Among-stock comparisons for improving stock assessments of data-poor stocks: the 'Robin Hood' approach

An approach is outlined for conducting stock assessments in which parameters are estimated for multiple stocks at the same time. Information from data-rich stock assessments, e.g. trends in fishing mortality, and values for parameters of selectivity functions are provided to data-poor assessments in the form of penalties on the estimated parameters, which leads to stock assessments for the most data-poor stocks being informed by those for the most data-rich stocks. The method is applied for example purposes to data for nine stocks in Australias southern and eastern scalefish and shark fishery. The results of the application confirm that results for data-rich stocks are little impacted by being assessed in conjunction with data-poor stocks and that the results for data-poor stocks can be qualitatively different when information for data-rich stocks is taken into account.