Content area
Full Text
DOI: 10.1007/s00191-004-0212-1J Evol Econ (2004) 14: 407429Jeffrey P. Carpenter1, Peter Hans Matthews1, and Okomboli Ongonga21 Department of Economics, Middlebury College, Middlebury, VT 05753, USA(e-mail: {jpc,pmatthew}@middlebury.edu)2 Gradute School of Business, Stanford University, Stanford, CA 94305, USA(e-mail: [email protected])Abstract. Recently economists have become interested in why people who face
social dilemmas in the experimental lab use the seemingly incredible threat of
punishment to deter free riding. Three theories with evolutionary microfoundations have been developed to explain punishment. We survey these theories and
use behavioral data from surveys and experiments to show that the theory called
social reciprocity in which people punish norm violators indiscriminately explains
punishment best.Keywords: Social dilemma Punishment Norm Evolutionary game theory
ExperimentJEL Classication: C91, C92, D64, H411 IntroductionEconomists have become interested in analyzing, in the experimental lab, something known to eld researchers for quite a while, that people who face social
dilemmas (i.e. situations in which group and individual incentives are at odds)
sometimes control free riding locally by the use of social, economic, and/or physical sanctions.1 The existence of schemes by which people monitor each other and We thank Carolyn Craven, Corinna Noelke and two referees for comments, and Middlebury College for nancial assistance. In addition, Carpenter acknowledges the support of the National Science
Foundation (SES-CAREER 0092953).Correspondence to: J.P. Carpenter1 Economic punishment experiments include Fehr and Gaechter (2000), Bochet et al. (2003), Bowles
et al. (2001), Carpenter (2002b), Carpenter and Matthews (2002), Masclet et al. (2003), and Sefton et
al. (2000). Relevant eld research is summarized in Ostrom (1990) and Ostrom et al. (1994). A specic
example of eld research is Acheson (1988).c
Springer-Verlag 2004Why Punish? Social reciprocity
and the enforcement of prosocial norms[star]408 J.P. Carpenter et al.punish those who free ride is problematic for standard economic theory. Why?
First, in non-repeated interactions, any theory assuming that agents simply want to
maximize their material gain can not reconcile the cooperative behavior needed to
obtain socially efcient outcomes because free riders always do better. Free riding
when others contribute avoids the costs associated with contributing yet returns the
benets of cooperation and free riding when others free ride prevents one from being taken advantage of. Hence, no self-interested person would ever cooperate. The
same argument can be made for not punishing free riders because punishment,...