Essays on social insurance and allocation of resources

Abstract

The thesis is composed of four stand-alone essays analyzing economic problems pertinent to social insurance and allocation of resources in three applications to development, education and labor. The first essay investigates co-operative patterns of farmers in rural India engaging into informal insurance and public irrigation provision. I demonstrate theoretically, empirically and quantitatively that these two margins of co-operation reinforce each other, if the irrigational infrastructure is managed by local societies. On the other hand, management by external government agencies is associated with excessive crowding-out of informal insurance. The second essay investigates constrained efficiency of a model with educational investments subject to uninsurable dropout risk, moral hazard and an endogenous college wage premium. I show that the laissez-faire equilibrium is constrained inefficient and is characterized by under-education. To this end, I show that an optimally designed student loan program with graduation-contingent repayment rates can attain the allocative efficiency of second best. In the third essay, I show in a simple model of lumpy educational investments that subjective pessimism over returns to education can be self-confirmed in equilibrium. This leads to two empirical implications: (i) both the degree of human capital concentration and the degree of educational investments misallocation may be increasing in the rigidity of the education system’s design; and (ii) commonly pursued methods may not identify the true underlying skill distributions. The fourth essay uses a quantitative model of labor search with unemployment insurance and voluntary quits to study welfare consequences of a policyreform giving entitlement to workers quitting their jobs in the US. Structural results show that pursuing a generous entitlement policy for quitters may allow for significant welfare gains through improved insurance and allocation of workers. Moreover, I employ the assumption of monetary search costs and show that it can explain the empirically documented unemployed search behavior.