Monday, March 06, 2006

on Social Dilemmas...

Queries and thoughts regarding this paper:
Kollock, P. (1998) Social Dilemmas: An Anatomy of Cooperation, Annual Review of Sociology 24: 183-214

Summary of the paper
Social dilemmas are situations where individuals don’t cooperate due to the immediate benefit of defection or the lack of immediate benefit of cooperation. Such individually reasonable behavior, or dominate strategy, leads to a “deficient equilibrium” in which everyone is worse off.
This paper reviews several models of social dilemmas and possible solutions for them. The Prisoner’s Dilemma has two variations: Assurance Game has two equilibria, mutual cooperation and mutual defection, based on the trustfulness. Chicken Game has another two equilibria, unilateral defection and unilateral cooperation. In both types of games, one’s best outcome is determined by the partner’s choice. The N-person dilemmas can be distinguished by the social fence (immediate cost) and social trap (immediate benefit), which features are captured by the public goods game and the tragedy of the commons, respectively.
Three types of possible solutions for the social dilemmas are studied: Motivational solutions assume that individuals care about their partners’ outcome, and this can be influenced by personal traits, communication, group identity and competition. Strategic solutions assume egoistic individuals may cooperate based on conditions such as the expectations of reciprocity, available information, choice of partners, grim triggers, and so on. Structural solutions resolve the dilemmas by introducing external conditions such as reputation, payoff, perception of efficacy, authorities and sanctions.

Queries and thoughts
This paper presents a clear, structural analysis of cooperation in terms of models and solutions. However, there are some important issues that haven’t been discussed:
1. Interaction among individuals (and context of cooperation): Individuals can contribute to the group at variant degree and in different ways. Cooperation is not just a binary option but a type of interaction that needs to be situated in its context. It is more important to understand how an individual cooperates with others than whether or not s/he cooperates. For example, in a learning group, a student who only asks and doesn’t answer questions seems to take more advantages from others, but a student who helps others can also benefit by refining and representing his understanding about a problem. Discovering the meaning of interaction can also encourage individuals’ cooperation.
2. Structure and components of cooperation: A cooperative group can be composed by homogeneous or heterogeneous members. Usually people avail themselves greatly from loose, changeable cooperative relationship. For example, we often learn more when joining a new or diverse group. Selecting an appropriate structure of cooperation to maximize self-benefits often plays a key role in individual rationality of cooperation.
3. Purpose of cooperation: I believe that the benefit of group can be maximized only when the individuals’ needs are satisfied. Any strategy that can boost cooperation but results in some individuals’ loss should be considered carefully. Otherwise, the cooperation would not be stable. For example, sanction against defectors can lead to a distrust society where a valuable cooperative relationship is difficult to establish or maintain. Therefore, a solution to the social dilemmas is effective only if it takes individuals’ benefit and loss into account.

1 comment:

Anonymous said...

"Cooperation is not just a binary option but a type of interaction that needs to be situated in its context. It is more important to understand how an individual cooperates with others than whether or not s/he cooperates"

That is true, and experiments show different degrees of cooperation.

"A cooperative group can be composed by homogeneous or heterogeneous members. Usually people avail themselves greatly from loose, changeable cooperative relationship."

This is an interesting observation. Most of the experiments are done with rather homogenous groups (undergraduate students).

"...the benefit of group can be maximized only when the individuals' needs are satisfied. Any strategy that can boost cooperation but results in some individuals' loss should be considered carefully. Otherwise, the cooperation would not be stable."

A lot of recent experiments focus on this issue, especially the importance of fairness. They also find that option of sanctioning can lead in some situations lead to lower cooperation, as you predicted.