An Experiment on Deception, Reputation and Trust∗

We report results from an experiment on a repeated sender/receiver game with twenty periods in which one period has higher weight. The sender communicates about the realized state in each period, the receiver takes an action matching his belief about the state, and then learns whether the sender lied. Receivers are matched either with malevolent (human) senders who prefer the agents to take wrong decisions or with benevolent (machine) senders who always tell the truth. Our findings do not support the predictions of the Sequential Equilibrium. The deceptive tactic in which malevolent senders tell the truth up to the key period and then lie at the key period is used much more often than it should and it brings higher expected payoff. We suggest that our data are well organized by the analogy-based sequential equilibrium (ABSE) in which three quarters of subjects reason coarsely when making inferences and forming expectations about others’ behaviors. ∗We thank Maxim Frolov for assistance on the experimental design, Guillaume Frechette, Guillaume Hollard, Frederic Koessler, Dov Samet, Jean Marc Tallon, and the participants of the Neuroeconomic Workshop, the Extensive form Games in the Lab Workshop, LSE-UCL workshop, LSE behavioral economics seminar, The first Socrates workshop, the ASFEE conference, the IHP, Dauphine, Cerge-EI, HEC-Polytechnique, Paris 1, Technion seminars’ participants for helpful comments. Jehiel thanks the European Research Council for funding and Ettinger the Governance and Regulation Chair for its support. †Université Paris-Dauphine, PSL Research University, LEDa and CEREMADE, 75016 Paris, France and CIRANO ; [email protected] ‡PSE, 48 boulevard Jourdan, 75014 Paris, France and University College London ; [email protected]