ii
Executive SummaryFishery managers throughout the world are concerned about finding ways to reduce undesirable and nonmarketable bycatch (i.e., bycatch discards) and excess harvesting capacity. Both of these issues are associated with economic waste in the form of unnecessarily high production costs, potential reductions in future harvest levels, or unnecessary utilization of factors of production to discard undesired catch. The Food and Agriculture Organization (FAO) of the United Nations, the National Marine Fisheries Service (NOAA Fisheries), and various member nations of the FAO adopted a voluntary code entitled The Code of Conduct for Responsible Fisheries in 1995, which promotes the reduction in bycatch discards and excess capacity in commercial fisheries.Extensive progress has been made towards reducing bycatch of undesired species through new regulations and modifications to traditional fishing gear. Substantial progress has also been made towards assessing and reducing harvesting capacity in fisheries. Unfortunately, these efforts have been made in isolation. That is, efforts to reduce bycatch have not concurrently considered the ramifications of bycatch reduction on harvesting capacity, and efforts to assess capacity have not incorporated how capacity would vary with reductions in undesirable outputs.Without proper attention to the relationship between reducing undesirable outputs and the maximizing desirable outputs, it is quite likely that the estimates of capacity used to help develop capacity reduction programs may be subject to error.We examine four approaches for estimating and assessing both capacity and technical efficiency of production activities that involve the production of both desirable and undesirable outputs. Although we primarily focus on estimating capacity while explicitly recognizing the need to allow desirable outputs to expand and undesirable outputs to contract, we also consider several other options for changing the direction (expansion and contraction) of desirable and undesirable outputs.All four methods considered in the report are based on data envelopment analysis ( DEA), which is a mathematical programming approach for estimating technical efficiency (TE) and capacity output. We first examine the more traditional DEA approach for estimating capacity; this is an output-oriented approach, which only takes desirable outputs into account and ignores undesirable outputs. We then introduce and summarize (2) a directional distance function approach, which permits desirable outputs to increase and undesirable outputs to decrease by the same proportion. Next, (3) a hyperbolic approach is then presented and discussed; this approach allows desirable outputs to expand by a scalar and undesirable outputs to contract by the inverse of the scalar. Last, we present (4) iii the approach of Seiford and Zhu (2002), which is an output-oriented approach but allows desirable outputs to increase and undesirable outputs to decrease. We then apply the various models to a data sample from fis...