Open-Access Fishery Models: Relaxing a Constraint and Removing an Econometric Obstacle

Over the past 30 years, a widely accepted model of "open-access" fisheries has been developed, yet empirical tests of the standard model have been relatively few. One difficulty is that fish stocks, the levels of which affect the rate of catch, are not directly observable. Simplifying assumptions are generally required, such as the assumption that catchability does not changeover time. Estimation on the basis of the standard model also raises difficultiesin specification, if contemporaneous correlation of the error term with one ofthe regressors is to be avoided. This paper describes an algorithm that imposes a less restrictive pattern (than constancy) on catchability, yet does so in an econometrically acceptable fashion. It also reports on an application of this algorithm to the Flemish Cap groundfishery over the period from 1971 to 1985.