Improving the utilization of fishery resources as well as reversing the declines in resource levels are recognized as urgent and of high priority by resource managers from around the world (FAO 1996). For many nations, improving resource utilization and restoring resource levels will require considerable downsizing of existing fishing fleets. The exact level of reduction and the best programs for reducing capital and productive capacity, however, are unknown. These levels must be determined by the various nations in accordance with well-specified goals and objectives for fishery resources. The two basic questions in this regard, posed by Hannesson (1987), Clark, Clarke, and Munro (1979), and others, are: (1) how much of the existing capital stock to utilize in a given time period and (2) whether this capital stock should be increased or decreased by investment or disinvestment.
An initial stumbling block for determining the optimum utilization of fishery resources is the definition and measurement of capacity, capital, capital utilization, capacity utilization, and variable input utilization. This report provided several possible definitions and measurements. In addition, an extensive effort was made to offer approaches which would permit the actual measurement and determination of the measures given the objective of matching capital to resource levels in fisheries.
The concepts of capacity and capacity utilization this report recommends are those consistent with economic theory. The concepts of capacity and capacity utilization are short-run in nature, since the capital stock (and possibly other inputs) is fixed. Then, the notion of capacity is predicated on some measure of potential output, not inputs or fishing effort. Capacity utilization is the ratio of actual to capacity output. The technologically-derived primal definition of capacity is the maximum possible output level that can be produced in the short run given the variable inputs, the stocks of fixed inputs such as capital and the resource, and the state of technology. The corresponding measure of capacity utilization indicates how much the plant is producing relative to the technological maximum for which is was designed. The economics definition of capacity is the economic optimum level or target level (such as total allowable catches) of output that is produced in the short run given the stocks of fixed inputs such as capital and the resource, variable input prices, and the state of technology. Economic capacity utilization then indicates whether there are incentives to invest or disinvest in any time period. Thus, the second basic question posed above pertains to economic capacity utilization.
Managers often desire information on capacity utilization with respect to the total allowable catch or TAC. That is, managers want to know whether or not a fleet has excess catching capacity relative to a specified TAC. In this latter case, we may define capacity utilization as the ratio of the TAC level to the maximum potential output of the fleet. Overcapacity is defined to be the situation when capacity output exceeds a desired or target level of output (e.g., the fleet can harvest well in excess of a target harvest level equal to maximum sustainable yield). Overcapitalization is the case when the actual capital stock exceeds the optimum capital stock which is just necessary to produce a desired target output level at minimum cost.
The concept of utilization rates of inputs is relevant. There are utilization rates for variable inputs, the capital stock, or effort. In particular, the first question posed above is that of capital utilization. One definition of capital utilization is the ratio of the optimal to optimal capital stock and a second definition is the ratio of capital services to the capital stock.
The report recommends retention of the standard economics usage and nomenclature for capacity and capacity utilization as well as for capital and the utilization of capital, effort, and variable inputs. This consistency of terms and concepts with economics will help prevent the considerable confusion and proliferation of terms that currently exists in the fisheries literature.
The review of the economics and fisheries literature indicated that fishing capacity is generally used in the fisheries literature and capacity reduction programs to mean the maximum available capital stock in a fishery that is fully utilized at the maximum technical efficiency in a given time period. In sum, in many instances fishing capacity refers to what the economics literature calls capital utilization. In some instances, a technologically-derived primal measure of maximum or capacity output is intended.
The report also observed that capital stock measures are apparently often used to measure capacity and capacity utilization under the implicit assumption of a fixed optimal capital-output ratio, so that capacity output can be expected to vary directly with the observed capital stock. The two measures of capacity and capacity utilization coincide only if there is but one fixed input (capital) and if production is characterized by constant returns to scale.
The fisheries literature employs a myriad of terms for capacity. These terms include capacity, harvest capacity, available fishing effort, catch capacity, maximum effort utilization, maximum potential effort, fishing power, the capability to catch fish, and potential fishing capacity. The comparable fisheries terms adopted in this study were available fishing effort, or equivalently, maximum potential effort.
Another term commonly found in the fisheries capacity literature is fishing power. In these instances, fishing power is ultimately used to mean the stock of heterogeneous capital rather than the original meaning of relative catch rates or relative technical efficiency as developed by Garstang, Beverton and Holt, and Gulland. When fishing power is used to capture the capital stock, fishing effort frequently becomes a flow measure of capital services formed by the product of the capital stock and fishing time (the latter often called activity). To cut down on the confusion and terms with dual meanings, the report recommends that the term fishing power be reserved for the original intention.
The capital stock is comprised of multiple capital goods, such as the vessel, engines, gear, and equipment. The issue then arises of measuring this heterogeneous capital stock and in most instances, trying to form a single-valued measure of this capital stock. As discussed in the report, a single measure of capital stock is required for a meaningful measure of capacity and utilization rates of capacity and capital. A physical measure of capital stock, such as vessel sizes or numbers, is more meaningful to capacity reduction programs than a monetary value. However, this does raise the issue of how to properly aggregate the heterogeneous capital stocks (a monetary measure automatically does this). Standardization of fishing power also enters into the picture here. The report has suggested several untried approaches to standardize fishing power, based on panel data econometrics and technical efficiency.
Multiple products and multiple quasi-fixed or fixed factors - including the resource stocks, raise considerable complexities for measures of capacity, capacity utilization, and capital or effort utilization. When there are multiple products, primal measures of capacity and capacity utilization can ideally be formed only under rather restrictive conditions: (1) an aggregate output (requiring homothetic separability of outputs from inputs and the resource stocks); (2) a ray measure, in which all outputs are in fixed proportions (Leontief separability); or (3), the capacity measure conditional upon one of the (disaggregated) outputs. In contrast, the dual economic measure of capacity utilization readily accommodates multiple outputs, since the measure uses a scalar measure of the cost gap that exists when actual outputs differ from the capacity outputs, so that single-valued measures are still involved. (These measures are ratios of shadow to total costs or revenues or profits.)
When there are multiple outputs and multiple fixed factors, measures of capacity and capacity utilization are not generally possible. The issue is capacity pertaining to which capital stock - the vessel, engine, gear, or equipment? One general condition for a rigorous economic measure is that the number of outputs must be less than or equal to the number of fixed inputs. When only a single measure of the capital stock is employed, meaningful measures of capacity and capacity utilization are possible if the results are conditional upon the existing levels of the resource stocks.
The report also recommends other measures, depending on the goals and objectives of nations and capacity reduction programs as well as practical concerns such as data availability. These other measures include available or maximum potential fishing effort, capital utilization (often endogenous), effort utilization (often endogenous), variable input utilization, and the frontier output and corresponding input levels.
The report considers both conventional technologically-derived primal and economic-derived primal and dual approaches to measuring capacity and the utilization rates of capacity, capital, variable inputs, and effort and maximum potential effort. Ideally, economic-derived measures are preferred. This approach incorporates firms economic responses to markets and resource stocks and their conditions, and readily handles multiple outputs. This approach also readily accommodates regulatory constraints on inputs or outputs, such as limits on effort, employment, and total allowable catches. An economic-derived measure also incorporates the greatest amount of information available, most notably cost and revenue data. However, the demanding data requirements, especially cost data, in many instances make these economic-derived measures practically infeasible or costly to implement. Moreover, in situ systematic comprehensive programs to collect costs and input prices are infrequently found and are difficult to establish. Nonetheless, the judicious application of triennial cost surveys, such as those by DFO Canada, is promising and would allow economic-derived measures to be applied.
We recommend systematic cost-and-earnings data collection programs, such as those of DFO Canada, for capacity reduction programs. Problems of inconsistent questions from questionnaire to questionnaire, no capital services prices or vessel acquisition prices or current market valuations of vessels, and no fuel prices and/or quantities should be resolved with the otherwise excellent DFO data collection program. In addition, to the extent possible, the data should be collected as panel or longitudinal data, where the same vessels are tracked over time, accounting for entry and exit.
The difficulties - particularly data - encountered in forming economic-derived measures suggest a reliance, in many instances, on technologically-derived primal measures of capacity and capacity utilization. These technologically-derived primal measures require physical quantities of inputs, outputs, and resource stocks. These data are more routinely collected, in the form of landings statistics. Nonetheless, even in these instances, insufficient and/or untimely data for an accurate and comprehensive analysis are often collected on the physical quantities of the heterogeneous capital stock and variable inputs.
The peak-to-peak approach is the technologically-derived primal approach that demands the least data, and in that regard, is the most widely applicable. It is the simplest primal approach. The peak-to-peak approach requires data on landings and vessel numbers and some indication of a technological change trend. Some care should be given to the proper aggregation of all species to an aggregate output.
Two technologically-derived primal approaches with great promise are the stochastic production frontier and data envelopment analysis (DEA). These approaches should provide sufficient information to determine the maximum potential harvest of a fleet as well as identify the individual operating units which might be eliminated. Either approach should be short-run, with the capital stock fixed (rather than specified as capital flows). A critical element is to develop the analyses at the firm rather than fleet level. This places the analysis at the micro instead of macro level. If nations desire to remove economically inefficient operating units, both the DEA and dual-based stochastic frontier can provide the necessary information. The DEA approach is considerably easier to use. The DEA approach, because it is a form of mathematical programming, readily accommodates bycatch and socio-economic concerns, such as employment levels. With either the stochastic production frontier or DEA, an economic measure of capacity and capacity utilization can be developed and operating units and the economic efficiency of each operating unit may be determined.
In practice, however, there are many cases in which available data will be insufficient for using either the stochastic production frontier or DEA approach. Often, only information on landings and number of operating units will be available. In such cases, the DEA or stochastic frontier may be used, but the peak-to-peak approach - which is very simple to use, easy to apply, and least demanding of data - will provide estimates of maximum potential harvest and subsequent number of operating units.
If some nations desire to only determine the maximum potential effort, regardless of allowable catches, of an existing fleet, the DEA approach offers a cost-effective method of determining the maximum level of effort and its utilization rate. However, independent surveys and interviews or examination of detailed trip level data will be necessary to ascertain if operating units made as many trips as possible or fished as many days as they could during an operating period (e.g., a year). The output-based DEA measures allows the maximum output of each operating unit to be determined. The additional information allows a determination of other days or trips which could have been taken during a year. The two together permit maximum output and subsequently overall output capacity along with maximum potential effort to be determined. The solution, alone, however, does not facilitate measures of capacity units as done in the UK; the solution will allow determination of which vessels or operating units might be targeted for decommissioning.
If nations want to develop measures of relatively homogenous operating units, regardless of the level of effort, which might be done for the purpose of fleet or effort consolidation or to obtain meaningful measures of capacity and utilization of capacity or effort, numerous options are available. Individual firm-level production frontiers can be estimated and used as a basis for developing groupings of homogenous operating units (e.g., conduct cluster analysis of frontiers and group on length, horsepower, and other vessel/gear characteristics). Or a simple cluster analysis of catch per unit effort could be run and observations then grouped on vessel/gear characteristics. Alternatively, fixed effects panel data models could be specified and estimated in which each group was an arbitrarily identified according to vessel characteristics (e.g., vessel tonnage: < 50, 51 to 100, etc.).
Overall, there are a wide array of definitions and measurements of capacity and capacity utilization which may be considered for fisheries. In this report, we have not only offered a broad number of possible definitions and measurements of capacity and capacity utilization, but we have also offered a wide range of related issues which need to be addressed when developing measures of capacity and capacity utilization in fisheries. Foremost among the related issues is the need to determine the optimum configuration of fishing fleets. The configuration must be done from not only the perspective of a given fishery but also from the perspective of the entire fishing industry. We also have proposed a wide array of possible methods for determining measures of capacity and capacity utilization. Based on data typically available for fisheries and the needs of many nations for information on capacity and capacity utilization, the DEA approach appears to offer the most to help nations achieve their stated goals and objectives relative to matching capital to resources.
The authors are grateful for comments and suggestions received from: Dominique Gréboval, Fisheries Department, FAO; Pamela Mace and Matteo Milazzo, National Marine Fishery Service, USA; Gordon Munro, Department of Economics, University of British Columbia, Canada; Kathy Segerson, Department of Economics, University of Connecticut, USA; Kjell Salvanes, Department of Economics, Norwegian School of Economics and Business Administration, Bergen, Norway; Niels Vestergaard, Danish Institute of Fisheries Economic Research, Esbjerg, Denmark; and Jim Wilen, Department of Agricultural Economics, University of California, Davis, USA.
1 Here we are abstracting from the issue of vintage of heterogeneous capital stock, in which capital goods purchased in different time periods embody different states of technology and hence efficiency. Instead, for simplicity, we focus on uniform vintage.
2 The more efficient skippers are expected to earn intra-marginal rents, and consequently face a higher opportunity cost of exit from the fishery, all factors equal. Hence, they are expected to remain in the fishery.
3 The standard approach becomes to rely upon the notion of ray average costs, in which outputs are held in fixed proportion for different output levels, effectively imposing Leontief output separability, giving an aggregate output measure. See Baumol, Panzar, and Willig (1982) for further discussion of ray average costs. Segerson and Squires (1990) develop the corresponding ray measure of multiproduct capacity utilization.
4 This approach differs from the two-stage production process outlined above, which requires all primary inputs (capital, labor, fishing time, etc.) To be combined into a single separable and aggregate input, fishing effort. Instead, some or all primary inputs may be used in the production of many or all intermediate inputs. For example, crew can be combined with the vessel and fishing time to produce an intermediate input akin to fishing effort conceived as an aggregate input, but crew can also directly affect on-board operations, such as maintenance and processing. This approach has yet to be applied to fishing.
5 See Diewert and Shephard (1970) for further discussion. Shephard called the factor requirements function an inverse production function. Grilliches first called it the factor requirements function (Diewert 1974).
6 The factor requirements function, g, need not be separable as specified; separability between inputs and outputs is not a required condition of the factor requirements function.
7 Greene (1993) and Coelli (1995) provide recent comprehensive surveys.
8 DEA also permits an assessment of technical and economic efficiency from a non-orienting or base perspective.