Existence values in benefit-cost analysis and damage assessment.
This paper addresses theoretical issues relating to existence values.(1) Earlier contributions (including Krutilla 1967; Randall and Stoll 1983; McConnell 1983; Madariaga and McConnell 1987; Smith 1987; Boyle and Bishop 1987; Freeman, forthcoming) have already considered a number of important aspects including the definition of existence value, how to characterize existence for theoretical purposes, where to demarcate the boundary between use and existence values, the motivations that might lead people to hold existence values, and the relationship between existence values and weak complementarity. Still, important theoretical questions remain unresolved.
The issues to be addressed here can be most easily described by referring to results from an actual study. Boyle and Bishop (1987) used contingent valuation to estimate that preventing extinction of the striped shiner in Wisconsin would be worth about $12 million annually to Wisconsin taxpayers. The striped shiner, a small minnow inhabiting the turbid depths of the Milwaukee River, has no known present or likely future uses. The $12 million is therefore to be interpreted as a pure existence value.
Such existence values raise at least five conceptual issues which have not been fully addressed in the previous literature. First, what can be said about the "existence" of existence values in a case like this? As Freeman (forthcoming) has pointed out, there is a growing consensus among economists that people may place positive values on important natural assets they never plan to use. A commonly mentioned example is the Grand Canyon. But, the striped shiner is quite another matter, not only because of its seeming uselessness, but also because nearly all of Boyle and Bishop's respondents had never heard of it prior to receiving the contingent valuation survey. Are existence values theoretically possible for obscure and previously unknown resources like the striped shiner? The second issue will be termed the "project selection problem."(2) Suppose there is a project that would mean the difference between survival and extinction of the striped shiner in Wisconsin. If costs would be less than $12 million, would there be a strong economic case for undertaking the project? A third issue will be termed the "adding up problem." If striped shiner values were added to other values that Wisconsin citizens might express, for example, existence values for grizzly bears, prevention of acid rain, preservation of wilderness, and other "environmental good things," then the sum would quickly become unbelievable. Granting this point, is it grounds for dismissing the shiner value a priori? A fourth issue can be summarized by supposing that a chemical spill wipes out the shiners. Assuming for the sake of argument that contingent valuation worked well, would we conclude that Wisconsin citizens have sustained "damages" of $12 million annually? This issue boils down to a question of property rights. The fifth issue is the problem of high per unit values. Suppose that there are 1,200 striped shiners left in the state. Then, they would be worth an average of $10,000 each! Is is plausible that individual members of very obscure species could be worth that much on average?
We intentionally set aside for the time being questions about the accuracy of contingent valuation, not because such issues are unimportant, but because theoretical issues must be given a higher priority. Theory must guide empirical measurement and interpretation of results. If existence values are theoretically doubtful, questions of empirical validity are irrelevant. Using the shiner results as an example, we explore here some theoretical nuances of existence values and draw inferences about the design and interpretation of contingent valuation studies, but postpone questions about whether existence values can be measured accurately once the theoretical foundations are laid. We ask readers to take the $12 million as an accurate figure for purposes of argument only, suggesting that they put questions about the accuracy of contingent valuation, which must eventually be asked, temporarily "on hold."
Questions relating to valuation under uncertainty are also postponed. Existence value and option value are sometimes linked together as "nonuse values." Option value is beyond the scope of this paper, and we assume certainty throughout.
We shall show that existence value creates no new conceptual problems for resource valuation under certainty. However, some problems that are well known on a more general level take on added importance when existence values are interpreted in the context of public decision making and damage assessment.
II. EXISTENCE VALUES FOR THE OBSCURE AND UNKNOWN
We will use the term "resource existence values" to mean values for natural resources that are motivated from sources within the individual's utility function other than personal use. In the case of wildlife, use values might include values associated with recreational or commercial hunting, fishing, or trapping, the so-called nonconsumptive uses such as bird-watching, and so on. Other resources, such as wilderness areas, marine parks, plant resources, and wetlands require somewhat different categories of use values, but the underlying principles would be the same. Use values involve either in situ contact with the resource in question or personal consumption of products derived from the resource. Existence values involve neither personal consumption of derived products nor in situ contact.
Why should consumers value something that they do not personally consume? Altruism toward friends and relatives or others who may be users, toward future generations of users, or toward animals themselves could motivate existence values (Randall and Stoll 1983). Feelings of environmental responsibility might also enter in (Bishop and Heberlein 1984). As previously noted, most economists would probably not rule out the theoretical possibility of existence values for major natural assets.(3) But, the shiner has no present nor known future uses and was unknown to respondents prior to the survey. Some might argue that existence values for the obscure and unknown should be ruled out a priori.
However, we would argue that there are compelling reasons not to rule out existence values in such cases. Even on the abstract level of basic theory, we distinguish between preferences and information. When consumers express their preferences subject to a budget constraint or firms maximize profits under a given production technology, perfect knowledge is often assumed. For consumers, this means that we assume each is aware of the characteristics of the market products that can be purchased and the various ways in which these goods and services can be combined to generate utility. In reality, consumers engage in numerous activities to inform their choices. One way of rationing scarce information-gathering resources is to ignore information that is not relevant to current choices. Suppose a new product becomes available. If initial information shows that the new product may be capable of satisfying preferences, then the consumer has an incentive to acquire more information about it. Or suppose that dryland farmers are presented with the prospect of a new irrigation project in their area. They may not have previously had any incentives to learn about irrigated farming technologies. The prospect of increasing profits and the necessity of choosing whether or not to sign contracts for water provide incentives to seek information on relevant production possibilities.
Just because respondents to a contingent valuation questionnaire do not have full knowledge of a given resource does not necessarily mean that they have no preferences with respect to that resource. Rather, it may simply reflect the fact that no choice problems involving that particular resource have yet been encountered; they may have felt little need to acquire specific information about it. Thus, contingent valuation scenarios must provide respondents with information needed to make sound choices (Fischoff and Furby 1988). The same motivations that yield existence values for well-known resources (e.g., altruism toward animals and feelings of environmental responsibility) could yield positive existence values for previously unknown resources. Lack of knowledge cannot be taken as evidence that the existence of such resources lacks the ability to satisfy preferences. It could simply indicate the lack of past choice opportunities to motivate information gathering. In the case of the striped shiner, it is possible that people are concerned about the fate of endangered species, even obscure ones.
Therefore, it seems to us that positive existence values for obscure and unknown parts of nature must be considered a theoretical possibility. Generalizing such values from samples to the population does raise concerns because such values are conditional existence values in the sense that they would exist for specific objects only if the population as a whole were fully informed about those objects. Such conditional existence values should be admitted as valid possible welfare measures because of the limited ability of real world people to obtain and process the vast amounts of information necessary to be fully informed on all parameters affecting their welfare. Members of a sample drawn for a contingent valuation exercise might be viewed as delegates of the population. They bring the general preferences of the population to bear on the specific details of the resource valuation issue at hand. On behalf of the population as a whole, they review information about the resource to be evaluated and, based on that information, evaluate the resource. The goal of the contingent valuation scenario is to convey information that can help respondents link the resource characteristics or service flows being valued to their preferences, which are in turn representative of the preferences of the population.(4) More succinctly stated, the goal should be to obtain the values that the population as a whole would express if it were more fully informed.
Affirming the theoretical possibility of existence values for natural and environmental resources, including the obscure and unknown, does raise additional issues about how those values should be interpreted in the policy arena. Can such values simply be added to other benefits and costs in evaluating proposed policies, regulations, and projects? Should existence values be treated as equivalent to use values in assessing the damages, under CERCLA and the Oil Pollution Act, associated with spills of oil and hazardous substances into the environment? Or, do existence values have special characteristics which should be considered when interpreting them for such purposes?
III. THE PROJECT SELECTION PROBLEM
To address these questions, we will introduce a formal model of existence value. Our model is not greatly different from those currently in the published literature, except that it will suit our purposes to have two resources with existence values in the model. Other theoretical modeling efforts, such as Randall and Stoll (1983), Boyle and Bishop (1987), Smith (1987), Madariaga and McConnell (1987), and Freeman (forthcoming) have focused on defining the relationships between use and existence values and related issues. Our model is a special case of earlier models in that we will assume that only the existence of the resources being studied matters to the individuals whose welfare is being evaluated. Use values of the resources in question are not an issue. A model focusing on pure existence values will facilitate the exposition here and in the next section.
Suppose that the welfare of n consumers is affected by the existence of populations of two different species, the striped shiner and the Higgins-eye pearly mussel, another endangered species with habitat in Wisconsin. Let us assume, following what was just said, that there are no known uses for either species. To simplify the exposition we will take the extreme case where, without public sector intervention, both species will become extinct in the near future. Suppose that the ith consumer has a quasi-concave, twice-differentiable utility function of the form
|U.sub.i~ = |U.sub.i~(|X.sub.1~, |X.sub.2~, |Y.sub.i~), |1~
|X.sub.1~ = the population of striped shiners, |X.sub.2~ = the population of Higgins-eye pearly mussels, and |Y.sub.i~ = the income of the ith consumer, a measure of consumption of market goods.
Market prices are assumed constant. Since |X.sub.1~ and |X.sub.2~ are exogenously determined and the consumer spends all income on "other goods and services," maximum utility can be determined directly by plugging in the values of the parameters, without the need to resort to maximization subject to a budget constraint.
We choose the simplest notation here to express the assumption that the populations of shiners and mussels "matter" in some sense to the individual in question. Following what was said in the preceding section, it is assumed that underlying preferences include a desire to avoid extinction of species, a desire that shiners and mussels can satisfy in part. The populations of shiners and mussels become arguments in the utility function based on these underlying preferences and on information about these species received from some source. The convention of including the populations of animals in the utility function is receiving wide acceptance (for a recent overview, see Brown and Plummer 1989).(5) As a theoretical abstraction, simply including the populations of shiners and mussels in the utility function is probably acceptable, but we suspect that more complex relationships may exist in reality. For example, we could let I = g(|X.sub.1~, |X.sub.2~, ...), where I is an index representing the degree to which the future of plant and animal species is assured. Utility would then have become a function of I and income. We suspect also that a change in an animal population is likely to be valued differently depending on the circumstances. For example, if 20 percent of the shiner population were lost because of some natural phenomenon like a severe winter, the loss might be smaller than if an equal number of fish were lost due to anthropogenic causes such as a chemical spill or vandalism. We could, of course, have specified a more complex utility function. However, at the current level of analysis, few insights would be added.
We will assume that
|Mathematical Expression Omitted~
|Mathematical Expression Omitted~
|Mathematical Expression Omitted~
|Mathematical Expression Omitted~
Duality applies in the usual way to yield the expenditure function,
|E.sub.i~ = |E.sub.i~(|X.sub.1~, |X.sub.2~, |U.sub.i~) |6~
where |E.sub.i~ is the minimum expenditure required to achieve specified utility level |U.sub.1~, given species population levels |X.sub.1~ and |X.sub.2~. If the level of utility achievable with populations at levels |X.sub.1~ and |X.sub.2~ is given by
|U.sup.12~.sub.i~ = |U.sub.i~(|X.sub.1~, |X.sub.2~, |Y.sub.i~ |7~
|E.sup.12~.sub.i~ = |E.sub.i~(|X.sub.1~, |X.sub.2~, |U.sup.12~.sub.i~) = |Y.sub.i~.
Marginal willingness to pay for increases in the species populations is given by
|Mathematical Expression Omitted~
We will assume that
|Mathematical Expression Omitted~
i.e., that marginal willingness to pay for |X.sub.j~ declines as |X.sub.j~ increases.
We will assume that shiners and mussels are viewed as substitutes; in mathematical terms, this can be stated as
|Mathematical Expression Omitted~ |11~
Marginal willingness to pay for |X.sub.1~ declines as |X.sub.2~ increases and vice versa. Consumers who are concerned about the extinction of obscure species may be willing to pay less for improvements in the population of shiners the more secure are the prospects for mussels as measured by the population of mussels. Complementarity is also a theoretical possibility, and whether such species are complements, substitutes, or neutral could be investigated empirically.
A theoretical definition of existence value is straightforward. Suppose that both the striped shiner and the mussel will become extinct in the absence of projects to save them. Further, suppose that a public project would maintain the shiner at its current population, but would have no effect on the mussel's survival prospects or on consumption of market goods. Using compensating surplus as the welfare measure and symbolizing existence value of the shiner as E|V.sup.1/0~.sub.i~, we have that
E||V.sup.1/0~.sub.i~ = |E.sub.1~(O, O, ||U.sup.00~.sub.i~) - |E.sub.i~(|X.sub.1~, O, ||U.sup.00~.sub.i~), |12~
||U.sup.00~.sub.i~ = |U.sub.i~(O, O, |Y.sub.1~). |13~
Notice that this value is conditional on extinction of the mussel. E|V.sup.2/0~.sub.i~, the existence value from a project that would prevent extinction of the mussel, but leave the plight of the shiner and market consumption unchanged, can be defined in similar terms:
E||V.sup.2/0~.sub.i~ = |E.sub.i~(O, O, ||U.sup.00~.sub.i~ - |E.sub.i~(O, |X.sub.2~, ||U.sup.00~.sub.1~). |14~
While the model we are proposing here does not include a time element, it could be extended easily to do so. Making the model intertemporal would allow the consumer to hold a positive value for avoiding temporary loss of some part of the population even if recovery were fairly rapid. Our discussion of possible motives for holding existence values implies that it would be a mistake to rule out, a priori, existence values for avoiding temporary losses of part of the population. Sympathy or empathy for animals, for example, could lead to such existence values for temporary losses. We see no theoretical reason to limit existence values to cases where natural resource losses are irreversible.
The model can be used to define and analyze the project selection problem in the following way. Suppose that the costs of a project to save the shiner are equal to |C.sub.1~. Let the sum of E|V.sup.1/0~.sub.i~ over all n consumers be E||V.sup.1/0~.sub.i~ and suppose that E||V.sup.1/0 |is greater than~ |C.sub.1~. How compelling would the economic case be for completing the project?
On one level, it must be recognized that having benefits greater than costs is never a terribly compelling case for project completion. A different design of the same sort of project or another project entirely might be better. In general, benefits in excess of costs only assures that the gainers from a project could compensate losers and still be better off; nothing can be concluded about the optimality of the project based on the sign of benefits minus costs. On this level, there is always a project selection problem whether or not existence values are thought to be an issue.
But, suppose that obscure endangered species are highly substitutable. Project selection would then become an especially relevant issue. It would be incumbent on the analyst to investigate cost-effective alternative means to save obscure endangered species rather than blindly accept the first project that happens to have positive net benefits. Given that, in the real world, there could be high substitutability not only among endangered species, but also between existence of endangered species and existence of other environmental "good things" (and possibly non-environmental "good things" as well), the economic case for any particular project based only on comparison of its existence values to its costs would seem to be weak.
Whether our hypothesis of high substitutability for the existence of many environmental good things is valid is an empirical question that needs to be investigated. If it proves valid under empirical scrutiny, then the project selection problem would be less severe in two situations. First, uniqueness is obviously an overworked term in dealing with environmental preservation, but it seems likely that some environmental assets have fewer substitutes than others. Mussels are probably a much better substitute for striped shiners than for bald eagles. Existence values for resources within major national parks and for well-known animals like whales and bears can probably be used in applied welfare analyses with fewer concerns for project selection problems than would arise for more obscure life forms.(6)
The second place where the project selection problem would seem to be muted is in the area of damage assessment for spills of toxics and oils. Existence values do raise some other issues in this context as we will discuss below, but some sort of analogue to the project selection problem does not seem to be present. This is because the "project" has, in effect, already been selected. When the toxic substance or oil was spilled in a given location, the definition of the thing to be evaluated was set.
IV. THE "ADDING-UP" PROBLEM
Closely related to the project selection problem is the adding-up problem. One commonly expressed concern is that there must be something conceptually wrong with existence value since there are hundreds of environmental good things that could be the subject of existence value studies. If we added up the existence values of each of them for any given member of society, the sum would become implausibly large. Consider an example. If the striped shiner is worth $4 to the average Wisconsin taxpayer (the average value per taxpayer used to calculate the $12 million figure for the state as a whole) and there are 100 obscure endangered species in Wisconsin, then would it follow that there is a value of $400 per taxpayer for all obscure endangered species?
The fallacy in this argument would be apparent immediately if one tried to apply it to market goods. It is well known (Just, Hueth, and Schmitz 1982; Hoehn and Randall 1989) that valid measures of the welfare impacts of multiple price changes are in variant with respect to the order in which the price changes are evaluated. However, when the goods are either substitutes or complements, the value of any particular price change will depend on the order in which the price changes are evaluated. Likewise, adding up values for a number of species ignoring that they might be substitutes or complements could lead to theoretically invalid welfare measures. If species are substitutes, for example, and consumers believe that several of them are about to become extinct, then simple adding up will overestimate their combined value.
Consider again the two-species case. Suppose that, in addition to a project that would save the shiner, there is a project that would save the mussel, but leave the shiner's situation and consumption of market goods unaffected. Suppose that if the shiner project is completed, the population of striped shiners will be |X.sub.1~ |is greater than~ O, and that it will be zero otherwise. Likewise, if the mussel project is completed, the population will be |X.sub.2~ |is greater than~ O, and zero otherwise. Suppose that the costs of the two projects are |C.sub.1~ and |C.sub.2~, respectively. Define the existence value if both species are saved as follows:
E|V.sup.12~.sub.i~ = |E.sub.i~(O, O, |U.sup.00~.sub.i~ - |E.sub.i~(|X.sub.1~, |X.sub.2~, |U.sub.00~.sub.i~). |15~
Note first that if, as we have hypothesized, obscure endangered species are substitutes, then adding up would involve a theoretical fallacy. Substitutability would mean that
E||V.sup.1/10~.sub.i~ + E||V.sup.2/0~.sub.i~ |is greater than~ E||V.sup.12~.sub.i~. |16~
Further mathematical justification for this assertion is provided in an appendix. The left-hand items could result from doing two contingent valuation surveys. In the first, respondents would be asked to value the benefits of the shiner project, ignoring the mussel project; in the second, they would be asked to value the benefits of the mussel project in isolation. The right-hand side expresses respondents' values in a survey where they are asked to value saving both the shiner and the mussel from extinction. Adding up, as on the left-hand size, would indeed overestimate the combined value of shiners and mussels.
Note next that such adding up could lead to an aggregation problem. Letting E|V.sup.1/0~ be the sum of the E|V.sub.1/0~ over all n consumers and likewise for E|V.sup.2/0~ and E|V.sup.12~, it is possible that
E|V.sup.1/0~ |is greater than~ |C.sub.1~ |17~
E|V.sup.2/0~ |is greater than~ |C.sub.2~, |18~
E|V.sup.12~ |is less than~ |C.sub.1~ + |C.sub.2~. |19~
The two projects, evaluated in isolation from each other, would pass the benefit-cost test, but a single project combining the needed measures to save both the shiner and the mussel would not. E|V.sup.12~ is the theoretically correct measure of the benefits of the combined project.
It is worth noting that this is not a problem peculiar to existence values. It could occur for environmental use values or market goods if inadequate attention is given to the relationships among goods and services at issue. The point is that dismissing contingent existence values based on adding up arguments will not work. Nothing can be said about the value of 100 species based on the striped shiner value of $4 per taxpayer. In a broader context, it is not valid to argue that the shiner value is implausible because there are hundreds of endangered species and a multitude of other environmental good things about which consumers may be concerned.
Those who worry about adding existence values for striped shiners to existence values for grizzly bear, wilderness, etc., may really be suggesting that we must think beyond individual projects, particularly those involving non-unique environmental assets, if we are to come up with cost-effective ways of satisfying demands for environmental improvements. This does not make existence values wrong or irrelevant, but it does make them more difficult to interpret for policy.
As a final comment on adding up, it is interesting to note that the direction of the error from incorrectly aggregating individual values could be in the opposite direction for damage assessment. That is, incorrectly aggregating individual values could underestimate combined values. To see this, suppose that a contingent valuation study is done that accurately measures the amount that would be required to compensate consumers for the loss of the shiner. At the same time, a separate study is undertaken that measures the compensation required for loss of the mussel. Let the values for typical consumer i be E||V.sup.1/2~.sub.i~ for striped shiners, given that the mussel continues to exist at population level |X.sub.2~ and E||V.sup.2/1~.sub.i~ for mussels, given that striped shiners still exist at population level |X.sub.1~. Formally, these values would be defined as
E||V.sup.1/2~.sub.i~ = |E.sub.1~(|X.sub.1~, |X.sub.2~, ||U.sup.12~.sub.i~) - |E.sub.i~(O, |X.sub.2~, ||U.sup.12~.sub.i~) |20~
E||V.sup.2/1~.sub.i~ = |E.sub.i~(|X.sub.1~, |X.sub.2~, ||U.sup.12~.sub.i~) - |E.sub.1~(|X.sub.1~, O, ||U.sup.12~.sub.i~) |21~
where ||U.sup.12~.sub.i~ is defined as before.
Now suppose that spills entirely wipe out both species. The measure of the loss would be E||V.sup.00~.sub.i~ defined implicitly by
E||V.sup.00~.sub.i~ = |E.sub.i~(|X.sub.1~, |X.sub.2~, ||U.sup.12~.sub.i~) - |E.sub.i~(O, O, ||U.sup.12~.sub.i~). |22~
Let E|V.sup.1/2~, E|V.sup.2/1~, and E|V.sup.00~ equal the respective values aggregated over all n consumers. Notice that the way we have set up the problem means that compensation demanded will be represented by negative values. If |X.sub.1~ and |X.sub.2~ are substitutes, then it would follow that
/E|V.sup.1/2~ + E|V.sup.2/1~/ |is less than~ /E|V.sup.00~/. |23~
Using the sum of the two losses measured with the other species still in existence would underestimate the true loss.
V. THE PROPERTY RIGHTS PROBLEM
Recall how this problem was illustrated in the introduction: If a toxic spill completely wiped out the striped shiner, making the species extinct in Wisconsin, has the citizenry sustained damages of $12 million per year?(7) One's first reaction is to say, yes. If, as has been postulated in this paper, striped shiners are an argument in the utility functions of some Wisconsin citizens, then such a spill would cause them to suffer a welfare loss that we as economists are bound to recognize. This follows from one of the fundamental postulates of welfare economics; the individual is the sole judge of his or her well-being and it is not for the economist to judge which tastes and preferences are worthy of consideration as part of society's welfare and which are not.
Unfortunately, there is an issue here to which economists have given too little attention. Society does not view all changes in utility as equally valid for consideration in setting damages or deciding on public policies. Some people may get disutility from, for example, the hair styles of adolescents, but society may choose to ignore their losses in favor of allowing people the freedom to wear their hair as they please. In a similar way, society might choose to ignore people's concerns about the existence of resources they do not use.
This principle is particularly clear in a legal environment. Alpha may or may not be liable for a loss of utility inflicted on Beta, depending on many factors. For example, Beta probably cannot collect damages if someone unrelated to him/her is struck down by an automobile in Beta's presence no matter how much disutility he/she suffers as a result of the experience.
In the environmental area, there is an important set of issues here associated with property rights. In the context of litigation, the process of determining whether damages to resources that people do not use are compensable is closely linked to the process of deciding whether those holding existence values for resources have property rights in those resources. Property is a benefit stream, and property rights constitute the assurance of the state that it will protect that benefit stream.(8)
Wildlife resources are a good example to illustrate the issues. It is well established that wildlife resources are the property of the state to be held in trust for the benefit of the people. Historically, this status for wildlife was clearly related to preservation of food sources (Environmental Law Institute 1977), and it is not clear that the state is to act as a trustee for the interests of those holding existence values. The status of property rights of nonusers in other resources such as beaches and other inanimate public resources seems even less clear.
Nevertheless, an interesting trend is developing in favor of including existence values in assessments of natural resource damages, in effect granting property rights in public resources to those who do not currently use them. Under CERCLA and the Oil Pollution Act, as amended, parties responsible for the discharge of hazardous substances and oil into the environment are liable for resulting damages to natural resources. The U.S. Department of the Interior was assigned to promulgate rules for assessing such damages under CERCLA. The original rules for damage assessment, published by Interior in the Federal Register on August 1, 1986, allowed the inclusion of existence values, as we are using the term here, in damage estimates, provided that use values could not be measured.
Several aspects of the original damage assessment rules were challenged in court in the State of Ohio v. the Department of the Interior (880 F.2d 432 |D.C. Cir. 1989~). In a 1989 ruling on this case, the U.S. Court of Appeals for the District of Columbia Circuit further strengthened the property rights of nonusers in natural resources suffering damages from oil and toxics. The decision states, in a way that rather obviously refers to economic arguments, that nonuse values "may represent 'passive' use, but they nonetheless reflect utility derived by humans from a resource, and thus prima facie, ought to be included in damage assessment." This same decision rejected the rather odd rule that nonuse values could be counted only if use values could not be measured.
Proposed modifications in the Department of the Interior rules that would implement the Court of Appeals decision (U.S. Department of the Interior 1991) explicitly recognize that public resources damaged by oil or toxics may have "compensable values" that include existence values. The new Interior rules are of interest in the current context because they seem to take it for granted that nonusers have property rights and focus attention instead on where the damages are likely to be large and small. Most observers expect that new rules specifically addressing oil spills, which are to be promulgated by the Department of Commerce under the Oil Pollution Act of 1990 (P.L. 101-380), will permit the inclusion of existence values where they can be shown to be significant. Unless this trend is reversed, it appears that people will have property rights in public resources they do not use, at least so far as damages from oil and hazardous substances are concerned.
While such property rights issues are most visible on the negative side, where society is trying to judge whose "ox has been gored," the same set of issues arise in more subtle ways when one tries to evaluate the benefits of alternative steps that society could take in a positive direction. Bishop, Boyle, and Welsh (1987), for example, considered total valuation of Great Lakes ecosystems. The Great Lakes have suffered tremendous environmental insults over the years. While empirical evidence is still scarce, we would be surprised if existence values for restoration of those ecosystems, including the native fish populations, were not substantial. Such values, if they were counted in the policy debate over lake management, would probably conflict somewhat with current sport and commercial fishing interests. The sport fishery, in particular, is based to a substantial degree on stocking exotic salmon and trout for a "put-grow-and-take" fishery. If existence values for restoration of natural ecosystems turned out to be large, this could point toward a major reorientation of public programs in directions that would foster rehabilitation of native species and reduce sport fishing benefits. Commercial fishing might also have to be curtailed in some places. Society may have to decide how much weight to assign the values of nonusers in making decisions about how to manage the Great Lakes. We raise this issue here because it is basically an issue of who "owns" the fish-producing capabilities of the lakes. In a similar vein, it is interesting to speculate about debate that could be generated over whether to count existence values of animal rights advocates in the management of furbearers and wildlife resources used for sport hunting. To ignore existence values in such areas is to invite the criticism that we economists are overly narrow (or, more bluntly, "biased") in choosing whose welfare to count--and whose to ignore--in benefit-cost analysis. In essence, we would have made implicit assumptions about property rights that may not be valid.
VI. THE HIGH-PER-UNIT-VALUE PROBLEM
As an illustration here, suppose, again, that there are only 1,200 striped shiners left, so that they are in effect worth $10,000 each. Is this cause for alarm or disbelief?
The answer is, of course, rather straightforward. Existence is a public good, which means that one fish can simultaneously satisfy demands of many people. The average of $10,000 per fish only seems preposterous because we are used to thinking of fish in their roles as private goods. Furthermore, as Randall and Stoll (1983) have pointed out, such high per unit values are conditional on the current size of the population and overall status of the species. Presumably, the marginal value of additional striped shiners--and hence the average value--would decline rapidly if the population were to expand and other requirements for its long-term viability were satisfied.(9)
It is the public goods character of existence that makes existence value such a potentially important concept. Public goods deserve special attention because use of relatively small quantities of resources can produce large benefits by providing non-rival consumers in large numbers with increases in welfare. Also, public goods need special attention because of the well-known result that private, individual property rights and market exchange often result in the provision of suboptimal levels of public goods.
So far, theoretical enquiry has identified few barriers to including existence values in the formal welfare theory that forms the basis for benefit-cost analyses and damage assessments. There would seem to be no basis for ruling out existence values on a priori grounds even for resources that are not widely known or for temporary damages to resources. Stated differently, it is theoretically possible that existence values could "exist" for obscure and previously unknown resources as well as unique natural assets. Though large per unit values do take one's breath away at first, they seem quite consistent with the public goods nature of existence. Where environmental assets are highly substitutable, project selection should involve careful attention to alternatives that may satisfy closely related demands. However, the project selection problem will be of less concern for unique resources and should not enter in at all in damage assessments. The alleged adding-up problem is a nonissue simply because adding values without careful attention to the relationships between items being valued and the order of aggregation is theoretically unjustified. It will be up to the broader mechanisms of society to determine when people have a property right in the existence of environmental and other assets. Without such property rights, existence values may be of little importance for benefit-cost analysis and damage assessment. The institutional status of existence of natural assets should not be viewed as static, however. If preferences and concerns that express themselves as existence values are becoming more evident over time and can be quantified, and if the welfare effects associated with existence are indeed substantial, property rights may evolve to give existence values greater clout in public decisions. We have noted evidence that such property rights are indeed evolving, at least with respect to damages to public resources from spills of oil and hazardous substances.
If ways can be devised to measure existence values with sufficient accuracy, and they are given full consideration in environmental management decisions and resource damage assessments, the full ramifications might be far-reaching. It is conceivable that a major realignment of national priorities in the direction of environmental protection and rehabilitation could result. To facilitate sound social choices about the future role of existence values in benefit-cost analysis and damage assessment, theoretical research is needed to build consensus on definitions, on theoretical relationships between value categories, and on correct approaches to empirical estimation. Laboratory and field experiments are needed to learn more about the validity of contingent valuation in general, and of contingent existence values in particular. Finally, applied studies should continue to estimate existence values in real world situations and explore their potential implications for public decisions.
Resource economists may well be accused of disciplinary imperialism as they seek to incorporate ever wider sets of phenomena under the umbrella of dollar valuation. Is nothing sacred? Must everything be reduced to the metric of unrighteous mammon? However discomforting this criticism may be, we economists find ourselves in a "damned-if-you-do, damned-if-you-don't" predicament. To ignore existence value would be to court the equally damning criticism of having made a thinly masked value judgment in favor of use values as the only true economic values. Having come this far in the valuation of natural resources, do we dare turn away from this new challenge?
The claim is that:
|Mathematical Expression Omitted~
Recall the definitions of existence values:
E||V.sup.1/0~.sub.i~ = |E.sub.i~(O, O, ||U.sup.00~.sub.i~) - |E.sub.i~(|X.sub.1~, O, ||U.sup.00~.sub.i~) |A2~
E||V.sup.2/0~.sub.i~ = |E.sub.i~(O, O, ||U.sup.00~.sub.i~) - |E.sub.i~(O, |X.sub.2~, ||U.sup.00~.sub.i~) |A3~
E||V.sup.12~.sub.i~ = |E.sub.i~(O, O, ||U.sup.00~.sub.i~) - |E.sub.i~(|X.sub.1~, |X.sub.2~, ||U.sup.00~.sub.i~). |A4~
The question at hand is the sign of difference between the sum of the two individual measures and the overall measure. It can be shown that the sum of the two individual measures exceeds the last measure if and only if the two species providing the existence values exhibit a substitute relationship as defined in the body.
E||V.sup.1/0~.sub.i~ + E||V.sup.2/0~.sub.i~ - E||V.sup.12~.sub.i~
= |E.sub.i~(|X.sub.1~, |X.sub.2~, ||U.sup.00~.sub.i~) - |E.sub.i~(|X.sub.1~, O, ||U.sup.00~.sub.i~)
+ |E.sub.i~(O, O, ||U.sup.00~.sub.i~) - |E.sub.i~(O, |X.sub.2, ||U.sup.00~.sub.i~). |A5~
However, note that the first two members on the right-hand side of |A5~ represent the value of species two given that species one is preserved. The last two members of the right-hand side of |A5~ represent the value of species two if species one is allowed to go extinct. Furthermore, if we let |w.sub.2~ represent the derivative of the expenditure function with respect to the second species, this equation can be rewritten as:
|Mathematical Expression Omitted~
By the definition of substitutability we know that:
|Mathematical Expression Omitted~
Consequently the bracketed term on the right-hand side of |A6~ is non-negative and the value of the integral will be greater than zero. That is to say, if the individual's utility function exhibits substitutability between the two species:
E||V.sup.1/0~.sub.i~ + E||V.sup.2/0~.sub.i~ |is greater than~ E||V.sup.12~.sub.i~. |A8~
Respectively, professor in the Department of Agricultural Economics and the Center for Resource Policy Studies, University of Wisconsin-Madison, and senior project manager, HBRS, Inc., Madison, WI.
An earlier version of this paper was presented at a meeting of Western Regional Committee W-133, Molokai, HI, February 20-23, 1990. It benefitted from the discussion of that group and particularly from comments of its discussant, Anthony Fisher. We are also grateful to faculty and students from the Department of Economics, the Department of Agricultural Economics, and the Institute for Environmental Studies at the University of Illinois-Urbana, who participated in a seminar based on this paper on September 26, 1990. All errors are, of course, the sole responsibility of the authors.
The research was supported by the College of Agricultural and Life Sciences, University of Wisconsin-Madison; by the University of Wisconsin Sea Grant College Program through grants from the National Oceanic and Atmospheric Administration, U.S. Department of Commerce, and from the State of Wisconsin federal grant, NA 80 AA-DD00086, project R/PS--33; and by HBRS, Inc., of Madison, WI.
1 In his seminal paper on the economics of natural environments, Krutilla (1967) suggested that wilderness might have a value to those who simply enjoy knowing that it exists. He drew a distinction between existence value and bequest value, the latter relating to the desire to leave natural amenities and other public goods to heirs and to future generations in general. The distinction between existence and bequest values persists in parts of the literature to this day (e.g., Loomis 1989), but here we shall use a more general definition of existence value, where the desire to bequeath environmental assets to future generations is just one of several possible motivations for holding existence values.
2 The term project will be used in its broadest sense. We mean not only capital projects, but also changes in public policies, regulations, and services provided by the public sector.
3 Similarly, the possibility of negative existence values in some instances would probably be acknowledged by most economists who have examined the issues. For example, some parents might have negative existence values for grizzlies in Yellowstone National Park if their children backpack there. While the argument here is developed in terms of positive values, symmetrical arguments for negative existence values are implicit.
4 The distinction we are drawing between preferences and information may not be as sharp as we are making it out to be here, but such a distinction is surely justified. The experience of participating in a contingent valuation exercise may lead a person to have a somewhat different outlook than he or she had before, but the potential impact on preferences of spending a few minutes or even an hour or two participating in a survey would be easy to exaggerate. It seems much more plausible that people bring preferences to a survey that are relatively stable. Preferences are used to interpret information in order to respond to survey questions.
5 Existence could be a binary variable in the sense that a species either exists (i.e., is viable in the long run) or it does not exist, but this is not a necessary assumption and, for reasons stated in Randall and Stoll (1983), would not seem to capture the essence of the problem.
6 Geographical proximity may play an important role in determining substitutability. For example, the Adirondack Mountain lakes that are susceptible to damage from acid rain might be viewed us having few substitutes by people in New York, but many substitutes by people in California. If so, perhaps New York residents' existence values for acid rain reductions in the region should be taken as more compelling for policy than comparable existence values expressed by Californians. Another basis for differentiating between relatively unique environmental assets and general environmental good things might be based on underlying motives. One might view Adirondack lakes and wilderness in Alaska as very substitutable good things if other things are equal, but as very different products if one's electricity consumption affects the lakes and there is no comparable link to Alaskan wilderness.
7 Let us agree that here, for once, we will not quibble over whether compensation demanded should be used instead of willingness to pay, as reflected in the $12 million figure. The focus here is on a different set of issues.
8 We are indebted to Daniel W. Bromley for helping us to clarify this point.
9 We are indebted to an anonymous reviewer for calling this point to our attention.
Bishop, Richard C., and Thomas A. Heberlein. 1984. "Contingent Valuation Methods and Ecosystem Damages from Acid Rain." Staff Paper 217. Department of Agricultural Economics. University of Wisconsin, Madison.
Bishop, Richard C., Kevin J. Boyle, and Michael P. Welsh. 1987. "Toward Total Valuation of Great Lakes Fishery Resources." Transactions of the American Fisheries Society 116:339-45.
Boyle, Kevin J., and Richard C. Bishop. 1987. "Valuing Wildlife in Benefit Cost Analysis: A Case Study Involving Endangered Species." Water Resources Research 23:943-50.
Brown, G. M., and M. L. Plummer. 1989. "Methods for Valuing Acidic Deposition and Air Pollution Effects." State of Science/ Technology Report 27, Part A, National Acid Precipitation Assessment Report, Washington, DC.
Fischoff, Baruch, and Lita Furby. 1988. "Measuring Values: A Conceptual Framework for Interpreting Transactions with Special Reference to Contingent Valuation of Visibility." Journal of Risk and Uncertainty 1:147-84.
Freeman, A. Myrick, III. Forthcoming. "Non-Use Values in Natural Resource Damage Assessment." In Valuing Natural Assets: The Economics of Natural Resource Damage Assessment, eds. Raymond J. Kopp and V. Kerry Smith. Washington, DC: Resources for the Future.
Environmental Law Institute. 1977. The Evolution of National Wildlife Law. Prepared for the Council on Environmental Quality. Washington, DC: U.S. Government Printing Office.
Hoehn, John P., and Alan Randall. 1989. "Too Many Proposals Pass the Benefit Cost Test." American Economic Review 79 (June):544-51.
Just, Richard E., Darrell L. Hueth, and Andrew Schmitz. 1982. Applied Welfare Economics and Public Policy. Englewood Cliffs, NJ: Prentice-Hall.
Krutilla, John V. 1967. "Conservation Reconsidered." American Economic Review 57: 777-86.
Loomis, John. 1989. "Test-Retest Reliability of the Contingent Valuation Method: A Comparison of General Population and Visitor Responses." American Journal of Agricultural Economics 71 (Feb.):76-84.
Madariaga, Bruce, and Kenneth E. McConnell. 1987. "Exploring Existence Values." Water Resources Research 23:936-42.
McConnell, K. E. 1983. "Existence and Bequest Values." In Managing Air Quality and Scenic Resources at National Parks and Wilderness Areas, eds. R. D. Rowe and L. G. Chestnut. Boulder, CO: Westview Press.
Randall, Alan, and John R. Stoll. 1983. "Existence Value in a Total Valuation Framework." In Managing Air Quality and Scenic Resources at National Parks and Wilderness Areas, eds. R. D. Rowe and L. G. Chestnut. Boulder, CO: Westview Press.
Smith, V. Kerry. 1987. "Nonuse Values in Benefit-Cost Analysis." Southern Economic Journal 51:19-26.
U.S. Department of the Interior. 1991. "Natural Resource Damage Assessment: Notice of Proposed Rulemaking." Federal Register 56 (82): 19752-73.
|Printer friendly Cite/link Email Feedback|
|Author:||Bishop. Richard C.; Welsh, Michael P.|
|Date:||Nov 1, 1992|
|Previous Article:||Estimating the structure of industrial water demands: the case of Canadian manufacturing.|
|Next Article:||Site aggregation in a Random Utility Model of recreation.|