Downside Risk and Fishing Quotas. Alistair Milne, Department of Economics, University of Surrey, email: a.milne@surrey.ac.uk
The exploitation of fishing stocks is characterised by a sharp asymmetry between upside and downside risks: overfishing leads to sharp declines in fish populations which then take several years to replenish. This paper explores the implications of this asymmetry for the determination of fishing quotas. In the absence of certainty equivalence it is not possible to summarise an estimate of the optimal rate of harvest as a single rate of harvest with appropriate margins of statistical error. Instead it is necessary to present a schedule of returns for different levels of harvest, with rapidly declining returns as harvest is increased above the point estimate of the optimal rate of harvest. It is also necessary to allow for the possibility of a departure from optimal policy in the future. These corrections to standard analytical practice have direct implications for the current procedure for setting European fishing quotas. Failure to adequately factor in for downside risk is one reason why quotas are set persistently too high, leading to the current destruction of European fish stocks.
A simple stochastic model is set out in which the evolution of the stock of a fish species depends non-linearly on the rate of harvest, the current level of stock and a (possibly serially correlated) random disturbance. Harvesting policies is first analyzed according to a best case scenario in which all future policy is optimal. Here a recursive equation is an appropriate tool for analyzing the expectation and distribution of the present value returns to each level of harvest. With further assumptions about the demand for fish this can then be translated into a cost-benefit analysis for a range of harvest levels. The analysis is extended to allow for the further risk of sub-optimality in future policy: with future harvest always set at higher than the optimal harvest levels in all future periods. This leads to a shrinkage of confidence limits in the sense that if policy makers choose, under pressure from industry, to systematically set quotas at the 95% upper confidence limit for the optimal level of quota, then the confidence limits themselves must be set much closer to the point estimate of optimal harvest level.
The recursive equation is calibrated using
data on European fish stocks. Numerical computations is made
of the expected balance of costs and benefits for different harvest
levels with 5% and 95% confidence limits. Comparison with quota
levels set for various species in European markets under the common
fishing policy, to see how actual quotas compare with the calculations
and to investigate to what extent they can be defended as lying
within confidence levels for the optimal level of harvest.