Dynamic Discrete Choice Modeling: Monte Carlo Analysis

Recent work on spatial models of commercial fishing has provided insights into how spatial regulatory policies (i.e. Marine Protected Areas) are likely to alter the fishing location choices of commercial fishermen and the efficiency of these policies. The applied studies have spanned a diverse range of fisheries, from sedentary to highly migratory species. This literature has largely ignored the inter-temporal aspects of commercial fishing site choice at the cruise level. Therefore, these models depict fishermen as if they are ignoring how a location choice on the first day of a cruise may have potentially important consequences for the rest of the cruise. For many fisheries, particularly highly migratory ones, fishermen might choose a dynamically optimal cruise trajectory rather than myopic day-by-day strategies. An econometric model that ignores the inter-temporal aspects of location choice will likely lead to erroneous conclusions regarding a vesselÕs response to spatial regulatory policies. A dynamic discrete choice model is developed herein that utilizes the same information conventionally used in static models but is entrenched in the principals of dynamic optimization (BellmanÕs principle). Using Monte Carlo analysis, we evaluate the relative performance of this estimator as compared to the conventional static model for a variety of conditions that mimic different fishery types.