In its review, in chapter 5, of biological data upon which wildlife management decisions in Alaska are based, the committee did not attempt to address questions of the social or economic implications of those decisions. Alaskan environments and the life-history traits of the focal species were described in purely scientific terms. The committee evaluated the success of past control efforts according to the stated goal of increasing moose and caribou for human harvest, without assessing whether the goal itself is economically or socially desirable. That question is addressed in this chapter, in terms of what is known about the attitudes and values of the public in Alaska and elsewhere, about the economic impacts of predator control in Alaska and elsewhere, and about how this information has been brought to bear on decision-making regarding predator control in Alaska.

There are 2 reasons for raising the issues considered in this chapter. The first is that they were specifically identified in Governor Knowles' request to the National Academy of Sciences; the statement of task directed the committee to examine what studies and research methods could be used to evaluate the full economic costs and benefits of predator control. The second is that in the committee's view economic impacts are central to modern wildlife management and depend, to a large degree, on public values. In democratic societies, government programs, such as wildlife management, should be based on widely shared public goals that legitimize the actions taken and the allocation of public financial and human resources to them. Funds for wildlife management are largely from

license sales and federal aid from excise taxes on firearms and ammunition. Nevertheless, some general public funds are appropriated for wildlife management, and their expenditure means that the public implicitly agrees to forgo the benefits of other uses of the funds.

Any analysis of the social and economic dimensions of wildlife management programs must recognize that people differ in their attitudes, their values, and their economic situation. Consequently, there will be differences in how they view the programs. Accordingly, the committee organizes its discussion on the basis of the different groups that might be affected by wildlife management programs in distinctive ways—Alaska Natives, other Alaska residents, and people outside Alaska who visit it or otherwise take an interest in its wildlife. The economic and social implications of predator control programs cannot be assessed without reference to people's attitudes and values. Therefore, each case begins with an analysis of what is known about attitudes toward wolves, bears, and their primary prey followed by a consideration of the economic implications of the attitudes for predator management and control options.

The first section of this chapter considers general North American attitudes toward the gray wolf and, to a lesser extent, bears, and related views of predator management. The next section focuses on views of Alaskans and visitors to Alaska. The remaining sections analyze the economic costs and benefits associated with wolf and bear control in Alaska for non-Native residents of Alaska, nonresidents, and for Alaska Natives, and the social and economic impacts of decision-making procedures themselves.

Much of North America's historical treatment of wolves, bears, and other large carnivores focused on suppressing them, as reflected most dramatically in extensive efforts to reduce and even eliminate their populations (Lopez 1978; Matthiessen 1988). The wolf was particularly targeted because of its perceived threat to livestock and human settlement and its presumed competition with people for other wildlife (Dunlap 1988). The prevailing image in colonial America of wolves and, to a lesser extent, bears is reflected in John Adams's remark of 1756: "The whole continent was one continuing dismal wilderness, the haunt of wolves and bears and more savage men. Now the forests are removed, the land covered with fields of corn, orchards bending with fruit and the magnificent habitations of rational and civilized people" (Kellert 1996:104). Widespread negative perceptions persisted well into the 20th century, as illustrated by the remarks of the historian and trapper Stanley Young; an early director of the US Biological Survey, E.A. Goldman; and the first president of the New York Zoological Society, William Hornaday: "There was sort of an unwritten law of the range that no cow man would knowingly pass by a carcass of any kind without

inserting in it a goodly dose of strychnine sulfate, in the hope of killing one more wolf" (Young 1946:27). "Large predatory animals, destructive of livestock and game, no longer have a place in our advancing civilization" (Dunlap 1988:51). "Of all the wild creatures of North America, none are more despicable than wolves. There is no depth of meanness, treachery or cruelty to which they do not cheerfully descend" (Kellert 1996:105).

Those unsympathetic attitudes contributed to the extirpation of wolves and brown bears from much of their range in the 48 states and their eventual listing under the terms of the US Endangered Species Act of 1973 (Kellert 1996). Radically changing views of wildlife and large carnivores during the second half of the 20th century, a view of wolves and bears as imperiled species, and increasing knowledge of these species all contributed to more-positive attitudes toward these animals. This change has been especially notable among urban, and younger Americans, and those with more extensive formal education (Kellert 1991, 1994).

A national study conducted in 1980 showed Americans roughly equally divided in positive and negative views of wolves and bears (Kellert 1985, 1980). More recent studies in Michigan, Minnesota, Idaho, Montana, and New Mexico and among visitors to various protected areas show widespread appreciation and concern for wolves and bears among a majority of respondents (Kellert 1996). A substantial minority express strong protectionist views toward the species. The wolf, in particular, has emerged as a powerful symbol for protecting and restoring wildlife and wilderness (Kellert 1996; Naess and Mysterud 1987). Popular books, films, and other media have celebrated the positive attributes and qualities of wolves and bears. Many new and established nongovernment organizations have marshaled considerable resources and public support for protecting and restoring wolf and bear populations. Studies among visitors to Yellowstone and Glacier National Parks have strongly endorsed the existence value of those species and supported the reintroduction and augmentation of populations in these areas (Bath 1987, 1991; Biggs 1988; Jobes 1991; McNaught 1987; Reading and others 1994).

Considerable ambivalence remains, however, among most demographic groups regarding wolves and, to a lesser extent, bears (Arthur and others 1977; Bath 1991; Buys 1975; Herrero 1970; Book and Robinson 1982; Jonkel 1975; Kellert 1996; McNamee 1984; Shepard and Sanders 1985). Largely hostile and exploitative attitudes occur among resource-dependent and rural groups, especially livestock producers (Herrero 1978; Kellert 1986, 1994, 1996; Pelton and others 1976; Stuby and others 1979). Negative, albeit less-antagonistic views have also been observed among the elderly and people with less formal education. Recreational hunters and commercial trappers often reveal positive attitudes toward and greater knowledge about these species than most groups, but are inclined to support their harvest and control. Various studies suggest that positive sentiments toward wolves and bears typically increase with geographic distance from either extent or proposed reintroduced populations (Llewellyn 1978).

Attitudes toward and knowledge of wolves and bears are often surprisingly

independent (Burghardt and others 1972; Kellert and others 1996; Murray 1975; Peek and others 1991; Petko-Seus 1985). Both people who are strongly in favor and those who are strongly opposed to protection and restoration often possess relatively high levels of knowledge of the species, but most people have limited factual understanding of either animal.

North American attitudes toward controlling wolf, bear, and other large-predator populations—especially to minimize livestock depredations—have been examined (Arthur and others 1977; Braithwaite and McCool 1988; Brown and others 1981; Buys 1975; Colorado Division of Wildlife 1989; Hastings 1986; Herrero 1978; Jope and Shelby 1984; Kellert 1985; McCool and others 1990; Petko-Seus and Pelton 1984; Trahan 1987). The majority of the general public favors limiting control to situations where substantial depredations have been clearly and convincingly demonstrated. Issues of culpability and humaneness (with respect to perceived pain and suffering) tend to be more important considerations than the cost of control. Much of the general public disapproves of indiscriminate population reductions that are independent of an individual wolf's or bear's guilt. They also strongly oppose using poisons, denning, or aerial gunning. Livestock producers and other resource-dependent groups are far more accepting of these latter control methods.

A number of economic implications relevant to wolf and bear management in Alaska can be drawn from those findings. Considerable potential consumer demand appears to exist among many groups of North Americans for visits to areas that have populations of wolves, bears, and other large carnivores. Results from the Boundary Waters Canoe Area, Glacier and Yellowstone national parks, and other protected areas suggest that the presence of those species constitutes a major attraction. Intensive wolf or bear management in Alaska conflicts with this potential consumer demand; additional data on Alaska tourism presented in the next section confirm this view.

Widespread public support for wolf protection and restoration in the lower 48 states and a general perception among the public of wolves and brown bears as imperiled species are likely to encourage advocacy groups to oppose wolf and bear control in Alaska. That could produce economic actions, such as support of a tourism boycott. A majority of the general public supports geographically limited predator control when a clear and compelling need has been demonstrated, when indigenous populations suffer from the absence of control, and if humane control methods are used. The need for convincing scientific data to support control programs and the use of animal-specific and humane control methods are both likely to increase management costs.

The Alaskan population is now largely urban: 67% of Alaskans live in cities. About 16% of the population is Alaska Native, the majority of whom reside in

rural Alaska. These Alaska Natives are economically and culturally dependent on subsistence hunting. Immigrants from other states amounted to an average of roughly 10% of Alaska's population each year during the decade 1980–1990; these immigrants tend to be much younger and to have more formal education than long-term residents of Alaska.

Various surveys that have assessed attitudes of the Alaskan people are discussed below. These surveys were conducted by different groups, including the Alaska Department of Fish and Game (ADFG), other government agencies, trade organizations, and citizen interest groups. They varied considerably in sampling techniques, survey methods, and data-collection procedures, so their quality and representativeness vary. Most relied on telephone data-collection procedures, which often underrepresent rural, especially Alaskan Native, population groups. All the studies were cross-sectional (conducted at one point in time) and thus reflect bias associated with the time of the year when the studies were conducted and the particular issues and events occurring then. Nevertheless, the results collectively present a crude approximation of the views and perceptions of the Alaska public and provide a useful perspective of prevailing wildlife attitudes and values in Alaska.

According to the 1991 National Survey of Fishing, Hunting, and Wildlife-Associated Recreation, 41.2% of the Alaskan population age 16 and older participated in hunting or fishing; this is a higher percentage than that of any other state; and much higher than the national average of 21%. Some 20% of Alaskans purchase a hunting license in an average year, but the percentage has been slightly and steadily declining for more than a decade. A majority of Alaskans report that they have at some point purchased a hunting license. Less than 10% of Alaskans oppose hunting. The proportions of Alaskans who have ever hunted and who object to the activity are, respectively, considerably higher and lower than in the contiguous 48 states. In 1991, about 62.1% of the adult Alaskan population participated in some form of primary nonconsumptive wildlife activity (viewing, photographing, or feeding fish or wildlife); this is substantially higher than that in most other states and much higher than the national average of about 40%.

The harvesting of furbearers (for example, more than 1,000 wolves each year) and the amount of subsistence hunting are far greater in Alaska than elsewhere in the United States (ADFG 1994; ADFG 1994–95; Miller and McCollum 1994a). Much of the latter activity is associated with the relatively large Alaska Native population.

The nonconsumptive value of wolves and bears is suggested by both statistical and anecdotal findings among residents and visitors to Alaska. According to one study, for example, the most popular and economically important species for viewing purposes were bears, wolves, and marine mammals, with a higher preference for viewing bears than wolves (Miller and others 1997). Alaska is the only place in North America where opportunities for viewing brown and black bears are readily available.

Some 96% of Alaskans believe that wildlife adds considerably to their ''enjoyment of living in Alaska" (Miller and McCollum 1994b). A majority of the Alaskan population report a desire for more state-managed areas for wildlife-viewing purposes, although only a minority of resident hunters support creating more wildlife-observation areas. Nearly three-fourths of resident hunters object to establishing such areas if it will result in closing areas to hunting. Tolerance of large carnivores among the Alaskan public is suggested by the fact that nearly half approve of bears living near urban areas, although only a minority of hunters support this idea (Miller and McCollum 1994a).

Most Alaskans endorse the hunting of bears, although a large majority object to baiting and attracting bears with food, and a majority disapprove of trophy hunting (the main objective of most bear hunting). Recreational wolf hunting generates little interest among Alaskan residents or nonresident visitors (Miller and McCollum 1994a), but wolves are extensively harvested for their pelts, and most Alaskans support this use. As with bear hunting, issues of "fair chase" and humane harvesting of wolves are important among the Alaska public. Most Alaskans object to using airplanes, denning, and poisons as methods of harvesting wolves, as discussed later (Alaska Wildlife Alliance 1992; Kellert 1980; Miller and McCollum 1994a,b,c).

Surveys of public attitudes toward controlling wolves and, to a lesser extent, bears to increase moose and caribou populations generate varied responses, depending on the wording of the questions and the group surveyed (Alaska Wildlife Alliance 1992; Miller and McCollum 1994a). In one study, a near majority (47%) of the Alaska public supported and 37% disapproved of killing wolves in some areas to increase moose and caribou populations. A majority of Alaska nonconsumptive wildlife users and female residents object to killing wolves to increase ungulate populations, but more than 60% of resident Alaskan hunters support this notion.

Strong support for wolf control has been found in areas that have experienced substantial decreases in moose or caribou populations presumably because of predators (Anderson 1995, Wolfe 1991; Gardner and Taylor 1992; Wolfe and Walker 1987). Some 90% of residents in game management unit 19D east, for example, support a reduction in wolf populations to increase moose numbers; more than 70% favor decreasing wolf populations for more than 5 years to maintain high moose numbers. A slight majority of residents of this area support brown bear control to increase moose numbers. Those views, as well as more ethnographic and anecdotal data from other areas of the state, indicate that a majority of rural, native, and especially subsistence groups support wolf control and harvest. The food, economic, and cultural importance of subsistence hunting and trapping is generally associated with support among Native populations for limiting wolf and, to a lesser extent, bear populations in areas where ungulates have declined presumably because of predators.

A majority of nonindigenous Alaskans, particularly residents of urban areas

and females, object to reducing wolf numbers (Alaska Wildlife Alliance 1992; Miller and McCollum 1994a). Those groups reluctantly endorse wolf control only in specific areas where clear and convincing data reveal substantial declines in moose and caribou numbers because of wolf predation. Moreover, one survey found that 43% of Alaskans believe that the overall number of wolves killed for control purposes should be reduced, whereas only 8% favored increasing the number of wolves killed for this reason. A majority of Alaskans remain unconvinced that reducing wolf numbers will increase moose and caribou hunting opportunities.

Public attitudes toward different methods of wolf and bear control have also been examined (Alaska Wildlife Alliance 1992; Kellert 1980; McCollum and others 1996; Miller and McCollum 1994a). Most data reveal an aversion to aerial techniques, as the outcome of a recent statewide referendum corroborates. Results from Alaska, as well as other areas of the United States, suggest that issues of fair chase and humaneness are major determinants of public opinion and often override economic considerations. For example, most people oppose poisoning, denning, and aerial gunning even when these techniques appear cost-effective. Far greater support exists for ground-based hunting, trapping, and snaring to kill wolves, although majorities of female, urban, and antihunting groups object to these control methods as well. Most Alaskans support wolf and bear control by local hunters rather than by professional wildlife personnel. Many, especially rural and native groups, believe that wolf and bear control should generate practical gains, such as meat and furs.

The views of key Alaskan groups suggest the following conclusions. Majorities of urban Alaskans, nonconsumptive users, younger people, females, and people with higher levels of formal education tend to object to wolf control and, especially, bear control unless a substantial problem of declining moose and caribou populations has been clearly and convincingly demonstrated in geographically specific areas. Those groups often object to control methods perceived as inhumane (for example, likely to cause pain and suffering) or unfair (for example, violating presumptions of a fair chase). The views of Alaska sport hunters tend to be more diverse. Strong support exists among resident sport hunters for wolf control and sometimes bear control in areas where substantial declines in large ungulate populations have occurred and that afford relatively easy road access for hunting purposes. Most Alaska sport hunters object to indiscriminate wolf reduction throughout the state and to control methods regarded as unsportsmanlike (for example, airborne hunting). Subsistence hunters, especially rural and native peoples, generally support wolf and bear control for cultural, economic, and food-gathering benefits.

For 2 reasons, the cost of a wolf and bear control effort is likely to increase substantially if prevailing attitudes among the Alaskan general public, especially urban residents and nonhunters, persist. First, most of the public objects to wolf and bear control except in specific geographic areas where a serious decrease in

ungulate populations has been clearly and convincingly related to wolf or bear predation and where serious economic or cultural harm has been inflicted on resident sport or subsistence hunters; demonstrating and communicating such a magnitude of effect will inevitably increase management costs. Second, most of the Alaskan public objects to control methods perceived as inhumane or unfair—for example, aerial control, denning, toxicants, and trophy hunting. Other control methods tend to be more costly, labor-intensive, time-consuming, or technically challenging. A partially offsetting economic effect might be associated with widespread support for wolf control by local people who derived economic and practical benefits from the activity.

Alaska has experienced a slight but consistent decline in the number of licensed hunters during the last decade and more. It might be due to decreasing ungulate numbers in areas readily accessible to large urban populations. But absolute and proportionate declines in the number of licensed hunters have been observed throughout the United States. If the decline in ungulate numbers in areas readily accessible to hunters has contributed to a decline in hunting participation, wolf and bear control could produce positive economic benefits. Brown bear control could also generate economic returns if conducted by nonresident trophy hunters, although widespread opposition to trophy hunting suggests that the meat and possibly pelts should be distributed to local populations. A greater harvest of ungulates resulting from wolf and bear control would produce substantial economic and cultural benefits for subsistence and native populations in rural Alaska.

Most nonconsumptive wildlife users and urban wildlife enthusiasts believe that intensive wolf and bear control reflects an entrenched bias among Alaska game management officials and hunting interests. That view encourages the use of legislative and judiciary methods for promoting policy change, rather than attempts to exercise influence on administrative agencies. The success of the 1997 ballot initiative involving aerial hunting of predators will probably reinforce the perception, which is likely to result in substantially greater management-associated costs because of the heavy economic burden created by using legislation and litigation as a means of achieving policy goals. Many nonconsumptive wildlife users also view ADFG as less interested in providing recreational viewing than hunting opportunities; this perception is likely to encourage opposition to user fees and taxes on nonconsumptive users as a way of generating additional wildlife agency revenues.

From an economic perspective, Alaska has at least 3 very striking features. First, it is a huge area—more than one-half million square miles—with a population of about 610,000, most of whom reside in and around 3 cities (Alaska Department of Labor 1996). Second, about 60% of the land is in federal ownership,

28% is in state ownership, and 12% is Native lands (figure 6.1). Only 1% is private, non-Native land. Third, natural resources and resource-based activities, such as petroleum extraction, tourism, and hunting, play a dominant role in the state's economy. The petroleum industry accounted for more than one-third of the gross state product (GSP) in 1993. More than one-fourth of all employment in the state is government-related, including federal, state, and local government employees. The seafood industry is the state's largest private employer and the tourist industry is the second-largest private sector employer, annually attracting nearly 1 million visitors who spend more than $1 billion of a total GSP of about $24 billion. Alaskans themselves are estimated to have spent about $89 million on hunting, $239 million on fishing, and $144 million on primary nonconsumptive wildlife activity in 1991.

Although hunting and tourism provide an important contribution to the Alaskan economy, there are 2 reasons why the data on expenditure and employment can be misleading in assessing the economic benefits of wolf control or the economic consequences of a tourism disruption triggered by public opposition to wolf control. The first reason has to do with the distinction between total and marginal economic impacts on the economy. The total impact is the entire contribution of a sector. The marginal impact is the increase or decrease in the contribution associated with some expansion or contraction in the activity within the sector. If what is at issue is the total elimination of a sector, whether hunting or tourism, the total impact is the appropriate measure to use. But if the change is smaller than that, the correct measure is the marginal impact of the change. For example, if it involves some increase in hunting or some decrease in tourism, the appropriate quantity is not the total value of all hunting or all tourism, but the value of the particular increase or decrease expected to occur. The marginal impact is harder to determine because it involves forecasting the magnitude of the specific increase or decrease in activity, as well as assessing the economic value associated with it. Nevertheless, this is the appropriate approach for evaluating the economic implications of predator control policies.

The second reason is that expenditure is generally not a correct measure of value, and the increase or decrease in expenditure is generally not an appropriate monetary measure of the increase or decrease in well-being. There is a basic distinction between financial or accounting revenues or costs and economic benefits or costs. Accounting revenues and costs reflect an assessment of monetary inflows and outflows according to the principles and conventions of accounting rules. Economic costs and benefits are intended to reflect a broader social perspective based on real costs, in the sense of resources actually used up, and real benefits, in the sense of actual changes in people's well-being. Although not unrelated to financial revenues and costs, economic benefits and costs represent a different standpoint for assessment.

The correct concept to use in an economic assessment of the costs of wolf control is the economic concept of marginal cost (Freeman 1993, Turner and others 1993). This should not be confused with average cost. The average cost is simply the total cost for some item divided by the quantity of the item; marginal cost is the change in cost per unit change in the quantity of the item. If all units of a commodity cost exactly the same, there is no difference between average and marginal cost; if not, there is a difference. Such differences can arise for several reasons, including the presence of fixed costs. By definition, where there are only fixed costs, the marginal cost is zero—since cost does not vary with the quantity of the item, there is zero increment in cost when the quantity is raised, and zero decrement when it is lowered. But, although the marginal cost is zero where there are fixed costs, the average cost is positive. It can make a substantial difference, therefore, whether one uses average or marginal cost. Marginal cost is generally the relevant concept for an economic analysis because it reflects the marginal impact; for the reasons given above, this is what the analysis should aim to measure. For example, if a hunter already owns some hunting equipment such as a rifle, the marginal cost of an extra hunting trip is the ammunition and the cost of gas needed to make the trip but not the cost of the rifle. Conversely, if somebody who did not previously hunt or own a rifle is now induced to take up hunting by an improvement in the moose population, say, then the cost of a rifle counts as part of the marginal cost of the increase in hunting associated with this person.

The economic concept of marginal cost also includes costs that are implicit rather than explicit. These are known as opportunity costs and they reflect the notion that the real cost of an item might be not what one pays directly to obtain it, but rather what one forgoes to obtain it. For example, in the case of a government agency with a fixed budget and a fixed staff, part of the cost of providing a new activity—say, wolf control—might be the other services that have to be reduced or postponed when agency personnel are diverted to work on the new activity. There could be no direct increase in payroll, because no extra staff are hired, but substantial opportunity cost in terms of something else forgone. Another example of opportunity cost is the time spent by a hunter in traveling to and from a site and hunting there. Depending on the situation, he might otherwise have spent his time in earning income at work or doing something else of value; if so, the opportunity cost of his time counts as part of the marginal cost of the hunting trip from an economic perspective.

Similarly, with benefits the correct concept to use is the marginal benefit, that is, the equivalent monetary value of the change in people's well-being as a

result of the program (Freeman 1993, Turner and others 1993). With marginal benefit, as with marginal cost, there are several important points of difference between economic and accounting concepts. A crucial difference is that, in economics a person's expenditure on an item is generally not a measure of the item's value to him. To be sure, there are some circumstances where expenditure is a measure of value. For small changes in marketed items that are freely divisible in quantity, the price a person pays for the last unit consumed is equated to the unit's marginal value. But for items that are not marketed or not freely divisible and for large changes in items that are marketed, the amount paid provides a bound on the value to the consumer but is not the same as this value. In economics, the difference between what an item is worth to a consumer and what he actually pays for it is known as the consumer's surplus from the item. The change in consumer's surplus is the economic measure of marginal benefit—the lost consumer's surplus, which is the economic measure of loss if the item is taken away, or the extra consumer's surplus, which is the gain if more of the item is provided.

In economics, there are 2 ways to express in monetary terms what an item is worth to a person, that is, 2 ways to define consumer's surplus: the compensating variation and the equivalent variation—more colloquially, willingness to pay (WTP) and willingness to accept (WTA). The WTP value of an item is defined as the maximal amount that a person would be willing to pay to ensure his continued access to the item over and above what he does pay for it; the WTA value is the minimal amount of money that he would be willing to accept to forego the item, above and beyond what he does pay for access to it. In general, the 2 measures are not equal; how close or far apart they are depends partly on whether things that money can buy are good substitutes for the item in question (Hanemann 1994).

All monetary measures of value in economics can be shown to be either WTP or WTA measures. The logic of the economic measure of value is that it expresses value in monetary terms through a determination of equivalence. Regardless of whether the item itself is a market good or can be obtained by spending money, its monetary value is defined to be the change in monetary income that would have an equivalent effect on the person's well-being. Therefore, changes in income are the currency in which economic welfare is measured.

How does this relate to expenditure? Suppose a person is able to go hunting at a cost of $150 per trip and elects to take 3 hunting trips per year. (This example assumes that the hunter can take as many trips as he wants at this price, with no constraint such as when there is only a limited number of permits for a particular type of hunting trip.) Thus, he spends $450 annually for hunting trips. However, he might be willing to pay something more than this if it were necessary. To make 3 trips, say, he might be willing to pay $500; his WTP measure of consumer's surplus would be $50. Or he might be willing to pay $1,000; his WTP measure of consumer's surplus would then be $550. In general, there is no

specific relationship between the amount of expenditure and the magnitude of the consumer's surplus. However, the amount of consumer's surplus on the last unit consumed is likely to be small; the last trip is likely to be worth something close to $150. It could hardly be worth less, because if it were, he presumably would not have elected to take it. It is unlikely to be worth substantially more, because if it were, he might have taken more trips than this. Hence, the value of the last unit to the consumer is likely to be close to its cost, and the consumer's surplus is likely to be small. But for units other than the last, the value can be considerably larger than the cost, and the consumer's surplus might be substantial. All that can be said in general is that if the cancellation of a predator control program makes it less attractive for a person to go hunting, the loss of consumer's surplus should be larger when the reduction in hunting activity is greater. But the loss bears no specific relationship to the expenditure on the lost trips; rather, it captures the hunter's loss of utility in addition to the expenditure foregone. (The discussion here deals with just the definition of the concept of monetary value in economics rather than its measurement, which is taken up in the following section.)

How is all this related to jobs and employment? Lay discussions of economic effects are often framed in terms of jobs gained or lost. Moreover, there is a type of economic analysis that is often performed in these circumstances to determine direct and indirect effects on employment, known as economic-impact analysis or, more technically, input-output analysis. This type of analysis can be reconciled with the economic concept of value discussed above up to a point. It should be noted that most discussions of employment focus on total employment, as opposed to the marginal increase or decrease in employment that would be associated with implementing or canceling a predator control program. For the reasons given above, it is the marginal impact on employment that is appropriate in an economic assessment. Furthermore, framing the discussion in terms of jobs versus personal income from employment and other sources involves mainly a difference in choice of units, not a substantive difference in the underlying item that is being measured. In addition, discussions of jobs are often confounded by assumptions about whether the workers would otherwise be employed elsewhere and whether the same needs would ultimately be fulfilled elsewhere.

For example, suppose that an analysis reveals that because of the abandonment of a wolf control program and the subsequent reduction in hunting activity, 1,500 jobs for hunting guides and outfitters would be lost in Alaska. Before one can reach any definitive conclusion, it must be determined whether any of those workers can be expected to obtain other jobs, as opposed to being permanently unemployed. To the extent that they do obtain other jobs, the relevant measure of loss is the net loss of personal income from employment in the new jobs versus the old, not their gross income in the old jobs. A similar issue arises with respect to nonresident hunters and visitors to Alaska. To the extent that they reduce their visits or hunting trips in Alaska as the result of a government program, from a national perspective the relevant measure of loss is the net loss of enjoyment they

receive as consumers after allowing for the other activities that they engage in elsewhere and the net loss of income from employment as a result of the reduction in their visits to Alaska after allowing for the income from employment generated by their increased activity elsewhere.

In summary, an economic assessment of a predator control program is couched in terms of the net gain or loss in personal income for the various parties affected by the program, allowing for the effect of employment and business profits on personal income, plus the net gain or loss in utility from the consequences of the program expressed in monetary terms as the change in income that is equivalent to this change in utility.

Assume that a predator control program does indeed trigger an increase in moose and caribou populations. If one thinks in terms of a balance sheet of the potential effects of a predator control program, there would be 3 main items: the marginal benefit of whatever increase occurs in the populations of prey species, such as moose and caribou; the marginal cost of planning and implementing the predator control program itself; and any marginal reduction in benefit associated with the decrease in the predator population. The benefits can be divided into five components:

1. The gain in utility for residents from increased recreation associated with moose and caribou.

2. The gain in utility for nonresidents from increased recreation associated with moose and caribou.

3. Any other gain in utility for residents arising from the increase in moose and caribou populations.

4. Any other gain in utility for nonresidents arising from the increase in moose and caribou populations.

5. The gain in personal income from employment and profits for residents resulting from increased recreation or tourism in Alaska by residents or nonresidents.

The costs similarly can be sub-divided into:

6. The costs to government agencies and others for planning and implementing the predator control program.

7. Any loss of utility for residents associated with the reduction in predator populations.

8. Any loss of utility for nonresidents associated with the reduction in predator populations.

9. Any loss in personal income from employment and profits for residents resulting from a reduction in tourism in Alaska by nonresidents triggered by their reaction to the predator control program.

Those benefits and costs are discussed further in the following 2 sections.

Because of their cultural and economic differences, the committee treats non-Natives and nonsubsistence residents separately from Native and subsistence peoples, and nonresidents separately from residents. This section presents an economic assessment of the impact on non-Native residents and nonresidents; the next section deals with Native and subsistence peoples. The first part of this section covers the impact of a predator control program on recreation associated with the prey species, moose and caribou. The succeeding parts deal with the impact on recreation associated with the predator species, wolf and bear; non-recreation impacts, including what are known as nonuse or passive-use values; the impact on resident non-Native Alaskans' personal income from employment and business profit; and the costs of predator control programs. In each case, we consider both what economic information is available and how it appears to have been used by ADFG in its decision-making process.

With respect to recreation by non-Native residents, the main activities likely to be affected by predator control programs are hunting and nonconsumptive wildlife-associated recreation (such as wildlife viewing and photographing). In each case, the increase in consumer's surplus associated with enhanced recreation is composed of a gain from an increase in recreation activity and a gain from an increase in enjoyment of the recreation activity. For example, if there are better hunting opportunities for moose, more people might take up hunting in Alaska and existing hunters might take more trips, go to different locations where conditions have improved, and obtain greater enjoyment from hunting at the locations where they normally hunt. Thus, there are changes both in the number of hunting trips and in the consumer's surplus per hunting trip.

Both those types of changes and the resulting change in aggregate consumer's surplus can in principle be measured by estimating a behavioral model of people's participation in hunting and choice among hunting sites, known in economics as a travel cost model. The model is so called not because it generates economic values on the basis of travel costs but rather because it is used to analyze hunter choices of how often to go hunting and where to hunt on the basis of factors likely to affect those choices, including the costs of hunting at various locations and the quality of the hunting experience there. By analyzing the implicit tradeoff between site choice and travel cost, it is possible to infer the monetary value of the hunting experience in terms of the WTP or WTA measures of consumer's surplus. This type of analysis is known in economics as revealed preference analysis—when one makes particular assumptions, people's behavior reveals their

preferences. In this context, travel costs are important not because expenditures are a measure of value, but because costs shed light on the tradeoffs that people face when they make recreational choices; it is their choices from which estimates of value are inferred. The use of travel costs for this purpose was first suggested by Hotelling (1931); the first to implement this approach were Trice and Wood (1958) and Clawson and Knetsch (1966). Behavioral models of recreation choices based on travel cost are now a standard component in environmental valuation (see McConnell and others 1990 or Freeman 1993).

A travel cost model of hunting or wildlife-viewing recreation could be used to predict both the change in hunter behavior (total hunting activity and allocation of the activity among alternative sites) and the change in consumer's surplus as the result of a predator control program that causes changes in the hunting season or the abundance of animals at various locations. For sportfishing in southcentral and southeastern Alaska, models like this were developed for ADFG by Carson and others (1990). Such models apparently have not yet been developed for hunting or wildlife viewing in Alaska.

In fact, it is not clear what type of economic analysis was performed by ADFG in connection with its wolf control programs in 1993 and 1995. Nevertheless, there is an active economics research program in the ADFG Division of Wildlife Conservation, and staff economists have collected data on the economics of wildlife-based recreation and prepared estimates of the consumer's surplus from hunting and wildlife viewing in Alaska (McCollum and Miller 1994). This has been applied to an economic assessment of game management strategies by Miller and others (1997).

We first summarize the findings from those studies and then comment on their applicability to the problem at hand. The estimates of consumer's surplus per trip are shown in table 6.1, which is based on material supplied to the committee by ADFG. The estimates were derived from 2 1992 surveys conducted by ADFG about wildlife in Alaska. One survey was administered to a large sample of resident voters (intended as a proxy for the adult resident population) and focused on their attitudes to wildlife and their involvement in nonconsumptive wildlife use in Alaska in 1991, especially wildlife-viewing trips. The other survey was administered to a large sample of resident and nonresident hunters who had taken out Alaskan hunting licenses in 1991; it focused on their attitudes to wildlife and their involvement in consumptive wildlife use in Alaska, especially big-game and waterfowl hunting. Both surveys selected a specific trip by the respondent and asked whether the respondent felt that the trip had been worth the cost (in the sense that they would take it again). If the answer was yes, the respondent was asked how much more they would have been willing to pay for the trip. That question uses an open-ended format, also known as a continuous-response format. An alternative approach presents respondents with a specific cost increment and asks whether they would have been willing to take the trip at

that price; this is known as the closed-ended or discrete-response format. Both formats were used in the course of the ADFG survey.

This is what is known in economics as a contingent valuation question. Contingent valuation generally is the process of measuring monetary value by eliciting it directly through some form of survey. It is an alternative to valuation through inference from behavior, such as the travel cost model approach mentioned above. Its use for this purpose was first suggested by Ciriacy-Wantrup (1947), who suggested that, just as people vote on taxes or bonds for state and local government programs, they could be asked how they would vote or what they would be willing to pay for something promoting the use or protection of the environment. This approach was first implemented by Davis (1964) to value recreation in the Maine woods. With the use of travel cost, it has become the main method of environmental valuation (Carson and others 1990; Carson and

Mitchell 1989). In recent years, there has been some controversy about the reliability of contingent valuation to measure what are called nonuse values—these are discussed further below (Diamond and Hausman 1994 and Hanemann 1994). But most critics view contingent valuation as valid for measuring the value of activities like recreational behavior (Desvousges and others 1993).

The estimates of WTP in table 6.1 are broken down by wildlife species, by type of recreation (hunting, primary nonconsumptive, and secondary nonconsumptive), and by type of participant (resident and nonresident). For moosehunting trips, the average consumer's surplus was $181 per trip for resident hunters and $393 for nonresident hunters—these are what hunters would have been willing to pay beyond what they actually paid. For residents viewing moose on a nonconsumptive recreation trip, the consumer's surplus was $123 per trip. For caribou hunting, the average consumer's surplus was $168 per trip for resident hunters and $432 for nonresident hunters; for residents viewing caribou, the consumer's surplus was $135 per trip. (It is not possible to tell from the information made available to the committee whether any of the differences in value between moose and caribou are statistically significant.) For wolf hunting, the average consumer's surplus was $1,500 per trip for resident hunters and $351 for nonresident hunters; for residents viewing wolves, the consumer's surplus was $195 per trip. These estimates of consumer's surplus per recreation trip are broadly consistent with other published estimates of the consumer's surplus from similar types of recreation. They are also broadly consistent with other types of information on the value of hunting in Alaska, such as what people are willing to pay for permits to hunt on private land; on some privately owned lands, use fees exceed $2,000 per hunter (Boyce and others 1995).

It should be emphasized that the values in table 6.1 are not differentiated by where the trip occurred or the type of hunting or viewing conditions (including animal abundance) at the site of the trip; they are averaged over different sites and different conditions of animal abundance. Thus, they should be thought of as estimates of the average consumer's surplus per trip for a typical trip of each kind. That has 2 important consequences. First, one cannot directly infer an estimate of the total consumer's surplus from wildlife viewing or hunting in Alaska. The consumer's surplus per trip almost certainly varies across trips: the last trip is generally less valuable than the first trip. If one multiplied the values in table 6.1 by the total number of trips of each type, this would almost certainly understate the total consumer's surplus from wildlife viewing or hunting in Alaska. Second, there would be some inaccuracy in using these values alone to estimate the recreational benefits of a changed pattern of wildlife viewing and hunting resulting from game management or predator control programs in Alaska. That is because the values in table 6.1 do not account for changes in the consumer's surplus per trip resulting from improved conditions at a given site and site substitution whereby a person wanting to hunt or view wildlife switches from a site that he would have visited before to another site that now becomes

more attractive as a result of the increase in wildlife population at that location. When there is site substitution of this form, the gain is the increase in consumer's surplus resulting from having to travel less far because one takes advantage of a newly improved site or resulting from obtaining enhanced enjoyment at the site where the activity is pursued. This increase could be more or less than the existing average consumer's surplus per trip and cannot be inferred from it.

In addition to uncertainty about the increase in consumer's surplus per trip, there would also be some uncertainty about the impact of a game management or predator control program on the pattern of recreational behavior, both the total volume of wildlife viewing or hunting trips and the allocation of this total among alternative sites—uncertainty that can best be resolved with the aid of a model like the travel cost model described above.

Most existing analyses of game management or predator control programs appear to have used other approaches that are questionable. One unreliable approach is to assume that there is a net increase in recreation activity that is directly proportional to the increase in animal population—if the animal population rises by 10%, there will be 10% more viewing trips or hunting trips. That approach rests on a chain of assumptions, each of which is dubious:

• The assumption that if the animal population at a site increases, the attractiveness of that site for recreation increases in the same proportion—it becomes 10% easier for viewers to spot an animal or for hunters to catch one. Instead, however, there might be non-linearities or threshold effects that lead to a nonproportional impact on the attractiveness or quality of the recreation experience.

• The assumption that if the attractiveness of a site increases, visits to that site increase in the same proportion. However, consider 2 sites of equal attractiveness that split a recreation market 50:50. Suppose that one of the sites becomes perceptibly superior to the other because of a 10% increase in the chance of viewing wildlife there. Conceivably, the entire recreation market could now switch its patronage to the improved site, leading to a doubling in visits there. Moreover, the enhanced attractiveness of the recreation opportunity would conceivably induce new people to take up the activity or existing participants to increase their participation. Those are merely examples that illustrate the general point that the impact on visitation can be disproportionally larger or disproportionally smaller than the change in site attractiveness.

• The assumption that a proportional increase in visits to a site implies an increase in total participation in the activity. Some of the increase might represent substitution of existing trips away from other sites, as in the above example.

For all of these reasons, therefore, one cannot generally assume that the total volume of recreation activity will change in proportion to the total size of the animal population. For consumptive uses, such as hunting, a variant of that approach is to estimate an allowable increase in the total harvest on the basis of

biological considerations and then assume a proportional increase in hunter effort (recreation activity). Alternatively, there could be retrospective data on harvest by hunters, but not on hunter effort. Boertje and others (1996) used retrospective data on the increase in harvests in areas with wolf control programs to infer an increase in hunting activity. But, that assumes that the catch rate (or its inverse, the amount of hunter effort per moose harvested) remains constant. A firm foundation for the assumption does not exist.

The approach of inferring the level of hunting activity from data on harvests is sometimes defended on the basis of the observation that the number of applications for special hunting permits greatly exceeds the available supply. It is argued from that that the availability of animals is the limiting factor on hunting activity and that therefore the level of hunting activity should increase in proportion to the size of the allowed harvest. It is certainly true that when access to particular types of hunting is regulated through some sort of quota system, such as a special permit, it is very likely that there will be some increase in hunting activity if the allowed harvest is increased. But there is still the question of substitution: some of the increased hunting activity might represent a diversion of trips away from other types of hunting or outdoor recreation. Moreover, the magnitude of the excess demand for regulated permits could yield a distorted impression of the true size of the ''reservoir" of frustrated latent demand for hunting. If applying for a permit and not getting one is relatively inexpensive, and if a hunter is at all risk-aversive, he will apply for more permits than he actually expects to use. For example, he might apply for 4 permits but expect to take only 1 or 2 trips.

In this connection, some of the people who testified to the committee linked the decline in the sale of resident hunting licenses in Alaska over the last decade to the decline in moose and caribou populations in some areas favored by hunters. There might be such a causal connection, but it is likely to be confounded by some other factors—over the last decade, there has been some decline in participation in hunting throughout the United States, including areas where animal populations have not declined. Some of the decline reflects a generational change in habits and interests away from hunting and other consumptive uses of wildlife. And some of it might reflect increasing demands on people's time that make hunting more difficult for them. The presence of such confounding factors makes it all the harder to predict how much of an increase in hunting to expect if animal populations are increased.

Even when there are direct retrospective data on changes in the volume of hunting activity, there can be important pitfalls of interpretation. Suppose, for example, that there are 2 areas, 1 in which a wolf control program is conducted and the other with no wolf control program. Suppose, further, that there are data on moose-hunting activity in the 2 areas that show that hunting declined in the area with no wolf control, but remained roughly constant in the area with wolf control. Reid and Janz (1995) confront a scenario like this on Vancouver Island.

They compute the benefits of wolf control for black-tailed deer hunting by assuming that without the program the volume of deer hunting in the wolf-control area would have declined at the same rate as it did in the no-wolf-control area. They calculate the extra days of deer-hunting activity in the wolf control area by using this assumption and multiply the result by an estimate of the consumer's surplus per day of deer hunting. Their analysis requires an assumption that the 2 areas are sufficiently similar that any difference in deer hunting can be attributed solely to the presence of the wolf control program in 1 area. More important, the analysis ignores the question of substitution among hunting sites. It is possible, for example, that some or all of the increase in moose hunting in the wolf-control area is simply a diversion of hunting activity that otherwise would have taken place in the no-wolf-control area. In that case, although there is still a gain in utility for deer hunters, it is likely to be some fraction of the consumer's surplus per hunting trip rather than the full consumer's surplus, as assumed by Reid and Janz. What should have been calculated is the increase in consumer surplus per deer-hunting trip diverted from other areas to the wolf-control zone.

In summary, the 2 core problems in assessing the economic benefits of any increase in ungulate populations for hunting and wildlife viewing are to predict the change in recreation behavior—both the total level of recreational activity and its allocation among sites—and to estimate the increase in consumer's surplus from recreation that is associated with this change in activity. Neither of those can be assessed with biological models of animal populations. Instead, both require a model of people's preferences and behavior along the lines of the travel cost model mentioned above.

It might be useful to give an example of how such a model can be used, on the basis of application of the ADFG model of sportfishing in south-central Alaska for a case study of the economic impact of closing the Kenai River to sportfishing for king salmon during the last week of July (Carson and others 1990). Fishing for king salmon on the Kenai River accounts for about 8.5% of all fishing at that time of year by residents of south-central Alaska. It was estimated, however, that the closure would reduce their number of fishing trips by only about 1.5%, because most of them would still go fishing but switch to other species and other sites. With the closure, it was predicted that salmon-fishing trips would fall from about 53.5% of all resident fishing trips at that time of year to about 50%. Among salmon-fishing trips, anglers would switch from king salmon to other types of salmon. Anglers still choosing king salmon as their target species would switch to other sites; for example, the share of king salmon trips going to the Kasilof River on the Kenai Peninsula would increase from 11% to 45%, and the share of king salmon trips to the ocean off Deep Creek on the Kenai Peninsula would increase from 4.5% to 13%. The main impact of the change is a reallocation of trips to other species and sites, and the loss of consumer's surplus is smaller than it would be if the main impact were a reduction in total fishing itself. If the impact were a reduction in total fishing, resident anglers'

loss of consumer's surplus from closing the Kenai River to king salmon fishing in the last week of July would amount to about $1.2 million; with the substitution of other species and sites, the loss drops to about $482,000.

Several thousand nonresidents visit Alaska each year to hunt—at least 7,162 in 1991. At the conceptual level, the economic issues that arise in estimating a monetary measure of the gain in utility for nonresidents who benefit from improved recreation for prey species, such as hunting and wildlife viewing, are the same as those discussed above for residents. The core problems are to predict the change in recreation behavior—both the total amount of their recreation in Alaska and its allocation among alternative sites—and to estimate the change in their total consumer's surplus from recreation in Alaska. The actual change in behavior, however, and the resulting change in consumer's surplus are likely to be quite different for nonresidents and residents because they are different types of people and engage in different types of recreation. For a nonresident, visiting Alaska to hunt or to view wildlife is a fairly substantial undertaking that might occur only once or twice in a lifetime. At home, most of a hunter's trips might be 1-day or 2-day trips; but he will plan a much longer visit to Alaska, probably involving hunting at several locations. Moreover, most or all of the hunting trips at home are without a guide, whereas nonresidents commonly hire a guide to hunt in Alaska; indeed, it is required by state law for nonresident hunting of brown bear, Dall's sheep, and mountain goats. The transportation costs and the use of a guide make hunting in Alaska much more expensive for nonresidents than for residents. On a brown bear hunting trip to Kodiak Island, for example, ADFG found that nonresident US citizens spend an average of $9,545 per trip, of which nearly $6,500 is for the guided hunt, about $1,000 for transportation, and another $1,000 for taxidermy (Loomis and Thomas 1992). A study by Yu (1991) found that 90% of the clients served by guides in Alaska were nonresidents and 92% of guides' revenue came from nonresidents.

Because hunting and wildlife viewing in Alaska are different types of recreational experiences for nonresidents and residents, it is necessary to estimate separate models of recreation participation and site choice for nonresidents and residents and to conduct separate analyses of the recreational impacts of a predator control program. ADFG has a separate model of sportfishing by nonresidents in south-central Alaska (Carson and others 1990) but not for other recreational activities.

The effect of a predator control program on resident or nonresident recreation involving predators needs to be factored into analyses with the recreational

impact involving prey species. The general issues are to predict the change in the level of recreation participation, the impact on site choice, and the resulting change in consumer's surplus. In the case of predators, however, some types of recreation might be affected differently—viewing of wolves might be reduced whereas, depending on how the control program is implemented, hunting of wolves increases. The reduction in opportunities to view wolves could be important, in that ecotourism and wildlife viewing are the fastest growing sectors of the Alaska travel industry. For example, more than 40% of tourists to Alaska cite wildlife viewing as important to them; 10% report being frustrated by not seeing the numbers and types of wildlife that they had hoped to see. At the Alaska Wolf Summit, Bob Dittrick, of Wilderness Birding Adventures, stated: "A large percentage of the inquiries we receive ask about the potential of seeing wolves. And many place it as their main personal goal for the trip even though I am selling birding trips. The tourists coming to Alaska want to see all the different wildlife, but from my experience wolves top the list. Given the choice, these people would give up seeing lots of wildlife for the chance of seeing one wolf. This is also true of Alaskans who vacation in the state." As noted earlier, data from the Boundary Waters Canoe Area and Glacier and Yellowstone national parks suggest that the presence of wolves, bears, and other large carnivores is a major attraction for visitors. Moreover, the sport-hunting value of brown bears, largely frustrated in the lower 48 states by Endangered Species Act restrictions, could motivate a small number of trophy hunters to visit Alaska if additional opportunities were provided.

The information available to the committee suggests that, in planning the wolf control programs in 1993 and 1995, ADFG did not conduct any analysis of the economic impacts on wolf-related recreation by residents or nonresidents.

In addition to the impact on both recreation involving both predator and prey species and on personal income from employment and profit in Alaska, to be discussed later in this chapter, people might be affected by predator control programs in various other ways that reflect their concerns and involvement with these animals. Use of the animals as a source of food or hides will be discussed in the following section, but both residents and nonresidents might experience a gain or loss of utility from a predator control program for reasons separate from their interest in recreation, meat, hides, or other such uses of the animals. For example, they might care about the prey species and want it to be preserved, regardless of whether they themselves plan to view it, hunt it, or eat it. Or they might care about the well-being of Native peoples whose way of life depends on the prey species. If people feel like that, they will experience a gain or loss of utility from a predator control program that must be accounted for in an economic

analysis, just as gains or losses associated with impacts on recreation, food, and personal income are counted.

In economics, this is known as existence value, nonuse value, or passive-use value—the value that people place on an item for motives that are separate from their interest in using it. The concept was first suggested by Krutilla (1967), who observed that people with no interest in using an item might still believe that it should be preserved and be willing to pay something to ensure its preservation. Their value will not be reflected in their use of the item, but it could be reflected in other forms of behavior, such as engaging in political or charitable activities or engaging in a consumer boycott to express displeasure. Consequently, existence value cannot be discerned by analyzing people's demand to use the item through the revealed preference approach described above. Conventional market data are inadequate to measure this value. Instead, one must use an approach that can directly elicit people's preferences.

In the case of predator control in Alaska, evidence strongly suggests the existence of substantial nonuse values, at least for some segments of the resident and nonresident populations. The widespread public support for wolf protection and restoration in the lower 48 states and a general perception of wolves and brown bears as imperiled species, as well as the public controversy about wolf control within Alaska, point to that. However, the magnitude of the nonuse values which exists for both prey and predator species, and includes concern by non-Natives for the well-being of Native peoples in the absence of predator control, is unknown.

Some critics have argued that nonuse values cannot be reliably measured through contingent valuation surveys. The question was considered by a panel of experts, headed by Nobel laureates Kenneth Arrow and Robert Solow, convened by the National Oceanic and Atmospheric Administration (NOAA). The panel asked to advise NOAA on whether contingent valuation is capable of providing reliable information about nonuse values in the context of assessing damages from oil spills. The panel concluded that contingent valuation studies can produce estimates reliable enough to be the starting point for a judicial or administrative determination of natural-resource damages, including lost nonuse value (Arrow and others 1993). It stipulated that, to be acceptable for this purpose, contingent valuation studies should adhere to guidelines dealing with various aspects of survey design and implementation.

No contingent valuation study on nonuse values associated with predator control has been conducted in Alaska. However, a contingent valuation study on nonuse values associated with wolf recovery in Yellowstone National Park was commissioned by the US Fish and Wildlife Service for its environmental impact statement on wolf reintroduction into Yellowstone and central Idaho (Duffield and Neher 1996). A telephone survey was conducted of 2 random samples of households, a national sample and a sample drawn from the 3-state region (Idaho, Montana, and Wyoming) where wolf reintroduction was to occur. Because of the

divided nature of opinion on the matter, an initial screening question was used to elicit whether the respondent favored or opposed efforts to reintroduce wolves. Depending on the response, two separate lines of questions were asked. Those in favor of wolf reintroduction were asked whether they would be willing to buy membership in a trust fund being established to support wolf reintroduction; those opposed were asked whether they would be willing to buy membership in a trust fund being established to oppose wolf reintroduction. In each case, a discrete-response format was used—respondents were told that membership in the trust would cost a specific amount and asked whether they would be willing to pay it. From the yes or no responses, the researchers derived an estimate of the mean or median WTP to promote or oppose wolf reintroduction among members of the population.

The survey found considerable differences of opinion nationally and in the local region. In the local region, opinion was quite closely divided—49% favored wolf reintroduction, 43% were opposed to it, and 8% did not know. National opinion was more lopsided—57% favored wolf reintroduction, 29% were opposed to it, and 14% did not know. The estimate of mean WTP was higher among supporters than opponents of wolf reintroduction and higher in the local region than nationally. Estimated mean WTP for wolf reintroduction among supporters was $20.50 per person locally and $8.92 nationally; estimated mean WTP to oppose wolf reintroduction among opponents was $10.08 per person locally and $1.52 nationally. As a further precaution, the researchers reduced those values by about 70% to allow for survey evidence that, with charitable contributions, people typically pay only about 30% of what they say they would pay.

Two other findings from the study of wolf reintroduction at Yellowstone are that use values are a minor and unimportant component of the benefits; instead, the conclusion is determined by nonuse values. Second, the geographic scope of the analysis and whose values are taken into account make a substantial difference. If the analysis is restricted to the local region, the costs of wolf reintroduction in fact outweigh the benefits, and the program would not be economically justified. If the analysis is broadened to the national population, the costs stay the same but the benefits are greatly expanded, and the conclusion overwhelmingly supports wolf reintroduction.

The conclusions of a study for Alaska are unlikely to be the same as those for Yellowstone, because the circumstances are different. In particular, there is no counterpart at Yellowstone of the impact on Natives' cultural values and lifestyles, which could turn out to be an important component of the evaluation of predator control in Alaska. As with Yellowstone, though, the conclusions of a study in Alaska are likely to be sensitive to the geographic scope of the analysis and the determination of whose values should be taken into account.

According to a survey conducted in December 1992 just before the wolf control program was suspended, around 80% of Alaskans thought that tourism and the visitor industry are critical to the well-being of the state (AWRTA 1995). Also, 40% of residents believe that the negative effect of the boycott on the state's tourism industry was a sufficient reason to cancel the program. The national boycott of tourism in Alaska certainly appears to have played a role in Governor Hickel's decision to suspend wolf control in 1992. The actual effect of the boycott on the Alaskan economy might have been quite small, but there were 2 mitigating factors: the governor moved quickly to stop the wolf control program, and it was ended before the height of the tourist booking period, which is mid-January through February. Surveys of tourist-related businesses conducted during and after the boycott suggest that, had it continued for longer, it would have caused a substantial reduction in tourism in 1993.

Depending on the circumstances and the method of implementation, a similar boycott of tourism in Alaska is likely if a wolf control program is again authorized. Because the situation is so variable, it is impossible to predict with precision the extent of a future boycott or the magnitude of its effect on the state's economy. Nevertheless, even if a boycott turned out to have only a modest effect on the economy, one cannot know this in advance. From an economic point of view, the measure of loss is the net reduction in personal income, whether from employment or from profit. Some of the tourist expenditures represent costs of produced inputs rather than personal income from employment or profit. In the case of sportfishing in south-central Alaska, which was estimated to generate expenditures in the state by residents and nonresidents amounting to about $93 million annually, only about $18 million was personal income in Alaska (Carson and others 1990). In addition to that direct impact, there are indirect and induced impacts in the rest of the Alaska economy, estimated for sportfishing to amount to another $47 million of personal income in Alaska. Those estimates are subject to the comments above about marginal versus average coefficients and offsetting changes in economic activity elsewhere in the economy. The same would apply to recreation losses from a tourist boycott. To the extent that jobs and income lost from a tourism boycott are made up by jobs and income gained in the activities that tourists substitute for a trip to Alaska, the net loss is reduced. If the substitution occurs out of state, from an Alaskan perspective there is a loss; from a national perspective, there is not.

In addition to any loss of personal income from changes in tourism and any loss of utility for residents or nonresidents associated with reduction in predator

populations, there are costs to government agencies and others associated with planning and implementing a predator control program. ADFG has prepared summaries of its costs for previous wolf control programs, but not specifically for the 1992 or 1994 program.

Three points should be noted about the calculation of such costs. First, the relevant concept is the increase in cost due to the program, that is, its marginal cost. Second, in this context opportunity costs might be more important than direct costs. As noted earlier, for an agency with a generally fixed budget and staff, part of the increased cost of providing a new activity might be the other services that have to be reduced or postponed when agency personnel are diverted to work on the new activity. To the extent that a program creates troublesome precedents for an agency or limits its future range of options, these too are opportunity costs associated with the program. Third, if fees are used to finance a program, they must be treated with caution because fees might be a transfer, rather than a real social cost.

This section examines the possible impacts of wolf control on rural Alaskans, those living in generally isolated communities where subsistence activities can be an important element of the community's economic base. About 20% of Alaska's population—about 124,000 people in 1995—lives in rural areas (Wolfe 1996). Those people reside in about 225 communities, most of which are off the road network and have fewer than 500 people. Only about half the rural population is Alaska Native, but most small villages are predominantly Alaska Native.

This section begins by considering how rural Alaskans might be affected by predator control. It requires an overview of the social, economic, and cultural setting for this population—particularly the importance of subsistence activities. The next part examines the economic value of wildlife resources to rural Alaskans. The section ends with a discussion of those issues from other (noneconomic) social-science perspectives, in particular applied anthropology.

Rural Alaskans support themselves through fishing, hunting, gathering of wild foods, and small-scale cash employment. Wolfe (1984, 1996) has characterized this as a mixed subsistence-cash economy. Wildlife resources are more than food sources and part of the economic base of these communities. It should be noted that from the Alaska Native perspective, subsistence is not just taking food from the local environment, but is an essential part of the Alaska Native way of living and entails a wide range of practices and beliefs, including spirituality, community, sharing, connection to and respect for other life, connection to the seasons, and teaching the young and moving them into their full inheritance

(Merculieff and Soboleff 1994). Predator control affects rural residents directly and in several ways; not only are prey species, such as moose and caribou, used and valued, but so are the predators themselves. The impacts on urban residents can be usefully categorized as impacts on consumptive or nonconsumptive recreational uses and passive use, but the impacts of wildlife management on some rural residents are more pervasive, in that wildlife-related subsistence activities constitute their life and livelihood.

Residents of rural Alaska make greater use of wild foods than communities anywhere else in the United States. Wolfe (1996) estimated that about 43.7 million pounds of wild food are harvested each year in rural Alaska. That amounts to 375 pounds (usable weight) per person annually in rural Alaska and 22 pounds per person in Alaska's urban areas (Wolfe and Bosworth 1994). By comparison, the average American uses about 222 pounds of store-bought meat, fish, and poultry each year. Wolfe and Walker (1987) summarized the data from the ADFG Division of Subsistence harvest surveys in 98 of the rural communities. Per capita harvests vary widely from relatively urbanized communities like Kenai (38 pounds) to remote villages like Hughes on the Koyukuk River (1,498 pounds). The most productive areas are along the interior rivers and the Arctic Coast. The composition of harvests varies widely across the state, from predominantly caribou and sea mammals in the north to salmon and moose in interior villages.

Fishing and hunting for subsistence provide an economic base for many rural regions. Case studies of rural Alaska's mixed subsistence-cash economy include Behnke (1982), Fall and others (1986), and Wolfe (1981, 1984). Similar analyses have been applied to the Canadian North (Usher 1981). In those economies, fishing and hunting are the central activities in the community and are conducted by family groups with efficient small-scale technologies, including fishwheels, gill nets, motorized skiffs, and snowmobiles (Wolfe and Walker 1987). Subsistence is augmented by cash employment commonly in commercial fishing, trapping, public-sector employment, services, and construction. The cash income enables families to purchase snow machines, fishing nets, rifles, and ammunition. The monetary incomes typically earned in villages are not large enough to support the family unless a portion is used in subsistence fishing and hunting (Wolfe 1991).

Subsistence resources are clearly important as food and an economic base in rural Alaska, but they also have a cultural meaning. One aspect of this is that these economies differ markedly from most modern market economies in terms of distribution. Harvests of subsistence resources are not sold in markets but are distributed on a kinship, status, role, or other basis to people in the community and nearby communities. Sharing is an important basis for reciprocity relationships, status, and identity.

Subsistence is broadly integrated through links with traditional beliefs and practices. Each village has a unique "seasonal round" of activity that is driven by the natural cycle of availability and quality of food. In the interior, this can

include the greening of the first plants, return of waterfowl, early king salmon runs, summer runs of chum and sockeye, berrying, and fall runs of silver salmon. In the fall, moose are fat and in prime; caribou might be migrating through. The winter can be a time when pelts are prime and trapping is most productive. Subsistence practices as a system combine many aspects of life, such as work and sport, that are separate for most contemporary Americans. Hensel (1994) reports that when older Yup'ik men in western Alaska were asked "What do you do for fun?" and "What do you do for work?" one generally got the same answer: ''hunting and fishing."

Some villages are more dependent than others on species that can be affected by wolf or bear predation. Predation by wolves or bears has minimal impact on major food sources for coastal communities that depend largely on marine resources or river-based communities with abundant runs of anadromous fish. But some communities are potentially vulnerable to the influence of predation by wolves or bears. Stokes and Andrews (1982) provide a discussion of the importance of moose to Nikolai and Telida, small villages in the upper Kuskokwim with populations of about 90 and 30 people, respectively, in 1981. At that time, Telida had no store, and the store in Nikolai depended on delivery by air from McGrath. In the fall of 1981, 52 moose were harvested by the 37 households of these 2 communities. As in most years, moose harvested in the 1981 season made up the major portion of the winter protein needs of people living in the area. Moose is, by far, the largest and most consistent and culturally important source of protein in the local food supply (Stickney 1981).

Residents expend a substantial cost in terms of both effort and money to acquire moose, in some cases up to 20% of annual household income just for operating expenses (primarily gas for boats). As Stokes and Andrews note, the prominence of moose at traditional ceremonial occasions, such as potlatches, is an indication of the special cultural values ascribed to moose. Caribou populations in the early 1980s were quite low in the area. Salmon contribute to the local diet but are in relatively poor condition in the upper Kuskokwim as they approach the spawning areas. In addition, ADFG regulations in the middle 1960s eliminated the traditionally used king salmon fence and trap, and that has resulted in lower harvests (Stokes 1983). Stickney (1980:11) summarized the importance of moose to Nikolai as follows: "Because moose is the most important item in the village diet the villagers will tolerate an expenditure of their limited cash resources to subsidize their hunting venture. This is not a luxury to them, but rather a necessity they don't question. Without a moose they must fall back on the store's expensive commodities without the sufficient cash base to provide a sufficient protein…diet. Due to supply problems the store cannot always be relied on to provide a nutritious adequate diet even if the cash were available."

As Klein (1989) has observed, fish and wildlife are essential to the subsistence way of life of many northern peoples. It is obvious that without these local food sources, the subsistence communities could not exist. The "importance" of these resources includes their economic contribution to the communities that use them and the more subjective psychological well-being "derived from a sense of economic security, and the cultural traditions or spiritual values that are interwoven with the resources" (Klein 1989:99).

In principle, the economic value of an incremental change in these resources is the equivalent in monetary income that would make a given person just as well off. The problem is to find a method to measure the equivalence. For marketed resources, the value of the last unit of a good that is consumed can be closely approximated by the market price. A person is observed to have made the tradeoff of money and the good, and one can infer that the good is worth at least as much as the price of the last unit taken. However, many or most subsistence resources are not traded in established markets. Subsistence resources are distributed by sharing, rather than sale, so local markets do not exist. Aside from problems of isolation, there are laws in most states, including Alaska, that prohibit the sale of most wildlife resources.

Even where some important resources are available in markets (such as markets for salmon or halibut), the resulting prices are generally not relevant for those who do not participate in these markets. This includes most residents of rural Alaskan villages. The price of king salmon in Anchorage provides little information on what the salmon is worth to a resident of Nikolai and is perhaps no more relevant than the price of sockeye salmon in Tokyo. Nonetheless, the most common approach to valuing subsistence resources is to use the replacement cost based on market prices. That is clearly inappropriate, in that the cost of the replacement might have little or nothing to do with the actual change in economic well-being that a person realizes by having or not having the given resource. Although the approach is in principle wrong, estimates based on it have been used in the context of environmental impact assessment (Usher 1976), in litigation (Duffield 1997), and descriptively (Wolfe and Bosworth 1994). For example, Wolfe and Bosworth (1994) assume a replacement cost of $3–$5 per pound and estimate the replacement cost of all wild food harvests in Alaska at $131.1–$218.6 million. Although those numbers are questionable, they indicate the relative importance of wild foods to Alaska Natives. The per capita cash value of subsistence foods in the rural interior is estimated to be $3,063 per person, or about half the 1990 per capita income for Native families of $6,205. Of course, for any analysis of a change in management, the relevant issue would be not total values, but the marginal value of changes.

The problem with interpreting replacement cost as the marginal value of a

subsistence resource is compounded by the problem of choosing the appropriate replacement or substitute. Many subsistence products have no commercially available equivalents, and available "substitutes" might be judged inferior on the basis of taste, texture, amount of fat, and possible presence of pollutants (Hensel 1994). The use of a range of $3–$5 per pound in some applications is based on retail prices of beef or salmon, which might or might not be equivalent to the resource at issue. In the Exxon Valdez case, the issue of using exact replacement of all subsistence resources, including marine mammals, led to average prices in 1994 of $10–$14 per pound in the settlement of the Alaska Native claim (Duffield 1997).

Brown and Burch (1992) reviewed the sparse literature on the economic value of subsistence harvest. They suggested a microeconomic model of subsistence harvests that has 2 components of value: the demand for participation in the activity and the demand for the product. Whether or not that is a useful way to frame the issue, the problem of estimating the components remains. The authors reviewed the range of possible methods, including travel cost and contingent valuation, and concluded that the latter, at least in principle, could provide relevant measures.

Wolfe and Walker (1987) analyzed a database that included both per capita subsistence harvests and income. They developed a model that shows a systematic tradeoff between those 2 components of the mixed cash-subsistence economies and provided an interpretation concerning factors that affect economic development in Alaskan rural villages. Not surprisingly, in areas where subsistence harvests are quite high, cash incomes tend to be low and vice versa. That is consistent with what one would expect because people have a fixed amount of time to allocate to different activities and the types and value of available work differ among communities. Duffield (1997) used the Wolfe-Walker data to calculate the marginal value of subsistence harvests based on the observed tradeoff of income and subsistence harvest. That approach, which used wage differentials across communities as a method for valuing site specific amenities, yielded estimated values of subsistence harvests in Alaska of about $30 per pound.

These analyses assume that Native peoples who live by subsistence hunting and fishing do so by choice, rather than out of necessity, and that they perceive themselves better off than if they were fully absorbed into the market economy. There is some support in the literature for that perspective on the part of Alaska and northern Canada Natives (Berger 1985; Klein 1989; Usher 1976). Hensel (1994:10) provides anecdotal evidence for this view in the following quotation from an interview with a Yup'ik woman professional in Bethel in July 1992:

"You know, I have had some really good offers, more money and advancement if I would move to Anchorage or Juneau, but I've wanted to stay here so that I could continue to do subsistence.

"One time when my supervisor was out from Anchorage, we were having her over to dinner and serving her moose, strips, salmonberry and blueberry pie. I

told her how important it was for me to be able to eat this food, and hunt and fish. I thanked her for all her help with my career but told her that I wanted to stay here."

As the preceding discussion suggests, converting changes in individual welfare to monetary measures for subsistence users is difficult. Assistance can be provided by other disciplines, such as cultural anthropology, that take account of the profound cultural differences associated with wildlife and wildlife management. In fact, anthropological studies are a valuable approach used by ADFG Division of Subsistence for evaluating management issues.

The committee has attempted to identify the marginal value of additional information on the basis of the likelihood that it would change a management decision. Several examples of ADFG Division of Subsistence community-level studies by that criterion have proved valuable in leading to apparently positive management changes. This type of social and economic analysis might be most appropriate for the remote rural communities because of the relatively homogeneous and distinct cultures in these communities; the relative autonomy and self-governance of these communities due to isolation, history, and the legal definition of subsistence rights to wildlife resources; and the recognition and acceptance by the larger society of the unique cultural and economic setting of the communities.

The ADFG Division of Subsistence, which was established around 1980, is unique in ADFG in that social science studies are important in its work. Some of these studies use an economic perspective (for example, Wolfe and Walker 1987) but a number of community-specific harvest surveys, which are very different from the Division of Wildlife Conservation hunter surveys, have been conducted. Hunter surveys focus on the primary game animals (moose, caribou, and so on) and provide a time series for harvests and hunter participation for all licensed hunters from 1983 to the present (Robert J. Walker pers. comm. 1997). In contrast, the Division of Subsistence studies are for all subsistence resources (from roe on kelp, to gull eggs, to berries, and so on), are at the community level, and typically collect data for 1 year only. Thus, the Division of Subsistence surveys provide a detailed picture of all subsistence activities for one time for the given community.

In addition to the harvest surveys and related reports, the division also has undertaken social-impact studies that focus on a specific resource in a given region, sometimes from the perspective of cultural anthropology. Reports of 2 of those studies are discussed here as examples of social science studies that have led to wildlife management changes in the communities of interest. The methods and findings might have elements of benefit-cost or other economic analyses but are predominantly cultural studies in which the behavior at issue appears to be

largely motivated by ethical or cultural values. The first study summarized below concerns brown bear use in northwestern Alaska by village residents. It provides an example of where management changes were based almost entirely on cultural considerations and financial concerns were not an issue. The second study concerns changes in moose-hunting regulations in the upper Kuskokwim River drainage (part of GMU 19D), particularly affecting the villages of Nikolai and Telida. Here, the change involved resolution of conflict between subsistence users and fly-in hunters and was made mainly on equity grounds. Both of those policy changes appear to have been in the public interest, and the social science analysis, although it involved no formal economic analysis, was relevant and appropriate to the problem at hand.

Loon and Georgette (1989) examined subsistence use of brown bears by residents of northwestern Alaska, particularly the Kotzebue Sound region (GMU 23) and to a lesser extent Norton Sound (GMU 22). The use of brown bears for food and raw material was prevalent in all the inland study communities, but coastal communities rarely used bears for food, because the flesh of coastal bears, which feed on carcasses of sea mammals has an unpleasant flavor. The authors note that northwestern Alaskans have an extensive array of traditional laws and lore regarding human-bear interactions. The traditional practices cover hunting strategies, butchering processes, personal conduct, methods of defense, and appropriate attitudes. Because brown bears are believed to have keen hearing, Inupiat hunters do not openly discuss their bear hunts. There are believed to be severe consequences to the hunter and his family if these laws are not heeded. Hensel (1994:2), in a study for the Association of Village Council Presidents in western Alaska, makes similar observations for Yup'ik, noting that "the hunting of brown bears, and even discussing these animals, is potentially dangerous."

Loon and Georgette discuss the incongruity between the brown bear hunting regulations in place at the time and the customary and traditional hunting practices. The state regulations in the late 1980s presume that the primary use of a brown bear is as a trophy, whereas subsistence hunters' rules presume that the primary use is as a source of food and raw material. For example, state regulations in the late 1980s required that a person who kills a bear must personally present the skin and skull to an authorized representative of the ADFG for sealing within 30 days after taking. A person with a bear had to keep the skin and skull together until a representative of the department removed a rudimentary premolar tooth from the skull. In contrast with the treatment of most big-game species, discarded brown bear meat is not considered waste under the regulations.

Almost all of these brown bear hunting regulations were at odds with customary subsistence hunting practice. Some subsistence hunters leave the head in the field at the kill site as a sign of respect, a practice that is in conflict with the

sealing requirements. Requiring hunters to salvage the skin and skull does not accommodate those who hunt for meat and fat only; to subsistence hunters not requiring the salvage of bear meat is objectionable. A bag limit of 1 bear every 4 years is inconsistent with the fact that in most villages only a small number of men actually harvest bears, and these hunters share their harvests throughout the community. The strong prohibition on speaking openly about brown bears includes not even speaking about one's intentions to hunt. Requiring local hunters to purchase a hunting tag before hunting brown bears, and thereby deliberately making their intentions known, is incompatible with traditional hunting practices. The regulatory hunting seasons at that time, April 15 -May 25 and September 1 -October 10, also conflicted with traditional hunting times, which can begin as early as March when some bears first come out of their dens. Those many differences between the hunting regulations and traditional practices are culturally based, "learned differences which derive from the social values of the hunter's community" (Loon and Georgette 1989:49).

Not surprisingly, given the mismatch between the regulations and traditional practices, Alaska Natives generally ignored the regulations. As a result, the state was able to collect only very limited information on harvest and use of bears. Loon and Georgette estimate that of the bears killed by interviewed hunters over the previous decade, only 3% were reported. They concluded that hunters would be more likely to report their bear harvests if regulations accommodated their hunting practices and if the reporting procedure were simple.

The study by Loon and Georgette led to changes in the management of bear hunts in some areas (Hensel 1994). For example, a western Alaska brown bear management area (covering essentially the Yukon-Kuskokwim delta and the Aniak and Togiak drainages) was established as a cooperative agreement of the Association of Village Council Presidents and the US Fish and Wildlife Service. In that area, tagging and sealing requirements were suspended after the fall of 1991. Residents were required to register in advance of brown bear hunting but paid no fee. The season was extended to run from September 1 to May 31.

In the early 1980s, the ADFG Division of Subsistence produced a series of reports concerning subsistence use in the upper Kuskokwim River drainage, particularly communities of Nikolai and Telida (Andrews and Stokes 1984; Stickney 1980, 1981; Stokes 1984; Stokes and Andrews 1982), which depend heavily on subsistence resources. The studies began in 1980 when the McGrath local Fish and Game Advisory Committee asked the Board of Game to implement a controlled-use zone around the 2 villages. According to Stickney (1980), the proximity of the region to the municipalities of Anchorage and Fairbanks created a situation in which the game resources, especially moose, were subject to competition from urban hunters, as well as boat hunters from down-river

Kuskokwim communities. In Stickney's opinion, hunting by outsiders, annual village take, marginal moose habitat, and a high wolf population all acted to keep moose at a low density in the area. Moose was the most important food item in the local diet, and alternative subsistence resources (salmon, whitefish, and caribou) were not plentiful enough in the area to be dependable buffers for inadequate moose harvests. Stickney's assessment of the situation was that the survival of the villages was at stake. With regard to Nikolai, "the villagers hunt for meat for the winter and they will not return empty handed if possible even if their prey does not conform to the State's regulations." With regard to Telida, "in this case …the regulations imposed by the State have apparently little bearing on what the village of Telida faces as a reality. The villagers will ensure their own survival even if the regulations compel them to operate outside the permitted system." Stickney concluded that as of 1980 there were not enough data to make firm recommendations to resolve the problem.

At the spring 1981 Board of Game meeting, a controlled use area for moose harvest was established for part of GMU 19D, partly on the basis of the report presented by division staff (Stickney 1980, 1981). Stokes and Andrews (1982) monitored the controlled use area during the 1981 season and concluded that probably no aircraft entered the region for hunting. The institution of the controlled-use area was looked on favorably as a management tool by local residents. In 1981, the season was also changed to include a winter hunt in late November and early December. Stokes and Andrews noted that some traditional spring hunting of moose is also important for spring and summer food needs. In March 1982, the Board of Game adopted an extended winter moose season in portions of the controlled-use area from December 1 through February. Stokes (1983) and Andrews and Stokes (1984) evaluated the 1982-1984 winter moose seasons and concluded that they were successful in providing Nikolai residents the opportunity to harvest moose legally during times compatible with local environmental conditions and with cultural needs.

Harvest of wildlife by Alaskans on a sustained yield basis for subsistence use, or as a replacement for store purchases of foods, presents an admirable model of human compatibility with the world environment through its minimal contribution to depletion of the world's nonrenewable energy and mineral resources, and the environmental pollution associated with their use (Harbo 1993). This is in contrast to most people in developed countries who are dependent upon food produced by intensive agriculture. The amount of energy and minerals consumed in the production of commercial food stuffs, especially in the United States, represents a major component of all nonrenewable resource use. This includes production of farm equipment, use of petroleum products for fertilizer production, operation of farm machinery, processing of foods, their storage, and

shipment to markets. The contributions from agricultural production to pollution of the atmosphere (including "greenhouse" gases), the waters, and lands of the world are immense. Alaska's contribution to the nation's energy needs through the production of oil from the North Slope reserves is appreciated by many Americans. However, most Americans are not aware of the contribution that Alaskans make to the conservation of energy and other resources through their sustainable harvest and use of wild meats.

In its review of the social science methods that can be used to assess the social and economic impacts of a predator control program and by describing what is known about those impacts in relation to wolf control in Alaska, the committee identified major gaps in existing information. The next chapter discusses the general context in which wildlife management decisions must be made and the constraints on those decisions.

ADFG (Alaska Department of Fish and Game). 1994–95. Trapper questionnaire statewide report.

ADFG (Alaska Department of Fish and Game). 1994. Subsistence in Alaska: 1994 Update. Division of Subsistence, Juneau, AK.

Alaska Department of Labor. 1996. Labor Department Estimates Population Trends. Research and Analysis Section.

Alaska Wildlife Alliance. 1992. Survey among Alaska residents regarding wolf hunting.

AWRTA (Alaska Wilderness Recreation and Tourism Association). 1995. Results of economic survey of AWRTA tourism business members on the effects of the tourism boycott. AWRTA, Wolf Issue/Tourism Boycott: Survey Results.

Anderson DB. 1995. The Alaska Department of Fish and Game public opinion survey on predator control in Game Management Unit 19D East. Report to the Alaska Board of Game. Division of Subsistence and Division of Wildlife Conservation, Alaska Department of Fish and Game, Juneau, AK.

Andrews EL and J Stokes. 1984. An overview of the Upper Kuskokwim controlled use area and the use of moose by area residents, 1981–1984. Technical Paper No. 99. Alaska Department of Fish and Game Division of Subsistence.

Arrow K, R Solow, E Leamer, P Portney, R Hadner, and H Schuman. 1993. Report of the NOAA panel on contingent valuation. Federal Register 58: 4601–4614.

Arthur LR, E Gum, E Carpenter, and W Shaw. 1977. Predator control: the public viewpoint. Pp. 135–155 in Transactions of the 42nd North America Wildlife Conference.

Bath AJ. 1987. Attitudes of various interest groups in Wyoming toward wolf reintroduction in Yellowstone National Park. MA thesis, University of WY, Laramie. 123 Pp.

Bath AJ. 1991. Public attitudes in Wyoming, Montana and Idaho toward wolf restoration in Yellowstone National Park. Pp. 91–95 in Transactions of the 56th North American Wildlife and Natural Resources Conference. Wildlife Management Institute, Washington, DC.

Berger TR. 1985. Village journey: the report on the Alaska Native Review Commission. Inuit Circumpolar Conference. New York: Hill and Wang.

Biggs JR. 1988. Reintroduction of the Mexican wolf into New Mexico—an attitude survey. MA thesis. NM State Univ, Las Cruces, NM. 66 Pp.

Boertje RD, P Valkenburg, and ME McNay. 1996. Increases in moose, caribou, and wolves following wolf control in Alaska. J Wildl Manage 60:474–489.

Boyce JR, DW McCollum, and JA Morrison. 1995. An economic impact analysis of the big game hunting guide industry in Alaska.

Braithwaite A and S McCool. 1988. Social-normative influences and backcountry visitor behavior in occupied grizzly bear habitat. Symp Soc Sci in Res Mgt, Univ Ill., Urbana, IL.

Brown TL, DJ Decker, and DL Hustin. 1981. Public attitudes toward black bears in the Catskills. New York State Department of Env. Conservation. Albany, NY. Pp. 108.

Brown TC and ES Burch, Jr. 1992. Estimating the economic value of subsistence harvest of wildlife in Alaska . In GL Peterson and others, Eds. Valuing wildlife resources in Alaska. Boulder: Westview.

Burghardt GM, RD Hietala, and MR Pelton. 1972. Knowledge and attitudes concerning black bears by users of the Great Smoky Mountains National Park. Int Conf Bear Res Mgt 2:255–273.

Buys C. 1975. Predator control and rancher's attitudes. Environ & Beh 7:81–89.

Carson RT and RC Mitchell. 1989. Using surveys to value public goods: the contingent valuation method. Johns Hopkins University Press, Baltimore, MD.

Carson RT, WM Hanemann, and D Steinberg. 1990. A discrete choice contingent valuation estimate of the value of Kenai King Salmon. J Behavior Econ 19:1–15.

Ciriacy-Wantrup SV. 1947. Major economic forces affecting agriculture with particular reference to California. Hilgardia; a journal of agricultural science 18:2–21.

Clawson M and J Knetsch. 1966. Economics of outdoor recreation. John Hopkins University Press, Baltimore, MD.

Colorado Division of Wildlife. 1989. Attitudes about wildlife and black bears. Denver, CO.

Davis RK. 1964. The value of big game hunting in a private forest. Trans N Am Wildl and Nat Resour Conf. 29:393–403.

Desvousges WH, AR Gable, RW Dunford, and SP Hudson. 1993. Contingent valuation: the wrong tool to measure passive-use losses? Choices: the magazine of food, farm and resource issues. Second Quarter 8:9–11.

Diamond PA and JA Hausman. 1994. Contingent valuation: is some number better than no number? J Econ Persp 8:45–64.

Duffield JW and CJ Neher. 1996. Economics of wolf recovery in Yellowstone National Park. Trans. 61st N Am. Wildl and Nat Resour Conf 61:285–292.

Duffield JW. 1997. Nonmarket valuation and the courts: the case of the Exxon Valdez forthcoming Contemporary Economic Policy.

Dunlap T. 1988. Saving America's wildlife. Princeton: Princeton University Press. 222 Pp.

Fall JA, JC Schichnes, M Chythlook, and RJ Walker. 1986. Patterns of wild resource use. In Dillingham: Hunting and fishing in an Alaskan regional center. Technical Paper No. 135. Alaska Department of Fish and Game, Division of Subsistence.

Freeman AM. 1993. The measurement of environmental and resource values. Washington, DC: Resources for the Future.

Gardner C and K Taylor. 1992. Results of the wildlife management survey in the upper Tanana/Fortymile region. Alaska Department of Fish and Game. Fairbanks, AK.

Hanemann WM. 1994. Valuing the environment through contingent valuation. J Econ Persp 8:9–44.

Harbo S. 1993. Environmental sanity: think globally, act locally. Paper presented at the Wolf Summit. Fairbanks, Alaska, January 1993.

Hastings B. 1986. Wildlife-related perceptions of visitors in Cades Cove, Great Smoky Mountains National Park. PhD dissertation, Univ TN, Knoxville.

Hensel C. 1994. Brown bear harvests in the western Alaska brown bear management area, 1992/1993: Statistical Information and Cultural Significance. Manuscript.

Herrero S. 1970. Man and the grizzly bear. Biosci 20:1148–1153.

Herrero S. 1978. People and grizzly bears: the challenge of coexistence. Pp 167–179 in CM Kirkpatrick, Ed. Wildlife and people. The 1978 John S. Wright Forestry Conference. Purdue Res Foundation, West Lafayette, IN .

Hook R and W Robinson. 1982. Attitudes of Michigan citizens toward predators. In F Harrington and P Paquet, Eds. Wolves of the world. NJ: Noyes Publishing.

Hotelling H. 1931. The economics of exhaustible resources. J Polit Econ 29:137–75.

Jobes PC. 1991. The greater Yellowstone social system. Cons Biol 5:387–394.

Jonkel CJ. 1975. Opinion—of bears and people. Western Wildlands 2:30–37.

Jope K and B Shelby. 1984. Hiker behavior and the outcome of inter-actions with grizzly bears. Leis Sci 6:257–270.

Kellert SR. 1980. A national study of American attitudes, knowledge, and behaviors toward wildlife: Phases I, II, III. US Govt Printg Off, Supt Doc 024-010-00-625-1/2/4.

Kellert SR. 1985. Public perceptions of predators, particularly the wolf and coyote. Biol Conser 31:167–189.

Kellert SR. 1986. The public and the timber wolf in Minnesota. Transactions of the 51st North American Wildlife and Natural Resources Conference. Wildlife Management Institute, Washington, DC.

Kellert SR. 1991. Public views of wolf restoration in Michigan. Transactions of the 56th North American Wildlife and Natural Resources Conference. Wildlife Management Institute. Washington, DC.

Kellert SR. 1994. Public attitudes toward bears and their conservation. In JJ Clear, C Servheen, and LJ Lyons, Eds. Proceedings of the Ninth International Bear Conference. U.S. Forest Service, Missoula, MT.

Kellert SR. 1996. The value of life: biological diversity and human society. Island Press, Washington, DC.

Klein DR. 1989. Northern subsistence hunting economies. Pp. 96–111 in RJ Hudson and others, Eds. Wildlife production systems. Cambridge University Press, New York.

Krutilla JV. 1967. Conservation reconsidered. Amer Econ Rev 4:777–786.

Llewellyn L. 1978. Who speaks for the timber wolf. Transactions of the 43rd North American Wildlife Conference. Wildlife Management Institute, Washington, DC.

Loomis JB and MH Thomas. 1992. Pricing and revenue capture: Converting willingness to pay into state and private revenue . Pp. 255–274 in GL Peterson and others, Eds. Valuing wildlife resources in Alaska. Westview Press, Boulder, CO.

Loon H and S Georgette. 1989. Contemporary brown bear use in northwest Alaska. Alaska Department of Fish and Game, Division of Subsidence. Technical Paper No. 163.

Lopez B. 1978. Of wolves and men. New York: Scribner's.

Matthiessen P. 1988. Wildlife in America. New York: Viking.

McCollum DW and SM Miller. 1994. Alaska voters, Alaska hunters, and Alaska nonresident hunters: their wildlife related trip characteristics and economics. Alaska Department of Fish and Game.

McConnell KE, IE Strand, and NE Bockstael. 1990. Habit formation and the demand for recreation: issues and a case study. Pp. 217–235 in A Link, Ed. Advances in applied micro-economics, Volume 5. JAI Press.

McCool S, A Braithwaite, and D Cikanek. 1990. Beliefs about bears. Wildl Soc Bulletin.

McNamee T. 1984. The grizzly bear. New York: Alfred A. Knopf.

McNaught DA. 1987. Wolves in Yellowstone Park? Park visitors respond. Wildl Soc Bulletin 15:518–521.

Merculieff L and W Soboleff. 1994. Letter to the Alaska Federation of Natives and the participants of the Elders and Youth Conference from participants in the Alaska Native traditional knowledge and ways of knowing workshop, September 13–14, 1994.

Miller SM and DW McCollum. 1994a. Alaska voters, Alaska hunters, and Alaska nonresident hunters: their characteristics and attitudes towards wildlife. Alaska Department of Fish and Game.

Miller SM and DW McCollum. 1994b. Tradeoffs between uses and users in natural resource management: A case study of who gains, who loses, and what it means. Alaska Department of Fish and Game.

Miller SM, R Sinnot, and D McCollum. 1994c. Differentiating among wildlife-related attitudinal groups in Alaska. Transactions of the 59th North American Wildlife and Natural Resources Conference. Pp. 512–523.

Miller SM, SD Miller, and DW McCollum. 1997. Attitudes toward and relative values of Alaskan brown and black bears to resident voters, resident hunters, and nonresident hunters. Int Conf. on Bear Res and Management, Vol 10.

Murray A. 1975. Influence of education programs on wolf conservation in Canada. In DH Pimlott, Ed. Wolves. Supplementary. 43 Pp. World Conservation Union, Morges, Switzerland.

Naess A and I Mysterud. 1987. Philosophy of wolf policies. I: general principles and preliminary exploration of selected norms. Cons Biol 2:16.

Peek JM and others. 1991. Restoration of wolves in North America. The Wildlife Society, Technical Review 91–1. Wildlife Society, Washington, DC.

Petko-Seus PA and MR Pelton. 1984. Backpacker bear knowledge in Great Smoky Mountains National Park. Hum Dim Wildlife Newsletter 3:8–12.

Petko-Seus PA. 1985. Knowledge and attitudes of campers toward black bears in Great Smoky Mountains National Park. MS. thesis, Univ TN, Knoxville, TN.

Reading R, TW Clark, and SR Kellert. 1994. Attitudes and knowledge of people living in the greater Yellowstone ecosystem. Soc and Nat Resour 7:349–365.

Reid R and D Janz. 1995. Economic evaluation of Vancouver Island wolf control. Pp. 515–521 in LN Carbyn, SH Fritts, and DR Seip, eds. Ecology and conservation of wolves in a changing world. Canadian Circumpolar Institute, Occasional Publication No. 35, Edmonton.

Shepard P and B Sanders. 1985. The sacred paw: the bear in nature, myth, and literature. Viking Penguin, Toronto, Canada.

Stickney A. 1980. Subsistence resource utilization: Nikolai and Telida—interim report. Technical Paper No. 20. Alaska Department of Fish and Game, Division of Subsistence.

Stickney A. 1981. Subsistence resource utilization: Nikolai and Telida—interim report II. Technical Paper No. 21. Alaska Department of Fish and Game, Division of Subsistence.

Stokes J and E Andrews. 1982. Subsistence hunting of moose in the Upper Kuskokwim controlled use area, 1981. Technical Paper No. 22, Alaska Department of Fish and Game, Division of Subsistence.

Stokes J. 1983. Winter moose season in the Upper Kuskokwim controlled use area, 1982–1983. Technical Paper No. 72. Alaska Department of Fish and Game, Division of Subsistence.

Stokes J. 1984. Natural resource utilization of four Upper Kuskowin communities. Technical Paper No. 86. Alaska Department of Fish and Game, Division of Subsidence.

Stuby R, E Carpenter, and L Arthur. 1979. Public attitudes toward coyote control. USDA Economic Statistics and Cooperative Service. ESCE-54. Washington, DC.

Trahan RG. 1987. Wilderness user's attitudes, information use and behaviors in relation to grizzly bear dangers in the backcountry. Yellowstone National Park, National Park Service.

Trice AH and SE Wood. 1958. Measurement of recreation benefits. Land Econ 34:195–207.

Turner RK, D Pearce, and I Bateman. 1993. Environmental economics. Johns Hopkins University Press, Baltimore, MD.

Usher PJ. 1976. Evaluating country food in the northern native economy. Arctic 105–20.

Usher PJ. 1981. Sustenance or recreation? The future of native wildlife harvesting in northern Canada. Pp. 56–71 in Proceedings, First International Symposium on Renewable Resources and the Economy of the North, M Freeman, Ed. Assoc of Canadian Universities for Northern Studies, Ottawa.

Wolfe RJ. 1981. Norton Sound/Yukon delta sociocultural systems baseline analysis. Technical Paper No. 59. Alaska Department of Fish and Game, Division of Subsistence.

Wolfe RJ. 1984. Commercial fishing in the hunting-gathering economy of a Yukon River Yup'ik Society. Etides/Inuit Studies Supplemental Issue 8:159–183.

Wolfe RJ and RJ Walker. 1987. Subsistence economies in Alaska: productivity, geography, and development impacts. Arctic Anthropology 2:56–81.

Wolfe RJ. 1991. Trapping in Alaska communities with mixed subsistence-cash economies. Technical Paper No. 217. Alaska Department of Fish and Game, Division of Subsistence.

Wolfe RJ and RG Bosworth. 1994. Subsistence in Alaska: 1994 update. Alaska Department of Fish and Game, Division of Subsistence, 4 Pp.

Wolfe RJ. 1996. Subsistence food harvests in rural Alaska and food safety issues. Paper presented to the Institute of Medicine. National Academy of Sciences Committee on Environmental Justice, Spokane, WA.

Young S. 1946. The wolf in North American history. Caldwell, ID, Caxton Printers.

Yu X. 1991. Valuing the impact of Alaska's guiding industry on the state's economy. M.Sc. thesis, University of Alaska, Fairbanks.