Theoretical Perspectives on Social Capital, by Thomas Ford Brown

INTRODUCTION


One of the most influential theoretical constructs emerging from the New Economic Sociology has been the notion of "social capital". But what precisely is social capital? The term quickly pervaded the lexicon in social science and policy circles, but clear definitions have been scarce. Few of those employing the concept have elaborated a detailed theory of social capital. Scholars have begun to criticize undertheorized and oversimplified usages of the concept, but most attack the problem by constructing their own variant conceptualization of social capital. Many of these variants represent progress in the middle range, and it is reasonable to customize one's definitions to the empirical question at hand. Still, the greater challenge remains -- to construct a theoretical explication of social capital that: 1) is not problem-specific; 2) is not restricted to a particular unit of analysis; 3) is theoretically consistent across levels of analysis; 4) is conceptually distinguished from correlate theories of agency, motivation, and social structuration; 5) can encompass, organize, and articulate those existing conceptualizations of social capital that have already proven viable.


To overcome the theoretical vagueness and disarticulation that have plagued social capital scholarship, we need a conceptualization that is precise, comprehensive, and that specifies problems for empirical investigation. My goal is two-fold. First, I will elaborate a conceptualization of social capital at each of three levels of analysis -- micro, meso, and macro -- incorporating and integrating the best of existing theory. Second, I will show how the three perspectives constitute the basis of a general theory of social capital, and discuss how the three perspectives work together to inform and provoke empirical inquiry.


The resulting general theory is not a set of substantive assertions to be proved or disproved through empirical investigation; rather, it is a conceptual tool for facilitating empirical inquiries by generating interesting, comparatively-oriented questions that will engender substantive theories in the middle range (see Appendix A). My hope is that this will constitute a step towards integrating social capital research into a coherent, theoretically consistent body of knowledge.


My ontological starting point is systemism: social capital is a processual system for allocating resources across a social network according to the pattern of relations among the individual egos that comprise the network. Systemism implies a three-dimensional analysis of the system's components, structure, and environment. In a system of social capital, the components are the individual egos that comprise the social network [1]. The system's structure is the pattern of relationship ties among the egos. The system's environment is constituted by the greater social ecology in which the system is embedded. In conceptualizing systems of social capital, I map components, structure, and environment directly onto the micro, meso, and macro levels of analysis.


The micro level of social capital analysis I call the embedded ego perspective. At this level, we are considering an individual ego's potential to mobilize resources through the social network in which the ego is embedded. Here, our focus is on individual nodal outcomes, but considered within the context of a particular social structure.


The meso level of social capital analysis I call the structural perspective. At this level, we are considering the structuration of a specific network of social capital, the patterning of ties between egos in that network, and the ways that resources flow through the network as a consequence of its particular structure. Here, our focus is on the process of network structuration and its distributional implications, rather than on the individual egos that comprise the network.


The macro level of social capital analysis I call the embedded structure perspective [2]]. At this level, we are considering the ways that a particular network of social capital is embedded in a system of political economy, and embedded in greater cultural or normative systems. Here, our focus is on "external" cultural, political, and macroeconomic influences on the nature of social ties in a network, influences on the structure of the network, and influences on the dynamics of the network's construction, change, and devolution.


The three levels of analysis are not mutually exclusive. Instead, they work together, the lens aperture opening and closing as necessary to focus on a specific aspect of the problem under consideration. Processes visible only at one level of analysis inevitably shed light on the workings of processes visible only at the other levels. Any given problem needs to be engaged at all three levels of analysis in order to construct a complete understanding of social capital as a dynamic process of social structuration.


A general theory of social capital must meet two criteria. First, the conceptualization must remain theoretically consistent across levels of analysis, which I accomplish here by relating the three levels to the "components, structure, and environment" dimensions implied by systemism. Second, the theory of social capital must articulate with theories of agency, motivation, and social structuration. These correlate theories must be disentangled from the theory of social capital by explicitly specifying the articulations among them. This allows social capital to serve as a self-sufficient theoretical construct that is compatible with various approaches to the problems of agency, motivation, and social structuration. This way, we can select for our analytical tool kit the theory of agency or of social structuration that is most apropos for the specific problem at hand, without having to rebuild our understanding of social capital from the ground up. By conceptualizing social capital in terms of systemism, we allow for causal interactions that flow vertically across levels of analysis. Consequently, the general theory of social capital is compatible with a variety of theoretical approaches to correlate problems.


THE EMBEDDED EGO PERSPECTIVE


The micro level of social capital analysis is the embedded ego perspective. At this level, we are considering an individual ego's potential to mobilize resources through the social network in which the ego is embedded. Here, our focus is on individual outcomes, but considered within the context of a particular social structure.


James Coleman


The first major theoretical statement on social capital was proposed by James Coleman in 1988. Coleman made a crucial error in defining social capital in terms of its function: "Social capital is productive, making possible the achievement of certain ends that in its absence would not be possible" (1988, p. 98). This unfortunate statement is largely responsible for so many of the undertheorized, oversimplified, and weakly conceptualized arguments that have plagued social capital scholarship to date, as researchers have accepted Coleman's definition uncritically and thus replicated his error.


To define a social phenomenon in terms of its outcome is to confound antecedent with consequent. It is conceivable that different antecedents can lead to the same outcome, or that the same antecedent can have different outcomes under different conditions. Sorting out these possibilities demands empirical investigation. But first, antecedents and consequences need to be disentangled in the fundamental conceptualization of the problem. Unfortunately, many researchers have mistakenly assumed that social capital can be understood in terms of its outcome, or that social capital always leads to "productive" outcomes. While most of Coleman's work on social capital is insightful and valuable, his definitional error demands correction.


Coleman's primary goal is not a theory of social structure, but a rational choice theory of agency that admits a social context. In discussing social capital, Coleman aims to mitigate the undersocialized nature of the rational choice approach by explaining how social structure shapes and constrains rational action, and how social structure is emergent consequent to individuals maximizing their self-interest. Rather than limiting his focus to strict methodological individualism, Coleman's approach to rational action is through a consideration of the patterns of relations among people.


Coleman addresses the source of trust as a central problem in rational choice theory. The existence of trust contradicts the opportunism implied by the assumption that individuals maximize their self-interest. While rational choice theorists have attempted to explain trust through an evolutionary account, as something that develops over repeated interactions between two self-interested actors in a dyadic relationship, the existence of more diffused trust dispersed throughout a larger social system remains problematic. Coleman observes that differences in the nature of social networks can affect the levels of trust among individuals in those networks. Consequently, the sociostructural context must be an important factor in constructing rational action. This realization mandates further inquiry, moving beyond the limits of strict methodological individualism.


Coleman observes that the creation of outstanding obligations between two individuals constitutes a tie between them. The existence of outstanding obligations also constitutes a resource upon which people can draw in times of need, thus explaining the formation of social capital in terms of rational action -- it is rational and self-serving to create social capital that you can draw upon in the future, analogous to opening a savings account to reserve financial capital for future use. Further, the creation of outstanding obligations -- when repeated across time between actors, and repeated across social space among actors in a given network -- constitutes a basis for generalized trust among members of a common network.


This gives us an explanation for diffused trust throughout a network, an explanation that is consistent with rational choice theory. The problem with this explanation is that it does not address the relevance of alternative motivations for forming social ties and trusting people, such as preexisting cultural systems of norms and values, or -- to remain within a rational choice framework -- the influence of preexisting rules of transaction that obtain within a social system and serve to foster trust.


Coleman also observes that social capital can be a by-product of activities engaged in for other purposes (1994, p. 312). This opens the door to cultural and normative explanations for the formation of social capital, although Coleman does not pursue these alternatives to rational choice. Instead, he explains the formation of norms as an outcome of rational action, as a "means of reducing externalities" (1994, p. 317). While this approach does offer an explanation for the origin of norms, it once again does not admit the effects of preexisting cultural and normative systems on the structuration of social capital (or, to remain within the rational choice framework, the effects of preexisting rules of transaction). Nor does it allow for dynamic interaction between the actions of individuals and their social context. Instead, the dynamics of the formation and destruction of social capital are explained as the outcome of aggregated individual actions -- a unidirectional conception of causation from micro to macro.


But trust may not necessarily be an outcome of networking. The existence of social capital networks could also be explained as a consequence of preexisting widespread trust and generalized expectations of reciprocity in a society. Coleman's conceptualization is weakened by his assumption of unidirectional micro -- > macro causation. I would argue that the "chicken and egg" question needs to be problematized; which comes first -- trust, or social capital networks? While Coleman does make some provisions for the embeddedness of individuals within a network, his reductionism constitutes a major weakness when it comes to assessing the embeddedness of meso-level social structures within preexisting macrostructural regimes, as I will demonstrate below.


Alejandro Portes


Alejandro Portes offers a more finely-grained and comprehensive conceptualization of the origins and nature of social capital (Portes & Sensenbrenner 1993; Portes 1995). Portes defines social capital as:


"...the capacity of individuals to command scarce resources by virtue of their membership in networks, or broader social structures... The ability to obtain [social capital] does not inhere in the individual... but instead is a property of the individual's set of relationships with others. Social capital is a product of embeddedness" (Portes 1995, pp. 12-13).


While this definition is ego-centered, Portes goes on to expand the conception of social capital to include characteristics of the social network itself. Borrowing from Granovetter (1990), Portes distinguishes between relational embeddedness and structural embeddedness. Relational embeddedness describes dyadic expectations of reciprocity based upon the expected ability of the other actor in the dyad to enforce sanctions. But when both actors are part of a broader network -- and hence structurally embedded -- trust is increased because of mutual expectations that the broader community will enforce sanctions. Portes calls this "enforceable trust" (1995, p. 14).


Portes holds that expectations of reciprocity and enforceable trust are both instrumentally motivated by fear of sanction. He thus invokes a rational choice theory of agency motivating the structuration of these two types of social capital. But more importantly, by moving from social ties regulated by expectations of dyadic sanction, to social ties regulated by expectations of enforceable trust, Portes expands his conception from an ego-centered definition to one that admits more of the social network's structural influence. This serves to articulate the micro and meso levels of analysis, sustaining theoretical consistency when changing the lens focus.


Portes identifies two additional types of social capital that move beyond the rational choice explanation to admit cultural and normative motivations. First, he notes that internalized values and norms, or "value introjection", can motivate one to establish social ties or transfer resources to others because of general moral imperatives (1995, p. 15). Next, he introduces the notion of "bounded solidarity", which can motivate one to establish social ties or transfer resources to others because of identification with in-group needs and goals (1995, p. 15). Here again, in moving from value introjection to bounded solidarity, Portes begins with an egocentric approach and then expands it towards a meso-level explanation, thus articulating the relation between ego-centered and structural causation. Further, he admits macro-level causation by incorporating, within the concept of bounded solidarity, the influence of systems of particularism.


Portes's conceptualization of social capital is valuable for two reasons. First, he has elaborated differences in the characteristics of the social ties among egos, explaining these differences as an outcome of the differing characteristics of the social network in which they are embedded, and as an outcome of the degree or type of embeddedness into a network. This allows us to conceive of social capital as processual, a dynamic outcome of reciprocal causation between an ego and the social structure in which the ego is embedded. Second, Portes has specified different types of motivation behind the structuration of social capital, thus expanding upon Coleman's unidimensional rational choice explanation. This allows us, from the embedded ego perspective, to articulate our conceptualization of social capital with a variety of theories of agency, motivation, and social structuration. We are no longer restricted to the rational choice approach, although it is still available as one component of our analytical tool kit.


Egocentric Power


Portes and Coleman both fail to adequately address the role of unequal power relationships among egos as a factor influencing the creation, maintenance, and destruction of social capital. Egos with greater power will realize more opportunities to create, control, and thus exploit network ties. It is important to distinguish between the power that the ego realizes as a consequence of his personal characteristics, and the power that the ego realizes as a consequence of his network position. From the embedded ego perspective, we are concerned with the first -- ego's ability to command resources by virtue of ego's being able to command linkages or command transactions. From the structural perspective, we are concerned with the second -- ego's ability to command resources by virtue of ego's position within a network. These dimensions of power are differentiated only for analytical purposes. In practice, they will work together. A complete account will observe the interactions between the two levels of analysis.


From the embedded ego perspective, we can identify three ideal types of power that obtain as a consequence of personal characteristics from which an ego can benefit within a system of social capital. Coercion describes ego's ability to compel action on the part of others. This includes the ability to enforce sanctions (as Portes observes), to compel the creation or destruction of network relationships, and to mobilize resources across network ties. This type of power derives from ego's preexisting control of the resources necessary to wield the types of coercion he needs to create and exploit a social tie. Charisma describes ego's ability to attract and exploit network relationships with others as a consequence of personal magnetism. Authority describes ego's ability to attract and regulate network relationships as a consequence of authority conferred by personal characteristics, as determined by the greater cultural or normative systems in which the network of social capital is embedded. Figure 1 displays the four types of social capital described by Portes, plus the three additional types of social capital deriving from egocentric power.


The Downside of Social Capital


Portes also observes that social capital has a downside (Portes 1995, pp. 14-16; Portes & Landolt 1996). Social relations that allow an ego to mobilize resources across a network can also imply obligations towards other network members. This can create two types of problem for the individual ego. First, the ego is subject to claims inspired by those obligations, thus exposing him to free-riding by other network members upon his own resources. Second, community norms can place constraints on individual entrepreneurship or innovation. Membership in a tightly-knit network can subject one to restrictive social regulations and sanctions that limit individual action.


We can conceptualize this problem as an analog of the fundamental accounting identity describing financial capital:


Capital = Assets - Liabilities


Following Portes & Landolt's argument, we can then identify social capital from the embedded ego perspective as the difference between social assets and social obligations (Babones 1996). This observation differs sharply from the assumptions of many meso and macro-oriented analysts of social capital who disregard the inherent costs of maintaining social ties and mistakenly assume that social capital is only and always positive. This accounting identity gives us a starting point from which to proceed, from the embedded ego perspective, with an analysis of the distributional outcomes of an ego's embeddedness within a particular system of social capital.


Portes & Landolt's conception of the downside is primarily focused from the embedded ego perspective, addressing individual outcomes rather than group outcomes. But while some types of social capital may create negative individual outcomes, they may not necessarily be harmful to the interests of the group as a whole. For example, types of social capital that constitute a public good may also constitute unwanted costs on the individuals called upon to maintain the system. Conversely, another possibility is a type of social capital that is beneficial to individuals in the network but harmful to the greater society in which the network is embedded, such as organized crime networks or systems of political patronage. Such vertical interactions between the micro, meso, and macro levels underscore the necessity of analysis at the macro perspective, in which we consider the embeddedness of the networks themselves within larger social systems.


THE STRUCTURAL PERSPECTIVE


The meso level of social capital analysis is the structural perspective. At this level, we are considering the structuration of a specific network of social capital, the patterning of ties among egos in that network, and the ways that resources flow through the network as a consequence of its particular structure. Here, our focus is on the process of network structuration and its distributional outcomes, rather than on the individual egos that comprise the network. At the meso level, the major theoretical statement on social capital is Ron Burt's structural holes theory (1992; 1993). For Burt, social capital is owned jointly by all parties to a relationship; hence, a structural perspective is mandated. Burt's theory incorporates ideas previously proposed by Simmel, Coleman, and Granovetter.


Coleman describes a characteristic of social networks that he calls "closure" (1994, pp.318-320). Consider a three-actor network. When a two-way tie is established between each of the dyadic pairs in the triad, the system is deemed to have closure (Figure 2a). We assume that in a closed system, there is the potential for any information or resource to flow directly to any node in the network by the shortest possible route. Hence closed systems magnify systemic interactions, engender the emergence of norms and systems of regulation and sanction, and experience inflationary or deflationary spirals of trust.


But if we break one link in that triadic system, the situation changes in important ways (Figure 2b). The actor who has links with both of the other two actors has a distinct advantage in competitive situations. By virtue of his central location between unconnected nodes, he now has direct access to all of the resources in the system, while each of the other two actors must access resources through him. Simmel called this position the tertius gaudens, the "third who benefits" (Wolff 1964, p. 154). By occupying the structural location between otherwise unconnected nodes -- what Burt calls a structural hole -- the tertius gaudens realizes greater returns on the social capital extant within his network. This insight forms the foundation of Burt's structural holes theory.


The other fundamental idea inspiring structural holes theory is Mark Granovetter's description of "the strength of weak ties" (Granovetter 1973). Granovetter argued that access to new information is obtained through an ego's weak ties to nodes at a distance from his own local network. Information within his local network is widely shared locally, hence most of his local contacts are redundant. New information comes from non-redundant ties.


The tertius gaudens in our example above has access to "non-redundant" resources available within the network through both of his ties. Burt argues that in more complex networks, nodes who occupy a central location with connections to dispersed, non-redundant clusters of nodes, have access to more of the network's resources at less cost to themselves (Figure 3b). Redundant ties to multiple nodes within the same cluster require more effort to maintain (Figure 3a). By creating ties only to clusters of nodes in which you do not already have a contact, you increase your access to resources without redundantly increasing your maintenance and transaction costs. Burt's innovation is to argue, contra Granovetter, that the crucial causal element here is not the strength or weakness of the ties, but rather their redundancy or non-redundancy in relation to the ties you have already established.


Further, Burt observes that structural holes within a network not only have greater non-redundant access to resources, but are also strategically located to control the flow of resources between the clusters of nodes that the structural hole links together. Thus Burt constructs a conceptualization of power relationships within a system of social capital. For example, considering our three-actor model once again (Figure 2b), you can see that the tertius gaudens controls the flow of resources between the other two nodes, placing him in a position of greater power.


Burt conceptualizes social capital as the degree to which the structure of a network confers access to resources and control over resources to various nodes in the network. The outcome for any particular node can only be assessed as a function of the network's overall structure. But Burt resolutely avoids admitting the attributes of the actors involved, or the nature of the ties between them. This elision is both a strength and a weakness of his theory.


Burt argues that the "who" one accesses resources from is secondary to the "how" one accesses those resources over the network structure (Burt 1993, p. 69). He assumes that individuals who form dyadic ties will share extensive social similarities in interests, wealth, power, values, etc. And since socially similar people form ties, an ego's close contacts will have redundant resources. Serious problems arise from assuming away the import of "who". First, in considering interpersonal relations, you can easily develop examples of redundant social ties between people of widely differing social status or power. Second, Burt's theory is not limited to interpersonal networks, but purports to describe the functioning of social networks in general. You could argue that in business networks, for example, ties between similar firms are the exception, not the norm -- similar firms are in competition with each other. Further, ties are made between firms precisely in order to maximize direct access to a particular resource. In this situation, the "who" is crucial.


Burt ignores the "who" for two reasons (1993, pp. 69-70). First, he argues that focusing on network structure increases the generalizability of his theory. He wants to focus on aspects of social capital that generalize beyond the characteristics of the specific individuals connected by a relationship. My criticism here is that particularism is a feature of most networks, and that the question of whether particularism manifests only within isolated dyadic ties or as a feature that integrates the network as a whole is a crucial systemic-level variable, as we have seen in Portes's conception. Consider for example the role of particularism in organizing Asian business networks, e.g., the Japanese keiretsu, or the kinship-based Chinese networks (Granovetter 1994). Could, would these systems even exist if the actors selected network contacts out of purely individualistic motivations? Here, the "who" is the most important determinant of network structuration. The influence of particularism on a given network's process of structuration is best observed from the macro-level embedded structure perspective, and may not be apparent from Burt's meso-level perspective.


Second, Burt argues that the "who" is highly correlated with the "why" and the "how", and hence they are largely redundant in an analysis. He claims that actors who know how to structure their personal network also know with whom to network. This implies that actors are networking solely in order to maximize their self-interest, or that actors who are maximizing are the only ones worth studying. My criticism here again is with Burt's unidimensional theory of agency and motivation.


Burt's empirical demonstrations all locate agency in rational actors motivated by self-interest. This is consistent with his argument that the nature of network ties is secondary in importance to their structural location. When making and breaking a tie is based solely upon the players' competitive access to resources within the network, as implied by a rational choice theory of agency, then ties are valued by actors only along that single dimension. The choice of making or breaking a tie is decided by simple, easily-modeled equations. This parsimony gives Burt's conception its explanatory power, to the extent that his reductionism is plausible for a given empirical analysis.


But as we have seen above, there are manifold motivations behind network structuration, some of which do not adhere to rationalized, competitive maximizing of self-interest. Burt does not deny the existence of such motivators, only their import on a competitive outcome. His argument is that those who do not compete on the basis of self-interest are merely scenery, and hence not of theoretical interest (Burt 1993, p. 98).


I would argue that even the scenery constitutes part of the network structure through which the competitive players must maneuver. An ego's competitiveness is not a binary-state variable, but instead falls along a continuum. Network actors' location on the continuum of competitiveness will need to be problematized in many empirical inquiries. Further, even non-competitive scenery takes part in creating community regulations and meting out sanctions when expectations of reciprocity and trust are violated. And I would argue that the very existence of a given network may stem precisely from normative or particularistic motivations, rather than from pure self-interest on the part of the individual actors. Consequently, ignoring the scenery may be risky. Again, the import of the scenery may best be apprehended from the embedded structure perspective, and may not be apparent from Burt's meso-level perspective.


Further, in a rational choice framework, the problem of trust remains. Does Burt implicitly assume that the social similarities leading individuals to form ties with one another will be enough to engender trust? If so, that would contradict the rational choice assumption of opportunism. If he is not assuming dyadic solidarity on the basis of social similarity, then where do trust and reciprocity enter into the picture? Certainly, trust is a function of the strength of a tie, but Burt has dismissed the import of the tie's strength or weakness. He simply assumes away the problems of trust and reciprocity when it comes to forging and sustaining network ties.


The "chicken and egg" question of causal flow remains: is it the existence of generalized trust and generalized expectations of reciprocity that makes networks possible, or does the existence of networks generate trust and expectations of reciprocity? Burt's reluctance to elaborate the articulation between his theory of agency and his theory of network structuration is problematic. Too much is assumed or ignored.


For the purposes of his theory, Burt assumes that trust is person-specific (1992, p. 16). A more comprehensive conception of social capital would hold that trust may also be a function of network characteristics, or of the macrostructural regime in which the network is embedded. Portes has specified this as "bounded solidarity" and "enforceable trust". Further, when a player minimizes network maintenance and transaction costs by sustaining only non-redundant ties, then it becomes crucial that those ties be trustworthy. This suggests that the value of non-redundancy in a tie is a function of that tie's trustworthiness -- thus making it risky to ignore the import of trust. Burt does note that strong trust in a dyadic tie takes time to develop, which argues for a processual perspective on trust (Burt 1992, p. 215). Here, he reluctantly begins to admit the importance of the strength of a tie.


While Burt's reductionist approach is flawed as a general theory of social capital, it still makes for powerful, insightful theory that is useful in empirical analyses in which it is: 1) plausible to assume that actors are motivated by competitive self-interest rather than by normative systems or particularism, and 2) plausible to assume that the network is bounded by variables exogenous to the competition. But when we admit cultural and normative motivations for making and breaking ties, the ties become much stickier, and the dynamics of the network's structuration become much more complexly motivated and indeterminate. The advantage of Burt's model lies in its parsimony and generalizability. Its weakness lies in Burt's reluctance to admit environmental causal factors that influence the dynamic, processual aspects of the network's structuration.


My criticisms above all speak to structural holes as a general theory of social capital. However, when we divest Burt's theory of its rational choice assumptions and consider what remains as a sub-component -- the meso perspective -- of a more comprehensive general theory of social capital, then structural holes performs admirably. In this light, we are demanding less of structural holes theory. We do not ask it to come pre-packaged with a theory of agency and motivation -- we will develop that at either the micro level, if we want a rational choice framework, or at the macro level, if large-scale social influences seem more relevant to the problem at hand. Network analysis is compatible with a variety of theories of social structure and agency, as long as networks are seen as meso-level outcomes of action on the micro and macro levels, rather than as generative forces in themselves.


Structural holes theory is a remarkable and impressive achievement, when considered purely as an approach to the network structure of social capital at the meso level of analysis, employed in conjunction with well-articulated analyses at the micro and macro levels. It offers a strong conceptualization of social capital at the structural level, and it can be operationalized using existing methods of network analysis. Structural holes theory offers an explanation of how a system of social capital allocates resources as a consequence of the system's structure. Further, at the macro level of analysis, considering networks themselves as units of analysis in a structural holes perspective can help us understand why resources are channeled into certain networks of social capital while excluding others.


THE EMBEDDED STRUCTURE PERSPECTIVE


At the macro level of analysis, we are considering how a network in which social capital is created, manifested, and deployed is embedded within larger systems of political economy, or within larger cultural and normative systems. If we eschew rational choice motivations for the structuration of social capital, then we must move to the macro level of analysis to articulate our understanding of social capital with our theories of agency, motivation, and social structuration. Eschewing methodological individualism, the macro level is where we observe important motivating forces. Here is where we perceive macrosocial regimes that influence who forges ties with whom, influence how those ties manifest, and influence the distribution of resources both within networks and among networks. Here we perceive the macrosocial logic that commands the very existence of many kinds of social capital networks.


Even if we remain within the rational choice framework, an embedded structure analysis is still mandated by the existence of predetermined rules of transaction. Even if we assume that the social systems encompassing a network of social capital are an outcome of aggregated individual maximizing, we must still consider that frictionally ossified macro-level trading regimes are working to shape and structure individual transactions at the micro and meso levels of analysis. Clearly, an embedded structure perspective is relevant no matter with which general theories of society we articulate our theory of social capital.


The embedded structure perspective on social capital has been little explored. The relevant theoretical comments are found in cultural sociological approaches to the economy, and do not directly focus on social capital (Brown 1997; Zelizer 1988; Zukin & DiMaggio 1990). Thomas Ford Brown (1997) presents a theory of industrial structuration that incorporates a consideration of cultural factors which influence social capital networks on the demand side. Viviana Zelizer criticizes economic sociology in general for its "structural absolutism" -- reducing everything to sociostructural relations and networks (Zelizer 1988, p. 629). Zelizer also objects to cultural absolutism, and calls for a balanced approach that considers structural, economic, and cultural factors. Zukin and DiMaggio (1990, p. 17) introduce the concepts of: 1) political embeddedness, which observes that economic action is always carried out in a greater political context; and 2) cultural embeddedness, which observes that economic assumptions, rules, and rationality are limited and shaped by culture.


Following these scholars, the embedded structure approach to social capital observes that networks of social capital are embedded in overlapping systems of politics, economy, and culture. These macrosocial regimes may: 1) determine the types and amounts of resources available to the network over time; 2) describe with whom actors may forge ties, thus bounding and structuring the network; 3) legitimate and regulate transactions; 4) construct and implement sanctions in response to violations of the regulatory system; 5) describe and regulate social status within the network; 6) construct the motivations underlying network transactions; 7) construct and regulate competition between different networks.


The types of embeddedness described above are in framed in terms of macro -- > meso and macro -- > micro causation. A complete analysis from the embedded structure perspective will also consider micro -- > macro and meso -- > macro causation, and further consider the import of reciprocal causation across levels. An excellent example of reciprocal causation across levels would be "conspiracies against the public", a term coined by Adam Smith (1986) to describe collusive networks of businessmen who conspire to enrich themselves to the detriment of the public good of the greater population in which the network is embedded. Other examples of conspiracies against the public would include criminal networks and systems of political patronage. Portes & Landolt (1996) suggest that networks whose membership is bounded by particularism can also constitute conspiracies against the public, in that network resources are reserved to in-group members while access is denied to the general public.


In the above examples, macro -- > micro causation is apparent in the ways that macrostructural regimes construct the fundamental motivations for wealth, status, and power that drive the structuration of the collusive networks. Macro -- > meso causation is apparent in the fact that network structuration is shaped by its illegitimacy of purpose, which creates a need for forms of organization and transaction that can evade the macrostructural regulatory regimes. Meso -- > macro causation is apparent in the ways that the macrostructural regimes are shaped by their need to devise regulatory measures in response to the anti-systemic threat posed by the collusive networks.


We can specify three ideal types of social capital networks, categorized according to the prevailing motivation behind the network's transactions: 1) "Economic" describes a network of egos motivated by economic considerations; 2) "Status" describes a network of egos motivated by reputational considerations; 3) "Sociability" describes a network whose egos forge ties according to altruistic or particularistic motivations.


We can further specify three ideal types of macrosocial regimes in which social capital networks are embedded: 1) macroeconomic; 2) cultural/normative; 3) political. By cross-categorizing the ideal types of network motivation with the ideal types of macrostructural influence, we can construct an ideal typology of embedded structures (see Figure 4).






                                          Macrosocial Regime
Network motivation
            Macroeconomic    Cultural/Normative    Political
Instrumental-economic        1        2        3
Instrumental-status        4        5        6
Sociability            7        8        9

Figure 4: Ideal Typology of Embedded Structures



Ideal Typology Of Embedded Structures


1) Network of instrumentally-motivated egos optimizing on economic utility functions, embedded in a macroeconomic system that influences the types andamounts and types of economic resources available over time. One example of this ideal type of embedded structure would be a network of traders in a commodity market. Charles Smith (1989) describes how regular buyers in some established auctions divide access to resources according to what amount and quality is available at any given time. In times of scarcity, buyers with greater prestige within their local network get first choice of goods. Here, the macroeconomically-determined supply influences local processes of distribution that cross-cut with how the buyers and sellers are organized within their own network of social capital.


2) Network of instrumentally-motivated egos optimizing on economic utility functions, embedded within a culturally-determined system of norms that bounds the network by influencing with whom ties can be forged, that regulates transactions, that constructs sanctions in response to violations of the normative system. Here again, Smith's account of networks of auction traders exemplifies the importance of cultural factors in determining network membership, regulation, and organization. Another example would be the ethnic economic enclaves described by Coleman (1988) and Light (1994), in which economic transaction networks are bounded, organized, and regulated according to ethnicity and reputational considerations.


3) Network of instrumentally-motivated egos optimizing on economic utility functions, embedded within a political system that legitimates and regulates transactions, that implements sanctions in response to violations, and that influences the types and amounts of resources available over time. This ideal type is best exemplified by politicians who use their position within the legislature to benefit themselves economically through establishing and exploiting network ties that can bring them board memberships, speaking engagements, post-political career employment, or campaign contributions. Their self-seeking economic interests are pursued within the context of a particular network of social capital that is shaped and regulated by the political system in which the actors are embedded.


4) Network of instrumentally-motivated egos optimizing on reputational utilities, embedded within a macroeconomic system that influences the status constructed in the cultural system. Examples of this system can be found at the highest levels of finance capitalism, in which actors who are already exceedingly wealthy still compete to amass even more wealth, more to increase their power and social status than to simply accrue wealth as an end in itself. Networks of social capital can be found in this realm in exchanges of board memberships, political fund-raising initiatives, and other deal-making relationships in which economic position influences status and reputation.


5) Network of instrumentally-motivated egos optimizing on reputational utility functions, embedded within a culturally-determined system of norms that describes social status, that determines the types of transactions that affect status, that regulates status transactions. A classic example would be the "kula ring" trading system found in historic New Guinea, in which symbolic prestige goods of negligible economic value were traded by men according to culturally-determined rules of transaction in order to bolster and increase their social status (Malinowski 1922). Another example would be the "potlatch" institutions found in the American Indian societies of the Pacific Northwest (Murdock 1936). In these systems, men competed to distribute goods to others in order to increase their own social standing.


6) Network of instrumentally-motivated egos optimizing on reputational utility functions, embedded in a political system that describes social status, that determines the types of transactions that affect status, that regulates status transactions. This ideal type can be found in many political realms, such as the favor-trading networks common within democratic legislative bodies of modern industrial society. Politicians capitalize on their network position to trade votes and otherwise influence policy-making in order to increase their power and status within the legislative body and within the polity as a whole.


7) Network of sociability embedded in an economic system that influences socioeconomic status within the network. An example would be high-society social circles, such as debutante balls and various upper-class charity endeavors, in which participation is predicated on one's position within a particular social network whose boundaries are circumscribed by economic class location. One might observe economic class location serving to bound many social circles of friends at all levels of income. A wide variety of resources can be mobilized through networks of sociability -- economic, informational, relational, etc.


8) Network of sociability embedded within a culturally-determined system of norms that constructs the motivation for forging outstanding obligations, that determines the types of transactions subject to its regulatory domain, that constructs sanctions in response to violations of the normative system, that determines with whom ties are to be forged, that thus bounds the network according to particularism, that determines what types of transactions are within its regulatory domain. This ideal type is exemplified by socializing practices in which party-throwing and party attendance function as means of constructing and maintaining network ties through creating and fulfilling obligations. Another example would be holiday gift exchange networks. A third example would be the ethnic sociability networks commonly observed in most American educational institutions that are integrated, as students organize into social circles on the basis of ethnicity and mobilize resources through those networks.


9) Network of sociability embedded within a political system that provides access to resources, that produces or destroys norms of trust and reciprocity, that enforces norms of trust and reciprocity. Again, this ideal type can be found in many political realms, such as the legislative bodies of any modern industrial society. This ideal type, however, describes legislators motivated by altruistic ideals of "public service". Clearly, this ideal type will often coexist with ideal types #3 and #6, as individual legislators may potentially act from both instrumental and altruistic motivations. A different example would be the "occupational ghettoes" in which a particular ethnic group gains a foothold in a governmental office or department, and takes advantage of this niche to benefit coethnics through granting jobs or other political favors and benefits. This is an example of bounded solidarity, as altruistically-motivated actors distribute resources to their coethnics according to in-group goals and needs.


In empirical research, we would expect to encounter networks embodying combinations of characteristics from various ideal types. Individuals within a single system of social capital may operate from differing motivations at different times with different network contacts, influenced by different macrostructural forces. Motivations within the network may vary according to the type of resource being distributed and exchanged. Combinations of motivators may even be multiplexed over a single network tie or transaction. And network structuration is a dynamic process, consequent to the dynamics of ego motivation and the dynamics of macrostructural causal influences. The ideal typology, consequently, is an heuristic tool. It is not intended to represent actual systems of social capital, but rather to inspire interesting questions about macrostructural influences on network motivation and structuration, about the flow of causation, and about the processuality of network structuration.


CONCLUSION


I began by defining social capital as a system for distributing resources across a social structure according to the relations among the egos that constitute the structure. I then conceptualized the system in terms of its components, its structure, and its environment. By mapping these three analytical dimensions onto the micro (embedded ego), meso (structural), and macro (embedded structure) levels of analysis, I have been able to construct a conceptualization that is theoretically consistent across levels. Further, I have incorporated and integrated within one heuristic framework the major theoretical statements on social capital to date.


While the conceptualization presented here has been developed in terms of the social ontology of systemism, and informed by the relational approach to social theory, it is theoretically self-sufficient. It contains built-in "hooks" that can articulate with a variety of theoretical approaches to correlate problems of agency, motivation, and social structuration.


The goal has been a general theory of social capital. The theory is not a provisional end-product, not a set of substantive statements that can be proved or disproved through empirical investigation. Rather, the theory is a means toward greater understanding of social capital in comparative perspective. The theoretical vagueness and disarticulation that has plagued social capital scholarship is corrected through detailed and comprehensive conceptualization. I present here an heuristic tool designed to inspire and lay the groundwork for the construction of substantive theories in the middle range. It is a tool that allows investigators to generate interesting, comparatively-oriented questions about systems of social capital that can be tested empirically (see Appendix A), thus facilitating research that can be integrated into a coherent, theoretically consistent body of social capital scholarship.


APPENDIX A -- HEURISTIC QUESTIONS TO GUIDE INVESTIGATION


1) How is the network bounded? What are the influences bounding the network? Are they local in origin, or are they part of a macrostructural regime?


2) What kinds of resources flow through the network as a consequence of the network's relational structure?


3) Are egos relationally embedded or structurally embedded?


4) What motivates tie formation? Is the motivation instrumental or altruistic? Is the motivation generated at the ego level or at the macrostructural level? Is the motivation normative, particularistic, or rational maximizing?


5) What determines "who" forms ties with one another?


6) What motivates tie destruction?


7) How persistent are ties in time and strength?


8) How does the ego benefit as a result of his embeddedness? How is his embeddedness to his detriment? Can his social capital be calculated in terms of the accounting identity?


9) Where are the structural holes? Are they "white holes" or "black holes"?


10) How dynamic is the system? At what speed does it change? What are the forces influencing the system's dynamism?


11) Does the existence of the system of social capital impose a cost on the greater society in which it is embedded, or does it constitute a public good, or is the outcome negligible for the greater society?


12) How do shifting macrostructural regimes of culture or economy impact upon the network's structuration? How do they impact upon the resources flowing through the network?


13) How does network structuration influence the structuration of the macrostructural regimes in which it is embedded?


14) How might macrostructural regimes influence the distribution of resources among competing networks, and how might that competition between networks influence the structuration of the network being analyzed?


FOOTNOTES


[1] The word "ego" refers to a specific node in a social network. That node may be an individual, an organization, a nation-state, or some other social entity.


[2] The meanings of "meso" and "macro" in this framework are relative to the unit of analysis. For example, when we are considering the social capital that pertains in a network of nation-states, the structure of the inter-state network would be analyzed at the "meso" level, while the embeddedness of network transactions within a pre-existing global regime would be analyzed at the "macro" level.


REFERENCES


Babones, Salvatore. 1996. Personal communication.


Brown, Thomas Ford. 1997. "Consumer Demand and the Social Construction of Industry." Working Paper Series, Program in Comparative and International Development. Johns Hopkins University.


Burt, Ronald. 1992. Structural Holes: The Social Structure of Competition. Cambridge: Harvard University Press.


Burt, Ronald. 1993. "The Social Structure of Competition." Pp. 65-103 in Explorations in Economic Sociology, edited by Richard Swedberg. New York: Russell Sage Foundation.


Coleman, James. 1988. Social Capital in the Creation of Human Capital. American Journal of Sociology 95:S95-S120.


Coleman, James. 1994. Foundations of Social Theory. Cambridge: The Belknap Press of Harvard University Press.


Granovetter, Mark. 1973. "The Strength of Weak Ties." American Journal of Sociology 78:1360-1380.


Granovetter, Mark. 1990. "The Old and the New Economic Sociology: A History and an Agenda." Pp. 89-112 in Beyond the Marketplace: Rethinking Economy and Society, edited by Friedland, Roger and A.F. Robertson. New York: Aldine de Gruyter.


Granovetter, Mark. 1994. "Business Groups." Pp. 453-475 in The Handbook of Economic Sociology edited by Neil J. Smelser and Richard Swedberg. Princeton: Princeton University Press.


Light, Ivan & Stavros Karageorgis. 1994. "The Ethnic Economy." Pp. 647-671 in The Handbook of Economic Sociology, edited by Neil J. Smelser and Richard Swedberg. Princeton: Princeton University Press.


Malinowski, Bronislaw. 1922. Argonauts of the Western Pacific. London: G. Routledge & Sons, ltd.


Murdock, George Peter. 1936. Rank and Potlatch Among the Haida. New Haven: Yale University Press.


Portes, Alejandro. 1995. "Economic Sociology and the Sociology of Immigration: A Conceptual Overview." Pp. 1-43 in The Economic Sociology of Immigration: Essays on Networks, Ethnicity, and Entrepreneurship, edited by Alejandro Portes. New York: Russell Sage Foundation.


Portes, Alejandro and Patricia Landolt. 1996. "The Downside of Social Capital." The American Prospect 26, May-June:18-21.



Portes, Alejandro and Julia Sensenbrenner. 1993. "Embeddedness and Immigration: Notes on the Social Determinants of Economic Action." American Journal of Sociology. 98, May:1320-1350.


Smith, Adam. 1986. The Wealth of Nations. New York: Viking Penguin Classics.


Smith, Charles. 1989. Auctions: The Social Construction of Value. New York: Free Press.


Wolff, Kurt. 1964. The Sociology of George Simmel. New York: The Free Press.


Zelizer, Viviana. 1988. "Beyond the Polemics of the Market: Establishing a Theoretical and Empirical Agenda." Sociological Forum 3:614-634.


Zukin, Sharon and Paul DiMaggio. 1990. The Structures of Capital: The Social Organization of the Economy. New York: Cambridge University Press.

[此贴子已经被烝民于2005-4-11 0:39:06编辑过]