Doctor of Philosophy (PhD)


Agricultural Economics

Document Type



The overall goal of this research is to empirically analyze shrimp fishermen behavior to help improve the management of the Gulf of Mexico shrimp fleet. Given that optimal management requires consideration of more than the net benefits derived from shrimp harvesting, this research also seeks to provide an empirical framework that would allow future investigators to measure benefits lost through bycatch-related management actions. This paper expands on previous fishing behavior literature by focusing on two of the most important short-run decisions confronting Gulf of Mexico shrimp fishermen (where to fish and how long to fish). A better understanding of these factors can provide useful information to policy makers in designing and implementing more effective policies. This study uses panel data for up to 15 years, which is a combination of the Coast Guard Vessel Operating Unit File and the Shrimp Landing File from National Marine Fisheries Service. In the location choice analyses, the U.S. Gulf of Mexico is divided into three areas: FL, LAM, and TX. For each area a conditional logit and mixed logit based on Random Utility Model are run to analyze the influence of fishermen’s past choice decision on current choice (state dependence), and the fishermen’s difference in preferences (preference heterogeneity). The results show that past experience does affect current decision, but the influence dies out fast. In addition, fishers are different in their preference in many aspects. Also, it seems that fishermen’s risk attitudes can change over the years. Their tolerance towards congestion exhibits changes over time too. As for their trip length decision, it seems that diesel price is negatively related to the length of days fished, so is the price difference between large and medium sized shrimp. Further, there seems to be a pattern that the trip length is increasing over the years. The incorporation of unobserved heterogeneity into the location choice and duration models corrects the potential biasedness in estimates and improves the goodness-of-fit considerably, aside from provides intuitive economic interpretations.



Document Availability at the Time of Submission

Release the entire work immediately for access worldwide.

Committee Chair

Walter R. Keithly