Viewpoint: On the generalizability of lab behaviour to the field


  • Both authors are members of NBER. Thanks to seminar participants at the 2005 International Meetings of the ESA for useful suggestions. Excellent suggestions from James Andreoni, Nicholas Bardsley, Gary Becker, Gary Charness, David Cooper, Dan Gilbert, Uri Gneezy, Glenn Harrison, Reid Hastie, John Kagel, Dean Karlan, Dan Levin, Jayson Lusk, Ted McConnell, Kevin Murphy, Andreas Ortmann, Charles Plott, Jesse Shapiro, Andrei Shleifer, Robert Slonim, and Richard Thaler greatly improved the study. Colin Camerer, Ernst Fehr, and Alvin Roth provided detailed comments and made suggestions that have resulted in many improvements, although not as many as they would have wished. Seminar participants at Brigham Young University, University of Chicago, Laval University, McMaster University, Oberlin College, and the University of Nevada-Reno also provided useful feedback on bits and pieces of this research. Financial support came from the National Science Foundation and the Sherman Shapiro Research Fund. This study was part of our original working paper titled ‘What do laboratory experiments tell us about the real world?’ Email:;


Abstract.  We can think of no question more fundamental to experimental economics than understanding whether, and under what circumstances, laboratory results generalize to naturally occurring environments. In this paper, we extend Levitt and List (2006) to the class of games in which financial payoffs and ‘doing the right thing’ are not necessarily in conflict. We argue that behaviour is crucially linked to not only the preferences of people, but also the properties of the situation. By doing so, we are able to provide a road map of the psychological and economic properties of people and situations that might interfere with generalizability of laboratory result from a broad class of games.


Il n'y a pas de question plus fondamentale en économie expérimentale que de savoir si et en quelles circonstances les résultats de laboratoire peuvent être généralisés à ce qui se passe sur le terrain. Dans ce texte, on développe les résultats de Levitt et List (2006) pour les appliquer à une classe de jeux dans lesquels les résultats financiers et ≪faire la bonne chose≫ ne sont pas nécessairement des choix conflictuels. Le comportement n'est pas une simple question de préférences des gens mais aussi un écho des propriétés de l'environnement. Il devient alors possible de proposer une image des propriétés psychologiques et économiques des gens et des situations qui pourraient empêcher qu'on généralise les résultats de laboratoire à une classe plus vaste de jeux.