Household risk versus individual risk: an experimental study
Astrid Hopfensitz  1@  , Francois Cochard  2@  , Hélène Couprie  3@  , Jiakun Zheng  1@  
1 : Toulouse School of Economics  (TSE)  -  Website
Toulouse School of Economics
Manufacture de Tabacs, 21 allées de Brienne 31000 Toulouse -  France
2 : Centre de REcherches sur les Stratégies Economiques  (CRESE)  -  Website
Université de Franche-Comté
IUT 30 Avenue de l'observatoire 25009 Besançon Cedex -  France
3 : Théorie économique, modélisation et applications  (THEMA)
CNRS : UMR8184, Université de Cergy Pontoise
33, boulevard du Port 95011 Cergy-Pontoise Cedex -  France

Many decisions taken by households involve risk and are taken jointly by multiple members of the family. There is increasing evidence that joint decisions can not be estimated as a convex combination of individual preferences. Moreover it seems that different members of the household have different weights on the overall outcome.

We experimentally study a joint decision task taken by 100 established couples (from Toulouse, France). In a first part couples had to chose between two options, that varied overall household risk while keeping individual risk and expected payouts constant. In a second part one of the two spouses (either the man or the woman) was ‘insured' against risk in one of the two options. We investigate behavior in these two parts and relate it to socioeconomic variables and individual risk preferences.

We find that overall couples show a preference for the option that presents lower risk on the couple level (even if this implies accepting larger individual risk). Secondly we investigate the case where one partner is insured: couples were men are insured stay with similar preferences, which can be interpreted as women not being negatively affected by holding solely a risk. However couples where women are insured show a switch in preferences to an option that increases risk for the household. Thus men and/or couples seem to be averse to the idea of solely holding risk in the household.