Intra-household efficiency: An experimental study from Ethiopia

A common feature of many intra-household models is the assumption of efficiency. This paper uses an experimental design to directly test this. Data from 1,200 married couples in Ethiopia were collected using treatments that vary initial endowments of spouses, final allocation rules and information in a voluntary contribution mechanism (VCM). Efficiency is decisively rejected in all treatments. Information improves efficiency only in some treatments, suggesting that the role of information is context dependent. Husbands' expectations of their wives' contributions are higher than their wives' actual contributions, and wives' expectations of their husbands' contributions are lower than their husbands' actual contributions. These systematic errors in expected and actual behaviour imply that the attainment of equilibrium as in a game theoretic framework is unlikely. Statistical tests indicate that instead of efficiency, other considerations are likely important. Overall, most of the empirical results cast doubt on models of the household that assume Pareto efficiency.