Why cooperate? An economic perspective is not enough.

Cooperation is usually explained from an economic perspective focused mainly on the tangible outcomes received by individuals that are also dependent on the behavior of others, with little reference to the actual behaviors used when cooperating. The potential consequences of social dimensions associated with cooperative behaviors are minimized in Skinnerian and game-theory models by means of anonymous subjects that behave individually while physically isolated in separate chambers. When cooperation and non-cooperation occur in the real world, however, they are often associated not only with different outcomes but also with different behaviors. Unlike non-cooperation, cooperative behaviors are usually intrinsically social, influenced by the presence and behaviors of familiar partners. Research is described that addresses whether the social dimensions of cooperative actions go beyond mere description of behaviors to also explain why cooperation occurs. One way to resolve the relative importance of economic and social factors for explanations of cooperation is to measure choice between the options of cooperation and non-cooperation. The economic perspective, linked to models derived from game theory, frames the question as a choice determined by differences in tangible outcomes such as food or money that, in evolutionary terms, are surrogates for gains in fitness. From a behavioral perspective, the choice between cooperation and non-cooperation is also determined by social dimensions associated only with cooperation. The influence of social cooperation on preference was examined by means of two rectangular chambers interconnected by a T-maze. In one chamber, pairs of laboratory rats were reinforced with saccharin solution for coordinating back-and-forth shuttling; in the second chamber, a single animal was reinforced for back-and-forth shuttling performed in isolation. With outcomes equalized between the two options, cooperation was preferred by the majority of subjects. Moreover, variation in the relative rate of reinforcement during cooperation was not a strong predictor of choice whereas the level of intra-pair coordination was positively related to preference. Implications of this result are discussed for both method and theory, including the hypothesis that the preference is influenced by intrinsic reinforcements evoked by cooperating. The consequences for evolutionary fitness would then arise not only from tangible outcomes but from the relationships that develop when cooperating even when immediate and tangible pay-offs are absent, insufficient or sub-optimal. The impact of cooperative relationships on fitness may therefore not occur immediately but in the future, and perhaps in another context, when they influence outcomes that have a significant impact on survival and reproduction.