Toward Behavioral Modeling of Alaska Groundfish Fisheries: A Discrete Choice Approach to Bering Sea/Aleutian Island Trawl Fisheries

Article excerpt

I. INTRODUCTION

Management of modern multispecies fisheries subject to binding catch quotas raises a number of issues that are of interest both from the standpoint of practical policy formation and as conceptual challenges. The presence of highly versatile and mobile fishing fleets, harvest technologies that are imperfectly selective, and a limited understanding of how joint catch patterns vary by time and area all combine to make the simultaneous achievement of multiple aggregate harvest quotas a tall order for fisheries managers to fill.

The effects of regulations on fishermen's choice sets and subsequent fleetwide response regarding species targeted, area fished, and resulting changes in catches are also not currently well understood in fisheries management. It is not uncommon for managers to regulate catch in one fishery and be surprised by unanticipated consequences in related fisheries.

Better knowledge of what broadly can be termed "behavioral response" in fisheries is needed for at least two reasons. For inseason quota management, better prediction of joint harvest rates through time reduces the likelihood of premature closure; of a fishery due to quota attainment in a second, technologically related fishery, and the associated social welfare losses. This occurs because of limited knowledge of technological interdependencies among species within fisheries and of substitution responses to regulations by fishermen. Frequently, regulatory responses are single-species in nature, such as season or area closures in response to catches of a particular species, which can exacerbate the multispecies quota management problem.

A better knowledge of fleet response to regulation is also needed for policy evaluation, i.e., for helping to decide which regulations to enact in the first place. Regional fishery management councils are mandated by the Magnuson-Stevens Fishery Conservation and Management Act (PL 94-265) and by Executive Order to promote "optimum use" and consider, to the extent practicable, both economic efficiency and the minimization of bycatch in promulgating regulations (National Marine Fisheries Service, 1996). As most American fisheries are now considered to be fully or overutilized, unintended catches from one fishery can impose significant opportunity costs on other segments of the fishing fleet and society by reducing the catches available in other fisheries. Without a reasonable understanding of how fleets respond to regulation, it is difficult if not impossible to predict the welfare gains or losses associated with policy alternatives.

Recent literature has begun to fill some knowledge gaps. Several papers have estimated the production technologies for multispecies fisheries. Squires (1987a, 1987b) was among the first to apply modern dual methods in an application to New England fisheries. More recent examples include Dupont (1991), Kirkley and Strand (1988), Thunberg et al. (1995), and Campbell and Nicholl (1995). In the North Pacific, Larson et al. (1998) identified the technology for the Bering Sea/Aleutian Islands (BSAI) midwater pollock fishery by estimating the associated quasirent share equations.

In characterizing fleet response to regulation, Wilen (1985) was among the first to develop models of the interaction of regulator and regulated in fisheries, and applications have been made to management of the Pacific halibut fishery (Homans and Wilen, 1997) and to the West Coast sablefish fishery (Squires and Kirkley, 1991), among others. Discrete choice models of fishery participation have been implemented by Bockstael and Opaluch (1983) for the New England groundfish trawl fishery and by Evans (1997) for the California troll salmon fishery.

The purpose of this paper is to develop and present a model of weekly fishery choice in the BSAI trawl fishery conducted by catcher processor and mothership-catcher vessel (CPM) operations. It is one of several steps that must be taken to build practicable models of fleet behavior to evaluate the consequences of regulatory actions. …