The Effects of Incomplete Information in Stochastic Common-Stock Harvesting Games

Abstract

Here the dynamic fishery harvesting game is generalized to a stochastic environment in order to examine the implications of incomplete and asymmetric information. The main emphasis is on a split stream version of the game: At the beginning of each harvest season the initial fish stock (or “recruitment”) divides into two streams, each one accessible to harvest by just one of the two competing fishing fleets. The fleets simultaneously harvest down their streams, achieving net seasonal payoffs for the catch. After harvest, the residual sub-stocks reunite to form the broodstock for the subsequent generation. The strength of this subsequent generation is determined by a specified “stock-recruitment relation,” and the cycle repeats. In this cyclic process, both natural environmental factors (stream-split proportions and stock-recruitment relation) and economic factors (harvest costs and benefits) will incorporate Markovian stochastic elements. At the beginning of each season, both fleets know the current recruitment and also have some (generally incomplete or delayed, and often asymmetric) knowledge of the current values of the stochastic elements. The knowledge structure of each specific game version is held in common by the competitors. In the dynamic game each fleet sets its harvest policy with the objective of maximizing the expected discounted sum of seasonal payoffs, and conditional on the extent of its current knowledge and of the anticipated policy of its competitor.

The implications of alternative knowledge structures are explored, through dynamic programming and simulation. Both information structures and the stochastic characteristics of bioeconomic parameters are varied continuously to explore their interplay. The asymmetric trade-offs among them are examined. The focus is on demonstrating the often unex pected, and sometimes counter-intuitive, effects that knowledge enrichment may have in these incomplete-information, common-property games.

This material is based on work supported by the U.S. National Science Foundation under Grant No. 9708475.