Multi‐Output, Nonfrontier Primal Measures of Capacity and Capacity Utilization

Abstract

This article implements an econometric approach for generating primal capacity output and utilization measures for fisheries. In situations where regulatory, environmental, and resource conditions affect catch levels but are not independently identified in the data, frontier-based capacity models may interpret such impacts as production inefficiency. However, if such inefficiencies are unlikely to be eliminated, the implied potential output increases may be unrealistic. We develop a multi-output, multi-input stochastic transformation function framework that permits various assumptions about how output composition may change when operating at full capacity. We apply our model to catcher-processor vessels in the Alaskan pollock fishery. Copyright 2004, Oxford University Press.

Estimating Fishing Vessel Capacity: A Comparison of Nonparametric Frontier Approaches

“…16 Except for the resurgence of the use of the deterministic frontier (i.e., parameters corresponding to a specified frontier function are estimated via mathematical programming but the estimates are adjusted by corrected ordinary least squares), DEA and the stochastic frontier appear to be the two primary approaches used to estimate and assess technical efficiency and capacity, with DEA being the primary approach used to estimate capacity (F@re et al, 1993;Kirkley et al, 2002;Felthoven and Morrison-Paul, 2004;F@re et al, 2006). In the next section, we introduce the various approaches and methods but primarily focus on the use of DEA, the stochastic frontier, and the modified deterministic frontier.…”

“…This is because a primary concern of the present research is the assessment of capacity in fisheries for which economic data necessary to estimate the economic concept of capacity are seldom available. We also present a comparative framework based on the stochastic frontier developed by Kirkley et al (2002) and a nonfrontier approach offered by Felthoven and Morrison-Paul (2004 F@re et al (1989b) later offered a more completed development of the framework and included a procedure for estimating an unbiased measure of capacity utilization. F@re (1984) and F@re et al (1989b) offered procedures for estimating three important metrics related to capacity: (1) capacity output, (2) an unbiased measure of capacity utilization, and (3) a measure of variable input utilization.…”

“…There remains the problem, however, of omitted variable bias. Felthoven and Morrison-Paul (2004) offer an alternative specification and estimation of capacity. They adopt an approach similar to that of Kirkley et al (2004) but do not consider the stochastic production frontier.…”

Estimating Heterogeneous Capacity and Capacity Utilization in a Multi-Species Fishery

“…Figures 1 and 2 illustrate these differences when one assumes a homogeneous versus heterogeneous model and the degree of over/under measure of capacity generated by the homogeneous model assumption, when there exists two distinct production technologies, 10 There are number of issues that must be addressed when defining capacity measures, for a more detailed discussion of these issues see [20]. 11 Alternatively, we could estimate capacity utilization…”

“…To define capacity we base our measure of capacity on the technological-economic approach [11]. This measure defines capacity as the maximum feasible output that can be produced given the current level of technological, environmental, and economic conditions.…”