Communication and production in the family farm: Theory and evidence from rural Togo

We test the existence of cheap talk between husbands and wives working together in a family business. Our setting is the farm household. We designed an experiment, contextualized as an input allocation game, in which the husband chooses the amount of resources to invest on his own plot and on his wife's land. The return from the land managed by the wife is higher. We experimentally vary whether the returns from the wife's plot are communicated to the husband (i) by the experimenter, (ii) by the wife herself, and (iii) by the wife herself but with the possibility for the husband to verify the accuracy of the information from the experimenter after he makes his allocation decision. Male producers allocate too few inputs to their wife's plot across all experimental conditions. We rationalize these ndings in a setting with limited enforcement of marital agreements and derive additional predictions. First, allocative ineciencies in production are worse when women hold private information compared to the full information treatment. This e ect is stronger for households for which resource-sharing under full information is lower. Second, communication between spouses can only compensate for damages from private information when the information is veri able ex post.