The first part of this paper dealt mainly with the first problem I identified: that incentives alone are a fragile base on which to erect explanatory structures. Even this relatively micro-level point moves the initial analytic focus away from individuals, since the crucial explanatory complements to incentives – trust, power, norms
and identity – are enacted in horizontal and vertical relations. Only by confining analysis to individuals can one easily sustain a narrow instrumentalist view. I then moved to the second problem, identifying social spaces and institutions or institutional sectors within which people act, and sketching arguments about how such spaces arise, are coupled or decoupled, and how resources flow among them.
Now I want to sketch how we might draw together these micro and macro strands– how individual actions, conditioned by incentives, trust and cooperation, power and compliance, and norms and identities that affect these states and actions, are shaped by and themselves reshape larger institutional configurations. As before, the issues in economic action have a family resemblance to those in theories of political action. For example, in “The Strength of Weak Ties” a paradox, that residents of Boston’s West End, devoted to their neighborhood, and horrified at the prospect of its demolition for “urban renewal”, nevertheless failed to resist by uniting and mobilizing behind local leaders. It was also argued that working-class culture, with its distrust of self-seeking leaders, sharply discouraged membership in political groups. The instrumental theorist might instead see garden-variety free-riding – every individual hoping that others would bear the cost of mobilization. My riposte applies also to the free-rider argument: much might be explained by social structural constraints. I proposed that the neighborhood consisted of cohesive network clusters which were, however, highly decoupled from one another, and that this fragmentation made mobilization difficult, whatever the intentions of individuals.
In more current terms, I suspected a deficit of individuals who could sit astride the West End’s structural holes and send out weak ties into various cliques in order to mobilize resources and claim a leadership role. Mobilization did occur within cliques, but could not spread beyond them. Distrust of leaders beyond one’s network may have stemmed from the lack of a short chain of social relations between ego and such leaders. Where such chains exist, reassurance about the leader’s intentions flow along them, plausible in part because one has the possibility of exerting influence through the chain in ways that restrain self-seeking. Here in fact we see issues of trust and power combined as they may often be, through the overall configurations of horizontal and vertical ties.
Economic formations should follow similar principles. Silicon Valley success lays special emphasis on the openness of networks and the free flow of people, ideas and capital across the porous boundaries of firms. Her argument highlights an extraordinary amount of trust among companies and individuals nominally in competition, in part because loyalties lay more with occupational groups; in part because rapid mobility meant that people in separate firms had often once worked together; in part because the culture of engineers stressed heavily the macho goal of exhibiting technical prowess to one another, often more important to self-esteem than high salary or job security.
By contrast, the Route 128 complex in the Boston metropolitan area shows an uncomfortable resemblance to West End, a collection of what is called (in a German context) “cathedrals in the desert”, trying to be self-sufficient, avoiding the sharing of ideas or personnel, and ultimately finding this strategy self-defeating in a fast-moving technical environment. The successful model resulted from complex intersections of firms, occupational groups and social networks, and a mobilization of goals that were a mixture of personal pride, social standing and financial gain, harnessed to one another in ways that led to achievements no single goal alone could have sustained.
Given the extensive network connections, structural holes are few in such a setting. There is correspondingly little in the way of power centers among Silicon Valley industrial firms, even though some have grown large and important in revenues. But in the supporting infrastructure, such as finance and law, there is much more striking stratification and hierarchy of power. Though the systematic research remains to be done on structures of status and influence, informal accounts suggest that having the right venture capitalist (e.g. Kleiner, Perkins, Caulfield and Byers, or KP in local parlance) or the right law firm (e.g. Wilson, Sonsini, Goodrich and Rosati) is a great advantage in one’s industrial progress; such firms therefore have the power to dictate terms favorable to themselves. Historical accounts suggest that this dominance traces back to earlier periods when these leading firms faced a fragmented resource base and were unusually successful in mobilizing across separate networks and sources, as I briefly discussed earlier under the heading of the relative coupling and decoupling of finance and industry.
Another example illustrates and further develops these themes. Richard Locke’s Remaking the Italian Economy (1995) analyzes contrasting outcomes from the 1970s and 1980s re-structuring of the two major automakers Fiat and Alfa-Romeo. Fiat restructured by vigorously repressing labor unions, creating so much industrial conflict that the entire region suffered. Alfa Romeo had a more complex negotiated process with a happier regional economic ending. Locke attributes this difference to how networks of political actors and associations were structured in Turin and in Milan. In Fiat’s base, Turin, political actors and associations mainly clustered in two opposing camps, one associated with business and the other with labor, having strong internal links but few connections between — a pattern he refers to as “polarized networks”. Milan’s pattern was, instead, “polycentric”, in which associations and interest groups form a dense network and are linked to one another through many horizontal ties. In polycentric regions, he argues, frequent communication and the larger number of intermediaries mute conflict and keep lines of communications open. In such a structure, trust is facilitated, whereas the absence of intermediaries in Turin aborted attempts by moderates on both sides to reach compromise. The intermediaries humanize the other side by familiarity with it, and provide a line of communication for tentative discussions. In their absence, as in Turin, to express a sentiment of compromise toward the other side looks implausible and even treasonous. Such overtures would falter in any case, since there would be no obvious known and trusted interlocutors to receive them. So the structural situation creates cognitive and normative pressures which reinforce the separation and make conflict more likely.
In his analysis of the textile industry, Locke uses these distinctions to understand why the widely heralded success of small-firm networks in Italy in fact seems subject to sharp regional variations, increasingly failing in some areas while flying high in others. He suggests that whether such a form works is not an abstract matter, but depends on its compatibility with the local social and political networks. In particular, the rather polarized and hierarchical networks of Prato turned out to be much less fertile ground in the longer run for this form than those of polycentric areas such as Biella.
Here we veer into the territory of “social capital”, but the puzzle from the point of view that all these cities had a rich associational life, supposedly the progenitor of the norms, networks and trust that compose this capital. The difference was that Turin’s and Prato’s associations were structured vertically, with few ties across to other types of association, but with further vertical ties reaching out of the region to national parties or other organizations; Milan or Biella, by contrast, were richer in horizontal ties, of the sort which muted conflict in one case and facilitated the myriad details of inter-firm cooperation in the other. So it is not just the density of associational life that matters for economic (or political) outcomes, but the structure of its ties as also emphasized, in the theory of social capital.
These points link to an older tradition of thought that might be called “neo-Tocquevillian”, which emphasizes the importance for community, democracy and other political and economic outcomes, of associations and “cross-cutting ties”. Multiple and politically inconsistent affiliations, loyalties and stimuli reduce the emotion and aggressiveness involved in political choice. For example, in contemporary Germany, a working-class Catholic, pulled in two directions [i.e., toward his class and toward his religion], will most probably vote Christian-Democratic, but is much more tolerant of the Social Democrats than the average middle-class Catholic. …the chances for stable democracy are enhanced to the extent that groups and individuals have a number of cross-cutting, politically relevant affiliations.
More should be said in comparison of these older and newer theoretical traditions, but for now a couple of points seem interesting. One is that while the midcentury literature on “cross-cutting ties”, emerging as it did from a structural-functional view, stressed their role in conflict reduction, the existence of cross-cut, which I would characterize as some level of coupling among discrete networks or institutions, also provides channels through which a strategic actor may leverage weak attachments across segments so as to assemble resources into a larger social entity. If that entity is a political structure, we might challenge the idea that such a pattern enhances democracy, since political entrepreneurs might find this the most fertile ground on which to assemble empires or other autocratic systems; if the larger entity is an economic organization, such as a business group, conglomerate or strategic alliance, then we are talking about the organization of economic influence, such as that possessed by the Schumpeterian entrepreneur. Here we might think of Alfred Sloan pulling together the bits and pieces assembled earlier but only lightly coupled by William Durant, into General Motors.
Thus, we may distinguish three kinds of structures and corresponding potentials: the highly decoupled structure, without crosscutting ties, might be more prone to conflict when interests collide, but less likely to ever be pulled together into a social phenomenon of larger scale. The weakly coupled structure may lead to more consensual outcomes when conflict arises, but in the presence of an active entrepreneur may most lend itself to the amassing of power or influence over a large social entity. The highly coupled structure has, in effect, less structure. It may be the most amenable to a high level of cooperation, but even less likely than the first type to ever be highly coordinated from a center.
This rough typology has the advantage of stressing structure yet leaving an important role for agency. I take the structures and their connectedness as given, but this can only be for convenience of exposition. Certainly one of the most interesting issues is where these patterns originate. One of the most problematic aspects of early social capital formulations was the idea that current political outcomes are determined by the communal patterns of eight hundred years earlier. But to overcome this requires some focused historical argument about what determines network structures, and to what extent they may be altered by strategic actors who understand how to assemble resources.
