We simulate in a randomised lab experiment the effect of Cash Mobs on consumers’ behaviour in an original variant of the multiplayer Prisoner’s dilemma called Vote-with-theWallet Game (VWG). The effect is modelled in a sequential game with/without an environmental frame in which a subset of players (cash-mobbers) is given the opportunity to reveal publicly (in aggregate without disclosing individual identities) their cooperation decision. We find that the treatment has a positive gross effect, that is, the share of cooperators is significantly higher in treated sessions and this is mainly due to the higher share of cooperators among cash-mobbers. Our results suggest that cash mobs-like mechanisms can help to solve social dilemmas with entirely private solutions (not based on punishment but on positive action) without costs for government budgets.
Keywords: vote with the wallet, prisoner’s dilemma, randomised experiment
JEL Classification: C72 (Noncooperative games), C73 (Repeated games), C91 (Laboratory, Individual behavior), M14 (Corporate culture, Social responsibility).