. w. swnimmumfiwwt . 331933. . a 4. $33.. I In” I . wkwufi. it it : .. tvr.\rl \hi \ 2:13, 3.3. 1 3.... r: . A....:. Wan-L.“ l 4613‘ .n This is to certify that the dissertation entitled Exploring Social Capital and Its Political Consequences: The Case of Taiwan presented by J ong-Tian Wang has been accepted towards fulfillment of the requirements for PhD degree in Political Science I Major pxbfessor Date 06/l2 /2002 MS U is an Affirmative Action/Eq ual Opportunity Institution 0-12771 UBRARY Michigan State University ‘ PLACE IN RETURN BOX to remove this checkout from your record. To AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE FEB 2 4 2004 6317s a ~. SEPKJ $5511 29007. 'Ur l 6/01 c:/ClRC/DateDue.p65—p.1 5 EXPLORING SOCIAL CAPITAL AND ITS POLITICAL CONSEQUENCES: THE CASE OF TAIWAN BY J ong-Tian Wang A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Political Science 2002 ABSTRACT EXPLORING SOCIAL CAPITAL AND ITS POLITICAL CONSEQUENCES: THE CASE OF TAIWAN By Jong-Tian Wang This dissertation examines the meaning of social capital, suggests frameworks for studying social capital, and conducts empirical research on the political consequences of social capital in the case of Taiwan. Literature has shown two different usages (meanings) of social capital. The author refers to sociologists’ view of social capital, i.e., resources embedded in social relations, as private social capital while political scientists’ understanding of social capital, i.e., social trust and network of civic engagement, has been called public social capital in this dissertation. Based on the general analytical frameworks provided here, the author uses empirical data collected from Taiwan to explore the political consequences of both types of social capital. Results show that for private social capital, those with social capital characterized as having weak ties are more likely to participate in local community activities as opposed to those who rely only on strong ties. Regarding public social capital, change of regime from authoritarian to democratic does not promote social trust at the individual level. On the contrary, the level of social trust has declined due to the uncertainty of the transition period. In addition, there is no statistically significant relationship between social trust and membership in voluntaiy associations as indicated in Putnam’s theory. Furthermore, public social capital does not necessarily sustain democratic beliefs. The results from Taiwan indicate the need to further current public social capital theory. Copyright by J ONG-TIAN WANG 2002 To my parents, Ififlfi and i‘f—Rfisffi Acknowledgments I would like to express my gratitude to Professor Darren Davis, who has served as my dissertation committee chair. Without his encouragement and support, this dissertation would not have been completed. I have also greatly appreciated the helpful advice of Professors Paul Abramson, Brian Silver, and Kenneth Williams, my dissertation committee members. Professor Steve Gold from the Department of Sociology has reminded me of sociologists’ concern for the issue of inequality of social capital. I also wish to thank Professor Tse-min Lin of University of Texas at Austin for pointing me to the idea of social capital. The Office of Survey Research of Academia Sinica was kind enough to assist in this research project by providing two series of surveys: “Taiwan Social Change Survey” and “Social Image Survey in Taiwan.” Taiwan National University’s “Legislative Election Survey” also benefited me greatly. Mr. Gabriel Musk’s timely help in editing my manuscript brought me much relief. I feel much indebted to my wife, Mei-Fang. She not only took the (demanding) job of being a wonderfiil at-home-mom but also taught me how to balance my roles as a husband, father, and social scientist. My daughter, Grace, is my source of joy. And finally, my parents’ unlimited love and care will be always in my heart. vi TABLE OF CONTENTS LIST OF TABLES ................................................................................... x LIST OF FIGURES ................................................................................ xii CHAPTER 1 INTRODUCTION ................................................................................... 1 The Study of Social Capital ................................................................ 2 The Purposes of This Study ............................................................... 6 Outline of Chapters ......................................................................... 8 CHAPTER 2 SOCIAL CAPITAL: ITS INTELLECTUAL AND SOCIAL ORIGINS .................... 18 Introduction ................................................................................. 18 What Is Social Capital .................................................................... 19 The Intellectual and Social Origins of Social Capital ................................. 24 Intellectual Origins of Social Capital .......................................... 24 Social Origins of Social Capital ................................................ 33 The Definitions of Social Capital: Sociologists vs. Political Scientists ............ 36 Conclusion ................................................................................. 48 CHAPTER 3 ANALYTICAL FRAMEWORKS FOR SOCIAL CAPITAL ................................. 51 Introduction ................................................................................. 5 l Studying Private Social Capital: A Theoretical Perspective ......................... 53 The Nature and Sources of Private Social Capital ............................ 53 Mechanisms of Private Social Capital ......................................... 56 The Consequences of Private Social Capital ................................. 62 An Analytical Framework for Private Social Capital ........................ 66 Studying Public Social Capital: A Theoretical Perspective ......................... 70 The Nature and Origins of Public Social Capital ............................ 72 The Maintenance of Public Social Capital .................................... 78 Consequences of Public Social Capital ....................................... 85 An Analytical Framework for Public Social Capital ........................ 92 Conclusion ................................................................................. 96 CHAPTER 4 METHOD, DATA, AND MEASUREMENT ................................................... 97 Introduction ................................................................................ 97 Methods ..................................................................................... 97 Data .......................................................................................... 98 Measurement .............................................................................. 101 vii A Review of Measuring Social Capital ...................................... 101 The Measurement of Social Capital in This Research ..................... 104 CHAPTER 5 STRUCTURAL SOURCES AND CONSTRAINTS OF THE SOCIAL CAPITAL: THE CASE OF TAIWAN ....................................................................... 110 Introduction ............................................................................... 1 10 Taiwan’s Cultural Heritage and Social Capital ...................................... 110 Chinese Society: A Highly Structured One .................................. 111 How Do Chinese Interact With Others: The Power of Guanxi ........... 114 Associational Life in Chinese Societies ..................................... 118 Horizontal Social Ties in Traditional Chinese Societies ................... 120 Summary .......................................................................... 122 Taiwan’s Historical Development and Social Capital ............................... 123 Taiwan’s History Before 1945 ................................................. 123 Taiwan’s History After 1945 .................................................. 126 Summary ......................................................................... 133 Taiwan’s Political Institutionalization and Social Capital .......................... 134 Political Party System .......................................................... 135 The State-Society Relations ................................................... 141 Electoral Politics and Local Factions ......................................... 147 Summary. ........................................................................ 152 CAHPTER 6 PRIVATE SOCIAL CAPITAL AND ITS POLITICAL CONSEQUENCES ............ 154 Introduction. .............................................................................. 1 54 Recapitulate The Concept of Private Social Capital and The Literature Review .................................................................................... 154 Strength of Ties Thesis and Its Application To Taiwanese Society ............... 158 The Strength of ‘Strong’ Ties in Taiwanese Society ...................... 159 Private Social Capital, Its Mechanism and Political Consequences: An Empirical Research ...................................................................... 165 Hypotheses ....................................................................... 165 Data, Measurement and Method .............................................. 171 Results and Discussion ......................................................... 177 Concluding Remarks ..................................................................... 193 Appendices ................................................................................ 195 CHAPTER 7 PUBLIC SOCIAL CAPITAL AND ITS POLITICAL CONSEQUENCES ............... 196 Introduction ............................................................................... 196 Critical Literature Review ............................................................... 197 The Paradoxical Relationship Between Democratic Polity and Trust: Tension vs. Reinforcement .................................................... 198 Voluntary Associations and Social Trust .................................... 203 viii Consequences of Public Social Capital: Maintaining Democratic Stability? ......................................................................... 208 Public Social Capital and Political Consequences: The Case of Taiwan .........211 Hypotheses ....................................................................... 212 Method and Data ................................................................ 213 Measurement ..................................................................... 214 Results and Discussion ......................................................... 220 Concluding Remarks ..................................................................... 247 Appendices ................................................................................ 249 CHAPTER 8 CONCLUSION .................................................................................... 251 Revisiting The Two Forms of Social Capital ........................................ 251 Summary and Research Findings ...................................................... 255 The Popularity and Ambiguity of The Concept of Social Capital ........ 255 Two Forms of Social Capital .................................................. 257 Analytical Frameworks for Studying Social Capital ........................ 257 Structural Sources of Social Capital in Taiwan ............................. 258 Political Consequences of Private Social Capital ........................... 259 Political Consequences of Public Social Capital ........................... 260 Where Should We Go From Here ...................................................... 262 BIBLIOGRAPHY ................................................................................. 264 ix 2-1 5-1 5-2 5-3 5-4 7-2 7-3 7-4 7-5 LIST OF TABLES A Sample of the Definition of Social Capital Percentage of Respondents Who Think Personal Achievement Is More Important Than Stable Family Life The Importance of the 2-28 Incident to Different Ethnic Groups National Identity and Ethnic Background Attitude Toward Social Movements and Related Groups Frequencies and Percentages of Major Independent Variables The Distribution of Respondents’ Characteristic of Private Social Capital Community Involvement and Types of Private Social Capital Effects of Social Positions on the Patterns of Private Social Capital: Logistic Regression Private Social Capital and Potential Local Community Involvement: Logistic Regression Private Social Capital and Local Community Involvement: OLS Regression Reciprocity and Social Resources: OLS Regression Number of Associations in Taiwan (1983-1998) Trends of Membership in Voluntary Association in Taiwan (1983-1998) Trends of Social Trust in Taiwan (1984-1998) Frequencies and Percentages of Major Independent Variables The Distribution of Membership in Voluntary Association Effects of Membership in Voluntary Association on 38 113 130 133 147 178 181 183 184 186 189 192 221 223 224 227 230 232 7-7 7-8 7-9 7-10 Interpersonal Trust: OLS Regression (1998 Taiwan Social Change Survey) Effect of Membership in Voluntary Associations on Interpersonal Trust: OLS Regression (1996 Taiwan Social Change Survey) Correlation Coefficients (Pearson’s r) of Public Social Capital and Political Tolerance, Democratic Beliefs, Political Participation, Political Trust, and Political Efficacy Effect of Public Social Capital on Political Tolerance and Support for Democratic Values: OLS Regression Effects of Public Social Capital on Political Trust, Political Participation, and Political Efficacy: OLS Regression xi 235 237 240 243 3-1 3-2 7-1 LIST OF FIGURES Michael Woolcock’s Typology of Bottom-Up Dilemmas of Development An Analytical Framework for Private Social Capital An Analytical Framework for Public Social Capital Trends of Public Social Capital In Taiwan (1983-1998) xii 66 68 93 226 CHAPTER 1 INTRODUCTION ‘Social capital’, being a relatively new analytical concept, has attracted scholars from multifarious disciplines to research a wide variety of subjects in the past two decades, especially in the 905. Most of the works focus either on how to solve ‘collective action problems’ or how to attain individual goals by using resources derived from social relations. For instance, why do countries with similar endowments of natural and human resources achieve very different levels of economic development? Why do countries or areas having similar institutional arrangements vary greatly in performance? Why are some schools more effective and their teachers more committed? Why do some immigrants better adapt to a new environment and thus succeed? The idea of ‘social capital’ provides us with valuable insight into the social aspects of these questions. Despite its popularity nowadays, social capital is not a new concept in social science. It also suffers ambiguity of meaning. This dissertation research aims to provide a broader view on the studying of social capital. Specifically, I will try to clarify the definition of social capital, synthesize current literature and suggest general analytical frameworks, piecing together important themes from social capital literature. And finally I will use Taiwan as a case to explore the relationship between social capital and its consequences, with special attention to its political consequences in Taiwan.' This proposal is five-fold: introduction, a brief review of the study of social capital, the purpose of this dissertation research, chapter outlines, methodology, data resources, and measurement issues. THE STUDY OF SOCIAL CAPITAL The contemporary meaning of social capital originated from sociologists. Two prominent advocates, James Coleman and Pierre Bourdieu, presented the basic idea of social capital. According to Coleman, social capital refers to: “A variety of different entities, with two elements in common: they all consist of some aspect of social structures, and they facilitate certain actions of actors—whether personal or corporate actors—within the structure...Unlike other forms of capital, social capital inheres in the structure of relations between actors. It is not lodged either in the actors themselves or in physical implements of production.” (Coleman, 1988298) This general definition points out the role of social structure in generating resources for the use of its members. Pierre Bourdieu, a French sociologist, defined social capital in terms of class or networks of relationships and called it, “The aggregate of the actual or potential resources which are linked to possession of a durable network of more or less institutionalized relationships of mutual acquaintance and recognition—or in other words, to membership in a group—which provides each of its members with the backing of the collectively-owned capital, a ‘credential’ which entitles them to credit, in the various sense of the wor .” (Bourdieu, 1986:248-249) ' 1 will explain in the third section of this proposal why I chose Taiwan as my case. Bourdieu’s definition highlights the importance of ‘connection’. People possess different volumes of social capital; in proportion to the size of the network of connections he can effectively mobilize and amount other forms of capital (economic, cultural or symbolic) possessed by members of that network. These two sociologists lay the foundation for contemporary study of social capital, which started in the 808 and began to thrive in the 908. The topics of social capital literature range from ethnic entrepreneurship (Portes, 1995; Light, 1984; Zhou, 1992), education (Coleman and Hoffer, 1987), health (Dreze and Sen, 1995) to poverty (Collier, 1998), economic development (Fukuyama, 1995; Knack and Keefer, 1997), management of common resources (Ostrom, 1990), and institutional performance (Putnam, 1993b), to name a few. Even the World Bank has paid attention to the notion of social capital. T Social Capital Initiative (SCI) was established in October 1996 as the Social Development Department of the World Bank to promote and strengthen social capital, with special efforts on social capital helping developing countries solve social and economic problems. Given the popularity of social capital, the fundamental problem facing this new field is the ambiguity of its definition. Scholars from various disciplines may refer to different ideas when speaking of social capital. In chapter two, a limited survey on the definitions of social capital in literature is conducted. The results show that there are two major camps attempting to use the concept of social capital in dealing with very different issues. One camp, mostly sociologists, tends to regard social capital as a resource embedded in social relations or social structures. The amount of social capital an individual may possess depends on the other members’ resources in his or her personal networks. And the forms (values) of social capital will shift according to the changes of sociopolitical and economic context.2 In addition, the norms or values held by individuals become social capital only if they facilitate others’ actions. Outside the group or network these norms or values may be of little value. Therefore social capital depends on a context. Furthermore, most sociologists highlight the instrumental dimension of social capital. In other words, social capital is a rational investment in an individual’s social relations with the aim to fulfill their personal goal(s), regardless of the fact that those goal(s) might be beneficial or harmful to others or society as a whole. The other camp, consisting mostly of political scientists and development theorists, also shows tremendous interest in the concept and roles of social capital. Political scientists differ from sociologists concerning the concept of social capital and the scope of its function.3 Most of them regard social capital as a resource to solve collective action problems or developmental problems in general instead of as a resource that individuals can appropriate for their own interests. Accordingly, they focus on subjects such as norms of reciprocity, trust, especially interpersonal trust, and civic associations (voluntary associations) and their functions in bringing about spontaneous cooperation among citizens. They further claim this positive effect will overcome the dilemma of collective action and/or lower transaction costs and hence produce desired socioeconomic development and stable democracy (Putnam, 1993b; Fukuyarna, 1995 & 2 The value of a specific form of capital, e.g., financial, social, cultural, or human, is determined by the socioeconomic and political context in which it operates. For various examples on this account, see Edward & Foley (1998: 129-130). Fukuyama also expressed a similar argument about the contingent values of social capital (1999:18-19). 3 Putnam is believed to be the leading figure in introducing and adopting social capital into the study of 1999; Evans, 1996; Knack & Keefer, 1997; Newton, 1997; Paxton, 1999; Pye, 1999; Rose, 1995; Stolle & Rochon, 1998; Sullivan & Transue, 1999; Tendler & Freedheim, 1994; Whiteley & Seyd, 1997; Booth & Richard, 1998). Thus, social capital becomes a context independent factor in political scientists’ viewpoint. More specifically, certain social norms, e.g., cooperative norms, and values, i.e., honesty, which are regarded as having the ability to facilitate collective actions, become a capital (resource) for the aggregate actors, whether it is groups, organizations, communities, regions, or nations. Despite the differing ideas, we can still find similarities to bridge these two camps. First, they stress the importance of stable and reliable social relations in conditioning either individual or collective actions. Second, most of them see social capital as a resource and focus on its positive consequences, although it has its ‘downside’.4 In order to differentiate these two ideas of ‘social capital’, I will use private social capital to refer to most sociologists’ view of social capital, i.e., the resources embedded in an individual’s social relations. On the other hand, public social capital will be used to refer to those social norms, values, and networks of civic engagement cherished by political scientists and development theorists in helping solve collective action dilemma. Both private and social capital will be addressed in this dissertation. The next two sections will describe my intentions in dealing with these two subjects in the context of Taiwan. political science. In Making Democracy Work, Putnam used sociologist James Coleman’s idea of social capital as a basis and highlighted its role in inducing social cooperation (Putnam, 1993b: 167). ‘ Portes and Landolt (1996) remind us that social capital has at least four negative consequences: exclusion of outsiders, excess claims on group members, restrictions of individual freedoms, and downward leveling norms. 1 will address these issues in greater detail in chapter three. THE PURPOSES OF THIS STUDY This research synthesizes the current social capital literature and applies it to the case of Taiwan. This research has three main aims. First, I would like to provide a broader view of social capital theory. As mentioned before, the idea of social capital is popular at present, but from this popularity arise problems which may confuse us. The most fundamental is the definition of social capital. Ironically, no consensus exists so far on what social capital is although almost everyone thinks it important. Furthermore, most literature seems to focus on particular aspects (or dimensions) of social capital- seldom does any work try to connect these dimensions organically. In this dissertation, I will do some preliminary work in hopes of suggesting general frameworks for studying social capital from a theoretical point of view. I believe this should be the first step for anyone interested in this new largely unexplored topic. Therefore, I will spend two chapters addressing the concept of social capital and its theoretical issues. Second, I want to address in greater detail the structural sources of private and public social capital in Taiwan. Literature has provided us with several plausible answerss, but we have not seen more comprehensive explanations on the sources of social capital in general or in specific societies. Therefore, in synthesizing the literature I argue that decisive historical events, cultural heritage, and other structural factors such as the 5 There are at least three approaches to explaining why social capital is easy or difficult to develop in a society. The first concerns social structural factors such as inequality or conflict (Boix & Posner, 1998). The second focuses on the cultural dimensions of society, arguing that shared cultural elements, e.g., symbols, meanings, and assumptions enhance or constrain people’s efforts to engage in social action (Wood, 1997). The third is an evolutionary approach. According to it, “cooperation could be triggered not political atmosphere are the basic dimensions in explaining and shaping the types of social connections among Taiwanese, which either facilitate or hinder their political intentions. This effort will also serve as background knowledge for the analyses, mostly quantitative, that follow in chapters five and six. Third, with the basis from social capital literature and the general framework that I will propose in chapter three, I would like to explore the political consequences of both private and public social capital in the case of Taiwan, keeping comparative perspective in mind. I will use survey data as my primary test of the controversial arguments over public social capital. I will also explore how public social capital influences people’s political attitudes and behaviors as well as how private social capital affects an individual’s involvement in community affairs. Detailed descriptions will be presented in the next section. As to the selection of the case subject, Taiwan holds several advantages from the perspective of both social capital and comparative politics. First, Taiwan is the only Chinese society in the world so far that has gone through the process of democratization successfully. It provides us with an opportunity to observe the relationship between social capital, especially public social capital, and regime types. Does the level of social capital increase or decrease as the regime type changes from authoritarian to democratic?6 In addition, does the level of (public) social capital in a newly democratized country correlate with people’s beliefs toward democratic principles, which are the bedrock of by trust, but simply by a set of fortunate practices, random at first, and then selectively retained (with varying degrees of learning and intentionality)” (Bateson, 1988). 6 A cross-national analysis of democracy and social capital revealed a positive relationship between trust and democracy (Paxton, 1999: 103). However, the case of Taiwan provides us with a better chance to observe the change of stock of social capital in two different regime types while being able to control as stable democracy? I will provide some preliminary thoughts and answers which contribute to the present research cannon. Second, since social capital theory was developed in the West, the subjects and scenarios of social capital research are mostly western.7 From the perspective of theory building, it would be interesting to see to what extent societies influenced by oriental culture fit the theory. The case of Taiwan is a pioneer work. Third, to the best of the author’s knowledge, there is little research on private social capital and its political consequences.8 As a political scientist, I am interested in knowing whether private social capital affects political attitudes or behaviors. The survey data collected in 1997 by the Institute of Ethnology Academia Sinica, Taiwan, Republic of China provides us with some related materials, which can be used for this explorative purpose.9 The following section briefly describes what each chapter in this dissertation deals with. OUTLINE OF CHAPTERS Chapter one provides an introduction to this research, including purposes and chapter outlines. many variables as possible. 7 For examples, see Putnam’s works on Italy and United States (1993b, 1995), and Hall’s research on Great Britain (1999). 8 Most research done by sociologists is focused on the relationship between private social capital and goal- oriented (instrumental) actions, such asjob searching (Lin, Ensel, and Vaughn, 1981) and increasing economic returns (Fernandez, Castilla, Moore, 2000). 9 This data set allows us to test the relationship between private social capital and people’s involvement in community affairs. For details, please see ‘chapter outlines’ and ‘measurement’ in this proposal. Chapter two explores the intellectual and social origins of the idea of social capital. This chapter provides a fundamental knowledge of how we have reached the modern idea of social capital. Many excellent literature reviews on the origins of social capital have led to the current views,'0 however they tend to focus on its intellectual roots and seldom address some unique features of contemporary society that may reveal the salience of social capital. Therefore, chapter two gives a more balanced view on the origins of social capital. Chapter three aims to suggest two analytical frameworks for private and public social capital respectively. A more detailed theoretical discussion will be presented. The format I use for the discussion is organically-oriented. I will discuss both forms of social capital, private and public, starting from their nature and sources; mechanisms by which social capital exerts its influences; and the maintenance (or reproduction) of social capital to the general consequences of social capital. Finally, I will synthesize the literature and my discussions in this chapter and propose general analytical frameworks for analyzing private and public social capital. These general frameworks will accordingly serve as guidelines for the following chapters of this dissertation. Chapter four introduces the methodology used in this research. In addition to the conceptualization of social capital and providing analytical frameworks, the remainder of this dissertation consists of empirical studies according to the frameworks. Thus, it is necessary to give an overview of the methods, the nature of the data, and measurements of major concepts in this chapter. '0 For instance, Michael Woolcock’s “Social Capital and Economic Development: Toward a Theoretical Synthesis and Policy Framework,” (1998); and Alejandro Portes’ “Social Capital: lts Origins and Applications in Modern Sociology,” (1998). Chapter five discusses the structural sources of social capital in Taiwan. Here structural sources mean the long-term factors influencing the configurations and formation of social capital. Thus this chapter serves two important purposes: providing background knowledge of Taiwan’s historical, cultural and political development with particular emphases on their impacts on social capital formation; and supplying information from which hypotheses can in the latter chapters be derived. 1 will mainly focus on three broad sources, which affect social capital formation in Taiwan, i.e., cultural heritage, major historical events, and politics after 1945. I argue that culturally Chinese societies, including Taiwan, are primarily relation-oriented. This characteristic makes it especially appropriate to be examined under the scrutiny of social capital theory.ll In addition, this relation-oriented characteristic plays an important role in interpersonal conflict solving, which accordingly reduces their reliance on formal organizations in dealing with conflicts. Furthermore, interpersonal trust, an important component of public social capital, is low due to the cultural heritage.'2 Regarding historical development, I single out two decisive events, which happened after 1945,'3 i.e., the incident of February 28th in 1947 and massive " For example, the norm of reciprocity is extremely important in Chinese social networks. The term Chinese use to refer to reciprocity is called ‘Bao”, meaning you should reciprocate at least as much as you receive from others. Otherwise, you will lose your ‘face’ and trustworthiness in your personal networks, and hence being isolated by others. More details are in chapter five. '2 Confucianism, the dominant guideline to Chinese moral behaviors, proposes ‘five cardinal relations’ to teach people the basic orientations towards others. It makes no mention about how to treat strangers. Coupled with the emphasis of interpersonal networks and family bonds, this implies a dichotomous treatment to strangers: you either pull the strangers into your personal networks by connections (quanxi) and trust them or treat strangers as untrustworthy. '3 The geographic location of Taiwan is peripheral to Mainland China. The majority of residents before the 17'“ century are the so—called aborigines. Starting from 16705 to 17505, Taiwan underwent a complex process of Sinicization whereby numerous Han emigrants colonized the native tribes, claimed new land, and developed marketing networks with the coastal mainland (Chen, 1999). Taiwan was ceded to Japan as the result of the Sino-Japan war in 1895. Having been colonized by Japan for 50 years, Taiwan was returned to China in 1945 when World War 11 ended. Therefore, the year of 1945 marks Taiwan’s return to China. For 10 immigration of mainlanders into Taiwan.l4 I argue that these historical events bring two fundamental results in terms of the development of social capital. First, due to the language barrier, insufficient information and unpleasant initial encounters between Taiwanese and newly immigrated mainlanders, sheng-chi wen-ti (problems of different provincial origins) evolved as an important issue, which has very broad impacts on social, economic, and political spheres in Taiwan thereafter.15 Second, related to the first, mutual distrust between mainlanders and Taiwanese not only constrains the expansion of (public) social capital but also obstructs the formation of national identity between the two groups. As far as the political dimension is concerned, the general image of political development before 19903'6 in Taiwan can be characterized as “strong state, weak society.” Ever since the KMT took over Taiwan, it has faced problems of legitimacy. Therefore it has to find ways on the one hand to strengthen its control over society and on the other to provide some inclusive institutions to gain the support of Taiwanese. Around this consideration I will highlight three areas, which affect the development of social capital. The first one concerns the party system. Since the nationalists took over Taiwan after World War II, the KMT has been the only political party to rule for over 50 years.'7 The one party system results in a culture lacking cooperation and trust. The second pertains to the KMT’s way of controlling society. It is generally recognized that the relationship between state and society in Taiwan before democratization can be simplicity, 1 will limit my analysis of Taiwan’s historical development to events after 1945. ‘4 The nationalists (KMT) were defeated by the Chinese Communist Party (CCP) in the civil war in 1949, and an estimated million plus people, mostly soldiers, bureaucrats, and their families, followed Chiang Kai- shek, leader of the KMT, to Taiwan. '5 More details will be addressed in chapter five. '6 Taiwan underwent political liberalization in the mid 19805. This process progressed to democratization in the 19905. The first presidential election by the general public in 1995 marked the completion of democratization. Taiwan was regarded as a democratic country thereafter. characterized as “authoritarian corporatism” or “state corporatism” (Ding, 1999). This institutional arrangement has three important characteristics: first, according to the law, most occupations are required to form their own occupational group and every one in that occupation should participate in that group. Therefore the government controls the interest aggregations of the society. Second, voluntary associations cannot be formed without government permission. In addition, voluntary associations, including the above- mentioned occupational groups, similar in nature may not coexist in the same area. These regulations resulted in many small-scale associations, which lack the ability to integrate horizontally. From the perspective of the state, it can avoid horizontal linkage among voluntary associations and therefore ensure its strong autonomy. Third, the KMT strongly monitors the personnel and finance of most voluntary associations, especially those related with economic or social functions, such as Farrner’s Associations, Worker’s Unions in different levels, or different business associations. That way, the state can not only make sure these associations act according to state policies, but also prevent social turmoil (Wade, 1990). The third factor regards the KMT’s ruling strategies in electoral politics. As I mentioned before, the nationalists faced legitimacy issues after takeover in 1945. In addition, the KMT decided to enforce the policy of local self-government. Therefore they needed local elites’ help to mobilize the Taiwanese on various levels of local elections.'8 Given the existing political factions at that time among Taiwanese politicians, the KMT finally adopted the strategy of accepting and supporting these local '7 Marshal law was enacted in May 1949, which prohibited the formation of new political parties. '8 Immediately after the takeover, the KMT tried to destroy the three major local factions in Taiwan. But they could only get rid of leaders of those factions; the actual networks related to those factions had already spread all over Taiwan. Thus the KMT decided to accept the existence of local factions and use them as tools to mobilize people, especially in local elections (Chen, 1995). 12 factions in exchange of their help in electoral mobilization. However, the KMT did not allow these local factions to form horizontal linkages (coalition) in elections. Local factions can only exert their influences at their own county level or below. Besides, the KMT tried to support two or more local factions in the county level and purposely provoke distrust between factions. This so-called ‘bifactionalism’ was designed to create check—and—balance between local factions in order to limit their further development into uncontrollable powers (Chen, 1995). These unique features of the patron-client relationship between the KMT and local factions form a political culture not conducive to the expansion of social capital. Chapter six concentrates on private social capital and its political consequences in Taiwan using survey data.'9 The methods I use are primarily network-oriented and quantitative in nature.20 The level of analysis is individual. The idea is inspired by Granovetter’s thesis of “strength of weak ties.”2| There are three major themes in this chapter: first, I will test the relationship between private social capital and community involvement. An individual’s instrumental ties will be used as the indicator of his or her private social capital.22 These instrumental ties can be further categorized as either strong ‘9 The source and purpose of the survey data will be given in the next section. 2° 1 will briefly mention how I measure a respondent’s social network using survey data later in this proposal. In 1973, Mark Granovetter published a paper called “The Strength of Weak Ties”. The purpose of this paper was to explain how interaction in small groups aggregates to form large-scale patterns through the analysis of processes in interpersonal networks. In this paper he emphasized ‘weak ties’, i.e. our acquaintances, do not cause alienation. Rather, they are in general vital for an individual’s integration into modern society. The significance of weak ties is that they are far more likely to bridge social circles different from one’s own than are strong ties, i.e., close friends or relatives. This function can be useful on both the individual and aggregate level. For example, weak ties better spread ideas or information, which can be used for job searching or collective movements. Granovetter’s arguments received high attention and many works were inspired by his ideas (Granovetter, 1973; 1982). More detailed descriptions of strong ties and weak ties will be addressed in chapter six. 22 The reasons 1 want to select instrumental ties as the indicator of private social capital will be discussed later in the ‘measurement’ section. ties or weak ties, depending on the closeness of the people involved in the interpersonal ties. My major task in this regard is to generate from Granovetter’s theory hypotheses about the strength of weak ties. His general argument relating to macro phenomena is that individuals with few weak ties were unlikely to be able to mobilize effectively for collective action within their communities23 (Granovetter, 1982: 126). Therefore, I would expect individuals having weak ties in their instrumental networks to be more likely to involve themselves in community affairs and also to be more likely to participate in more group activities than those having only strong ties in their instrumental networks. Second, I test the relationship between frequency of reciprocity and tie strength. Putnam regards the norms of reciprocity as one component of social capital. I argue in chapter three that norms of reciprocity are better seen as the mechanism that maintains social capital. Given the importance of norms of reciprocity in social capital literature, we seldom see any empirical analysis on this subject. My intention here is to conduct preliminary research on the actual operation of this mechanism. Literature indicates that social actors’ strong ties have greater motivation to assist and be available as opposed to their weak ties (Granovetter, 1982:113). In the context of reciprocity, people with pure strong ties in instrumental relations have higher frequencies of reciprocation. In addition, I explore the relationship between frequency of reciprocity and help-givers’ social resources. Although reciprocity is important to the maintenance of social capital, the frequency of reciprocity should not be regarded as symmetrical unless the two parties involved in exchanging favors have equal amounts of social resources. Thus those who 7'3 Granovetter concurred with Karweit’s argument that the lack of bridges between weak ties tends to leave most without even indirect access to leaders of the particular organizations in question. Thus alienation and a sense of helplessness contribute to the decision-making process (Granovetter, 1982:123). l4 own more social resources than their help-givers tend to give more help to people in his instrumental ties than those who own less social resources.24 That is because help-givers with more social resources may have less chance to ask for others’ help.25 In addition, from the perspective of accumulating reputation or credit slip from those he or she has helped, imbalanced transactions are actually a rational choice for both sides engaging in social exchange. Chapter seven deals with public social capital in Taiwan and its political consequences. I use primarily statistical methods in testing the hypotheses that I propose in this chapter. The data I use are secondary survey data. Interpersonal trust and participation in voluntary association will be two major indicators of public social capital. There are two themes in this chapter: first, whether the change of regime type has any effect on the stock of public social capital for society as a whole. From chapter five I trace three major sources that condition the development of macro level social capital in Taiwan. Most of these factors point to negative effects on the formation of social capital. Thus, I argue that authoritarian regimes are less capable of producing social capital compared to democratic regimes. In the case of Taiwan, 1 hypothesize that the aggregate level of public social capital in Taiwan before democracy was less than it is in the era of democracy. Second, I will explore the political consequences of public social capital at the individual level. There are several issues I want to address. 1) I will treat social trust as a 2‘ Here ‘more’ or ‘less’ mean the frequency of reciprocity, not the quality of reciprocity, which is difficult to measure in quantitative research. 2’ Of course this does not mean people with more social resource are all altruists. I will use the example of patron-client relationship in the local faction politics to argue that people rich in social resources give favors to those needed in exchange of respect and other potential supports, such as electoral mobilization. In other 15 dependent variable and see if participation in voluntary organization really increases social (interpersonal) trust in Taiwan.26 I believe that participation in voluntary associations may not be always be beneficial in terms of interaction with others. Besides, the cultural heritage and the influence of the KMT’s authoritarian corporatism relegates the role of social groups including voluntary associations lower than in both Putnam’s Italy and USA. Thus, I expect either no relationship or very little, if any, between social trust and participation in voluntary associations. 2) By treating public social capital as independent variable, I explore its impacts on several political attitudes and behaviors thought to be the bedrock of stable democracy. They include beliefs in democratic values, political participation, political trust, political tolerance, and political efficacy. Specifically, beliefs in democratic values will be broken down into four dimensions: support of rule of law, support of check-and-balance, support of right of freedom and support of equal rights.27 Political participation will be examined in the forms of voter participation and non-electoral participation. Political trust will be divided into trust in politicians and trust in government as a whole. words, the ‘desire for recognition’ may serve as incentive for giving favors. This idea is borrowed from F ukuyama (Fukuyama, 1995). 26 There are debates on the relationship between social trust and participation in voluntary associations, two most important components of public social capital. Putnam, for example, argued “people who join are people who trust. . .the causation flows mainly from joining to trusting” (1995b: 666). Others, like Tocqueville (1968), and M111 (1910) also stressed that networks of voluntary activity create trust and cooperation. Nevertheless, some scholars criticized this argument as ignoring the possibility of social conflict that voluntary associations may provoke (Whittington, 1998; Olson, 1982). Still others, like Sullivan & Transue, question the assumption of this argument, which generalizes attitudes from one domain to people in general. Rich examples from psychological studies of perception show that people can compartrnentalize their attitudes (Sullivan & Transue, 1999: 648). In addition, participating in voluntary association may not be always good in terms of interaction with others. Thus the effects of membership in voluntary association on social trust should be either zero or very small. 27 My data allows me with the help of factor analysis to sort democratic beliefs into several dimensions instead of talking about the general and abstract idea of democratic beliefs. This way, 1 can test the effects of social capital on each of the dimensions. A more comprehensive view on the effects of social capital on beliefs in democratic values is therefore possible. By examining the effects of public social capital on these issues a more thorough understanding of the roles of public social capital in supporting democracy may be attained. Most social capital literature holds positive views on the role of social capital in sustaining democracy.28 However, I argue, according to the framework I propose, that public social capital may not always be good for every dimension of democracy, although it is widely seen as easier to cultivate under democratic system.29 I suspect public social capital may have mixed, i.e., positive, negative or no, effects on those political attitudes and behaviors thought desirable for stable democracy. I propose hypotheses according to the literature, test them using survey data, and hopefully thus demonstrate public social capital has mixed effects, both positive and negative, in terms of sustaining democracy.30 This intention implies my work dealing with public social capital and its political consequences is primarily explorative instead of explanatory. Chapter eight concludes this dissertation. Findings of previous chapters, comments on these findings and suggestions for future studies will be summarized and discussed. 2' For instance, Eric M. Uslaner argued that trusting people are more tolerant and acceptant of minority cultures. (1999:141) Orlando Patterson contended that greater generalized trust (social trust) leads directly to greater political trust. In addition, political trust, combined with other socioeconomic factors, determines the level of political participation and activism (1999: 196). 29 For example of arguments that the democratic system better fosters public social capital, see Sztompka, 1999; Levi, 1996; and Muller & Seligson, 1994. 3° For example, social trust may have positive effects on respondents’ trust in government, and this accordingly may have negative effects on their attitudes toward the principle of check and balance because they see less need for a trusted government to be checked by formal institutions. 17 CHAPTER 2 SOCIAL CAPITAL: ITS INTELLECTUAL AND SOCIAL ORIGINS INTRODUCTION ‘Social capital’, being a relatively new analytical concept, has spurred scholars from various disciplines to research a wide variety of subjects in the past two decades, especially in the 903. Most works focus either on solving ‘collective action problems’ or attaining individual goals by using resources derived from social relations. For instance, as previously posed, why do countries with similar endowments of natural and human resources end up achieving very different levels of economic development? And why do countries or areas having similar institutional arrangements vary in performance? Why are certain schools and teachers more effective? Which immigrants will adapt to new environments and succeed? The social aspects of these questions highlight the role played by social capital. Again, despite its popularity nowadays, social capital is not new to social science, and it also suffers from ambiguity. Thus in this chapter I intend to briefly introduce the meaning of social capital, its intellectual and social origins, and approaches widely used in the study of social capital. The theoretical issues related to social capital will be discussed in the next chapter. 18 WHAT IS SOCIAL CAPITAL? The current definition of social capital originated from sociologists, of whom James Coleman and Pierre Bourdieu are among the most prominent in defining social capital. According to Coleman', social capital refers to: “A variety of different entities, with two elements in common: they all consist of some aspect of social structures, and they facilitate certain actions of actors—whether personal or corporate actors—within the structure...Unlike other forms of capital, social capital inheres in the structure of relations between actors. It is not lodged either in the actors themselves or in physical implements of production.” (Coleman, 1988:98) This broad definition points out the role of social structure in generating resources for the use of its members. For example, “a group within which there is extensive trustworthiness and extensive trust is able to accomplish much more than a comparable group without that trustworthiness and trust”(Coleman, 1988:101). It also implies that social capital may not always be beneficial as many wrongly assume because “a given form of social capital that is valuable in facilitating certain actions may be useless or even harmful for others” (Coleman, 1988:98). It would be safe to call Coleman’s social capital as ‘neutral’, for it is only a resource embedded in the social relations. As to what area it will be used depends on the actor’s goals, whether it is socially desirable or not. Regarding social structures favoring the creation of social capital, Coleman’s ' In his article, Coleman tries to demonstrate that Catholic and other religious schools tend to have lower dropout rates compared to other nonreligious private schools. The social capital generated from multiplex relationships between religious structures and educational structures is the main reason for this difference. 19 points can be boiled down to two mechanisms: norms and closure. With appropriate norms like obligations, reciprocity and trustworthiness, people can rely on each other’s help and sanction defectors. If the norms are internalized in members, they become a resource, which can be appropriated by others. Therefore, members of the community or network can all be the holders of social capital. For example, they can give their loans without wonying about nonpayment, or they can send their children to play in the streets without concern. This mechanism is further enhanced by closure in social networks, which induces people in the network to interact in more than one arena. An example often cited from Coleman’s article is the wholesale diamond market in which a merchant will hand over to another merchant “a bag of stones for the latter to examine in private at his leisure, with no formal insurance that the latter will not substitute one or more inferior stones or a paste replica. Such free exchange of stones for inspection isimportant to the function of this market. In its absence, the market would operate in a much more cumbersome, much less efficient fashion.” It is the close ties, through family, community. and religious affiliation, i.e., closure in social networks, that provide the insurance that is necessary to facilitate the transactions in the market (Coleman, 1988:99). In other words, closure offers the structure necessary to the effectiveness of norms. Pierre Bourdieu, a French sociologist, defines social capital in terms of A class or networks of relationships and calls it, “The aggregate of the actual or potential resources which are linked to possession of a durable network of more or less institutionalized relationships of mutual acquaintance and recognition—or in For detail, see Coleman (1988). 20 other words, to membership in a group—which provides each of its members with the backing of the collectively-owned capital, 3 “credential” which entitles them to credit, in the various sense of the word.”2 (Bourdieu, 1986:248-249) For Bourdieu, social capital is never independent of other forms of capital. People possess differing volumes of social capital according to both the size of the networks of connections he can effectively mobilize and the amount of other forms of capital (economic, cultural or symbolic) possessed by members of that network. This highlights the importance of ‘connections’. Networks of relationships facilitate the conversion of different forms of capital. For example, in order for economic capital3 to be converted into social capital, considerable labor is needed, including expenditure of time, attention, care, and concern. Bourdieu thinks of these efforts as forms of social exchanges, often ignored by economists, which will eventually bring profits in the long run, either monetary or other forms. In other words, an important characteristic for social capital highlighted here is that it is a rational investment in the social context, especially when goods or services are not available to actors through the use of economic capital. People should invest in sociability and devote labors mentioned above in order to access those goods or services (Bourdieu, 1986:252-253). Again, these two sociologists lay the foundation for the study of social capital. 2 Bourdieu tried to identify three forms of capital—economic, cultural, and social—and focused on the convertibility of the three types. According to him, economic capital is the root of the other two, and capital has the tendency to reproduce itself. Given the inequality of economic capital in society, the accumulation of social and cultural capital in fact reproduces the inequality of social structure, which perpetuates economic inequality. Actually, his work was an effort to express Marxian ideas about the reproduction of classes using the economic notion of capital transformation. For a detailed discussion, see Bourdieu (1986) and James N. Baron & Michael T. Hannan (1994:1117). 3 Economic capital, according to Bourdieu (1986:243), refers to resources, which are immediately and directly convertible into money and may be institutionalized in the form of educational qualifications. 21 The roles social relations and social context play in conditioning human economic and political behaviors thus regain prominence. This trend has greatly influenced political scientists and given them a new analytical tool in explaining a variety of political phenomena. Robert Putnam, among them the most influential, has with Making Democracy Work (1993b), stimulated a great deal of ongoing discussion on the idea of social capital and its relationship with democracy, civil society, and development.4 This book is even reviewed by The Economist as being the most important work of social science since Pareto and Max Weber (Harriss & Renzio, 1997: 922). In Making Democracy Work, Putnam seeks the preconditions that account for the superior government performance of the northern and central regions in Italy as opposed to their southern counterparts. According to Putnam, the key lies in the different stocks of social capital held in those regions. Social capital here refers to “features of social organization, such as trust, norms, and networks, that can improve the efficiency of society by facilitating coordinated actions” (Putnam, 1993b:167). In a latter article, he summarized social capital as “social connections and the attendant norms and trust” (Putnam, 1995b: 664-665). Putnam argued that those areas with a higher level of social ’ To name just a few, for example, Margaret Levi, “Social and Unsocial Capital,” Politics and Society 24(1): 45-55, 1996; Margaret Levi, “A state of Trust” in Trust & Governance, edited y Valerie Braithwaite and Margaret Levi (New York: Russel Sage Foundation, 1998); Sidney Tarrow, “Making Social Science Work Across Space and Time: A Critical Reflection on Robert Putnam’s Making Democracy Work,” American Political Science Review 90(2): 389-397 (1996); Louis Puttemran, “Social Capital and Development Capacity,” Development Policy Review 1(13):5-22, 1995; Jonathan Fox, ”How Does Civil Society Thicken?: The Political Construction of Social Capital in Rural Mexico,” World Development 24(6): 1089-1 103, 1996; Francis Fukuyama, Trust: The Social Virtues and the Creation of Prosperity, New York: The Free Press, 1995; Robert p. Weller, “Horizontal Tie and Civil Institutions in Chinese Societies,” in Democratic Civility, edited by Robert W. Hefner, New Brunswick: Transaction Publishers, 1998; Sandra Hofferth, Johanne Boisjoly and Greg J. Duncan, “The Development of Social Capital,” Rationality and Society 11(1): 79-110, 1999; and J. L. Sullivan and IE. Transue, “The Psychological Underpinnings of Democracy: A Selective Review of Research on Political Tolerance, Interpersonal Trust, and Social Capital,” Annu. Rev. Psychology. 1999 50:625-650. 22 capital performed consistently better, either in the economic or political sphere, than those without. The reason was simple: social capital facilitates coordination and cooperation for the mutual benefit of the members of the association. Based on the brief review of the definition of social capital from the above three prominent authors, we can roughly capture the essence of social capital: first, social capital is a resource embedded in social relations. Second, social capital can facilitate certain actions for members in the social network. Third, social capital has its moral and instrumental components. It is those norms and obligations in the networks that make up the moral component of social capital. On the other hand, social capital is a resource into which other resources can be invested with an eye on future returns. Fourth, the concept of social capital can be not only applied to individuals but also to society as a whole. Different societies may have different levels or stocks of social capital just like individuals may possess different amounts of social capital. And when we talk about social capital at the societal level, we usually mean those traits and structures, such as voluntary associations, norms of reciprocity and interpersonal trust, inducing people to cooperate to solve collective action problems. The above summaries may raise more questions than they can clarify on the meaning of social capitals. I will in the next chapter discuss in greater detail the theoretical problems affiliated with the concept of social capital. For now, it is enough to conclude that the study of social capital and its influences on human behavior have been emerging as a promising intellectual field despite its infancy. 5 A more comprehensive list of definitions of social capital will be provided later in this chapter. Next chapter will have a more detailed theoretical discussion on the sources, maintenance, and consequences of social capital. 23 Given its immense popularity and wide application in research nowadays, social capital as an analytical tool is no stranger to social science. In the following section, I would like to trace its intellectual and social origins. That way, we will have a firm grasp on this term and know why this idea has been so popular recently. THE INTELLECTUAL AND SOCIAL ORIGINS OF SOCIAL CAPITAL Intellectual Origins of Social Capital As the term “social capital” indicates, based on the above brief discussion, it is a social form of capital (resource). Therefore, it is reasonable to believe its origins are both sociological and economical. The following review will focus on the main sources that generate social capital. The study of the relationship between humans and society has long concerned modern sociologists. Based on different assumptions, sociology is divided into two schools: holistic and individualistic. The holistic view sees society as an independent subject, having its own characteristics and always preceding its members in terms of causality. One of the dominant sociologists in this school is Durkheim. For him, society as a whole does not equal the sum of its members, and “the group thinks, feels and acts entirely differently from the way its members would if they were isolated. If therefore we begin by studying these members separately, we will understand nothing about what is taking place in the group” (Durkheim, 1982:129). This kind of structural perspective also tends to give precedence to the whole [society] in understanding the action of an actor 24 who has been socialized within that whole and whose values depend on the roles that he or she plays there. For instance, social facts, as Durkheim put it, “consist of manners of acting, thinking, and feeling, external to the individual, which are invested with a coercive power by virtue of which they exercise control over him” (Durkheim, 1982252). Following this line of thinking “value introjection”6 plays a critical role in conditioning people’s contractual relations. The group rituals, such as values, moral imperatives, and commitments, act as a guarantor of social or economic exchange. Hence, the expectation of repayment from the other, for instance, is not based on knowledge of the actors involved, but on the embeddedness of both actors in a common social structure. Another theoretical tradition within this holistic sociology, which also exerts influence on the emergence of social capital, is Marx’s analysis of “class consciousness”. According to Marx, industrial revolution compels the proletariat to form a mass (or class) and become more and more concentrated. In opposition to the bourgeoisie, the workers founded “permanent associations in order to make provision beforehand for these revolts” (Marx & Engels, [1848] 1973:76). This identification with each other in the same group forms the so-called “bounded solidarity”, in which members of the same group can then appropriate such dispositions to their desired goals.7 As to the other school of sociology, they focus on individuals’ social actions, e. g., what people do, how they behave individually and collectively in social contexts. Weber, for example, holds that it is possible to deduce human action not so much from the 6 The term “value introjection” was introduced by sociologist Alejandro Portes and it refers to the idea that values, moral imperatives, and commitments precede contractual relations and inform individual goals other than the strictly instrumental. For details, see Woolcock (1998:161). 7 For example, this “bounded solidarity” can be the source of social capital, leading wealthy members of a church to anonymously endow church schools and hospitals; members of a suppressed nationality to 25 ‘ . ”‘tO' lip-‘5“ »..,. i .943 v.»- ’J . 1- I T’ functions performed by the actor as from the values to which he subscribes (Weber, 1968:4). Thus he argues that “sociology rests upon the study of the action of one or more separate individuals and must consequently adopt a strictly individualist method.”8 Another defender of this approach is Popper, who claims, “all social phenomena, and particularly the functioning of all social institutions, should always be studied as the result of the decisions, actions, attitudes, etc. of individuals. . .one must never be satisfied with explanations of a collective nature” (Popper, 1945198). This “methodological individualism” finds its company in and blends with classical economies, which relies on a very simplified assumption of individual action, i.e., rational choice or utility maximization, and a simple mechanism, i.e., market equilibrium, to derive macro-level phenomena from the aggregation of individual behaviors.9 This intellectual trade between individualistic sociology and economics has several inspirations on the emergence of social capital. First, social exchange theory, based on utilitarian or instrumental assumptions, has emerged and challenged traditional sociological accounts of social order. From classical analysis of social exchange by Homans (1958), Simmel (1964) and Blau (1964) to Coleman’s (Coleman, 1990) “rational choice sociology”, social exchange contains the added element of resource transaction rather than just social interaction, which relies on shared norms and values. Recently this voluntarily join life-threatening military activities in its defense; and industrial proletarians to take part in protest marches or sympathy strikes in support of their fellows (Portes, 1998). ‘ 8 This is quoted from M. Mommsen, “Max Weber’s Political Sociology and His Philosophy of World History,” International Social Science Journal 17:25, 1965. 9 For example, Adam Smith, one of the founders of classical political economy, emphasized the role of individual self-interest and its effects on guaranteeing both economic growth and social order and said, “every individual is continually exerting himself to find out the most advantageous employment for whatever capital he can command. It is his own advantage, indeed, and not that of society, which he has in view. But the study of his own advantage naturally, or rather necessarily leads him to prefer that employment which is most advantageous to the society ” (Bimbaum, 1988:17). 26 line of argument, emphasizing economic rationality in social exchange, has converged with the traditional anthropologist argument, which highlights the significance of relational rationality in social exchange.l0 Hence, transactions, whether balanced or imbalanced, are regarded as “means to maintain and promote social relation, create social credits and social debts, and accumulate social recognition” (Lin, 2001:152). Social capital thus is the gains obtained in social exchanges. A second inspiration for social capital in this tradition is from the literature broadly covering “collective action problem”. Mancur Olson’s classic The Logic of Collective Action marks the first systematic account of the potential problems affiliated with group behavior. An economist, Olson tried to challenge the common wisdom in sociology and political science in which groups of individuals with common interests usually attempt to further those common interests. The thesis of his book can best be summarized by Olson’s introduction in the book: “Indeed, unless the number of individuals in a group is quite small, or unless there is coercion or some other special device to make individuals act in their common interest, rational. self- interested individuals will not act to achieve their common or group interests. ” (Olson, 1971 :2) The reason is that a rational individual who is not excluded from enjoying the '0 Lin suggested two different rationalities involved in social exchange. Those value-oriented considerations, i.e., the calculation of transactional gains and costs in exchanges, are called transactional rationality, and the consideration favoring the maintenance and promotion of the relationship is called relational rationality. He further argued that “while both rationalities are enacted by actors in most exchanges, for a given society at a particular time, institutions favor one rationality over the other, allowing moral judgment on the relative merits of one type of capital (economic or social) over the other” (Lin, 2001 : 151). 27 benefits of a good that is collectively produced will feel a strong impulse to take free rides. If more people take free rides in a group, as more likely in large group, then it will make all those who value the good of collective action worse off. Olson’s work implies the difficulty and fragility of cooperation in human societies. This widely praised and much-discussed book stimulated a great amount of advanced works on the nature of cooperation. For example, Robert Axelrod investigates the evolution of cooperation among individuals who pursue their own self-interest without the aid of a central authority (Axelrod, 1984). Elinor Ostrom examines whether and how the exploration of common pool resources can be organized in a way that avoids both excessive consumption and administrative cost (Ostrom, 1990). Francis F ukuyarna, based on the evidences of contemporary evolutionary biology, argues against the classical liberal accounts of “state of nature”ll and emphasizes that cooperation is not so difficult as rational choice theory expects. Rather it has a natural or genetic basis (F ukuyarna, 1999). Most of the themes mentioned above are closely related with yet another school of literature: the study of trust. Trust is believed to be able to enhance the operation of social exchange or to solve the problems associated with collective action. For instance, Trust can lower the transaction costs by eliminating unnecessary contracts. It is trust, again, that makes reciprocal exchange possible. For great uncertainty exists among people involved in social exchange lacking trust; one knows not whether the other will reciprocate in the future. Furthermore, it is trust that is said to cure collective action problems. If actors in the collective movement trust each other to do their part then all of will be better off by ” For instance, Hobbes, Locke, and Rousseau all assumed individuals in the state of nature have no natural inclination for society, and cooperate only as a means of achieving their individual ends (Fukuyama, 1999: 166). 28 attaining their goals; if every citizen trusts their fellows to pay taxes then no one will be cheating and no monitoring cost is needed by government. Trust is a lubricant of cooperation (Dasgupta 1988:49); trust is the emotional basis of cooperation (Barbalet 1996:77). There is yet another theoretical development in the early 19708 that has impacted the latter social capital literature. “Social network analysis” or “structural analysis”'2 focuses on relationships among social entities and on the patterns and implications of these relationships. This approach shares four main assumptions about social actors, their relations and the resulting structure. First, actors and their actions are viewed as interdependent rather than autonomous units. Second, relational ties between actors are channels for transfer or “flow” of resources. Third, the network structural environment provides opportunities for or constraints on individual action. Lastly, network structure is regarded as preserving patterns of relations among actors (Wasserman & Faust, 1994:6- 7). Researchers in this field have recently developed the so-called “social resource theory”, aiming to relate “tie strength” and the provision of resources to self-interest oriented actions. In other words, actors in the social network are regarded as “work entrepreneurs” instead of passive players, exploiting their network position to further their own interests (Lin, 1986; Granovetter, 1982). Related to social network theory and the above mentioned exchange theories, there has been an effort to integrate the “exchange theories” advanced in economics and sociology. Scholars involved in the “new economic sociology” focus on how the relevant combinations of social ties and resources are assembled to meet a given economic '2 Social network analysis is sometimes called structural analysis by sociologists. This structural analysis is 29 challenge. For example, some of these works address the economic success of East Asia and point to the importance of dense interpersonal or interorganizational networks in business.13 In brief, this field emphasizes the embeddedness of economic actions.'4 The concept of “capital” derived from economics also sheds light on the meaning of social capital. Although this term is widely used in the modern world, its meaning is not clear even among economists. Originally the idea of capital referred to money put to use as a loan apart from the interest on that loan. The meaning of capital became blurred when economists in the eighteenth century extended it to in not only financial resources but any resources that can be used to generate future values (Albrow, 1999215). This ambiguity is further reflected in the distinction between “capital” and “capital goods”. The standard economics textbook refers to “capital” as having two different meanings. First, “capital goods” refers to machines, buildings and other specific means of production. Capital goods are designed to help produce other goods. Second, “capital” refers to the money and other financial assets used to buy and sell capital goods (Stiglitz, 1993:18; Black, 1997:47). Despite the distinction of two forms of capital, economists study capital within a strict theoretical framework, including other concepts, such as rates of return, opportunity costs, and funding of investments. However, the extended usage (or conceptual stretch) of capital in other fields does not follow such stringent fashion and therefore more easily causes conceptual ambiguity. The social capital coined by sociologists is a good example. Serious consideration by sociologists to use the concept of capital in their research not the same as the structuralism used in anthropology. See Johanson (1994:114). ‘3 For example, see Orru, Biggart & Hamilton (1997) and Greenhalgh (1988). " For the idea of “embeddedness” and its significance in the theory, see Granovetter (1985). 30 began in the 19608. Becker’s human capital theory may be one of the most influential works in this respect. He explained how increases in human capital fuel economic development through the growth of scientific and technical knowledge (Becker, 1964). Sociologists then borrowed the term human capital to refer to personal characteristics that have value in labor markets, with less attention to specifying the structure of investment or depreciation (Baron & Harman, 1994:1123). Ever since human capital appeared, many other types of “capital”, e.g., linguistic capital, cultural capital, social capital, consumption capital and self-command capital, have also been invented and become new interests in the discipline. However, most of these “new” types of capital proposed above suffer from lack of theoretical precision and other related issues.'5 The conception of social capital also has its roots in political science. In the early 605, Almond and Verba’s The Civic Culture was a landmark study of the “soft” factors of political life, such as values, beliefs, and their effects on the stability of democracy in five countries (Almond & Verba, 1963).'6 They found citizens in stable democracies, i.e., United States and Great Britain, showed consistently higher levels of civic competence, interpersonal trust, political efficacy and reported more pride in their political institutions than their counterparts in three other countries with shorter continuous democratic rules, i.e., West Germany, Italy, and Mexico. Their work reaffirmed with more advanced empirical methods cultural theses proposed by ancient political philosophers, such as '5 As to the criticism of “the plethora of capitals”, please see Baron and Hannan (1994). '6 This book earned landmark status due to the empirical methods it employed, such as survey and statistical techniques, which were not generally available to scholars before. As to the modern culturalist approach on the study of economic and political development, Max Weber and Alexis de Tocqueville are the two forerunners. Weber’s The Protestant Ethic and the Spirit of Capitalism ([1905] 2000) explored the linkage between Protestant Ethic and the emergence of modern capitalism, implying the appropriate cultural attitudes and beliefs are the necessary conditions for economic development. Alexis de Tocqueville, on the other hand, attributed the maintenance of stable democracy in the United States to the “manners” (habit of 31 £31 m: l1? thl‘ r .. 9‘: v...‘ \n. Plato and Aristotle”. This political-culture approach,‘8 stressing socio-psychological attributes of citizens in sustaining democratic stability, can be seen as an attempt to seek resolutions to political collective action problems. Latter works such as Putnam’s Making Democracy Work, which stresses civic virtue as a social capital in explaining both economic and institutional/political performances, are really an extension of this approach.19 In this section I mentioned the intellectual origins of the idea of social capital. In sociology, we see social norms, study of trust, social exchange theory and network analysis; in economics, we see transaction cost theory, logic of collective action, and the concept of capital; and in political science, we see the political culture approach, civil society, and search for the stability of democracy. The interplay of these intellectual currents brings up contemporary social capital literature. In addition to intellectual inspiration, social capital also arises as a response to the heart) of the people (Tocqueville, 1968). '7 Almond and Verba appreciated Aristotle’s mixed-govemment model and his emphasis on several attitudes conducive to sustaining stability, such moderation, interpersonal trust, and even a certain diffidence regarding political participation. Their empirical evidences derived from the five countries supported these arguments (Almond & Verba, 1989:17). '8 Almond and Verba do not consider political culture a theory. Rather, “it refers to a set of variables which may be used in the construction of theories” (Almond & Verba, 1989:26). '9 Actually, in part 111 of The Civic Culture, Almond and Verba linked civic culture with social and psychological conditions and argued that an individual’s nonpolitical attitudes and nonpolitical affiliations, i.e., the attitudes toward his social and interpersonal environment, play an important role in determining his view of the specific political nuances of his environment (Almond and Verba, 1963:261-262, 300). Thus, for example, they argued “people who frequently engage in group activities and who place a high value on outgoing character qualities would also tend to place a high value on outgoing character qualities would also tend to view the human environment as safe and responsive” (lbid: 266). Similarly, what they called ‘civic cooperation’—the propensity to work with others in an attempt to influence the government — is also influenced by one’s view of the social environment, mainly by the level of interpersonal trust (lbid: 269- 273). In addition, in chapter eleven they concentrated on the roles of one particular aspect of social structures, i.e., secondary or voluntary associations, and demonstrated that voluntary association membership affects an individual’s political attitudes and political competence. Putnam’s works on social capital in Italy and the United States should be seen as the elaboration of Almond and Verba’s pioneering work, although they did not invent the concept of social capital to refer to the social relations and structures that form civic culture. 32 social and political trends in contemporary western societies. The following section is devoted to the brief discussion of these trends. Social Origins of Social Capital Some unique features of contemporary society render the salience of social capital. Our world has become extremely interdependent given the division of labor and development of technology. Thus cooperation among individuals, communities, or nations is desperately needed for the sake of all involved. However, as our dependence on others’ cooperation increases, so does our vulnerability to the possible failures of others’ to fulfill their expected roles. Although we are in need of cooperation to improve our collective well-being, we as rational individuals sometimes face social dilemmas, which produce collective outcomes undesirable to the whole group. How can we find the solutions that can help rational individuals transcend collective action problems on the one hand and secure collective outcomes that they prefer on the other hand? It is not surprising that the idea of social capital, regarded as a prominent solution to collective action problems, has received much attention in recent years. The second issue stimulating the research of social capital is related to the first. Since the 19605, the West has experienced a series of social movements aiming to free individuals from the constraints of many traditional social norms and social rules (Fukuyama, 1999:13). This dominant culture of “unbridled individualism” not only caused the loss of shared values but also destroyed communities. Consequently, we saw “[the] grth of crimes, broken families, parents failing to fulfill obligations to children, 33 neighbors not looking out for each other, and citizens opting out of public life” (Fukuyama, 1999:15). Responding to the above mentioned culture and its problems, economists and economic historians have together contributed to the rise of the “new institutionalism”, in which they argue that “persons facing collective dilemmas might prefer to live in a world in which the freedom to choose is constrained.”20 For institutions, especially with centralized mechanisms, will force individuals to take into account the social costs and benefits of their private actions. In other words, the institutions serve to modify the incentives of rational individuals by imposing sanctions and coercions so their self- interests and the collective welfare converge (Bates, 1988:392,396). But this relatively 9, “undersocialize account of collective action problems ignores the role of social relations in solving collective dilemmas.2| The uncertainty of others’ preferences, the structure of social interactions, and communities are factors to be further explored to make our understanding of the relationship between individual and collective outcomes more complete. The study of trust, personal ties, and social norms, which have been discussed often in recent social capital literature reflect exactly this trend. The general historical movement of the Western societies, especially the United States and Great Britain, in the last quarter of the twentieth century experienced a cyclical change, in which “state”, “market”, and “society” have each struggled to dominate the others. These changes also impact the study of social capital. Specifically, the failure of 2° Note here that the new institutionalism still holds the assumption of utility-maximization of individuals. 2' Actually, sociologists are probably most skeptical of the free-rider issue concept suggested by economists. For example, groups in which social networks are closed often face the problem opposite to flee rider, i.e., excess zeal, see Coleman (1990). In addition, institutions alone can’t stem force or fiaud. Some degree of morality, especially trust, must be assumed to operate the institutions. See Granovetter, 34 the “welfare state” in the 19705 resulted in the rise of new conservatism in the 19805, advocating classical conservative remedies based on the deregulated market. Mediating structures (voluntary associations or third sectors) were thought to be able to deliver or replace public programs to address social problems (Edward & Foley, 1998; Zijderveld, 1998; Couto & Gutheir, 1999; Hall, 1999). This resurgence of “civil society” coupled with late 805’ pro-democratic movements around the world paved the way for finding and treating social capital as both a remedy for social problems and a safeguard to democracy.22 Even the World Bank has paid attention to the notion of social capital. In October 1996, the Social Capital Initiative (SCI) was established in the Social Development Department of the World Bank to promote and strengthen social capital, with special emphasis on how social capital can help developing countries solve social and economic problems.23 In sum, the conception of social capital can not only be traced back to its intellectual origins but also has its roots and policy implications in contemporary social contexts. That is why we have seen more and more scholars from different fields addressing their own questions by applying the common framework upon which social capital is centered. The following section will briefly review recent social capital (1985: 489). 22 ‘Civil society’ became a fashionable subject after the collapse of communism in Eastern Europe because many scholars have recognized that post-communist societies were characterized by a particular deficit of social structures that were a necessary precondition of stable democratic political institutions. On the other hand, Hungary, Poland, and the Czech Republic are the post-communist countries that appear to have greater chances than others to become successful democracies: their retained civil societies throughout the communist period are thought to be the key factor (Fukuyama, 1995a: 8; 1995b: 361). 23 Ever since Putnam’s Making Democracy Work, a new school of thoughts has been blending the civil society thesis and the civic culture approach. Some called this school the “neo-Tocquevillean tradition”. See Edwards & Foley (1998). Recently, the World Bank has encouraged the academic study of social capital and urged the formation of social capital in developing countries to solve their social and economic problems. For a detailed information on the World Bank’s position on social capital, please check their website at www.worldbank.org/poverty/special/whatsc.htm. 35 literature with special concern for definitions, the first step for anyone concerned with the study of social capital. THE DEFINITIONS OF SOCIAL CAPITAL: SOCIOLOGISTS vs. POLITICAL SCIENTISTS The contemporary use of social capital first appeared in the study of community. For example, Hanifan argued that “social capital...refers to ...those tangible assets [that] count for most in the daily lives of people: namely good will, fellowship, sympathy, and social intercourse among the individuals and families who make up a social unit” (Hanifan, 1920278). Jacobs’ study of community pointed out the decisive role played by the networks of strong, crosscutting personal relationships developed over time in generating the basis for trust, cooperation, and collective action in communities (Jacobs, 1965). Others later employed this term to show the significance of resources embedded in social structures in shaping individuals’ success in society (Loury, 1977; Coleman, 1988). Therefore we could say that ever since the emergence of the concept of “social capital” its definition has been ambiguous. This confusion and imprecision stems mainly from its appeal to scholars in different fields. The views of different authors on the concept of social capital appear to depend on their disciplinary background and the research questions they would like to address. What exactly is social capital? How many definitions of social capital can we find? There is likely no consensus in social capital literature. It is therefore necessary to survey the definition of social capital and glean any similarities or patterns. That way we 36 will have a general idea about what various authors had in mind when they spoke of social capital. From the limited sample I have, table 2-1 shows all their definitions on social capital. 37 TABLE 2-1 A Sample of the Definitions of Social Capital Authors Definitions Adler & Kwon Albrow Baker Boix & Posner “a resource for individual and collective actors created by the configuration and content of the network of their more or less durable social relations” (2000:93). “the institutions which provide the infrastructure of social order, community organization and reliability in social relations on which future value depends” (1999:17). “a resource that actors derive from specific social structures and then use to pursue their interests; it is created by changes in the relationship among actors” (19902619). “a set of institutionalized expectations that other social actors will reciprocate co-operative overtures. This expectation generates cooperation by making otherwise uncooperative actors willing to undertake those overtures in the first place” (1998:686). 38 TABLE 2-1 (continued) Authors Definitions Bourdieu Boxman, et a1. Brehm & Rahn Briggs “the aggregate of the actual or potential resources which are linked to possession of a durable network of more or less institutionalized relationships of mutual acquaintance or recognition” (1986:248-249). “the number of people who can be expected to provide support and the resources those people have at their disposal” (1991: 52). “the web of cooperative relationships between citizens that facilitates resolution of collective action problems” (1997:999). “resources stored in human relationships, whether casual or close...As a resource for action, social capital works at various levels: family, neighborhood, city, and society” (1997:] 12). 39 TABLE 2-1 (continued) Authors Definitions Burt Coleman “the relationship with other players... through whom you receive opportunities to use your financial and human capital. . . [It] is the final arbiter of competitive success [under imperfect competition)” (1992:8-1 0). “a quality created between people [it] predicts that returns to intelligence, education, and seniority depend in some part on a person’s location in the social structure of a market or hierarchy” (1997:339). “[social capital is] a variety of different entities having two characteristics in common: they all consist of some aspect of social structures, and they facilitate certain actions of actors—whether personal or corporate actors —within the structures. . .Unlike other “forms of capital, social capital inheres in the structure of relations between actors. It is not lodged either in the actors themselves or in physical implements of production” (1988:98). 40 TABLE 2-1 (continued) Authors Definitions Couto & Guthrie “[social capital] expresses the moral resources of trust and cooperation” ( 1999:64). Foster-Fishman & Frank “the processes of building social relations, the accessing of resources within those social relations, and the use of these resources and relations for specific benefits or outcomes” (20002). F ukuyama “the ability of people to work together for common purpose in groups and organizations” (1995:10). “a set of informal values or norms shared among members of a group that permits cooperation among them” (1999:16). “the ability to work together in groups” (1999: 174). Hall “the propensity of individuals to associate together on a 41 TABLE 2-1 (continued) Authors Definitions Lin Loury Montgomery Nahapiet & Ghoshal regular basis, to trust on another, and to engage in community affairs” (1999: 41 7). “At the core of the conventional definition is membership in voluntary association” (19992420). “investment in social relations by individuals through which they gain access to embedded resources to enhance expected returns of instrumental or expressive actions” (1999:39). “naturally occurring social relationships among persons which promote or assist the acquisition of skills and traits valued in the marketplace” (1992:100). “the cumulative capacity of social groups to cooperate and work together for a common good” (1997: 30). “the sum of the actual and potential resources embedded 42 TABLE 2-1 (continued) Authors Definitions Paxton Portes Putnam within, available through, and derived from the network of relationship possessed by an individual or social unit” (Nahapiet & Ghoshal, 2000:121-122). “Social capital involves two components: trust and associations” (1999:97). “the ability of actors to secure benefits by virtue of membership in social networks or other social structures” (2000248). “features of social organization, such as trust, norms, and networks, that can improve the efficiency of society by facilitating coordinated actions”(1993b: 167). “Most forms of social capital, such as trust, are moral resource” (1993b: 169). “features of social life—networks, norms, and trust—that enable participants to act together more effectively to 43 TABLE 2-1 (continued) Authors Definitions pursue shared objectives...Social capital, in short, refers to social connections and the attendant norms and trust” (1995b: 664-665). Putterman “an individual’s investment in social tie” (1995:7). Rose “the stock of formal or informal social networks that individuals use to produce or allocate goods and services” (199823) Uslaner “a system of values, especially social trust” (1999a: 122). “a set of ‘moral resources’ that leads to increased cooperation” (1999b: 2] 5). Sidney “relationships among persons, groups, and communities that engender trust and/or mutual obligations” (1996: 7) 44 TABLE 2-1 (continued) Authors Definitions Smith, Harlan “[That] consists of those operative and effective social norms that facilitate social trust and cooperation for the common good” (1999: 1 74). Stolle & Rochon “the networks and norms that link citizens to each other and that enable them to pursue their common objectives more effectively” (1 998 :47). Pye “networking and learning to work together on the basis of trust” (1999:764). Walker, Kogut & Shan “a means of enforcing norms of behavior among of individual or corporate actors and thus acts as a constraint, as well as a resource” (Walker, Kogut & Shan, 2000:228). Whiteley & Seyd “social capital refers to citizens’ trust in other people” (1997:125). 45 TABLE 2-1 (continued) Authors Definitions Woolcock “the information, trust, and norms of reciprocity inhering in one’s social networks”(1998:153). “the nature and extent of a community’s personal and institutional relationships” (1998: 1 82). World Bank “the norms and social relations embedded in the social structures of societies that enable to coordinate action to achieve desired goals.”24 From these various definitions of social capital we can sum up their major similarities and differences. There are two major similarities among these definitions: first, they imply the importance of stable and reliable social relations in conditioning either individual or collective actions. Second, most of the definitions regard social capital as a resource and focus on its positive consequences although it has its “downside”.25 2" This definition is quoted from the World Bank’s website: http//www.worldbank.org/poverty/scapital. 2’ Portes and Landolt remind us that social capital has at least four negative consequences: exclusion of outsiders, excess claims on group members, restrictions on individual freedoms, and downward leveling norms. For details, see Portes & Landolt (1996). I will also address this issue in greater detail in the next chapter. 46 Concerning the differences, we see two major camps attempting to use the concept of social capital in dealing with very different issues. One camp, mostly sociologists, tends to regard social capital as a resource embedded in social relations or ”social structures. The amount of social capital one may possess depends on the other members’ resources in one’s personal networks. And the forms (values) of social capital will shift according to the changes of sociopolitical and economic context”. Also, the norms or values held by individuals become social capital only if they facilitate others’ actions. Outside the group or network, these norms or values may be of little value. Therefore, social capital is context-dependent. In addition, most of the sociologists highlight the instrumental dimension of social capital. Social capital is a rational investment in an individual’s social relations with the aim of fulfilling their personal goals, regardless of whether they be harmful or beneficial to others or aggregate actors as a whole. The other camp, consisting mostly of political scientists and development theorists, also shows tremendous interest in the concept and roles of social capital. But political scientists hold their own views about social capital and the scope of its functions.27 Most of them regard social capital as a resource to solve collective action problems or developmental problems in general instead of regarding it as a resource that individuals can appropriate for their own interests. Accordingly, they focus on subjects such as norms of reciprocity, trust (especially interpersonal trust), and civic associations 2° The value of a specific form of capital, e.g., financial, social, cultural, or human, is determined by the socioeconomic and political context it occupies. For various examples on this account, see Edward & Foley (1998:129-130). Fukuyama also expressed a similar opinion about the contingent values of social capital (1999:18-19). 27 Putnam is believed to be the leading figure in introducing and adopting the social capital framework into 47 (voluntary associations) and their functions in bringing about spontaneous cooperation. They further claim this positive effect will overcome the dilemma of collective action or lower transaction costs and hence produce desired socioeconomic development and stable democracy (Putnam, 1993b; Fukuyama, 1995&1999; Evans, 1996; Knack & Keefer, 1997; Newton, 1997; Paxton, 1999; Pye, 1999; Rose, 1995; Stolle & Rochon, 1998; Sullivan & Transue, 1999; Tendler & Freedheim, 1994; Whiteley & Seyd, 1997). Thus, social capital seems to become a context independent factor in political scientists’ eyes. Those norms, values, and the ability to associate, which can facilitate collective actions rather than individual actions, become a capital (resource) for the aggregate actors, whether it be groups, organizations, communities, regions, or nations. Contrary to this value-added approach, sociologists like Coleman and Bourdieu treat social capital as a social relational and structural resource not necessarily beneficial to society as a whole.28 CONCLUSION In this chapter, I have addressed the meaning and origin of social capital, both intellectually and socially. The popular appeal of this concept indicates the limits of economic models, especially rational choice theory, in analyzing and explaining social and political life. However, being a complementary approach, social capital analytical framework is still in its infancy. From the aspect of theoretical construction, the study of the study of political science. In Making Democracy Work, Putnam used sociologist James Coleman’s idea of social capital as a basis and highlighted its role in inducing social cooperation (Putnam, 1993b, 167). 2” Sociologists, e.g., Coleman and Bourdieu, treat social capital as a morally or ethically neutral resource. Everyone can appropriate social capital to any personal goals, whether beneficial or harmful. While Putnam and other neo-Tocquevilleans have stretched the original concept of social capital to represent socially 48 social capital must be furthered greatly. On the whole, we face a divide in the study of social capital. More communication among scholars from different disciplinary backgrounds is urgently needed. Further, each camp mentioned above faces problems. Although sociologists’ view on social capital is more theoretically refined, they are still not clear about the actual structure of social capital. For example, some regard social networks as an access or a bridge to specific resources, and these resources available to any actor are contingent on the resources available to other individuals socially proximate to the actor. Others see the social networks themselves as a form of social capital. They contend that network range, indicated by size, is the primary measure (Burt, 1992:] 1-12). Still others see social relations as social capital. It is the optimal combination of “embeddedness” and “autonomy” of the social relations that creates social capital (Woolcock, 1998:182). As to who owns social capital, some (Coleman, 1990; Burt, 1992) say social capital is owned jointly in a relationship, while others (Portes, 1996) regard social capital as an individual resource roughly analogous to other individual assets. Political scientists face even more confusions in applying this concept to their work. Some leaders in the field are ambiguous about social capital itself and its sources. Putnam, for example, thinks social trust, norms of reciprocity, and networks of civic engagement are all components of social capital. The relationship among them is that civic engagement and norms of reciprocity foster social trust, and civic engagement fosters robust norms of reciprocity (1993b, 171 &173). Given that social trust is the core component of Putnam’s social capital, other components such as networks of civic engagement are actually the generator (source) of social capital instead of social capital desired norms and values. See Greeley (1997) and Edwards & Foley (1998). 49 itself. Similarly, Fukuyama regards cooperative norms like honesty and reciprocity as social capital. While he also states “[The] norms that produce social capital...must substantively include virtues like truth telling, the meeting of obligations, and reciprocity” (Fukuyama, 1999:16-17). Therefore, we cannot discern whether the cooperative norms are sources of social capital or social capital itself. If the former is true, than what is social capital? 1f the latter is true, we end up wondering where social capital comes from. In the next chapter, I will review social capital literature in greater detail, concentrating on the conceptual understanding of social capital. In synthesizing current literature I will also propose two general analytical frameworks for studying private and public social capital respectively. 50 CHAPTER 3 ANALYTICAL F RAMEWORKS FOR STUDYING SOCIAL CAPITAL INTRODUCTION In chapter two we have seen that there are two camps of social capital literature nowadays. One focuses on the individual’s social relations (personal networks) and their ability to provide necessary resources for goals achievement. The other centers on the norms and values within certain societies and their power to solve collective action dilemmas and hence make the society better as a whole. This chapter aims to synthesize literature from each camp and develop an analytical framework from which we can not only capture the general picture of social capital but also develop hypotheses to be tested. Since two basic viewpoints on social capital exist, it is necessary for the sake of theoretical precision to either: strive to integrate them under a general framework and find some mechanisms to bridge them; or to treat them as two different forms of social capital literature and therefore give them different labels. The first task would be more difficult than the second. First, we don’t know whether the mechanism really exists or not.1 Second, no one to the best of my knowledge has ever tried attempted this. Perhaps Michael Woolcock’s work can be considered the most similar to this challenge. From the perspective of economic sociology, he employs two major concepts, embeddedness and autonomy, to form his analytical framework, which can be applied both at the micro and ' For instance, sociologist Alejandro Portes holds the belief that the theory of social capital should be limited to cases at the individual level because there is less care and theoretical refinement in redefining social capital as a structural property of large aggregates (Portes, 2000: 62). 51 macro levels of analysis in economic development. By ‘embeddedness’ he means social ties that are integrated within social actors at the micro level or synergies between state and society at the macro level. ‘Autonomy’ refers to extra-community networks at the micro level and institutional competence or organizational integrity at the macro level. It is the combination of these two forms of social capital2 that determines the development outcomes (Woolcock, 1998). Strictly speaking, Woolcock’s fiamework falls under the sociological view, i.e., social-relational attribute of social capital. He did not address much pertaining to the social-psychological attributes of social capital in which political scientists are interested. Thus, we still lack the appropriate bridge between the two camps. I take the other approach when dealing with this problem: treating the two camps as two distinct social capital literatures. One the one hand, as I mentioned in the first chapter that most sociological views of social capital are focused on the instrumental dimension of social capital and are center at micro level analysis. Social capital is thus regarded as rational investment in an individual’s social relations with the aim to fulfill their personal goals, whatever they may be. Therefore, I will refer to this current of social capital as private social capital. On the other hand, those norms, values, and networks of civic engagement cherished by political scientists and development theorists in helping solve collective action dilemma will be called public social capital here because it is believed to be a capital (resource) for the aggregate actors, be it groups, organizations, communities, regions, or nations. This way we will not be confused with the different forms of social capital when engaging social capital literature. 2 Woolcock uses social capital to refer to at least two different entities in different contexts. For instance, he treats embeddedness and autonomy as two key forms of social capital (1998:185), while he also implies that social capital consists of two distinct social relations (19982180). 52 STUDYING PRIVATE SOCIAL CAPITAL: A THEORETICAL PERSPECTIVE In this section, I will synthesize the literature regarding social relations as an individual’s resource and propose a general analytical framework for studying it. I will attempt to bring the pieces of social capital research together and organize this section following an organic approach. The nature, sources, mechanisms, and outcomes of social capital will be discussed in order.3 In addition, social capital in this section will refer exclusively to private social capital unless otherwise defined. The Nature and Sources of Private Social Capital Private social capital is the resources embedded in an individual’s social relations. In order to benefit from these resources an individual must be a member of a certain group or different social structure or have the ability to reach members of other groups or structures.‘ This is the structural dimension of social capital. A child, for example, being a 3 In criticizing Putnam’s Making Democracy Work, Levi insists “we need a more complete theory of the origins, maintenance, transformation, and effects of social capital” (1996:52). Newton also emphasized for in-depth inquiry into the nature, causes, and consequences of social capital (19972584). Their critiques of social capital literature inspired me to organize my review as their demands. 4 My use of private social capital is based primarily on Alejandro Portes’ (2000) definition of social capital. Nonetheless, in Portes’ definition, i.e., the ability of actors to secure benefits by virtue of membership in social networks or other social structures, he seemed to imply each actor is capable of identifying resources in his networks when in need. In other words, he holds the individual’s ability to recognize and appropriate potential resources embedded in the social relations as constant. I do not think the ability to recognize potential resources in the social networks is constant across actors. Some people are easier than others to recognize potential resources to them and expand their social networks. That is why I add individual ’5 endowment as another condition to benefit social capital. 53 member of a family, can normally enjoy the unconditional love, care, and financial support of his or her parents that would otherwise not be available to him. Or for an immigrant planning to start a new business, being a member of his or her ethnic group (community) can win the help-ranging from getting considerable economic resources, tips about business opportunities, to access to markets, and a pliant and disciplined labor force that are not available outside the community due to the exclusion from the mainstream financial and civic institutions- of fellows of the same ethnic identity (Woolcock, 1998:173-174; Portes, 2000:54). In addition to the inherent or naturally-occurring relations ascribed above and their related social resources, we often observe other types of activities that are not totally dependent on the ascribed relations. For instance, job-seeking activities sometimes need more than just relatives’ help. The ability to reach someone with higher social status, assumed to hold more social resources’, becomes an important factor both in determining whether a job can be secured and the level of the job. Studies show that there is a strong relationship between social resources of the respondents, as indexed by contact status, and their attained job statuses“. This structural aspect of social capital generates two complimentary arguments: first, from the perspective of embeddedness,7 an individual’s initial (inherent) social position, e.g. gender, race, and family background, provides certain types of social resources that he or she may appropriate. The level and types of 5 Social resources here refer to those resources accessible through one’s direct and indirect social ties. See Lin, Vaughn, and Ensel, 1981). 6 Lin, Vaughn, and Ensel researched how people find their first and last jobs among those who used personal contacts. The zero-order correlation between contact status and the attained status of their firstjob was 0.65, and between contact status and the last job status the correlation was 0.68. See Lin (1982:140). 7 In fact, most social capital literature (Bourdieu, 1986, Coleman, 1988, 1990; Granovetter, 1982,1985) views the resources engendered by social capital from the lens of embeddedness. 54 social resources vary across different social positions. The flip side of this thinking is that it is exactly the initial social position which may constrain an individual’s choices when faced with decisions, especially those related to goal attainment. Second, in order to solve the dilemma that embeddedness poses we must allow the autonomy of an individual to play a role in goal achievement. Given an initial social position, an individual could also try to break the boundary of social structure and reach to other domains of the society that are initially inaccessible. That is exactly what the strength of tie argument states. According to Granovetter, weak ties“, far from producing alienation, are vital for an individual’s integration into modern society. Furthermore, weak ties9 have a special role in individuals’ potential mobility since: “[there is a] structural tendency for those to whom one is only weakly tied to have better access to job information one does not already have. Acquaintances, as compared to close friends, are more prone to move in circles different from one’s own. Those to whom one is closest are likely to have the greatest overlap in contact with those one already knows, so that the information to which they are privy is likely to be much the same as that which one already has.” (Granovetter, 1974:52-53) In other words, social mobility is possible through the aid of an individual’s personal networks. The realization of this mobility will ultimately depend on a social actor’s ability to recognize the person (contacts) he or she needs and the way to access them. This ability varies from person to person. Of course, this ability is also constrained, 8 ‘Weak ties’ refers to our acquaintances as opposed to our close friends and kin. See Granovetter (1982). 9 In fact, it means weak ties that function to bridge. Specifically, weak ties do not necessarily positively affect instrumental activities; only bridging weak ties, which help one connect to social circles different 55 to some degree, by the given social networks each individual occupies.lo In brief, the sources of private social capital are embedded in the social relations each individual possesses. The nature of private social capital involves at least three related issues. First, a person’s social relations include not only inherent relations, e.g., family and kinship, but also those achieved social relations. Second, the ability to recognize the location of resources embedded in those relations differs across individuals. Third, the ability to access resources via given connections can not only fulfill personal goals but also provide opportunities to reshape the given social structure. Mechanisms of Private Social Capital Given the social resources embedded in individuals’ personal networks, the next logical question concerning private social capital should be how these resources may be appropriated by the actors involved in the networks. In other words, what are the mechanisms that make social capital real? I believe this is where norms and values, especially norms of reciprocity, enter the discussion of social capital. Scholars in social capital literature all recognize the importance of norms and values in the making of social capital, but there is no discussion on where we should put social norms in the theory of social capital. So far it seems that most of them treat norms, other rule-like components, or even culture either as the source of social capital or as the social capital itself. I argue from one’s own, do. See Granovetter (1974; 1982:] 12). '0 The degree of autonomy in a network should not be considered as a constant influence across different structures. The closure of linkage between the actor and his network should impose some constraints on the actor’s choice to expand his social relations. Generally speaking, the more closed the linkage between actor and his network, the more difficult for him to reach further horizons. For a discussion of the interplay 56 that social norms are better treated as the mechanism rather than the source of private social capital. A brief discussion on the role of social norms may be a good starting point. The origins of social norms vary greatly." Some are spontaneously generated and implicit, e.g., incest taboo, and some are forcefully imposed and clear, e.g., formal rules. Despite the different origins, they all serve a very important function: that of maintaining social order. Without norms, formal or informal, we would live in a situation where uncertainty predominates human social interactions, which accordingly leads to infinite conflicts and chaos. Douglass North, from an economist’s perspective, argues that norms and rules as an “institution” contribute to the lowering of transaction cost in human social interaction (North, 1990). Similarly, Granovetter contends that social norms (or generalized morality) are vital to the function of formal institutional arrangements because these implicit agreements assure our actions will have predictable consequences (Granovetter, 1985). I have already shown social capital originates from social relations (or social networks). Social capital can become a type of resource only if both parties adhere to the existing rules, either formal or informal. This is so because social norms set the durable standards for how to behave and what to expect when conducting social exchanges. Therefore social norms should not be considered social capital itself although the relationship between the two is close.” The question, then, is how should we consider the relationship between the two? Here I prefer to think of social norms and rules as the between embeddedness and autonomy, see Woolcock (1998), Portes (2000). ” For a detailed discussion of origins of social norms, please see Fukuyama (1999). '2 About this some social capital theorists do not feel certain. Putnam, for example, thinks of social norms, such as norms of reciprocity, as a component of social capital itself (Putnam, 1993b: 171). Fukuyama holds similar view (Fukuyama, 1999:16). 57 mechanism between social capital and its final products. 0 in economic terms, they can be seen as the ‘capital goods’ aimed to produce final usable products from capital resources, just like machines and factories do for industrial production. Like I mentioned earlier, in Portes’ wbrds, being a member of certain group or different social structure provides potential resources that an individual can appropriate. The possibility that an actor take advantage of these social resources lies in the norms internalized in members of the groups or social structures. This internalization of norms makes people willing to “pay their debts in time, give alms to charity, and obey traffic rules because they feel and obligation to behave in this manner” (Portes, 2000:48). It is this moral dimension of norms that can be appropriated by others as a resource. It can not only generate a feeling of obligation to help others but also impose sanctions on those who violate the rules. Existing social norms are not the only ones that can help facilitate individual’s goals. We can also find norms created for specific purposes. Thus from the perspective of capital resources and capital goods, social capital as a resource can be invested in the creation of new norms to reduce production costs, just as machines do in factories. For example, in a study of learning, Stephen L. Morgan regards the dense social networks among parents as social capital (resources). These resources can then be invested in “capital goods” such as norms of discipline, responsible monitoring, and regularized information exchange through which the goals of fostering citizenship, quality of life, psychological sustenance, extracurricular achievement, and learning among students can be achieved at lower costs. In addition, the norms created among parents can substitute for other capital goods, e.g., textbooks and facilities, when financial capital resources are 58 lacking and still attain the desired goals.l3 Thus, social capital, via the help of capital goods (social norms), can produce positive products to the benefit of individuals who engage in social exchange. Recently, economists of the so-called “new institutionalism” have been also interested in studying norms and rules, but their primary focus is the function of social norms and rules in fulfilling market exchange." While economists begin to recognize the importance of norms, they hold very different views from sociologists on the nature of social exchange. They try to recast social relations in economic terms. Specifically, based on the assumption of individual utility maximization, people act altruistically only because they have calculated and expect to get benefit from such behavior. Following this undersocialized line of thinking, social capital is possible because people’s altruistic actions create a sense of obligation for beneficiaries to reciprocate in the future. This sense of obligation comes from a special form of social norms call “reciprocity”. Actually, the norm of reciprocity is the most important social norm discussed in social capital literature (Putnam, 1993b: 172). In fact, the norm of reciprocity is usually studied together with the problem of cooperation. The logic is clear: given the assumption of rational individuals in economic theory, social order or cooperation is difficult to achieve." In order to explore the best strategy conducive to human cooperation, Robert Axelrod invited game theorists, '3 This is from the abstract of Stephen L. Morgan’s paper “ Social Capital, Capital Goods, and the Production of Learning”, presented at the conference of the Social Capital: Bridging Disciplines, Policies and Communities, Michigan State University, April, 1998. ” For example, Douglass North said “formal rules in even the most developed country, make up a small (although very important) part of the sum of constraints that shape choices. In our daily interactions with others, ..., the governing structure is overwhelmingly defined by codes of conduct, norms of behavior, and conventions” (North, 1990:36). '5 Economists usually cite the Prisoner’s Dilemma game to refer to the situation in which the pursuit of 59 economists, and political scientists to participate in a computer tournament and concluded that “tit for tat”"’ is the most successful strategy: “What accounts for TIT FOR TAT’s robust success is its combination of being nice, retaliatory, forgiving, and clear. Its niceness prevents it from getting into unnecessary trouble. Its retaliation discourages the other side from persisting whenever defection is tried. Its forgiveness helps restore mutual co-operation. And its clarity makes it intelligible to the other player, thereby eliciting long- term cooperation.” (Axelrod: 1984:54) Axelrod’s study highlights the importance of iteration (repeated interaction) in inducing cooperation. Without expected repeated interaction, an individual will likely hesitate to cast his trust in another person whom he does not know and will never see again. Thus the norm of reciprocity will never emerge. Biologists also concur with Axelrod’s findings and add that species, including human beings that experience repeated interactions, have relatively long lives, and have the cognitive capabilities to distinguish cooperators from betrayers based on a host of subtle signals tend to develop norms of reciprocity (Fukuyama, 1999: 172). Besides, reciprocity has two different modes: balanced reciprocity, often seen in market exchange, and generalized reciprocity in social exchange. Balanced reciprocity refers to “a simultaneous exchange if items of equivalent value,” while generalized reciprocity refers to “a continuing relationship of exchange that is at any given time unrequited or imbalanced, but that involves mutual expectations that a benefit granted now should be repaid in the future” (Putnam, 1993b: 172). What social individual self-interest has led to an outcome that is deplored by both. For details, see Laver (1997). '6 The strategy of “tit for tat” is a process in which a player starts by cooperating with an opponent and then does whatever the other player did on the previous move (Axelrod, 1984). 60 capital literature focuses on is generalized reciprocity because it is the bedrock of social exchange and can reconcile self-interest and solidarity in the long run. The two mechanisms indicated above - repeated interaction and social norms /rules — provide opportunities and stable structure conducive to the feeling of security and certainty, which are accordingly the preconditions of appropriating social capital. But the ultimate users are the people involved in social exchange. Thus the individual’s ability to recognize the potential resources embedded in his or her social networks are decisive in determining the actual outcomes of using social capital. A scenario comes to mind in which two people share the same social networks, but due to differing personal endowment and personality, these two may end up with unequal levels of achievement by resorting to their social capital. Specifically, people vary in willingness to invest in their social relations, in the ability to recognize the potential resources available to them, and in the ability to access the actual resources once recognized. In sum, the mechanisms that help transfer social capital resources into actual products (outcomes) consists of three elements: repeated interactions, which is manifested in the forms of network engagement; social norms, which aim to provide stable social order; and personal endowment, which determines the ultimate values of social capital. Through these three mechanisms the potential resources embedded in social relations can possibly be transformed into the actual usable assets. However, the creation of usable assets through social relations may not always benefit social or aggregate actors. This leads us to the last aspect of private social capital analysis: the consequences of private social capital. 61 The Consequences of Private Social Capital 1. Positive Consequences In chapter two I introduced both the intellectual and social origins of social capital study. It is logical to expect that scholars in this field will mostly focus on the positive side of social capital because the term “capital” implies the valuable merit of social relations. There is also a biased tendency for sociologists to view good things as emerging from sociability (Portes, 2000:56). Thus it is not surprising that we are overwhelmed by the many works singing the praise of social capital. These works are done primarily from a social resources perspective. The major theme of this good-outcomes argument can be boiled down to the claim that social capital, through network ties and norms of reciprocity, can lower the transaction costs involved in social exchange and hence increase the efficiency of action. Empirical evidence shows that the social capital present in family and community is said to have a positive impact on reducing the probability of public school students dropping out (Coleman, 1988). Basley, Coate, and Loury showed that informal financial institutions, i.e., rotating savings and credit associations (ROSCAS), based on social relationships among members could provide access to individuals lacking access to credit markets to improve their welfare (Basley, Coate, and Loury, 1992). Scholarly studies of ethnic entrepreneurship also present us with examples of how strong intra-community ties can benefit its members (Woolcock, 1998). Fernandez, Castilla, and Moore’s case study finds employers who take advantage of workers’ social connections in hiring new workers via employee referrals yields 62 significant economic returns (Fernandez, Castilla, and Moore, 2000). In addition, some argue that social capital facilitates the development of intellectual capital'7 by affecting the conditions necessary for exchange and combination to occur (Nahapiet and Ghoshal, 2000: 132). Although individuals or groups can benefit from appropriating social capital, the consequences are not necessarin desirable in the long run or for the community or society as a whole. If social capital is defined as potential resources embedded in social relations, we should also be aware of the potential negative effects that it may incur. More specifically, being a member of a certain group or social structure may help one access certain resources that you otherwise could not. However, one may also face pressures or obligations due to norms of reciprocity, to repay what one is not willing to or what might impede one’s firture plans. Therefore, social capital itself should be regarded as neutral. The positive or negative consequences that social capital engenders depend on whether one sees it from a social resources perspective or a social control one. Level of analysis and different time frames also play a role in dealing with the outcomes of social capital. The following section will elaborate this point. 2. Negative Consequences Most social capital literature sees social relations from the perspective of social resources. However, there are at least four negative consequences, according to Portes (2000), that the so-called social capital may produce. The core argument from which '7‘ Intellectual capital’ here refers to the knowledge and ability of a social collectivity, such as an 63 these negative consequences are deduced is the strength of intra-community or group ties. Generally speaking, the stronger the inner ties among members, the more likely that it will exclude outsiders, encourage free riding, demand conformity from members, and impede the outreaching capability of its members.'8 Thus, once again the neutral nature of social capital is confirmed. In addition, the outcomes of social capital for a given individual or a group are relative to whether we look inward or outward for our level of analysis. The four consequences that Portes proposes are limited to the individual level. If we include the macro level in the analysis, then we tend to observe positive consequences for members in its group and negative consequences for the society as a whole. For example, those groups that are deviant from the mainstream society, e.g., hate groups, cults, mafia, etc., may have a high level of solidarity among members. Active participation in such groups may provide many opportunities for members, but this participation also reduces the possibility for members to expand their social ties to the larger community or society because of the label they carry. Or in a society with high- context culture, people tend to resort to personal connections rather than legal institutions as a way of solving conflicts. This may result in criminals evading punishment due to their connections to high- ranking officials. This application of private social capital serves its desired purpose at the expense of social justice! Just as Coleman said, “[A] given form of social capital that is useful for facilitating certain actions may be useless or harmful for others” (Coleman, 1990:302). Beside the difference in level of analysis, we can also identify the different outcomes of social capital from a structural perspective. It organization, intellectual community, or professional practice. See Nahapiet and Ghoshal (2000: 124). '8 Portes used “exclusion of outsiders”, “excess claims on group members”, “restrictions on individual freedoms”, and “downward leveling norms” respectively to refer to these negative consequences. For 64 is from this perspective that Woolcock tried to fill the theoretical gap between positive and negative consequences of social capital. Borrowed from economic sociology and comparative institutionalism of development, Woolcock characterizes social capital as having two dimensions: embeddedness and autonomy. The concept of “embeddedness” embraces three ideas: first, all forms of social exchange are inherently enmeshed in social relations; second, embeddedness itself could take different forms, i.e., social ties, cultural practices, and political context, all of which work to shape opportunities and constraints which individuals face as they seek economic advancement; third, the many benefits gained by embeddedness in a given networks are not without corresponding costs (Woolcock, 1998:163). Actually, the concept “embeddedness” embodies most of the idea that social capital indicates. However, in order to account for the different outcomes social capital may generate, both positive and negative, the concept of “autonomy” must be introduced into the theory. Otherwise, we will either tend to focus on the positive side of them or fail to explain why some undesired outcomes appear after taking advantage of social relations as resources. The idea of “autonomy” in comparative institutionalism generally refers to the state’s ability to insulate itself from the influence of society (Evans, 1995). Woolcock borrowed this idea to depict, at the micro-level, the individual’s ability to reach extra- community networks or, at the macro-level, institutional coherence, competence, and capacity (Woolcock, 1998:168). Using “intra-community ties” as embeddedness and “linkage” as autonomy, a 2 x 2 table can be made to predict the four ideal-type consequences social capital may produce. It is shown in Figure 3-1. details, see Portes (2000:56-59). 65 Figure 3—1 Michael Woolcock’s Typology of Bottom-Up Dilemmas of Development.” High Anomie Social opportunity LINKAGE (Extra-community networks) Low “Amoral individualism” “Amoral familism” INTEGRATION (Inna-community ties) Sources: Woolcock (1998: l 72). Through Woolcock’s typology we have a more balanced view on the possible consequences of social relations. Furthermore, the development of social capital from this perspective is not a static process. An individual’s social capital may become his “social deficit” once the greater environment changes between two different time frames or when the present goal is achieved.20 An Analytical Framework for Studying Private Social Capital In this section I will present a general framework, based on the literature review of '9 For a detailed description of the four ideal types of consequences of social relations, see Woolcock (1998:171-175). 2° The simplest way to demonstrate this argument is to consider that human sociability begins with family and kinship. Familial attachment bonds family members and works to support each member. While this resource can be a help it will also discourage its members from engaging in more complex social networks when this familial attachment is too strong. If the member is granted a certain degree of autonomy then he can expand his social networks to include extra communities and achieve more successes. Edward Banfield (1958) used “amoral familism” to refer to such phenomenon. Literature of economic sociology also provides many similar cases. 66 the previous sections, for studying private social capital. It is shown in Figure 3-2. 67 Figure 3-2 An Analytical Framework for Private Social Capital Sources of Private Social Capital (Being a member of certain networks or diflerent social position; inhere or invest) Private Social Capital (Resources embedded in social relations) _———’ Mechanisms (Repeated interaction, Social norms and rules, and Personal endowment and Personality) Filters (Levels of analysis and Difference of time frame) Consequences 68 This framework has several strengths worth recapitulating. First, it regards social capital as a neutral resource embedded in the individual’s social relations. Second, it differentiates social capital’s sources from its consequences." Third, it considers both the structural and relational dimensions as the possible resource of social capital. Fourth, there are three main factors serving as the interrelated mechanisms that convert social capital (resources) into actual goods (consequences). Network ties (closure or repeated interaction) provide individuals opportunities to engage in social exchanges, through which reputations, sanctions, and certainty can be created. Norms and rules offer a stable frame of reference for individuals to gauge each other’s obligations and rights, which form the potential resources for future utilization. In other words, they provide structural opportunities or constraints for people engaging in social exchanges. Given the opportunities and resources, the ultimate actions should be taken by the individuals involved in the social exchange. Therefore, the ability to recognize the potential resources embedded in the social relations and to appropriate them plays a decisive role in shaping the actual outcomes. Fifth, it warns that the consequences of appropriating social capital, assumedly producing positive consequences from the perspective of users, may cause negative (undesired) consequences later for the same users or to the greater community as a whole. That is why I use level of analysis and difference of time frame as the filter for analyzing the possible outcomes of appropriating private social capital. 2' There is a common mistake made by some scholars in this field: the sources of social capital and the benefits (outcomes) derived from them are often confused. Defining social capital as equivalent to its sources will lead to logical circularity. For example, social capital should be understood as the ability to command resources embedded in social relations. The actual level of such resources is different from social capital itself. Thus, “a student who obtains the money necessary to pay for college tuition from her parents or relatives is thought to have social capital;. ..Such an inference does not take into account the possibility that the unsuccessful student also my have highly supportive social networks that simply lack the economic means to meet such an expense” (Portes, 1996219). 69 The second half of this chapter will discuss the other camp of social capital literature: social capital as social virtues and the ability to associate among social actors. STUDYING PUBLIC SOCIAL CAPITAL: A THEORETICAL PERSPECTIVE Public social capital here refers to the voluntary associations and those norms or values, which are conducive to cooperation among social actors.22 It is a desirable resource for a community or society as a whole because it can reduce the transaction costs of social exchange and hence help solve collective action problems.23 Among those 22 In fact, there is still no consensus among scholars on the definition of public social capital. The reason is perhaps Putnam’s ambiguous concept of social capital. He defined social capital as “features of social organization, such as trust, norms, and network, that can improve the efficiency of society by facilitating coordinated actions” (1993b: 167). He further delineated the relationship among these three essential components of social capital by saying “social trust in complex modern settings can arise fi‘om two related sources—norms of reciprocity and networks of civic engagement” (Ibid: 171). In another article, he argued “people who join are people who trust. . .the causation flows mainly from joining to trusting” (1995: 66). Further, he argued that “the greater the density of associational membership in a society, the more trusting its citizens. Trust and engagement are two facets of the same underlying factor—social capital” (1995:73). Also he wrote “[N]etworks of civic engagement foster robust norms of reciprocity” (1993b: 173). Therefore, we seem to be able to conclude that Putnam’s “network of civic engagement”, i.e., intense horizontal interpersonal communication, is the generator of social trust and norms of generalized reciprocity. Given this ambiguity between his definition and subsequent discussions, scholars influenced by Putnam tend to either: focus on networks of civic engagement as a society’s social capital and treat social trust and norms of reciprocity as its consequence (Fox, 1996; Evans, 1996; Woolcock: 1998); or treat social trust and social norms as social capital and regard the mechanism of repeated interaction, including networks of civic engagement, as the sources of social capital (Fukuyama, 1995 & 1999; Uslaner, 1999; Whiteley and Seyd, 1997). This problem of imprecision in definition needs to be resolved or integrated in order for the study of social capital to be furthered. 23 Collective action problem is usually represented by the prisoner’s dilemma game. However, there are at least two other types of collective action problems besides the prisoner’s dilemma game, i.e., the chicken game and the assurance game. The factors that differentiate these situations include the nature of the public good that created a collective action problem and the structure of the setting for its provision. In this chapter, the collective action problem refers mainly to the assurance game, in which most people think cooperation is their best choice but are unsure about others’ willingness to cooperate. For concise 70 advocating this idea, political scientists and developmental theorists are most cited in the broader social capital literature." For example, the leading figure of this field, Putnam, defined social capital as “features of social organization, such as trust, norms, and networks, that can improve the efficiency of society by facilitating coordinated actions” (Putnam, 1993b: 167). Fukuyama also defined social capital as “a set of informal values or norms shared among members of a group that permits cooperation among them” (F ukuyama, 1999:16). Hall regarded social capital as “the propensity of individuals to associate together on a regular basis, to trust on another and to engage in community afiaim” (Hall, 19992417). Pye’s social capital refers to “networking and learning to work together on the basis of trust” (1999:764). Still Whitely and Seyd treated social capital as “citizens’ trust in other people” (Whiteley & Syed, 1997:125). Most developmental theorists implicitly follow Putnam’s definition (Fox, 1996; Evans, 1996; Brown & Ashman, 1996). Since Putnam is one of the most prominent political scientists” leading social capital research in political science, it is helpful to focus on his major theses related to social capital as the starting point of this section. Specifically, I will try to synthesize current literature on public social capital in the order of the following themes: the nature and origin of public social capital, maintenance of public social capital, consequences of descriptions of these collective action problems, see Robert Bate (1988). 2‘ Sociologists seem reluctant to accept the idea of public social capital and think of it as a conceptual stretch from the original social capital idea (Portes, 1996 & 2000). However, a tremendous amount of research centering on Putnam’s idea of (public) social capital has accumulated. This trend also converges into an interdisciplinary business, since it covers a variety of topics such as the study of trust, civil society, rational choice theory, etc. It is an undeniable trend and deserves its own label. That is why 1 call this current the study of public social capital and treat it separately fi'om the sociological view of social capital. 2’ Robert Putnam’s influence on the spreading of the idea of social capital can be reflected in the passages in President Clinton’s State of the Union address in 1995. He was invited to Camp David to consult with President Bill Clinton after his “Bowling Alone” was published in 1995 and was also the subject of a profile 71 public social capital, and the mechanism relating public social capital to its consequences. In addition, ‘social capital’ in this section refers exclusively to public social capital unless otherwise defined. The Nature and Origins of Public Social Capital Among Putnam’s theses, the following seems to be particularly relevant to the nature and origins of public social capital. First, that the formation of social capital is path dependent. He summarized it as “where you can get to depends on where you’re coming from, and some destinations you simply cannot get to from here” (Putnam, 1993b: 179). In other words, the historically derived social contexts provide each society with different sets of Opportunities and incentives (Ibid). Second, that there are three essential forms of social capital in Putnam’s analysis: trust, social norms and networks of civic engagement (Putnam, 1993b: 167,169-170). Third, that social capital is a public good and must often be produced as a by-product of other social activities. Fourth, that most forms of social capital are “moral resources” in nature, which increase with use and diminish with disuse. Fourth, that “social trust” is the major form of social capital, which is derived from two other forms of social capital, i.e., norms of reciprocity and networks of civic engagement." In Making Democracy Work, Putnam presented his findings about regional in People magazine (Lemann, 1996). 2" Putnam’s three major components (forms) of social capital have specific meanings. Specifically, trust refers to social trust or generalized trust. Norms of reciprocity refers to generalized reciprocity, which involves mutual expectations that a benefit granted now should be repaid in the future. Networks of civic engagement represent intense horizontal interaction among social actors. For details, see Putnam (1993b: 72 governmental performance in Italy”. He showed us that the regional governments in northern and central Italy outperformed those in the South. His major contribution in this book is to explore the conditions that give rise to these differences. The answer, according to Putnam, lies in the presence of “civic virtue” (civic community) in the North and Central and the South’s lack of such attribute. Civic virtue is an ingrained tendency to form small-scale associations that create a fertile ground for building trust and cooperation in the citizenry. Citizens in such communities “do not and can not ride for free, because they understand that their freedom is a consequence of their participation in the making and acting out of common decisions.” In contrast, in a community with less of this attribute, “life is riskier, citizens are warier, and the laws, made by higher-ups, are made to be broken” (Putnam, 1993b: 111). The next logical question is “why do people in the South lack such civic community that those in the North and Central enjoy?” In answering the question of formation of civic community, Putnam followed a path- dependent approach and said the North-South difference dates back to the 11005, when the Normans established a centralized, autocratic regime in the South, and at the same time a series of autonomous republics arose in the North. The “vertical bond”, i.e., hierarchical relationship, characterizes the interpersonal relationship in the South and prevents the norms of trust, cooperation, and reciprocity from taking roots in southern regions. The lack of social capital dooms the South to less economic development and inferior institutional performance as opposed to her northern counterparts. This pattern has remained consistent for decades and exerted a powerful influence on the results of 171-176). 27 In 1970 Italy created local governments in its twenty regions and transferred many of the central government’s functions to them, providing a scenario in which to compare institutional performance. For 73 regional government reform in the 705. Putnam’s social capital theory is originally derived from sociologists’ concept of social capital, with a focus on the collective character of this concept. It is no surprise that most of sociologists don’t accept Putnam’s version of social capital, calling it a “conceptual twist” (Portes, 2000:59). For most sociologists have been treating social capital as an individual level resource and have not paid due attention to the possible nature of social capital as a structural property of the aggregate. The treatment of social capital as an aggregate level resource is further expounded by another prominent author, Francis F ukuyama. In Trust: The Social Values and the Creation of Prosperity, Fukuyama’s main argument is that the ability to cooperate socially and the level of trust in a society, two major components of social capital, condition the prosperity of the economy and structure the market. The market itself, according to F ukuyama, provides opportunity and incentive for people to cooperate for the sake of mutual enrichment (Fukuyama, 1996:356). This inclination to sociability will then reinforce democratic institutions. Actually, Fukuyama tried to use social capital, embedded in the premodem cultural habits such as trust, as an independent variable to account both the variations of market economy and democracy. Specifically, the economic success and sustainability of democracy both rely on the stock of social capital a society holds (356-358). Great similarities exist between Fukuyama’s and Putnam’s analysis of social capital. First, they all attribute a society’s stock of social capital to its historical and cultural roots. Once it is made, a society’s future can theoretically be determined at least in the economic and political domains. Second, relevant to the first, they seem to hold a details, see Putnam (l993b). 74 pessimistic view about the development of societies since path dependence28 conditions the opportunities a society may choose when facing developmental problems. A society lacking social capital due to limited opportunities provided by history will fall into the vicious circle, which also is a near-equilibrium, which is difficult to overcome. Third, both stress the roles of repeated interaction that foster norms of reciprocity and social trust. Putnam focus on the tradition of civic community to fulfill this function while F ukuyama focused on the cultural roots a society possesses. Fourth, they focus on the positive consequences of social capital, just like their counterparts of private social capital in sociology, and leave readers wonder whether or not their so-called social capital has any negative consequences. Fifth, they all think the ability to work together is a resource valuable to both society and individuals. This line of thinking coupled with path dependent argument implies that a society’s success in the economic or political realm may result from an evolutionary process. Accordingly, this poses a difficulty in further explaining why people in some areas had more opportunities to learn to cooperate than those in other areas.29 In sum, the public social capital theory’s account of origin of social capital is primarily historically and culturally oriented. This line of thinking implies a pessimistic view toward a society’s development, whether economic or political. Thus, the 2’ The concept of ‘path dependence’ is used mostly by economic historians. It points out the important role of past events in circumscribing contemporary choices. It’s main argument, in Putnam’s words, is that “where you can get to depends on where you’re coming from, and some destinations you simply cannot get to from here” (Putnam, 1993b: 179). In other words, history matters in determining the durable differences in performance between two societies, even when the formal institutions, resources, relative prices, and individual preferences in the two are similar (Ibid). 29 The social capital theory proposed by Putnam and Fukuyama highlights the concept of equilibrium, i.e., positives such as trust and cooperation will reinforce themselves, while undesirables such as distrust will also reinforce themselves, and hence either virtuous or vicious circles will result. The danger of thinking in terms of equilibrium is that “it makes it very easy to skirt the important issue of how the virtuous circle is 75 endowment of a society’s stock of social capital is difficult to purposely construct. Other scholars, however, try to supplement this theory by providing alternative means to produce social capital. Their accounts are mainly from the rational choice theory and new institutionalism.3o Using trust as a major form of social capital, Margaret Levi contended, “history and experience form the basis for initial starting points... these are the experiences a particular set of individuals has with another set of individuals or with a state. The vicious circle can, in principle, be broken, but only by attacking racism and other forms of interpersonal discrimination and only by institutions that ensure the credible commitments of governments” (Levi, 1996: 48). Actually, contrary to Putnam’s bottom-up thesis, Levi highlighted the role of the state and policy performance in reducing the uncertainty facing people’s everyday life. This top-down approach stresses ‘ the ability and importance of the central authority (third-party enforcer) in facilitating trust and cooperation among citizens.3| In a similar vein, developmentalists also point to the importance of synergy or cooperation of the state and society arguing, “a range of developmental outcomes is possible, depending on the prevailing combinations of the state’s organizational capacity, and its engagement with and responsiveness to civil society “ (Woolcock, 1998:176). Peter Evans also emphasized the constructability of social capital instead of Putnam and Fukuyama’s “endowment” thesis. According to him, a competitive political system, egalitarian social structures and robust bureaucracies all initiated in the first place” (Boix & Posner, 1998: 687). 3° This intellectual current highlights many issues concerning cooperation among human being, such as “prisoner’s dilemma game”, tragedy of the commons”, and “the logic of collective action”. All of these models assume the nature of human society is uncooperative (Hardin, 1968; Olson, 1965 & 1982). 3 ' For example, Levi cited new economic institutionalism’s argument that the state can build trust among citizens by establishing and enforcing property rights (Levi, 1996: 5.1). The general idea of new institutionalism is that institutions help individuals to overcome collective actions dilemma by providing incentives and coercion to induce trust and cooperation among individuals (Bates, 1988). 76 facilitate emergence of state-society synergy, which accordingly promotes the norms of cooperation and networks of civic engagement among ordinary citizens and can be used for developmental ends (Evans, 1996). Given the above discussion, we are still left to ponder the most basic question: are human beings cooperative in nature? Or put it in another way, can human societies generate social capital at all? This turns our attention to evolutionary biology’s account of how cooperation is generated in animal societies. Biologist Robert Trivers contends that reciprocal altruism would be most likely to develop in species enjoying repeated interactions, relatively long lives, and the cognitive capabilities to distinguish cooperators from betrayers. It is exactly these mechanisms that allow human beings to develop social capital (Fukuyama, 999:172). F ukuyarna further elaborated this argument by focusing on the psychological traits that evolve during the evolution of the human psyche, including the cognitive capabilities to distinguish cooperators from defectors, and emotions or instinct that ensure we play social interactions reciprocally (Ibid: 177). The information derived from this field runs against the assumption of rational choice theory, which postulates that human cooperation is difficult due to the tendency of individual’s utility maximization. Similarly, Boix and Posner thought, “[I]t might well be the case that co-operation emerges spontaneously. . .and that what truly requires explanation is the set of forces that block its continued growth” (Boix & Posner, 1998: 688). From the perspective of Marx’s class theory, their main argument is that those benefiting from the given social structures will do everything to prevent those disadvantaged from cooperation to take over their role. Therefore, they proposed that a community’s co-operative capacity is a function of the 77 degree of social and political inequality that the community has experienced over the course of its historical development (lbid: 687). In order to demonstrate their proposition, they reinterpreted the variation of social capital in Putnam’s Italy. They pointed to the historical events such as “the Norman invaders” and “steep social hierarchy” (social inequality) as the two major factors preventing people in southern Italy from acquiring co-operation equilibrium as compared to their counterparts in northern Italy. The authors also claim that their explanations can be applied to the case of Spain (lbid: 688-689). In sum, we seem to be able to reasonably conclude the discussion by proposing the following plausible argument: Every human group, community or society has the potential to generate social capital. Therefore, the development of social capital is natural to human societies. We can even assume every society has the same amount of potential to produce social capital. The variations of level of social capital that we experience in diflerent societies are results conditioned by the interactions among structural, cultural and historical factors in each society. The Maintenance of Public Social Capital There are two major theses from Putnam that are related to the maintenance of public social capital. First, civic traditions and therefore social capital have remarkable staying power and are nearly impossible to change due to the virtuous circle that social capital enforces (Putnam, 1993b: 171). Second, stocks of social capital tend to be self- reinforcing and cumulative (177). This mechanism helps evolve two social equilibriums, 78 i.e., virtuous circle and vicious circle, for all societies facing collective action problems. Again, I will use these theses as a starting point. There are at least two ways to consider how public social capital is sustained. We can think of it in a positive manner, i.e., asking what factors can reproduce public social capital; or we can think it in a negatively, i.e., what factors impede public social capital from reproducing itself. These two approaches should be regarded as complementary rather than mutually exclusive. 1. The Positive Approach The mainstream thinking regarding the maintenance or accumulation of social capital is that stocks of social capital tend to be self-reinforcing and cumulative and its supply increases rather than decrease through use and becomes depleted if not used. Albert 0. Hirschman, for example, contends that ‘love’ and ‘public morality’, such as civic spirit, trust, observance of elementary ethical norms, and so on, are important inputs in facilitating the functioning of economic system. But he criticizes economists’ assumptions on these inputs as scarce resources. For him, these are resources “whose supply may well increase rather than decrease through use”32 and “these resources do not remain intact if they stay unused” (Hirschman, 1992 155). Putnam concurs with Hirschman as well (l993b: 169). Garnbetta also argues, based on the assumptions that behavior spreads through learning and imitation, that trust is not depleted through use; it 32 Note that Hirschman did not propose a linear relationship between using these resources and increasing supply of them. On the contrary, there exists a limit at which increased practice of these love and public morality does conflict with self-interest and therefore reduces its supply (Hirschman, 1992:156-157). 79 may increase through use. Similarly, sustained distrust can only lead to further distrust (Gambetta, 1988: 234). Perhaps Sztompka’s remarks on the culture of trust can best summarize these authors’ arguments: “Once they [the culture of trust] spread, a self-amplifying process starts to operate: The syndromes are enhanced by imitation and mutual confirmation. Spreading in the population, such complex personality syndromes33 turn into a phenomenon of a macro-societal order, which may be called social moods.” (Sztompka, 19992126) Public Good vs. Private Good Given the above thesis, the issue now is how it is possible to increase the chance that people consume social capital so it can accumulate. In other words, what is the mechanism for social capital to reproduce itself? The current literature seems to be divided into at least two camps. Interestingly, however, both camps share the same basic assumptions on human beings, i.e., social actors are rational in mind, but come to very different conclusions. The first camp regards social capital as a public good and must often be produced as a by-product of other social activities (Coleman, 1988: 317-318; Putnam, 1993b: 170). The logic of this argument runs: given the rational nature of human beings, the actions taken by people can be the resources of others through the mediation 33 In fact, Sztompka argues for certain personality syndromes that are associated with trustfulness, including trusting impulse, general activism, optimism, future orientation, high aspirations, success orientation, innovative drive. Social trust may be the product of the interactions between structural opportunities and these personal traits. Although there is little knowledge about how the actual mechanism generates such trust, once it is there, it will reinforce itself. (Sztompka, 1999:125-126) 80 01 St at."- Ian-l \zfil' all] iti of social structures, e.g. closure of social ties, and social norms. And these resources cannot be manipulated by the beneficiaries of social capital.34 In other words, social capital is the resources generated from social relations in which social actors act on their own interests. People involved in certain social relations cannot control the production of social capital. They just take it as given. It ends with no one willing to produce resources purposely for other people because the benefits of actions that bring social capital into being are largely experienced by persons other than the actor. Thus nonmarket actors are the logical bearers for social capital to thrive. That is why Coleman stressed the role of family while Putnam emphasized the role of voluntary associations in supplying social capital.” The other camp, represented by F ukuyama, does not treat social capital as a public good. Instead it is a “private good pervaded by extemalities” (Fukuyama, 1999:257). This argument is particularly true, according to Fukuyama, for modern capitalism. Based on the same assumption of individual utility-maximization, Fukuyama thought it is in the long-term interests of selfish individuals to produce social capital because these virtues n 36 will eventually become “economic assets . He illustrated this point using examples mainly from commercial actors. For instance, in order to survive the market competition, 3’ Coleman uses several examples to illustrate this point. For instance, “[A]n individual who serves as a source of information for another because he is well informed ordinarily acquires that information for his own benefit, not for the others who make use of him” (Coleman, 1988: 5117). 35 Actually, voluntary associations or networks of civic engagement can be seen as a major intervening factor that reproduces social capital in Putnam’s theory. I will discuss the point later in this section. 36 Albert Hirschman also holds a somewhat similar view on what a rational individual should do when facing collective action problems. He argued that “a truly maximizing individual will attempt to be as activist as he can manage” instead of shirking and attempting to get a free ride. Since “ the output and objective of collective action are...a public good available to all, the only way an individual can raise the benefit accruing to him from the collective action is by stepping up his own input, his effort on behalf of the public policy he espouses” (Hirschman, 1992: 152). The main difference between Hirschman and Fukuyama is that Hirschman figured noninstrumental consideration into utility calculation, while Fukuyama 81 individual firms or corporations must show a high degree of honesty and civility in their customer service; develop corporate cultures among employers, which attempt to socialize workers in a firm into a series of norms that will improve their willingness to cooperate with on another; and build a sense of group identity (lbid: 256-257). These virtues, understood as social capital, not only become the economic assets for those who produce them but also hold a host of beneficial spillover effects for the broader society.37 Thus, from the perspective of private good, the modern market economy will produce social capital constantly, and this demand will intensify as we approach a more complex and technologically higher level of economy (Ibid: 255). Fukuyama’s praise of the role that market economy plays in producing social capital actually challenges conventional wisdom. However, overemphasizing the positive aspects of capitalism tends to hinder our ability to consider the potential dark side of its impacts on social capital. For instance, the efficiency of market itself must depend on fair court system and contract enforcement. These institutional arrangements are designed to make opportunistic behavior or free-riding costly. But these mechanisms ignore the roles of concrete personal relations and the obligations inherent in them in discouraging malfeasance. Hence, as Granovetter states, “substituting these arrangement for trust results actually in a Hobbesian situation, in which any rational individual would be motivated to develop clever ways to evade them; it is then hard to imagine that everyday focuses purely on instrumental consideration in producing social capital. 37 In fact, Fukuyama did not specify the actual mechanism of how self-interested behavior could be generalized for people. The possible answer found in his book might be, on the one hand, that “[W]e often start out obeying a norm for self-interested reason, but continue obeying it for what amount to moral considerations”. Thus it is difficult and often unreasonable to detach moral behavior entirely from self- interest. On the other hand, market exchange, as opposed to Putnam’s civic engagement, promotes norms of reciprocity, which can be carried from economic life on into moral life (Fukuyama, 1999: 260-261). 82 economic life would not be poisoned by ever more ingenious attempts at deceit” (Granovetter, 1985: 488). Thinking About the Ways of Solving Collective Action Dilemma Another line of thinking related to the maintenance of social capital is inspired mostly by rational choice theorists. They started by asking themselves what makes collective action problem persist and what may solve collective action problem. Only by exploring these questions can we find the essentials to production and maintenance of social capital. Michael Taylor and Sara Singleton provided us with a helpful insight into these questions. They saw the existence of collective action problems from the perspective of cost-benefit analysis. In general, if the transaction costs38 involved in solving collective action problem are higher than the expected benefits, the problems will remain unsolved. Taylor and Singleton then asked “why are some groups able to solve a CAP by themselves and others not?” Specifically, “what distinguishes those groups whose members manage to solve a CAP wholly endogenously from those whose members fail to solve very similar CAPs endogenously” (Taylor and Singleton, 1993: 198)? Their answers lie in several characteristics collectively possessed by group members, i.e. “stability of relations”, “multiplex relations”, “direct relations”, and “shared beliefs and preferences” (Ibid, 199). In essence, what Taylor and Singleton tried to 3’ There are three phases, according to Taylor and Singleton, involved in solving collective action problem, i.e., identifying the possibilities for cooperation, bargaining on one scheme of cooperation, assuring the agreed scheme and their cooperation remains monitored and enforced. Each of the three phases will incur its corresponding costs, namely search costs, bargaining costs, and monitoring and enforcement costs. These costs are generally called “transaction costs” (Taylor and Singleton, 1993:196-198). 83 demonstrate was that “community”, which meets all the above characteristics, can lower the transaction costs of solving collective action problem. In a similar vein, Bates also pointed out the problem of collective dilemma involved in solving collective action problems. That is, an institution may provide incentive for cooperation in collective action; however, the actors in the collective action may not be able to agree on which institution to choose (Bates, 1988: 397-398). Instead of this institutionalist solution to collective action problems, Bates suggested “soft” behavior such as expectation, signaling, and uncertainty in inducing cooperation in collective action.39 Community, according to Bates, provides the environment to fulfill these soft solutions. Putnam concurred with such line of thinking and incorporated the so-called ‘civic community’ into his social capital theory as a critical factor in producing and maintaining social capital. Putnam argued that networks of civic engagement or, broadly speaking, civic community provide an ideal environment for people to develop interpersonal trust and norms of reciprocity. The interactions involved in the civic engagement facilitate coordination and communication and increase information availability about the trustworthiness of other individuals. And this will lead to further participation in civic activities. Hence a ‘virtuous circle’ may sustain itself. In other word, civic community provides an enabling environment to economize on the need for intervention by external agency and to easily induce cooperation among members. 2. The Negative Approach 84 Aside from this positively oriented approach, we can also tackle the maintenance of social capital from the negatively oriented approach, i.e., which factors will impede the accumulation of social capital? So far, most works focus on social structural factors. They all share one similar assumption, i.e., social cooperation is a natural tendency in human societies. What must be explained are the variables that obstruct social cooperation (Boix and Posner, 1998; Woolcock, 1998; Fukuyama, 1999). Of those social structural factors, social inequalities, divided society, inappropriate law enforcement, undemocratic governance, and war are often cited as factors hostile to the maintenance and accumulation of social capital. In other words, the accumulation of public social capital needs some enabling environments. If those environments do not exist, we will face the so-called “scale-up” problem."o Consequences of Public Social Capital Like its counterparts of private social capital, most literature in public social capital draws our attention to the positive consequences it produces, whether directly or indirectly. In this section, I will briefly summarize these theoretical consequences and some empirical evidences supporting them. In addition, the possible negative 39 Bates seemed to base his arguments on the assumptions that rational people prefer to cooperate when facing uncertainties and that reciprocating other people’s behavior is natural for human beings. 4° Both Fox (1996) and Evans (1996) mention this problem. Essentially, it means that even in the later development there is a sufficient amount of public social capital in lower aggregate level. The question is how to scale it up. In order to expand the social capital, according to these development theorists, a series of institutional arrangements needs to be in place, such as a competitive political system, egalitarian social structures and robust bureaucracies. 85 consequences, largely ignored by scholars, will also be discussed. 1. Positive Consequences of Pubic Social Capital The theoretical consequences of public social capital are primarily derived from the mechanism of repeated social interactions.“l The general argument states that given the assumption that human beings have the cognitive ability to distinguish cooperators from betrayers and the emotional system to reward cooperators and punish defectors, repeated social interaction provides opportunities for people, organizations, communities, and societies to reduce uncertainties, increase information sharing, increase cooperative behavior, and impose sanction (or increase costs) against opportunistic strategies. These effects can in sum lower the transaction costs incurred in collective action problem, or broadly speaking, social exchange. As to the mechanism, i.e., repeated interactions, scholars have their own focus. Putnam emphasizes the networks of civic engagement, Fukuyama stresses the market system, and Evans and Woolcock draw attention to the relationship between state and society. Based on the general argument postulated above, public social capital is said to l) sustain economic dynamism and government performance; 2) reconcile self-interest and solidarity; 3) foster robust norms of reciprocity; 4) facilitate communication and improve the flow of information; 5) embody past success at collaboration; 6) increase the potential 4' There are of course many other factors influencing the actual outcomes of public social capital, especially historical events and social contexts. Here I sense that many desired consequences of public social capital discussed in literature focus on the results of repeated interaction, which accordingly induces those consequences, while holding other variables constant. Therefore, I point out that repeated social interaction is the primary mechanism underlying the popularity of public social capital. 86 costs to opportunism; 7) make citizens sophisticated consumers of politics; 8) contribute to effective governance by facilitating the articulation of citizens’ demands; 9) reduce the costs of enforcing and implementing governmental policies and regulations by shaping citizens expectations regarding the behavior of others; 10) foster civic virtue by enhancing citizens’ preferences for collective benefits; 1]) promote institutional effectiveness through its effects on the behavior of policy-making and bureaucrats; 12) foster accommodative practices among otherwise antagonistic elites; 13) help ensure efficient regular turnovers of power in democracy; 14) eliminate the need for third-party enforcers or insurers (Putnam, 1993b; Fukuyama, 1995, 1999; Boix and Posner, 1998; Paxton, 1999). Some of these positive theoretical expectations of public social capital have been supported by empirical data and can be sorted into two broad categories: socioeconomic development and institutional performance."2 Socioeconomic development Most social capital literature argues that the reason why social capital can produce or help to produce desired outcomes is that it can reduce transaction costs. We have already seen Putnam’s work on Italy. Actually, Putnam claimed in Making Democracy Work that civic community, as a form of social capital, is the precondition for economic growth (Putnam, 1993a; 1993b). F ukuyama argued that the level of trust inherent in society is the single most important factor determining a nation’s well being, as well as its ability to compete (F ukuyama, 1995). Stephen Knack and Philip Keefer, through a cross- ” These consequences are mostly analyzed at the macro-level. 87 national comparison, showed that trust and civic cooperation have significant impacts on aggregate economic activity. In addition, the relationship between level of interpersonal trust and economic growth is especially significant in poorer countries where formal institutions are less well-developed (Knack and Keefer, 1997). Still, much research has been done on the role of social capital in democratic transition in Eastern Europe since 1989. Richard Rose illustrated the incompetence of the Russian Federation in solving its social problems and the strength of informal networks to help Russians survive the hardships of transformation (Rose, 1995). Kolankiewicz George argues that the extensive networks left over from the communist regime, the norm of reciprocity, and trust help the transformation from socialism to capitalism since they help enforce agreements among potential entrepreneurs, their potential clients, and their creditors at times when proper institutions enforcing contractual obligations were still under construction (George, 1 996). Institutional performance Again, Putnam found regional institutional performance"3 in Italy to be positively correlated with the level of civic engagement, the latter being the precondition of the former (Putnam, 1993a; 1993b). Another case is from Judith Tendler and Sara Freedheim (Tendler and F reedheim, 1994). In a search for answers as to why a certain Brazilian State notorious for clientelistic and corrupt ways of governing suddenly performed so well, the ’3 Putnam used twelve indicators, including internal processes, policy pronouncements, and policy implementation in many policy sectors, to measure institutional performance. Factor analysis was used to create a composite measure of it. For details, see Putnam (1993b, 66-82). 88 authors found, in the case of rural preventive health program, that the health agents (workers) took on a variety of extra tasks voluntarily,“ and therefore gained the trust and respect from the people who are in the legacy of mistrust of anything that came from “government.” This embeddedness of public workers in the community they serve resulted in outstanding achievement by this public program. Similarly, Lam showed how government officials and citizen-users cooperate to achieve the successful management and governance of irrigation in Taiwan (Lam, 1996). 2. Negative Consequences of Public Social Capital So far the majority of the literature focuses only on the positive effects public social capital may provide. However, just like its counterpart, the same set of public social capital may have negative effects on the aggregate actors. I will limit my brief discussion to the following four issues: the roles of vibrant civil society, contextual factors in shaping actual consequences of public social capital, the roles of community, and the roles of trust in sustaining democracy. Again, Putnam’s social capital theory is a good starting point for my discussion here. In Putnam’s theory (Putnam, 1993b), social capital can improve institutional performance, but he was argues this by assuming the process of making public policies to be a constant. In other words, he only focuses on how public policies are implemented with the aid of social capital and ignores how public policies are made in the context of social capital. Olson, for example, argues that those groups which are more successfully ‘4 The incentive for participating in these extra tasks for these agents is from the state’s hiring process. For 89 manage their collective action processes can of course increase the aggregate welfare of the group, but this gain is at the cost of societal welfare, which implies the detrimental feature of successful collective organization (Olson, 1982). The dimension of social conflict underlying the networks of civic engagement is largely missing in social capital literature. Furthermore, Putnam’s and other neo-Tocquevillean scholars seem to disregard the contextual factors in shaping the actual outcomes of social capital (Edwards and Foley, 1998: 132). If we put contexts into consideration, the story may have changed. For example, in a situation where a society is deeply divided the social interaction (networks of associational life) that forms social capital may well intensify rather than ameliorate the given conflicts among those groups. This is because due to the lack of mutual trust the expected cost of showing cooperative endeavors across network boundaries is higher than the potential gain (Boix and Posner, 1998: 693). Therefore, we may see on the one hand the highly accumulated social capital within subgroups; on the other hand, it is exactly this unequal distribution of public social capital across subgroups that reinforces given communal divisions instead of building norms of reciprocity and social trust across groups. The situation may worsen when the central government shows an ambiguity about its purposes. State weakness not only invites the efforts of those who seek to exploit government for their own purposes but also invites social conflict through its own weakness and its apparent unwillingness to intervene to alter social dynamics (Whittington, 1998: 28). As we have seen, social capital literature seems to acclaim the roles of community details, see Tendler and Freedheim (1994: 1777). 90 as an effective cure for collective action dilemma. The stability of relations is said to be the necessary condition for the rationality of conditional cooperation in the prisoners’ dilemma game; multiplex relations in the community intensify the capacity for mutual monitoring of social ties; direction relations enable a community to tackle its own problems with the intervention of outsiders; shared beliefs and preferences smoothen the road to agreement in searching collective action solutions (Taylor and Singleton, 1993: 199; Portes, 2000: 57). But the same characteristics that form a community can sometimes make social actions a high-risk undertaking. Stability of relations over time can engender overwariness of offending others; multiplex and direct relationships can increase the risk involved in challenging another individual (Massey, 1994:429). The key issue here is the degree of autonomy of individuals in the community. Taylor and Singleton focus on the linkage within community members and the “endogenous sanctions” from the linkage, which will ensure participation in collective action. Massey emphasizes the autonomy of individuals in the community and asserts that a lack of anonymity in a strong community can actually make participation in a social movement more difficult, especially if it is a controversial one. The flip side of positive effects of social capital concerns the role of trust in maintaining a political regime. Most social capital literature praises trust as a “lubricant” that makes the running of any group or organization more efficient (Fukuyama, 1999: I6; Putnam, 1993b: 171). Following this view, distrust of government or wariness of others seems to be treated as “a matter of cynicism rather than political judgment” (Whittington, 1998:28). In fact, the emphasis of trust in inducing cooperation in collective action runs the risk of assuming linear relationship between trust and its desirable consequences. A 91 society in which most extend unlimited trust toward others or institutions may not go without negative consequences. A political system lacking checks and monitoring from civil society easily abuses its power. Similarly, in a society of angels, i.e., totally honest and trusting, the social order will not be stable since a tiny fraction of defectors would benefit unproportionally from this setting and “all it takes is one spectacularly successful opportunist to turn angels into ordinary, distrustful mortals” (F ukuyama, 1999: 176). Thus although a general distrust of others is an obvious social ill, vigilance and wariness about others, especially about institutions, may also constitute important components for a healthy society. This is why some even suggest, “distrust and suspicion may, in a very fundamental sense, constitute potent and important forms of social capital” (Kramer, 1999:590) An Analytical Framework for Studying Public Social Capital In this section, I will present a general conceptual framework that specifies the sources, benefits, and risks of public social capital based on the literature review above. It is presented in Figure 3-3. 92 Figure 3—3 An Analytical Framework for Public Social Capital Origins of Public Social Capital (Human Nature, Historical Events, and Cultural Tradition) Intermediary Factors (Formal Institutions’ Roles, Social Structural Factors, and a Public Social Capital Personal Endowments) (Social Trust and Networks of Civic Engagement) Filter (Contextual Factors, such as Social equality and cleavage) 1 Consequences 93 Several points need to be elaborated regarding this framework. First, it assumes that norms of reciprocity will emerge naturally given the assumption of repeated interaction and that human beings have the cognitive ability and emotions to distinguish traitors from cooperators." Second, historical events and cultural tradition provide collective memories and a frame of reference for actors to partake in social interactions, and hence they act as background variables in shaping the possibility of forming public social capital. Third, social structure’s role in shaping the stock of social capital lies in the stability and continuity it presents society. Basically, stability and continuity cultivate feelings of safety and a predictable future. Formal institutions can also foster the formation of social capital by fulfilling several functions. For instance, they can provide a fair and transparent environment in which people may interact. In addition, the different institutional designs can help or impede the formation of social capital, e.g., a totalitarian ligjflfiléfiupposed to repress the networks of civic engagement. Also, most of the formal institutions are designed neither to foster nor destroy social capital. The extemalities of these institutions therefore exert influence on the social capital formation. Personal endowment involves psychological and material resources that people hold. The more resources one enjoys the more likely that one will be willing to reach out for connections due to the reduced cost likely encountered in social exchange. In sum, the predictability and fairness of environment, as well as sense of control in social interactions encourage ‘5 This assumption originates from my reading of Fukuyama’s book (Fukuyama, 1999). However, a similar view has been widely accepted by other scholars. For examples, see Malinowski (1926), Gouldner (1960), and Levi-Strauss (1965). 94 people to trust and engage."6 Fourth, an adequate amount of social trust and the ability to associate among social actors are regarded as the main component of public social capital. However, the relationship between the two is not so clear in current literature.“7 Fifth, the actual consequences that public social capital engenders depend on the configuration of social structures, which I call the filter or mechanism. The same set or form of social capital may produce different outcomes due to different structural configurations of mechanisms each group or society faces. For example, a high level of social capital may exist in each of the subgroups of an ethnically divided society. However, due to severe hostilities towards each other, the stocks of social capital can not meld and be scaled up to the higher level. The given conflict could also worsen on account of the high level of social capital in each camp. On the other hand, a high level of ’6 According to literature on trust, the factors that can induce trust can be boiled down to the resources and certainty (or predictability). That is, other things being equal, the more resources one holds the easier one can place trust in others. Similarly, the more certain an actor is of the environment the more likely she will choose trust. For detailed reasons, see Sztompka (1999), Kramer (1999), Patterson (1999), Luhmann (1979), and Giddens (1991). As to the “sense of control” in inducing cooperation among people, Watabe and his colleagues conducted a series of experiments using a sequential prisoner dilemma They found that if participants were assigned to make decision before his or her partner and know the partner knows his or her decision, those participants will tend to show a higher level of willingness to cooperate than under other logically equivalent experimental conditions. They attributed this difference to the participants’ sense of controlling the outcomes. For details and comparable experiments, see Hayashi, Ostrom, Walker and Yamagishi (1999). Similarly, Uslaner argues that sense of control will lead to a sense of optimism for the future and will replace particularized trust with general trust in others (Uslaner, 1999: 144-145). ‘7 There are debates on the relationship between social trust and participation in voluntary association, two most important components of public social capital. Putnam, for example, argues that “people who join are people who trust...the causation flows mainly from joining to trusting.” (1995: 666) Others, like Tocqueville (1968), and Mill (1910) also stress that networks of voluntary activity create trust and cooperation. Nevertheless, some scholars criticized this argument as ignoring the possibility of social conflict that voluntary associations may result. (Whittington, 1998; Olson, 1982) Still others like Sullivan & Transue question the assumption of this type of argument, which generalizes attitudes from one domain to people in general. Rich examples from psychological studies of perception show that people can compartmentalize their attitudes. (Sullivan & Transue, 1999: 648) In addition, I will argue that the experience of participating in voluntary association may not always be good in terms of interaction with others. Thus the effects of membership in voluntary association on social trust should be expected to be either zero or very small. 95 subgroup public social capital is more easily scaled up to cross-group social capital if that society is less divided or enjoys higher level of social equality. CONCLUSION In this chapter I have developed two analytical frameworks for private and public social capital respectively. Based on these frameworks, I will use Taiwan to test hypotheses derived from the frameworks. This task is limited to the exploration of social capital and its possible political consequences. However, before going to the empirical research, it is helpful to get grasp the culture, history and political development of Taiwan. For it not only provides us with background knowledge of the case we are studying but also supplies necessary information from which hypotheses in the latter chapters can be derived or reasonable explanations can be expected after hypotheses tests. Thus, the next chapter will explore the structural sources of social capital in Taiwan. 96 CHAPTER 4 METHODS, DATA, AND MEASUREMENT INTRODUCTION As stated in chapter one, in addition to the clarification of the concept of social capital (chapter two) and the suggested analytical framework for studying social capital (chapter three), the rest of this dissertation, chapter five through chapter seven, is an empirical study exploring social capital and its political consequences in Taiwan. Thus, there is a need to introduce the methods, data, and measurement that will be used in this study. METHODS This dissertation project employs both qualitative and quantitative methods. Historical analysis will be the main tool used in chapter five, where the structural resources of Taiwan’s social capital are analyzed. To measure the political consequences of both private and public social capital, statistical methods, both descriptive and inferential, are mainly used. In general, the descriptive' statistics, such as frequency and percentage distribution of major variables will be presented first. Bivariate analysis, such as crosstabs with Chi-Square statistics (12) or correlation coefficients will then be presented as measures of association between variables in question. Finally, multivariate 97 analysis, such as OLS regression or logistic regression, will be performed as the ultimate test to the hypotheses proposed in each chapter. DATA The arguments of the historical analysis in chapter five will be based on various literature related to Chinese culture, Taiwan’s history and its political development. The empirical study which is the core of chapter six and seven will employ three different sets of national survey data. The first is “Taiwan Social Change Survey,” conducted by the Institute of Ethnology Academia Sinica, Taiwan, Republic of China. The second data set is called “The Social Image Survey in Taiwan”, collected by the Sun Yat -Sen Institute for Social Sciences and Philosophy, Academia Sinica, Taiwan, Republic of China. The third one is “Legislative Election Survey”, conducted by The Electoral Behavior Research Group of the Department of Political Science at National Taiwan University. Since two data sets of the Taiwan Social Change Survey, 1997 and 1998, contain survey questions that most fit my research interests, I will use the Taiwan Social Change Survey as my primary data source and the other two as auxiliary sources when needed. Before further introducing the two major data sets that I use in the following chapters, a brief introduction of the two other survey projects will be provided as well. The Legislative Election Survey is a nation-wide survey project conducted after every corresponding year’s legislative election. The available data are from 1984 to 98 1996.1 All respondents are at least 20 years old and eligible to vote. The sampling method used is multi-stage stratified sampling. Probabilities proportional to size (PPS) are employed to ensure every citizen has the same probability of being sampled.2 The main reason for using this survey series is that it consistently asked people if they participated in voluntary associations in each survey, and the question wording (format) is the same across surveys. Thus it can be used to measure the people’s level of participation in voluntary associations, which is one of the components of public social capital here. The Social Image Survey in Taiwan is also a nation-wide survey project that began in 1990 focusing on understanding Taiwanese' attitudes or opinions regarding major social events and new behavioral patterns that emerged in the early 905 when rapid social and political transformation occurred. Since then as many as 17 surveys have been conducted, each having a specific theme, ranging from measuring people’s understanding or attitudes toward contemporary specific social events and political issues, to educational, psychological, and economic issues. Multi-stage stratified sampling is used. The population is peOple living in Taiwan between 20 and 64 years of age. The available data cover from 1990 to 1994. As far as Taiwan Social Change Survey is concerned, it is the result of a long-term program called “the fundamental research project of Taiwan’s social change” fimded by the National Science Council, Republic of China. It started in 1984 when the National Science Council gathered scholars from various fields in social science to establish a ' The legislative election is held every three years. It is a national level election where people elect legislators representing their own district for the Legislative Yuan. 2 For details, please see Hu et al, 1993. 99 dependable and large survey data bank. The aims of this program are, one, to provide data for social scientists engaging in theory building; two, to aid government understanding of people’s opinion or attitudes toward various issues in order to serve as references for future policymaking. The major difference between Taiwan Social Change Survey and The Social Image Survey In Taiwan is that the former is conducted on a five year cycle tracking long term developments while The Social Image Survey In Taiwan is more focused on measuring the impact of sudden social events or issues that accompanied the democratization of Taiwan. Thus, it is more flexible in terms of capturing the social attitudes of Taiwanese people. However, the program stopped in 1994. Of those surveys in the past fifteen years, two are most related to my purpose. One is the 1997 survey, which conducts Taiwan’s first national sample investigating various aspects of individuals’ social networks. At the same time this survey tried to measure respondents’ involvement in local community affairs. Therefore, it provides us with some usefirl data to observe the patterns of private social capital and its political consequences. The population is adults between 20 and 74 living in Taiwan. Cases are selected by multi- stage stratified sampling with probabilities proportional to size (PPS). The sample size is 2,835. Chapter six will employ this survey as empirical evidences to test the hypotheses formulated in the chapter. The other survey is conducted in 1998. The theme of this survey is people’s long- term political attitudes, including those toward significant controversial issues facing Taiwan as well as the general political attitudes toward the operation of democracy. Specific subjects cover democratic values, cynicism, political efficacy, trust (both social and political), participation (both social and political), and ethnic identification. The 100 population is citizens in Taiwan ages 19 to 75. The sampling method is similar to the one used in the 1997 survey. The sample size is 1,737. Since this data contains both components of public social capital identified in this dissertation, it will be used as primary data in chapter seven to test hypotheses of public social capital and its political consequences. Since I also trace the trend (macro-level) of public social capital before and after Taiwan’s transition to democracy, several surveys will be employed as well to generate aggregate data for this purpose. Specifically, the available survey data that provide the measure of membership in voluntary association are from Taiwan University’s Legislative Election Survey, which includes the years of 1983, 1986, 1989, and 1992. In addition, the Taiwan Social Change Survey also provides four data sets, which cover the years of 1993, 1996, 1997, and 1998. As far as the trend of interpersonal trust, Taiwan Social Change Survey provides five available data sets, which are the years of 1985, 1990,1991,1996, and 1998. The Social Image Survey in Taiwan also has two data sets (1991 and 1993) that can be used to measure social trust. MEASUREMENT A Review of Measuring Social Capital The measurement of social capital poses a serious problem for researchers employing quantitative approaches: conceptual ambiguity.3 We have seen a variety of 3 For various definitions of social capital, please see chapter two. 101 indications of this in literature. For example, Robber Putnam, following James Coleman’s definition, treats as social capital any features of social organization that can improve societal efficiency facilitating coordinated actions. He emphasizes trust, norms and networks (Putnam, 1993b:167; 1995az67; 1995b:664). Stephen Knack and Philip Keefer use ‘social trust’ as their primary indicator of (public) social capital. ‘Norms of civic cooperation’ (civic norms) and ‘membership in voluntary associations’ are two other indicators (Knack and Keefer, 1997). John Booth and Patricia Bayer Richard choose ‘political knowledge’ and ‘interpersonal trust’ to represent (public) social capital. John Brehm and Wendy Rahn treat (public) social capital as the results of the reciprocal relationship between ‘interpersonal trust’ and 'civic engagement.’ By civic engagement, they mean membership in civic and political organizations (Brehm and Rahn, 1997). Dietlind Stolle and Thomas R. Rochon so far provide the broadest set of indicators of (public) social capital. In their research, (public) social capital indicators range from political action, engagement in community, interest in politics, generalized trust, credit slips to political trust, political efficacy, tolerance, free riding and optimism (Stolle and Rochon, 1998:63). Peter Hall operationalized social capital using ‘formal or informal patterns of sociability’ and ‘social trust’ (Hall, 1999: 420). Pamela Paxton used ‘trust in others and trust in institutions’ and ‘individual’s associations with other individuals and through groups’ as two dimensions or indicators of social capital (Paxton, 1999). Brown and Ashman used the existence of local organizations as indicators of the presence of social capital among local groups (Brown and Ashman, 1996). Some even treated political participation and volunteerism as indicators of (public) social capital (Putnam, 19953; Ladd, 1996). 102 Elli I‘ll COT Although scholars have not reached a consensus on what indicators to use in measuring public social capital, no doubt interpersonal trust (or generalized trust) and associational membership are the two most used indicators of public social capital. Thus these two indicators will also be employed in my research as the measures of public social capital. Regarding private social capital, measurement is even more difficult due to its broader definition as any resources embedded in accessible to the ego’s social networks. I will limit my discussion of the measurement of private social capital to that literature concerning the use of network resources to fulfill instrumental goals. According to this research paradigm, social capital takes two different forms in a continuous process of status attaimnent. First, an ego’s social ties combined with human capital and ascribed position (parental statuses) is expected to affect his attained statuses such as occupational status, positions of authority, sectors or earnings (Lin, 2001182). Thus, the extensity of social network, i.e., how far an ego can reach, becomes an ‘accessed social capital’ (Ibid). In addition, social contacts, their available resources and the tie strength between ego and the contact, i.e., strength of ties, are also a form of capital called ‘mobilized social capital’ (Ibid). These two forms of social capital are measured differently in literature. For accessed social capital researchers use either the method of ‘name generators’ or ‘position generators’ to measure the contacts’ diversity, range of resources and access to structural positions.4 For mobilized social capital, the contacts’ occupational prestige is used most often as the indicator of social capital (Ibid: 86). Strength of ties is measured either by closeness with contacts, frequency of visits or calls, 103 ii ‘ij Pul VOIL alii Soc [if pox this iii its; Ill." wit F0 Sir this "it. ..‘io “lift or by content of the relationships The Measurement of Social Capital in This Research Public Social Capital The public social capital used here refers to social trust and membership in voluntary associations. They are measured in several different ways according to the availability of data. The following describes the operationalization of these two measures. Social trust: according to the series of “Taiwan Social Change Survey” and “The Social Image Survey in Taiwan”, seven data sets ask questions about social trust. Unfortunately, the wording was not always the same in various questionnaires; this may pose a threat to the interpretation of results.6 However, the three different wordings in these surveys are distributed consecutively and can still be used as a preliminary comparison.7 Specifically, in the Taiwan Social Change Surveys of 1984, 1990, and 199] respondents were asked if they agree with the statement: there is no longer reliable and trustworthy relationships among people. Yet in 1991’s and 1993’s Taiwan Social Image Survey and 1996’s Taiwan Social Change Survey, respondents were asked if they agree with the following statement: it has become a tendency that people cannot trust each ’ For details and examples, please see chapter six of Lin (2001). 5 Since this concept will be used as the primary indicator of private social capital in this dissertation, how this concept is measured will be described later. 6 For the discussion of question wording effects, please see Schuman and Presser (1996). 7 Note that the purpose of comparison is to observe the change of social trust level in different time fiames, which is one of the questions this dissertation aims to answer. For details, please see chapter seven. 104 other. 8 Still, respondents were asked in 1998’s Taiwan Social Change Survey if they think “most people can be trusted?” The percentage of those who answer in favor of trusting people is used as the aggregate level indicator of social trust in these years in order to compare the level of public social capital in different regime types that Taiwan has experienced. As far as individual level analysis of public social capital and its political consequences, which consists of the second part of the analysis in chapter 7, a factor analysis of the following two survey questions in 1998’s Taiwan Social Change Survey was conducted and produced one significant factor (eigenvalue=1.28), which explains 64 percent of the variance: “do you agree that people can be trusted?” and “do you agree that most people are willing to help and only few people are selfish?” The standardized factor score will be used as the primary indicator of social trust. Another measure of social trust in chapter 7 is made from 1996’s Taiwan Social Change Survey. The reason for another measure of social trust is that 1996’s data set provides a measure for the effect of depth of involvement in voluntary associations on the level of social trust, which cannot be measured using 1998’s Taiwan Social Change Survey.9 When using 1996’s data set to test the relationship of these two components of public social capital, social trust is measured with two survey questions: “can most people be trusted or can you not be too careful in dealing with people?” and “Most people are willing to help or they only mind their own business?” A significant factor was extracted 8 1996’s Taiwan Social Change Survey asked positively instead of negatively. It asked respondents if they agree with the statement: “it has become a tendency that people can trust each other”. 9 1998’s Taiwan Social Change Survey can only help us test the relationship between membership in voluntary associations and social trust, too tenuous to disprove the effect of voluntary associations on social trust if no relationship is found in the results of regression. 105 Sun 1'1 01: s 1158 3550- iii If}: iii it Hit from the factor analysis (eigenvalue=1.19), which explains 59% of the variance. Note that the two questions’ wording is similar to those used in 1998’s analysis but not identical. Membership in voluntary associations: Taiwan University’s Legislative Election Surveys of 1983, 1986, 1989, and 1992 asked respondents if they participate in any social organizations. The question wordings are the same across these surveys. Therefore I will use these data sets to calculate the aggregate level of membership in voluntary associations in Taiwan. On the other hand, the Taiwan Social Change Survey also provides us with similar information. However, the question format is slightly different. For example, in 1993’s, 1996’s, and 1997’s Social Change Surveys they showed a list of different associations to the respondents and asked in which of those associations they participate. Thus, in order to calculate the percentage of membership we must check each respondent and see if he or she ever took part in any of those associations listed. Finally, in 1998’s Social Change Survey respondents were asked if they participate in any voluntary associations or maintain any recreational associations with their relatives, friends or colleagues. The percentage of those participating in voluntary associations will be used as another indicator of public social capital, and again these statistics will serve as the basis for comparing the level of change in public social capital. In the individual level analysis, thel998’5 Social Change Survey’s question regarding whether the respondent participates in any voluntary associations will be used as the indicator of membership in voluntary associations. As to the depth of involvement in voluntary associations, which can be measured in 1996’s Taiwan Social Change Survey, three levels of involvement are recognized: membership, participation in the activities, and volunteering in voluntary associations. 106 ik‘Pf‘O'. , " ‘ ‘ _ pp“, 2" -. . . .,,) 3",I'p'9001b.l". .2. . These three variables have been dichotomizedl0 by the author and will be used in regression models employing 1996’s Taiwan Social Change Survey data to test the effect of voluntary associations on social trust. As can be seen, one of the limits of the aggregate measures of public social capital is the continuity of question wordings. To solve this problem is beyond the author’s ability. However, due to the similarity of wordings, we can still perform a preliminary comparison on whether the level of public social capital has made a shift between different regime types. Private Social Capital As indicated in chapter three, private social capital in this research is referred as the resources embedded in an individual’s social relations. The 1997 Social Change Survey tried to measure two different types of individuals’ social networks: expressive and instrumental. In measuring expressive social ties, the respondents were asked to list at most five people they can talk to when having troubles in social relations with others. Then they were asked to provide information about the backgrounds of these people and the patterns of relationships with respondents, such as how did they get to know each other, how long has the relationship persisted, the most frequent means of keeping contact, how often did they contact each other, how close the relationship is, and how important is this relationship to their life. Similarly, in measuring instrumental social '0 The original questions ask the respondents 1) whether they participate in a variety of voluntary associations; 2) whether they were members of those associations; and 3) whether they had ever volunteered 107 networks, the respondents were asked to list at most five people to whom they resorted in the past year when troubles arose in their career, financial situation, family matters, legal problems and health problems. These help-givers’ personal backgrounds and how the relationship was kept with respondents were measured as well. Of the two types of networks, I will use instrumental networks as a respondent’s private social capital although we can never underestimate the importance of expressive ties in easing one’s life. The reasons are twofold: first, according to the definition used in this research, private social capital is a resource embedded in one’s social relations. This resource can be useful only when appropriated. It is also particularly relevant to the empirical analysis, for we can only make inferences from that which has already happened. By judging from the people one asked for help in order to reach certain goals we can better understand the concept of ‘capital’ in the context of social relations. Second, over 60% of the contacts made in the instrumental ties reappear in respondents’ expressive ties.ll I further categorize one’s instrumental networks into two different types of ties, strong ties and weak ties.12 Regarding the measure of strength of ties, the literature reviewed offers several ways to operationalize the concept.'3 Frequency of contact is sometimes used to measure the intensity of friendship. Some classified an individual’s relatives or friends as strong ties, and acquaintances as weak ties. Still some identified a respondent’s acquaintances or friends of friends as weak ties, and friends, relatives, or for those associations. Thus if a respondent had at lease one positive answer to any of the questions, he or she would be categorized as ‘yes’ in that level. ” The survey asked respondents whether the contacts listed under expressive ties reappeared in any of the five contacts for the instrumental ties. The author's results are based on this information. '2 As to the strength-of-weak-tie argument, please see chapter six for details. '3 Granovetter defined the strength of an interpersonal tie as “(probably linear) combination of the amount of times, the emotional intensity, the intimacy (mutual confiding), and the reciprocal services” (1973:1361). He admitted the difficulty of operationalizing the concept (Ibid). 108 neighbors as strong ties (Granovetter, 1973:1371;Granovetter, 1982:110,ll8). In the context of Taiwan one’s instrumental networks will be characterized as strong ties if all the people he listed as help-givers are either relatives or good friends; one’s instrumental ties will be categorized as weak ties if at least one of the help-givers listed is from other categories of relation to the respondents.'4 To summarize, the political consequences of private and public social capital will be analyzed in different chapters using two different data sets. In chapter seven, different periods of aggregate levels of public social capital will be compared. Due to the problem of continuity of question wordings, the results will be interpreted with caution. As to other variables used in the analyses, they will be given detailed descriptions in the appropriate chapters of how they are operationalized. '4 I also tried to construct this variable using frequency of contact as the criterion. This measure is positively correlated with the one constructed based on types of relationship at r=.2 1. According to the preliminary regression analysis, the results are very similar in terms of hypothesis testing. Therefore, I will use types of formal relations instead of frequency of contact as the criterion to dichotomize one’s instrumental ties. 109 CHAPTER 5 STRUCTURAL SOURCES AND CONSTRAINTS OF SOCIAL CAPITAL IN TAIWAN INTRODUCTION Chapter three covered the nature of both private and public social capital and proposed two general analytical frameworks for them. Before the empirical analysis, it is helpful to provide a profile of Taiwan’s cultural heritage, major historical events, and political development. These factors are necessary for understanding the case. They also serve as important sources of social capital in any society. In this chapter I will discuss the above three broad sources (or constraints) of social capital with special emphasis on their implications for the formation of private and public social capital. TAIWAN’S CULTURAL HERITAGE AND SOCIAL CAPITAL ‘Culture’ is a term generally used by social scientists in reference to the symbolic and learned aspects of human society (Marshall, 1998:137). It has at least two dimensions: one is the interpersonal ties and relational networks in which people are embedded, while the other includes the symbols, language, values and assumptions people use to interpret their world.I According to this conceptualization of culture, we can infer that culture in any society plays an important role in shaping its members’ ' In Laitin’s book, he suggested “two faces of culture”, mentioned above, to make the study of culture more 110 behavior in social interactions and in providing a frame of reference (or worldview) when facing uncertainties. These roles of course will affect the formation and level of social capital embedded in a particular society. In this chapter I will limit my discussion of Taiwanese culture to those aspects most related to the concept of social capital. Since Taiwan’s history and that of Mainland China once ran the same basic course, its cultural heritage is highly influenced by Chinese culture (Davison & Reed, 1998:34; Huang, 1998:32-36). Therefore, I will use Chinese culture as Taiwan’s cultural heritage.2 Chinese Society: A Highly Structured One Oriental, especially Chinese, societies and western ones, differ greatly. Chinese society is organized around a set of values called ‘luan li’3, which are deeply influenced by Confucianism. The focal point of this societal map is family. Family is not only the origin of interpersonal relations but also a school of social relations. Specifically, an individual learns the norms of interacting with others from parents, and these norms are expanded to the greater society as well. These norms or values can best be summarized as Confucius’ ‘Five Cardinal Relations’ (wu luan)’. Among these interpersonal relations, theoretically refined. l adopt his classification of culture and use it as my guide of discussion in this chapter. 2 As to the public perception on the difference between Chinese culture and Taiwanese culture, the Taiwan Social Change Survey of 1998 may provide a yardstick. Around 40% of the respondents replied ‘no difference’ and among those 60% who said ‘difference’, only 35% saw significant difference between the two cultures. In other words, around 80% of the total respondents thought there is no or little difference between the two cultures. The figures are calculated by the author. 3 ‘Luan’ means relations originating from family life, while ‘Li’ refers to the actual reflections (or expression) of the affects of the relations upon family members. See Liang (1967: 80-81). ‘ Wu luan specifies five kinds of relations that an individual will encounter in his life, including the ways to deal with parents, brothers or sisters, spouse, superiors or subordinates, and friends. As to the evolution of 111 three are focused on familial relations, namely, “father should be righteous, mother should be loving, elder brother should be friendly, younger brother should be respectful, sons should be filial” (Yueh, 1989:133). Based on these norms, Mencius, the second great Sage in Chinese history, expanded these social relations to include those between superiors and subordinates and friends. Thus Chinese are taught to show the appropriate attitude on five cardinal relations: affection between father and son, righteousness between ruler and minister, attention to appropriate roles between husband and wife, proper hierarchy between older and younger, and faithfulness between friends (Weller, 1998:243). In sum, wu luan has acted as the basic guideline for Chinese social behavior for thousands of years. The above norms guiding the Chinese’ social behaviors have several characteristics deserving more detailed discussion. First, family life5 is the center of social activities. Within the family, a hierarchy is imposed on the members, with the father having unchallenged authority. Due to this emphasis on family, almost all other social relations emulate this paternalistic family structure. For example, emperor is regarded as the father of his subordinates, local officials is also regarded as father in his district. Similarly, those subjects (ordinary people) at different administration levels regard themselves as the sons of their superiors. The teacher is also treated as father of his students. Ordinary people call themselves as brothers and sisters. The whole world is said to be originated from and should work as a big family. In brief, Chinese people live in a highly structured society with special emphasis on hierarchical social orders. Second, an individual is born within a given interpersonal networks, from which this idea, see Yueh (1989: 133-139). 112 lib ta; ll he is not allowed to deviate. Therefore social obligations outweigh social rights. Similarly, the sense of self gives way to the sense of family. Survey data confirm this. Table 5-1 shows that over 80% of the respondents think a stable family life outweighs personal achievement. Table 5-1 Percentage of Respondents Who Think Personal Achievement Is More Important Than A Stable Family Life" n % Strongly agree 34 2.19 Agree 209 13.47 Disagree 1 096 70.62 Strongly disagree 151 9.73 Source: The Social Image Survey in Taiwan, February 1993. ‘Original question is “In modern society is personal achievement more important than a stable family life?” Third, the state is never treated as an institution equal to family. This means public policies don’t receive much attention by the ordinary people. In addition, the ability to organize and demand civil rights or pursue self-interests is extremely lacking in Chinese society. Fourth, interpersonal relations are governed by rites6 instead of laws. Hence people interact with others not according to the rule of fairness but to the party’s social status relative to him, which results in fairly unequal interactions. But the third and fourth implications are the ones worth more detailed discussion under the framework of social capital analysis. 5 Family here refers to not only the concept of nuclear family, but also includes kinship ties. 6 ‘Rites’ refers to various grades of courtesy due to differences in the social ranks. 113 How Do Chinese Interact With Others: The Power of Guanxi It is not possible to grasp the essence of social relations in Chinese societies without knowing the concept of ‘guanxi’. Guanxi is broadly understood in Chinese context as any kind of connection that an individual can take advantage of for the sake of self-interests. Guanxi can be cultivated from a wide variety of possible ties, ranging from parental and sibling ties to other types of bonds such as neighbors, classmates, and co- workers (Bosco, 1992: 167). Why do Chinese people put so much weight on guanxi in their interpersonal relationship? As I mentioned, Chinese society is highly structured. Interpersonal networks impose complex social patterns for an individual to follow. Besides, paternalistic family structure nurtures a society ruled not by objective law but by the virtue of rulers. This pattern places the rulers in a better position to attain most social resources. Therefore, anyone who wants to survive must, to various degrees, attach themselves to resource holders. Guanxi, thus, becomes a must to survive. The next question is how does guanxi operate in this highly contextual society? There are basically three different social groups that an individual has to engage in social exchange: family (expressive ties), associates (mix ties), and strangers (instrumental ties).7 Chinese treat these three groups with different criteria due to the degree of acquaintance. For family members, the ‘need rule’ is applied, i.e., “every member should do his best for the family, and the family will in turn supply him the 7 Here I borrow Hwang’s basic conceptual framework (1987). 114 resources necessary for living.” (Hwang, 1987:950) Trustworthiness is therefore easy to establish among members. For strangers, Chinese are extremely cautious in dealing with people they do not know. That is because there is no rule or norms in Chinese culture telling people how to treat strangers.8 Given this, the logical solution is back to the ‘equity rule’ or the ‘law of the jungle’ when engaging social exchange. Here rational calculation predominates social exchange between strangers. Empirical evidence also supports this argument.9 Accordingly, this attitude, as opposed to Westerners, also results in the unwillingness of Chinese people to help strangers.lo Given this apathy towards strangers, how do Chinese people try to utilize strangers’ social resources once they feel the need? This is where the game of ‘guanxi’ comes in. The game typically begins with an individual (X), who needs a resource from a stranger (Y). X can achieve this goal by pulling on his relationship with an associate (Z), who also keeps relationship with Y. As long as Z in indebted to X and Y is indebted to Z, X is very likely to get the help of Y through Z (Bond, 1991 :59). This social exchange involves several critical concepts: first, it fulfills the norm of ‘bao’ (reciprocity). Z repays the debt he owes to X by allowing Y to 8 Notice that ‘wu Iuan’ (the five cardinal principles in social relations) did not include the category of strangers in social relations. Strangers have no place in Chinese social logic. It is in this vacuum that there are no constraints beyond self-interest. Accordingly, “[W]e may say that from birth, a Chinese person is enclosed by a network of interpersonal relationship which defines and organizes his existence, which controls his Heart-and Mind. When a Chinese individual is not under the control of the Heart-and Mind of others, he will become the most selfish of men and bring chaos both to himself and to those around him.” (Sun Long-ji, cited from Bond, 1991257) 9 In a cross-cultural experiment, Bond and Leung (1983) find that Chinese individuals tend to be particular about trifles and to behave rationally when interacting with strangers and be able to use more equitable strategy of relating inputs to outcomes than did Americans. '0 Hwang and Harris (1974) conducted an experiment including Taiwanese and Americans in New Mexico and found that Taiwanese are more unwilling to help a stranger send a letter for him than American counterparts. In addition, there is a well-known phrase advising people to “only shovel snow in front of your house, never mind the frost upon another’s roof.” This also reflects that Chinese people are selfish and apathetic in 115 repay the debt owed to Z indirectly by granting Z’s request to help X. Of course, since then X is indebted to Y and this could be used for next round of social exchange. Second, when X asked Z to request Y’s help, we call X’s behavior ‘la quanxi’ (pulling his social connections) Third, Y’s helping X is called giving ‘renqing’ (favor) to X.ll Fourth, after this social exchange, X and Y are longer strangers. They now become members of ‘weak tie"2 in each one’s social network. In brief, Chinese usually avoid direct contact with strangers. Rather, they use their social connections (guanxi) to pull those strangers who they think they have to encounter for whatever the reason is into their own social ties. It not only avoids the possible conflict or embarrassment due to the rejection of others'3 but also fulfills the norm of reciprocity among associates. This leads us to the scrutiny of social relations among associates. For the category of associates (or acquaintances), which stands between family and strangers in terms of intimacy, Chinese people use the rule of so-called ‘renqing’.l4 Renqing consists of a set of norms guiding people’s social interaction with their associates. There are at least three phases that characterize this game of renqing. First, it is rational for people to help each other within their social networks due to the strong dealing with strangers. " Specifically, ‘renqing’ here refers to “a certain kind of resource that can be used as a medium of social exchange.” (Hwang, 1987:954) '2 As to the origin and meaning of weak ties, please refer to chapters one and six of this dissertation. For now the most concise definition of weak ties includes those we are acquainted with but not deeply involved with in our daily life. '3 In this case, we call the one making request ‘losing face’ (diu lion) because the one who is asked to help does not ‘give face’ (gei lion) to him. Again, face here is close to the Western notion of ‘dignity’. ” Strictly speaking, there are at least three different meanings of renqing in Chinese context. First, it represents the emotional responses of an individual confronting the various situations of daily life. Second, it refers to the substantial assistance incurred in social exchange. Third, it implies a set of social norms that people must follow in social interactions among associates (Hwang, 1987: 953-954; Lu, 1996: 140-141 ). This chapter will only focus on the rational dimension of this concept. 116 mini 116ml Sociz alas emphasis on ‘bao’ (reciprocity)‘5 in Chinese culture. Second, it is exactly this expectation that make people more sensitive to the response of the beneficent. This concern frequently creates a dilemma for those asked to give help. For they cannot control the length of time needed for the recipient to reciprocate given the fact that they have to respond to the request, i.e., giving help, immediately.’6 Third, since Chinese people tend to place the harmony of the inner group higher than their immediate interests incurred in social relations”, interpersonal relations tend to focus on the allocation of resources by manipulating favors (renqing) and face, which resultantly forms a very complicated network in terms of strategic thinking. It also directs people to center their sight on specific individuals rather than on the issue involved. For example, if something happens, Chinese people are prone to first think of who is involved in it rather than whether it is justifiable or correct. Consequently, if people have troubles, they tend to think about whom they can ask for help even if it is not allowed by law. Thus the rule of renqing dominates and the rule of law is not so strictly enforced in Chinese society as it is in the West. What implications can we infer from the above-mentioned Chinese patterns of social relations in terms of the possibility of collective action? Are Chinese people good at association to achieve collective purposes? This leads us to the next section. '5 Although the norm of reciprocity is a universal one (Gouldner, 1960), Chinese society seems to be especially sensitive to this behavior, probably a result of the requirement of filial piety in family, people are expected to reciprocate favor they receive with more in quality or quantity. A famous saying is ‘if you receive a drop of beneficence from others, you should reciprocate with a fountain of beneficence.’ '6 As to the relative strategies dealing with the dilemma, see Hwang (1987) '7 I believe the concern of getting face, a combination of dignity and honor, in social life sometimes outweighs the immediate personal interests incurred in social relations and may explain why this happens. For empirical evidences, see Chu and Yang (1976) and Bond, Leung, and Wan (1982). 117 Associational Life in Chinese Societies The ability of people to organize or participate in civic organizations has been regarded as a form of social capital. It is claimed that this kind of public social capital can not only relieve the dilemma of collective actions but also create a virtuous cycle for the development of society.18 Does Chinese culture foster and support this kind of social capital formation? As I mentioned previously, Chinese societies are organized around the notion of luan li, which is in fact the extension of the family mode (Liang, 19632200). Social interaction follows the norms of family. Given this setting, the western notion of voluntary associations, aiming to pursue collective goals, is relatively strange to most Chinese. This is because an individual living in a society based on the notion of luan li will assume obligations on the one hand to people involved in his social networks and on the other hand will expect care from the same set of people when in trouble. In other words, the unique pattern of ‘from family to the universe, and induce the universe to the circle of family’ blurred the boundary of society, which accordingly reduces the need for voluntary associations to pursue public goods. In addition, any potential association that needs to ask its members to sacrifice the immediate interests of their families will not survive in the first place.19 This greatly reduces the chances of forming associations for public good. Another factor affecting the formation of voluntary associations is from the '3 For details, please see chapters two and three of this dissertation. '9 A simple explanation (example) of why Christianity cannot flourish in Chinese societies is because it forbids Christians to worship their ancestors, which clashes with traditional Chinese values. 118 traditional attitude towards ‘party’.20 We have to notice a lofty idea regarding the ownership of the world or universe in Chinese political thoughts, i.e., the world is owned by the public. Given the influence of luan Ii, individuals are encouraged formally” to behave unselfishly in social interactions. Hence, group activities involving any organizations or parties will naturally be regarded as ‘pursuing private interests’, which runs against public good. This may also explain why there is no such concept of civil society as the West has.22 Forming parties and pursuing private interests are always associated in Chinese daily language, which is called jie tang ying sy, especially when used in the political domain. My arguments above can be summarized in three points: first, the more the people in a society depend on the help of family (kinship), the less the need and significance of voluntary associations. Second, when family structure is stable and the functions of it are comprehensive, people tend to focus their lives on how to improve the well-being of the family, and hence public goods are ignored and remote, although lofty, in their life. Thus collective actions will be more difficult to achieve than otherwise. Third, ‘forming parties in order to pursue private interests’ is traditionally disdained by the general public in Chinese societies. In addition, there are many examples from Chinese history in which parties formed were oppressed by the central authority.23 2° The ‘party’ here refers broadly to gathering for self-interests. 2' I believe this requirement of unselfishness is an ideal pattern of social interactions praised and taught from generation to generation. However, it is not realistic to expect everyone to follow this idea. That is why I use the word ‘formally’. 22 I agree with Fukuyama’s argument (Fukuyama, 1995) that strong familism is the most important factor preventing China from having a strong civil society. But the traditional resentment of ‘parties’ should be another important determinant explaining why Chinese societies have no strong civil society like the West. 23 For instance, intellectuals in the late Ming Dynasty tried to form parties and collaborated closely with university-like institutions to monitor the central authority. Those efforts were severely oppressed by officials (Yueh, 1989: 107). 119 Horizontal Social Ties in Traditional Chinese Societies Chinese traditional culture does not encourage people to associate for the public good. But that does not mean Chinese societies lack such horizontal social ties as those called voluntary associations in the West. I will briefly sketch the profiles of these horizontal ties and their functions. In general, horizontal ties in Chinese societies are limited to the local level and are determined mostly by either lineage or native birthplace. There are five broad types of associations existing in Chinese societies. The first one is formed based on the lineage or same surname called tzong qing huey.” These organizations traditionally serve as semi- judiciary institutions in solving conflicts among its members and as intermediary agent in providing manpower, food, and taxes needed by the state (Yueh, 1989:85-86). The second is formed by people native to the same area. The smallest unit of this kind of organization is called Village. It could be naturally established (which is closely knitted with lineage or imposed by the administration). For those coming from the same native place but traveling or doing business in other places including overseas, they tend to form tong xiang huey as a place for exchanging social support. The third kind of association is established based on occupation. These organizations are voluntarily formed by those who are of the same occupation. They have to fulfill certain obligations, which are supposed to benefit all involved. The fourth kind of association is related to ‘same study’ 2" Strictly speaking, tzong qing huey refers to associations within which members not only share a surname but also traceable kinship. Another type of tzong qing huey is between those with the same surname but with no traceable kinship. This is because Chinese believe people with the same surname originated from the same ancestors, therefore should be regarded as the offspring of the same family regardless of 120 background. People usually feel close when they study in the same school or under the same teacher. School is the second most important place for Chinese to learn social values. Within the logic of luan li, school is treated as the second family of the individuals. The teacher normally plays the role of father in school. Students in the same school but different grades usually follow a hierarchy similar to the family, i.e., students with higher grader are respected by those with lower grader just like elder brother or sister is respected by their younger brother or sister in the family. This relationship exists formally in the name of tong xue huey or even informally after graduation. Thus solidarity among those sharing a study environment or experience forms a reserve of social support, especially in finding jobs.25 The last kind of association can be broadly categorized as belief groups. They include religious groups and secret societies. These associations, represented by temples, bang, or huey, operate mostly at local levels. They not only bring people together to worship their own gods, appease wandering ghosts, celebrate marriages and mourn deaths (Weller, 1998:234), but they also serve other political functions such as mobilization of the public.26 For secret societies, they are primarily voluntary pietistic associations whose members are from the lower class. Their functions range from self help to fulfilling justice in the name of Heaven. It is the latter that makes it a potential threat to the state (Weller, 1998:235; Yueh, 1989:109). geographical dispersion. 2’ This phenomenon is further supported by the fact that there are many factions in work places. Jobseekers tend to have more leverage than others if they know seniors in the company or government agency. In the long run, we are likely to see factions based on either same native place or same study background in the work place. 2" In the case of Taiwan, local temples are the center of gatherings. This is especially true for people living in rural areas. Each temple has its own committee led by local elites. Those local elites have a tremendous influence on the formation of public opinions. They also cooperate with local factions to mobilize people at 121 Iona llstor them CCIIIEI dtl e1 life I] IIIOIR’ 318i gin and of 311E lit] Summary Following is a brief summary of Chinese cultural influences on social capital formation: First, although Taiwan has been ruled by different countries in her 400-year history, the closest connection it has is with Chinese culture. Therefore, Chinese culture is the major foundation shaping Taiwanese thinking. Second, Chinese societies are organized under Luan Li, in which family is the center of gravity for social interactions. All other relationships radiate from the norms developed from the family life. Family matters in analyzing Chinese societies. Third, Chinese culture, represented Confucianism, emphasizes linkage of social life more than the autonomy an individual can enj oy.27 Fourth, the game of guanxi dominates Chinese social life, which make people more sensitive to interpersonal relationships. Fifth, strangers do not exist in Chines social relations guidelines. Thus Chinese are extremely cautious in dealing with strangers. Sixth, Chinese culture seems to discourage people’s ability to convene for public good. Lastly, culture is not immobile. It reproduces and transforms itself by both shaping and absorbing social life. Therefore, cultural background can only provide us an overview of what we should expect in a society. Other factors should also be included in our analysis of the sources of social capital. various levels of elections. 122 TAIWAN’S HISTORICAL DEVELOPMENT AND SOCIAL CAPITAL An important issue facing Taiwan now is the problem of national identity.28 It is especially significant after the Democratic Progressive Party (DPP), the largest political opposition party in Taiwan, won the presidential election in March 2000.29 This issue is further intertwined with sheng-chi wen-ti (problems of different provincial origins or ethnic cleavage). I believe these two issues will severely impede the future of social capital formation in Taiwan. In this section, I will present a brief history of Taiwan with particular focus on events leading to current problems of national identity and ethnic cleavage. Taiwan’s History Before 1945 Taiwan is an island about 245 miles long from northeast to southwest and 90 miles across at the widest point going from northwest to southeast. It is about 100 miles across the Taiwan Straits from China’s Fujian Province. Until the sixteenth century, the dominant groups living in Taiwan were aborigines, descent of Malay-Polynesia (Davison & Reed, 1998:6; Clark, 1989252). Although Taiwan was recorded in Chinese history as early as 230 AD, it received increased attention from Chinese people and other countries 27 For the meaning of linkage and autonomy in social capital literature, please refer to chapter three. 28 Please see Table 5-3. 2’ DPP is known as the advocator of Taiwanese independence. The party seeks to sever the connection between Taiwan and Mainland China, and their ultimate political agenda is to establish an independent country, i.e., Republic of Taiwan. 123 starting from the end of sixteenth century. Pirates and businessmen from China and Japan rushed to the island to do business with aborigines. At the same time European countries such as the Netherlands and Spain also showed high interest in the business and strategic advantage that Taiwan possessed. In 1642 Dutch forces won the battle with Spaniards and claimed jurisdiction over Taiwan. Chinese government at that time still had no intention to claim sovereignty over Taiwan. This situation changed when the Manchu ousted the crumbling Ming dynasty and founded the Ch’ing dynasty in 1644. Cheng Ch’eng kung, a Ming loyalist, resisted Ch’ing rule and decided to retake Taiwan from the Dutch and use it as a base to battle against the Ch’ing dynasty. He and his troops successfully ousted the Dutch in 1662. Within the next thirty years, Cheng and his son basically established the institutional, social and cultural foundations of Chinese civilization on Taiwan (Gold, 1986225). In 1683 the Ch’ing dynasty defeated Cheng’s regime in Taiwan and made Taiwan a prefecture (fu) of Fukien province. Since then, a greater numbers of Chinese started to immigrate to Taiwan. An estimated two million Chinese resided in Taiwan at the turn of the century (Clark, 1989254). Among those immigrants, the majority are from southern Fukien province, i.e., chyuan zhou and zhang zhou, and northern Kwangtung province, especially mei xian. Those from Fukien province came to Taiwan earlier than people from other provinces and therefore the term Taiwanese are traditionally reserved for them and their offspring. They also hold the largest population relative to other ethnic groups in Taiwan. For those originating in Kwangtung province, they speak a different dialect than the Taiwanese. They came to Taiwan later than the Taiwanese and therefore called themselves Hakka, i.e., guests. The settlers stayed together based on the same surnames, 124 ancestral place, dialect, or belief in a common deity. However, due to the lack of effective government control, competition for cultivating superior lands, and the language barrier, Taiwan in the nineteenth century was plagued by deep social unrest, i.e., armed battles, between cultivators. Thus hatred and strife divided Fukien immigrants, Kwangtung immigrants and the aborigines. In 1895, Taiwan was ceded to Japan as the result of Sino-Japanese War. During the 50 years of Japanese rule, Taiwan’s infrastructure and transportation were greatly improved.3°On the other hand, the colonialism established a structure of rigid control over Taiwanese society to prevent any further outbreaks from happening. Top governmental positions at all levels were held by Japanese (Gold, 1986:35; Davison & Reed, 1998219). The education system also discriminated against Taiwanese. This institutionalized inequality between Japanese and Taiwanese combined with the improved communication within the island lessened the hostility among ethnic groups and began to form a sense of local (Taiwan) identity as opposed to Japanese rule. Thus the ethnic conflict of the Ch’ing dynasty was replaced by nationalism and solidarity. Starting from 1919,31 many anti-colonialism movements appeared in Taiwan, which can be seen as the first sign of a quasi-civil society.32 However, these movements were terminated as a result of the war between Japan and China. Taiwan was militarized by Japan and dissent and social activism were suppressed (Gold, 1986244). 3° Japan’s purpose in Taiwan was to improve Taiwan’s agriculture in order to supply Japan’s domestic needs for foodstuffs and raw materials. For details, see Clark (1989) and Gold (1986). 3‘ In 1919, Japan sent the first civilian govemor-general, Baron Den Kenjiro, to Taiwan, which marked the liberalization of Japanese rule. Under the liberal climate, Taiwan’s non-armed anti-colonialism movements were encouraged to make their voice heard. Most of these movements’ appealed for political equality for 125 Taiwan After 1945 In August 1945, Japan uncondictionally surrendered to the Allies. Taiwan was returned to China according to the Potsdam Declaration of July 1945. The results were mixed for Taiwan. During the war Taiwan was subjected to heavy American bombing and thereby lost “three-quarters of industrial capacity, two-thirds of power, and one-half of the transport network” (Ranis, 19792209). Besides, the mass evacuation of the Japanese after the war also took away most of the skilled administrative and business leaders (Clark, 1989261). Initially, most Taiwanese people welcomed their Chinese ethnic kin to take over Taiwan, hoping to be treated equally. But the euphoria of reunion with the motherland soon shattered. Many factors contributed to the disillusionment of Taiwanese people and the unfortunate tragedies that happened. The fundamental cause lies in the clash between the expectation of Taiwanese on returning to China and the nationalist government’s strategic consideration on taking over Taiwan. Taiwanese showed deep respect to China, their cultural progenitor. Nevertheless, when they compared the fifty-years of Japanese rule with that of the newly arrived Chinese troops and bureaucracy, the result was unfortunately in favor of the Japanese rule. This needs to be explained in light of the situation facing the nationalist government and what they really knew and thought about Taiwan. Taiwanese (Gold, 1986243). 32 For details of these organizations and movements, see Hsiao, Huang, and Weng (1995). 126 The Chinese Government’s Attitude towards the Taiwanese The Chinese government, let by Chiang Kai-shek and his nationalist party, i.e., Kuomintang (KMT), was not serious enough in taking over Taiwan after defeating Japan. This is partly due to Taiwan’s history to China. Taiwan was regarded as culturally backward by the Nationalists. The fifty years of Japanese rule also made the Nationalists believe that the Taiwanese were not purely Chinese and needed to be educated and corrected. This stereotype made those Chinese troops and bureaucrats sent to Taiwan look down upon Taiwanese people in the first place.33 This is reflected in the arrangement of personnel in reconstructing Taiwan. Although most of the Taiwanese elites were willing to cooperate with the new government, their Japanese educational background was deemed as a highly suspect by the new government. Thus, in addition to the mainlanders, those Taiwanese who had spent most of their time living in Mainland China,34 especially during the eight years of war with Japan, occupied almost all the central level positions. Local Taiwanese were only staffed in the lowest positions (Gold, 1986250). The feeling of being ignored or even discriminated soon spread widely among Taiwanese. Another reason the task of rebuilding Taiwan was ignored is the armed battle between the KMT and the Chinese Communist Party (CCP). Preparing for the decisive battle with the CCP in Mainland China, Chiang Kai-shek could not send troops of quality to take over Taiwan. In addition, this situation left a political vacuum for opportunists, 3’ For details, see Jiang (1995). 3‘ Taiwanese with a study or work background in China during wartime were called pan shan (half- mountains) by their fellow Taiwanese in comparison to the mainlanders, which were called An-shan (mountain people). In fact, the ‘mountain’ refers to China since local Taiwanese, traditionally, like to call China Tang-shun (Mountain Tang). 127 both from China and Taiwan, to exploit Taiwan’s social assets (Wakabayashi, 1996:68- 69). Furthermore, the infamous inflation due to civil war in China spread to Taiwan, which deteriorated its economy and decreased its living standards drastically.35 All the aforementioned factors, including prejudice, distrust, and other contingent events presuppose the unhappy relations between the new government and the governed.36 A decisive historical incident eventually happened under this political climate. The February 28‘II incident (2-28 Incident) On the evening of February 27, 1947, a group of Monopoly Bureau agents in Taipei beat a woman selling cigarettes not produced by the Monopoly Bureau.37 A man who came to her defense was shot to death. Crowds of angry people gathered in the streets of Taipei protesting the violent behavior of government officials. Crowds clashed with police and troops. Some angry protesters even attacked innocent mainlanders as scapegoats. Starting from the first of March, major cities’ police stations were occupied by Taiwanese as their bases for confrontation with KMT troops. A settlement committee, composed of mostly provincial legislators, was formed on March 2 and came to a “Thirty-two Demands” asking Chen Yr, the Administrator-General and Garrison Commander, to engage in political reforms, including lifting marshal law, fighting 35 The net domestic product in Taiwan in 1946 was only 55 percent of the prewar level, with the year of 1937 as the base of comparison (Kuznets, 1979234). For other index, see also Koo (1973:402). 36 There were a lot of posters all over the island saying, “The dog (Japanese) has gone, but here comes a pig (mainlander).” Taiwanese intellectuals also started to question the legitimacy of the rule of nationalist government from China. See Wakabayashi, (1996269). 7 The quality of cigarettes then manufactured by the Monopoly Bureau was not welcomed by consumers. Many tried to smuggle foreign cigarettes and sell them on the underground market. Agents fiom the Monopoly Bureau and police station were sent to streets to investigate and arrest the suspects 128 corruption, implementing Taiwan self-govemance, and recruiting more native Taiwanese into the government (Ibid, 73-74). Chen Yi promised to make reforms and not to bring in additional tr00ps. However, on May 8 around 13,000 troops from mainland arrived in Taiwan.38 With the added military aid, Chen Yi broke his promise and launched systematic repression against native Taiwanese. A discernible pattern of this military repression is that “the Nationalists intended to liquidate the Taiwanese intellectual and social elite. They went after the Settlement Committee members, teachers, students, newspaper editors, lawyers, and anyone considered critical of the government” (Gold, 1986:51). The actual people who died in this incident are not knowable; different estimates ranged from 10,000 to almost 30,000 (Gold, 1986:51; Wakabayashi, 1996275). In the wake of this tragic incident, the KMT government strongly oppressed any attempt to discuss or memorize this incident publicly until the late 19805 when political liberalization began. But the history of the February 28th Incident has long been an important collective memory for the Taiwanese people. Table 5-2 shows that Taiwanese are significantly favor that 2-28 Incident be taught in schools and serve as an example to educate the next generation. This tragedy and its implication are still stressed by many Taiwanese politicians with an eye to siphon votes from Taiwanese as opposed to the KMT candidates in various levels of elections. (Wakabayashi, 1996:73). 3"The Chinese Communist Party openly supported Taiwan’s armed struggle against KMT’s government (J iang, 1995262). In addition, Chiang Kai-shek seemed to foresee the failure of the KMT’s against the CCP. KMT’s government needed a last resort just in case of losing the civil war. These factors increased the strategic value of Taiwan and hence the KMT did whatever it could to stabilize Taiwan’s social order. 129 Table 5-2 The Importance of the 2-28 Incident* to Different Ethnic Groups“ Taiwanese Hakka Mainlander Aborigine Not important 72(16.5)*** 61099) 68 (27.1) 6(28.6) in between 146(33.4) 97(3 1.7) 77(30.7) 8(38. 1) Important 219(50.1) 148(48.4) 106(42.2) 7(33.3) Chi-Square=13. l9, p<0.05. Source: The Social Image Survey in Taiwan, July 1994. ”fire survey question is “do you think the 2-28 Incident is worth being remembered and taught to the next generations?” Respondents were given a lO-point scale with 10 representing very worthy. For the convenience of analysis, the scale has been recoded by author into three categories as shown in the above crosstab. “Respondent’s ethnic background was decided by their father’s. Taiwanese and Hakka are sometimes called Taiwanese in the broadest sense for their ancestors came to Taiwan far earlier than mainlanders. In this analysis, I will differentiate the two unless otherwise defined. "*Parentheses show column percentage. Large Scale Immigration from China On December l0, Chiang Kai-shek flew to Taiwan as the result of losing the civil war to the CCP. Taiwan became the KMT’s last base to possibly fight back and reclaim Mainland China. In l945 Taiwan’s population was about six million. However, during the next four years, over one million refugees from Mainland China retreated to Taiwan as the result of civil war in China. Most were KMT troops, around 600,000 (Gan, Shiao, Chen, & Wu, 199825). This sudden influx of Chinese refugees further complicated the already soured relations between Taiwanese and their Chinese kin since the 2-28 Incident. In many ways, the KMT government basically followed the Japanese pattern of colonial rule although this was the result of many unintended historical events and may Military repression became an inevitable means. 130 not be what the government was originally expecting. The mainlanders seemed to replace the governing position of the Japanese, and Taiwanese were still treated as the governed. For example, top positions from the central government and public owned enterprises to local schools and administrations were all filled with mainlanders, just as their Japanese counterparts did. Second, the spontaneous segregation of living areas also deepened the mental chasm between Taiwanese and mainlanders (Gates, 1981:260-262). Specifically, Mainlanders occupied most of the houses left by the Japanese. Elementary school districts still followed the Japanese era. This further prevented mainland children from interacting with Taiwanese ones. Different dialectics also caused communication difficulties. Further, most of the dependents of the troops lived separately from the Taiwanese. They developed their own culture and had their own village called juan tsuen (military dependents’ village). They are closely controlled by the Ministry of Defense and were loyal to Chiang Kai-shek and the KMT.39 The apparent divide between Taiwanese and mainlanders began to blur due to the mandatory requirement of speaking kuo yu (Mandarin, or national language) in schools. Beside, as hopes of reclamation of the Mainland grew dismal, inter-racial marriage between mainlanders and Taiwanese increasedfofueling ethnic communication and integration. The third. boon on ethnic integration was the industrialization and ’9 For instance, residents of juan tsuen are traditionally regarded as an army of tie piao (iron votes) when it comes to the various levels of election. The majority of them are mobilized to vote for the ruling party, i.e., KMT, and are extremely hostile to the opposition party or political groups composed of Taiwanese. Because the dissenting groups or the opposition party, i.e., Democratic Progressive Party, always advocates overthrowing the KMT rule and building a new independent country. 4° After the KMT retreated to Taiwan, military personnel were not allowed to marry Taiwanese. The primary consideration was their determination to fight the CCP very soon hopes of reclaiming China. In other words, they had no intention of living in Taiwan for the rest of their lives. This situation changed when the KMT’s ambition became dismal due to the international politics, which will be mentioned in next section. 131 urbanization of Taiwan. It is a common perception in literature that urbanization and tolerance are positively related (Wirth, 1938; Stouffer, 1955). Urbanization increases the chances of interaction among people from different backgrounds, which accordingly imposes a pressure for people to accept heterogeneity and form a sense of relativism (Wirth, 1938:15). However, it is too soon and dangerous if we think, according to the three structural factors, that ethnicity would eventually become a minor issue in Taiwan. On the contrary, ethnicity and the related national identity problem became even sharper in the 19905 after the democratization of Taiwan. Table 5-3 shows that Taiwanese in 1994 were 3.6 times more likely to claim they are Taiwanese than the mainlanders; while the mainlanders are 3.3 times more likely to think they are Chinese than the Taiwanese. In addition, a pattern seems to emerge. That is, as of 1998, native Taiwanese lean toward the Taiwanese end of the spectrum while mainlanders are move toward the middle of the identity spectrum. I do not have newer data to track this changing pattern. But I do believe this issue will remain salient in the future. In brief, the memory of the February 28th Incident is still haunting the Taiwanese. The process of institutionalization of the political system in Taiwan exerts yet another counterbalancing force and keeps this issue as salient as any other in Taiwan. This will be discussed in the next section. 132 Table 5-3 National Identity* and Ethnic Background Taiwanese Hakka Mainlander Aborigine (94) (98) (94) (98) (94) (98) (94) (98) 1.Taiwanese 45.2%“ 49.6% 37.5% 30.5% 12.3% 10.8% 31.9% 54.5% 2. Taiwanese but 25.2% 31.4% 21.4% 35.3% 11.3% 32.8% 20.1% 9.1% also Chinese 3. Chinese but 14.0% 11.5% 15.5% 21.8% 23.9% 30.2% 16.8% 15.6% also Taiwanese 4. Chinese 15.6% 7.5% 25.7%12.4% 52.6% 26.3% 28.2%18.2% 1994: Chi-Square=21 1.3, p<0.01. 1998: Chi-Square=200.3, p<0.01. Sources: The Social Image Survey in Taiwan, July 1994 (N=1388) and Taiwan Social Change Survey 1998 (N=1751). _ ‘This crosstab is based on two survey questions with similar wordings that appeared in the above- mentioned surveys. The first question in the 1994 Social Image Survey is “what would be your answer if you were asked what your national identity is?” The other question in the 1998 Social Change Survey is “which of the following best fits your national identity?” Both questions offered the same options as shown in the table. I”Column percentages. Summary I want to highlight two points in summarizing Taiwan’s history. First, due to geographical location, Taiwan has played a marginal role in China’s history. Its experience of being ruled by different foreign countries makes its people suffer identity crisis. On the one hand, their Chinese origin makes them resist foreign invasion and forges a sense of being uniquely Taiwanese. On the other hand, the fifty years of Japanese rule successfully made most Taiwanese elites identify themselves as Japanese."l This 4' For example, former president Lee Teng-hui, a native Taiwanese, never felt ashamed to say he was 133 current is still fighting with the Chinese identity instilled by the KMT.42 To complicate matters, the new ruling government is run by the DPP party, an advocate of Taiwanese Independence, which represents less than 20% of the public opinion.43 Where the country will go is becoming a critical issue. Second, the 1945 take-over of Taiwan by the Nationalist government is a failure in terms of the initial reactions to slightly differing ethnicity. There were many gaps between the two sides, ranging from economic development to mutual understanding, which set the stage for the later clash. The Nationalist government possessed a prejudice toward the Taiwanese due to fifty years of separation, thinking of them as the Japanese emperor’s people instead of pure Chinese. In addition, fueled by the eight-year war with Japan, Chinese troops and bureaucrats assigned to take over Taiwan treated Taiwanese people as a scapegoat to which they channeled their hatred. Mutual distrust became widespread. The memory of unhappy encounters passes onto latter generations. It also lays the foundation for Taiwan’s political development. TAIWAN ’S POLITICAL INSTITUTIONALIZATION AND SOCIAL CAPITAL Political institutions are the results of culture and learning. They also act as rules for political games and affect people’s thoughts and behaviors in the long run. This Japanese before he was twenty years old. “’2 For details, see “The Power of Taiwan’s Senior Elites,” Journalist (in Chinese), Vol 751 (2001). ‘3 In fact, according to various surveys conducted in the past, people in Taiwan who really advocate an independent Taiwan have never exceeded 20%. Most of the people, around 60%, prefer to maintain the status quo. My argument is based on three different national surveys conducted in different periods, i.e., The 134 section is devoted to the discussion of Taiwan’s political institutional development, which is thought to be critical to people’s conceptions of trust and cooperation. Political Party System Before the late 19805, Taiwan’s political structure can be characterized as “one party authoritarianism.” The Kuomintang (KMT) is the unchallengeable ruling party. The party monopolizes power at almost all levels of government, including armed forces and the police force. Government is literally the administrative machine of the party (Tien, 1989:71). Inside the KMT, it is organized according to the Leninist principles of democratic centralism. On the national level, the delegates of Party Congress are responsible for electing the members of Central Committee, and it in turn elects the members of the Central Standing Committee and its chairman, the real paramount leader of the party.44 Through its administrative branches at various levels, the party’s decisions made by the Central Standing Committee can be effectively enforced all over the country.45 Many factors have contributed to the consolidation of the one party system in Taiwan. Internationally, Taiwan’s strategic position to block communist China fit the United States’ national interests in the early 505. Thus the US. had to support the KMT regime in Taiwan, for it was the sole party with the power to stabilize Taiwan. Social Image Survey in Taiwan, August 1991, August 1992, and the Taiwan Social Change Survey, 1998. The results for those who support Taiwan Independence are 11.8%, 18.9% and 15.6% respectively. 4" As of March 2000 only Chiang Kai-shek, Chiang Ching—kuo, and Lee Teng-hui have assumed this position. In fact, they were also the party chairmen when acting as the chairman of the Central Standing Committee. 135 Domestically, the after-shock of the February 28th incident left Taiwanese people politically alienated. There is no documentation of formal political parties in the era of Japanese rule (Hu, 1998:23). Further, the ongoing civil war also justifies the harsh rule by the KMT for the sake of national security. These factors have paved the way for KMT to establish a dominant position. The legal justification of Taiwan’s one party authoritarianism46 is based primarily on two laws, the Provision Amendments for the Period of Mobilization of the Communist Rebellion, and martial law.47 The former justifies the president’s almost unlimited authority to “take emergency measures to avert an imminent danger to the security of the state or of the people—without being subject to the procedural restrictions prescribed in Article 39 or Article 43 of the constitution” (Article 1 of the Provision Amendments, Tien, 1989:109). In addition, it also breaks the two-term limitation of the president set by the constitution so the president can be reelected as long as he wants to run for reelection.48 Martial law, on the other hand, greatly restricts people’s civil rights with the help of military power. Based on this law the Taiwan Garrison Command (TGC) was established in 1950, responsible for implementing martial law. Civil rights such as the right to publish, privacy of correspondence, the right to practice religion, free speech, free assembly, and the right to petition are all under the supervision of the TGC. Under martial law, no new political parties can be formed. Dissidents can only express their views under ‘5 For detailed KMT party structure, see Tien, 1989. “’6 In fact, there are two other small political parties, the Democratic Socialist Party (DSP) and the Young China Party (YCP) accepted as legal political parties by the KMT. But they relied on the KMT’s financial subsidy to survive (Tien, 1989: 64). Therefore, they should not be regarded as independent political parties competing with the KMT for political power. ’7 In fact, these two laws had already existed before the KMT retreated to Taiwan. They originated in the civil war period in Mainland China and applied to Taiwan automatically. ‘3 It is commonly believed that this was tailored exclusively for Chiang Kai-shek, then president of the 136 certain limitations, such as in the period of election campaigns.49 One party authoritarianism lasted for almost forty years and began to soften when the late president Chiang Ching-kuo decided to liberalize Taiwan’s political system as a response to international and domestic pressures.50 Martial law was lifted in July 1987 and replaced by the national security law. The immediate effect of this reform was the formation of Democratic Progressive Party (DPP) in September 1987, the first real opposition party in Taiwan since 1945.5 I In January 13 1988 Chiang Ching-kuo died and vice president Lee Teng-hui assumed the presidency immediately. Lee’s Taiwanese background and his determination to quicken the pace of democratization in Taiwan further complicated Taiwan’s political ecology and the ethnic issue. Specifically, inside the KMT, Lee won the power struggle over the mainlander elite for the position of chairman and got elected as the 8th president of the Republic of China (Taiwan). Lee’s policy of Taiwanization of the KMT and a lenient attitude toward the issue of Taiwan Independence place him closer to the DPP’s ideology than to the KMT’s. A greater number of mainlander elites in the KMT were threatened and forced to leave the KMT.” Under Lee’s leadership, the KMT split and its advantage in elections at various levels drastically decreased.53 Consequently, this trend peaked in the 2000 presidential election. Republic of China. ’9 In the period of election campaign, law enforcement agents were usually more lenient towards dissenting opinions. Many people called this period ‘holidays for democracy’. 5° For factors leading to political reform, see Gold (1986), Wakabaashi (1992) and Peng (1995). 5' In fact, the formation of the DPP was still illegal at that time. Although martial law has been lifted, the law regulating the people’s right to form groups, including political parties, was not effective until January 1989. However, the KMT did not take any steps to stop the formation of the DPP. It is widely believed that Chiang Ching-kuo really was serious about liberalizing Taiwan’s political system. 52 Some of them formed a new political party called the “New Party” and claimed the orthodoxy of the KMT. Still some others stayed in the KMT and became a faction called the “non-mainstream” as opposed to Lee’s Taiwanese elite who really controls the KMT. ’3 Two examples are enough to support this argument: the 1994 Taipei city (capital of the Republic of China) mayoral and 2000 presidential elections, in which the candidates of the DPP won due to the inner 137 The KMT lost the campaign and the DPP’s presidential candidate, Chen Shui-bian, was elected as the ninth president of the Republic of China (Taiwan). The KMT handed over its ruling legitimacy to the DPP and became the opposition party for the first time since its retreat to Taiwan. The fifty-year history of the one party system has impacted people’s thoughts and behavior profoundly. Here I will limit my discussion to its impacts on opposition parties, especially the DPP. The formation of the DPP is the result of over 10 years of opposition movements. Restricted by martial law, Taiwanese dissidents could not organize on a large-scale to integrate their resources. Hence, an elite involved in the opposition movement can only earn people’s support in elections by using his own resources, e.g., establishing personal service centers, fu wu chu, to provide favors to local residents. Once they became elected, they accumulated their own political resources mostly by themselves. Therefore, history made individualism precede the formal organization, which developed later.54 Given this, although the DPP borrowed the KMT’s organizational structure, the leaders at the central level could not have the real authority to command local elite behavior. In other words, the vertical and horizontal connections within the party organization did not operate as they were supposed to. Local elites’ concerns always outweigh the party’s overall policies. Another consequence of the one party system in Taiwan is reflected in the split of the KMT. The KMT nominee not only had to face the DPP candidate but also the challenge of former KMT member that was not satisfied Lee Teng-hui’s rule and run as either another party’s candidate or as an independent candidate. If the split had not happened, the KMT would not have lost such important battles in terms of the total number of pan-KMT supporters. For example, in the presidential election in 2000, the total votes that Lien, KMT candidate, and Soong, former KMT member, counted 59.94% of the I38 interaction between the KMT and the DPP or, broadly speaking, the relationship between ruling and opposition parties. During the democratization period, Lee had many chances to invite the DPP’s chairman and high-ranking staff to discuss ideas of how to amend the constitution. Those private talks aroused many suspicions from both parties. From the perspective of KMT members, it is unsettling to see their chairman shake hands and talk pleasantly with those who strongly oppose the KMT’s ruling style. For those in the DPP, many worried that their chairman and other elites would be assimilated into the KMT. This mutual distrust did not disappear even after the DPP won the presidential election in 2000. Although Chen was elected as the president,55 his party, the DPP, is still the minority in the Legislative Yuan, the parliament of Taiwan. According to the spirit of the constitution amended in 1997,56 Chen has to appoint a majority party leader, i.e., KMT, as the premier and form a coalition government to share political power. However, the DPP did not want to follow the constitution and thought the DPP is the ruling party and should have its own government. This winner-take-all vision conflicts with the spirit of the constitution. President Chen first appointed former minister of Defense, Tang Fei, who is a KMT member, as the premier. But Tang held a different view from president Chen as to whether to continue constructing a fourth nuclear power plant. Tang was soon replaced by total votes and the elected president, Chen, garnered only 39.3% of the popular votes. 5’ As to the origin and development of DPP, please see Huang, 1992 and Tien, 1989. 55 As indicated above, Chen only won less than 40% of the popular votes, given the allocation rule of plurality (relative majority) set in the constitution. The ROC’s constitution was amended several times during the leadership of Lee. According to the amendment of 1997, ROC’s government was shaped towards the one similar to, but not exactly the same as, France’s semi-presidential system. However, it allows the president to name a premier without the approval of the Legislative Yuan (parliament). By allowing the president this power, the KMT and the DPP delegates reached an constitutional agreement that the president should appoint the leader of the majority party or majority coalition in the parliament as the premier if president and the majority of the parliament are from different parties (Kao, 200 l ). I39 Chang Jun-shyong,57 a DPP member and a former legislator of the Legislative Yuan. Since then the DPP has become an unprecedented party, i.e., the three-fold minority government, which means a government with minority president, minority premier, and minority parliament.58 This pattern of government of course faces many objections from the ‘majority” opposition parties, i.e., the KMT, the New Party, and the People First Party (PFP).59 The major issues dividing the ruling DPP and opposition parties include the issue of national identity, cross-strait issue (views on how to deal with mainland China), and whether to continue constructing the fourth nuclear power plant. The conflicting views between the two camps impose a high degree of uncertainty on Taiwan’s future. The sour relationship between ruling and opposition parties are partly the result of the KMT’s long uncontested reign, which provided little chance for politicians to learn tolerance and cooperation. Even the ideologically similar parties have difficulty cooperating in elections. As I mentioned before, there are three major parties originating from the KMT. It would have been reasonable for them to cooperate in the year-end (2001) Legislative Yuan’s election in order to maintain the majority advantage, especially under the prediction that no single party would win the majority seats in this particular election. The same holds true for the county magistrate’s election: for opposing parties: no cooperation means little chance. However, evidence indicates that cooperation among ’7 Tang’s five-month stay in the ofiice makes him the premier with the shortest term in the ROC’s history. ’8 President Chen was elected with less than 40% of the popular vote, the premier is not backed by the majority of the Legislative Yuan (parliament), and the president and premier’s party is minority in the parliament, only one third of the seats. This term “three-fold minority government” was coined by Jan Lian, the new chairman of the KMT, in his public speech at Oxford University in England (Central Daily News, p3, June l3,200l). 9 The People First Party was formed right after the 2000 presidential election. The leader and chairman is Soong Chu-yu (James Soong). He was a charismatic leader in the KMT. Due to the personal clash with Lee, he failed to get nominated as the KMT’s presidential candidate. He decided to run the election as an independent candidate and lost to Chen by a close margin. His new party siphons off members mostly from I40 KMT originated parties has not met expectations. This involves not only the issue of electoral strategy but also competing for the right of control in the process of cooperation.60 Hence, the prison’s dilemma still haunts these opposition parties. The coming Legislative Yuan’s and county magistrate’s election will test whether and how well opposition parties overcome this dilemma. The State-Society Relations Scholars of political science have long debated the relation between state and society. I do not intend to review that debate, rather my focus in this section is the actual relations between state and society in Taiwan after 1945 and how it influences the formation of social capital. Background When the Nationalist government took over Taiwan in 1945, it also took over the KMT. 6° For the Legislative Yuan and National Assembly election, Taiwan adopts a unique electoral system called a Single Non-transferable Vote (SNTV). Voters vote for only one candidate on the ballot and the elected candidates in a, for example, four seat district, are simply the four with the most votes, regardless of party affiliation (Taagepera & Shugart, 1989:28). This allocation rule is thought to discourage cooperation among same party candidates because candidates from the same party will normally compete not only against each other but also with other parties’ candidates for several seats ranging from I to 17 (in Taiwan’s case) in the same district. Rather, given that the amount of votes needed to win are relatively few compared to single member districts, party members are prone to leaving their party and running independently if they are not nominated by their party. Regarding the competition for directing the cooperation process, evidence shows that the KMT and the PFP accused each other for lacking the sincerity to cooperate. For example, in this coming county magistrate and city mayoral elections, no parties seemed willing to support other parties’ candidates, whom they thought have better odds of defeating the DPP’s candidates. Rather, each party wants their own candidates to be supported by other parties (Central Daily News, p I, August 17, 200 l ). I4I many state-owned enterprises left by Japan. This provided a strong base for the financial autonomy of the KMT regime. In addition, the KMT’s control of military and police power and the tragedy of the 2-28 incident had made Taiwanese detract from politics rather than voice their thoughts. These factors paved the way for the relatively easy work for the KMT regime to stabilize Taiwan’s society. Given the favorable situation for KMT rule, the leaders of the KMT realized that in the long run they still needed to institutionalize the relations between state and society that they may not only control the society but also win the support of the Taiwanese. The two most important institutional arrangements were incorporating socioeconomic sectors into the state and under the close supervision of the state, and allowing local elections.“ The Operation of State Corporatism One of the major reasons for the KMT’s loss of the civil war is its ignorance of the power of mass mobilization, which was mastered by the CCP (Eastman, 1984:218). In August 1950, Chiang Kai-shek ordered the formation of a Central Reform Committee, which was charged with the “Reform Program of the KMT.” In this large-scale reform, the KMT reorganized itself to ensure the smooth implementation of the party’s polices. Its aim was to bring youth, students, farmers and workers under the state’s control and guidance. The KMT successfully won support of farmers by undertaking land reform in the 6' Electoral politics and its consequences will be discussed in the next section. I42 early 505.62 The success of land reform brought about conditions favored by the KMT. In the economic sphere, land reform largely reduces the unevenness of income distribution, which accordingly alleviates the potential pressures associated with social inequality. In the political sphere, it eliminated Taiwanese landlords’ power base and won the support of Taiwanese farmers. At the same time, the government organized peasants and rural residents into 340 KMT—controlled farmers’ associations, which offer credit, introduced technology, supply inputs, and serve as marketing cooperatives (Yang, 1970:407-411). The associations are hierarchical, with the provincial farmers’ association as the main managing unit. They operate more as quasi-govemmental institutions, in part to implement the agricultural and credit policies of the government, than as articulators of the farmers’ interests (Tien, 1989:48). As to organizational control, the rates of party membership among farmers’ association leaders are almost 100% (Tien, 1989:61-62). For the control of workers and other social groups, the legal base is the so-called “Organizational Law of People’s Groups in Time of Emergency.” According to this law, two basic requirements apply to the social groups. One, most of the occupations are required to form their own occupational group and everyone in that occupation is required to participate in that group. Two, voluntary associations cannot be formed without the government permission. In addition, voluntary associations, including the above- mentioned occupational groups, similar in nature cannot coexist in the same area. By the two requirements, the government not only controls the interest aggregation of the society but also prevents large-scale associations from arising, which precludes the possible spontaneous horizontal integration among voluntary associations (Ding, 1999:66-67). ’2 For a detailed discussion of land reform in Taiwan, please refer to Yang (1970) and Gold (1986). I43 However, this law is not strictly enforced (Ibid: 76). The actual means that the KMT uses to keep track of the socioeconomic groups lie in the control of their personnel and financial support.63 For example, the chief of the board of directors (Ii-shih-zhang) and the general manager (tsung-kan-shih) at the county or higher levels of labor unions are almost all members of the KMT (Li, 1992:120). Around 80% of the funds of the Chinese Federation of Labor (CFL), the highest level of legal labor union in Taiwan, are sponsored by the government, and its major function is to act as a quasi-diplomatic agent rather than a pressure group defending laborers’ own interests (Shen, 1994:184-186). The KMT also pays close attention to the political socialization of youth and students. Starting in 1952, a special mass organization known as the Chinese Youth Anti- Communist League (CYACL) was established by the KMT, which was under the supervision of the General Political Department in the Ministry of Defense. The CYACL has experienced several reforms,64 but its goal remains the same: providing large-scaled and monopolistic recreational activities to bring together students into the KMT and government in general and preventing potential competitors from doing the same. In sum, through state corporatism the KMT government enjoyed the monopolistic power of policy making for over forty years while at the same time repressing the development of voluntary associations. Recent Development since Liberalization The top-down pattern of state corporatism in Taiwan has become crippled since ’3 For detailed statistics, please refer to Ding (1999:72-74). I44 the liberalization of the mid80s.65 The state is no longer able to control tightly the socioeconomic sectors it once did. Of course political liberalization, including the lifting of martial law and related regulations, offers more space and tolerance for any possible social movements. After the lifting of martial law, and the KMT’s tolerance of the illegal formation of the DPP, people began to sense more space and decreased risk when voicing their concerns and dissatisfaction toward the authority. Thus, suddenly Taiwan seemed to be flooded by a great wave of social movements never seen before. For example, outdoor street protests rose from 337 in 1986 to 734 in 1987, when martial law was lifted, and 1172 in 1988 (Wang, 1993:82). The official number of assemblies and demonstrations in Taiwan also rose from 538 in 1986 to 1,233 in 1987 and even 7,775 in 1990 (DGBASEY, 1999:181). According to Hsiao’s summary, fourteen broad issues were involved in those spontaneous social movements, including consumers’ rights, local anti-pollution self- help, ecological protection, laborers’ rights, women’s rights, self-govemment on campus, discharged soldiers asking for return to mainland China, mainland soldiers asking for their welfare, teachers’ rights, aborigines’ rights, farmers’ rights, political dissidents’ rights, disability rights, religious rights (Hsiao, 1994;26:27). Besides, immediately after the passing of The Organization Law of People’s groups in 1989, the number of registered political parties surpassed 40 and increased to 86 in 1998. The common theme was to protest public policies. Thus a question about the sudden emergence of Taiwan’s various social 6’ For a detailed history of this organization, see Tien (1989:87-88) and Wakabayashi (1992:120-l22). 6’ I do not intend to discuss the factors leading to the liberalization and democratization of Taiwan. For those interested in this topic, please refer to Chou & Nathan (1987), Chen (1987), and Wakabashi (1992). I45 movements needs to be answered: why did so many horizontal associations flower in the months immediately after martial law was lifted? I believe we can answer this question from two perspectives: first, it was the ruling political elite’s willingness to recognize and accept the trend of democratization. But this factor only plays the role of catalyst. The other factor is the long-term effect of Taiwan’s economic development. Taiwan’s Per Capita GNP rose from 1976’s U.S.$1,132 to 1987’s U.S.$5,298. And the Per Capita Income also rose from N.T. (New Taiwan dollar) $ 3,9559 in 1976 to N.T.$154,229 (DGBASEY, 1999:12—13). Further the unemployment rate between 1976 and 1987 was relatively constant at around 2.2%.66 The percentage of senior high school graduate or above rose from 26.4% in 1976 to 40.5% in 1987.67 As people have become wealthier and less concerned with basic needs, a desire for recognition as “an adult with a certain basic human dignity” becomes more urgent (Fukuyama, 1995a: 22-23). From this perspective, we see the abrupt social movements since 1987 as an ongoing process rather than a sudden social phenomenon. Specifically, people began to reinterpret the existing social and political structures and redefine issues that were repressed in the past. With the help of the political liberalization, this undercurrent trend has asserted itself more and more. The horizontal linkages among social groups also emerged to accumulate enough political power to compete with the state and capitalists (Shen, 1994:308; Xu, 1989:123). In sum, the social movements and significance of voluntary organizations after 1986 signals Taiwan’s potential to form strong social capital given years of state corporatism. On the other hand, beneficiaries of the old system and the legacies left by state corporatism may still exert considerable force to counter social capital formation. In 6° Calculated by the author from the DGBASEY. 146 fact, Table 5-4 shows that people in Taiwan hold a relatively stable attitude toward the role of social movements. Over 80% of them think social movements improve social equality. Ironically, however, around 60% of the people in Taiwan who are not willing to participate in the groups think they have a positive role in society. Table 5-4 Attitude Toward Social Movements and Related Groups* 1 991 l 992 1994 1. Good for social reform“ 88.4% 82.6% 87.7% Not good for social reform 11.6% 17.3% 1 1.9% 2. Willing to participate in them by 6.1% 5.6% --*** giving money Willing to participate in them by 18.5% 14.1% -- volunteering Willing to participate in them by 20.3% 12.8% -- both means Not willing to participate in them 55.2% 67.5% -- Sources: The Social Image Survey in Taiwan, February 1991(N=l605); August 1992 (N=1523); and July 1994 (N=1437). ’This table is based on two survey questions which appeared in the three surveys. The first stated, “Overall, what kind of impact do you think recent social movements have on Taiwan’s social reform?” The second asked, “To what degree are you willing to participate in groups involved in social movements?” "The original question is on a four-point scale. It has been recoded by the author into a dichotomous variable. *"The question did not appear in this survey. Electoral Politics and Local Factions 67 Calculated by author from the DGBASEY. 147 Although unhappy about the DPP take-over, the KMT decided to establish Taiwan as a base for an anti-communism campaign through both economic and political means. Through land reform, the inequality of land distribution was alleviated and farmers’ living standards increased significantly. The success of the land reform increased Taiwan’s agricultural productivity, which in turn was used to provide labor, capital and foreign exchange for industrial development (Peng, 1995:259-266). And in order to sofien sour relations after the 2-28 incident, and to forestall possible anti-system movements from the Taiwanese, the KMT chose to adopt a limited democracy, i.e., allowing Taiwanese to elect their own local legislators and the county or city head. By successfully implementing economic policies and allowing local elections in Taiwan, the legitimacy problem that the KMT faced was also tactically solved. In this section, my main focus is electoral politics and its consequences that may influence the formation of social capital in Taiwan. Local elections were institutionalized in October 1950, when the first elections for county commissioners and city mayors and members of county and city councils were held. Due to the lack of social connections, the KMT chose to mobilize voters through the help of local Taiwanese elite.68 These elite was mostly gentry or landlord in the Japanese era. Their social prestige, due to knowledge and wealth, brought them many followers. Gradually, through the network of guanxi these elite and their followers form a patron- client relationship,69 broadly called local factions. ’8 In fact, the KMT government tried to destroy local factions led by Taiwanese elite. But they can only remove leaders of factions, while the actual networks related to those factions have already spread throughout Taiwan. Thus the KMT decided to accept the existence of local factions and use them as tools to mobilize people, especially in local elections (Chen, 1995). ’9 There are many explanations for the formation of patron-client relations in Chinese societies, ranging 148 The existence of local factions was further reinforced by the KMT’s ruling strategy. Specifically, backed by tremendous socioeconomic and political resources, the KMT established a higher level of patron-client relationship with local factions. On the one hand, the local factions were offered exclusive monopoly power to run local business, such as the operation of credit cooperatives, farmers associations, fishermen associations, and public transportation. According to the study by Chen and Zhu, among 89 local factions at county level, 90% of them own or control at least one monopoly business (Chen & Zhu, 1992:89). Accordingly, through the benefits of business and their members being elected as the legislators, the local factions provide their clientele constituents (followers) with particularized services both legally and illegally.70 On the other hand, the KMT adopted a ‘divide and rule’ strategy. This so-called bifactionalism keeps at least two factions in the same area and alternates its support between the factions across elections or across elective offices. In other words, through check-and-balance, the KMT, being a super patron and arbiter, can prevent local factions from developing into cross county or city coalitions. The role of local factions becomes even more important after political liberalization and democratization. Their influence has moved from the local to the national level. First, with the vanish of political strongman and the KMT’s loss of authoritarian power, elections are the only legitimate means of allocating political power. Thus the local factions have more leeway to coax political resources from the ruling party given the opposition parties’ electoral competition. Second, the power struggle after the fi'om authoritarian personality (Wen, 1972:47-75; Jiang, 1989:] 1-13), family-centered orientation, including emphasis on similarity (Tang) (Jacobs, 1979), to rational calculation (Pye, 1980). 7° The services that they provide to their constituents include meeting the budget for local infrastructures. conflict resolution, attending marriage or funeral ceremonies etc. In addition, faction leaders are also involved in law enforcement intervention, such as trying to lobby law enforcers not to punish their 149 death of Chiang Ching-kuo encouraged Taiwanese elite, led by Lee Teng-hui, to join with local factions in order to uproot the mainlander elite’s power bases (Chen, 1995:228-236). In sum, an informal social contract exists between the KMT and local factions. Local factions transform their Guanxi network into votes for candidates supported by the KMT or local factions itself. In return, the KMT allows local factions certain monopoly businesses. Thus both of the parties benefit from this patron-client relationship. However, this pattern bears several undesirable consequences in terms of social capital formation. First, although local factions are quasi-political groups, yet they do not have any fixed political beliefs (Ibid: 234). They exist and operate only through favor exchange. Specifically, faction leaders must display enough social and financial resources to attract followers. Followers exchange their loyalty and services for expected interests participation in factions. Thus the principle of right and wrong gives way to personal interests and faction labels. Emotional reaction and paying more attention to criticizing individual opponents than rational policy debate have become the life of local faction politics (Zhao, 1987:61-63). Second, the rivalry between factions is maintained or even escalated by elections. During the election campaign each faction tries to reduce the odds of other faction candidates by spreading rumors of personal scandals, which are denied by the opponents, or by any other means they demonize their opponents. As a result of years of confrontation, factions can only gather more supporters to fight against their rival counterpart(s). Supporters in each faction develop a sense of hatred (Liao, 1997:46; Chen, 1995:22). The general public is generally lost and does not know which party to trust.“ constituents who run illegal businesses. Illegal lobbying through renqin network really plague Taiwan’s society. For details, see Liao (1997:46). 7' Trust is not an issue for those involved in faction politics. But I would like to emphasize the effect of the 150 Third, the competition for public sector resources is mostly zero-sum games (Chen, 1995223). With the demise of any supreme leader in Taiwan, this further increases the difficulty of cooperation or tolerance among local politicians and their followers.72 As former chairman of the DPP Shih Ming-deb said, “politicians in Taiwan lack tolerance and respect. This is why factions are now above their parties and the interests of the people” (Central Daily News, p3, August 19, 2001). Accordingly, public policymaking is likely to deadlock. Last but not least, the SNTV electoral system provides reason for both inner party candidates’ mutual distrust in election and spaces for new political parties, neither of which is conducive to tolerance and cooperation that a stable democracy needs.73 On the other side, the KMT’s decision to institutionalize electoral politics, first at the local level then at the national level,74 really inspires people’s political awareness. It also raises people’s concerns about Taiwan’s political democratization (Hu, 1998:42). Although faction politics has undesirable influences on the development of social capital, as indicated above, its function to mobilize people to participate politically should not be underestimated. means factions use to vilify their opponents; most of them are illegal and irresponsible toward the general public. Whatever we call it- stretching the truth or telling a lie or spreading a rumor- it has already set unhealthy examples for our society, especially students who are learning moral values in school. 72 The KMT used to play such a role in its authoritarian era. For instance, it was able to promise each faction certain political benefits alternately or provide equivalent benefits for factions. 7’ For the operation of SNTV, see footnote 58 of this chapter. 7’ Due to increasing vacancies in the national representative bodies (the National Assembly, Legislative Yuan, and Control Yuan), in which members were elected in mainland China before 1949, the nationalist government decided to hold the first national elections in 1969 to fill the vacancies. But these members elected in Taiwan still acted as those elected before 1949 in China, i.e., they would stay in office until the KMT reclaim Mainland China. In 1972, after Taiwan was unseated from the United Nations, the KMT amended the temporary provisions and created an additional 51 seats for the Legislative Yuan, 53 for the National Assembly, and 15 for the Control Yuan. From then on, a national level election has been held 151 Summary Three points of this section can be summarized as follows: first, a long-term one- party system leads to a sense of distrust and intolerance among political actors. Taiwan now is facing a natural consequence of this ‘social deficit’: a minority ruling party, which does not intend to cooperate with the opposition parties. To make things worse, this ruling party, i.e., DPP, withholds a party platform that deviates from most Taiwanese’ vision of Taiwan’s future. How will the ruling party advocating Taiwanese Independence direct a country in which most of the people do not support this party’s ultimate political agenda? Thus the uncertainty of the future will bring further obstacles to the formation of social capital. Second, although modernization plays a decisive role in bringing about political democratization to Taiwan, we still see many legacies left by state corporatism and traditional culture. Among them, for example, people are not very interested in participating in voluntary associations. Third, electoral politics in Taiwan has both positive and negative functions in terms of forming social capital. At the local level, especially rural areas, local political factions help mobilize people to participate in politics and hence raise their political awareness. But the modes of competition among local factions set bad examples for mutual trust and willingness of cooperation. At the national level, the electoral system (SNTV) negatively affects political interaction, who also seta bad example for the general pubhc. every three years. For details, please refer to Tien (1989). 152 Based on the background provided in this chapter, in the next two chapters I will conduct an empirical research on both private and social capital using Taiwan. The political consequences of social capital will be the main focus. 153 CHAPTER 6 PRIVATE SOCIAL CAPITAL AND ITS POLITICAL CONSEQUENCES INTRODUCTION In previous chapters I have sketched a contemporary study of social capital and proposed frameworks for analyzing private and public social capital respectively. The cultural and historical factors structuring both private and public social capital have also been discussed in the context of Taiwan. In this chapter, I conduct a primarily empirical study using a quantitative method to explore some possible relationships between private social capital (individual’s social capital), as well as its political consequences. In addition, one of the operating mechanisms within individuals’ networks, i.e., norm of reciprocity, which comprises private social capital, will be explored as well. RECAPITULATE THE CONCEPT OF PRIVATE SOCIAL CAPITAL AND THE LITERALTER REVIEW Social capital is commonly regarded by sociologists as a source of network- mediated benefits beyond family and close friends (Portes, 2000253). Although the concept of social capital originates from a variety of intellectual sources,l from the 19705 an approach known as network analysis may have the most influence on the study of ' For details, please refer to chapter two of this dissertation. 154 private social capital. Network analysis focuses on relationships between actors instead of social positions of actors. The network-orientated approach aims to explore subjects such as “the manner in which individual actions create social structure; the manner in which social structure, once created, constrains individual and collective action; or the manner in which attitudes and behaviors of actors are determined by the social context in which action takes place” (Marsden & Lin, 1982:10). Among these concerns, social capital is particularly applicable as an analytical concept to study the way individual actors adapt to create social structure. The network-oriented approach has a tradition to see an individual’s social connections as instruments for advancing social actors’ private goals. (Portes, 200:53). Literature has shown many works using a network approach to treat social capital as an explanation of access to employment, mobility through occupational ladders, and entrepreneurial success (Granovetter, 1974; Loury, 1977; Lin, Ensel, and Vaughn, 1981; Portes, 1987). I do not intend in this chapter to comprehensively review literature on private social capital but rather the so-called “strength of weak ties” arguments proposed by Granovetter in the early 703. His theory received many attentions, mostly from sociologists. Many works have been inspired by his original thoughts. The primary goal of Granovetter’s ‘strength of weak ties’ theory was to build a bridge connecting micro- level interactions and macro-level patterns in a convincing way. The cutting point he used to achieve this goal is through the analysis of interpersonal networks, for he thought that they will help us grip the essence of how interaction in small groups aggregates to form large-scale patterns (Granovetter, 1973:1360). More specifically, at the micro level he emphasized the functions of weak ties as a crucial bridge between two closely knit 155 groups of friends. Due to less overlap of friendship in weak ties, several disadvantages may be attributed to individuals with few weak ties, including information deprivation from distant parts of the social system, a disadvantaged position in the labor market, difficulty organizing or joining politically based movements of any kind (Granovetter, 1982:106). While at the macro level, social systems lacking weak ties will be fragmented and incoherent (Ibid). Granovetter further contended that weak ties, “far from being productive of alienation, as one might conclude from the Chicago school of urban sociology—especially from Louis Wirth—are actually vital for an individual’s integration into modern society” (Granovetter, 1982:107). In fact, using Woolcock’s typology of social capital, Granovetter’s works and emphasis can be categorized on the linkage side of social capital, i.e., an individual’s autonomy to reach out for extra-community networks. In chapter three I mentioned that the other dimension of social capital, i.e., embeddedness, captures most of the idea that social capital indicates. Granovetter’s original work builds our insight into social capital although he never seemed to use the phrase ‘social capital.’ Related to Granovetter’s works, Lin proposed a theory of instrumental action in which he postulated, “the level of the initial position is positively related to social resources reach through a contact” (Lin, 1982:134). In other words, a society’s structural opportunity gives those whose ascribed positions are relatively high a better chance to reach more social resources than those whose initial social positions are low. This strength-of-positions proposition serves as an explanation for the maintenance of social inequality. However, to explain why social mobility is possible in some societies challenges the strength-of-positions proposition. Lin then adopted Granovetter’s strength- 156 of-ties argument and contended that by breaking out of one’s own intimate social circle, weak ties rather than strong ties tend to lead to better social resources. In addition, the assistance of weak ties has more effects on those whose initial social locations are low than on those with higher ascribed social positions (Lin, 1982:134-135). So far may piece together several elements of this strength-of -tie argument. First, this argument points us to the importance of linkage in an individual’s social networks. Although strong ties are readily available and easily accessed, yet weak ties serve as bridges for information and social resources not otherwise available. Hence social mobility is possible and explainable. Second, on the macro level, also due to the bridging role, collective action is easier for those who own weak ties. The reasons are two fold: first, without weak ties as a mediator, people tend to be alienated from and distrust their leaders. That is, “trusting in leaders is integrally related to the capacity to predict and affect their behavior” (Granovetter, 1973:1374). The existence of intermediary personal contacts who can assure another that a leader is trustworthy will make this easier if weak ties exist. This argument has been tested by Karweit using biracial school settings in the United States as an example and suggests that encouraging bridging weak ties has the effects of linking culturally heterogeneous groups and increasing general social solidarity (Granovetter, 1982: 123). The other reason that weak ties make possible collective action is the ‘complex role sets’ it induces and the need for individuals to develop ‘intellectual and cognitive flexibility’ (Granovetter, 1982:107-108). Since “the ability to construct and function in complex voluntary organizations may depend on a habit of mind that permits one simultaneously to assess the needs, motives, and actions of a great variety of different people” (Ibid: 108). Third, as to the sources contributing to the formation of weak ties, 157 several factors have been suggested, from the division of labor, work settings, the development of communications systems, bureaucratization, population density, the spread of market mechanisms, average family size, and differences in social class structure and sense of security (Granovetter, 1973:1375; Granovetter, 1982:107; Pool, 1980; Blau, 1977). Given the above theoretical introduction, I would like to insert the strength-of-ties theory into the context of Taiwan and see first to what extend this theory may be applied to Chinese societies; and second, how well this theory fits empirically the data collected from Taiwan. STRENGTH OF TIES THESIS AND ITS APPLICATION TO TAIWANESE SOCIETY Granovetter’s influential paper, “The Strength of Weak Ties” stimulated many studies to elaborate it or test its validity. However, most were conducted in Western societies, especially the United States, where individualism prevails. To the best of the author’s knowledge, there are very few if any studies using the strength-of-ties argument directly to observe Taiwanese society.2 Given the limited availability of survey data, I will try to test the validity of this theory in the second half of this chapter. But first I will discuss how we should expect the strength—of-tie thesis to operate in a society highly sensitive to the power of social networks or guanxi. 2 Empirical (quantitative) studies using the social network approach in Taiwan are relatively few. Most of those studies focus on social networks’ effects on the achievement of social position or other individual goals (Hwang, 1999:293). Most studies employing networks concepts are qualitative in nature. The topics range from the operation of faction politics among political elite (Chen and Zhu, 1992; Chen, 1995; Wang, 1996; Liao, 1997) to the economic domain such as the effects of kin ties and political networks on the business running of Taiwanese capitalists (Hsiao, 1992); and the dependence of small and median scale enterprises on their social networks for development (Chen, 1994). 158 The Strength of ‘Strong’ Ties in Taiwanese Society Granovetter’s argument placed a premium on the roles of an individual’s weak social ties in both enhancing personal goals and some other desirable macro level consequences. However, in chapter five the author argued that Chinese culture is very family-oriented. Family is the center of life. Social connections radiate from family and they finally return to support the reputation of the family.3 From this perspective, we would expect parents, kinship, and relatives to play the role of supporter, either emotional or instrumental, to the social actors in Chinese society. Relatively few literary works discuss strength of ties and the values of strong social ties. Granovetter, however, mentions two important roles of strong ties: greater motivation to be of assistance and more greater availability (Granovetter, 19822113). Based on the assumption of homophily (the tendency to select friends with similar characteristics to oneself), Blau argued that the lower one’s class stratum, the greater the relative frequency of strong ties (Blau, 1977). Pool also contended that the sense of security of individuals, e.g., level of wealth, determines the utility of weak ties. For example, a highly insecure individual is under strong pressure to become dependent upon one or a few strongly protective individuals. On the contrary, a person with resources on which he can fall back resists dependence on any other individual, and can explore more alternative options at his own wish (Pool, 1980). Although these authors provided us with insight on the use of different tie strength, they seemed to focus on the individual’s social 159 positions. I will add that a society’s culture may also constrain social actors’ choices in utilizing various tie strength. The role of family in Chinese society will first be covered and then how non-kin relationships operate, with the strength-of-ties argument in mind. The centrality of family in one’s life can never be underestimated in Taiwanese society. In a family, great personal sacrifice is made for the sake of constructing a secure and growing economic foundation in order to both honor parents and ancestors in general and to support the generations to come. After children reach adulthood, they accept the filial obligation to care for their parents. A Confucian dictum says, “When we are young, our parents nurture us; when our parents are old, we sustain them.” Even after the division of the family estate in the immediate family,4 adult children and parents still maintain close contact. From the perspective of parents, children’s behavior reflects the upbringing given by them, and it also affects their ability to be a model for their offspring. The paramount role of Chinese familism is best observed in the economic sphere. Those who study Taiwan’s business organizations reach the same conclusion: Taiwan’s privately owned businesses are overwhelmingly family-owned and family-controlled enterprises5 (Mark, 1972; Greenhalgh, 1988; Orru, Biggart, & Hamilton, 1997). For example, in an empirical study conducted by Taiwan’s Tunghai University showed that 84 of the top 97 business groups in 1983 could be strictly classified as family-owned 3 Note that family here refers not only the immediate family but also extended family, including kinship and relatives from the same identifiable ancestors. . ’ Generally speaking, any adult brother can demand his share of the estate at any time. But this division of family estate typically occurs soon after brothers marry. In some places, brothers usually delay this division until after their father dies. For more details, see Davison and Reed (1998: 173-175). 5 Most of the authors use this phenomenon to explain why the Taiwanese economy appears to have favored small-firm economic structure, but I use it as an example of how strongly the family bond ties Chinese people. 160 business groups. Among these groups, 23 are owned primarily by a single household head, and the remaining groups (61) have multiple family members. Out of these 61 business groups, more than 80% are of the following three types: fathers and sons, brothers, and brothers and their sons.6 These family firms follow a typical cycle of four stages formulated by Wong (1985) and modified by Hamilton (1996) in which the ‘emergent stage’ is normally led by the father (head) of the family. The capital accumulation is through loans from family and fiiends or trusted business acquaintances. In the second stage, ‘centralization stage’, the father will heavily reinvest the firm’s profits and centralize the control of decision making over money and strategy. In the ‘segmentation stage’, the founder either retires or dies and his sons assume control of the firm and other independent firms established by the father. In the fourth stage, ‘disintegration stage’, the sons of the founder decide to divide the family estate left by their father and run their own firms established by their father in the second stage. Even after the segmentation stage, the sons might cooperate as close colleagues while starting to build their own estates. As to the funding business, personal networks, which include family members, fi'iends, and business partners, serve as an important back-up that allows a more flexible use of capital. In fact, “most investment capital for new projects comes either from reinvested profits or from the entrepreneur’s personal network of family and fi‘iends. In most cases, the personal network is probably the more important source” (Hamilton, 1996:281). Enterprises based on non-kin relations also tend to be dominated by several 6 These results are from Peng’s doctoral dissertation (1989) Relationships among Taiwan Business Owners and Their Changes: A Sociological Analysis (Tunghai University). I cited them from Orru, Buggart and Hamilton (1997:259). I61 family blocks (Mark, 1972:15). Regarding interaction among Taiwanese enterprises, the personal networks, guanxi, of the managers play more important roles than the aid from formal channels. For instance, with guanxi, a kind of trusting relationship, the owner of the enterprise can lower the cost of finding trustworthy workers, mobilizing financial resources, dealing with bureaucratic demands, and finding the correct location of markets (Hamilton and Kao, 1990). In sum, we would expect Taiwanese to resort to family members as their first choice of social support. Beyond the immediate family, lineage and clan become another family-related factors constraining people to strong ties. According to the southeastern Chinese tradition,7 a number of related families will normally agree to establish an ancestral hall to honor an ancestor with whom all members of the family share blood ties. By doing this, these families can not only fulfill their responsibility to worship their ancestors but also form secure networks to improve their common good. Similarly, the clan is an organization established by people of the same surname.8 It emphasizes similaritiesg, which provides easy access and psychological security to ask for potential social support in a society where strangers and associates rarely mix. That is why “lineages, clans and hybrid forms live on to provide support networks for many in Taiwan today” (Davison and Reed, 1998:179). Regarding the non-kin ties and its operation in Taiwanese society, it is actually the focal point of Chinese guanxi networks. I do not intend to recapitulate the concept of 7 As to why southeastern China, especially Fukien province and Kuangtong province, has dense lineage groups, please refer to Freedman (1966). Taiwanese culture and customs, as I mentioned in chapter five, are largely influenced by migrants from especially Fukien. 8 Given that debate still exists over the difference between lineage and clan in the Chinese context (Davison and Reed, 1998:179; Yueh, 1989:81), the truth that they are an extended form of family is no question. 162 guanxi- delineated in chapter five- only that it’s the means Chinese people normally use to connect two seemingly unrelated persons, families or groups. Three commonly seen patterns accomplish this. First, members of two unrelated families get acquainted through marriage. Due to the marriage, members of the two families heap together their social resources. It is especially significant for those who engage in political or business activities. As I mentioned in chapter five, Taiwan’s local political factions play a critical role in various levels of elections. One efficient way to accumulate their social resources is to create a chance for their sons or daughters to develop romantic relationships with people from families with great political influence! If marriage is expected, it means these two families can take advantage of each other’s social networks and hence increase their chances of achieving any political goals they desire.'0 The same phenomenon is also popular with many families that run big businesses for the same consideration. Second, in addition to the paramount roles of family, surname, and lineage groups in Taiwanese society, people’s native place is also being seen as an important source for seeking social support. In chapter five I mention that organizations based on this regionalism is called tong xiang huey. These tong xiang huey are normally organized by migrants from the same village, town or county. Gallin and Gallin (1974) and Yin (1981), for example, provide us detailed facts about how these organization help its members achieve both economic success and mental peace with activities such as lending each other money, helping each other when in need, exchanging guanxi networks and running their own temples. The third way of expanding one’s guanxi network among non-kin relations is through a ritual called ‘swom brotherhood.’ It is basically a social extension of the 9 This is similar to the concept of homophily that sociologists and psychologists use. 163 kinship ideal, which originated on the Chinese mainland (Davison and Reed, 1998:180). The ritual is normally conducted in a temple and the friends swear an oath of mutual assistance and loyalty. Through this ritual their families become each other’s families and members of their families refer to one another in kinship terms, which of course fiirther expands their social resources. This ritual only applies to males- women in Taiwan traditionally enjoy less social prestige than men just as in most other areas of the world. However, they seem to be more likely to form informal ties based on mutual emotional support rather than a more comprehensive social support engaged in men’s social lives. However, we lack empirical data dealing with the differences of types of social support between genders in Taiwan. From the previous discussion, we can boil down some plausible statements: first, it is no doubt that Chinese people, including Taiwanese, are overwhelmed by the power of social connections, i.e., guanxi networks. Through the norm of reciprocity, two parties agree, mostly informally, to exchange favors even in situations where law demands neutrality.ll As Alston (1989:28) said “[T]he moral dimension operating here is that a person who does not follow a rule of equity and refuses to return favor for favor loses face and becomes defined as untrustworthy.” Thus, a person’s social connections work as his social capital that can be appropriated in almost any situation where a favor is needed. '2 '0 For empirical study of this type of connection, see, for example, Gallin and Gallin (1985: 10 l -1 l6). " For a detailed discussion of the operation of guanxi and renqing, see my chapter five. '2 Most literature treats the issue of using guanxi networks among Chinese as a source of many negative social consequences, e.g., corruption. I do not intend to address this issue. For a detailed discussion, please refer to Lu (1996). Second, given the impact of family on Chinese daily life, it is reasonable to expect those who are in need to resort to family members first and to non-kin ties only secondly. We can also infer that under the game of renqing weak ties will be utilized as mediators (bridges) to reach strangers who own social resources in need. Using Granovetter’s words, under the Chinese context, when it comes to asking for help, strong ties go first and weak ties second. Or from Woolcock’s perspective, Taiwanese people tend to live closer to one dimension of social capital, embeddedness,‘3 rather than the other dimension, autonomy. Third, given the ability of weak ties to reach distant strangers, who might be of great help, it is intriguing to ask who, in a society dominated by strong, is likely to utilize weak ties? And how are they different from those who use only strong ties? As a political scientist, not only do the factors determining the probability of using weak ties instead of strong ties interest me but also the political consequences that the difference may cause. I will empirical explore in the following section private social capital, its mechanism and its potential political consequences in Taiwan. PRIVATE SOCIAL CAPITAL, ITS MECHANISM AND POLITICAL CONSEQUENCES: AN EMPIRICAL RESEACHI Hypotheses According to the analytical framework suggested in chapter three, private social capital refers to resources embedded in one’s social relations. In the context of Taiwanese '3 For detailed definitions of embeddedness and autonomy, see chapter three. 165 society, it represents the help available by pulling on guanxi networks. We can further differentiate the guanxi network into strong-ties oriented or weak-ties centered. Thus, the first hypothesis to test is: H, : Most Taiwanese’ private social capital is characterized as strong ties rather than weak ties. Given the instrumental values of weak ties illustrated in literature, a fascinating question is ‘who is more likely to utilize weak ties to reach his or her instrumental goals?’ The answer to this question cannot only be used as a plausible explanation for personal achievement but also shed light on our understanding of the spread of information or ideas that may affect the degree of social integration. For instance, as mentioned in chapter five, the ethnic conflict in Taiwan creates a problem for its ongoing democratic consolidation. It may be helpful to know what backgrounds are more likely to use weak ties and check if the existence of different ethnic backgrounds in his weak ties can generate a more tolerant attitude towards many critical political issues. I argue that personal endowment and personality are very critical to the appropriation of his private social capital. The ability to recognize the potential contacts embedded. in his guanxi networks and the willingness to appropriate them decide the final outcomes. Since keeping weak ties requires risking the break up of renqing or reciprocity, those who can afford this risk or loss should be more likely to appreciate the instrumental value of using weak ties in his private social capital. Therefore, personal backgrounds, i.e., social positions, which can create a heightened sense of security of social relations, should lead 166 the individual to appropriate more weak ties in his social capital stock. Thus I next hypothesize: H 2 : Those who have a greater sense of security will be more likely to use weak ties as opposed to those who do not. In his influential article The Strength of Weak Ties (1973:1373-1376), Granovetter suggested that individuals with few weak ties were less likely to be able to mobilize for collective action within their communities. By studying the social ties of the initiators and initial recruits in five conflict groups mobilized around [different] educational issues in a suburban community, Steinberg (1980) found that theses initiators (seven) are only marginal individuals in the community. In addition, as Granovetter’s theory predicts, the group recruited on the basis of strong ties was linked to the fewest organizations. Among those more successful groups, i.e., groups formed on the basis of weak ties, they “were able to recruit some individuals who had occupied leadership positions...The evidence suggests, tentatively, that where innovations are controversial, a mobilization strategy based on the activation of weak ties is more likely to facilitate adoption of the goal and integration into the school decision-making structure” (Steinberg: 1980:25). Again, the argument lies in the ability of weak ties to reach out for sources otherwise unreachable. From the perspective of reducing the costs of collective actions, I am interested in the potential effects of weak ties on individuals’ degree of community involvement and associational activities. Are those who more often use weak ties more likely to participate in collective actions? In the context of community involvement, are they more actively involved in community activities than those who only concentrate on strong ties for social I67 support? Are any patterns recognizable between the two types of people in terms of participating in social or political organizations and related activities? In the context of Taiwanese society, there are at least two factors that need consideration before formulating a hypothesis on the above questions: first, the idea of local community is not as common as it is in the West. According to a nationwide survey, the 1997 Taiwan Social Change survey, over 75% of the respondents never used the word ‘our community’ to describe the place they live. Among these respondents, around 64% of them thought their place is either hard to regard as a community (8%) or is not a community at all.14 This can be explained by the fact that Chinese people are traditionally apathetic to local community affairs because they focus so much on family life. Second, when it comes to the social interactions, Chinese people are extremely sensitive to other people’s sentiment. Closely related is the tendency to avoid as many potential conflicts as possible.'5 Once conflict happens, Chinese people tend to opt for non-confrontational approaches to their resolution, such as resorting to mediators or arbitrators instead of face-to-face negotiation, which is widely believed to be the practical way to solve hostility or conflicts in the West (Bond, 1991 :66). In addition, the degree of autonomy in dealing with one’s social relations should provide some incentives for smooth entry into community affairs. Massey, in an article challenging Taylor and Singleton’s (1993) idea of the communal resource as a solution to collective action " These descriptive statistics are calculated by the author. The size of this survey is 2,835. '5 This argument is best illustrated by a study showing that Chinese students prefer social recognition to personal gain from social exchange. Specifically, when put in a team, those who performed poorly were willing to share the gain of the team work in proportion to their performance while those who performed excellent were willing to have equal share of the gain. In other words, keeping the harmony of the group is more important than the rule of fairness. 168 problems,I6 contended that “[I] n a situation in which people are more isolated and autonomous, it is easier to slip out of a movement, but it is also less risk to slip into one” 91994z428). Thinking strength-of-weak-ties, we will expect that those who cash their private social capital through the use of weak ties are more likely to become involved in community affairs: the more information and farther linkage they can reach, the more autonomy they have in dealing with the complex guanxi networks requisite to community affairs. Thus, the third hypothesis follows: H 3: Those who resort more to weak ties are more likely to participate in local community affairs than those who depend mostly on strong ties. The last issue I want to tackle is the actual mechanism operant in the guanxi network. The main purpose of this research is to demonstrate that although the norm of reciprocity is the most important social norm facilitating the realization of social capital, the reciprocal behaviors (repeated social exchanges) between the two parties are not as symmetrical as we would expect in the economic exchanges.l7 In a society where the importance of guanxi is so consistently recognized by its members,'8 it is expected that those holding more resources (both economic or social) will be more likely asked by '6 Taylor and Singleton (1993) argued in their article that the presence of ‘community’ will lower the transaction costs of collective action due to a set of characteristics that a community holds, i.e., stability of relations, multiplex relations, direct relations, shared beliefs and preferences. Massey, on the other hand, used a case of environmental protection in a small community to illustrate ‘how personal interdependence can combine with the lack of anonymity characteristic of small communities to make social action a high- risk undertaking” (Massey, 1994:423). '7 Although each economic transaction is not always balanced in terms of gain and cost between two parties, “it is strongly assumed that the balance of credits and debts will be achieved in the long run, but in a finite time frame, in repeated transactions” (Lin, 2001: 151). '3 I make this argument based on chapter five and empirical data. In the Taiwan Social Change Survey, 1992 Part I, a list of conditions was presented to the respondents, asking them if they feel it important to one’s career success. Regarding ‘being helped by an important person,’ the mean scores, calculated by 169 others to give favors, whatever the nature of the favors are. As to why people holding more social resources willingly act as others wish, i.e., giving more help than they will receive, it can be explained from the perspective of social exchange. After reviewing literature on human exchange,l9 Lin developed his own arguments concerning social exchange, synthesizing the two elements involved in human exchange: relationship and transaction of resources. He further differentiated two rationalities involved in human exchange: transactional rationality and relational rationality. Transactional rationality focuses on the calculations of transactional gains and costs in exchange, and relational rationality propels the calculations of relational gains and costs (Lin, 2001:150). Since these two rationalities weight on economic standing (wealth) or social standing (reputation) in exchanges, and since each society at a particular time may favor one over the other (Ibid: 151), it provides a useful conceptual tool to observe and understand the actual operation of reciprocal behavior and related issues. For Chinese societies, including Taiwan, where an individual’s social reputation (face) and maintaining social relations are paramount, we would expect relational rationality to dominate social exchanges. In other words, transactions in exchanges may be of secondary importance. Therefore, on the one hand, through the operation of guanxi and assumption of the pyramidal structure of social resources, those who own more social resources will be more likely to give help than receive it. On the other hand, through these imbalanced social exchanges those who give more help are actually accumulating more social capital, in Coleman’s words, credit slips. This capital is accumulated mainly through the author using SPSS, are statistically insignificant in terms of gender, ethnic backgrounds, age cohorts, level of education and occupations, which implies a wide consensus among Taiwanese people on this issue. ‘9 For details, please see chapter nine of Lin (2001). 170 spreading of a good reputation by the debtors in the imbalanced exchange.20 For example, the different levels of lawmakers in Taiwan own more social resources. They always claim to work for the best interests of their constituencies. In order to fulfill this promise, they have to respond to constituents’ requests, including activities such as participating in weddings, family firnerals, community affairs, or even protecting unlawful activities or businesses of their constituencies by pressuring law enforcement agents. Their favor giving is widely regarded as exchanging electoral support in return, which is, quantitatively speaking, disproportional to the help he or she has given. Therefore, my fourth hypothesis is as follows: H 4: Those who own more social resources tend to give more help than they will receive.2| The following section will use available survey data on Taiwan to test the four hypotheses. Data, Measurement and Method Data Source 2° For detailed discussion of such relationship between creditor and debtor, please see chapter nine of Lin (2001) 2' The amount of help is measured by the frequency and not by the nature of the help, which is difficult to quantify. I71 This empirical study is a secondary analysis based primarily on a nation-wide survey called No.3, the fifth Taiwan Social Change Survey in 1997. As to the detailed description of this data set, please refer to chapter four. Measurement I will briefly describe the variables used in this empirical study and how they are operationalized. Regarding the measurement of patterns of private social capital, i.e., strong tie and weak tie, please see chapter four for a detailed description. Controlled Variables The following variables will be used as controlled variables in this chapter’s multivariate analysis: Gender: male respondents are coded as 1; female respondents are coded as 0. Ethnic Background: respondents are assigned their ethnic background based on their father’s. There are four ethnic backgrounds in Taiwan: Taiwanese”, Hakka, Mainlander, and aborigine.23 In regression analysis, this variable will be broken into three dummy variables. Age: it measures the actual age of the respondents in 1997. This variable has been recoded by the author into five age cohorts, with each IO-years as the interval, starting 22 Taiwanese here refers specifically to those whose ancestors came from Fukien Province of China in Ch’ing Dynasty. They comprise the majority of the population in Taiwan. For more details, please see chapter five. 172 from 20 to 30 years old to 61 to 74 years old. Thus there will be four dummy variables in the regression analysis. Contacts Made Per Day: this variable measures how many people on average the respondent contacts a day, including via the Internet. It contains both those known previously by the respondents and those not known. I regard this variable as measuring the potential width one could reach in his or her social relations. Education: this variable measures respondents’ level of education. Original variable includes the highest level of schooling a respondent has attended. Six rank- ordered categories have been created from not receiving education to college and graduate. Five dummy variables are included in regression analysis. Income: measures respondents’ individual monthly income and contains 23 categories ranging from no income to over $300,000 (New Taiwan Dollars, roughly equal to around $10,000 US dollar). Occupation: original variable assigns respondents’ occupation based on a new chart of occupational categories (Hwang, 1999:236). For simplicity of analysis, this variable is recoded into four broader categories, including house wife/no formal job, skilled and unskilled worker, semi-professional/non-manual worker, and professional/managerial. Three dummy variables are included in the regression analysis. Years of residency: measures how long a respondent has lived in current residence. It is a rank-ordered variable ranging from within one year to above twenty years. 23 For general information about the demography of Taiwan, please refer to chapter five. 173 The Degree of community identity: it measures how strongly the respondents feel about membership in the community/surroundings. It is a five-point scale. Dependent Variables This research aims to study the effects of types of private social capital on local community involvement and the relationship between social resources and reciprocity. Thus local community involvement and reciprocity are the dependent variables. 1) Local Community Involvement In fact, local community organizations are not very common in Taiwan although their numbers are rising. For instance, as of 1998, the number of community development associations has risen from 2,312 in 1993 to 4,968 in 1998 (DGBASEY, 1999:180). However, the percentage of persons participating in these organizations was only around 3.6% in 1997 (Ibid). According to the 1997 Social Change Survey, around 31.5% of the respondents reported having formal local community organizations. A majority (nearly 70%) had no organization dealing with collective affairs at the local level except tsuen or li, the smallest formal administrative unit. For those having self-established community organizations, three variables measure the degree of community involvement: paying fees for the organization, participation in the activities of this organization, and oflicership in the organization. These three variables are recoded into dummy variables. As mentioned, the majority of the sample did not have formal local community organizations. It may be heuristic to explore their attitude toward this kind of organization. In fact, the 1997 Social Change Survey did ask them if they thought it 174 necessary to have a community organization handling collective affairs and in what way would they be willing to support this organization. There are five different ways to show support: paying fees to the organization, donation, oflicership in the organization, volunteering, and participation in the activities. These five variables are dichotomized and will be used as dependent variables to test H 3. In addition, these will be combined to form a single index measuring potential community involvement.24 There is a potential threat regarding the validity and reliability of this measure. Whether we actually measure the degree of local community involvement from those who has not had such local community organizations is not known. These respondents may inflate their willingness to participate in these hypothetical community affairs just because it is a social virtue to care about society. Regarding reliability, we do not have other comparable surveys to test the reliability of this measure. By using this measure, I simply assume ideas will guide actual behavior. 2) Reciprocal relations in instrumental ties The test of reciprocal relations in one’s instrumental ties involves two major questions. The first asks: “ how often do you help each of your instrumental helpers?”; the second asks: “comparing yourself with the help-givers you listed, who has more face, i. e., more social resources?” The answers for the first question included (1) never, (2) seldom, (3) sometimes, and (4) very often.25 An index of reciprocity is created to measure how often the respondent helps his or her helpers. It is calculated as the mean of at most five questions. If, for example, a respondent lists four helpers previously, then he would 2’ For detailed results of this factor analysis, please see Appendix6-2. 175 provide four answers to the above questions, one for each helper. A mean of these four answers is then calculated as the index of reciprocity. The original questions are on a four-point scale. Thus, the index can be regarded as a continuous variable ranging from 1 to 4. As to the comparative social resource,26 six answers were provided for the respondents: he/she owns (1) from much more social resoUrces than me to same amount, (2) a bit more than me, (3) similar amount, (4) less than me, (5) far less than me, i.e., mine is much more and (6) don’t know.27 I create an index of respondents’ average social resources relative to his instrumental helpers by: first recoding the five answers dichotomously, with 1 referring to help-givers own more social resources and 0 indicating the respondent has more social resources; and then calculating the mean score of these dichotomous variables. Third, the mean score is recoded as one if it is greater than .5 and zero if otherwise. Thus, a score of one indicates that the respondent’s help-givers on average hold more social resources; and zero shows that the respondents have more social resources. In fact, this index measures the relative social resources of the respondent within the instrumental network. Method I first conduct single variable analysis, including frequency and percentage distributions of the major independent variables. Then bivariate analysis is performed to test the relationship between dependent variables and independent ones. Lastly, 2’ l recoded the original variable such that the higher the number, the more help the respondent had given. 26 Note that this variable will serve as an important independent variable in testing the relationship between reciprocity and the holding social resources. 176 multivariate analysis tests the net effect of each independent variable on the dependent variable, controlling for other variables. Results and Discussion Table 6-1 presents the frequencies and percentages of major independent variables in this empirical research. As we can see, this sample consisted of almost equal percentages of males and females. Nearly half (49.2%) of the respondents of were between the ages of 31 and 50. Respondents with Taiwanese backgrounds dominate the whole sample as expected. Hakka and mainlanders share similar percentages of the sample. Aborigine consists of less than one percent of the sample. More than 50% of the respondents have at least a senior high school degree. Over 56% of them use only strong ties when help is needed. (Thus, hypothesis one is preliminarily proven.) Slightly over 50% of the respondents can be regarded as having more social resources than their contacts in the instrumental networks. Near half of them lived in the same area for over 20 years. Regarding community identity, near 40% of them report having no concern. 45.6% of the respondents show community identity. Bivariate analysis will be performed in the next section in order to quantify the relationship between dependent and independent variables. 27 The category of ‘don’t know’ is treated as a missing value. 177 Table 6-1 Frequencies and Percentages of Major Independent Variables n % Total Gender 2835 Male 1444 50.9 Female 1391 49.1 Age 2835 20 to 30 683 24.1 31 to 40 743 26.2 41 to 50 652 23.0 51 to 60 371 13.1 61 to 74 386 13.6 Ethnic Background 2824 Taiwanese 2208 78.2 Hakka 305 10.8 Mainlander 293 10.4 Aborigine 18 .6 Education 2833 No 268 9.5 Elementary 658 23.2 Junior high 419 14.8 Senior high 808 28.5 Junior college 375 13.2 College & above 305 10.8 Occupation 2830 House wife/no formal job 752 26.6 Skilled and unskilled worker 958 33.9 Semi-professional /non-manual worker 884 31.2 professional/managerial 236 8.3 Income Scale " 2723 Character of Private Social Capital 2664 Strong ties 1494 56.1 Weak ties 1 170 43.9 Social Resource Comparison 2603 Respondent has more 1377 52.9 Contacts have more 1226 47.1 178 Table 6—1 (Continued) n % Total People Contacted in Work 2825 0 to 4 245 8.7 5 to 9 561 19.9 10 to 19 784 27.8 20 to 49 671 23.8 50 to 99 314 11.1 Over 100 250 8.8 Residence History 2833 Less than 1 year 85 3.0 l to 4 years 338 11.9 5 to 9 years 338 11.9 10tol4 years 429 15.1 15to 19years 320 11.3 More than 20 years 1323 46.7 Community Identity 2825 Very not important 16 .6 Not important 416 14.7 Do not care 1105 39.1 Important 955 33.8 Very Important 333 11.8 a. It contains 23 categories, from no income to above NT$300,000. For the simplicity of displaying data, I will not list them in this table. However, I collapse those items into four categories: under NTS 1 0,000, between NT$10,000 to NT$30,000, between NT$30,001 to NT$50,000, and over NT$50,0001. The percentage distribution is 34.1%, 26.3%, 21.9%, and 17.7% respectively. From the bivariate analyses, listed in Table 6.2, we can tell that respondents resorting to only strong ties are more likely to be females (p<.000), elderly people (p<.000), less educated ones (p<.000), people with jobs requiring less professionalism (p<.000), people within lower income brackets (p<.000), ones having similar amounts of social resources as opposed to those in his or her instrumental ties (p<.003), and those who contact less people in the workplace (p<.000). On the contrary, respondents who have at least one weak tie in their instrumental networks tend to be younger generation males, those with a higher level of education, ones with professional jobs, people with 179 higher income, those possessing more social resources than his or her contacts in the networks, and the ones who generally contact more people in the workplace. Ethnic background does not influence the selection pattern of private social capital. Table 6.2 also provides some information necessary for my hypothesis tests. First, it further proves that Taiwan is a strong-ties dominated society. For example, controlling for the amount of social resources, strong ties are still preferred choices (52.7% and 58.8% respectively) when help is needed. The same is true if controlling for gender or age (see Table 6.2). Second, my hypothesis two (H2) is proven if we equate higher social positions with more social resources and hence a heightened sense of security.28 Maleness, a higher level of education, higher income, and professional jobs are commonly regarded as contributing to social prestige and hence characterize higher social positions. 2’ The principle of homophily renders an assumption that is widely accepted in sociology, i.e., the structural opportunity for reaching better social resources is much better for those whose initial positions are relatively high (Lin, 1982:133-134). As to the relationship between social resources and sense of security, 1 speculate these two concepts are positively related, i.e., the more social resources one owns, the more secure he will feel about and toward society. In order to support my speculations, I conducted a correlation analysis in the same survey, the 1997 Social Change Survey. In this survey, respondents were asked if they ever felt uneasy and could not trust others in the past week. The results show significant negative relationships between the amount of social resources and the two variables (r =-.O72 and -.057 respectively). That is, those who own more social resources felt less uneasy and were less likely to feel that other people cannot be trusted. 180 Table 6.2 The Distribution of Respondents’ Characteristic of Private Social Capital Strong ties Weak ties x 2 Gender 33.49, p<.000 Male 45.1% a (50.5%) ” 56.4%(49.5%) Female 54.9%(61 .7%) 43.6%(38.3%) Age 114.11, p<.OOO 20-30 19.0%(42.6%) 32.7%(57.4%) 31-40 26.8%(56.0%) 26.8%(44.0%) 41 -50 21 .6%(54.3%) 23.2%(45.7%) 51-60 15.3%(67.0%) 9.7%(33.0%) 61-74 l7.3%(74.4%) 7.6%(25.3%) Education 230.78, p<.000 No 13.1% 4.4% Elementary 26.9% 16.8% Junior high 17.5% 11.1% Senior high 27.3% 30.8% Junior college 8.9% 19.7% College and up 6.2% 17.1% Occupation 84.27, p<.000 House wife/no formal job 31 .9% 20.3% Skilled & unskilled worker 35.3% 31.1% Semi-professional /non-manua1 worker 26.9% 3 7. 1% professional/managerial 5.8% 1 1.5% Income " 114.45, p<.000 Low income 41 .7% 25.0% Lower middle income 27.1% 25.0% Upper middle income 18.7% 26.4% High income 12.5% 23.7% Ethnic Background ns " Taiwanese 79.2% 76.3% Hakka 10.5% 11.8% Mainlander 9.8% 1 1.3% Aborigine .6% .7% Social Resource Comparison 9.73, p<.003 Respondent has more 50.2%(52.7%) 56.3%(47.3%) Contacts have more 49.8%(58.8%) 43.7%(41 .2%) 181 Table 6.2 (Continued) Strong ties Weak ties x 2 People Contacted in Work 72.86, p<.000 0-4 10.5% 5.0% 5-9 22.7% 16.1% 10-19 28.1% 27.7% 20-49 22.7% 25.2% 50—99 9.1% 14.5% Over 100 7.0% 11.5% a. The entry in each cell is calculated as column percentage. b. Cell entries in parentheses are row percentages c. 23 original categories are divided into four broad income brackets. d. Not statistically significant. Regarding the relationship between community involvement and types of private social capital, Table 6-3 provides the information necessary for testing my third hypothesis (H 3 ). As we can see from the first three variables, for respondents who had voluntary community organization(s), the types of private social capital make no statistical difference in terms of community involvement. In other words, the strong or weak ties difference cannot explain the variation in community involvement. Hence H 3 is not supported at this time by the empirical data. However, as I mentioned previously, the majority of Taiwanese have not experienced local voluntary community life. How much would they be involved in local affairs if they had organizations that deal with collective affairs? Will different types of private social capital affect the degree of involvement? The last five variables on Table 6-3 appear to support H3. The 1997 Social Change Survey in Taiwan asked respondents who did not have local community organizations whether they would pay fees, make donations, take community office, volunteer and participate in as many of the activities of the organization as possible. Respondents with 182 weak ties are significantly more likely to give positive answers except for the making donations. Table 6-3 Community Involvement and Types of Private Social Capital Strong ties Weak ties x Paying community fees“ 1.02, us No 34.3% ” 30.8% Yes 67.7% 69.2% Participating in community activities .31, ns No 56.6% 58.6% Yes 43.4% 41.4% Taking community office 1.02, ns No 87.7% 85.3% Yes 12.3% 14.7% Paying fees " 7.58, p<.007 No 27.3% 20.2% Yes 72.7% 79.8% Making donation 1.18, ns No 39.0% 35.9% Yes 61.0% 64.1% Taking office 20.19, p<.000 No 79.9% 68.1% Yes 20.1% 31 .9% Being a volunteer for the community organization 14.08, p<.000 No 41.6% 30.7% Yes 58.4% 69.3% Participating in organization’s activities 6.45, p<.000 No 23.0% 16.8% Yes 77.0% 83.2% a. The following three items were measured among respondents who already have voluntary local community organization(s), which consisted of 3 1 .5% of the sample. b. Cell entries represent column percentage. 183 c. The following five hypothetical questions were asked to respondents, who have no voluntary local community organization(s), whether they would show these behaviors if a local community organization existed. In order to estimate the net effects of the independent variables on the dependent variables in my hypotheses or to make sure the relationships found in bivariate analysis are not spurious, multivariate analysis is needed. Tables 6-4 through 6-6 present these analyses for testing hypothesis two (H 2 ) and hypothesis three (H 3 ). The results of Table 6-4 further support H2 , i.e., people with higher social position or more sense of security will be more likely to use weak ties as their private social capital. For instance, other variables being constant, the odds of favoring weak ties are 1.23 times higher if the respondent is a male rather than a female. Specifically, controlling for other variables, males, and younger generations as opposed to those over 61 years old, people with higher socioeconomic status, and contacting more people in the workplace generally have a higher probability of selecting weak ties for help. Table 64 Effects of Social Positions on the Patterns of Private Social Capital: Logistic Regression Independent Variables Weak ties " Odds Ratio Gender Male .21* 1.23 (.09) C (reference: female) Age Cohort 20 to 30 .87" 2.38 (.16) 31to40 .28+ 1.31 (.16) 41 to 50 .52" 1.68 184 Table 6-4 (Continued) Independent Variables Weak ties " Odds Ratio (.16) 51 to 60 .20 1.22 (.18) (reference: 61 and above) Socioeconomic Status b .43" 1.54 (.05) People contacted in working place 5 to 9 .29 1.34 (.19) 10 to 19 .46* 1.59 (.18) 20 to 49 .33 + 1.39 (.19) 50 to 99 .62” 1.86 (.21) 100 & above .62" 1.85 (.22) (reference: 0 to 4 people) Constant -1.18** (.20) N 2,560 Model x2 (p) 256.87 (.000) a. Respondents with weak ties are coded as 1, strong ties as 0. b. For the sake of parsimony, l conduct a factor analysis combining respondents’ level of education, occupation, and income level. A principal components analysis of these three variables produced one significant factor: the standardized factor scores were used as the measure of socioeconomic status in this regression. For detailed results of this factor analysis, please see Appendix 6-1. c. Standard errors (s.e.) in parentheses. + p<.10, * p<.05, " p<.01. With regard to the relationship between potential community involvement and people with different types of private social capital, Table 6-5 provides five logistic 185 regressions with each one of the five variables measuring potential community involvement as a dependent variable. As the results show, holding other variables constant, the odds of favoring becoming a community staffer or volunteer, two of the highest level of community involvement, are e'43=1.54 and e'39=1.48 times higher respectively if the respondent’s private social capital contains at least one weak tie. Another way of testing the hypothesis is to combine all five variables as an index of community involvement and run an OLS regression to see if the types of private social capital still have effects on it. Table 6-6 shows the results of this OLS regression. Table 6-5 Private Social Capital and Potential Local Community Involvement: Logistic Regression Independent Variables I“ II" III" IV" V " Gender Male .27+ .35" 34* .07 19 (.17)" (.14) (.16) (.14) (.17) (reference: female) Age Cohort 20 to 30 .32 .21 1.02* .78" 1.09“ (.29) (.27) (.44) (.27) (.30) 31 to 40 .57+ .41 1.00* .89" 1.03" (.30) (.28) (.44) (.28) (.30) 41 to 50 .43 .32 1.00‘ .71' .83" (.30) (.28) (.44) (.28) (.30) 51 to 60 11 -.04 1.34" .47 51 (.31) (.30) (.46) (.30) (.31) (reference: 61 and above) Socioeconomic Status .48" .27" .26" .07 .15 (.10) (.08) (.08) (.08) (.10) Residence History 186 Table 6-5 (continued) Independent Variables l" 11” 111" IV" V" Less than 1 yr. -.03 .58 1.54" 1.12" .24 (.43) (.42) (.39) (.48) (.48) 1 to 4 yr. .75 ** .56* .37 .40* .64“ (.27) (.23) (.25) (.22) (.29) 5 to 9 yr. .70" .47* .10 .28 .70* (.26) (.22) (.24) (.22) (.29) 10 to 14 yr. .64" .28 .48* .27 .23 (.22) (.19) (.21) (.19) (.22) 15 to 19 yr. .92" .43* .30 .43“ .34 (.27) (.21) (.24) (.22) (.26) (reference: 20 yr. and above) Community Identification Low -1.19 .74 .29 1.56+ 1.63“ (1.14) (.78) (1.10) (.84) (.79) Medium -1.22 .81 .40 1.42+ 2.00* (1.13) (.77) (1.10) (.83) (.77) High -.65 135+ .96 2.11* 2.51" (1.13) (.77) (1.10) (.84) (.78) Very high -.16 1.76‘ 1.10 2.40" 3.20" (1.16) (.80) (1.10) (.86) (.83) (reference: very low identification) Types of Private Social Capital Weak ties .08 -.09 .43" .39" .15 (.16) (.14) (.15) (.14) (.17) (reference: strong ties) Constant 1.23 -1.16 -3.41 -2.3 -1.99 (1.15) (.80) (1.16) (.86) (.81) N 1070 1071 1070 1075 1076 Model x 2 113.12 79.85 88.40 82.38 80.39 (p) (.000) (.000) (.000) (.000) (.000) a. A dichotomous variable showing whether a respondent is willing to pay fees for the local community’s organization. b. A dichotomous variable showing whether a respondent is willing to make donation to the local community organization. c. A dichotomous variable showing whether a respondent is willing to be a staff member of an organization. d. A dichotomous variable showing whether a respondent is willing to be a volunteer of the organization. 187 e. A dichotomous variable showing whether a respondent is willing to participate in the activities of such an organization. f. Standard errors in parentheses. + p<.l, " p<.05, " p<.01. As Table 6-6 shows, the type of private social capital still affects the degree of potential community involvement, controlling for gender, age, socioeconomic status, residence history, and community identification. Specifically, people with weak ties in their instrumental networks have a higher degree of community involvement (p<.05) if there was a community organization dealing with their collective affairs. As to controlled variables, males, younger generations as opposed to those 61 and above, and people with higher socioeconomic status will have higher levels of community involvement if there was a local community organization. One interesting fact pertaining to the relationship between the years of residency and community involvement: we see from the Table that all categories of residency history showed higher levels of involvement as opposed to those who lived in the same area for over 20 years, which consisted of near 47% of the sample. Thus, the empirical evidence gives us inconclusive results in terms of testing hypothesis three (H 3 ). The sample from those who have already had community organizations does not support hypothesis three (H 3 ). Still, respondents who have not yet had formal community organizations, over two thirds of the sample, provide encouraging evidence in favor of hypothesis three (H 3 ). Thus, future studies are needed to explain why there is such a difference. 188 Table 6-6 Private Social Capital and Local Community Involvement: OLS Regression Independent Variables Potential Community Involvement " Gender Male .153" (.060) ” (reference: female) Age Cohort 20 to 30 .460" (118) 31 to 40 .508" (120) 41 to 50 .448" (120) 51 to 60 .307“ (.130) (reference: 61 and above) Socioeconomic Status ‘ .138* "‘ (.032) Residence History Less than 1 yr. .437" (.165) 1 to 4 yr. .327" (.094) 5 to 9 yr. .258" (091) 10to 14 yr. .245” (080) 15 to 19 yr. .295" (.090) (reference: 20 yr. and above) Community Identification Low .493 (328) Medium .521 189 Table 6—6 (continued) Independent Variables Potential Community Involvement " (.323) High .882" (324) Very high 1.05* (.33 1 ) (reference: very low identification) Types of Private Social Capital Weak ties .1 18* (.059) (reference: strong ties) Constant -1.371"‘ "‘ (336) R2 .156 F ratio 12.028" N 1061 a. The five original variables showing different degrees of potential community involvement, i.e., paying fees, making donations, taking organizational office, volunteering for a community organization, and participating in an organization’s activities are combined for a factor analysis. The principal components analysis yielded one significant factor. The standardized factor scores were used as the measure of potential community involvement in this regression. For detailed results of this factor analysis, please see Appendix 6-2. b. Standard errors (s.e.) in parentheses. c. Socioeconomic status is measured the same as the one used in Table 6-4. + P<.l, “ P<.05, ** p<.01. Is reciprocity symmetrical? 190 The last part of this chapter analyzes reciprocity. Although it is widely recognized as the most important mechanism or social norm to maintain social capital, little research has been devoted to the exploration of the actual operation of such mechanism. Given the assumption of pyramidal structure of human society and the positive relationship between social position and command of social resources (Blau, 1977; Lin, 1982), we expect that people of a higher social position or with more social resources will be asked to give more favors. Thus, due to unequal social resources embedded in both sides engaging social exchange, the norm of reciprocity will not run symmetrically. As to why the so- called creditor wants to maintain such a relationship? This is where Fukuyama’s ‘recognition’ thesis (1995) and Lin’s ‘reputation’ argument (2001) come into play.29 In testing hypothesis four (H 4 ) I wish to see to what extend the above argument can be proven in the context of Taiwan, and 1 want to show that social exchange is not as symmetrical as we normally think due to reciprocity. Table 6-7 shows that, holding other independent variables constant, respondents with more social resources than his or her instrumental contacts tend to give more help to the contacts.30 In other words, these people are more attractive to others as contacts in their own guanxi networks. In contrast, respondents who own less social resources than their contacts tend to ask more help than they give. However, their reciprocal behavior may differ in quality than that received from high social resource holders. For example, in order to win the next election, various levels of lawmakers in Taiwan (high social resource holders) have to respond to many kinds of requests from their constituencies as 29 Actually, both of the author’s arguments are similar in that they seek whether the giving of social resources can be returned as a metaphysical income for the creditor through the reciprocal mechanism, which can be further appropriated if needed. 191 mentioned previously. Normally, these lawmakers only ask for support in elections held every three or four years. Thus, the desire to be recognized may be the most important explanation for such unbalanced reciprocal behavior. In addition, holding other independent variables constant, respondents with pure strong ties tend to help each other more than those with weak ties, which supports Granovetter’s argument and the general findings of many ethnographic studies (Granovetter, 1982; Stack, 1974; Lomnitz, 1977). Table 6-7 Reciprocity and Social Resources: OLS Regression Independent Variables Index of Reciprocity " Gender Male -.056+ (.03) " (reference: female) Age Cohorts 2O t030 .247" (.052) 31 to 40 .171" (.052) 41 to 50 .126" (052) 51 to 60 .067 (057) (reference: 61 & above) Socioeconomic Status h .096" (017) Comparative Social Resource ‘ Contacts have more S.R -.O9l " (029) (reference: a respondent has more S.R.) Types of Private Social Capital Weak ties -.080** 30 This variable was coded as one if contacts, on average, have more social resources than the respondents. 192 Table 6-7 (continued) Independent Variables Index of Reciprocity " (.030) (reference: strong ties) Constant 2.99" (.05) R2 .041 F ratio 13.345" N 2507 a. See measurement section of this chapter for a detailed description. b. It is measured the same way as those used in previous tables. c. Please refer to the previous “measurement” section of this chapter. (1. Standard errors in parentheses. + p<.l, * p<.05, ** p<.01. CONCLUDING REMARKS This chapter has examined private social capital literature with special attention to that dealing with the concept of social networks. Inspired by the idea of ‘strength of ties’ and its close resemblance to the Chinese notion of guanxi, the author applied this line of thinking to one of the Chinese societies, Taiwan, to see how well this theory fits empirically the Taiwanese populace. Generally speaking, the case of Taiwan supports the ‘strength of weak tie’ theory. Due to its cultural influence, Taiwanese society is a strong-tie dominated society. However, those who have a higher sense of security and higher social position do have a higher tendency to form weak ties in their social networks. As to the hypothesis that 193 people with weak ties tend to participate in collective actions, the results are encouraging but not conclusive. Using local community affairs as indicators of collective action, the results do not support this hypothesis. However, for the majority of the people who did not have a local community nearby, the results support the hypothesis. Future study is needed to explain the difference. As far as the testing of the symmetrical aspect of reciprocity, the results also support the hypothesis that those who own more social resources tend to give more help than they will receive. Three considerations are believed to make the reciprocity skewed: the pressure of renqing, accumulating social capital, and the desire for recognition. 194 All 45?] APPENDICES 6-1. Factor Analysis of Socioeconomic Status Variables Factor Loading (1) Level of education (6- point scale) .712 (2) Occupation (4-pint scale) .845 (3) Income level (23-point scale) .829 % of variance explained 63.61 6-2. Factor Analysis of Potential Community Involvement Variables Factor Loading (1) Paying fees .706 (2) Making donation .687 (3) Taking this organization’s office .582 (4) Volunteering in this community organization .709 (5) Participation in community organization’s activities .666 % of variance explained 45.10 195 CHAPTER 7 PUBLIC SOCIAL CAPITAL AND ITS POLITICAL CONSEQUENCES INTRODUCTION Current literature on public social capital has placed the roles of civil society and trust in the political science spotlight. Public social capital here refers to a society’s voluntary associations and those norms or values that are conducive to cooperation among social actors.1 Thus the ability to associate and/or membership in voluntary associations, and interpersonal trust will be the two major indicators of public social capital discussed in this chapter. As I mentioned in chapter three, sociologists seem reluctant to accept the idea of public social capital and think it a conceptual stretch from the original social capital idea (Portes, 1996 & 2000; Lin, 2001). Nevertheless, a tremendous amount of research has centered on Putnam’s idea of (public) social capital. This trend has resulted in an interdisciplinary business, covering a variety of topics such as the study of trust, civil society, and issues of development. In this chapter I first conduct a limited critical literature review on public social capital and democracy, which is the center stage of public social capital literature. Then hypotheses are made based on the controversial issues pointed out in the literature review. The third part is an empirical test of the hypotheses using survey data collected in Taiwan. Results and interpretation will be based on the analytical framework proposed in chapter three. ' In fact, there are fewer consensuses on the definition of (public) social capital among political scientists than consensus on private social capital among sociologists. For reasons, please see my footnote 22 of chapter three. 196 CRITICAL LITERATURE REVIEW Robert Putnam is commonly recognized as the leading figure among social capital theorist.2 However, he is not clear about the concept of social capital in many of his works. Although his most widely cited definition of social capital is “features of social organization, such as trust, norms, and network, that can improve the efficiency of society by facilitating coordinated actions” (Putnam, 1993b: 167), he also delineated the relationship among these three essential components of social capital by saying “social trust in complex modern settings can arise from two related sources—norms of reciprocity and networks of civic engagement” (Ibid: 171). In another influential article Bowling Alone: America ’s Declining Social Capital, Putnam argued “people who join are people who trust...the causation flows mainly from joining to trusting” and “the greater the density of associational membership in a society, the more trusting its citizens. Trust and engagement are two facets of the same underlying factor—social capital” (Putnam, 199Saz73). In addition, he wrote “[N]etworks of civic engagement foster robust norms of reciprocity” (Putnam, 1993b: 173). Thus, we are seemingly able to conclude that Putnam’s “network of civic engagement”, i.e., intense horizontal interpersonal communication, is the generator of social trust and norms of generalized reciprocity. Given these vagueness of his theory, scholars influenced by him tend to either focus on networks of civic engagement as a society’s social capital and treat social trust and norms 2 In fact, the contemporary source of public social capital literature should be traced back to Almond and Verba’s The Civic Culture, in which they systematically explored the effects of social structures and political culture on democratic stability in five countries. 197 of reciprocity as its consequence (Fox, 1996; Evans, 1996; Woolcock: 1998), or treat social trust and social norms as social capital and regard the mechanism of repeated interaction including networks of civic engagement as the sources of social capital (fukuyama, 1995 & 1999;Uslaner, 1999; Whiteley and Seyd, 1997). As I discussed in chapter three, we cannot ignore the importance of norms of reciprocity in maintaining social capital, but it should not be deemed social capital itself. For norms of reciprocity will naturally occur with repeated social interaction given the assumption that humans have the cognitive ability and emotions to distinguish traitors from cooperators (Gouldner, 1960; Levi-strauss, 1965; Fukuyama, 1999). Thus, it is plausible to say that interpersonal trust and civic engagement should be deemed two major components of public social capital referred to by political scientists and developmental theorists. The following literature review will therefore focus on the relationship between public social capital and democracy. The Paradoxical Relationship Between Democratic Polity and Trust: Tension vs. Reinforcement The relationship between public social capital and politics has been the major focus of public social capital literature. However, there are several controversial views to be pointed out. The discussion properly begins with the question: ‘why is interpersonal trust a collective asset for politics, especially in the democratic polity?’ I argue that this question can be answered by looking to the two generic problems that any democratic polity faces: the balance between efficient government and people’s check on it, and the 198 contradiction between conflict and integration. At the macro level, i.e., how to check the government while keeping it efficient, the general principle is based on the institutionalization of distrust. A variety of mechanisms in the democratic system, such as periodical election, fixed terms of office, checks and balances, and rule of law, are designed to induce government accountability because they assume that trust can potentially be breached. If these mechanisms can be effectively enforced when needed then a spontaneous culture of trust will likely emerge (Sztompka, 1999:140). However, violations and abuse of democratic principles may eventually destroy the culture of trust and evolve into a vicious cycle of distrust (Ibid: 146). The question that follows is “ to what extent does this kind of ideal-type argument exist in real politics?” This is where the micro perspective of democratic dilemma enters the discussion. If we differentiate relations occurring in the political realm from those in the broader social realm by characterizing it as “conflicts over goods in the face of pressure to associate for collective action, where at least one party to the conflict seeks collectively binding decisions and seeks to sanction decisions by means of power” (Warren, 1999: 311), then we see politics as an authoritative way of solving conflicts in a society, and according to human history this is more likely to be a zero-sum game instead of a win-win situation. The cost of trust in political settings is relatively larger than those incurred in normal social relations. Thus political relations will not normally induce trust among people. But it is just the scarcity of trust in politics that makes it an asset (capital) to political relationships. Put to the above macro discussion, we easily see the value of trust in avoiding the violation and abuse of democratic mechanisms. Specifically if people distrust each other in democratic polity, then either institutional designs based on distrust will be purposely used by those 199 in power or extensively utilized by people, which results in social disorder. In addition, from the perspective of integration, a democratic polity, characterized as a pluralistic society, emphasizes communication and compromise through negotiation. As Parry stated: “Mutual communication required in a polyarchy best occurs where men trust one another” (Parry1976: 129). If partners involved in political relations do not accept the fundamental rules of the game, then the confrontational nature of politics cannot be transformed into a higher level of social integration through communication and compromise. In fact, political theorists have long been aware of the social-psychological requisites of democracy. Among them, interpersonal trust, termed in a variety of ways such as ‘open ego’ (Lasswell, 1948), ‘faith in people’ (Rosenberg, 1956), or ‘general social trust’ (Almond & Verba, 1963), is related to people’s belief in the democratic process. They basically treat interpersonal trust as a necessary condition for supporting democratic values and hence democratic stability. But they are not clear about whether the democratic system itself can beget this kind of trust. Therefore, in answering the question “can democratic polity itself beget interpersonal trust?” or more broadly, “how is public social capital maintained?” scholars have conflicting views. Putnam, for example, argues that civic traditions and therefore social capital have remarkable staying power and are nearly impossible to change due to the virtuous circle that social capital enforces. In addition, stocks of social capital tend to be self-reinforcing and cumulative (Putnam, 1993b: 171, 177). The major role-player, according to Putnam, is voluntary associations. Some scholars, especially political economists and economic historians, see this question from the institutional perspective and emphasize the positive 200 roles lama come little beca C011 $011 3111 SU 315 Sel roles of governments in facilitating interpersonal trust among strangers (Yamagishi & Yamagishi, 1994; Levi, 1996)3. Slightly different from Putnam’s theory, F ukuyarna contends that it is the market economy that produces public social capital. The long-term interests of selfish individuals will force them to induce and maintain public social capital because it will eventually become their ‘economic assets’ (Fukuyama, 1999).4 Conversely, societies with centralizing monarchies in the past, such as China, France, and southern Italy, undercut the autonomy of intermediate social institutions in their quest for exclusive power and thereby produce a low degree of generalized trust (Fukuyama, 1995: 361). From a sociologist’s perspective, Sztompka (1999:122-125) highlights the structural factors that democratic systems can better carry than other types of political systems in facilitating social trust, i.e., normative coherence, stability of social order, transparency of social organization, familiarity of social environment, and accountability of persons and institutions. By examining the democratic system’s mechanisms of preventing abuse of power, i.e., the institutionalization of distrust, Sztompka claimed that “all other things being equal, the culture of trust is more likely to appear in a democracy than in any other type of political system” (Ibid: 139).5 Some studies using empirical data also support the positive relationship between social trust and democracy (Muller and Seligson, 1994; Uslaner, 1999a; Paxton, 1999). There are, however, different views on the positive side of democracy in inducing social trust. Granovetter, for example, 3 Notice that they tend to agree that democratic polity can nurture interpersonal trust but it is the government instead of voluntary associations, argued by Putnam, that fulfills this task. 4 In his earlier book, Trust: The Social Virtues and The Creation of Prosperity, Fukuyama provided his theory of why modern liberal democracy and capitalist economy are interdependent. His main argument is that modern liberal democracy has transformed the human desire for dignity or respect into economic sphere (Fukuyama, 1995: 358-359). 5 This is also what he called ‘the first paradox of democracy’: the more there is institutionalized distrust, the more there will be spontaneous trust (Sztompka: 140). 201 criticizes the naive idea of new institutionalism being undersocialized.6 Clever institutional arrangements may render malfeasance too costly to engage in, however, “they do not produce trust but instead are a functional substitute for it” (1985:489). He further stated that “[S]ubstituting these arrangements for trust results actually in a Hobbesian situation, in which any rational individual would be motivated to develop clever ways to evade them” (Ibid) Warren, using the American political system as an example, showed the fragility of the trust between representative and constituency (Warren, 19992315-316). He argues mainly that political relations are distinct from other social relations in two parts: we cannot take shared interests for granted in political relations, and parties are less likely to be constrained in their actions by shared culture. Thus he concluded: “politics does not provide a natural environment for trust” (Warren, 1999219). In sum, there is little doubt that social trust is an asset to democracy. But can democracy itself nurture public social capital? 1 would add here, based on the analytical framework proposed in chapter three, that in order to nurture public social capital in democracy we must factor in the following three contingencies: first, democratic institutions should provide a fair and transparent envirornnent in which citizens may interact. Second, the social structure of a particular society matters. If the norms, network of groups, associations, institutions, and regimes are persistent then people will have more dependable reference points for social life and hence feel safer, a necessary for social trust to develop. Third, the physical well being of the people also determines the level of trust in others (Luhmann, 1979; Giddens, 1991; 6 Granovetter’s purpose of this article is to criticize the ignorance of social relation in new institutionalism’s analysis and emphasized the need to be wary of the concrete patterns of social relations in predicting social 202 Patterson, 1999; Sztompka, 1999; Kramer, 1999; Mansbridge, 1999). As I mentioned in chapter three, these resource-oriented arguments highlight the relatively low cost of trusting others should the trustee breach the psychological contract between them. Voluntary Associations and Social Trust The second issue regarding public social capital concerns the possible interaction between voluntary associations and social trust. In their classic The Civic Culture, Almond and Verba argue that although primary associations play an important role in the development of citizens’ political competence, larger institutions, especially secondary, nonpolitical voluntary associations are also a necessary part of the democratic infrastructure (19632 300). According to the authors, the main function of voluntary associations is to bridge individuals and the political systems in which they are embedded. They demonstrate that through participation in voluntary associations, democratic citizenship will be fostered and hence people will become more politically competent, active as well as more committed to democratic values7. Although they did not directly address the relationship between membership in voluntary associations and interpersonal trust, they believe there is a link between the two. In fact, they expect “people who frequently engage in group activities and who place a high value on outgoing character qualities would also tend to view the human environment as safe and phenomenon, especially economic behaviors. Although Granovetter did not intend to extend his argument to the political sphere, I found his argument heuristic in my discussion here. 7 Actually, they argued that these political attitudes are obtained from experience of social interaction within the organization and whether an individual is actively participating the association is the key to the effective 203 responsive” (1963: 266). In addition, Almond and Verba show that in the United States and Britain people with more interpersonal trust tend to believe more in civic cooperation, i.e., he can work with his fellow citizens in attempting to influence the government (1963: 284-285). Thus, there is a positive correlation between membership in voluntary association and interpersonal trust. On the other hand, Putnam follows Tocqueville’s social theory and contends, “associationism per se produces habits of cooperation and trust, social networks and norms that,...ultimately issue in this social trust and civic engagement that healthy democracies need” (Foley and Edwards, 1998: 12). He contends that “people who join are people who trust...the causation flows mainly from joining to trusting” and “the greater the density of associational membership in a society, the more trusting its citizens. Trust and engagement are two facets of the same underlying factor—social capital” (Putnam, 1995a273). According to him, people acquire trust and the skills required to act collectively through participation in horizontally organized voluntary associations. He cites Almond and Verba’s classic The Civic Culture to show that members of associations display more political sophistication, social trust, political participation, and “subjective civic competence” (Ibid: 90). Cross-nationally, Putnam also notes a strong bivariate relationship between trust and group memberships (Putnam, 1995b:666; 19982254). Uslaner echoes Putnam’s thesis and contends from his analysis of the 1987 General Social Survey that “[P]eople who play sports or who attend sporting events are more trusting. They are also, both directly and indirectly (through increased trust), more likely to join voluntary associations” (Uslaner, 1999: 145). Stolle and Rochon analyze survey role of voluntary associations. Further, they showed even passive members are more likely to be politically 204 data from the United Stated, Germany and Sweden and their results confirm Putnam’s assertion that associational membership creates generalized trust (Stolle and Rochon, 19992205). Patterson’s study on contemporary America using GSS (General Social Survey) data also reveals a positive relationship between the two (Patterson, 19992196). However, this more or less romantic vision has raised many questions as well as criticisms. For instance, the most serious seems to be Putnam’s failure to specify a convincing mechanism by which social trust is produced within a group and then scaled up to the general public. Using Cohen’s words, “Why does the willingness to act together for mutual benefit in a small group such as a choral society translated into willingness to act for the common good or to become politically engaged at all?” (Cohen, 1999: 220). Similarly, Sullivan and Transue point out that Putnam and other theorists focused on the psychological underpinnings of democracy held a questionable assumption, i.e., attitudes generated in one domain will generalize to people in general. Indeed, psychological studies of perception have shown how people can compartmentalize their attitudes (Sullivan and Transue, 19992 648). In addition, Eastis shows us “how complex the experience of membership is and how difficult it is to assess completely the effects of voluntary associations on American civil society” (Eastis, 1998: 76)8. Contrary to Putnam’s theory, she argues from her observation in two choral groups that “those who join are those who already trust in some way” because the organization encourages a type of self-selection by potential members (Ibid: 71). Knack and Keefer selected 29 market competent and active than nonmembers (19632305,318). 8 Eastis adopted an ethnographic approach to observe the membership in two choral groups. She found that “even in these ostensibly similar groups, the relationships and processes out of which networks, norms, and collective facilities are produced vary, being strongly influenced by organizational characteristics” (Eastis, 1998: 66). 205 economics from the World Value Survey and conducted a series of empirical analyses regarding the effects of public social capital on economic performance. An interesting finding related to the current issue is that horizontal associations have no significant effect on social trust when income and education are controlled for (Knack and Keefer, 19972 1281). In contrast, memberships in the more politically oriented 9 groups are positively associated with stronger social trust and cooperative attitudes (Ibid). In sum, Putnam’s thesis is based on two questionable assumptions: first, attitudes generated in one domain will generalize to people in general. Therefore the interpersonal trust produced within one group can be generalized to trusting strangers. Logic dictates to allow voluntary associations provide individuals an opportunity to interact continuously, which creates the so-called repeated game, one of the fundamental conditions requisite for cooperation and interpersonal trust. However, as we see from critics, a more convincing theory specifying the actual mechanism leading to the alleged causal relations is desired. Or are the in-group trust and out-group trust merely uncompromised? As Boix and Posner speculate, if we take social structure into account, “in deeply divided societies, [for example], networks of associational life tend to be segregated, and the social capital that these networks produce tends to be unsuitable for promoting co- operation across network boundaries. . ..The accumulation of social capital will therefore reinforce communal divisions and reduce the incentives for group leaders to compromise in their dealing with one another” (Boix Posner, 19982693). Second, horizontal voluntary 9 The authors used the average number of groups cited per respondent in each country as the indicator of Putnam’s notion of the density of horizontal networks in a society. They then tried to differentiate the so- called Olsonian groups, i.e., groups that more likely to act as distributional coalitions, from Putnam-esque groups. The reason behind it was that the authors thought Olson’s theory (1982) also focus on the horizontal 206 associations engender mostly benevolent interactions, which generates social trust and hence an effective government. In fact, Putnam’s thesis is heavily influenced by Tocqueville’s analysis of democracy in America in which he found voluntary associations were a necessary means for the successful resolution of common problems given the providers of public service were extremely weak (Tocqueville, 1960: 514). As Whittington pointed out: “[T]he neo-Tocquevillians have correctly emphasized Tocqueville’s belief that civil society serves a crucial political function, but they have underemphasized the problems that such an arrangement can create for the state and the democratic values it protects” (Whittington, 1998 23). Olson’s theory on the paradox of group behavior (1982) provides a counteraction to Putnam’s thesis. Stable societies, according to Olson, tend to accumulate rent-seeking collusive organizations and so-called “distributional coalitions”, or special interest groups generally reduce the efficiency of the government and make politics more divisive by lobbying for their preferential policies, which eventually reduce the society’s economic growth rate. In brief, Putnam emphasizes the positive side of associations in his social capital theory, while Olson stresses the negative side of groups in a politically stable society. Putnam admits the effort to reconcile the two different points of view worth trying. (Putnam, 1995az76). Thus we need a more balanced view on the roles of voluntary associations on social trust and social conflict. Besides, people who join the voluntary association may not necessarily get along with their fellow members as Putnam seemingly assumed. Every association has its own organization to run itself. Given the assumption of homophily, members with similar backgrounds tend to form factions within the organization. The practice of this pattern associations but reach a very different conclusion from Putnam’s. For details, see Knack and Keefer (I997: 207 may not lead to benevolent results such as solidarity. In other words, the potential collective action problem within specific groups is largely ignored in the analysis of Neo- Tocquevillian theorists. Consequences of Public Social Capital: Maintaining Democratic stability? In chapter three I propose an analytical framework to study public social capital. Public social capital in this framework is not always a boon. Depending on each society’s unique social structure, the same set of social capital may produce very different, even opposite, outcomes. In other words, social context should be considered when analyzing the consequences of social capital. However, the mainstream literature of public social capital, led by Putnam, seems to focus only on the positive side of public social capital (Edwards and Foley, 1998). They try to link public social capital and democratic theory through the concept of “civil society” and treat (or assume) voluntary associations as the generator of public social capital. Thus, the broader virtues- interpersonal trust, trust in government, political tolerance, and political participation- produced by participation in voluntary associations can all serve to maintain democratic stability. I do not intend to criticize this line of thinking here. Rather, I would like to highlight the controversial points in literature concerning the political consequences of public social capital. I am particularly attracted to the roles of trust in sustaining democracy. Specifically, can membership in voluntary associations and interpersonal trust as the form of public social 1271-1273; 1277-1283). 208 capital lead to political trust, which has long been thought to be associated with the health of democracy? It is commonly acknowledged that an adequate amount of political trust, although unspecified, is a necessary condition for democratic stability (Lipset, 1960; Almond and Verba, 1965). It indicates citizens’ affective support for the political system. As Miller said “A democratic political system cannot survive for long without the support of a majority of its citizens” (Miller, 1974: 951). Almond and Verba’s The Civic Culture also argues the virtues such as generalized social trust, political efficacy and trust in government should be treated as key ingredients in stable democracies (Foley and Edwards, 1998: 13). Under the influence of social capital literature, some even conceptualize political trust as “political capital” referring to the citizens’ trust in, and respect for, the institutions of the political system (Whiteley and Syed, 19972 125). Given this line of thinking, is it logical to speculate whether the two kinds of capital can take on the form of the other? The literature again shows different views on this question. Rosenberg’s 1950’s study shows a positive relationship between general social trust and politically relevant trust. In his American sample he finds that those who score high in ‘faith in people’ are less likely than the low scorers to be cynical about politics and politicians (Rosenberg, 1956). Although Putnam does not openly argue it, his research certainly supports the idea that a society with high levels of social capital will also have high levels of political capital (Whiteley and Seyd, 1997: 130). Patterson also contends that “the greater generalized trust of the upper SES (socioeconomic) groups leads directly to greater political trust, while the opposite happens in the lower SES 209 groups” (Patterson, 1999: 195)”). In addition, based on his empirical analysis, Patterson found a positive relationship between the number of organizations in which persons are members and the percent of persons who say they trust others, another evidence affirming Putnam’s thesis (196). Although the author suggests no convincing explanation on why interpersonal trust can be scaled up to political trust,ll the common sense inference is: “a lower willingness to trust others is likely to be associated with a lower willingness to trust public officials” (Hall, 19992454). Whiteley and Seyd’s study on British party members shows that “voluntary activity of all kinds helps to foster respect for the core institutions of the political system such as the House of Commons, the monarch and the armed forces” (Whiteley and Seyd, 19972141).12 However, Hall’s study on social capital in Britain, using data mainly from the World Value Survey, run contrary to the above claim. Both aggregate and individual data, according to Hall, point to the speculation that “associational life alone does not seem to maintain levels of political trust” (Hall, 19992454). On the other hand, “an active associational life does not inoculate a society against political distrust.” In other words, Hall reminds us that although there is association between social trust and political trust at both the aggregate and individual level, the two seem to respond to a range of factors beyond patterns of sociability (Ibid). '0 Patterson proposed a model studying trust and political participation in the American democratic system, in which the so-called “affective trust”, operationalized by the question “can people be trust”, combined with other socioeconomic variables has direct effect on people’s level of confidence in the federal executive and legislative branches of government (Patterson, 1999:171-187). ” The author only tapped this issue through citing Coleman’s argument about TV watching and its effects on lowering people’s political trust (Patterson: 192). TV watching, according to the author, also promotes generalized distrust by overemphasizing crime and violence (194). Thus TV watching seems to be the third variable leading to both interpersonal trust and political trust. In brief, I found the author lack of convincing argument about the relationship between the two kinds of trust. '2 The authors implicitly followed Putnam’s thread and inferred that the experience of cooperation within voluntary associations generates trust and therefore can be transmitted to the political realm. For details, see Whitely and Seyd, 19972 131. 210 Cohen, from the theoretical perspective, openly rejects the idea that political trust is the result of generalized or interpersonal trust. Instead, she argues that political trust is derived elsewhere (Cohen, 1999:222). For her, government or the state should be understood not just as a third-party enforcer but also in terms of institutionalized action- oriented norms. Only through the fulfillment of “professional ethics” can people develop the expectation that institutional actors will live up to and enforce the norms of the institutional setting in which they interact (Ibid). This view is similar to Offe’s so-called “top-down” strategy that can affect people’s trust in institutions, i.e., institutions themselves should develop an “impeccable recor ” to fulfill their constitutive norms (Offe, 1999). In fact, their arguments concur with the analytical framework proposed in chapter three. Public social capital’s effects on a society should be examined in the context of the particular social structure in which they are embedded. Whether a society can maintain stability of its social order, normative coherence, transparent social organizations while providing regularity of social environment and hold persons and institutions accountable together determine a people’s confidence in institutions.'3 In the second part of this chapter an empirical case study is conducted with special attention to how the case of Taiwan” can shed light on the above-discussed controversial issues. Several national surveys will be used as empirical evidence. PUBLIC SOCIAL CAPITAL AND ITS POLITICAL CONSEQUENCES: THE CASE OF TAIWAN '3 These five factors are borrowed from Sztompka’s book Trust: A Sociological Theory, in which he suggested a model explaining the cause of a trusting culture. I found them helpful to settle the controversial issue discussed here and also support my analytical framework. 211 Can democracy itself nurture public social capital? Does a vibrant civil society beget social trust? And can public social capital help maintain democratic stability? To what extent the case of Taiwan shed light on these answers is the main concern of this empirical study. My study of Taiwan’s public social capital is divided into two parts. First, two major hypotheses and one research strategy will be formulated based on the critical literature review in the first half of this chapter. Second, with the assistance of survey data, empirical tests will be conducted and results will discussed. Hypotheses (l) Hypothesis 1: although cross-national comparisons support the argument that democratic countries tend to hold more stock of public social capital, a newly democratized country will not experience an increase in social trust. Rather, we should expect at least a temporary decrease due to the unpredictable environment affiliated with transition to democracy. However, the democratic system will provide structural opportunity for the development of voluntary associations. (2) Hypothesis 2: participation in voluntary associations alone cannot beget social trust. Specifically, it is only one of many ways people may regularly meet and socialize. In addition, as I argue previously, the potential collective action problem within each specific group or association may cancel out any social trust generated from the '4 The reasons why I selected Taiwan as a case for the empirical study are stated in chapter one. 212 associational life. Rather, we have reasons to believe a vibrant civil society can bring potential negative effects (Olson, 1982). Therefore, all these considered in the context of Taiwan, we should expect very small or no effect of associational life on people’s level of social trust. No systematic inquiry into public social capital’s roles in maintaining democratic stability exists, let alone any theory specifying the relationship between them. Similarly, we have seen no related study in the context of Taiwan. Therefore, the author will explore public social capital’s effects on several attitudinal and behavioral factors thought to be the foundation of democratic stability. The results will be served as the base of future research. Method and Data The empirical study in this chapter consists of two parts: the first aims to track the aggregate level of public social capital before and after Taiwan’s democratization. It is designed to test Hypothesis 1. The ecological data on numbers of associations in Taiwan (Table 7-1) is from Statistical Yearbook of Interior 1998, The Republic of China. As to the levels (percentages) of interpersonal trust and membership in voluntary associations across different years (presented in Table 7-2 and Table 7-3), they are calculated primarily from three nation-wide surveys: “Taiwan Social Change Survey”, “The Social Image Survey in Taiwan”, and “The Legislative Election Survey.” Chapter four provides the detailed nature of these data sets, such as purposes of these projects, sampling 213 methods, and sample size. It also lists the available years of these survey data from which the two components of public social capital is calculated. The second part of this study is a multivariate analysis using individual level data, including testing Hypothesis 2 and exploring the political consequences of public social capital. The 1998 Taiwan Social Change Survey will be used as the sole data set for these purposes.'5 However, in order to make the test of Hypothesis 2 more convincing I also employ the 1996 Taiwan Social Change Survey, which instead of testing the relationship between membership in voluntary association and interpersonal trust provides us available data to test the effect of depth of involvement in voluntary association on interpersonal trust. Multivariate OLS regression will be the major analytical tool. Measurement 'Public social capital' as used in this dissertation refers to social trust and membership in voluntary association. The detailed operationalization of these two concepts is described in chapter four. Other major variables used in this analysis will be briefly described as follows. Controlled variables The following variables are from 1998’s Social Change Survey and will be used as controlled variables in most of this chapter’s multivariate analysis. 214 Gender: male respondents are coded as 1, and female respondents are coded as 0. Ethnic Background: respondents are assigned their ethnic background based on their father’s. There are four ethnic backgrounds in Taiwan: Taiwanese”, Hakka, Mainlander, and aborigine.l7 Age: the actual age of the respondents in 1998. I recode this variable into five age cohorts, with 10 years as the interval starting from 19 years old to 60 years old and above. Thus, there will be four dummy variables in the regression analysis measuring the effect of generation. Education: respondents’ level of education. The original variable includes basically the highest possible education the respondent might have received. Three rank- ordered categories have been created, from elementary and below to high school to college and above. Therefore, two dummy variables will be used in the regression analysis measuring the effect of education. Political attitudes and behaviors The following attitudes and behaviors are commonly considered necessary to the stability of democracy. My purpose is to examine whether public social capital has any effects on them. '5 Please see chapter four for details. '6 Taiwanese here refers specifically to those whose ancestors came from Fukien Province of China in Ch’ing Dynasty. They consist of majority of the population in Taiwan. For more details, please see also chapter five of this dissertation. '7 For general information about the composition of Taiwanese society, please refer to chapter five of this dissertation. 215 Political tolerance: long regarded as a desired virtue in liberal democracy, it implies not only the recognition of equal rights for all citizens but also the best mean toward the discovery of truth (Sullivan, Piereson, and Marcus, 1982: 24). However, there is debate about how to measure political tolerance. Starting in the 19505 Stouffer (1955) attempted to measure political tolerance in the US. by asking if people support the civil liberties of the then unpopular left wing groups. This measure was criticized as only measuring tolerance of citizens who feared and disliked left wing groups in American politics (Sullivan and Transue, 19992630). Instead, Sullivan et a1 (1982) suggested a two- step measurement procedure, in which the “least-liked political group” was first detected for the individual and then measured the extent to which the individual supports or opposes the political rights of that group. The 1998 Taiwan Social Change adopted Sullivan et al’s format and thus provided a measure for political tolerance. A factor analysis of the following three questions produced one significant factor (eigenvalue=2.22), which explains the 74 percent of variance, and will be used as the indicator of political tolerance: for those who you think would most harm our society, do you think they should be banned from making a public speech on TV, to teach in elementary and high school, and to take government positions? Belief in democratic values: there is a list of questions in the 1998 Social Change Survey that measure respondents’ attitudes towards democratic norms. These questions can better serve to measure different dimensions of democratic norms. By differentiating them, they will give us a more comprehensive measure of belief in democratic values instead of only asking the respondents “if you support the democratic system (or democracy)” 216 A factor analysis of the following four questions18 produced two significant factors”, which will be used as indicators of “supportive of rule of law” and “supportive of check and balance”: 1. As long as the government’s public policies are doing people good, they need not be approved by a majority of the people; 2. No matter how unreasonable the laws are we should still abide by them until they are be changed; 3. Government cannot do much for the people if it is checked too often by the legislative bodies; 4. No matter what a political party advocates we should accept it if it wins the election. Similarly, a factor analysis of the following five questions20 yielded two significant factors” and will be used as indicator of “supportive of right of freedom” and “supportive of equal rights”: 1. More knowledgeable and educated people should have more say on public policy making; 2. Brutal crime suspects should receive punishment immediately without waiting for long lasting trail from the court; 3. Government should have the right to decide which views can be circulated in our society; 4. Citizens should have the same weight in making public policies regardless of their level of education and knowledge; 5. Local peace will be adversely aflected if there are too many groups holding diflerent opinions in that place. '8 These four variables have been recoded by author such that the higher the score the more supportive of democratic norm. '9 Please see Appendix 7-1 for details. 2° These five variables are recoded by author such that the higher the score the more supportive of democratic norm. 2' Please see Appendix 7-2 for details. 217 Political participation: two patterns of political participation are differentiated in this analysis: vote participation and non-vote participation. The mean score22 of the voting behaviors in 1995’s Legislative Yuan’s election and presidential election is used as the indicator of respondent’s vote participation. As to non-electoral participation, respondent’s mean score of the following nine items23 is used as the indicator: If you felt the government makes an unreasonable and unfair decision in elections, would you try to take the following actions? 1. Give a full statement to the related department; Inform friends and relatives about the issue and discuss it with them; Write or call representatives; Sign jointly with others to make your voice heard; Write or call the newspaper; Show your dissatisfaction to your own political party; Show your objection in the next election; Participate in a related protest or demonstration; Other. WFNQMPP’P Political trust: political trust is normally understood as citizens’ faith in the governing institutions (Erikson, 19912 163). However, there debate over what the level of political trust really represents. Does it represent the political system as a whole or simply the performance of office-holders (Ibid: 165)? Thus the author conducted a factor analysis of the following five questions and produced two significant factors“, which will be used 22 l recoded the two original variables into two dichotomous ones, either vote (1) or not (0). 23 These nine items has been recoded by author into dichotomous ones, either yes (1) or no (0). 2’ Please refer to Appendix 7-3 for details. 218 as indicators of “trust in office-holders and politicians” and “ trust in government as a whole”: Do you agree with the following statement25 ? 1. Most decisions made by the government are correct; 2. People ’s representatives claim that they work for the interests of people. In fact, they work for their own interests; 3. Government ’s public policies are afiected by conglomerates. The interests of the majority are ignored; 4. Most of the time our government spends tax money in the right direction; 5. The majority of oflice-holders are more or less corrupt. Political Eflicacy: this refers to the notion that an active citizen can play a part in bringing about social and political change (Campbell et a1. 1954). It was originally developed to explain the variations of voter turnout in the US. (Erikson & Tedin, 19952 166). Starting in the 605, it has become an important theoretical factor in maintaining democratic stability (Almond & Verba, 1965). It is argued that the potential effectiveness of the public’s participation in the political process plays a major role in holding public officials accountable (Ibid: 346-347). The measure of this concept is still facing severe validity and reliability problems. Scholars have not reached a consensus on what questions should be used (Erikson & Tedin, 1995: 166). Based on the meaning of political efficacy, two questions from the 1998 Taiwan Social Change Survey will be used as the measure of political efficacy. A factor analysis of the following two 2’ Some of the questions were asked in positive way while others were asked in negative way. In order to simplify the original five-point scale and make all five questions conform to the same direction, the author recoded them into dichotomous ones such that trusting attitude is coded as l and non-trusting attitude as 0. 219 questions26 produced one significant factor (eigenvalue=1.1), which explains 52.8% of the variance: Do you agree with the following statement? 1. It can make a diflerence if People like you express your views on politics; 2. Politics is too complicated and too hard to understand. Results and Discussion Does the democratic system beget public social capital?: A macro level analysis Taiwan began her movement toward democratization in the late 808 and reached democratic consolidation in the mid 90s. This unique experience provides us with an opportunity to observe the relationship between regime type and the stock of public social capital. Past empirical research shows a positive relationship between social trust and democracy.27 But they may ignore the contextual factors at work in this relationship. Does democracy preclude public social capital or vice versa? The aggregate data gathered here may give us some insight on this issue. As we can see in Table 7-1, voluntary associations rose dramatically after the mid 80s when Taiwan was undergoing democratization. In fact, after the passage of “The Civil Organizations Law for the Period of Mobilization of the Suppression of Communist Rebellion” in January 29, 1989, citizens could literally form any civil organization. There 2" The author recoded the original variables in such a way that the higher the score the more sense of efficacy it is. 27 These researches are based on aggregate data and cross-national comparison. For examples, see Uslaner, 1999a and Paxton, 1999. 220 are basically three broad categories of civil organizations: social associations, trade associations, and political organizations. Among these, social associations increased almost threefold within 15 years. The average numbers of associations per 10,000 people also increased almost twofold, from 5.2 to 9.9. As my analytical framework suggests, formal institutions may provide different opportunity structures for public social capital to thrive. Here we see the liberalization of the KMT regime in Taiwan has made voluntary associations more visible and autonomous, which increases citizens' opportunities to become involved in associational life. Table 7-1 Number of Associations in Taiwan (1983-1998)28 1983 1986 1989 1992 1993 1996 1997 1998 Social associations 4933 5365 6719 8190 9089 11788 12825 13783 Trade associations 4868 5260 6427 6828 6914 7193 7533 7774 Political organizations . . 45 96 100 1 1 l 115 1 19 Numbers/10,000 5.2 5.4 6.5 7.2 7.7 8.9 9.4 9.9 Source: Statistical Yearbook of Interior 1998, The Republic of China. However, the above data did not provide us with information regarding the variation of levels of membership in voluntary associations before and after Taiwan’s democratization. The statistics of Table 7-2 make up for this insufficiency. Table 7-2 lists the percentage of respondents who say they participate in at least one voluntary 221 association in various nation-wide surveys from 1983 to 1998. From 1983 to 1992, one fourth to one fifth of the respondents reported participating in at least one association. After 1993 the percentage rose to one third of the samples and reached its highest, 39%, in 1997. If we compare the level of associationalism in Taiwan after democratization, the average of membership in voluntary associations in Taiwan is relatively low compared to the long-established Western democracies, such as United States and Great Britain. It is similar to the levels of France and Italy and higher than Spain’s, which also underwent democratization in the late 705.29 Thus, we see the transition to democracy in Taiwan creates opportunity structures for voluntary associations to thrive. However, the culture and history of a particular society also plays important roles in determining the level of associationalism once opportunity presents itself. As mentioned in chapter five, Taiwanese showed that vigorous social movements have reinforced the increase of associationalism because of the need for organizations. With regard to the other dimension of public social capital, i.e., interpersonal trust, available survey data indicate a decreasing trend from the period of authoritarian regime to democratic regime, which supports Hypothesis 1. Table 7-3 shows that social trust declined from 1984 to 1991. There were 6% more respondents supporting the idea that no reliable and trustworthy relationships were to be found among people. In other words, 28 The selection of these years in this table is based on the years in which survey data are available. Thus a meaningful comparison can be made. 29 According to 1990 World Value Survey, which has similar pattern of wording with 1993, 1996 and 1997’s Taiwan Social Change Survey regarding the measure of membership in voluntary association, 222 Table 7-2 Trends of Membership in Voluntary Association in Taiwan (1983-1998) 1983 1986 1989 1992 1993 1996 1997 1998 Yes 26% 25% 1 1% 21% 32% 34% 39% 23% No 74% 75% 89% 79% 68% 66% 61% 77% N 1,618 1,429 1,288 1,395 1,961 2,831 2,596 1,798 Sources: 1983 through 1992 are from Taiwan University’s Legislative Election Survey; 1993 through 1998 are from Taiwan Social Change Survey. citizens in Taiwan were less trusting in the period of democratization than they were under the authoritarian regime. When using differently worded questions to measure social trust in Taiwan from 1991 to 1996, a declining trend still persisted. The level of social trust decreased from 40% in 1991 to only 23% in 1996. The 1998 Taiwan Social Change Survey adopted wordings for measuring social trust which were similar to that used in the World Value Survey. Slightly more than one third of the respondents (35%) said most people can be trusted, which is similar to the ones measured in Italy and South Korea from the 1990 World Value Survey but far lower than Great Britain (43.7%), the United States (51.1%) and Denmark (87.7%). Figure 7.1 shows the trends of both membership in voluntary associations and interpersonal trust in Taiwan from the 80s to the end of the 903. In fact, we should be careful about concluding that Taiwan’s transition to democracy has decreased the level of social trust among citizens. On the one hand, the pluralism advocated by the democratic system should have puzzled many Taiwanese who United States and Great Britain have 69% and 50% of respondents respectively reporting being member of at least one voluntary association. France and Italy have 36% and 33% respectively. 223 were accustomed to the strong leadership of the central government. The transition to democracy Table 7-3 Trends of Social Trust in Taiwan (1984-1998) 1985* 1990 1991a 1991b" 1993 1996 1998*" Trusting 76% 73% 70% 40% 39% 23% 35% Not trusting 23% 27% 30% 60% 61% 77% 65% N 3,967 2,289 1,044 1,512 1,544 2,730 1,798 Sources: 1991b and 1993 are from The Social Image Survey in Taiwan; others are fiom Taiwan Social Change Survey. * 1985’s, 1990’s and 1991a’s survey measure social trust with the question “do you agree that no reliable and trustworthy relationships are to found among people?” " l99lb’s, 1993’s and 1996’s survey measures social trust with the question “do you agree that it has become a tendency that people can trust each other less?” *" 1998’s survey measures social trust with the question “do you agree that most people can be trusted?” made social order less stable than before. Specifically, although the emphasis on human rights and the sudden increase in individual and press freedom are regarded as universal values, they also open the door for more crime and challenge many traditional values. According to statistics provided by DGBASEY, for example, Taiwan’s crime skyrocketed by almost three times from 1985’s 60,707 to 1986’s 179,990, when political liberalization began. From then on the number of criminal cases increased steadily and reached 1996’s unprecedented high, 456,117 (DGBASEY, 19992116). Regarding the interplay of new democratic values and old orientation to authoritarianism, the empirical study shows that after eight years of democratic rule (as of 1996) Taiwanese people still did not form a majority to overcome authoritarian political tendencies, i.e., “desiring freedom from political Oppression while simultaneously wanting to be ruled by a strong leader” (Doh Chull Shin and Huoyan Shyu, 1997: 117). In other words, as my analytical framework 224 suggests, the value shock makes the social environment less familiar and more uncertain. Accordingly, people are less likely to trust others (Luhmann, 1979; Giddens, 1991; Kramer, 1999; Sztompka, 1999). On the other hand, if the above argument is correct, we still do not know when this transition of social environment and value change will mature and move back to a more stable situation. Even if this expectation is optimistic, i.e., democratic institutions will ultimately increase the level of social trust in a society, we will probably not see the level of social trust as high as those in the Western democracies.30 The last point can be made from the results of Table 7-3, a caution regarding the validation of cross-national comparison. As I pointed out in the literature review, there is a consensus regarding the positive relationship between regime types and level of social trust. However, although Taiwan’s case study is still inconclusive, from the single case study we at least know that transition to democracy also plays an important role in determining possible levels of social capital. In other words, contextual factors should be considered when making inference from results of cross-national comparisons. 3° 1 make this argument based on the cultural factors that discussed in chapter three of this dissertation. In fact, similar case can be found in Torcal and Montero’s study of the impact of Spain’s new democratic institutions on its social capital in which they found Spain’s level of social trust kept constantly low regardless of the change of regime and attributed this stable social trust as the consequence of“ long- standing process of cultural accumulation” (Torcal and Montero, 1999: I77). 225 (D 0 5’ c 411 § 0 O. _31 0.0 Years . 96 ofmembershlp I %ofsocialtrust 1992158315941935193619871939193919901931 1992193319941935199619'97159819'99 20'00 Sources: Please see Table 7-2 and Table 7-3 Figure 7-1 Trends of Public Social Capital In Taiwan (1983-1998) 226 In sum, the limited aggregate data from the case of Taiwan show that democratic institutions do provide opportunity for voluntary associations to thrive, and membership levels in voluntary associations in Taiwan also reveals a rising trend.3 1 On the other hand, the level of social trust shows a declining trend. Thus, the aggregate data of Taiwan do not fully support the mainstream belief that democratic systems positively impact the stock of social capital in a society. Furthermore, only the increase of membership in voluntary associations cannot induce more social trust for society as a whole. The next section deals with the relationship between the two components of social capital and their impacts on several attitudinal and behavioral variables at an individual leveL Does membership in voluntary associations induce social trust?: A micro-level analysis The data of individual level analysis on public social capital is from 1998’s Taiwan Social Change Survey.32 Table 7-4 presents the frequencies and percentages of major independent variables in this empirical research. Table 7-4 Frequencies and Percentages of Major Independent Variables n % Total Gender 1 ,798 3 ' Compared to the case of Spain, Taiwanese people do participate in more as the number of associations increase. Spaniards remain low in the rate of participation given the number of voluntary associations increase as well. For details, see Torcal and Montero (1999). 32 The 1996’s Taiwan Social Change Survey will also be used to test Hypothesis 2 because it provides us with the measure of depth of a respondent’s associational life. 227 Table 7-4 (continued) n % Total Male 884 49.2 Female 914 50.8 Age 1,797 19 to 29 350 19.5 30 to 39 539 30 40 to 49 446 24.8 50 to 59 208 11.6 60 and above 254 14.1 Ethnic Background 1,795 Taiwanese 1 ,275 71 Hakka 272 15.2 Mainlander 237 13.2 Aborigine 11 .6 Education 1 ,795 No/Elementary 484 27 High school 811 45.2 Junior college 266 14.8 College and above 234 13 Membership in voluntary 1798 associations 228 Table 7-4 (continued) n % Total No 1393 77.5 Yes 405 22.5 Interpersonal trust* (Social trust) Most people can be trusted 612 35% 1731 Most people cannot be trusted 1,1 19 65% Most people are willing to help 915 53% 1726 Most people are not willing to help 811 47% “ Measured as a standardized factor score from two survey questions: “do you agree that most people can be trusted?” and “do you agree that in our society most people are willing to help and only few people are selfish?” The frequencies showed in this table are for each variable. For simplicity of presentation, each variable has been reduced to a dichotomous one instead of the original four-point scale. As we can see in Table 7-4, this sample consists of almost equal percentages of males and females. A majority of the respondents (54.8%) are between 30 and 49 years of age. Respondents of Taiwanese background dominate the sample as we expect. Hakka and mainlanders comprised similar percentages of the sample (15.2% and 13.2% respectively). A majority of the respondents (73%) have at least a high school education. Less than a quarter of the respondents said they participated in at least one voluntary association. Regarding interpersonal trust, a majority of the respondents (65%) did not agree that most people can be trusted. As to whether they found most people willing to help, respondents are polarized into two numerically similar groups. Slightly more than 229 half of the respondents (53%) believe that most people are willing to help, while 47% of the respondents disagree. Table 7-5 presents the distribution of membership in voluntary associations. It tells us that men participate significantly more than women in voluntary associations. Older people (50 years old and above) also participate significantly more than younger ones do. Ethnic background and level of education do not show significantly different patterns in their sub-categories. About one fifth to one fourth of the sub-categories are members of at least one voluntary association. A majority of the respondents are non- members. Regarding the relationship between membership in voluntary associations and interpersonal trust, the crosstab shows no sign of a correlation (p>.05) between the two. In other words, participation in voluntary associations cannot explain the variation of respondents’ social trust. Table 7-5 The Distribution of Membership in Voluntary Association Member Non-member x 2 Gender 10.64, p<.001 Male 25.8% 74.2% Female 19.4% 80.6% Age 28.76, p<.000 19 to 29 13.7% 86.3% 30 to 39 21.0% 79.0% 40 to 49 24.7% 75.3% 50 to 59 30.8% 69.2% 230 Table 7-5 (continued) Member Non-member x 60 and above 27.2% 72.8% Ethnic Background 4.3, n.s. ” Taiwanese 22.8% 77.2% Hakka 25.0% 75.0% Mainlander 1 7.7% 82.3% Aborigine 27.3% 72.7% Education 5.4, n.s. No/Elementary 24.0% 76.0% High school 20.3% 79.7% Junior college 22.9% 77.1% College and above 26.9% 73.1% Interpersonal trust People can be trusted 38.5% c 34.4% 2.22, n.s. People can not be trusted 61.5% 65.6% People are willing to help 57.9% 51.6% 4.83, n.s. People are not willing to help 42.1% 48.7% a. Entry in each cell represents row percentage. b. Not statistically significant. c. Entries under interpersonal trust represent column percentages, implying membership in voluntary associations is the indicator of interpersonal trust. 231 In order to estimate the net effects (controlling for other variables) of membership in voluntary association on interpersonal trust, an OLS regression was conducted. The results are presented in Table 7-6. There are two models in Table 7-6. The second model adds one more new independent variable measuring respondents’ feelings on ethnic tension in Taiwan. The rationale of adding it to the model is based on the discussion in chapter five, which argued that ethnic conflict is becoming a serious problem for Taiwan. Ethnic encounters should affect people’s opinions as to whether others can be trusted. From model (1) of Table 7-6 we can see men are more trusting than women. Generation does not have an effect on social trust. Respondents with a junior college degree tend to be consistently more trusting than those without or only graduating from elementary school. Those having college degree or above are also more trusting than elementary school graduates or those without an education.33 Compared to respondents with a Taiwanese background, Hakka respondents are more trusting (p<.1). Those who scored high in attitude toward ethnic tension, i.e., thinking his or her ethnic group does not receive deserved respect, are less trusting. Membership in voluntary association has no significant effect on social trust. Table 7-6 Effects of Membership in Voluntary Association on Interpersonal Trust: OLS Regression (1998 Taiwan Social Change Survey) Independent Variable Interpersonal trust " (1) (2) Gender Male .12* .13** ’3 If the significant level is set at .10 instead of .05. 232 Table 7-6 (continued) Independent Variable Interpersonal trust " (1) (2) (.05) " (.05) (reference: female) Age Cohort 19to 29 -.16 -.14 (.10) (.11) 30 to 39 -.11 -.09 (.09) (.10) 40 to 49 -.09 --05 (.09) (.10) 50 to 59 -.O4 -.O4 (.10) (.11) (reference: 60 and above) Education High school .06 .02 (.07) (.08) Junior college .28" .23“ (.09) (.10) College and above .17 + .06 (.09) (.1 0) (reference: No/Elementary) Ethnic background Hakka .13 + .13 + (.07) (.07) Mainlander -.Ol -.02 (.08) (.08) Aborigine -.06 .03 (.30) (.33) (reference: Taiwanese) Membership in voluntary associations Yes .10 .06 (.06) (.06) (reference: No) 233 Table 7-6 (continued) Independent Variable Interpersonal trust " (1) (2) Attitude toward ethnic tension 6 -.10** (.03) Constant -.09 -.05 (.08) (.09) R2 .02 .03 F ratio 2.39" 3.03" N 1,684 1,516 a. The dependent variable, interpersonal trust, is the scale of two survey questions related to social trust. Please see the Measurement section of chapter four for details. The author has recoded the two variables in such a way that higher score represents higher (positive) trust b. Standard errors in parentheses. c. It is measured by using factor scores of the following three survey questions: 1) are there any big difi'erences among the four ethnic groups in terms of their thoughts and ideas?; 2) do you think your ethnic group suflers unfair treatment in our society?; and 3) do you think the history and culture of your ethnic group receive due respect? Necessary recoding has been made in such a way that higher score representing more ethnic tension the respondent feels. For details of factor analysis, see Appendix 74. + p<.l, “ p<.05, "p<.01 In order to further explore the effect of participating in voluntary association on social trust, we should deconstruct this behavior according to its intensity. In other words, we should be able to differentiate those being mere members of some voluntary associations from those participating extensively in those associations. The 1996 Taiwan Social Change Survey provides us with information useful to fulfill this task.34 Table 7-7 is the results of regression analysis using this data set. Again, as it shows, participation in and membership in voluntary associations do not significantly affect respondents’ level of 3’ A variety of voluntary associations, including ‘others’, were listed in 1996 Taiwan Social Change Survey and respondents were asked in each kind of association that if they ever participate in their activities, make 234 social trust. However, those who participate intensively in voluntary association, indicated by volunteering, are more trusting than those who do not (p<.07). In sum, through bivariate and multivariate analysis in the case of Taiwan, we seem able to conclude that there is no relationship between associational life and interpersonal trust. The only point that can be used to defend the neo-Tocquevillian argument is that the effect of associational life won’t appear until people become deeply involved in it. Strictly speaking, the empirical evidence from Table 7-7 only tenuously supports this argument (.05