-operating or “fake”
businesses. Also, only businesses from the capital city of Belgrade in Serbia, and
cities of Podgorica, Herceg-Novi, and Kotor in Montenegro were selected for the
sampling frame. This was a deliberate decision given that the same frame was
used to identify individuals to participate in the interviews. Selecting only these
cities allowed the investigator to control the area within which interviews would
take place so that the investigator could physically visit any of the businesses that
decided to participate in the study. Of the total number of businesses that
259
satisfied the above criteria, each was assigned a temporary identification number.
Then, every business whose identification number ended in 3 and 5 were selected
for the sampling frame.
From the sampling frame every twentieth business was chosen by the
investigator to receive a copy of the survey. If the name of the general manager
and/or owner was available, the survey was sent directly to the named individual.
If this information was missing (as was the case for most of the businesses in
Montenegro), the letter was addressed to the “General Manager”. The mailed
questionnaire included a letter to the respondent from the primary investigator,
an information sheet about the survey and the rights of the respondent, the
survey itself, and a self-addressed stamped envelope in which the survey was to be
returned to the investigator. The return address for all of the surveys was a single
P.O. Box in Montenegro where the investigator personally collected them.
B.2 Sampling Procedure
The pilot survey was mailed out in the summer months of 2005. The goal of the
pilot stage was to test the general willingness of the businesses to respond and
mail the completed surveys back to Montenegro. In addition, three business
owners were asked to go through the survey and identify any problems with
wording, length, or the order of questions. Based on the comments received from
the business owners that reviewed the questionnaire, some of the questions were
restructured and simplified. Due to the poor response rate in the pilot stage, the
survey was subsequently shortened. In addition, the letter to the respondents
from the investigator, as well as the information sheet about the survey, were
printed to include the the university affiliation and the logo. The improved
260
questionnaire was mailed out during the summer of 2006. The response rate
increased slightly from 2005 but it is difficult to specify the degree to which the
improvements to the survey played a role.
To compensate for the poor response rate, the 66 pilot study surveys and the
returned questionnaires from 2006 were included in the final pool of completed
questionnaires. Both of the surveys included essentially comparable information
although the order of the questions changed from 2005 to 2006. This complicated
the coding procedure since the 2005 questionnaires had to be processed to match
the 2006 coding. The randomness of the results ensured that pair-wise deletion of
cases was appropriate in analyses in which some elements from the pilot (2005)
surveys were missing compared to the surveys from 2006.
During the summer of 2006, when it became evident that the original database
obtained from the Chambers of Commerce contained faulty and incomplete
information, the database was supplemented with two additional sources
containing the contact information for SMEs in Serbia. The first was obtained
from a foreign bank subsidiary in Belgrade which offered a contact list for
previous or current clients in the city of Belgrade. The second source included a
database of SMEs managed by the agency for the promotion of exports from
Serbia. Approximately 15% of the surveys were mailed to the businesses identified
through these alternative sources. It is unknown, however, how many of the
completed surveys came from these sources. No additional databases were used to
supplement the information obtained from the Chamber of Commerce of
Montenegro.
261

Abstract (if available)

Abstract The study bridges transition scholarship with social network studies to examine whether the social embeddedness of economic actors in transition is a source of advantage or an impediment for private sector growth. Transition scholars have recognized the omnipresent role of informal personal exchange, reciprocal ties, social networks, and exchange of favors as socialist legacies and path-dependent cultural aspects in the post-communist context. However, they have been less successful in recognizing the role of these cultural factors as agents of institutional change. Adopting a structural approach to social capital, this analysis centers on social networks and the resources embedded in the relationships that form them. Original empirical evidence from small and medium size enterprises (SMEs) in Serbia and Montenegro is utilized to examine whether differences in business success can be explained through the differences in micro-level social capital and the nature of informal exchange relationships.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

The divergence of the economic fortunes of Hindus and Muslims in British India: a comparative institutional analysis

PDF

The influence of social networking sites on high school students' social and academic development

PDF

Social capital and community philanthropy: the impact of social trust and social networks on individual charitable behavior and community foundation development

PDF

The intersection of grit and social capital: a mixed methods examination of successful first-generation college students

PDF

The Catholic church in Latin America: an evaluation of the institutional and political impacts of progressive church reforms

PDF

The manifestation of social capital within the mathematics, engineering, and science achievement (MESA) program

PDF

Mejor vida/better life and day-to-day exchanges: Networks of social exchange in contemporary arts practice

PDF

The underrepresentation of African Americans in information technology: an examination of social capital and its impact on African Americans’ career success

PDF

A view from below: the development and role of organizational social capital in neighborhood regeneration in Los Angeles

PDF

The role of skin tone on candidate evaluation: an analysis of racial socialization among Latino immigrants

PDF

Essays on health and aging with focus on the spillover of human capital

PDF

Effects of economic incentives on creative project-based networks: communication, collaboration and change in the American film industry, 1998-2010

Asset Metadata

Creator Bozovic, Iva (author)

Core Title The impact of social capital on economic performance: lessons from small and medium size enterprises (SMEs)

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Political Economy

Publication Date 09/19/2007

Defense Date 06/04/2007

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag institutions,OAI-PMH Harvest,Serbia and Montenegro,SMES,social capital,social networks,transition

Place Name Montenegro (countries), Serbia (countries)

Language English

Advisor Kuran, Timur (committee chair), English, Robert D. (committee member), Hadfield, Gillian (committee member)

Creator Email bozovic@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m829

Unique identifier UC149547

Identifier etd-Bozovic-20070919 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-587062 (legacy record id),usctheses-m829 (legacy record id)

Legacy Identifier etd-Bozovic-20070919.pdf

Dmrecord 587062

Document Type Dissertation

Rights Bozovic, Iva

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu