European Journal of Personality Eur. J. Pers. 15: S37±S52 2001) DOI: 10.1002/per.420 Individual Differences in Cooperation in a Circular Public Goods Game ROBERT KURZBAN 1 * and DANIEL HOUSER 2 1 Economic Science Laboratory, University of Arizona, USA 2 Department of Economics and Economic Science Laboratory, University of Arizona, USA Abstract Research using the public goods game to examine behaviour in the context of social dilemmas has repeatedly shown substantial individual differences in patterns of contributions to the public good. We present here a new method speci®cally designed to capture this heterogeneity in play and classify participants into broad categories or types. Players in groups of four made initial, simultaneous contributions to the public good. Subsequently, players were sequentially told the current aggregate contribution to the public good and allowed to change their decision based on this information. The game continued, with players updating their contribution decision until the game ended at an unknown point. By looking at the relationship between players' contributions and the aggregate value they observed, we were able to cleanly classify 82% of our players into three types: strong free riders 28%), conditional cooperators of reciprocators 29%), and strong cooperators 25%). We also found that scores on some of the personality dimensions we investigated self-monitoring, self-esteem, neuroticism, and conscientious- ness) correlated with player type. Finally, males were found to be more likely to be strong cooperators than females. Copyright # 2001 John Wiley & Sons, Ltd. INTRODUCTION Over the last three decades, individual differences have taken on greater importance in the context of group cooperation games. Whereas it used to be thought that players' `dispositional qualities' would have only negligible effects in these environments Pruitt and Kimmel, 1977), more recently, psychologists and economists have begun to take seriously the idea that individual differences among participants might be important in understanding and modelling behaviour in experimental games see e.g. Liebrand, 1984; Yamagishi, 1986). Indeed, Rapoport and Suleiman 1993) suggested that Received 20 November 2000 Copyright # 2001 John Wiley & Sons, Ltd. Accepted 27 April 2001 *Correspondence to: Robert Kurzban, Department of Anthropology, 341 Haines Hall, Box 951553, UCLA, Los Angeles, CA 90095-1553, USA. E-mail: rkurzban@ucla.edu Contract/grant sponsor: International Foundation for Research in Experimental Economics.