Search results

AbstractWe provide new evidence about a positive correlation between the own amount sent and the own amount returned in the investment game. Our analysis relies on the experimental data collected under the strategy method. While the percentage returned is independent of the amount received for most of our subjects, it is strongly correlated to their amount sent as a trustor. Our analysis is based on a two-way classification of subjects: according to their trusting type and according to their reciprocal type. We show the existence of a strong positive relation between trusting types and reciprocal types within subjects.

We test a repeated principal–agent relationship with hidden actions in the laboratory. At each period, the principal offers payments to the agent in order to induce him to provide the high effort level. The agent can accept or reject the offer; if he accepts it, he can choose the high or the low effort level. We observe that offered payments are higher than predicted by the subgame‐perfect equilibrium, and higher in the partners' than in the strangers' design. Furthermore, agents strategically reject inequitable offers in early rounds in order to induce principals to increase payments in the subsequent rounds.

Résumé L'effet irréversibilité prédit qu'un agent qui se comporte selon l'hypothèse de maximisation de l'utilité espérée, et qui anticipe un supplément d'information préa­lablement à ses choix futurs, adoptera une position « plus flexible » (ou « moins irréversible ») à la période présente. Dans cet article nous présenterons les résul­tats d'une étude expérimentale dont l'objectif est de mettre en évidence l'effet irré­versibilité sur un plan empirique, sans préjuger du type de comportement adopté par les agents. Les résultats obtenus sont conformes aux prédictions théoriques, et suggèrent que l'effet-irréversibilité est compatible avec des comportements qui ne respectent pas nécessairement l'hypothèse de maximisation de l'utilité espérée.

Using a laboratory experiment, we behaviourally study the impact of a sudden increase in the common-pool size on within-group conflict, i.e., the paradox of the plenty. We also consider the potential role of governance in avoiding this paradox. In the first stage, a randomly-chosen leader of the group determines how much of the common-pool resource to protect from second-stage conflict. In the next stage, each group member allocates his private endowment between working or fighting for a share of the unprotected resource. We consider two treatments: anarchy (consisting of the second stage only) and with a leader deciding in the first stage. We find that the existence of institutions is not always better than anarchy. This is aggravated when the resource size is higher. Group conflict (income) decreases (increases) only when leaders chose the strongest resource protection. When leaders are malevolent, i.e., they chose weak resource protection, outcomes are worse than when institutions are absent.

We report the results of an experiment investigating the provision of a step-level collective good. This experiment compares the behavior of subjects in a public good game with the behavior of subjects in a club good game. In the club good game, players who do not contribute the membership fee are excluded from the consumption of the collective good. The introduction of a small membership fee has surprisingly strong effects: it increases the provision of the collective good, delays provision failures and increases the number of contributors. The experiment also reveals the limits of the introduction of a small fee. At a high threshold level, the membership fee no longer ensures successful provision. Adapted from the source document.