Corner solutions in empirical acreage choice models: an andogeneous switching regime approach with regime fixed cost

Corner solution problems are pervasive in micro-econometric acreage choice models because farmers rarely produce the same crop set in a considered sample. Acreage choice models suitably accounting for corner solution need to be specified as Endogenous Regime Switching (ERS) models. Micro-econometric ERS models are however rarely used in practice because their estimation difficulty quickly grows with the dimension of the considered system. Their functional form is generally quite involved and their congruent likelihood functions need to be integrated using simulation methods in most case of interest. We present here an ERS model specifically designed for empirically modeling acreage choices with corner solutions. This model is theoretically consistent with acreage choices based on the maximization of a profit function with non-negativity constraints and a total land use constraint. It can be combined with yield supply and variable input demand functions. Furthermore, the model accounts for regime fixed costs which represent crop specific marketing and management costs. To our knowledge, this is a unique feature for an ERS model accounting for non-negativity constraints. The proposed ERS model defines a Nested MultiNomial Logit (NMNL) acreage choice model for each potential production regime. The regime choice is based on a standard discrete choice model according to which farmers choose the crop subset they produce by comparing the different regime profit levels. The structure of the model and the functional form of its likelihood function makes the Simulated Expectation-Maximisation algorithm especially suitable for maximizing the sample likelihood function. The empirical tractability of the model is illustrated by the estimation of a five crop production choice model for a sample of French grain crop producers