5
Economic Methods of Valuation

Chapters 3 and 4 discussed a wide array of services and amenities that biodiversity provides for people who might or might not value its individual components—individual genes, species, and ecosystems—and the diversity of components. Some aspects of biodiversity are valued directly; while others are valued for their contributions to ecosystem support and, hence, to sustainable production of things that are valued directly. The economic value of biodiversity has its place in the policy-making process. Although biodiversity might well have substantial economic value, compared with alternative consumptive resource uses, economic value does not tell us everything we need to know about the value of biodiversity.

Economic valuation is an attempt to provide an empirical account of the value of services and amenities or of the benefits and costs of proposed actions (projects or policies) that would modify the flow of services and amenities. Economic valuation provides a utilitarian account, that is, an account of contribution to the satisfaction of human preferences (see chapter 4 for a detailed discussion). Therefore, it provides a particular perspective on value—in this case, on the value of biodiversity. Utilitarians might object to some aspects of the economists' utilitarian account: to produce an economic account of contribution to preference satisfaction, a particular kind of structure has to be introduced into the analysis, and utilitarians will not always endorse the process or the results. In addition, there are many nonutilitarian perspectives on value (see chapter 4), which deserve consideration on their own merits.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 87
5 Economic Methods of Valuation Chapters 3 and 4 discussed a wide array of services and amenities that biodiversity provides for people who might or might not value its individual components—individual genes, species, and ecosystems—and the diversity of components. Some aspects of biodiversity are valued directly; while others are valued for their contributions to ecosystem support and, hence, to sustainable production of things that are valued directly. The economic value of biodiversity has its place in the policy-making process. Although biodiversity might well have substantial economic value, compared with alternative consumptive resource uses, economic value does not tell us everything we need to know about the value of biodiversity. Economic valuation is an attempt to provide an empirical account of the value of services and amenities or of the benefits and costs of proposed actions (projects or policies) that would modify the flow of services and amenities. Economic valuation provides a utilitarian account, that is, an account of contribution to the satisfaction of human preferences (see chapter 4 for a detailed discussion). Therefore, it provides a particular perspective on value—in this case, on the value of biodiversity. Utilitarians might object to some aspects of the economists' utilitarian account: to produce an economic account of contribution to preference satisfaction, a particular kind of structure has to be introduced into the analysis, and utilitarians will not always endorse the process or the results. In addition, there are many nonutilitarian perspectives on value (see chapter 4), which deserve consideration on their own merits.

OCR for page 87
Theoretical Foundations Welfare-Change Measurement The foundation of benefit-cost analysis (BCA) is welfare-change measurement: the benefit from some proposed action is the money-related welfare change that it generates. The concept of benefit is an increase in welfare, that is, preference satisfaction; and welfare change is measured in terms of money. Valid money measures of welfare change can be defined conceptually and can be estimated with reasonable accuracy, precision, and reliability; and individual welfare changes to arrive at social benefits and costs can be added up. Skepticism about any of those claims, in general or in the specific application to biodiversity, suggests that caveats should be applied to the interpretation of benefit-cost information or its use in policy decisions. The conceptually valid measures of welfare change are willingness to pay (WTP) for benefits and willingness to accept (WTA) for costs. WTP is the amount of money that someone would willingly pay to get a desired good, service, or state of the world rather than go without; WTA is the amount of money that would induce someone to willingly give up the good, service, or state of the world. Those measures are readily defined in market terms—WTP is the buyer's best offer, and WTA is the seller's reservation price (the price at which the seller will hold rather than sell)—but they are by no means restricted to commodity markets. Some people are willing to pay substantial amounts of money for improvements in the quality of their life. Some would willingly accept a lower level of amenities if compensated with money; for example, some would willingly move to an undesirable location if promised a large enough pay raise. For BCA of a policy proposal, aggregate benefits are defined as the sum of WTP figures for all those who stand to gain from the proposal. Aggregate costs are the sum of WTA figures for all who would provide goods and services or bear disamenities if the policy proceeds. Some critics object to aggregating benefits or costs that accrue to individuals, on the grounds that individuals with greater income and wealth tend to have greater WTP (or WTA) and that simple aggregation makes no attempt to correct for this or to place extra weight on things that benefit the disadvantaged. Given that many proposals promise benefits and costs continuing well into the future, the "bottom line" of the BCA is expressed as net present value, that is, the difference between the sum of present and future benefits and the sum of present and future costs, all discounted to the present. The practice of discounting has been controversial in some circles, especially in the context of environmental projects and policies (for example, Daly and Cobb 1989), where it is claimed that discounting tends to trivialize the demands of future generations for present conservation (see box 5-1) . That argument has been winning fewer converts in recent years (Heywood 1995), as economists have been reminding us

OCR for page 87
BOX 5-1 Does Discounting Harm the Future? The discounting of future benefits and costs is a practice introduced from financial analysis to account for the productivity of capital. In recent years, some environmental economists have been swayed by critics who worry that discounting implies that the concerns of the future (perhaps only a few decades hence) count only trivially in the calculations of the present. Thus, we have the discounting paradox: we must discount, it is claimed, to avoid damaging the future by making wasteful commitments of capital to unproductive projects; and we must not discount, it is also claimed, to avoid trivializing future demands for present conservation. The paradox can be resolved in the following way. We can be reasonably confident of two propositions: if the problem is simply to determine the rate of consumption from an endowment (the "cake-eating problem"), a society with a positive discount rate will choose a consumption path relatively high at the outset and declining over time (Page 1977); and if capital is productive and the young need to borrow it to produce efficiently, equilibrium interest rates will be positive and a policy of repressing the interest rate (undertaken, one imagines, to protect the future) will actually depress the trajectory of future welfare (Farmer and Randall 1997). That is, in a cake-eating economy, discounting is destructive of future welfare, but in a productive economy it is not. This resolution of the discounting paradox directs our attention to the real question: are we, or are we not, operating in an economy that is ultimately cake-eating? If capital accumulation is sufficiently high, renewable resources are managed carefully, capital and renewable resources are adequately substitutable for exhaustible natural resources, and technological development tends to enhance the substitutability of plentiful resources for those which are most scarce, the cake-eating problem can be avoided. In that case, concerns that discounting inherently damages the future are misplaced. It can also be argued that such policy interventions as the safe minimum standard, which address directly crucial natural resources, provide a more appropriate response to conservation crises within otherwise productive economies than would repression of the discount rate. that realistically high discount rates discourage wasteful investments that would actually harm future prospects. Current writers are skeptical about the wisdom of using low discount rates to achieve policy goals, preferring more direct approaches to the concerns of environmentalists. For example, Howarth and Norgaard (1991) argue that balancing equity among generations should be addressed by intergenerational transfers of resources, and Farmer and Randall (1997) suggest that targeted conservation policies provide the appropriate remedy to the extent that particular natural resources are both necessary for human welfare and threatened with exhaustion. Cost-effectiveness analysis can be useful in guiding decisions toward the

OCR for page 87
most efficient way of meeting specified goals. However, it does not provide estimates of values. If there are several ways of accomplishing a particular and well-specified goal, cost-effectiveness analysis compares the costs of the various approaches; the most cost-effective is the one that accomplishes the goal at the lowest cost. If different approaches would achieve different quantitative levels of performance, cost-effectiveness might be expressed as cost per unit of performance (for example, cost per acre preserved or cost per nesting pair saved). If the policy-maker is confident that the different approaches are otherwise equivalent in terms of the results achieved, choosing the most cost-effective approach is justified. Categories of Value WTP and WTA for some natural resource or amenity are equivalent to its total economic value. However, humans use and enjoy natural resources and amenities (as they do other goods and services) in a variety of ways. At one extreme, natural resources can provide commodities that are purchased and consumed directly; at the other extreme, people might enjoy satisfaction that a particular habitat is being maintained at high quality. Both kinds of use generate economic value, but it is likely to be expressed in different ways and via different institutions for commodities, in terms of quantities taken and prices paid and for habitat quality, perhaps via voluntary contributions to conservation organizations. Total economic values include all the several kinds of economic values that have been identified by economists. Total economic value is the WTP for a change in the state of the world. To impose some order and consistency, the following relatively simple classification of economic value is gaining ground among economists. Use value is generated when a person uses an environmental service actively, typically by consuming it directly or combining it with other goods and services and the person's own time to "produce" an activity that generates utility. Recreation experiences, for example, are produced by combining on-site amenities with travel services, recreation equipment, and the participant's time. Use value is likely to be reflected (at least in part) in behavior such as purchases and visits. Use value, naturally, includes the expected value of future use. If uncertainty attends future availability of an amenity or future demand for it and potential users are risk-averse, use value under uncertainty can include option value, the value of assurance that things (such as biodiversity) that are available now will still be available when we need them, and quasi-option value, the value of waiting to decide on the disposition of an asset (such as whether to build houses on Camp Pendleton—see the case study in chapter 1) motivated by the possibility that we will be able to make a "better" decision later, perhaps because we will have more information. When institutions provide opportunities for individuals to secure options for future use, these kinds of value might be reflected in behavior.

OCR for page 87
Passive use value captures the idea that people might enjoy satisfaction from "just knowing" (that is, enjoying the assurance) that a particular habitat is being maintained in good condition. There is no general expectation that passive-use value involves overt activities or is reflected in behavior. However, contributions to voluntary organizations that provide habitat preservation and political support for pro-habitat policies are consistent with passive-use value. Together, use value, option value, and quasi-option value make up total economic value. It also includes bequest value, in that bequest motives assume that one's heirs will enjoy use or passive use. Total economic value includes all the kinds of economic value. There is no claim that economic value, however, constitutes the totality of value. As chapter 4 has made clear, there are many ways of valuing , but, total economic value then represents a comprehensive application of the economic way of valuing. Methods of Valuation Valuation relies on detailed information from the natural sciences . We might value an environment as an asset, in which case its value would be the net present value of the services that it provides and will provide. Alternatively, we might evaluate some proposed action (a project or policy); value would then be the net present value of the change in services that the environment will provide minus the cost of implementing the proposed action. Either way, valuation requires detailed knowledge of the service flows of the environment, of the costs incurred in preparing these services for human enjoyment, and of the responsiveness of service flows and costs to human interventions (Randall 1987 and NRC 1997 provide conceptual models of the valuation process). Much of that information must originate with experts whose specialties are far from economics, for example, ecologists and hydrologists. Economic valuation depends heavily on information that is fundamentally noneconomic. Valuation also requires evidence of WTP and WTA. Evidence of WTP and WTA varies along two dimensions of quality: consistency with the conceptual framework of welfare-change measurement and reliability of the data themselves. For example, data generated by market transactions are convincing in at least one respect—paying money is the sincerest expression of WTP and accepting money and relinquishing an amenity constitute the sincerest expression of WTA. But the data might, for a variety of reasons, fail to measure the correct value concept. Price typically indicates marginal value (literally, the value of the next unit more or less than the status quo quantity—a small change); but a proposal might involve nonmarginal (big) changes. In addition, market distortions of various kinds might distort prices, markets might be incomplete or otherwise imperfect, and the environmental service involved might be nonmarketed. Data generated by contingent valuation or contingent policy referendums often can (because a researcher controls the valuation context) be addressed to the right value mea-

OCR for page 87
sure, but still this might raise doubts as to whether contingent payments and votes are reliable predictors of behavior. Valuation researchers are often faced with one or another form of this dilemma: "harder" data might depart from the ideal, and conceptually valid measures might be "softer". In some cases where hard data depart from the ideal value concepts, economists have developed ingenious methods of inferring the ideal values; however, there is always a risk that they will be forced to substitute assumptions for evidence and structure for information. The resulting value estimates will be to some degree artifacts of the methods used and the research decisions made. Direct and Indirect Evidence from Markets It is hard to imagine a market for biodiversity as a whole, but its various components are routinely marketed. Consider a biodiverse forest. Timber, fuel wood, and some nonwood products can be produced and sold. The forest can provide catchment for water that is valued by downstream farmers and urban residents. The forest ecosystem can harbor genetic resources with commercial potential, for example, rare species that might be of pharmacological interest or wild species that are precursors of modern, commercially important plant varieties. Recreationists and nature-lovers can devote resources (money and time) to visiting the forest. People can buy homes near the forest to have access to its amenities. The productivity and value of these various activities depend on how the forest is managed, so proposals that affect forest planning and management will generate costs and benefits that are reflected, to various degrees, directly or indirectly in markets. The Pacific Northwest case study in this chapter provides a detailed example. Market Demands and Prices For commodities that can be sold in quantities that are small relative to the total market, the economic value that can be assigned to a decision to sell or preserve is simply the product of the market price of the commodity and the quantity. For example, if 20 acres of old-growth timber is reserved from the market to protect a pair of spotted owls, one estimate of the cost of this decision is the product of the volume of the timber and the market price per unit volume. That holds as long as the quantity is so small that its removal from the market does not affect the market price of timber generally. But consider the Pacific Northwest (see case study, this chapter), in which the area of federal forest taken off the market to protect the spotted owl and other threatened species had accounted for some 10% of the nation's softwood lumber production. The quantity of timber removed from the market was clearly large enough to affect the market price for timber in much of the country. For such a decision, an estimate of the economic cost must consider not only the change in

OCR for page 87
the quantity of timber marketed from that area, but also the change in the market price per unit volume of the remaining marketable timber (from before the decision to after the decision). Such an estimate must also consider what economists refer to as "substitution effects". The changes in quantities and prices that result from a decision of this magnitude affect the market price of timber in other areas. Timber producers in forests other than those immediately affected by the decision (for example, the southern United States) respond to the change in timber price by changing the quantities of timber that they put on the market, thus causing further changes in the price of timber. Additional complications include the effects of the changes in timber price on the marketing of such substitute products as steel and plastics and the modifying effects of time as these various factors work through the marketplace. In sum, calculating the effects of decisions that affect market prices is not easy, but it is conceptually feasible. The goal of valuation—measuring net present value—introduces two complications. First, "net value" requires that any costs associated with using the resources, such as, timber-harvesting costs, be subtracted from gross value. Second, "present value" requires prediction of future demands for environmental services. In the cases of timber, water supply, and genetic materials, the forest augments the supply of things that are valued as factors in production. So the demand for forest products is a derived demand, which complicates predictions of demand: the analyst needs to be concerned with demands for the final products (houses, irrigated crops, and pharmaceutical products) and with the supply of other things that might serve as substitute factors in their production. The idea of substitutes suggests another approach to valuation: when it is hard to observe market demands for forest products directly, the analyst might look to market evidence concerning substitutes. For example, the avoidance cost method might value improved water quality by observing the household water-filtration costs avoided, and the replacement-cost method might value increased water catchment by calculating the cost of additional reservoir capacity that would serve the same purpose. In both cases, the methods provide an upper-bound value for the particular services they address: the services cannot be valued at more than the cost of avoiding the need for replacing the service with a perfect substitute, but they could be valued at less than that, in the event that effective demand would not clear the market for these services at these prices. The Quabbin Reservoir and Lake Washington case studies in chapter 6 illustrate this. Travel-Cost Methods Recreationists spending their money and time to visit the forest leave a trail of indirect evidence about their WTP for the services and amenities that it provides, and travel-cost methods attempt to tease out this WTP. The weak-complementarity assumption, of course, limits the travel-cost method to estimating the

OCR for page 87
use values associated with site amenities. The simplest travel-cost models posit simply that the number of visits, at a given level of site quality, is a function of travel costs and socioeconomic variables, where travel cost is a proxy for the ''price'' of visits and includes costs of distance traveled and time spent in traveling. Substitute sites and activities typically are included in arbitrary fashion or assumed to be of little import (formally, this is accomplished via assumptions of separability in the utility function). A large literature attests to the difficulty that researchers have experienced in estimating the cost of travel time (Bockstael 1995), but this is symptomatic of a general difficulty: it is inherently difficult for researchers to observe the cost of a visit, that is, the value opportunities foregone to make the visit (Randall 1994). If one assumes a relationship between the quality of on-site amenities and the costs of goods and services used in traveling to the site, the value of an increment or decrement in site quality is measured as the integral between demands for visits at the with-proposal and without-proposal levels of site quality. The random-utility model (RUM) has become the travel-cost model of choice (Bockstael 1995) because its systematic treatment of substitute sites allows it to characterize site quality more completely. RUM models are therefore more useful than basic travel-cost models for valuing changes in levels of environmental amenities. Their disadvantage arises from their substantial information needs, which in practice often lead to the use of very large data sets and simplifying analytical assumptions that impose rigidities; thus, estimates based on travel-cost models are to some degree influenced by researchers' analytical choices. When travel-cost models are used to predict number of visits, validation is relatively simple, and several well-known models have performed well (for example, Bockstael and others 1987). However, direct validation of the value estimates obtained with travel-cost models is impossible; the best one can do is test for convergence of the results of travel-cost methods and the results of alternative approaches, such as contingent valuation, and such tests have provided some empirical evidence of convergent validity. Hedonic Price Analysis Hedonic price analysis separates the factors that contribute to prices to identify the contribution of those based on environmental amenities. Imagine a good with several important or desirable features, such as a house or automobile. It is a reasonable hypothesis that the price of a particular house or car reflects its particular characteristics. If a statistical analysis succeeds in explaining the price of a house as a function of its characteristics and one of those characteristics is the level of environmental amenities, then the marginal (small) impact of a change in an amenity level (a trait that makes it attractive) on the house price should provide evidence of this amenity value. This is the intuition behind hedonic price analysis. A hedonic price function, relating house prices to characteristics, is

OCR for page 87
estimated. Typically, three kinds of characteristics are used: on-site characteristics, such as the number of bedrooms; neighborhood characteristics, such as school quality; and environmental amenities, such as access to a biodiverse forest. The first derivative of the hedonic price function with respect to the environmental characteristic of interest is its hedonic price (or marginal implicit price), a measure of the marginal value of the amenity. The literature suggests that hedonic price analysis has succeeded, in a fairly wide range of circumstances, in generating plausible estimates of marginal hedonic prices for various housing characteristics, including environmental amenities. To value nonmarginal changes in amenity levels, however, it is necessary to estimate hedonic demands, that is, demands for amenities. The literature reports many attempts to find conceptually valid methods of identifying hedonic demands, but no method has proved generally acceptable. Hedonic price analysis is often effective for valuing marginal changes in the levels of environmental amenities that can be accessed via, say, choice of home site but cannot generally be used for valuing nonmarginal environmental changes. The assumptions underlying the method limit its application to a subset of use values; for example, a housing-price hedonic analysis will measure use values associated with home site amenities, but not values that can be accessed regardless of exactly where one lives. Evidence from Self-Reports If we design and ask the questions with enough care, perhaps people can provide reliable evidence of amenity values by telling us their WTP or WTA directly or by telling us what they would do (for example, buy or not buy or vote yes or no) if given well-specified choice situations that we construct to generate data that we can analyze to infer their WTP or WTA. That is the intuition behind contingent valuation and contingent-choice experiments. The great advantage is that the researcher controls the context of choice, which makes it possible to estimate total economic value, passive-use value, and various use values that can elude the methods that use market-generated evidence, directly or indirectly. A further advantage is that information can be obtained to value amenity levels beyond the existing range; if it can be described by the researcher and comprehended by the respondent, it can be valued. The potential disadvantages lie in the self-reported nature of the data: some people might seek to answer strategically, some might answer carelessly, and some might struggle mightily (but hopelessly in the end) to provide valid responses to questions that cannot be answered meaningfully. Economists, who are weaned on the admonition to "watch what people do, not what they say", approach these methods with a well-developed skepticism; yet the results, although mixed, have been encouraging enough to stimulate a proliferation of applications. The techniques require primary data collection in a survey or experimental

OCR for page 87
context. With rapid advances in information and communication technologies and increasing synergism among research programs in, for example, economics, social psychology, and marketing, it is reasonable to expect vigorous innovation in research design and data collection methods. In this report, we use the standard categories of contingent valuation (in which responses to one or a few choice questions provide the basic data for valuation) and contingent-choice experiments (in which value is inferred from responses to a sometimes long sequence of pairwise choices). The basic project underlying the methods is to learn about value from people's self-reports; and as development and testing of these methods proceed, we can expect new approaches to emerge and existing categorizations to become obsolete. Contingent Valuation The essential elements of a contingent-valuation (CV) exercise are a description of the default and alternative situations (respectively, what you get if the proposal fails and if it passes), the institutional environment, the valuation question, and the policy-decision rule: How does the answer to the valuation question affect whether the proposal passes or fails? (See the Grand Canyon flush case study below.) The valuation question can be continuous (or open-ended); for example, What would you be willing to pay? Or it can be in the form of a dichotomous choice; for example, Given the stated cost to you and the policy-decision rule, would you vote yes or no? (Alternatives in common use are, Would you buy it or not?; Would you donate to the trust fund or not?) The different forms of the valuation question require different analyses to estimate WTP or WTA; for example, the results of the dichotomous form are usually analyzed with some kind of RUM (Hanemann 1984). With different policy-decision rules, they imply different kinds of incentives for truthful responses (Hoehn and Randall 1989). There is already an extensive literature of CV applications, and attempts to validate CV include tests for internal consistency and tests of convergence with value estimates obtained with different methods. Encouraging results have been obtained (for example, Carson and others 1996; Smith and Osborne 1996), but critics have raised enough doubts (for example, Hausman 1993) for CV to remain controversial. A 1993 report by a prestigious panel (Arrow and others 1993) failed to settle the issues when it endorsed CV in principle, even for measuring passive-use values in environmental-damage litigation, but announced a long and demanding list of standards that a valid CV should satisfy. CV that would meet the panel's standards would be prohibitively expensive in most applications, and, as methodological innovation and the accumulation of evidence proceed, the process of rethinking the panel's recommendations is beginning. One of the panel's recommendations deserves highlighting here. In keeping with a good deal of professional opinion, the panel concluded that CV could not

OCR for page 87
be endorsed for estimating WTA directly—that whereas WTA is the appropriate measure of value for decrements in environmental services, considerations of reliability lead to the recommendation that, instead, self-reports of WTP to avoid loss can substantially understate WTA (Hanemann 1984). The panel's recommendation would have the effect, therefore, of undervaluing the losses from destruction of unique ecosystems. Contingent Choice Experiments Open-ended CV sets a rather difficult task for respondents (announcing a dollar value of some nonmarketed amenity), and dichotomous-choice CV sets a simpler task (announcing whether a proposal is accepted or rejected at a specified cost) but collects only one or two valuation data points. It might be argued that progress could be made by having respondents make a larger number of simple pairwise choices. That is the motivation for contingent-choice experiments, in which data generated by a sequence of pairwise choices are analyzed with RUMs to generate value estimates. As Adamowicz and others (1994) demonstrate, these methods have another potential advantage: contingent-choice and actual-choice data can be combined to extend the range of data points and to test for consistency between the two kinds of data. The methods also have disadvantages: a long sequence of pairwise choices can tax respondent's patience, and the RUM analytical framework imposes rigidities on the analysis. Contingent-choice experiments are a fairly recent development, so the evidence on their performance is rather thin. Initial applications have emphasized amenity-use values, but there is no inherent reason why they could not be used to estimate passive use and total values, and (given the CV controversies) current research in environmental-damage assessment is heading in that direction. Case Study: The Grand Canyon Flush Dams and reservoirs on major streams affect downstream conditions by changing water flows, water temperatures, sediment loads, and the character of stream-bottoms and beaches. A large-scale test of the potential for reestablishing stream-bottom and beach characteristics in the Grand Canyon of the Colorado River was conducted in the spring of 1996. A week-long surge of water through the canyon was provided by opening the gates on the Glen Canyon Dam just upstream of the Grand Canyon. Initial results of the experiment were evident almost immediately, even while the "grand flush" was in progress. A major purpose of the experiment was to determine whether sandbars along the river could be restored for recruitment of riparian trees and shrubs, an important element in the canyon's ecosystem; a finding before the experiment indicated that the dam-caused lack of springtime floods had reduced the energy in the river needed to lift bottom sediments onto adjacent sandbars. An additional goal was

OCR for page 87
a small group of specialists in each category of species to come to an agreement in a short time with only the currently available information. Perhaps years of additional research would lead to different estimates, but resource managers usually cannot wait that long. The approach provided reasonably adequate information for recognizing real differences in habitat outcomes among the management options that were considered. The economics-related measures of differences in effects among the options were also reasonable, although limited. The decision was political. Effects on employment and receipts from timber sales are relevant when federal forests are involved. But they are only partial measures of the values involved in maintaining some degree of biodiversity. Effects on communities in the region and the value of interagency and citizen collaboration in making these kinds of decisions were also recognized in the FEMAT report. And there was a short discussion of the economic effects of increased prices (due to decreased federal timber harvests) on consumers of wood products. But estimates of the various effects with the different options were not included in the report, largely because of conceptual and measurement problems. There was no single "currency" with which the value of biodiversity could be measured for this resource management decision even if there had been clear agreement on how to measure differences in biodiversity. The size of the region was based largely on the range of the northern spotted owl, which was the subject of the original court case. That meant it was a very large area for focusing on some kinds of effects, such as human-community impacts, which vary considerably from one locale to another whose measurement can easily be lost when impacts on individual communities are melded into state or multistate regional estimates. At the same time, once the issue was defined to include the impacts of forest management on salmon, it was difficult to leave out the effects on salmon of management of ocean fisheries, which extend the scale even more. In contrast, evaluation of habitat outcomes for some of the species that have restricted ranges, such as some of the mollusks, might have lost something by being part of a bigger analysis. This case shows of the incremental effects of a resource-management decision involving biodiversity. The FEMAT analysis provided input for the decision by showing what would be gained and lost for each additional increment of protection of old-growth forest habitat. The basic structure of the analysis was appropriate even if the analysis was limited by gaps in available information. Roles for the Various Valuation Methods Direct market evidence can be useful for valuing natural-resource commodities harvested from biodiverse environments, genetic resources useful (for example) in plant breeding for agriculture and forestry, pharmaceutical resources, and so on.

OCR for page 87
The travel-cost method is useful for evaluating recreational-use benefits. The growing demand for adventure travel and ecotourism suggests that such benefits will play an increasing role, especially in habitats where charismatic megafauna are present and tourism can be managed compatibly with species and habitat preservation. Hedonic price analysis can have a role in benefit estimation, for example, in cases where a market develops in land near habitat reserves so that land value reflects demands for amenities generated by biodiversity. Nevertheless, it is reasonable to expect the role of such analysis to be modest and occasional. The methods relying on direct and indirect evidence from markets have important limitations for valuing biodiversity: They have some conceptual and methodological problems. They are limited to a subset of use values; passive-use value and holistic total economic value are beyond their reach. It is difficult to implement conceptually valid piecewise valuation procedures that would give these methods a role in eclectic valuation schemes that use different methods to value different components of the suite of biodiversity services and amenities. Contingent valuation, although controversial, is the obvious method for valuing biodiversity because it is, at least in principle, capable of valuing nonmarket-use values, passive-use values, and total economic value. Nevertheless, biodiversity presents serious challenges for CV in that respondents often are asked (of necessity) to value relatively unfamiliar services and amenities. Of the more than 2,000 publications to date involving CV, relatively few have addressed biodiversity, habitats, or endangered species. Passive-use values, because they are nonrival (that is, passive users do not compete with each other for access), can be very large in the case of environmental services that appeal to a large number of people. In a CV of viewing of elephants in Kenya, Brown and Henry (1993) elicited WTP for maintenance of elephant populations. Their results show that visitors gain about $25–30 million per year in consumer surplus (value over what they actually pay) for viewing elephants, a proportion of which is likely to represent passive use value. On wider habitat protection, Moran (1994) shows that the consumers' surplus attached to nonconsumptive use of Kenya's protected areas by foreign visitors (as a subset of all users) is about $450 million. It is safe to conjecture that those values would be overshadowed by the passive-use values of nonvisitors if these had been measured. Given the broad applicability of CV to valuing biodiversity, it is important to address the criticisms that have been raised about CV: Validation of CV is inherently difficult. In the absence of convincing validation, and given the very large value estimates that can be expected for

OCR for page 87
passive-use value and total economic value of prominent biodiversity resources, CV and the value estimates that it generates will remain controversial. CV surveys of biodiversity or species-preservation issues often generate a relatively high proportion of protest or refusal responses, and some respondents indicate an unwillingness to address these issues in terms of trades for money. Good CV design—for example, structuring the CV as a referendum about spending more or less public money for preservation projects—can minimize the occurrence of protest or refusal responses. Nevertheless, some refuse to respond to even the best CV questions, and some of these nonrespondents are thoughtful people who draw on nonutilitarian moral philosophies when trying to resolve biodiversity issues. Chapter 4 makes clear that these are legitimate reactions, and they illustrate the limits of utilitarian CV in dealing with nonutilitarian concerns. CV comes in a variety of forms, each with its own communication and incentive properties, so blanket claims about the validity of CV are meaningless. But one constant is that the validity of any CV effort depends on respondents' understanding of what is being valued. In the case of biodiversity, citizen knowledge of the details of any particular case is likely to be quite low, so researchers will need to provide a good deal of case-specific information. Therefore, issues of communication and comprehension are likely to be prominent in criticism of many CV efforts directed at biodiversity. It is important to recognize that this problem applies also to any other approach or process that takes citizen opinion seriously. Contingent-choice experiments are still in their infancy, especially in contexts where passive-use values can well dominate. Nevertheless, one might expect increasing application of these methods. Examples Various estimates have been made of the value of aspects of biodiversity. They include in this report the estimates for ecosystem services in chapter 3, the estimates in the Pacific Northwest forests and Grand Canyon flush case studies in this chapter. In addition, the article by Costanza and others (1977) discussed in a later section, A Tempest Over Valuing the World's Ecosystem Services, has global estimates of average per hectare and total values of biodiversity for 17 ecosystem services and 17 biomes. In most of the cited examples, as well as in most of the numerous other published examples, the value estimates are for some particular element of biodiversity or for services that are related to some element of biodiversity. The estimates in the paper by Costanza and others (1997) are unusual in that the sum of the values for the 17 ecosystem services represents estimates of one-time annual values or the present value of the stream of expected future values.

OCR for page 87
BOX 5-2 Why Benefits and Costs Matter When called on to defend the systematic use of BCA in public decision processes, economists are likely to start talking about the need to impose a market-like efficiency on the activities of government (for example, Arrow and others 1996). However, the reasons are not immediately obvious. It can be argued coherently that, although society delegates many workaday decisions to the market, government is an institution that human societies invoke when they choose, for their own reasons, not to be efficient. It is not a frivolous point: efficiency is a harsh discipline and one that in practice tends to reinforce the distributional status quo; it is by no means clear that society ought to impose that discipline on everything that it does. So we must look elsewhere for good reasons to take benefits and costs seriously in public policy. Benefits and costs cannot count for everything. Hubin (1994) asks us to consider benefit-cost moral theory: that right action is whatever maximizes the excess of benefits over costs, as economists understand the terms benefits and costs. It is hard to imagine a single supporter of such a moral theory among philosophers or the public at large. Instead, we would find unanimity that such a moral theory is inadequate and an enormous diversity of reasons as to exactly why. Benefits and costs must count for something. This argument comes in two parts. First, benefits and costs provide a fairly good account of contribution to preference satisfaction (Hubin 1994). Second, preference satisfaction matters morally. One cannot imagine a plausible moral theory in which the level of satisfaction of individual preferences counts for nothing at all. If we examine a broad array of contending moral theories, preference satisfaction counts for something in each of them. Public Roles for Benefit and Cost Information Preference satisfaction to inform decisions, rather than to decide issues. Because preference satisfaction is a consideration under any plausible moral theory, Benefit-Cost Analysis in the Federal Government Benefit-cost analysis (BCA) is a generic term that can refer to nearly any comparison of benefits and costs as long as they are measured or estimated in comparable units. In the federal government, the term sometimes refers to protocols for comparing the "desirable and undesirable impacts of proposed policies" (see box 5-2) (Arrow and others 1996). An early use of a formal process of BCA was in the evaluation of federal water-resources development projects after enactment of the Flood Control Act of 1936. The act required that proposed projects be undertaken only "if the benefits to whomsoever they accrue exceed the costs." That, of course, is consistent with the progressive model of "scientific govern-

OCR for page 87
an account of benefits and costs might be used routinely as a component of some more comprehensive set of evidence, accounts, and moral claims to inform the decision process. This notion is congenial to many economists (for example, Arrow and others 1996, p 221). However, it leaves unanswered the question, How should an account of preference satisfaction be weighted relative to other kinds of information? And can the answer be principled, or must it always be circumstantial? Preference satisfaction subject to constraints. One alternative way of coming to terms with the idea that preference satisfaction counts for something in any plausible moral theory but cannot count for everything is to endorse preference satisfaction as the decision rule for issues in which no overriding moral concerns are threatened. Preference satisfaction could then be decisive within some broad domain, which itself would be bounded by constraints reflecting rights that ought to be respected and moral imperatives that ought to be obeyed. That would implement the commonsense notion that preference satisfaction is fine as long as it does not threaten any concerns that are more important. The general form of such constraints might be, "don't do anything disgusting." The basic idea is that the public decision-maker should adopt—or a pluralist society would agree to be bound by—a general-form constraint to eschew actions that violate obvious limits on decent public policy. That kind of constraint is in principle broad enough to take seriously the objections to unrestrained pursuit of preference satisfaction that might be made from a wide range of coherent philosophical perspectives. Examples of such constraints might include these: Don't violate the rights that other people and perhaps other entities might reasonably be believed to hold. Be obedient to the duties that arise from or could reasonably be derived from universal moral principles. Don't sacrifice important intrinsic values in the service of mere instrumental ends. In each of those cases, the domain within which pursuit of preference satisfaction is permitted would be bounded by nonutilitarian constraints. In the context of protection of habitats, species, and ecosystems, society could decide on the basis of preference satisfaction but subject to some kind of conservation constraint. ment", a model that achieved great influence in the first half of this century (Nelson 1987). Guidance for analyses of federal water-resources projects grew in a series of documents (the "Green Book" of 1947, Senate Resolution 148 of 1957, Senate Document 97 of 1962, and so on) that ultimately provide for a "four accounts" model for water-resources planning: the national economic-development account (basically, a BCA); the regional economic-development account, which focuses on income and employment effects; the environmental-quality account, which overlaps with the national economic-development account to the extent that it proves feasible to determine the benefits and costs of at least some of the antici-

OCR for page 87
pated environmental-quality changes; and the quality-of-life account. Both benefits and costs were to be measured in terms of estimated market values. The approach appears straightforward on the surface, but numerous issues arise in trying to apply the general guidelines to proposed projects, including defining the appropriate scope of the analysis, deciding how to account for regional and local effects, and especially deciding how to treat real benefits and costs for which market values do not exist. Concerns in the last 2 decades with the presumed high costs of federal regulations has broadened the use of BCA by the federal government. Executive order 12291 of 1981 required analysis of benefits and costs of all new federal regulations with major economic and other effects. It stated that regulatory actions should maximize "net benefits to society" and should not be undertaken "unless the potential benefits to society . . . outweigh the potential costs to society." The order, taking note of the difficulty of estimating both benefits and costs in monetary terms, required that proposed rules include descriptions of benefits and costs that cannot be quantified and that the determination of net benefits include "an evaluation of the effects that cannot be quantified in monetary terms.'' Executive order 12866 of 1993 replaced executive order 12291. It, too, required that assessments of benefits and costs include qualitative measures of those which are difficult to quantify. A recent article by several prominent economists noted that estimates of benefits and costs of regulations should be accompanied by a description of the uncertainties surrounding the estimates and that the analyses should also identify distributional (equity) consequences (Arrow and others 1996). But the executive orders provide few clues about how these qualitative measures are to be made and used in analyses. BCA for federal regulations continues to be treated on a largely ad hoc and project-by-project basis, especially where the market provides little guidance on values. With the fading of the progressive dream, a pluralistic, participatory process has emerged. Instead of trusting in the experts to get things right, citizens seek access to the decision-making process. BCA has a somewhat different role in such a process: it will have influence to the extent that citizens are convinced, first, that benefits and costs are relevant considerations and, second, that the particular BCA is reasonably accurate and reliable. The second of those concerns—essentially, the quality of benefit-cost information—is a serious concern in the context of biodiversity, but it has already been discussed here at some length. Here, we address the first concern: Can we give good reasons why benefits and costs are relevant considerations in policy decisions? A BCA is an account—not a perfect account but a fairly good account—of the prospective contribution of some proposed action to the satisfaction of human preferences. Because preference satisfaction cannot logically count for everything but also cannot logically count for nothing, benefits and costs will be

OCR for page 87
relevant considerations in policy decisions but will not be the only relevant considerations (Hubin 1994; Randall 1999). One appealing answer to the question of when BCA counts and how much it counts is that an ethically pluralistic society of thoughtful moral agents would have good reasons to agree to choose on the basis of benefit-cost considerations when nothing more important is at stake and to impose the most important of these "more important" considerations as constraints on what can be chosen (Randall 1999). In the case of biodiversity, the appropriate constraints can well include safe minimum standards of conservation for critical species and habitats (see box 5-3) (Farmer and Randall 1998). BOX 5-3 Safe Minimum Standard of Conservation Ciriacy-Wantrup (1968) proposed the safe minimum standard (SMS) of conservation. The stock of a renewable resource (for example, a species or ecosystem) would be maintained at a level at least high enough to protect it from potential extinction unless the costs of so doing were "immoderate" (Ciriacy-Wantrup) or "intolerably high" (Bishop 1978). The SMS is designed not to serve as a comprehensive conservation policy, but to impose a constraint on "business as usual'' to protect against disasters. Some philosophers and economists have criticized the SMS, charging that it is an ad hoc policy switch that cannot be grounded in a single, clear objective statement; in other words, whatever justifies "business as usual" cannot also justify the SMS. Farmer and Randall (1997) have responded that standard utilitarian, contractarian, and Kantian accounts of societal ethics encounter irresolvable problems when extended to conflicts between existing generations and potential future generations. In the absence of philosophies that resolve intergenerational conflicts in an internally consistent manner, such safeguards as an SMS constraint make sense. There is disagreement among ethical systems as to the interpretation of the escape clause: How high would the cost of maintaining the SMS have to be, to be judged intolerable? In some utilitarian interpretations, the intolerable cost is quite low, so the SMS constraint becomes little more than a reminder to perform BCA carefully, to take option and existence values seriously, and to give preservation the benefit of any doubt. Kantians, however, could argue coherently that the intolerable cost should be high enough to impose genuine hardship. Finally, to what species and ecosystems should the SMS be applied? At one extreme, there can be broad agreement that things that are essential for human welfare should be subject to the SMS. At the other extreme, the "precautionary principle" (that we should apply the SMS to every species ''just in case") is supported by relatively few and is impracticable if taken literally. Clearly, commitment to an SMS constraint in principle leaves some important issues unresolved.

OCR for page 87
A Tempest over Valuing the World's Ecosystem Services In May 1997, Robert Costanza and a long list of coauthors published a paper titled "The Value of the World's Ecosystem Services and Natural Capital" in Nature. They estimate the annual value of the world's ecosystem services to be about $36 trillion, compared with an estimate of about $18 trillion for the world's annual gross product. The reaction of neoclassical economists is best typified by V. Kerry Smith's response (Mispriced Planet, Summer 1997, Regulation): "Their results should not be used in any form—whether as measures of the importance of changes in natural resources to human welfare; as yardsticks for future project appraisals; or as sources of a road map for future research." Although there are many technical issues for debate, the core of the argument arises because of the claim by Costanza and others (1997) that their estimate of the value of ecosystem services is based on the concept of WTP. A thought experiment is useful. Imagine that evil aliens orbit Earth and threaten to destroy ecosystem resources one by one unless we pay blackmail in the form of an annual fee for each service. Costanza and his colleagues are quickly assembled to value each category of ecosystem services. The first resource threatened is forests, which generate $4.7 trillion per year, on the basis of the estimated WTP of the world's countries for the forests' total ecosystem services (Costanza and others 1997). On the basis of the group's recommendations, Earth agrees to pay $4.7 trillion each year. Next the aliens threaten the coastal shelves, worth $4.3 trillion. However, having already agreed to pay $4.7 trillion for forests, reducing available world gross product for human consumption from $18 to $13.3 trillion, Earth opposes the Costanza estimates because "we cannot afford $4.3 trillion more; we are much poorer now." In other words, the demand and value for coastal shelves is reduced because available gross product from which to pay is reduced. If we follow this line of argument, the world's total annual gross product ($18 trillion) is the most that could be paid as a bribe to save the world's ecosystem services without reducing the accumulated value of the the world's capital. The value estimates of Costanza and others (1997) are based on separate studies of the values of individual components, each of which assumes that people's incomes remain at current levels. The problem has been termed the independent valuation and summation problem by Hoehn and Randall (1989), who argue that it is inappropriate to simply add the values obtained from independent studies, because aggregate values will be overstated. It is clear from the way that Costanza and others construct their estimates that their work does, in fact, suffer from the independent valuation and summation problem. However, the story is not over. Costanza and others (1997) respond to Smith with a substantive counter-argument of their own. Because ecosystem services are, for the most part,

OCR for page 87
unpriced, the sum of the world's gross product underestimates world income. Furthermore, the actual value of the world's ecosystem services would increase through proper management if the resources were properly priced. A simple example will illustrate their argument: Because of overfishing, the North Atlantic fishery is now capable of contributing little to the world's gross product. With proper management (which might include putting a price or tax on each fish taken from the sea to discourage overfishing), the fishery would be restored, the sum of the world's gross product would go up, and our ability to pay a bribe to protect coastal shelves from alien destruction would increase. The utility of the paper by Costanza and others (1997) is not in its estimates of the value of the world's ecosystem services, but rather that it initiated a visible discussion of the difficulties of estimating such values, whether on a global or on a more localized basis. As was pointed out in chapter 2, biological systems are complex. The debates over the Costanza paper point to the complexity and interactions of economic systems. These debates have contributed to a better public understanding of the difficulties in estimating economic values, especially in the absence of market-price information. As long as the value of most ecosystem services is not subjected to a market test, such debates will continue, and in the end they will advance understanding not only of the issues, but also of the values that are involved. Summary Economists have developed an array of tools for estimating values when the lack of ordinary markets precludes use of their favored measure, market-determined prices. These are powerful tools for informing decisions involving biodiversity. But they have limitations. Estimates of value based on them should be treated with careful attention to the assumptions that have been made in obtaining them. Support for their veracity can be indicated by the degree to which results obtained from various estimates converge. Particular care should be taken as the scale of the decisions for which estimates of value are made diverges from the normal scale of market processes. The economist's usual view of market decisions as being made at the margin—that is, for small changes in quantities and prices—is a key assumption for most estimates of value. References Adamowicz WL, Louviere J, Williams M. 1994. Combining revealed and stated preference methods for valuing environmental amenities. J Envir Econ Mgmt 26:271–92. Arrow KJ, Solow R, Portney PR, Leamer EE, Radner R, Schuman J. 1993. Report of the NOAA panel on contingent valuation. Washington DC: GPO. Arrow KJ, Cropper ML, Eads GC, Hahn RW, Lave LB, Noll RG, Portney PR, Russell M, Schmalensee R, Smith VK, Stavins RN. 1996. Is there a role for benefit-cost analysis in environmental, health, and safety regulation? Science 272(5259): 221–2.

OCR for page 87
Bishop RC. 1978. Economics of endangered species. Amer J Agric Econ 60:10–18. Bockstael N. 1995. Travel cost models. In: Bromley DW (ed). The handbook of environmental economics. Cambridge MA: Basil Blackwell. p 655–71. Bockstael N, Hanemann ME, Kling CL. 1987. Estimating the value of water quality improvements in a recreational demand framework. Water Resour Res 23:951–60. Brown GM, Henry W. 1993. The economic value of elephants. In: Barbier EB (ed). Economics and ecology: new frontiers and sustainable development. London UK: Chapman & Hall. Carson R, Flores N, Martin K, Wright J. 1996. Contingent valuation and revealed preference methodologies: comparing the estimates for quasi-public goods. Land Econ 72:80–99. Ciriacy-Wantrup, S von. 1968. Resource conservation: economics and policies, 3rd ed. Berkeley CA: Univ Calif Div Agric Sci. Costanza R, d'Arge R, de Groot R, Farber S, Grasso M, Hannon B, Limburg K, Naeem S, O'Neill RV, Pareuelo J, Raskin RG, Sutton P, van den Belt M. 1996. The value of the world's ecosystem services and natural capital. Nature 387:253–60. Daly HE, Cobb JBJ. 1989. For the common good: redirecting the economy toward community, the environment, and a sustainable future. New York NY: Beacon City Pr. Farmer M, Randall A. 1997. Policies for sustainability: lessons from an overlapping generations model. Land Econ 3:608–22. Farmer M, Randall A. 1998. The rationality of a safe minimum standard of conservation. Land Econ 74:287–302. FEMAT [Forest Ecosystem Management Assessment Team]. 1993. Forest ecosystem management: an ecological, economic, and social assessment. Washington DC: USDA, Forest Service, and USDOI, Fish and Wildlife Service, and others. Hanemann WM. 1984. Welfare evaluations in contingent valuation experiments with discrete responses. Amer J Agric Econ 66:332–41. Hausman J. 1993. Contingent valuation: a critical assessment. Amsterdam Netherlands: Elsevier Science. Heywood CH (ed). 1995. Global biodiversity assessment. Cambridge UK: Cambridge Univ Pr. Hoehn J, Randall A. 1989. Too many projects pass the benefit cost test. Amer Econ Rev 79:544–51 Hoehn J, Randall A. 1987. A satisfactory benefit cost indicator from contingent valuation. J Envir Econ Mgmt 14:226–47. Howarth RB, Norgaard RB. 1990. Intergenerational resource rights, efficiency, and social optimality. Land Econ 66:1–11. Hubin DC. 1994. The moral justification of benefit/cost analysis. Econ Phil 10:169–94. Kant I. 1991. Philosophical writings. New York NY: Continuum. (Behler E [ed]). Moran D. 1994. Contingent valuation and biodiversity conservation in Kenyan protected areas. Biod Cons 3(8): 663–84. NRC [National Research Council]. 1996. River resource management in the Grand Canyon. Washington DC: National Acad Pr. 226 p. NRC [National Research Council]. 1997. Valuing ground water: economic concepts and approaches. Washington DC: National Acad Pr. Nelson RH. 1987. The economics profession and the making of public policy. J Econ Lit 25:49–91. Norton B. 1988. Commodity, amenity, and morality: the limits of quantification in valuing biodiversity. In: Wilson EO, Peters FM (eds). Biodiversity. Washington DC: National Acad Pr. p 200–205. Page T. 1977. Conservation and economic efficiency. Baltimore MD: Johns Hopkins Univ Pr. Randall A. 1999. Taking benefits and costs seriously. In: Tietenberg T, Folmer H (eds). The international yearbook of environmental and resource economics. Cheltenham UK, Northampton MA: Edward Elgar. Randall A. 1994. A difficulty with the travel cost method. Land Econ 70:88–96. Randall A. 1987. Total economic value as a basis for policy. Trans Amer Fisheries Soc 116:325–35.

OCR for page 87
Shapard, R. 1996. A grand experiment brings spring floods to the canyon. Amer City Country 111:26 Smith VK, Osborne LL. 1996. Do contingent valuation estimates pass a Scope test? A meta analysis. J Envir Econ Mgmt 31:287–301.