Diagnosing the Problems
The preceding chapters have established that many marine fisheries are over-exploited and that fishing has adversely affected many marine ecosystems. At the most basic level, the problem is that too many people are catching too many fish, sometimes in the wrong place at the wrong time, and sometimes using equipment and techniques that damage natural ecosystems. The world's fishing capacity greatly exceeds what is needed to catch the sustainable yield. Finally, with a growing human population and increasing industrialization, there is growing pressure and ability to catch fish. This diagnosis highlights the need for an analysis of the major factors that have contributed to the current state of affairs. In broad outline they can be divided into three major categories: scientific matters, management matters, and socioeconomic incentives.
Scientific matters can be further divided into two major categories: lack of adequate scientific information and failure to use existing scientific information appropriately. We begin with the information itself and then discuss its use.
Assessments of Stocks and of Fishing Mortality
A fundamental premise of fishery management is that the productive potential of a stock and a fishery on it is a function of the abundance and biomass of the animals present in the stock and their life-history characteristics. These characteristics include the age distribution, natural mortality rate, age at maturity, and fecundity as a function of age. Most fishery-management programs depend to a degree on assessments of the stock and its productive potential (NRC 1998a).
The primary purpose of a stock assessment is to determine the abundance of individuals in the stock. In addition, fishery scientists estimate the mortality rate
and partition it into components from fishing and natural causes, allowing the exploitation rate to be calculated. Based on the stock's growth potential and mortality rate, its productive potential is estimated. Usually, estimates of the fishing mortality rate and stock size that can optimize (or maximize) the catch on a sustainable basis are provided. Thus, a determination of the status of the stock relative to present fishing intensity and the stock's ability to sustain additional fishing are two outputs of an assessment.
Assessing a fish stock is an inherently difficult problem. Except for anadromous fish (like salmon) migrating up rivers, it is impossible to actually count individuals, requiring that sampling methods be applied. They include fishery-dependent samples and data obtained from the fishery itself and fishery-independent samples from various survey techniques (e.g., trawl surveys, acoustics sampling, mark-recapture experiments). The sampling data usually provide the information that goes into the stock assessment. Because it is impossible to count all individuals in the stock, there is no perfect method to confirm the accuracy of the assessment techniques. Indeed, there are many biases: fishing and sampling gears selectively catch fish of certain sizes and behaviors. And the overall vulnerability or availability of fish to sampling gears can change with size and age (Ricker 1975; Gulland 1983). Fisheries that have been established for many years are difficult enough to assess, but recently established fisheries are even harder because long-term data are lacking.
A common problem in assessment is the nonuniformity of catchability, the probability that an individual will be caught by a unit of fishing gear. In schooling species such as herring or krill, or very large species such as whales where a single individual is worth pursuing, the relationship between catch and fishing effort is usually nonlinear (Ulltang 1980, Csirke 1988, Miller and Hampton 1989, Gulland 1983). In such fisheries, experienced fishers often can maintain high catch rates when the stock is declining in abundance. Under that circumstance, fishers—and sometimes fishery managers—can underestimate the mortality rate of the stock, making it easier to severely overfish it before regulations are instituted. This explains, for example, why a highly mobile and technologically sophisticated offshore fleet was able to maintain high catches of northern cod off Newfoundland while the less-mobile inshore fleet could not (Neis 1992, Steele et al. 1992, Finlayson 1994, Chapter 2 of this report). Walters and Maguire (1996) identified this problem as one that led to inaccurate stock assessments of northern cod, and indeed it is a reason that fishery-independent surveys are preferred in stock assessments.
Estimating fishing and natural mortality rates is also difficult, in large part because there is no perfect way to estimate abundance. Discards, mortality of escaped animals, and unreported catches are difficult and expensive to estimate precisely. However, they are part of fishing mortality, and the inability to estimate them accurately can lead to underestimates of fishing mortality, as discussed in more detail in Chapter 3 and in Chapter 2 for northern cod. Natural
mortality, the sum of all mortalities not a consequence of fishing, also is difficult to determine and contributes significantly to inaccuracies in assessments.
None of the above invalidates stock assessment as a fishery-management tool. However, in any industry where profit is a relatively small percentage of investment, as it usually is in fishing, it is difficult to forgo those small profits by reducing or stopping catches when there is a chance, but not a certainty, that the assessment indicates overfishing. The uncertainty cannot be reduced to zero, or even close to zero. Thus, the basic scientific uncertainty that inevitably accompanies estimates of stock size (abundance), productivity, and fishing mortalities is a reality that must be taken into account by any sustainable management program.
Environmental variability cannot be precisely predicted, and that leads to scientific uncertainty. Many fish populations fluctuate substantially from year to year and from decade to decade. In some cases these fluctuations are related to environmental fluctuations (Chapter 3); in other cases the causes are poorly known or unknown. Whether or not their causes are known, the fluctuations are often not predictable more than a year in advance, if that long. This unpredictable variability results in a rather fundamental scientific uncertainty: the future size of a stock is often unknowable, although probability distributions of future stock sizes are often estimated. Fluctuations in the population sizes of many commercially important marine species can occur at much shorter time scales than typical responses of the fishing industry, whose responses often depend on processes (such as shipbuilding, repayment of loans, and vocational training) with much longer time scales. The natural variability of many marine stock sizes interacts with uncertainties in current stock assessments to make precise planning impossible.
In some fisheries, landing statistics (the number, kind, and size of fish landed) are fairly accurate; in others they are less so. Even when landing statistics are accurate, however, they can bear an uncertain relationship to the number of fish killed by the fishery (Chapter 3). These uncertainties contribute to uncertainties in assessments of stock size, fluctuations, productivity, and fishing mortality. A fishery that tries to extract the last "surplus" animal (i.e., one that tries to maximize yield not over the long term but each season) is flirting with danger: many of the uncertainties described above work to deplete the fishery rather than to increase it. Despite the excellent work of many biologists in fishery agencies and universities, there will always be scientific uncertainty concerning how heavily a fishery can be exploited.
Uncertainties Concerning Socioeconomic Information
The uncertainties associated with landing statistics were mentioned earlier. Some uncertainty derives from uncertainties and lack of good information about human behavior. Indeed, the history of fishery management is full of examples of surprising human reactions to regulations—reactions that would not be surprising if managers had better and more detailed information. Limitations on boat lengths, for example, can lead to grotesquely wide boats, as in Bristol Bay (Alaska), where salmon boats are restricted to 32 ft in length, but have no restrictions on their beam. In Alaska, fishers responded to fleetwide halibut quotas not by reducing their fishing power but by fishing harder, so they could catch as many fish as possible before the overall quota was reached. That made the open seasons shorter and shorter until a "derby" fishery resulted, with hundreds of boats racing for fish in openings that ended up lasting less than a day (NRC 1994c, Buck 1995, Pennoyer 1997). A system of individual quotas was implemented in 1995, largely because of this distortion of fishing effort (Pennoyer 1997).
Despite considerable study, there remain major uncertainties in how fishers, their communities, and markets will respond to such management options as individual transferable quotas (see e.g., McCay 1995a), community development quotas, comanagement programs, international treaties, and so on. More needs to be learned about how fishery scientists and managers respond to uncertain information and to information that does not fit their scientific paradigm, such as traditional knowledge (Neis 1992, Finlayson 1994). The uncertainties regarding less-industrialized countries are even greater than for countries in North America, Europe, and Australasia. Mariculture has also introduced uncertainties by affecting the prices of wild-caught products, usually depressing them. Technological innovations also can have profound effects on fishing and the behavior of fishers, processors, and marketers. For example, the connection of Seattle to the east coast of the United States by the transcontinental railroad in the late nineteenth century provided a market for Pacific halibut, which resulted in an enormous increase in fishing pressure (Thompson and Freeman 1930). Other technological innovations, such as onboard freezers, also affected halibut (Bell 1978) and other fisheries (e.g., NRC 1992a) by allowing ships to stay at sea with their catches for weeks instead of only a few days.
How Scientific Information is Used
Because scientific information concerning fisheries is to some degree uncertain, there is always a temptation to assume the best and treat the fishery as though the uncertainty will work to benefit rather than hurt the fishers (Ludwig et al. 1993). Thompson's (1919) insight that scientific information will have to be overwhelming to change sport and commercial practices remains as true in 1998
as it was in 1919. The fishery literature is replete with examples of misuse or even lack of use of scientific information. For example, shrimp trawlers off the U.S. southeast and Gulf coasts long rejected the conclusions of the National Marine Fisheries Service that they were largely responsible for killing endangered sea turtles (NRC 1990). Because the information contained some degree of uncertainty, the shrimp trawlers were able to resist attempts to use turtle-excluder devices in their trawls until the National Research Council's reanalysis of data clarified their contributions to turtle mortalities. The uncertainty does not always work against the fishers. Recently, the International Pacific Halibut Commission increased the allowable catch of Pacific halibut because it learned that its stock assessments had been too pessimistic: apparently, there are really more halibut than the stock assessments had indicated (Ana Parma, IPHC, personal communication, 1997). Knowledge that uncertainties can occasionally benefit the fishers by leading to more fish than were expected makes it that much more difficult to routinely forgo catches in the face of uncertainty. Many cases of overexploitation of fishery resources result from this cause. NMFS (1996a), for example, described many examples of U.S. fishery resources that have been ''excessively fished" for many years. In other words, those resources have been exploited at higher rates than the scientific information—widely disseminated and never seriously questioned—supported, and no amount of scientific information would have changed this outcome.
Sometimes information is not used by policy makers and other stakeholders because it is not communicated to them in a way that is relevant and understandable. An important factor in communicating scientific information to managers and the public is to acknowledge and account for differences in the cultures of scientists, managers, policy makers, and the public. Disciplinary barriers, differences in operational constraints, and institutional differences are obstacles to good communication. There is a great deal to be learned from the experience and science of risk communication (e.g., NRC 1994f, 1996c).
Fishery management as a whole process (i.e., not necessarily individual fishery managers) is frequently blamed for failing to deal with the uncertainties in scientific information described above and for failing to take a conservative, or risk-averse, approach. But the fishery-management process includes many actors outside the management institutions themselves. In addition, fishery-management institutions often operate at time and space scales that do not match those of an individual fishery (e.g., NRC 1996a, 1996b). For example, the population fluctuations of some species occur so quickly that they cannot be determined until well into the fishing season, but most management and industry responses take longer than a season, especially those involving significant capital investments, such as boats and gear. In other cases, environmental variations affect the
stock size of target species (Chapter 3) in often unpredictable ways. In many cases the target species ranges over very large areas, often crossing several management jurisdictions and sometimes several nations (e.g., NRC 1994b, 1994c, 1996a, 1996b), yet the jurisdiction of fishery-management institutions often reflects political boundaries. In other cases a watershed (NRC 1996b) or a large reef or bank might be the appropriate management unit. Progress has been made in dealing with such difficulties (Chapter 5), but responding at the appropriate time and space scales remains a challenge for management.
Another problem is that many managers are trying to balance diverse, even conflicting, but unarticulated goals (Rothschild 1983, Pikitch 1988, Policansky 1993b, Hutchings et al. 1997). Another aspect of this problem is that a variety of political agendas and potential conflicts of interest complicate fishery management (NRC 1994c, 1996b). The challenge of making fishery management an inclusive yet balanced and fair process is a daunting one, as discussed in Chapter 5. There is a tendency to think of fishers as a monolithic group, all intent on taking as many fish as possible, but this oversimplified view is not realistic. Even in small fisheries there are usually a variety of sectors with different goals and interests. For example, in the northern cod fishery, the inshore and offshore fisheries are very differently constituted and have very different interests (Finlayson 1994). Processors can have very different interests from fishers. Recreational anglers—especially in the United States—are another important interest group. Even within recreational fisheries, there can be an enormous diversity of goals and interests, as described by Merritt and Criddle (1993) for Kenai River chinook salmon (Oncorhynchus tshawytscha).
Multiple fisheries on a single species or population and single fisheries on multiple species or populations (mixed-population fisheries) immensely complicate managers' difficulties. For example, Pikitch (1988) suggested that maintaining the structure of a community unchanged might not be compatible with any catch rate. As another example, Pacific halibut are managed by the International Pacific Halibut Commission, but approximately one-quarter of halibut landings in 1990 were taken as bycatch in other groundfish fisheries (Thompson 1993). That situation complicates the management both of halibut and of the other groundfish, whose capture can be severely limited by restrictions on the halibut bycatch.
The problem of mixed-stock or mixed-species fisheries is that some species and stocks (populations) are more productive or less susceptible to fishing (catchable) than others, so if fishing pressure is low enough to protect the least productive or most catchable population, others are "underharvested" and there is great pressure to allow an increase in catch rate. When the catch rate increases, the less-productive populations (NRC 1996b) or species (Pikitch 1988, Roberts 1997) or more catchable populations (Clark 1990) or species (Brander 1981) are depleted. This problem can be quite serious; it appears to have contributed or led to the loss of species from some environments (e.g., some salmon populations or species in streams in the Pacific Northwest [NRC 1996b]) and might even cause
the local extinction of a species, for example the common skate (Raja batis) in the Irish Sea (Brander 1981); the barndoor skate (Raja laevis) may be near extinction throughout its range in the Northwest Atlantic (Casey and Myers 1998). But the precise nature and timing of the interactions related to multispecies fisheries are very hard to predict (Pikitch 1988).
Finally, many fishery-management agencies have mandates and goals that are potentially in conflict. They are often asked to promote fishing and the fishing industry and to protect the ecosystem and the individual species in it. Sometimes, a goal—often unspoken but occasionally explicit (e.g., Task Force on Atlantic Fisheries 1983)—is the preservation of a fishing community's way of life. How fishery management affects fishing communities is a major issue all over the world. In the United States, the Sustainable Fisheries Act of 1996 requires fishery impact statements that assess the likely effects of management measures on fishing communities. The difficulty of dealing with goals that are not made explicit—much less agreed on by most of the parties involved—is common to many resource agencies, not only fishery agencies. While it can dealt with, it often does not receive the attention it deserves.
Enforcement is often a difficult problem for fishery managers and is related to many of the scientific uncertainties described above (involving biological and social sciences). The incentives to bend the regulations or to cheat are many, and there are so many participants in most fisheries that it is impossible to prevent or catch all violations. Sometimes the regulations themselves are confusing or not well disseminated, resulting in unintentional violations. Recreational fisheries are particularly difficult to monitor, and to some degree they depend for compliance with regulations on an honor system. The problem is well known, and because it involves illegal activities, solutions are made more difficult. We provide one example, that of whaling.
International whaling is controlled by two organizations, the International Whaling Commission (IWC), which allocates catch limits, and the Convention on the International Trade of Endangered Species, which regulates the import and export of endangered species. Both organizations tightly restrict the hunting and trade of all baleen whales plus sperm whales. The IWC adopted a moratorium on commercial whaling in 1986 (Marine Mammal Commission 1998), although some species like the humpback whale and the blue whale have been fully protected since the mid-1960s. About 500 to 600 whales are taken each year for research or aboriginal use, the bulk of which are minke whales taken under scientific permit to Japan. The purpose of the IWC is to sustainably manage the whaling industry and as such it represents an important example of international fisheries control for the benefit of sustainable exploitation of global natural resources.
Whale meat is important commercially in Japan, where prices range up to $150 per kilogram. Until recently, it has not been possible to compare the makeup of the retail market with expectations based on IWC rules because most whale products are unidentifiable once they reach local markets (e.g., sliced bacon, dried marinated whale jerky, canned meat). However, DNA sequences can now be used to identify the species of whale on international markets (Baker and Palumbi 1994, 1996), making it possible to evaluate compliance of the retail market with international expectations. These DNA results show that enforcement of international regulations is too lax to adequately protect the world's whale populations. About half of the whale products on the Japanese market are from the expected minke whale populations. The other half is made up of unprotected species of small toothed whales (dolphins, beaked whales, porpoises) and prohibited baleen whales (Baker et al. 1996). To date, most of the world's baleen whales have been found in the retail market, including humpback whales, blue whales, fin whales, Bryde's whales, and northern minke whales (legal only since 1994). Blue whales are particularly threatened, with an estimated worldwide population of 4,000 animals, yet they continue to be part of the retail market.
These animals are entering the market under the cover of legal products, but they are probably taken by illegal whaling and shipping. Recently, information from the former Soviet fishing fleet has shown that tens of thousands of whales were taken illegally in the 1960s and 1970s without notice by the international community (Yablokov 1994). Whale-meat smuggling may be a lucrative business, and reports of confiscation of illegal whale shipments surface regularly (Baker and Palumbi 1996).
International agreements have led to recovery of many of the world's whale populations. However, enforcement of existing regulations is hampered by the scale of the oceans and the complexity of international shipping. Illegal activities dilute efforts to manage whale populations scientifically and threaten the balance of national priorities on which international agreements are based.
The problems of the socioeconomic incentives in fisheries have been widely discussed. One problem is uncertainty, which can lead to risk-prone behavior. Another related (and better-known) problem is that of the "commons," where a lack of clear property rights leads to a difference between individual and short-term interests on the one hand and societal and long-term interests on the other. The classic statement of this "open-access" problem came from an analysis of fishing (Gordon 1954): it is in the interest of an individual fisher to increase catch, but it is not necessarily in the interest of the whole fishing community. Other problems, such as overcapitalization, derive from this one. A further problem is that some incentives, such as those derived from personal and societal goals, culture, and lifestyle, are not easily identified or incorporated into the bioeconomic models often used to develop fishery-management plans.
The conflict between individual short-term goals and long-term broader societal goals is strikingly illustrated by the plight of poor hungry people who depend on fishing for food. People in such straits cannot afford to worry about food tomorrow or food for others in the community if they do not get food today. However, even though they cannot avoid the strategy of satisfying today's needs at the expense of future needs, their plight will only get worse without some kind of intervention, because the fisheries will have less and less capacity to produce food. Some intervention is required to develop sustainable fisheries in such cases.
The problem of overfishing described in preceding chapters lies not so much with an inherent "greed" on the part of fishers but because, in most fisheries, fishers have faced an economic incentive system and a regulatory regime that lead almost inevitably to overexploitation and economic waste. Fishery resources are a form of natural capital. They can be seen as assets that can yield a stream of economic returns to society (Clark 1990, Clark and Munro 1994). They differ from human-made capital in that we receive an initial endowment of the capital assets from nature. It is possible to invest and disinvest in natural capital just as in human-made capital. Refraining from fishing and enhancement activities are investments in the resource. Fishing in excess of sustainable yield, and thus depleting the resource, is disinvestment.
Economic motivations for investments and disinvestments in fishery resources are similar to those for what might be termed conventional capital (e.g., plant and equipment). Investment in the resource requires a current sacrifice—deferred profits or costs of enhancements in the hope of a future return—while disinvestment provides immediate economic returns at the cost of lower future returns.
A key factor in investment decisions is the rate at which the investor discounts future economic returns as compared with current returns. The lower the discount rate, other things being equal, the greater will be the incentive to invest; higher discount rates lead to a lower incentive to invest. (The relative value of money today as compared with its value at some future time is known as the time preference of money and it is measured by the discount rate. Discount rates can reflect objective estimates of known relationships—e.g., depreciation of equipment, inflation rates, or the knowledge that if one doesn't fish soon, in a competitive-allocation fishery there might be no fish later—or they can reflect subjective time preferences.)
The payoff from the investment is enjoyed in the future, so resource investment, like other investment, involves uncertainty about the future. As a result, most individuals and societies give less weight to (risky) future than to current returns. In other words, risk is one cause of the future's being discounted, which reduces the incentive for investing. If the risk to future returns is high enough, the
incentive to invest can disappear and disinvestment will occur. These changes caused by the presence of a positive discount rate were not considered explicitly by Gordon (1954) in formulating his concept of the economically optimal fishing effort that results in maximum economic yield. Gordon implicitly assumed that the appropriate social discount rate is zero, i.e., that future economic returns from the resource should be given the same weight as current returns (Clark and Munro 1982).
The above has made capture fisheries difficult to manage in economic terms, mainly because of poorly defined property rights (Gordon 1954). Even when coastal states claim property rights to fishery resources in their waters, it is often difficult or expensive to vest property rights in the resources to fishers on an individual or collective basis (Gordon 1954, Munro and Scott 1985). Fish are mobile and not easily observable before they are caught. This contrasts with agriculture, in which property rights to the basic natural resource—land—are well defined.
The poor definition of property rights has two major consequences. First, individuals have a strong incentive to discount future returns heavily and not to invest (Clark and Munro 1982). If they attempt to invest by not fishing, they might do no more than increase their competitors' catches. Instead, the incentives lead them to increase their own share of the available resource by fishing more, rather than less. The second major consequence, which follows from the first, is fleet (and perhaps processing) overcapacity. These conditions—increased fishing effort and increased capacity—make overexploitation more likely.
Practical Considerations for Management of Fisheries
In practice, fishery management has two critical elements. The first—related to conservation—is intended to limit fishing mortality and is implemented through input controls and output controls.1 The second can be explicit, or implicit or by default, and that is related to the processes by which access to the resources are allocated. The default situation usually involves no specific action. If managers attempt to prevent overexploitation by limiting fishing effort without changing a competitive allocation scheme, the limited catch will become a common pool. Each fisher will be competing with others for a share of the total resource, which is now limited. This competition often leads to increased capital investment in fishing effort (gear, boats, human resources, and so on), a phenomenon called capital stuffing. Soon there is more fishing capacity than needed to catch the limited resource, which leads to a race for the fish. In addition, as regulations are successful in increasing the size of the resource (increasing fish populations), profits from fishing will increase and make additional investment
worthwhile, which also results in overcapitalization. The investment of individuals in more fishing capacity is entirely rational, even though the total catch will not increase.
The primary alternative to a default or competitive allocation process is share-based or rights-based allocation. This approach also can provide incentives for conservation because participants (rightsor share-holders) have a stake in the future of the resources and because some rights can provide incentives for efficiency (output reduction). The promotion of efficiency occurs because in the absence of competition for shares of the resource, economic success is represented by fishing efficiently. These are the main reasons why this committee has encouraged the development and use of share-based allocation systems to replace competitive allocation schemes.
The following paragraphs briefly describe some of the theory and experience of fishery management from an economic viewpoint. For more detailed discussions, the reader is encouraged to read Clark (1990) and OECD (1997) and references therein. The OECD publication in particular contains many recent references and ample examples of fishery management in practice.
In theory, the imposition of conservation measures strong enough to be effective, either through input controls such as gear limitations or seasonal and areal closures, or through output controls such as catch limits (usually total allowable catches or TACs), allows allocation methods to be considered independently of conservation measures. In practice, conservation and allocation methods can become dependent if competitive allocation drives up fishing costs to the point where rent is dissipated or marginal. The management of Pacific halibut before the implementation of ITQs is perhaps the best example of this: the resource was protected for decades, but the race for fish became excessive and dangerous with many adverse social and economic consequences, as described earlier in this chapter.
If the system were well balanced, or theoretically perfect, the above might not be a problem; when rents approached zero, investment would decline to reflect that. But in practice, three important factors cause problems. The first is imperfect information about the current and future size of the fishery resource, which can lead to costs (inputs) that exceed what the resource can sustain. The second factor is natural variability, which can have a similar effect. The third factor is the cost of complying with regulations (for example paying for and using specified gear modifications). The costs of compliance as well as of licenses or other fees can increase the costs of fishing, which can affect profitability, especially if other suppliers of the market do not bear the same costs.
The first two factors are particularly problematic when a fishery develops on a resource that has a high standing stock. In that case, the investment is often made in response to the standing stock, analogous to financial capital, rather than in response to the productivity of the resource, analogous to return on investment. In such cases one observes a so-called ratchet effect (Ludwig et al. 1993) and the
investment overshoots the appropriate level for the productivity of the resource and excess capacity is a result.
Excess capacity is a form of economic waste and makes a fishery vulnerable to resource shocks (e.g., reduced availability of fish or regulations to reduce catch) or to economic shocks (e.g., falling prices or increasing costs, such as happened in the 1970s with the oil embargo). Excess capacity is difficult to estimate quantitatively, but a few quantitative and many qualitative estimates—ranging as high as 75 percent for some fisheries—were described by Mace (1997). For Alaska groundfish fisheries, Pennoyer (1997) estimated that overcapacity was 300 to 400 percent before the introduction of individual quotas. Often, the vessel and processing capital cannot be used for purposes other than the specific fishery. Vessel and factory owners usually have debts, such as mortgages, to be serviced. As the excess capacity dissipates economic returns from the fishery, the incentive for fishers and processors to press for liberal catch quotas increases as does the political influence of the users: if their catches are significantly reduced, they often face bankruptcy. The industry often uses scientific uncertainty or natural variability to argue against reduction of effort (OECD 1997). The pressure for liberal catch quotas can be very strong—often involving important political figures—and risk-prone management often results (Sissenwine and Rosenberg 1993, Rosenberg et al. 1993). Even if managers resist pressures to make risk-prone decisions, the existence of a large, chronically undersatisfied fleet exacerbates monitoring, control, and surveillance (Dupont 1996). If the fishing capacity of the fleet is held to levels at or below that required to produce the maximum sustained yield or the maximum economic yield, then the possibilities of overfishing are substantially diminished (OECD 1997). The more the fishing capacity exceeds that necessary to produce the desired sustainable catch, the greater the potential for overfishing. Indeed, overcapitalization is often cited as the most important factor in overexploitation of the world's fisheries (e.g., NMFS 1996b, FAO 1996b, Christy 1997, Mace 1997, WWF 1997). Nonetheless, there are examples of fisheries that have extreme excess capacity in which strong management has prevented overfishing even in the absence of share-based allocations (e.g., the Pacific halibut fishery before implementation of ITQs).
These practical matters cause pressures that often lead to or exacerbate overexploitation. If there is political will, they can be dealt with, but often that will has been lacking. No method of reducing fishing mortality to achieve conservation and thus sustainable fishing will be economically or socially painless; financial investments and jobs will be lost. However, if sustainable fishing is to be achieved, reducing effort in the short term is necessary. The options lie in deciding how and when to reduce effort so as to reduce economic and social disruption. The options, however, can be exercised only if decisions are made before the resources are depleted.
The problem of overcapitalization, which has been described in this report as a serious contributor to overfishing, is probably seriously aggravated by government subsidies. Estimates of the total subsidies in marine fisheries are extremely hard to obtain and vary widely, from as much as $46 billion annually to $11 billion annually (FAO 1993b, Garcia and Newton 1997, Milazzo 1997, Porter 1997). All authors agree that the estimates are very imprecise. Even the lower estimate, however, is very large, especially considering that the total gross revenues of the world's marine fishing fleet are estimated at about $70 billion (FAO 1993b) or $80 billion (Milazzo 1997) per year.
A recent symposium in New Zealand on fisheries and international trade devoted substantial time to subsidies and concluded that subsidies in fisheries are large and pervasive and that the motivations and impacts of the subsidy programs are variable and poorly understood (Pacific Economic Council Task Force on Fisheries Development and Cooperation 1997). Until our knowledge of these matters improves, it is difficult to make specific recommendations for dealing with them, although it is clear that they complicate attempts to reduce fishing effort by reducing overcapitalization (Porter 1997, Milazzo 1997); it is also clear that much additional research is needed. Milazzo (1997) suggested that environmental subsidies (e.g., vessel and fishing permit buybacks, stock enhancement, research and developments in "clean" fishing gear, perhaps others) are preferable to "conventional, effort- and capacity-enhancing" subsidies and should be given higher priority. The design and implementation of environmental subsidies need improvement, and certainly research is needed to understand their potential better.
A great number of scientific, management, and socioeconomic uncertainties and difficulties contribute to the overexploitation of marine fisheries and their ecosystems. Fisheries science cannot provide precise estimates of fish abundance or of the impacts of fishing, and the information that science can provide is not always well used. Environmental variations introduce a great deal of variability in fish populations and uncertainties in managing them. Pervasive and powerful economic and social incentives lead to overexploitation of fisheries. Some improvement is possible in each of these three areas, but it seems certain that the kinds of incremental improvements that have characterized recent decades will not by themselves reverse the trend toward increased overexploitation, although maintaining progress in those areas is essential. Creating more appropriate incentive systems and developing management institutions that can accept and deal with variability and uncertainty are crucial to establishing sustainable fisheries. Without them, populations of individual species and the structure and functioning of marine ecosystems are likely to continue to decline.