How Institutions Shape Preferences: Experimental Evidence from a Large-Scale Property Rights Reform Implemented as Randomized Control-Trial

textabstractI investigate how a major institutional reform of property rights over land affects cooperation and trust preferences in a society where agricultural land is the main households' asset. The reform took place in hundreds of West African villages in Benin and consists in registering customary tenure rights over agricultural land that are traditionally characterized by collective property and informal possession. With the reform, registered plots acquire a new legal status akin to private ownership, making it possible to claim property in court and sell or use them as collateral. Identification capitalizes on the randomized control-trial implementation of the reform that used a public lottery to select villages who had the reform implemented. Those villages not selected as of today maintain the customary land tenure. As of today, this is the only example of large scale land rights reform implemented as a randomized control-trial. Seven years after the reform implementation, I performed two studies conducting lab-in-the-field experiments to collect data on cooperation and trust choices from a total of 546 participants in 32 villages. The results of the first study, conducted in a densely populated province where villagers are richer, more educated and live closer to markets and paved roads compared to other country's rural areas, show that the formalization of land rights significantly increases participants' contribution in a public good game (40%) and trustor's transfer in a trust game (35%). The second study replicates the first one but it is conducted in a different rural area that experienced the same reform and shares the same set of formal institutions as the province in the first study, but that is characterized by markedly different socio-economic characteristics -- the lowest population density, income, education levels and highest distance from paved roads in the country. In contrast with the first study, for participants belonging to the least market integrated and socio-economic developed villages, the reform determines a reduction of cooperation and it has no effects on trust levels