Essays in expected utility : agricultural production under uncertainty
This thesis presents how uncertainty affects decision making, especially in terms of efficiency and productivity, and analyzes the primary sources of suboptimal behavior. Alternatives to the prominent Expected Utility framework are examined, with special focus on the State-Contingent approach, which is used to produce adequate measures and a state-decomposition of the Malmquist productivity index. On the empirical part, the non-parametric methodology of Data Envelopment Analysis (DEA) is applied on real data in agriculture, with special attention in the significance and proper use of distance functions. Moreover, a new approach in the calculation of distance functions is proposed, solving the issues of slacks and input-output differences. Finally, a theoretical model for decisions under uncertainty is proposed and compared against mainstream models of Expected Utility and leading behavioral models.