On implementing maximum economic yield in commercial fisheries

MEY was developed originally as an equilibrium concept. In reality, fisheries are not in equilibrium, nor are species caught in fixed proportions. Operationalizing MEY requires developing models that take the dynamics of stocks, costs, and prices into account. Maximizing the net present value of profits over time is a more appropriate objective and is consistent with the concept of MEY. In the past, dynamic bioeconomic models capable of such analyses generally have been relatively simplistic in their assumptions because the results were more illustrative of the benefits of moving toward an economic target than an attempt to identify the target per se (12, 13).

In these and many other dynamic bioeconomic models, considerable attention is given to how the size of the fish stock changes over time (e.g., as a result of natural and fishing mortality, growth, and recruitment). For example, in the case of the NPF, the population dynamics are modeled on a weekly time-step, account is taken of key biological processes, and bycatch of other prawn species is modeled explicitly (11, 14). However, in most dynamic bioeconomic models, fleet dynamics are represented in a crude or ad hoc manner or are missing entirely. Little prior consideration has been given to how to estimate or select desired levels of inputs and outputs in multispecies fisheries where fishers can alter the combinations of species caught (at least to some extent) by changing when, where, and how they fish, even though most fisheries are of this type. Detailed models of how fleets adjust over time are difficult to formulate and parameterize. Over time, vessels may enter and exit a fishery, invest and disinvest (affecting their technical efficiency, fishing power, and cost structure), and change their behavior in response to changing economic incentives (e.g., changing both the level and spatial allocation of fishing effort). Although models exist to capture at least some of these factors (15, 16), no bioeconomic model has been able to include all of them comprehensively. In the case of the NPF, the industry recently had undergone a major restructuring, and license numbers were limited, so the assumption that no vessels would enter or leave during the transition period to MEY was considered reasonable. (This assumption is examined later.) However, other behavioral or vessel/gear changes may occur and are not captured in the model analysis.

In most countries where economic performance is identified as an objective of fisheries management, the objective usually is related vaguely to maximizing or improving the returns to society from the use of the resource. However, MEY is a partial-equilibrium optimum and relates only to the fishery. In most fisheries, vessel numbers need to be decreased substantially to achieve MEY. The associated reduction in crew numbers and in regional economic activity associated with the fishing industry may result in a net economic loss even though the fishery experiences a substantial increase in economic profits (8). The traditional economic response to this scenario is that the loss in regional economic activity is ephemeral, because the resources previously consumed in fishing are freed up to be used more productively in other sectors. However, short-term factors have been highly influential in management decision making in some major fisheries (17). In the case of the NPF, this effect on associated economic activities was not considered an issue because considerable fleet adjustment had already taken place, and the trajectory to MEY was not expected to require further fleet reductions.

Without constraints on fishing activity, a dynamic bioeconomic model will suggest that it is economically optimal to reduce fishing effort in the short term to achieve a higher, longer-term stream of benefits when biomass is currently less than the economic optimum and the discount rate is finite. This scenario may involve closing the fishery or substantially reducing harvest for several years. However, this approach does not account for the costs associated with effort reduction or fishery closure. Closing the fishery would be optimal only if the vessels had a viable alternative use or if the stock were severely depleted (such that any fishing effort would prevent recovery) (18). Most alternative fisheries in Australia and the rest of the world already are closed to new entrants, so the capital has few alternative uses following fishery closure. Fixed costs still will be incurred, and fishers will need a means of covering these costs. Labor also is problematic. Displacing labor is less difficult than displacing capital, but regaining a skilled labor force when a fishery reopens or requires increases in effort can be difficult (19). Given these considerations, closing a fishery for a short period, although potentially optimal in the “model world,” is generally unacceptable in real life.*

We can go further and suggest that forcing a fishery to be unprofitable in the short term is also impractical, even if longer-term gains would be realized. Although a theoretical economist would argue that fishing is still worthwhile, the ability of fishers, many of whom may have experienced low levels of profitability in the past, to survive a period of negative profits may be limited because vessels still need to cover their variable costs. To impose such a situation knowingly on a fleet when an alternative path may exist in which profits are consistently nonnegative is likely to lead to litigation and delays. Consequently, limits on effort reduction in any particular year must be imposed, and the resultant constrained optimal trajectory will diverge from the unconstrained optimum. In the case of the NPF, as would be expected, setting a minimum constraint that fleet profits must be nonnegative in each year resulted in a different estimate of the optimal trajectory of catch and effort than in the unrestricted case (Fig. 1), although the optimal equilibrium yields of the species were similar (Table 1). (More detailed model outputs for each simulation are presented in the SI Text.)

Similarly, setting a minimum effort level for each fleet—a suggestion made by the industry and incorporated into the current management system—provided an outcome similar to the profit-based constraint.

The ability of fishers to apply fishing effort is limited also. Although underutilized capacity exists in most fisheries, once this capacity is fully used, individual vessels are technologically unable to apply more fishing effort and take more catch, or doing so is not economically viable (i.e., the additional costs exceed the value of the additional catch). The optimal trajectory in the previous analyses assumed in effect that vessels were unrestricted in their ability to fish, with unrealistic levels of fishing effort being needed for some weeks during the fishing season. Restricting total effort to a realistic level, given the season length, resulted in an entirely different effort trajectory (Fig. 1) and set of optimal yields (Table 1). Allowing new boats to enter the fishery results in a lower effort trajectory and set of optimal yields than given in the scenario in which effort is unrestricted, because additional fixed costs are involved. Profits also are lower than in the unrestricted case but are greater than when vessel numbers are fixed but vessels are restricted in their annual levels of effort.

From the experiences in the NPF, as well as economic theory, optimal effort, biomass, and yields are functions of both input (e.g., fuel) and output (e.g., prawn) prices as well as of biological parameters (4). Although historical biological parameters can be assumed to be valid for the immediate future, economic parameters may vary considerably over time. For most fisheries, prices generally are assumed to be invariant with the quantity landed.^†

Consequently, most previous bioeconomic models have assumed that prices and costs remain constant (in real terms) over time (12, 13, 15–17). However, when prices are expected to change over time, the optimal trajectory and final MEY will depend on the future price, and this price must be anticipated well in advance (21).^‡ Hence, expectations about future price movements cannot be ignored if model results are to be used for management purposes. Accurate forecasts of these variables are critical but in most instances will be subject to high levels of uncertainty. The economic environment can change substantially even over a short time period, as seen in the substantial oil market fluctuations during 2008 and the recent economic crisis, rendering previous forecasts invalid. Fluctuations in exchange rates affect both costs of imports and the prices received for exports. Reduced demand in importing countries may result in the domestic market being flooded, with subsequent price implications.

This uncertainty is likely to increase with the length of the forecast period. For the NPF analysis, which is based on a relatively short-lived species, a 7-year time period was considered necessary to enable stocks to recover to the optimal levels. For long-lived species, we would expect forecasts for a relatively long period to be required, because MEY would take longer to achieve. Although the importance of future profits is reduced as a result of discounting, uncertainty around the MEY estimates could be substantial.

Assuming we have accurate cost information, what costs do we include in analyses? In the long run, all costs are variable. However, in the short term many costs are fixed, so the estimate of the optimal output can be based on variable costs alone, and fixed costs can be ignored. However, some sizeable costs, such as repairs and maintenance, are quasi-variable (or, alternatively, are quasi-fixed). That is, they are affected by the level of fishing activity in a given year, but some level of cost is incurred irrespective of the level of fishing activity or may depend on the cumulative activity over several years. Treating these costs as variable costs may underestimate the short-term optimal yield, whereas treating them as fixed costs may overestimate this yield.

The impact of different assumptions regarding the treatment of prices and costs on the estimation of MEY was examined. Assuming constant prices and costs (in real terms) resulted in estimates of optimal catches being around 5% lower for the 3 main species, whereas treating the quasi-variable costs as fixed (rather than variable) resulted in optimal catches being around 2–3% higher (Table 1). Furthermore, the optimal effort trajectory (and resultant “equilibrium” stock size) required to achieve MEY differed, depending on the assumptions regarding costs (Fig. 1).

The uncertainty about future prices and costs suggests the need for regular revision of the costs and prices on which management advice is based and hence of the catches and levels of fishing effort that are estimated to maximize new present value. For the NPF, a decision has been made to re-estimate the trajectory of catch and effort every second year; this decision was based on some additional modeling work that indicated that this revision would not compromise profits and sustainability but would reduce interannual variability in management advice.

At best, model estimates of transitional paths indicate what should happen rather than what will happen. Fishers respond to the changing economic conditions and to the set of incentives generated by the management system under which they operate. Maximization of discounted profits cannot be achieved without changes in governance structures, such as those that remove the incentives to race to fish. However, changes in these structures may result in changes both in cost structures of the fleet and in the price received (22, 23), adding an additional level of complexity to the estimation of MEY. Additional models of fleet behavior must be considered, and a range of alternative cost assumptions also may be required to incorporate these factors. As noted earlier, the optimal yields are highly sensitive to these cost assumptions.

Even the best models of fisher behavior are unlikely to provide a perfect representation of how fleets are likely to respond to management changes, and even the best governance structures are unlikely to guarantee that the optimal trajectory is achieved. Given that the optimal trajectory is state-dependent, a new trajectory will need to be re-estimated regularly, taking into consideration what the management system actually achieved in terms of catch and effort in the fishery during the transition period, as compared with the goals it was aiming to achieve. For the NPF, the decision was made to have a continuously rolling transition period of 7 years, so that each time MEY is estimated (taking into account also the latest price and cost forecasts), the target trajectory is over a constant period rather than an ever-decreasing period. This rolling transition period may mean that a single-point estimate of MEY may never be achieved, but it is expected that the fishery will be able to come as close as possible to achieving optimal profits over time.

Fishers and managers need to be able to participate in the decision-making and implementation phases of managing a fishery in countries where comanagement is seen as key to good fisheries management. However, profit maximization and MEY are concepts that are not well understood, even by scientists working in the fisheries field, and often are viewed with suspicion by industry, which equates lower catches with lower (rather than higher) profits. In contrast, fishers, scientists, and managers are reasonably aware of the strengths and weaknesses of MSY-type reference points and harvest strategies. MSY is a constant, irrespective of price and cost assumptions, reducing the management uncertainty faced by fishers. Although a range of effort trajectories may exist so that stock size fluctuates around that corresponding to MSY—with differing economic consequences—the end point generally is more stable. Fishers therefore are well able to participate in the decision-making and implementation phases of managing a fishery. The discussions focus on which harvest-control rules are preferable, and, once the rules are agreed upon, fishers are reasonably confident that the effort trajectory and biomass target will remain relatively unchanged and are able to plan their activities accordingly. In contrast, MEY is a moving target, because it changes with predictions of costs and prices, and practical considerations such as how to set harvest strategies against such a target have yet to be determined.

In the NPF, industry fully supported—and indeed proposed—the move to MEY, but many fishers were not aware of what this change actually entailed. This lack of understanding is not surprising, given the paucity of fisheries that could be looked at for examples. Once some of the complications and implications became apparent, even though the work was undertaken in a strong collaborative environment, some industry members started to have second thoughts about MEY. Without the strong institutional arrangement in the fishery, it is unlikely that implementation would have been possible. Time will cure these challenges as more examples similar to the NPF become available. Education of stakeholders about the reason for using MEY as a management target is critical, as is sharing of knowledge and experiences in modeling MEY.