License Buyback Programs in Commercial Fisheries: An Application to the Shrimp Fishery in the Gulf of Mexico

Abstract

This dissertation provides a thorough analysis of the costs associated with, and efficacy of, sequential license buyback auctions. I use data from the Texas Shrimp License Buyback Program - a sequential license buyback auction - to estimate the effects of a repeated game set-up on bidding behavior. I develop a dynamic econometric model to estimate parameters of the fisherman's optimal bidding function in this auction. The model incorporates the learning that occurs when an agent is able to submit bids for the same asset in multiple rounds and is capable of distinguishing between the fisherman's underlying valuation of the license and the speculative premium induced by the sequential auction. I show that bidders in the sequential auction do in fact inflate bids above their true license valuation in response to the sequential auction format. The results from our econometric model are used to simulate a hypothetical buyback program for capacity reduction in the offshore shrimp fishery in the Gulf of Mexico using two competing auction formats: the sequential auction and the one-time auction. I use this simulation analysis to compare the cost and effectiveness of sequential license buyback program relative to one-time license buyback programs. I find that one-time auctions, although they impose a greater up-front cost on the management agency - are capable of retiring more fishing effort per dollar spent then sequential license buyback programs. In particular, I find one-time license buyback auctions to be more cost effective than sequential ones because they remove the possibility for fishermen to learn about the agency's willingness to pay function and use this information to extract sale prices in excess of the true license value.