Aggregation dynamics

The social world starts with social individuals. So how do we get more complex social outcomes out of the actions and thoughts of independent individuals? How do the actions and thoughts of individuals aggregate into larger social happenings? How did the various religious, political, and relational attitudes of rural Kenyans aggregate to widespread ethnic violence a few years ago? What sorts of conditions lead to interactions that bring about unexpected outcomes? And, of course, how do larger social happenings impinge upon individuals, leading to characteristic kinds of socialized behavior?

These are questions I've usually addressed from the other angle -- the "dis-aggregation" angle. I've asked how large social entities and processes can be disaggregated into actions and relationships pertaining to socially situated individuals. This is what we're aiming at when we ask for microfoundations for things like state power or the changing impact of Islam. Here, though, let's look at the upward-facing processes.

Here are some examples of the kinds of happenings and situations that we might be interested in:

• collective social behaviors: riots, uprisings, boycotts, protests, panics, runs on the bank
• coordinated actions by differentiated but organized groups: design of a new model vehicle at General Motors, movement of a tank squadron through a battle zone, actions by squads of fire fighters in a large forest fire
• shifts of valuations and attitudes: styles, slang, inter-ethnic attitudes, diffusion of religious beliefs, dissemination of rumors through a population
• preservation of stable organizations and structures: maintenance of the rules governing shop-floor behavior in a factory, sustaining command authority within a military organization, keeping a sustainable level of parent involvement in a cooperative childcare center, keeping a reasonably high level of taxpayer compliance within a mass society

How do the reasonings, thoughts, and choices of socially situated individuals aggregate and link together in ways that constitute these higher-level phenomena?

Some social outcomes are essentially agglomerative. A riot sometimes occurs when a small group of people have a grievance and a provocation; a larger group share sympathy with the original rioters; others see the behavior and join for their own reasons; and in a few hours thousands of people are breaking windows, setting cars on fire, and blocking the streets.

Second, some outcomes are the result of strategic behavior by a number of individuals. Competition leads agents to tailor their actions in such a way as to take advantage of the expected behavior of others; cooperation leads individuals to attempt to coordinate their behavior so as to bring about an outcome that the group prefers. These are the kinds of aggregative processes that Thomas Schelling describes in Micromotives and Macrobehavior; and, of course, some of the outcomes of both competitive and cooperative behavior are paradoxical with respect to the interests and intentions of the individuals who take part in them.

Third, some outcomes come about because individuals have constituted themselves as collective agents; they have come to identify themselves as members of a group and they behave deliberately out of consideration for what other members in the group will do. Their behavior is oriented to the behavior and intentions of the others in the group. They specifically strive to coordinate their actions with others through a shared set of social goals, identities, and attitudes. Raimo Tuomela explores this perspective in The Philosophy of Sociality: The Shared Point of View.

Other outcomes are largely structured by imperatives (rules or commands) conveyed through delineated social relationships from a group of decision makers to a group of agents. This is the key subject matter of the study of organizations and institutions. The Midwest account manager for a national bank issues instructions to loan officers throughout the region to do this or that; communications are sent out, supervisors are instructed to observe compliance, and loan officers change their behavior. This process too works through the incentives and purposes of a set of dispersed agents; but in this case there is a substantial degree of top-down coordination that is secured through communication channels and a supervisory structure.

The dynamics of competition and cooperation are well understood; the ideas of a prisoners' dilemma, tragedy of the commons, and race to the bottom capture some of the logic of independent individuals each seeking to maximize their own interests (Mancur Olson, The Logic of Collective Action: Public Goods and the Theory of Groups, Thomas Schelling, Micromotives and Macrobehavior). But cooperation too has its logic; even when a group of individuals are seeking to find a common strategy to further their group interests, it is easy to spot the organizational obstacles that stand in the way: lack of consensus, imperfect communication, etc. These are aggregation dynamics that derive from the logic of individual agency.

And the dynamics of organizations and systems of rules likewise have some obvious characteristics. Costs of enforcement mean that an institution's rules are only imperfectly enforced. Conflicts of interests among powerful actors lead to inefficient behavior within the organization (e.g. ministry of steel and ministry of railroads). Well known organizational shortfalls like the principal-agent problem, imperfect and costly communications transmission, and costly information all lead organizations to fail in somewhat predictable ways. Robert Bates (Markets and States in Tropical Africa: The Political Basis of Agricultural Policies) and Charles Perrow (Normal Accidents: Living with High-Risk Technologies, The Next Catastrophe: Reducing Our Vulnerabilities to Natural, Industrial, and Terrorist Disasters) each capture some of the imperfections of organizations in their works. These are aggregation dynamics that derive from more structural features of systems of rules, roles, and actors.

Let's look at an example of an aggregate-level phenomenon a bit more closely. Take style change over space and time. There are several different types of actors: street-level people who wear clothes with some degree of intentionality (want to look cool); street-level innovators (the first person to wear droopy jeans); clothing designers who want to influence the market for clothes through innovations that will be adopted by the public; "cool" finders who are looking for emerging innovations in hot places; marketers who specifically design strategies for proliferating new designs at Banana Republic or Abercrombie and Fitch. In this space of actors, consumers, and creators it sometimes happens that a new fashion emerges and quickly spreads to a larger population; how does this occur? What are the mechanisms of diffusion and adoption that lead to the success of a new look in denim jackets deriving ultimately from Seattle skateboarder culture? Stan Lieberson looks at some of these surprising dynamics in A Matter of Taste: How Names, Fashions, and Culture Change.

I won't try to catalogue here all the pathways through which individual actors aggregate into larger social happenings. But it seems clear that several factors are key to any such discussion: communication channels, organizations, and social networks.