Nos tutelles





Accueil > Actualités

Working papers

publié le , mis à jour le

Working papers


Can public and private sanctions discipline politicians ? Evidence from the French Parliament

Maxime Le Bihan, Benjamin Monnery , Working paper GATE 2018-08
This paper investigates the effects of sanctions on the behavior of deputies in the French National Assembly. In 2009, the Assembly introduced small monetary sanctions to prevent absenteeism in weekly standing committee meetings (held on wednesday mornings). Using a rich monthly panel dataset of parliamentary activity for the full 2007-2012 legislature, we study the reactions of deputies to (i) the mere eligibility to new sanctions, (ii) the actual experience of a salary cut, and (iii) the public exposure of sanctioned deputies in the media. First, our diff-in-diff estimates show very large disciplining effects of the policy in terms of committee attendance, and positive or null effects on other dimensions of parliamentary work. Second, exploiting the timing of exposure to actual sanctions (monthly salary cuts versus staggered media exposure), we find that deputies strongly increase their committee attendance both after the private experience of sanctions and after their public exposure. These results suggest that monetary and reputational incentives can effectively discipline politicians without crowding out intrinsic motivation.

Embezzlement and Guilt Aversion

Giuseppe Attanasi, Claire Rimbaud, Marie Claire Villeval , Working paper GATE 2018-07
Donors usually need intermediaries to transfer their donations to recipients. A risk is that donations can be embezzled before they reach the recipients. Using psychological game theory, we design a novel threeplayer Embezzlement Mini-Game to study whether intermediaries suffer from guilt aversion and whether guilt aversion toward the recipient is stronger than toward the donor. Testing the predictions of the model in a laboratory experiment, we show that the proportion of guilt-averse intermediaries is the same irrespective of the direction of the guilt. However, structural estimates indicate that the effect of guilt on behaviour is higher when the guilt is directed toward the recipient.