Editors' Conclusion to American Behavioral Scientist, 42, 2 (September 1998), "Beyond Tocqueville: Civil Society and Social Capital in Comparative Perspective," Special issue edited by Bob Edwards and Michael W. Foley.

Please send comments to

SOCIAL CAPITAL AND CIVIL SOCIETY BEYOND PUTNAM

Bob Edwards and Michael W. Foley

Abstract

Both "civil society" and "social capital" have proven useful heuristics for drawing attention to neglected nonmarket aspects of social reality and constitute a needed corrective for predominant economic models, especially rational choice theory. Social capital is the most recent in a string of efforts --including human capital and cultural capital-- to address failures of economic and "economistic" models, while recognizing the importance of "capitalized" resources in human interactions. However, both concepts break down, though in different ways, when treated as the basis for elaborating testable hypotheses and further theory. Civil society is most often posed in polemical or normative contexts, while uses in sectorial models often become embroiled in unresolvable boundary disputes and definitional refinement over just what constitutes "civil society" and what differentiates it from "state" and "market." Efforts to adapt "social capital" to prevailing theoretical and empirical models in political science by adding moral and ethical value have undermined its potential as an empirical analytic concept. We argue that for purposes of empirical research social capital should be divested of any social-psychological "value added" and treated as a more limited social-relational concept characteristic of social networks and organization. In so doing we criticize recent work on the US by Robert Putnam for turning the useful social relational concept of James Coleman and Pierre Bourdieu into just another label for the norms and values associated with the empirical democratic theory of the 1950s.

SOCIAL CAPITAL AND CIVIL SOCIETY BEYOND PUTNAM

by Bob Edwards and Michael W. Foley

The wide resonance achieved by the notions of civil society and social capital among both social scientists and political practitioners suggests that these terms serve crucial purposes of analysis and prescription in contemporary social thought. Nevertheless, widespread adoption does not ensure that a concept enjoys conceptual clarity or coherence or provides analytical purchase. In our judgement, while the notions of civil society and social capital play an important heuristic role in calling attention to aspects of social reality and distinctions often overlooked in contemporary theory, each concept breaks down in parallel, but distinct, ways when treated as the basis for empirical analysis. In the next few pages, we sketch both the heuristic value of these concepts and their failure to provide coherent support for developing testable hypotheses and further theory.

THE MANY CIVIL SOCIETIES

The notion of civil society owes its birth, both in earlier incarnations at the dawn of the liberal state and more recently, to its usefulness as a polemical concept. Whether as the "ethically obtainable ideal" of the Scottish moralists (Seligman 1992, p. 26), as "society against the state" in the Polish and Latin American conceptions, or as a sphere of social autonomy and democratization from below among the German Greens and the French "Second Left," the concept served to juxtapose a sphere of voluntary, purposive association to the forces of chaos, oppression or atomization of the time. Competing concepts of civil society thus almost invariably bear the marks of the political struggles within which they were born. Considerable overlap in the sorts of social actors identified as central to "civil society" among these conceptions gives the notion an air of universality -- suggesting that if only we could come to agreement about just who and what is included under its umbrella we could achieve a comprehensive theory of state-society relations. Yet, the real purchase of the notion of civil society is polemical and normative and tied closely to the context for which each version was shaped.

In each of the cases mentioned above, "civil society" crystallized projects of social autonomy over against the dominant power(s) of that time and place. In the Poland of the 1970s, civil society was more a wish than a reality until the explosive growth of the Solidarity movement in 1980-81. Prior to Solidarity's emergence the context of political opportunities in Eastern Europe under the Brezhnev Doctrine constrained the possibilities for action and helped shape the emerging conception of civil society as a third way between reform of the Communist regime from above and open revolt from below. After 1981 events in Poland outstripped theory, and what started as simply an assertion of social autonomy in the face of Communist power quickly became a highly charged and dangerous political movement (Pelczynski 1988).

Latin American conceptualizations reflected both the struggle against the military dictatorships of the 1970s and 1980s and a widespread conviction that conventional party politics had failed these societies. Latin American activists and thinkers thus framed civil society not only as "society against a repressive state" but as society in place of the parties (Garreton, 1989; Fals Borda, 1992). The leftist orientation of many anti-military activists led to an identification of civil society with the so-called the "popular sector" -- including a wide variety of lower-class and leftist groupings under this umbrella, but generally excluding even those sectors of the business and professional classes who eventually joined in the opposition.

In Western Europe during the 1970's and 1980s proponents of civil society developed their thinking against the backdrop of the neo-corporatist arrangements that had incorporated organized labor and its parties into institutionalized patterns of governance, but afforded little access to other constituencies. Civil society was conceptualized in opposition to the status quo of this neo-corporatist political settlement. By establishing "action spaces" within society in which to create new social and cultural organizations and institutions, spokespersons for "civil society" sought to embody alternative ways of accomplishing collective goods and a conceptualization of democracy at least implicitly critical of traditional representative forms (Melucci 1989; Dalton and Keuchler 1990; Edwards 1995).

In the United States, the term often appeared imported, with each school of thought tending to emphasize those aspects of the concept that best suited their particular ideological or theoretical purposes. The apparent crisis of the "welfare state" in the 1970's contributed to a search for a new paradigm in the United States, as elsewhere, but the initial result was a renewal of classical liberal remedies based on a relatively deregulated market and confidence in the ability of market economies to eventually "lift all boats." Academic and think tank economists, central banks and international lending institutions led by the World Bank and the International Monetary Fund, as well as the U.S. government from Reagan through Clinton have promoted economic liberalization and the unfettered play of market forces as the most efficient and humane means of providing public goods. The uneasy fit between civil society, with its implicit critique of purely market-driven approaches, and these prescriptions is evident in the World Bank's tardy attempt to incorporate civil society into its program of economic restructuring and state reform in the developing world. The generally conservative and depoliticized versions of civil society that have been proffered in the United States must be seen in this context.

As an analytical concept, civil society, and the sectoral models to which it is attached, suffer from acute definitional fuzziness. At least two related factors underlie this lack of clarity. Above all, the definitional confusion stems from variations across the "many civil societies" that have been the implicit empirical basis for the varying conceptualizations of the term and its deployment in specific polemical and normative contexts. Furthermore, the notion's precision suffers, as we have argued elsewhere (Foley and Edwards 1996), from the sheer historical and even transnational sweep of its application and the consequent temptation, despite the variation referred to above, to treat civil society and the "sectors" to which it is juxtaposed as ideal types (Weber 1949:90-105).

The problems of analytic schemes based on ideal types are apparent in the more home-grown versions of the civil society argument. John Van Til (1989) and Norman Uphoff (1993), for example, have each tried to unravel a peculiar organizational logic behind the sectorial breakdown into state, market, and "third" or "voluntary" sector by laying emphasis on the "voluntary" character of civil society organization. Uphoff's version is the most schematic and, accordingly, most representative of the problem of ideal typical constructions of this sort. For Uphoff, each sector operates out of a distinctive logic: for the state, that of hierarchical authority; for the market, that of profit and loss; for the voluntary sector, that of voluntarism. These logics determine the peculiar form of action and the particular problems of action in each sector. Typical of such efforts, Uphoff's scheme ignores the degree to which each of these organizational logics plays a role in all "sectors." Considered empirically, for example, business functions as much as a hierarchical organization as it does in response to market signals, as numerous analysts of the firm have pointed out. "Voluntarism" also may play a significant role in the evolution of the market. To take one example: Microsoft's adoption of its "open architecture" strategy, which allowed software developers advance access to new features and general permission to adopt broad elements of Microsoft's proprietary code without fees, pushed forward the software industry, and the fortunes of Microsoft and Intel, immeasurably faster than the sort of closed, proprietary strategy maintained by Apple. Netscape -- following the lead of a host of "freeware" and "shareware" developers -- took the approach one step further, allowing users free use of "beta" versions of its software in exchange for the (voluntary) testing and feedback that resulted. Similar examples of mixed organizational logics could be adduced for each of the other sectors.

But worse than the difficulties of identifying the essential characteristics that separate ideal types is the tendency in such theorizing to busy itself in definitional disputes about what fits which type and to what extent. Such "boundary maintenance" efforts generally come at the expense of empirical enquiry into how social phenomenon actually work. Ideal types are useful deductive propositions that enable investigators to make provisional sense out of complex social realities. However, once that initial analytic step is taken continued use of ideal types becomes counter-productive, and they must be intentionally left behind rather than subjected to further refinement. Nevertheless the heuristic leverage of that initial step tempts all too many analysts to push further, to load ideal types with more analytical freight than their heuristic capabilities can bear. Examples abound in the social sciences: ideal typical theorizing has plagued efforts to delineate organizational types (Michels, 1959; Rothchild-Witt, 1979), religious groups (Troeltsch, 1960; Stark and Bainbridge, 1979), forms of social organization (Tonnies, 1940), corporatism (Schmitter, 1979), or broad societal sectors of the sort discussed here. Such efforts move inexorably toward debates over the boundaries between types and often exhibit a sort of classificatory angst driven to specify the degree to which observed cases approximate a given type. Pushing an ideal typic analysis beyond its heuristic limits and pursuing the mirage of definitional refinement leads to ever more fine-grained description, but rarely to verifiable hypotheses and seldom even to noncircular explanations of the case at hand.

The case of "third sector" research in the US illustrates what is both heuristically valuable and empirically deficient about the civil society argument generally. The US debate generally recognizes the failings of strictly economic models ("market failure" as the basis for the emergence of nonprofits), yet accepts uncritically their implicit sectorial logic, only to become distracted by concerns with sectorial boundary specification and maintenance.⁽¹⁾

This difficulty is evident in the debate over the bounding of the non-state, non-market "sector" in the US context, which is alternatively labeled and distinctively defined as the "nonprofit sector" (Salamon and Anheier, 1992), the "independent sector" (Hodgkinson and Weitzman, 1989), the "voluntary sector" (Wuthnow 1991), or the "third sector" (Van Til, 1989). Here, as elsewhere in efforts to delineate civil society, numerous categories of groups are unduly neglected, from unions to public interest groups to churches. At the same time, boundaries often established to conform with the dictates of the nonprofit tax exemption manage to group under one heading organizations as diverse as the Metropolitan Museum of Art and Earth First!, the Mayo Clinic and the Animal Liberation Front.

All these sectorial schemes share a common effort to delineate a population of groups separate from market and government that arguably play a significant role in American social, political and economic life. While largely conceived in response to the perceived failures of the twentieth century state, the rediscovery of "civil society" or the "third sector," in Europe as in the United States, on the left as well as on the right, also springs from a rejection of the dominant economism of modern social thought. In this they are all critiques of the ways that predominant economic models, whether Marxist or neo-classical, unduly limit the roster of forms of social organization that "count" to those that fit clearly under either "market" or "state" rubrics. Such efforts are of considerable heuristic value in that they direct attention to neglected collective actors and forms of action, but when applied in a more fine-grained way they tend to lead toward the mirage of sectoral boundary maintenance over whether or to what extent specific types of groups are either "in" or "out."

In the last analysis, concepts such as civil society," "state" and "market" are too coarse-grained to help very much in sorting out the variety of social initiatives that make up a vital democracy. Political structure and political context can go a long way toward shaping both the kinds of organizations represented in society and their impact on the behavior and attitudes of citizens, as we have argued elsewhere (Foley & Edwards, 1996) and Booth and Richard (this volume) have shown, rendering generalizations on "the" role of civil society highly suspect. More generally, social initiatives depend crucially on a context set jointly by "state," "market" and "civil society"; and the logic of organization in such initiatives often intermingle legal coercion, authoritative decision-making and implementation, political manipulation, economic ends, and the voluntary pursuit of group and individual goals. But if this is the case, then it will be very difficult to specify in the abstract just what characteristics of "civil society" per se (or, for that matter, "the market" or "the state") contribute to healthy democracy and which do not, because the specific roster of beneficial characteristics would vary cross-nationally and over time along with the socio-economic and political context.

THE MANY FORMS OF CAPITAL

The heuristic value of the social capital concept in recent debate has lain in its calling attention to crucial aspects of social relations that impinge upon economic and political life and that are neither easily nor convincingly incorporated into an explanatory model based on the rational pursuit of individual self-interest. The growing popularity throughout contemporary social science of the rational actor model and its relative neglect of norms, values, social networks, organization and other context dependent resources explains social capital's heuristic leverage for a growing number of analysts. Thus, social capital can be seen as the most recent in a string of efforts --including human capital and cultural capital-- to amend or overcome the failure of the predominant economic model to incorporate nonmarket factors into its accounts of the political and economic behavior of individuals and groups. The concepts of cultural and social capital did not originate as an analogic extension of economic concepts, rather from the inability "to account for the structure and functioning of the social world unless one reintroduces capital in all its forms and not soley in the one form recognized by economic theory (Bourdieu 1986:256,242).

Social capital's heuristic value in challenging overly narrow economic models of development and democracy is exemplified in Robert Putnam's Making Democracy Work (1993). There, Putnam argued that greater reserves of social capital -- conceptualized primarily as dense associational networks that cross-cut traditional social cleavages -- exerts the dual effect of enhancing both governmental and economic performance. In taking up this line Putnam explicitly counters Mancur Olson's argument -- a generalization of public choice theory's view of non-market actors as little more than "rent-seekers" (Olson, 1983; Buchanan, Tollison and Tullock, 1980)-- that dense networks of strong organizations have the opposite effect, making economic performance increasingly sluggish and governance problematic as strong groups use their strength to curry favor from government at the expense of other groups or even the national interest. "Strong society," Putnam argues, "strong economy; strong economy, strong state" (p. 176). In a similar vein, the democracy enhancing side of these arguments is tied to an effort to counter, enhance or complete the economic model of democracy that sees democratic functioning as simply the aggregate outworking of individual interests rationally pursued. In this respect, at least, Putnam's work can be seen as a return to political culture variables.

James Coleman's articulation of the notion of social capital was explicitly conceived as the latest in a series of theoretical amendments --including human and cultural capital--to the predominant economic model (with which, however, it is not necessarily incompatible. See Bourdieu 1986; Becker 1993). The key empirical difference between human and social capital, in Coleman's rendering, is that social capital inheres in relations between individuals and groups, not in individuals per se. By contrast, human capital does inhere in individuals, so that when people move in and out of various social contexts their human capital, whether formal education or organizational skills, goes with them rather than remaining embedded in the context.

Greeley (1997) has argued that for Coleman social capital is a structural category and not something social psychological. At first blush, this seems to unduly narrow Coleman's conceptualization which did, after all, encompass norms of reciprocity and trust (Coleman, 1990, p. 304). Yet what Coleman had in mind was not the norms and values of individuals per se, but norms and values available as resources to those individuals who share access to that particular social context. For example, the norms of reciprocity and trust characteristic of a specific group of diamond merchants facilitates commerce among members. Individual traders take on those norms within that specific context. When business is concluded and those merchants walk out the door, those norms of reciprocity and trust do not necessarily go with them into other social contexts. No matter how much trust exits among traders within the confines of the diamond market, individual traders may not accept a used car dealer's description of a vehicle without first having it checked out by a mechanic or contacting its prior owner. Norms and values held by individuals become "social capital" only insofar as they facilitate action by others. And in this respect, they are context specific; outside that situation they may be of little or no value. Coleman also recognizes that norms and values may be internalized or mainly external, in the latter case representing how I must behave or what I must claim to believe to be accepted in a given setting. In either case, however, they represent a form of social capital for those who rely on their presence among others.

By conceptualizing such resources as forms of capital, both Bourdieu and Coleman direct analytic attention to the ways in which the products, characteristics and competencies of individuals are capitalized, that is converted into a form available for use by others.⁽²⁾

But from such an approach, it should be clear not only that each form of capital depends upon specific contexts for its realization, but that both capital and access to it is unevenly distributed in any society. No form of capital is distributed evenly in American society, nor do all individuals have equal opportunity to access the various forms of capital. Not all forms of social, cultural or human capital, finally, are equally valuable as resources to facilitate individual or collective actions.

We have taken up this argument with respect to social capital elsewhere (Edwards and Foley 1997), but a few points bear elaborating in this context. Access to various forms of capital is shaped substantially by inequalities of social location, whether these are patterned along lines of race, class, sex, geography or other salient factor (see Crowley, 1997; Oliver, 1997). Socio-economic and political context plays a prominent role in determining the value of a specific form of capital --financial, social, cultural or human-- and how that value varies, depending upon the goal toward which it is expended. For example, because of regional variations in supply and demand, a law degree is less valuable in the Washington, D.C. area than in eastern North Carolina. Yet network connections into the local legal communities would likely be of relatively high value to an attorney seeking to establish a business in either locale. Similarly, some human capital in the form of job skills may be of little or no value because they relate to an economically obsolete occupation. The social capital of having wide-ranging network connections into a dying industry would likewise be of little value.

To take another example: Particular forms of cultural competency such as a linguistic facility with Ebonics or Spanish are quite valuable within African American and Latino communities respectively, and might be "capitalized" in those contexts; but both are devalued by mainstream America. The devaluation of facility with Ebonics or Spanish, moreover, like the precise contours of "standard" English, shifts over time as a function of broad, sometimes contradictory, changes in socio-political and economic context. As corporate America awakens to the consumer purchasing potential of Hispanics, the cultural competency represented by fluency in Spanish and understanding Latino communities is being accorded more value in the market, even as there is a movement afoot in the country to cut such skills out of civic life by passing state level laws establishing English as the "official" language.

To assume that we can simply aggregate "social capital" to produce some measure of the resources available to "society" or "the polity," moreover, is to make the same mistake that economists make in using GNP per capita as a valid indicator of national economic "development". The same level of GNP per capita can mask vastly different distributions of income, wealth and opportunity, among subgroups within a society, each with its own potential for generating different levels of wealth, inequality or immiseration. In both cases, analysts inappropriately mix levels of analysis by aggregating measures (social capital, income) that have their primary meaning only at the level of the individual, networks or subgroups of a society and using them as indicators of the health of the whole. Like those students of political culture, from Almond and Verba to Inglehart, who treat scores on attitudinal variables aggregated at the national level as measures of "culture", the resulting configurations bear almost no relation to the concept being measured.

THE LIMITS OF THE CURRENT DEBATE

We would argue that Putnam's use of social capital, particularly in "Bowling Alone" and subsequent publications (1995a, 1995b) unduly narrowed the concept, focusing primarily on associational membership (and then only as a "source" of social capital) and "norms of reciprocity and trust," treating these as something akin to human capital (i.e., by vesting them in individuals, not social relations), and, finally, limiting attention to only the "good" stuff -- social capital as producer of "civic engagement". So defined social capital loses the distinctive benefits associated with Coleman's original formulation with its emphasis on socially embedded (and context specific) resources and its attention to social networks and organization. Important parts of the social capital argument we have sketched above fall by the wayside as one begins to operationalize it, as Putnam and others have done, with extant survey research. Indeed in doing so social capital is seamlessly transformed into a characteristic of individuals acquired through social networks, above all, the associations of civil society. This runs contrary to Bourdieu who argued that the effects of social capital are clear at the individual level it "cannot be reduced to a set of properties individually possessed (1986:256). But if social capital is as context specific as we have indicated, then it cannot be measured directly in extant longitudinal survey data like that of the General Social Survey. Nor is it legitimate, as we argued above, to treat the resulting aggregate measures as reliable indicators of the "social capital" available to the polity.

Social capital, in Coleman's conception, is a morally and ethically neutral resource that facilitates all manner of individual and collective endeavors. It can just as easily enhance the operations of a community improvement organization or human rights advocacy efforts as it can a drug gang or death squads. Putnam's adaptation has added moral and ethical value to the concept, and many neo-Tocquevillean researchers have followed his lead. This, we believe, seriously undermines its usefulness for empirical research. Efforts to further specify the concept in recognition of this conceptual shift unduly narrow the range of empirical phenomena the original concept pointed to and thereby undermine the heuristic value that motivated its adoption in the first place. In taking this "value added" approach, researchers limit what constitutes social capital to either those resources that lead to "good deeds" or, more narrowly, "good" resources. In either case infusions of beneficial social capital are expected to help restore democracy in America to a vision of "goodness" characterized by civility, cooperation and (apparently) minimal conflict.

Both Stolle and Rochon and Booth and Richard (both in this volume) take versions of the value added approach. By emphasizing that "public social capital" is the core of the social capital argument, Stolle and Rochon constrain what really counts as social capital to include components of "civic culture." Specifically those believed to undergird a distinct vision of the democratic process as civil, cooperative and tolerant.

When Stolle and Rochon operationalize their concept of "public social capital" as tolerance and cooperation they beg several fundamental questions which remain inadequately addressed within this "civic culture" perspective. Cooperate with which kinds of groups? Tolerant of whom? In what kinds of socio-political contexts? Even with a value as widely admired as "tolerance," we must ask what kind of tolerance is actually crucial for democratic polities? Is it the polite, civil sort that puts on a good face and stoically accepts the presence of those found to be personally repugnant? Or a less gentile, but more significant, sort of tolerance that acknowledges the right of other groups to participate as citizens -- and as adversaries? Tolerance of the latter sort is crucial for any democratic polity, while the former is the peculiar attribute of an "ideal" democracy without contention.⁽³⁾

For example, in the U.S. the religious (cultural) right is not particularly tolerant of the gay and lesbian community, and the reverse is quite often the case. Members of each group often mistrust each other, might not want to associate with each other and tend to ignore the other as much as possible in everyday life. They may very well not want to live next door to each other. However, unless one or the other group seeks to exclude the other from political representation and participation, as with Jim Crow in the South following Reconstruction, the democratic process is not imperiled.

CONTEXT DEPENDENCE AND THE NEO-TOCQUEVILLIAN ARGUMENT

Recent researchers, following Putnam's lead, have tended to treat social capital as if it were context independent. Efforts to apply the concept in diverse social and historical settings --ranging from Putnam's studies of Italy to the articles in this volume --repeatedly point up the difficulties in maintaining this assumption. Not all social capital is alike in its impact on either group or individual fortunes. Yet how and to what extent surrounding conditions -- political repression, economic distress, intense social conflict, and shifting patterns of cultural hegemony -- might force groups in upon themselves or project them outward in defensive and offensive action is rarely explored in the neo-Tocquevillean line of argument. One exception is the article by Booth and Richard in this volume. The authors compare divergent political contexts --more or less repressive-- and demonstrate that context matters: In post- repressive regimes participation is constrained and trust, both generalized and of government, is withheld.

In treating individual traits such as tolerance, trust, or membership in associations as consistent indicators of "social capital" regardless of context, researchers are adopting the flawed assumption that the elements of social capital are context independent. Consider for example individual memberships in voluntary associations. Stolle and Rochon (this volume) show that all voluntary associations are not alike with respect to both the amount and the sort of social capital "produced" and that some kinds of associations are more productive than others. Nevertheless, as they also correctly point out, voluntary association memberships per se have become the measure of choice for operationalizing social capital. Eastis' ethnographic analysis (this volume) shows very nicely, moreover, that mere membership in one or another category of voluntary association is too crude a measure to capture empirically the complex experience of membership. Members of both the choirs that Eastis analyzed could report very extensive participation, yet still come away from the experience with a rather different mix of human, cultural or social capital. Such variation owes much to characteristics of the groups and the structure of relations between their members, not to participation per se, nor to the type of group per se. Such differences cannot be measured utilizing the current stock of survey questions.

SOCIAL CAPITAL AND DEMOCRATIC THEORY

Following Putnam's lead many in the neo-Tocquevillean tradition have seized upon Coleman's discussion of norms and values in relation to social capital and operationalized the broader argument in terms of long standard measures of "social trust" and participation. The social capital argument, operationalized in this way, has increasingly become identified with social psychological, rather than the sort of relational and social structural variables emphasized by Coleman. This psychologized version has by now become a standard feature of social scientific inquiry into the quality of participation in American democracy and beyond. The social psychological turn to the argument has been criticized by Andrew Greeley in the previous volume (1997) for its failure to do justice to Coleman's fundamental insight. In Kenneth Newton's terms (1977), it lays emphasis on the "norms and values" aspect of the argument at the expense of the "networks," which tend to be seen simply as the bearers and sources of such individual level attributes (and even then, only as operationalized by "membership in associations").

Central to the ready acceptance by political scientists of "social capital" thus conceived has been its convenient operationalization in existing longitudinal survey research like the General Social Survey or the World Values Survey and the conformity of its general line of reasoning with the still widely accepted notions of the empirical democratic theory of the 1950's. Brehm and Rahn (1997), for example, argue that higher levels of "social capital" should be measured not only by higher levels of "generalized social trust" and participation in civic affairs, including voting, but by high levels of "trust in government." Their argument harks back to the democratic theory of the 1950s, best captured in Almond and Verba's five-nation study, The Civic Culture (1987 [1963]), where "trust in government" was taken to be a vital component of stable democracies and, therefore, of "civic culture."

That genuine democrats could number among the opponents of existing governments is evidently outside the conceptual ken of ademocratic theory born of the Cold War, in which "anti-system" parties and a willingness to employ "non-conventional" methods of protest were both seen as inherently "anti-democratic." These notions continue to be alive even among comparativists studying the still shaky regimes of Central America, witness the recent work of Mitchell Seligson (e.g., Muller and Seligson 1994). Yet, as Keith Whittington notes in his analysis of the social capital argument in the context of Jacksonian America (this volume), "Given the possibility of social conflict, distrust of government and of others can be a reasonable political choice, and not simply the product of a weak society" or, we would add, a defective democracy.

CONCLUSIONS

The emphasis on the "cultural endowment" of individuals so prominent in this line of research, and so much at variance with Coleman's broad conception of social capital as a characteristic of social relations and structure, reflects certain predilections of American political science and political thought generally. Though "political culture" has long evaded reliable conceptualization, it remains the favored explanatory gambit in American attempts to analyze the domestic political scene, explain the rest of the world and rationalize what is often seen as American exceptionalism (Lipsett 1996; Muller and Seligson 1994; for a critique, see Jackman 1996a, 1996b). Next to voting studies (which themselves rely heavily on survey research), public opinion research (often understood as giving empirical access to "political culture") is perhaps the best populated sub-field of American political science. A by now standard battery of survey questions, readily available through such reliable and relatively long-term endeavors as the General Social Survey and the National Election Survey, are taken to operationalize a scarcely questioned theory of democratic participation first developed in the 1950's. In this tradition, as among the neo-Tocquevilleans, individual attitudes and their demographic or cultural correlates are what is of interest, almost all that is of interest.

Even so sophisticated an undertaking as Verba, Schlozman, and Brady's Voice and Equality purports to provide a "comprehensive" account of civic and political participation in the United States by way of an analysis of individual responses to a survey (1995, p. 2). Their findings, conceptualized in terms of what they call the "civic voluntarism model" broadly sustain the neo-Tocquevillean thesis that "both the motivation and the capacity to take part in politics have their roots in the fundamental non-political institutions with which individuals are associated during the course of their lives" (p. 3). Whether, to what degree and, above all, why such relationships might (or might not) obtain in other societies is beyond the scope of their study. But just for that reason it is difficult indeed to weigh the importance of the factors identified as crucial in citizen participation beyond the borders of the United States, and perhaps even here. Only careful, cross-national research -- including research that is attentive to differences in political and economic context among nations --preferably longitudinal in character, could test the more general claims advanced by the neo-Tocquevilleans about the role of civil society in democracy.

By contrast, those who emphasize the autonomous, at times oppositional or conflictive, character of civil society press questions about the ways in which the political system might be made more responsive to organized societal demands, more accountable to societal scrutiny, more open to profound differences within society. The question is much less one of the development of proper attitudes on the part of citizens than of scope for the exercise of democratic rights and the fulfillment of democratic principles. Civil society may nurture a democratic culture, but the real question from this point of view (with which we largely sympathize) is how to reshape political institutions to let democracy flourish. Analysis is much more likely to focus on the ways in which constitutional and political settlements structure state-society relations; how, why and with what success groups mobilize to attain greater responsiveness and accountability on the part of the state; or the impact of global economic restructuring on the ability of both state and civil society to respond to citizen demands.

Putnam's work linked the concepts of civil society and social capital and prompted a ripple of empirical research, mostly with survey data, and a wave of theoretical debate. Our conclusion from this series is that the recent debate inspired by Robert Putnam's work on the U.S. in particular has added little new to the analytic repertoire available to contemporary researchers. Ironically, Putnam's lumping of social capital together with other forms of capital, on the one hand, and his narrowing its scope by adding moral and ethical "value," on the other, have blunted the fundamental heuristic value which was the basis of the concept's initial appeal. Furthermore, the opportunistic operationalization of social capital in extant survey data, treating it primarily as a social-psychological characteristic of individuals, has led the social capital argument back to something rather old --the richly appointed and populous analytic cul-de-sac of the empirical democratic theory of the 1950s.

By contrast Coleman's and Bourdieu's formulations of social capital do add something new by amending more narrowly construed economic models. To retain the greatest conceptual and empirical leverage from social capital, we would argue, two steps are required. First, social capital should be conceptualized more narrowly as a social relational and structural resource characteristic of social networks and organization, leaving aside the "norms of reciprocity and trust" which inspired Putnam but which might more properly be considered a sort of "cultural capital". Secondly, social capital should be considered in conjunction with other forms of capital, including financial capital, human capital (Becker 1993) and cultural capital (Bourdieu 1986).

By taking this approach we are suggesting that Coleman's formulation of social capital, best conceptualized in terms of relational or structural factors like networks or organization, brought a much needed amendment to prevailing economic analytic models, while preserving the insight that individuals and collectivities depend upon a variety of resources, "capitalized" in a variety of ways and contexts. However defined, any such concept is liable to misuse if appropriated without regard to nuance. We need to be much more careful in thinking about the kinds of effects associated with each form of capital. We must consider each separately in a way that leaves open the possibility that they may have different, even contradictory effects, indeed, that, depending upon context, any given form of capital might have quite different meaning for the individual and groups involved. Similarly, it is important to pay close attention to the circumstances in which forms of capital are created and deployed, avoiding the assumption that a single, global variable ("associational membership") can be taken to predict meaningful differences in social, cultural or human capital. At the same time, these reflections strongly urge that we rethink the question of the sort of cultural resources that genuinely facilitate economic and political vitality. The widespread adoption of Weber's dubious theory of the links between Protestantism and the rise of capitalism and the gross simplifications to which Weber's theory has been subject should alert us to the dangers inherent in such an endeavor. The same can be said for the continued reliance of political scientists on the highly questionable postulates of empirical democratic theory.

Several times we have criticized extant survey research for providing an inadequate empirical basis for robust and fruitful investigation of issues at the core of the recent social capital debate. We should make clear we are not arguing that such questions are beyond the reach of survey research per se. However, remedying this situation will require more than simply writing better questions for existing longitudinal surveys of individuals. Three related avenues for future research seem promising. First, the recent debate has shortchanged networks and social relations within networks in favor of social-psychological variables. Empirical research on all aspects of social capital would benefit from paying serious attention to networks and from explicitly network analysis along the lines of recent work by Heying (1997) and Diani (1997). Second, subsequent survey research should gather data on individuals and the specific organizations in which they participate. Doing so would enable researchers to take into account differences between the various organizations in which individuals participate and enable them to map the organizational field that comprises much of the context of citizen participation. In so doing variations in the appropriation and effect of social capital derived from face to face participation in proximate organizations as well as the more distant forms of participation (see Smith, this issue; Minkoff 1997) could be systematically examined as well. Third, this literature in general would benefit from more theoretically informed case analyses like those by Eastis and Warren (this volume) and Wood (1997).

We have argued that the initial value of social capital, like that of civil society, is heuristic and that recent efforts to add moral, ethical or cultural value to it have actually undermined its potential as an empirical analytical concept. We argue that for purposes of empirical research social capital should be divested of any social-psychological "value added" and treated as a more limited social-relational concept appropriate to social networks and organization. We also argue that for this more limited construction of social capital to have robust analytic value and theoretical coherence, it must be considered in concert with, but differentiated from, other forms of capital. In so doing we find ourselves much closer to James Coleman's and Pierre Bourdieu's original conceptions than to the neo-Tocquevillian version of Robert Putnam in which social capital becomes little more than a 1990's version of the attitudes, norms and values at the center of the empirical democratic theory of the 1950's.

In different, but parallel, ways the concepts of civil society and social capital each address perceived failings of predominant economic models for explaining the social and political behavior of individuals and groups within contemporary societies. In Europe, the rediscovery of civil society on the part of thinkers associated with the left answered to increasing discomfort with the narrow economism of much Marxist theory. In the United States, the currency of both concepts comes at a time when rational choice theory, built as it is upon the assumptions of a rather "bare-bones" micro-economic model, is making a serious play to become the dominant paradigm in American political science and gaining substantial strength within sociology and anthropology. The widespread and interdisciplinary enthusiasm for social capital and civil society indicates the range of constituencies for whom narrowly construed economic models, especially the rational choice manifestation, provide an inadequate basis for understanding and analyzing social and political life.

REFERENCES

Almond, G. & Verba, S. (1963). The civic culture: Political attitudes and democracy in five nations. Princeton, NJ: Princeton University Press.

Bourdieu, P. (1986). The forms of capital. Trans. R. Nice. Handbook of theory and research for the sociology of education. Ed. John G. Richardson. New York: Greenwood, 1986.

Bourdieu, P. (1990). The logic of practice. Trans. R. Nice. Stanford: Stanford University Press.

Brehm, J. & Rahn, W. (1997, July). Individual-level evidence for the causes and consequences of social capital. American Journal of Political Science, 41(3), 999-1023.

Buchanan, J.M., Tollison, R.D. and Tullock, G. (Eds.) (1980). Toward a theory of the rent-seeking society. College Station: Texas A&M University Press.

Cohen, J.L. & Arato, A. (1992). Civil society and political theory. Cambridge: The MIT Press.

Coleman, James S. (1988) Social capital in the creation of human capital. American Journal of Sociology. s95-1120.

Crowley, J.E. (1997). Separate but unequal: Class segregation in American civil society. Paper prepared for delivery at the 1997 Annual Meeting of the American Political Science Association. Washington, D.C., August 28-31.

Diani, M. (1997). Social Movements and Social Capital: A Network Perspective on Movement Outcomes. Mobilization 2(2), 129-48.

Dalton, R. and M. Kuechler, eds. (1990). Challenging the political order: New social movements in western democracies. New York: Oxford University Press.

Edwards, B. (1995). Organizational style in middle-class and poor people's social movement organizations: An empirical assessment of new social movements theory. Ann Arbor, MI: UMI.

Edwards, B. & Foley, M.W. (Eds.) (1997, March/April). Social capital, civil society, and contemporary democracy. American Behavioral Scientist, 40(5).

Edwards, B. & Foley, M.W. Social capital and the political economy of our discontent. American Behavioral Scientist, 40(5), 669-78.

Fals Borda, O. (1992). Social movements and political power in Latin America. In A. Escobar & S. E. Alvarez (Eds.), The making of social movements in Latin America (pp. 303-316). Boulder: Westview Press.

Foley, M.W. & Edwards, B. (1996, summer). The paradox of civil society. Journal of Democracy, 7(3), 38-52.

Garreton, M.A. (1989). Popular mobilization and the military regime in Chile: The complexities of the invisible transition. In S. Eckstein (Ed.), Power and popular protest: Latin American social movements (pp. 259-277). Berkeley & Los Angeles: University of California Press.

Greeley, A. (1997). Coleman Revisited: Religious Structures as a Source of Social Capital. American Behavioral Scientist, 40(5), 587-94.

Hardin, R. (1995). One for All: The Logic of Group Conflict. Princeton, NJ: Princeton University Press.

Heying, C. 1997. Civic Elites and Corporate Delocalization: An Alternative Explanation for Declining Civic Engagement. American Behavioral Scientist (40)5, 657-668.

Hodgkinson, V. and M. Weitzman. (1989). Dimensions of the Independent Sector: A Statistical Profile, 3rd edition. Washington, D.C.: The Independent Sector.

Jackman. R.W. and Miller, R.A. (1996a). Renaissance of Political Culture? American Journal of Political Science 40, 632-59.

Jackman. R.W. and Miller, R.A. (1996b). The Poverty of Political Culture. American Journal of Political Science 40, 697-716.

Lipsett, S.M. (1996). American exceptionalism. New York: W.W. Norton & Company.

Melucci, A. 1989. Nomads of the Present. Philadelphia: Temple University Press.

Michels, Robert. 1959. Political Parties. New York: Dover.

Minkoff, D. (1997). Producing Social Capital: National Social Movements and Civil Society. American Behavioral Scientist 40(5), 606-619.

Muller, E. & Seligson, M. (1994). Civic culture and democracy: The question of causal relationship. American Political Science Review, 88(3), 635-52.

Newton, K. In Edwards, B. & Foley, M.W. (Eds.) (1997, March/April). Social capital, civil society, and contemporary democracy. American Behavioral Scientist, 40(5).

Oliver, E. (1997). Civic involvement in suburbia: The effects of metropolitan economic segregation on participation in local civic affairs. Paper prepared for delivery at the 1997 Annual Meeting of the American Political Science Association. Washington, D.C., August 28-31.

Olson, M. (1982). The rise and decline of nations: Economic growth, stagflation, and social rigidities. New Haven, CT: Yale University Press.

Pelczynski, Z. A. (1988). Solidarity and 'the rebirth of civil society'. In J. Keane (Ed.), Civil society and the state (pp. 361-380). London: Verso.

Putnam, R.D. (1993). Making democracy work: Civic traditions in modern Italy. Princeton, NJ: Princeton University Press.

Putnam, R.D. (1995a). Bowling alone: America's declining social capital. Journal of Democracy, 6(1), 65-78.

Putnam, R.D. (1995b). Tuning in, tuning out: The strange disappearance of social capital in America. PS, 28(4), 664-683.

Rothchild-Witt, Joyce. 1979. "The Collectivist Organization: An Alternative to Traditional Bureaucratic Models." American Sociological Review 44:509-527.

Salamon, L.M. & Anheier, H.K. (1992a). In search of the nonprofit sector. I: The question of definitions. Voluntas 3(2).

Salamon, L.M. & Anheier, H.K. (1992b). In search of the nonprofit sector. II: The problem of classification. Voluntas 3(3).

Schmitter, P. (1979). Still the century of corporatism? Review of politics, 36:85-131.

Seligman, A. B. (1992). The idea of civil society. New York: The Free Press.

Smith. J. (1998). Global Civil Society? Transnational Social Movement Organizations and Social Capital. American Behavioral Scientist (this issue).

Stark, R. and W.S. Bainbridge (1979). Of Churches, Sects, and Cults: Preliminary Concepts for a Theory of Religious Movements. Journal for the Scientific Study of Religion 18:117-131.

Troeltsch, E. (1960). The Social Teachings of the Christian Churches. New York: Harper and Row. (Original work published in 1911.)

Tonnies, F. (1940). Fundamental Concepts of Sociology (Gemeinschaft and Gesellschaft), translated and supplemented by C.P. Loomis. New York: American Book Company.

Uphoff, N. (1993). Grassroots organizations and NGOs in rural development: Opportunities with diminishing states and expanding markets. World Development 21(4), 607-622.

Van Til, J. (1988). Mapping the third sector: Voluntarism in a changing social economy. New York: The Foundation Center.

Verba, S., Schlozman, K.L. & Brady, H.E. (1995). Voice and equality: Civic voluntarism in American politics. Cambridge: Harvard University Press.

Weber, M. (1949). The Methodology of the Social Sciences.

Translated by E.A. Shils and H.A. Finch. New York: The Free Press.

Wuthnow, R., ed. 1991. Between State and Markets: The Voluntary Sector in Comparative Perspective. Princeton: Princeton University Press.

Notes

1. ^1. Economics, of course, boasts its own sectorial model which, despite its many gaps and inadequacies, has gained a level of acceptance and legitimacy that is not likely to be replicated in sectorial approaches to state-society relations. One reason for its success is that it is backed by an arbitrary, but internationally accepted, scheme of classification, which permits extensive tracking of economic change and cross-national comparison. Nevertheless, both the pretended scope of the sectorial model and the arbitrary character of the system of industrial classification make for considerable distance between the model and the reality it is supposed to track. The "service sector," for example, embraces everything from street vendors to investment bankers; and measures of sectorial shifts at this level of aggregation, though widely accepted, are scarcely meaningful. Whatever the advantages and failings of its economic counterpart, however, sectorial models of the modern polity are unlikely to achieve such legitimacy, precisely because of the largely polemical and normative usages to which they have typically been put.

2. ^2. Bourdieu further distinguishes "cultural capital" from "social capital." The latter refers to "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" (1986:248). The capital of such memberships and the relations they entail are informal much like a reputation that provides individuals with the backing of the group's resources entitling "them to credit in the various senses of the word" (1986:249). Cultural capital includes the full range of a society's symbolic resources, from the norms and values that individuals bring to or encounter in interactions with others, to the religious, philosophical, artistic and scientific understandings that frame and interpret reality (Bourdieu 1990, p. 125). According to Bourdieu, it exits in three forms. In its "embodied" form cultural capital inheres in individuals as long-lasting dispositions of mind and body in which external wealth is converted --typically through family and class socialization-- into an integral part of the individual (1986: 244-45). In its "objectified" state cultural capital is "capitalized" in cultural artifacts like writings, art, monuments, and media products. The value of these cultural goods can be appropriated symbolically by anyone with access to them who possesses the means of "consuming" them (1986:246-47). Thirdly, "institutionalized" cultural capital takes the form of credentialing and the conferral of institutional legitimacy (1986:247-48).

3. ^3. An ideal, we might speculate, based on a secularized vision of the harmony and goodwill characteristic of Kingdom of God in Judeo-Christian thought. The discomfort with conflict and contention in so much of American culture may be closely tied to the perfectionist and millenarian strains in our religious tradition.