Copyright (c) 1999 The Columbia Law Review

Columbia Law Review

 

May, 1999

 

99 Colum. L. Rev. 941

 

Environmental Regulation, Cost-Benefit Analysis,

and the Discounting of Human Lives

 

Richard L. Revesz


Introduction

 

 The use of cost-benefit analysis has become commonplace in environmental and other health-and-safety regulation. Such analysis is now mandated by Executive Order 12,866 for all major regulations,   n1 and may eventually be required by statute if Congress passes one of the various regulatory reform bills that have been pending for some time.   n2 The primary benefit of many important environmental statutes, as determined by the dollar value assigned by cost-benefit analysis, is the human lives that are saved.   n3 Thus, in determining whether a particular regulation can be justified on cost-benefit grounds, the central questions revolve around  [*944]  the value assigned to the lives that would be saved by the program. Probably the most vexing problem concerning these valuations has been whether to discount the value of a life saved to account for the fact that the loss does not occur contemporaneously with the exposure to certain contaminants.

With respect to this issue, two opposing camps have developed among regulators, judges, and academics. A similar controversy has arisen in connection with other regulatory programs,   n4 as well as with the provision of medical services.   n5 Supporters of discounting argue that the value of human life must be treated in the same manner as the value of any other benefit or cost: because other benefits and costs are normally discounted to present value when they occur in the future, the value of life should be discounted as well.   n6 In contrast, opponents of discounting claim, generally by appeals to notions of ethics and morality,   n7 that lives saved in the future are no less valuable than lives saved in the present. As a result, they argue that discounting is inappropriate.   n8

 [*945]  The debate, which is not confined to the United States,   n9 has taken on a relatively high profile, including discussion in the popular press.   n10 For example, the issue played a role in the Senate's scrutiny of the unsuccessful nomination of Judge Douglas Ginsburg to the Supreme Court of the United States in 1987,   n11 and attracted the attention of Vice President Albert Gore during the 1992 presidential campaign.   n12

The discussion of the propriety of discounting human lives often conflates two different sets of problems.   n13 In the first, the benefits will not accrue until the future because the harm has a latency period. For example, an individual exposed to a carcinogen faces an increased probability of dying at some point in the future, perhaps twenty or thirty years later. In the second, the benefits of controls accrue primarily to future generations. Climate change caused by the presence of anthropo genic gases in the atmosphere is a prominent example of this phenomenon.

 [*946]  The question of how to value lives threatened by latent harms was starkly posed in a regulatory proceeding that took place in the late 1980s in connection with a partial ban on the use of asbestos promulgated by the Environmental Protection Agency (EPA).   n14 The Office of Manage ment and Budget (OMB), which is responsible for reviewing regulations to ensure their consistency with cost-benefit principles,   n15 strongly urged discounting the value of human lives over the period of latency of the harm; under its then-existing policy of discounting environmental bene fits at a 10% discount rate, the value of saving a life would have been reduced to only about $ 22,000.   n16 EPA withstood OMB's pressure and published final regulations that essentially rejected the concept of dis counting. The EPA's regulation was invalidated by the Fifth Circuit, partly for this reason.   n17

A recent article by Lisa Heinzerling shows how much rides on whether the value of human lives is discounted over a latency period.   n18 She shows that many environmental and health-and-safety regulations promulgated since the 1970s have acceptable cost-benefit ratios if the value of lives is not discounted, but fail cost-benefit analysis if those values are discounted.   n19

Discounting issues play an even more critical role in connection with harms to future generations, particularly with respect to the effects of climate change. Because of the long lag until many of the harmful effects of excessive anthropogenic gases in the atmosphere are felt, how much our society is willing to spend on measures to prevent climate change may well depend on how the question of discounting is resolved.   n20

 [*947]  Opponents of discounting adduce vivid statistics to illustrate what is at stake. For example, Derek Parfit notes: "At a discount rate of five per cent, one death next year counts for more than a billion deaths in 500 years."   n21 Even economists who do not oppose discounting acknowledge its striking effects: "When time horizons are very long, all benefits are discounted to zero using any positive discount rate, so that a death pre vented in the distant future is worth nothing at the present time."   n22

This Article seeks to shed light on what has become a shrill and un productive debate. The polar positions on both the latency and future generations issues are analytically unsound and overlook important components of both problems. Moreover, the latent harm and future generation situations are analytically distinct: what one concludes with respect to discounting in one context says little about the appropriate treatment of discounting in the other.

Part I addresses the problem of latent harms. Because there are essentially no empirical studies of the value of lives threatened by latent harms, regulatory analyses must adapt valuations derived from threats of instantaneous death in workplace settings. This Article argues that it is necessary to discount this value, to reflect that the years lost occur later in a person's lifetime. It also argues, however, that such discounting must be accompanied by countervailing upward adjustments, to account for the involuntary nature of exposure to environmental carcinogens, the dread such exposure causes, and the higher income levels of the victims. By not performing these adjustments, OMB may be undervaluing lives by as much as a factor of six, or even more for particularly long latency periods. Correcting this undervaluation, as this Article urges, could have an important impact on the regulatory process by allowing more stringent regulations to satisfy the requirements of cost-benefit analysis.

Part II deals with harms to future generations. It shows that the use of discounting in that case is ethically unjustified. As a result, it argues  [*948]  that discounting approaches should not replace the principle of sustainable development, which is used in the major international environmental law agreements to measure our obligations to future generations. The discussion shows, however, that the principle of sustainable development is also problematic, and sets forth the principal elements of an attractive theory of intergenerational obligations. The practical implications can be enormous: the rejection of discounting may lead to a far more stringent response to environmental problems, such as climate change, that have long time horizons.

The Article underscores the extent to which discounting raises analytically distinct issues in the cases of latent harms and harms to future generations, even though these two scenarios have generally been treated as manifestations of the same problem.   n23 In the case of latent harms, one needs to make intra-personal, intertemporal comparisons of utility, whereas in the case of harms to future generations one needs to define a metric against which to compare the utilities of individuals living in different generations. The case of latent harms gives rise to a problem that is primarily technocratic: determining how an individual trades off the utility derived from consuming resources at different times in her life. In contrast, the case of harms to future generations raise a difficult ethical problem. It is therefore not surprising that the appropriateness of discounting would be resolved differently in the two contexts.

The Article does not address the role that cost-benefit analysis should play in environmental regulation - a subject that has spawned a large academic literature.   n24 Rather, its goal is more targeted. It assumes, consistent with current practice,   n25 that an important set of environmental and health-and-safety regulations will be evaluated under principles of cost-benefit analysis, and that human lives will be valued as part of this analysis. Given these practices, it seeks to determine the best way to ac count for the fact that certain losses do not occur contemporaneously with the exposure to a contaminant.

A central goal of this Article is to move the regulatory process to wards a more thoughtful valuation of human lives threatened by environ mental carcinogens, and away from OMB's deeply flawed technique of taking valuations from the workplace setting and reducing them by an inflated discount rate.   n26 The Article also seeks to move the discussion of how to treat future generations beyond a focus on discounting, which is unlikely to provide an ethically defensible account of our obligations to future generations.

 [*949] 

I. Latent Harms

 

 The discussion begins in Section A by reviewing the central role that the debate over discounting played in the Corrosion Proof Fittings case and the extent to which, despite the court's resolution in that case, the issue remains unsettled in the public policy arena. Section B explains that the valuations of human life in the economics literature have been con ducted almost exclusively in the context of industrial accidents, where workers face a probability of instantaneous death. In contrast, as a result of understandable methodological complications, there have been essentially no valuations of risks to life with a long latency period, such as those posed by environmental carcinogens. Thus, it is necessary to construct a second-best valuation of a life threatened by a contaminant with a latency period, using as a starting point the valuations from the existing empiri cal studies on instantaneous deaths.

Section C begins the task of constructing a second-best valuation, relying on temporal models that describe the value of life by reference to a stream of utilities that individuals receive if they are alive in particular time periods. When an individual faces a threat to life that manifests it self only after a latency period, she loses fewer life-years than when the threat is instantaneous. Moreover, on average, the loss of life-years occurs further into the future. Downward adjustments to account for these two factors are therefore appropriate.

Section D examines the plausibility of the assumptions underlying the temporal models explored in Section C. It also shows that the dis counting of future utilities is conceptually different from the discounting of money flows.

Section E turns its attention to three important upward adjustments that need to be made when extrapolating from the case of instantaneous deaths to that of carcinogenic harms. These adjustments are necessary as a result of the relationship between an individual's income and the value that she places on life, the involuntary nature of exposure to environmental carcinogens, and the dread people suffer from carcinogenic risk.

Section F focuses on the choice of an appropriate discount rate. It shows that the emerging consensus in the economics literature calls for the use of a rate of 3% or less and takes issue with OMB's policy of pre scribing a 7% rate.

Section G estimates the undervaluation of life that results from OMB's approach of taking valuations from the workplace setting and, without further adjustment, mechanically reducing them by an inflated discount rate. Over a twenty year latency period, the OMB approach can lead to an underestimation by a factor of about six, with a factor of about two being attributable to the choice of discount rate.

Section H argues that discounting the value of life in the context of latent harms does not pose significant moral or ethical dilemmas that are distinct from those raised by cost-benefit analysis in general and the valuation of human life in particular. It is simply one defensible adjustment  [*950]  in the process of constructing a second-best valuation, using workplace valuations as a starting point. Discounting, however, cannot be the only such adjustment.

Before proceeding further, it is useful to underscore that Part I focuses on harms that an individual suffers as a result of an earlier exposure to an environmental contaminant.   n27 The term "latent" could be used to describe other phenomena as well: for example one might think that an environmental exposure producing a harm to future generations gives rise to a latent harm as well. As used throughout this Article, however, the term "latent" is used to describe only situations in which the exposure and the harm accrue to the same individual.

A. The Debate Over Discounting

 

 The appropriateness of discounting the value of human lives first received sustained attention in the regulatory proceeding that led to EPA's partial ban on the manufacture, importation, and processing of asbestos under the Toxic Substances Control Act (TSCA), and the chal lenge to this regulation in Corrosion Proof Fittings v. EPA.   n28 The question was highly controversial even before EPA's publication of the notice of proposed rulemaking in 1986.   n29 As required by Executive Order 12,291 (the Reagan Administration's predecessor of Executive Order 12,866),   n30 EPA submitted the draft rule to OMB for review before its publication in the Federal Register. In a March 1985 letter to A. James Barnes, EPA's acting Deputy Administrator, OMB raised questions about whether the benefits of the rule exceeded its costs.   n31 In performing a cost-benefit analy sis, OMB used a value per cancer case avoided of $ 1 million and discounted this amount at a rate of 4% for the length of the latency period.   n32 (At the time, an OMB guidance document provided for discounting of costs and benefits at a rate of 10%,   n33 but OMB instead used the rate contained in EPA's guidance document on cost-benefit analysis.)   n34

 [*951]  The following month, the propriety of discounting the value of human lives became an issue in connection with Barnes's Senate confirmation hearings:

 

 I have a great deal of ethical difficulty with a concept of applying a discount factor to human life. The lives of my three children are worth every bit as much to me 10 years from now as they are now. I personally reject that notion. I have talked to [EPA Administrator] Lee Thomas about it; I know that it is not one that finds favor with him.   n35

 

 In October 1985, a subcommittee of the U.S. House of Representatives chastised OMB for its insistence on discounting the value of human lives.   n36 It noted that discounting at OMB's 10% discount rate over a forty year latency period would reduce the $ 1 million value per life saved to just over $ 22,000.   n37 Thus, on cost-benefit terms, one could not justify a current expenditure of over $ 22,000 to save a life forty years in the future. Even at a 4% discount rate, the $ 1 million value of life would be reduced to about $ 208,000.   n38

The subcommittee referred to the testimony of Don Clay, Director of EPA's Office of Toxic Substances, that EPA "never had used discounting over the latency period of a chronic hazard," and that, by reducing the value of benefits to such an extent, OMB's approach would prevent EPA from regulating any carcinogen with a long latency period.   n39 The subcommittee further reported that Clay "personally opposed the discounting of lives in the asbestos case on ethical grounds."   n40 It concluded that OMB's position with respect to the discounting of the value of life was "simply an outrage" and urged EPA to "reject the use of discounting over the latency period of diseases caused by chronic hazards."   n41

EPA published the proposed rule on the asbestos ban in January 1986.   n42 The proposal did not quantify the value of life or undertake any  [*952]  discounting of this value over the length of the latency period.   n43 EPA took a different approach, however, when it promulgated the final rule in July 1989.   n44 It assigned a value to human lives, but discounted it at a rate of 3% from the time of the promulgation of the regulation until the time of the exposure to the carcinogen.   n45

The use of asbestos products does not necessarily result in immediate exposure; instead, exposure occurs when the product containing the asbestos begins to disintegrate. For example, some exposures occur when asbestos fibers are released into the air from the weathering of air conditioning products.   n46 Exposure is the first step of a process that might later lead to the incidence of cancer and subsequently to a death from cancer. EPA did not discount the value of human life from the time of exposure until the carcinogenic death, as OMB had urged, or even until the first manifestation of cancer.

In its response to comments accompanying the final rule, EPA at tempted to defend this decision. EPA noted that comments had been written on both sides of the discounting issue:

 

 Some commenters argued that EPA, in the proposal, improperly failed to discount benefits to be derived from the rule, and in support of documents for a final rule, only discounted benefits until the time of the exposure that results in the cancer rather than until the occurrence of the disease. Other commenters argued that EPA should not discount benefits, stating that discounting the benefit of saving human life is inappropriate methodology for this rulemaking.   n47

 

 EPA's response revealed a degree of ambiguity on this question and pro vided at best a lukewarm defense of its course of action. It stated:

 

 Arguments can be made that estimating benefits without dis counting is preferable in cases like this one where the primary benefits derived is [sic] the avoidance of human cancer cases. However, arguments also can be articulated supporting the discounting of benefits.   n48

 

 EPA was more categorical in defending its view that if discounting was appropriate at all, it was appropriate only until the time of exposure:

 

 Since the benefit of a regulation to control a hazardous sub stance occurs at the time of the reduced exposure, EPA has concluded that the appropriate period over which to discount is un til the time of exposure reduction. This approach was used in this case after extensive review of applicable literature and an  [*953]  examination of the inherent biases and features of other approaches.   n49

 

 This position has an important corollary for environmental problems in which the regulation leads to an immediate decrease in the exposure of individuals as is the case, for example, with airborne air pollutants. For such pollutants, no discounting of the benefits of the regulation would be performed under EPA's approach, except perhaps for discounting from the time of the preparation of the cost-benefit analysis to the implementa tion of the regulation.

Though EPA's explanation is not a model of clarity, one can surmise that its approach was not to discount for the period between the exposure and the death, when the harm was latent. Instead, the discounting that was performed affected only the period before the harm became latent.

In October 1991, the Fifth Circuit vacated the regulation and remanded in Corrosion Proof Fittings v. EPA,   n50 in part because of EPA's treatment of the discounting issue. The Fifth Circuit took the position that discounting was necessary in order to provide for a fair comparison of costs and benefits accruing at different times:

 

 Although various commentators dispute whether it ever is appropriate to discount benefits when they are measured in human lives, we note that it would skew the results to discount only costs without according similar treatment to the benefits side of the equation.... Because the EPA must discount costs to perform its evaluations properly, the EPA also should discount benefits to preserve an apples-to-apples comparison, even if this entails discounting benefits of a non-monetary nature.   n51

 

 The Fifth Circuit went on to hold that EPA had used an improper period for discounting, and that the value of human life should have been discounted to the time of injury.   n52 It noted:

 [*954] 

 

 The EPA's approach implicitly assumes that the day on which the risk of injury occurs is the same day the injury actually occurs. Such an approach might be appropriate when the expo sure and injury are one and the same, such as when a person is exposed to an immediately fatal poison, but is inappropriate for discounting toxins in which exposure often is followed by a substantial lag time before manifestation of injuries.   n53

 

 The court did not specify, however, whether it considered the injury to be the first manifestation of cancer or the death from cancer. The detection of carcinogenic cells is a serious injury, but if death does not follow it is not clear why it would be appropriate to attach to this injury the full value of life, rather than the value of the resulting morbidity.   n54

Finally, the Fifth Circuit upheld EPA's choice of a 3% discount rate. It implicitly assumed that the correct discount rate was the real rate of interest (the nominal rate of interest minus the rate of inflation) and stated that, historically, this rate has fluctuated between 2% and 4%.   n55

Despite the court's holding, the question of discounting the value of human life has continued to be controversial. For example, the Senate Report accompanying the Comprehensive Regulatory Reform Act of 1995,   n56 which would require the use of cost-benefit analysis in regulatory proceedings,   n57 contains a statement by Senator Leahy railing against such discounting:

 

 Cost/benefit analysis assumes that benefits that occur in the future have very little value. After determining the value of human life, cost/benefit analysis applies a "discount rate" to benefits that will occur in the future. Benefits of the lives saved in the future by a regulation are reduced by 6-7 percent per year.... This business evaluation tool does not make sense when applied to the protection of human life.   n58

 

  [*955]  The regulatory debate over the appropriateness of discounting of human lives, stated in conclusory terms and virtually devoid of any sus tained analysis, fails to shed light on the important issues underlying this question.   n59 After providing a brief overview of the economic approach to valuing human life, the remainder of Part I seeks to fill this void.

B. Valuations of Human Life

 

 Since the 1970s, willingness-to-pay studies have become the standard economic technique for placing a value on human life.   n60 By far the most common method for performing such valuations focuses on the choices that workers make in accepting risky jobs.   n61 The approach begins by de fining sets of jobs that require comparable skills and offer comparable non-monetary amenities, except that one exposes the worker to a higher risk than the other.   n62 Presumably, a rational worker would not accept the riskier job unless she obtained sufficient compensation for the additional risk. The resulting wage differential is the compensation that the worker obtains for the additional probability of death that she faces as a result of having taken the riskier job.   n63 An extrapolation, consisting of dividing the wage differential by the additional probability of death, is then per formed to determine the value of life.   n64

 [*956]  Willingness-to-pay studies of the value of human life have been con ducted almost exclusively in the context of industrial accidents, where the worker faces a risk of being either fatally injured by a piece of machinery and dying instantaneously, or surviving unscathed.   n65 In any time period, there is a probability that a fatal accident will occur. This probability is ascertained from industrial safety statistics.   n66

One could use the same approach to determine the willingness-to- pay to be free from risks with long latency periods.   n67 As long as workers understood the additional probability of, say, dying of cancer from a riskier job, and knew the length of latency period, they could figure out how much additional compensation to demand in order to accept the job with the higher risk. From this wage differential, one would extrapolate to determine the value of the life. The fact that the harm would accrue only in the future would be reflected in the wage differential. For example, other things being equal, an individual with a comparatively high discount rate would demand a comparatively low wage differential. We would then have measured exactly what we wanted to see, and there would be no need to perform any discounting.

It is likely that such studies have not been conducted for three principal reasons. First, the industrial statistics on deaths resulting from la tent harms are not as extensive as those for instantaneous accidents. The federal government became extensively involved in the regulation of  [*957]  workplace and environmental safety only in the 1970s (and prior state efforts in these areas were relatively modest).   n68 For example, if the federal government began to compile statistics on the risk of various work place settings in the mid-1970s, it would have immediately had a data set on instantaneous accidents. In contrast, for carcinogenic risks with a twenty-year latency period, comparable statistics on such risks would not be available until the mid-1990s, unless retrospective studies could be per formed. Moreover, while accidents on the job are relatively easy to track, statistics on mortalities associated with latent harms require much more difficult tracking of the health status of individuals after they leave their jobs. Further, while the cause of on-the-job accidents typically is relatively easy to identify, the causal link between occupational exposure and fu ture harms from carcinogens can be difficult to establish.

Second, in order for willingness-to-pay studies to yield meaningful results, individuals must be able to properly understand the nature of the risk; otherwise, they cannot determine what sum of money properly compensates them for the risk. Some commentators doubt that our cognitive capacities are sufficiently developed to perform such valuations in the case of future harms.   n69

Third, this problem is compounded by the fact that exposure to carcinogens may have a differential impact depending on an individual's characteristics, including, for example, whether she smokes. In order to decide how to respond to a wage premium, individuals would need to understand not only the "pure" carcinogenic risk of the job, but also the magnitude of any synergistic interactions that might result from such characteristics.

In summary, the task of directly performing a willingness-to-pay study of the value of life in the case of latent harms is fraught with difficulties, perhaps insurmountable ones. Instead, to obtain such a valuation, resort to a second-best approach is necessary.

C. Discounting as a Second-Best Approach

 

 As a result of the difficulty of obtaining a direct willingness-to-pay measure of the value of a life threatened by a latent carcinogenic harm, economists have devoted considerable attention to defining a relation ship between the value of a life lost today and the value of a life lost years from now. Such temporal models, also known as life-cycle models, study the distribution of an individual's utility throughout her life.   n70

 [*958]  The discussion that follows focuses, for illustrative purposes, on three different valuations: first, the life of a 40-year old that is lost today, for example, from an industrial accident; second, the life of a 60-year old, also lost today; and third, the life of an individual who is currently 40 years old but dies in twenty years as a result of exposure today to a carcinogen with a twenty-year latency period.   n71 For this discussion, V[in'j,k'] denotes the value attached to the life of an individual exposed to a harm at age j who dies at age k. Thus, the values of the three lives described above can be expressed as V[in'40,40'], V[in'60,60'], and V[in'40,60'], respectively. To keep the discussion simple, it assumes that these individuals, if not exposed to the industrial or carcinogenic risk, would die of natural causes at age 80.   n72

The three valuations differ in two important ways.   n73 First, the forty- year old dying immediately loses 40 years of life whereas the sixty-year old dying immediately and the forty-year old dying in twenty years lose only twenty years of life.   n74 Second, the individual exposed to the carcinogen does not lose these twenty years of life immediately, but twenty years later.   n75 Let u<l> denote the utility that an individual derives in year l from living that year. So, for example, for the forty-year old exposed today to the latent harm, u<60> is the utility that the individual would derive in twenty years from living in the year following her sixtieth birthday. In contrast, for the sixty-year old killed today in an industrial accident, u<60> is the utility that the individual would have derived this year if the accident had not occurred.

If these utilities were simply monetary payments as opposed to the well-being that comes from living, they could easily be compared with one another by discounting the future stream of benefits by a means of a discount rate. Discounting reflects the fact that it is more desirable to get a payment sooner rather than later. It is important to stress that this preference is not a function of the existence of inflation. In comparing monetary flows occurring at different times, the effects of inflation can be adjusted by converting all amounts to constant dollars. But even in an inflation-free world, it is best to get a given amount of money as soon as possible. Having the money sooner gives one the option of either spend  [*959]  ing it immediately or saving it for later, whereas getting it later (absent borrowing) rules out immediate spending. The rate used to discount amounts in constant dollars is typically known as a "real" discount rate.   n76

Given a discount rate of r, the present value of a payment P that is paid t years from now is [1/(1 + r)[su't']]P.   n77 I am not suggesting at this point that discounting to present value the utility that an individual derives from living for a year is equivalent to discounting a monetary payment, and will return to this issue later.   n78 Instead, I am showing the relation ship among the values of the three different lives if such discounting were appropriate.

Then,

V[in'40,40'] = u<40> + [1/(1 + r)]u<41> + ... + [1/(1 + r)[su'38']]u<78> + [1/(1 + r)[su'39']]u<79>

The loss for the forty-year old killed by the industrial accident is the utility of living in the year following the individual's fortieth birthday, plus the utility of living one year later discounted for one year, plus the utilities of living in all subsequent years until age 80 (when the individual would have died anyway), with each utility discounted for the number of years elapsed since the present.

In turn,

V[in'60,60'] = u<60> + [1/(1 + r)]u<61> + ... + [1/(1 + r)[su'18']]u<78> + [1/(1 + r)[su'19']]u<79>

Here, the loss takes the same form, except that the first year of loss of utility is the year following the individual's sixtieth birthday.

Finally,

V[in'40,60'] = [1/(1+r)[su'20']]u<60> + [1/(1 + r)[su'21']]u<61> + ... + [1/(1 + r)[su'38']]u<78> +

[1/(1 + r)[su'39']]u<79>

Only years following the individual's sixtieth birthday are lost, and these losses are discounted by the number of years from the present.

The relationship between V[in'60,60'] and V[in'40,60'] should now become appar ent. The latter value is simply the former discounted by twenty years.   n79 In other words, both individuals lose the same years of their lives - those following their sixtieth birthdays - but the latter individual loses them twenty years later than the former. Thus,

V[in'40,60'] = [1/(1+r)[su'20']]V[in'60,60']

Under this approach, the value that should be attached to the life of a forty-year old who is exposed to a carcinogen with a twenty year latency period and who dies at age 60 is equal to the value of the life of a sixty- year old who dies instantaneously in an industrial accident, with the latter  [*960]  value discounted for the twenty years that elapse before the carcinogenic victim dies.

So far, in fact, the discussion suggests that the OMB approach actually overestimates the value of the loss resulting from exposure to latent risks. The OMB procedure takes V[in'40,40'] and discounts it back to present value to account for the latency period.   n80 In fact, the correct approach would be to discount V[in'60,60'] instead,   n81 which is lower than V[in'40,40'] because of the twenty fewer years of life loss.   n82 As explained later, however, this overvaluation is outweighed by the substantial undervaluation that results from other elements of OMB's approach.   n83

D. Plausibility of the Model

 

 The model presented in the previous section relies on two important assumptions. First, it assumes that an individual's utility function can be expressed as a sum of utilities over the various periods comprising one's lifetime. Thus, one's enjoyment of life in one period is not affected by the resources available for consumption in prior periods,   n84 but only by the resources in that period.   n85 Under the model, an individual's utility in one period is not affected by the resources available for consumption in prior periods.   n86 So, for example, whether an individual was able to afford a quality education in a prior period does not affect the utility that she derives from a given level of consumption in subsequent periods. This assumption is clearly debatable. Indeed, John Broome, in a related con text, terms the assumption "dubious,"   n87 though he acknowledges that it is commonly made in economic analysis.   n88

 [*961]  Moreover, an individual facing death from cancer may focus on the fact of the death and on its cause, without paying particular attention to the death's timing. One's willingness-to-pay to avoid the risk may then be relatively unaffected by the length of the latency period. A number of studies show that individuals of different ages exhibit different willing nesses-to-pay to avoid instantaneous deaths, suggesting, consistent with the model, that their valuations are indeed affected by the number of life- years that they would lose.   n89 It is possible, however, that such behavior would not extend to carcinogenic risks as a result of the dread associated with such deaths.   n90 As a result of the paucity of studies of the willingness- to-pay to avoid carcinogenic risks,   n91 it is not possible to make empirically grounded claims concerning this hypothesis.

Second, the model uses a constant discount rate.   n92 So, for example, the same rate would be used to discount the utility of living twenty years in the future as would be used to discount the utility of living next year. As Donald Shepard and Richard Zeckhauser put it, the model assumes that "an individual's utility over lifespans of different length can be represented as a weighted sum of period utilities, the weights declining geo metrically with time."   n93 Shepard and Zeckhauser label this assumption "heroic."   n94

If, for example, I did not currently value at all the utility of living beyond the year 2010, I would be applying an infinite discount rate to the utilities that I would derive if in fact I were alive beyond that year. The present discounted value of those utilities would be zero. There is no mechanism by which I could transfer any life-years beyond the year 2010 to someone with a lower discount rate, in return for a higher present utility. In contrast, in the case of financial flows, if I undervalued relative to the market the stream of payments that I would receive on my Treasury bond after the year 2010, I could increase my utility by selling that stream of payments at the market price.   n95

 [*962]  There is little attempt in the literature to validate the constant discounting feature of the model through experiment or observation.   n96 One study of the implicit discount rates reflected in individuals' contingent valuation of the disutilities of various illnesses led the authors to question whether the conventional discounting model properly describes individual preferences.   n97

These problems with the assumptions underlying the temporal models for the valuation of lives threatened by environmental carcinogens should not lead to the conclusion that the models are inappropriate. At present, such models are the state of the art in economic analysis. It is therefore proper to continue to use them, absent a further refinement or an empirical falsification. But as the regulatory process seeks to construct appropriate second-best valuations for lives threatened by environmental carcinogens, it must pay further attention to the plausibility of the assumptions underlying temporal models.

E. Necessary Adjustments

 

 It is time now to scrutinize with more care some of the assumptions made implicitly in the model described in Part I.C. Such scrutiny reveals, for several reasons, that one cannot simply take an estimate of the value of life from an industrial accident (whether V[in'40,40'] or V[in'60,60']),   n98 discount it, and obtain a plausible estimate of the value of life from exposure to an environmental carcinogen with a latency period.   n99 Many adjustments need to be made for the estimate to be at all meaningful. These adjustments all lead to assigning a higher value to the life lost.

This section examines the principal adjustments that need to be per formed. It focuses primarily on differences between the valuations for instantaneous and latent harms that have been the subject of empirical examination.

1. Impact of Income on the Valuations of Life. - In the temporal model presented in Part I.C, the utility that an individual derives in a particular year is a function of the level of resources available for consumption that year. Economists have estimated that the elasticity of the value of life with respect to earnings (the percentage change in the value of life for a one percent change in earnings) is approximately one. Thus, for example, a  [*963]  10% increase in income would lead to a 10% increase in the value of life.   n100 The impact of income on the valuation of life calls into question several of the implicit assumptions made in Part I.C.

a. Increases in Income Over Time. - That model assumes implicitly that the valuation of a particular year of life, say the year following one's sixty-fifth birthday, is independent of the age of the individual making the valuation. Thus, for example, u<65>, the utility of living in the year fol lowing one's sixty-fifth birthday, is the same for both a forty-year old and a sixty-year old. The only difference related to the valuation is that the forty-year old discounts this utility for the twenty-five years that it will take until this utility is realized, whereas the sixty-year old discounts the utility for only five years.

A correction needs to be made, however, if income adjusted for infla tion rises over time. In comparing V[in'40,60'] with V[in'60,60'], one must account for the fact that by the time the forty-year old is sixty, her income, in real terms, will be higher than the sixty-year old's income is today.

If income rises in real terms over time, the relationship between V[in'40,60'] and V[in'60,60'] becomes different than that posited in Part I.C.   n101 Let g be the yearly increase in the individual's real income. Then,

V[in'40,60'] = [(1 + g)/(1 + r)][su'20']V[in'60,60']

Thus, V[in'60,60'] now needs to be subjected to two adjustments.   n102 First, it is increased by a factor of (1 + g)[su'20'] to account for the fact that the years of lost life will occur twenty years later for the forty-year old, and that for each of the years of life lost, the utility lost twenty years from now to the individual who is currently forty years old will be (1 + g)[su'20'] greater than for the individual who is currently sixty years old. Second, it is decreased by a factor of [1/(1+r)][su'20'] to discount to present value the utilities that the current forty-year old would enjoy twenty years later. To a first approximation,   n103 the relationship between V[in'40,60'] and V[in'60,60'] simplifies as follows:

V[in'40,60'] = [1/(1 + r - g)][su'20']V[in'60,60']

For example, if the real discount rate is 3% but income is rising at a yearly rate of 1% in real terms, then the effective rate at which V[in'60,60'] would be discounted to arrive at V[in'40,60'] would be 2%. Moreover, if r and g were equal, then V[in'40,60'] and V[in'60,60'] would be equal as well.   n104 The increase in the  [*964]  valuation of V[in'40,60'] to account for rising real incomes would exactly counteract the decrease resulting from the time lag in the enjoyment of utilities.

Table I presents the changes between 1982 and 1996 in mean and median incomes for workers fifteen years and over. The figures are presented in constant 1996 dollars.   n105

[SEE TABLE IN ORIGINAL]

 

 

 The table reveals that median and mean income grew at compound rates of 0.95% and 1.01% per year, respectively.   n106

b. Age-Dependent Nature of the Valuation. - A different issue is raised by life-cycle changes in levels of income. For example, Donald Shepard and Richard Zeckhauser analyze the valuations of a typical individual who enters the work force at age twenty, sees steadily rising income up to age 50, then experiences a small decrease in income until age 65, and loses all income as a result of retirement at age 65.   n107 The economics literature assumes that people value their lives as a function of their current income (and resulting consumption), not on the basis of projections of  [*965]  future income.   n108 Richard Zeckhauser has labeled this phenomenon as "temporal myopia."   n109

Shifts in an individual's income across time would not make a differ ence to the valuations of life if borrowing were available to equalize the amounts available for consumption. Typically, however, there are serious roadblocks to borrowing based on the expectation of higher incomes in the future.   n110 And, to the extent that such borrowing is possible, for ex ample through credit cards, the interest rates are prohibitively high.

Shepard and Zeckhauser calculate the impact of age on a person's valuation of life for two different scenarios, to which they attach "Robin son Crusoe" and "Perfect Markets" labels. In both cases, the individual supports her consumption from her own income and wealth, and has no heirs or dependents. In the Perfect Markets scenario, the individual can borrow in the capital markets, in order to support a higher level of con sumption earlier in life, and can purchase annuities to insure against variability in her lifespan. In contrast, in the Robinson Crusoe scenario, ac cess to these two markets is unavailable.   n111

The authors show that in the Robinson Crusoe model an individual's valuation of life reaches its peak at age forty. A forty-year old values her life 2.5 times as highly as a 20 year old (that is, returning to the notation previously used, V[in'40,40'] = 2.5V[in'20,20']). At first glance, this result might appear counterintuitive. After all, the twenty-year old loses twenty more years of life than the forty-year old. The reason that the forty-year old's valuation is higher, however, is that her income is more than three times higher, and this effect more than counteracts the shorter remaining life.   n112

In turn, in the Robinson Crusoe world, the forty-year old values her life almost twice as highly as a sixty-year old (V[in'40,40'] = 1.98V[in'60,60']).   n113 Two different effects are at play here. Most obviously, the sixty-year old has fewer years to live. But another factor is depressing the sixty-year old's valuation of her life. Beyond age forty, income continues to rise until age fifty, but consumption begins to fall. The reason is that at age forty, the individual begins to save for retirement and therefore has fewer resources available for current consumption. Indeed, even though income at age  [*966]  sixty is comparable to income at age forty, consumption is about 25% lower.   n114

The situation is more straightforward under the Perfect Markets scenario. There, the valuation of life is highest at age 20, and then falls continuously through the life cycle. In this model, the forty-year old's valuation is about two-thirds higher than that of the sixty-year old.   n115 Here, the difference between V[in'40,40'] and V[in'60,60'] is attributable exclusively to the different number of years of remaining life.

To the extent that the assumptions underlying the Robinson Crusoe model are at least partly realistic,   n116 one needs to worry about the procedure described in Part I.C in which the sixty-year old's willingness-to-pay to avoid an immediate death, V[in'60,60'], was used as a proxy (and then dis counted) for a forty-year old's willingness to pay to avoid a death twenty years later, V[in'40,60']. Given the levels of income and savings analyzed by Shepard and Zeckhauser, using V[in'60,60'] as a proxy for V[in'40,60'], as was done in Section I.B, will result in an undervaluation of the willingness to pay to avoid death of about 25% (as a result of the lower level of consumption at age 60).   n117

This undervaluation, however, may have decreased over time. Shepard and Zeckhauser relied on data from the late 1970s.   n118 Certain legal changes since that decade, particularly the end of mandatory retirement and the strengthening of protections against age discrimination, are likely to have affected the impact of age on income. In particular, it is possible that the peak income is received later in life and that the assumption that individuals receive no income after the age of sixty-five is now unrealistic. These changes would result in increasing the ratio of the sixty-year old's consumption relative to that of the forty-year old and thereby diminishing the difference in the valuations of V[in'40,40'] and V[in'60,60'] in a Robinson Crusoe economy.

In summary, the discussion in this subsection is presented only to illustrate the underlying methodological issues that must be resolved to obtain a plausible estimate of the value of life. More work needs to be done to determine the plausibility of the Robinson Crusoe model and the effects of changes in workplace patterns and legal protections since the 1970s.

c. Distribution of Income Across Occupations. - Individuals who take risky jobs generally have lower-than-average income.   n119 Thus, there is a  [*967]  problem in extrapolating from the willingness-to-pay studies conducted in high-risk occupations to the broader population affected by environ mental carcinogens.

One threshold issue concerns the definition of the population af fected by the different environmental programs. In principle, for every environmental regulation, one could attempt to determine the identity, age profiles, and economic characteristics of the affected population. One could then construct program-specific valuations of life that took into account the distribution of ages and incomes of the affected population, as well as of the latency period of the carcinogen subject to the regulation.

There are good reasons why one might not want to undertake such an evaluation. First, the informational requirements are likely to be daunting. For every environmental program, in addition to estimating the number of affected individuals, one would need to determine their demographic and economic characteristics.   n120

Second, an effect of particularized valuations based on levels of in come would be to justify, on cost-benefit grounds, more stringent regulation when the affected population is wealthier. Such a policy would be inconsistent with the central tenet of the increasingly influential environ mental justice movement, which calls for environmental regulation to be no less (if not more) responsive to the needs of communities that are disproportionately poor, or disproportionately populated by people of color than to the needs of wealthy, white communities.   n121

As a result, it is reasonable for EPA to use uniform valuations of life across environmental programs. These valuations would be based on representative characteristics of the population of the United States.   n122 Thus, to the extent that the subjects of the empirical studies involving  [*968]  industrial accidents have relatively low incomes, an upward adjustment in their valuations of life must be performed before translating these figures to the environmental context.

The U.S. Census provides median and mean earnings for all workers and for various occupational categories.   n123 The category including operators, fabricators, and laborers might be a good proxy for workers in risky occupations who are the subjects of empirical studies concerning the value of life. In 1996, the median and mean earnings of all workers 15 years of age and over were $ 20,716 and $ 27,366, respectively.   n124 The corresponding figures for operators, fabricators, and laborers were $ 16,883 and $ 19,981.   n125 Thus, the overall median earning is 22.7% higher than the median for workers in risky occupations, and the overall mean is 36.8% higher.

2. Involuntary Nature of the Harm.

a. Comparative Valuations of Voluntary and Involuntary Risks. - There is an extensive literature suggesting that individuals assign greater value to avoiding risks that are thrust upon them involuntarily than risks that they incur voluntarily.   n126 As Richard Zeckhauser points out, "this ten dency would introduce a downward bias in the implicit life valuations of those who voluntarily assume risks."   n127

The risk assumed by individuals who take risky jobs and subject themselves to a non-trivial possibility of industrial accidents is generally thought of as a risk assumed voluntarily.   n128 In contrast, the risk of expo sure to environmental carcinogens, for example, as a result of toxic air pollution, is generally thought of as involuntary.   n129

 [*969]  As a result, there will be a systematic undervaluation if one takes the willingness-to-pay to avoid voluntary harms and imports that figure into the context of environmental regulation. Determining the extent of the undervaluation, however, is complicated.

The economics profession strongly favors "revealed preference" valuations, under which the value assigned to a good can be observed through a market transaction. Willingness-to-pay studies of wage differ entials needed to compensate individuals for accepting a risk of death are a prominent example of a revealed preference technique.   n130 Revealed preference approaches are poorly suited for determining the valuation of involuntary harms because they are based on the existence of market transactions, and such transactions are generally seen as voluntary.   n131

Thus, in order to estimate how the valuations of involuntary and voluntary risks differ, one needs to resort to a different approach. In recent years, a great deal of attention has been devoted to the implicit valuations of human life derived from dividing the total cost of an environmental program by the number of lives saved. The result, for environmental pro grams that do not have significant other benefits, is the implicit value that the regulatory program has assigned to each life. The range of implicit valuations for regulatory programs is enormous, from around $ 100,000 per life to a number in the billions of dollars.   n132 To reach any worthwhile conclusions from these implicit valuations, one would need to make the  [*970]  heroic assumption that social expenditures in fact are reflective of public preferences.

Thus, a more promising alternative is to directly question individuals about the relative value that they attach to avoiding voluntary and involuntary harms.   n133 In the most comprehensive study of this type, Maureen Cropper and Uma Subramanian conducted a nationwide telephone sur vey of 1,000 households, asking interviewees to compare an environmental program and a public health program designed to address a particular risk, such as respiratory illness or cancer.   n134 The interviewees were first told that the two programs would cost the same amount of money and save the same number of lives, and were asked to determine which pro gram was best for society.   n135 Then, they were told that the program that they had found less attractive would in fact save x times more lives than its counterpart. The authors computed the number of lives saved by each program that made the median respondent indifferent between the two programs.

The interviewees were also told to describe some qualitative characteristics for the risk addressed by each of the programs, and, for each characteristic, to place the risk on a ten-point scale. One of these characteristics was the ease with which the risk could be avoided,   n136 which is a measure of the risk's voluntariness.   n137 In each case, the public health risk was deemed to be more voluntary than the environmental risk.   n138

For the purposes of this Article, the most relevant pair examined by the researchers was radon control in homes and a pesticide ban on fruit. Radon control, like workplace hazards, is a paradigmatic voluntary risk: an individual can avoid the risk by making a monetary sacrifice. In con trast, pesticide control, like other environmental risks, generally cannot  [*971]  be addressed effectively absent some level of social coordination. For this reason, the risk should be regarded as involuntary.   n139

The respondents were asked to assess, on a ten point scale, the ease with which the respective risks could be avoided. The mean ratio of the ease with which the radon risk could be avoided to the ease with which the pesticide risk could be avoided was 1.31.   n140 When respondents were told that the two programs would save the same number of lives (and cost the same), 72% chose the pesticide ban and only 28% opted for the radon control.   n141 The median respondent was indifferent between saving 100 lives by means of the pesticide ban and 213 lives through radon con trol.   n142 Thus, the median respondent implicitly found the life saved imperiled by the involuntary risk to be twice as "valuable."

More generally, the authors found, across the six pairs of risks that they studied, a consistent, statistically significant preference for addressing the less voluntary risk.   n143 Moreover, a significant minority of respondents - between 20 and 30% - always preferred addressing the involuntary risk, regardless of how many more lives would be saved by transferring the resources to addressing the voluntary risk.   n144

b. Unrepresentativeness of the Population Exposed to Workplace Risks. - Another type of adjustment needs to be made when using valuations of life in workplace settings as a second-best measure of the appropriate value of life for environmental programs. Individuals who take relatively risky jobs have a comparatively low willingness-to-pay to avoid the risk.   n145 Indeed, individuals with higher valuations would demand greater wage differentials to take a riskier job over an otherwise comparable job that was less risky. The employers, however, would not need to pay this higher premium if they could fill their jobs with workers who had lower valuations.

This concept can be illustrated by reference to an auction. The employer with the risky jobs offers a low wage premium and sees how many workers are willing to take the positions. If it does not fill all the vacancies, it offers a somewhat higher premium, and continues this process  [*972]  until it is able to fill all the jobs. Any workers who place a higher value on avoiding the risk end up not getting the job.

As a result, the willingness-to-pay valuations derived from the study of risky jobs are not the valuations of the mean or median member of society. Instead, they are the valuations of a relatively small subgroup with a disproportionate tolerance for risk.

In contrast, environmental risks in general affect a far broader sector of society. Moreover, because they are involuntary, there is no easy mechanism for individuals to self-select for such risks based on their lower- than-average valuations of risk.   n146 Thus, an appropriate correction needs to be made when extrapolating from the workplace to the environmental arena. No empirical literature, however, sheds light on the magnitude of this correction.

3. Dread Nature of the Harm. - There is also an important difference in the nature of deaths resulting from industrial accidents on the one hand and from environmental exposures to carcinogens on the other. The former occur instantaneously and without warning. The latter often occur following a long and agonizing ordeal. As Cass Sunstein pithily notes: "All deaths are bad. But some deaths seem worse than others."   n147

A far greater level of social expenditures is devoted to combating toxic risks like cancer than risks of instantaneous deaths. A recent, admirably comprehensive study by Tammy Tengs and a number of co-authors compares the cost-effectiveness of various risk reduction regulations.   n148 The authors first determine the cost per life saved by dividing the direct costs of the regulation by the number of lives saved. Then, they divide this cost per life saved by "the average number of years of life saved when a premature death is averted" to obtain the cost per life-year saved.   n149

The comparison of costs per life-year saved reveals enormous disparities. The median medical and toxin control measures cost $ 19,000 and $ 2,800,000 per life-year, respectively; the overall median is $ 42,000 per life-year.   n150 The authors also found a wide disparity in occupational in terventions depending on the nature of the death. The median occupational intervention designed to avert a fatal injury costs $ 68,000 per life- year, whereas the median occupational intervention involving the control of toxins costs $ 1,400,000 - more than twenty times as much.   n151

But as in the case of the comparison between voluntary harms and involuntary harms, one cannot draw strong conclusions from these dis  [*973]  parities because public expenditures may well not reflect people's preferences.   n152 Instead, a more direct measure of the difference in valuations is preferable.

A study by George Tolley, Donald Kenkel, and Robert Fabian at tempts to quantify the values attached to the avoidance of unforeseen, instantaneous deaths on the one hand and carcinogenic deaths on the other.   n153 For each of these risks, the authors define a low estimate, a medium estimate, and a high estimate, and present their figures in 1991 dollars. For unforeseen, instantaneous deaths, the respective estimates, derived from a survey of willingness-to-pay studies, are $ 1 million, $ 2 mil lion, and $ 5 million, respectively.   n154

Because, as indicated earlier, there are no willingness-to-pay studies estimating the value of life lost from a disease with a long latency period,   n155 the procedure used by the authors for estimating the value of carcinogenic deaths is more complicated. As their starting point, the au thors use the estimates for instantaneous deaths. Then, for their low estimate, they add a component for the value of the morbidity period pre ceding the death.   n156 This value is derived primarily from contingent valuation rather than revealed preference approaches.   n157

As the authors note, this estimate is conservative for two reasons. First, it understates the value of morbidity preceding mortality because conditions that eventually become fatal are more serious than nonfatal, chronic conditions. Second, it does not account for the dread aspects of carcinogenic deaths.   n158 The authors account for these two components in their medium and high estimates, relying primarily on a survey of how individuals compare deaths from cancer to deaths from other causes,   n159 and on contingent valuations of periods of severe limitations of activity preceding death. The authors' low, medium, and high estimates of the value attached to a life threatened by cancer are $ 1.5 million, $ 4 million, and $ 9.5 million, respectively. Thus, the medium valuation of life in the  [*974]  case of carcinogenic exposure is twice as high as the corresponding valua tion for an unforeseen, instantaneous death.   n160

F. Choice of a Discount Rate

 

 Parts of the preceding discussion have already hinted as to why the choice of the discount rate used in connection with the valuation of lives is more complicated than merely picking the discount rate used for monetary flows.   n161 I can invest $ 100 today at a 3.5% interest rate and have about $ 200 in twenty years. I cannot invest the utility that I derive from living a year at present and obtain, twenty years later, the utility that I would then derive from living two years.   n162 Similarly, I can sell the right to get a payment of $ 200 in twenty years for a present payment of about $ 100. I cannot engage in a comparable transaction with respect to the utility that I would derive from living in twenty years. As W. Kip Viscusi notes, "One cannot trade health ... across time .... If we value our health at forty-five but do not at twenty-five, then we cannot simply shift health status across time in the same way that we would shift monetary resources."   n163

This section undertakes two separate tasks. First, it reviews empirical evidence suggesting that, despite the conceptual difference between the two, there is no statistically significant difference between the discount rate that individuals apply to future health risks and the discount rate that financial markets apply to flows of money. Second, it criticizes OMB's approach with respect to discounting, especially as applied to future health risks, showing that OMB employs a rate that is inappropriately high.

1. Discounting Health Risks v. Discounting Financial Flows. - Thought ful analysts have recognized that the discount rates applied to financial flows cannot be applied mechanically to the discounting of the utility that comes from living in the future.   n164 The most extensive empirical work in this area is that of Michael Moore and W. Kip Viscusi, who seek to deter  [*975]  mine whether the rates of discount for health risks differ from the financial rates of time preference.   n165

In their most recent article on the subject, Moore and Viscusi estimate the implicit discount rate exhibited by workers facing a probability of instantaneous death as a result of job risks.   n166 They employ a temporal model that assumes that all life years are valued equally,   n167 and attempt to determine the relationship between wage premiums and job risks as a function of the remaining years of workers' lives (and other relevant characteristics).   n168

For example, consider two workers who have the same life expectancy and are otherwise also identical, but who demand different wage premiums for undertaking a risky occupation. The worker with the higher valuation (who therefore demands the higher wage premium) has a lower discount rate and therefore values more highly than her counter part the years that she will lose in the future. Alternatively, if two workers who have different life expectancies but are otherwise identical were to demand equal wage premiums, the worker with the shorter life expectancy will be exhibiting a lower discount rate: she will be valuing the future years more highly than the other individual.

On the basis of an empirical study of 1463 workers, Moore and Viscusi calculate a real discount rate of 2%.   n169 The authors note that this real rate "accords roughly with financial market interest rates for the period, once these nominal rates are adjusted for inflation."   n170 Their results, therefore, "provide no empirical support for utilizing a separate rate of discount for the health benefits of environmental policies."   n171

 [*976]  Moore and Viscusi reach this conclusion despite their earlier studies, which had found discount rates in the 10-12% range.   n172 They maintain that the confidence limits around these estimates were sufficiently large that the results should be thought of as "quite similar."   n173 The authors conclude:

 

 In each case the confidence intervals for the discount rate esti mates overlap available market rates of return. Moreover, since the point estimate of the discount rate falls short of the market rate in one case and exceeds the market rate in two cases, we find no clear evidence of systematic differences between dis count rates for health and financial rates of time preference.   n174

 

 With respect to the control of environmental carcinogens, it is relevant that the authors found that education has a large effect on the discount rate. In a study that found an overall real discount rate of 11%, the rates for workers with eight years of schooling and college-educated workers were 15% and 5.5%, respectively.   n175 Thus, to the extent that workers in risky occupations have a lower-than-average level of educational attainment, a downward adjustment on the discount rate would need to be made. For environmental carcinogens, this factor strengthens the authors' conclusion that the discount rate exhibited by financial markets is appropriate.   n176

 [*977]  To conclude, it is worth noting that the methodology used to estimate the rate at which individuals discount future utilities may lead to an overstatement of this rate. Recall that Moore and Viscusi assume that all life years are valued equally.   n177 This assumption is consistent with the standard approach in life-cycle models, in which the utilities derived from living in particular years are a function solely of the level of consumption available in those years.   n178 It is plausible, however, that such utilities are affected also by one's age, and that they fall (for a given level of consumption) with increasing age, as a result of the deterioration of one's physical capacity.

For example, at age fifty, one might not be able to engage in the full range of pleasurable activities that one could have undertaken at age thirty. Thus, the choices on how to convert consumption resources into utility at age fifty would be more constrained.   n179 If this were the case, part of the lower valuation attributed to later years in one's life would result from the lower utility derived from living during those years, rather than from discounting to reflect the passage of time. As a result, the discount rate estimated from a model in which utilities are constant across time (or a function only of the magnitude of resources available for consumption) would overestimate the actual discount rate.

2. Selecting an Appropriate Rate. - The choice of a discount rate is a key variable in the cost-benefit analysis of many environmental regulations. Because the costs of regulatory programs are typically borne around the time that the regulations go into effect but the benefits, in the case of latent harms, do not accrue for decades into the future, the higher the discount rate, the less desirable the regulation will seem. Re call, for example, that in the Corrosion Proof Fittings case, the present discounted value of the benefits would have been approximately ten times greater under a 4% discount rate than under a 10% discount rate.   n180

 [*978]  The OMB policy on discount rates does not address specifically the issue of how to discount health risks.   n181 Thus, these risks are discounted at the rates used in the evaluation of government projects in general, and government regulation in particular.

Until 1992, OMB employed a discount rate of 10% pursuant to a policy contained in its Circular A-94.   n182 In 1992, OMB amended this cir cular to mandate a real discount rate of 7%.   n183 OMB justifies this rate as "the marginal pretax rate of return on an average investment in the pri vate sector in recent years."   n184

The OMB policy, however, uses a different discount rate for cost- effectiveness analysis - that is, to determine which of several programs yielding identical benefits has the lowest cost in present discounted terms. For this purpose, OMB employs the real return on long-term government debt - the interest rate on long-term government bonds minus the rate of inflation.   n185 In recent years, this figure has fluctuated between 3% and 4%.   n186

The use of different rates for cost-benefit and cost-effectiveness analysis can produce perverse results. For example, consider two policies that have the same benefits, which are designed to address a future risk. Policy A costs $ 700,000 at present whereas Policy B costs $ 1,200,000 in ten years (the figures are in constant dollars). At a 3% discount rate, the present discounted value of the cost of Policy B is higher than $ 700,000, and thus Policy A would be preferred on cost-effectiveness grounds. On the other hand, at the discount rate of 7%, which would apply to cost- benefit analysis, Policy B would be more attractive.

Cost-effectiveness analysis can be used as a short-cut to cost-benefit analysis where the benefits of two policies are the same. But logic compels that the policy with the most attractive cost-benefit ratio also be the most cost-effective. This consistency requirement can be violated when the discount rates used for cost-benefit and cost-effectiveness analysis are different. Otherwise a trivial difference, say of one dollar, in the benefits of the two policies (so that cost-benefit analysis rather than cost-effective ness analysis must be used) would alter the choice between two policies that are essentially identical.

More fundamentally, however, there appears to be a growing consensus in the economics literature that the appropriate real discount rate for  [*979]  government projects is the real return on long-term government debt - the interest rate on long-term government bonds minus the rate of inflation. The underlying issues are quite complex, but can be simplified considerably for the purposes of this discussion.   n187

When the government undertakes a regulatory project, it is trading costs and benefits on behalf of its citizens. As Frank Arnold notes, "it then seems reasonable to discount the future benefits to the present using the same rate that the affected citizens would use, for it is on their behalf that the project is undertaken."   n188 This rate, often referred to in the literature as the "consumption" rate of interest,   n189 is generally taken to be the after-tax rate of return, adjusted for inflation,   n190 on relatively risk-free financial instruments,   n191 such as government bonds. In recent years, the economics literature has generally called for the use of a real discount rate of 2-3%.   n192

There is a complication, however. Consider initially two environ mental projects undertaken directly by the government, one financed by taxes and the other by borrowing. In the case of the project financed by taxes, the taxes will reduce the consumption of goods, so discounting the benefits at the consumption rate of interest is the appropriate procedure: individuals are simply trading off less consumption now, as a result of the taxes, for future benefits flowing from the project.   n193

The situation is potentially different if the government finances the project through borrowing. In a closed economy, with no capital flows into the country, the borrowing would displace money available for private investment. Because the returns from this investment yield taxes, its displacement would produce a loss to the government, equal to the fore gone taxes.   n194

An analytically analogous situation is posed by environmental regulation that imposes costs on firms, if these costs cannot be shifted to con  [*980]  sumers. In a closed economy, such investments would displace other private sector projects.   n195

The appropriate discount rate under these circumstances is the mar ginal pre-tax rate of return on private investment - the rate used by OMB.   n196 After this return is taxed by the government, the remaining return must be sufficient to cover the consumption rate of interest. If the return on the government's project was lower, social welfare would be enhanced by not undertaking the government project and thereby not displacing the private investment.   n197

In summary, traditionally, the literature on cost-benefit analysis inquired as to whether the project under consideration displaced consumption or private investment. It used the consumption rate of interest in the former case and the rate of return on capital in the latter.   n198

In recent years, however, the assumptions underlying this bifurcated approach have been called into question. In particular, increasing globalization has led to the integration of capital markets and the open ing of the U.S. economy to foreign investment.   n199 As a result, our econ omy can no longer realistically be viewed as closed. In an open economy, the level of taxable investments is unaffected by environmental regulation because no capital projects are displaced; the government therefore does not lose the corresponding tax revenues. Under these conditions, the consumption rate of interest is the appropriate discount rate.   n200

Consistent with this view, the consumption rate of interest is currently used as the discount rate by the General Accounting Office (GAO) and the Congressional Budget Office (CBO).   n201 Even EPA, which must submit its proposed and final regulations to OMB for review under  [*981]  Executive Order 12,866, has used a 3% discount rate in connection with a proposed regulation designed to address lead-based paint hazards.   n202 Other agencies, however, have explicitly linked their discount rate to OMB's.   n203

G. Estimating the Undervaluation of Lives Under OMB's Policy

 

 Section E explains the nature of the corrections that need to be made to intelligently translate the existing valuations of life from industrial accidents to appropriate valuations for environmental harms in general and carcinogenic harms in particular. Section F discusses how to choose an appropriate rate to discount the utility of life-years saved at the end of a latency period. The purpose of this section is to obtain a rough estimate of the underestimation of the value of human life that results from the OMB approach of taking valuations from workplace settings and mechanically reducing them by an inappropriately high discount rate over the length of the latency period. Because of OMB's role as the arbiter of regulatory analysis under Executive Order 12,866, this undervaluation has important public policy consequences.

Once again, the focus is on comparing the valuation of two different forty-year olds: one who faces a probability of instantaneous death in an industrial accident, V[in'40,40'], and the other who faces a probability of death at age 60 from an environmental carcinogen with a twenty-year latency period, V[in'40,60']. Recall the two factors that make V[in'40,60'] smaller.   n204 First, assuming for the sake of simplicity that these individuals would otherwise die at age 80, the number of life-years lost from the carcinogenic risk is only half. Second, the years lost from the carcinogenic harm occur later, and discounting is therefore appropriate; at a discount rate of 3%, the discount factor is 0.55. So, using round numbers, if these two corrections were the only relevant ones, V[in'40,60'] would be about one-quarter of V[in'40,40'], reflecting reductions of about one-half each on the account of the discounting and the difference in the life-years saved, respectively.

One should not overlook, however, the corrections on the other side, particularly those resulting from the involuntary nature of the environmental harm compared to the voluntary nature of the workplace harm, and the dread nature of deaths from environmental carcinogens compared to the non-dread nature of deaths from instantaneous industrial accidents. With respect to the first adjustment, the Cropper and Subramanian study, which compares deaths from voluntary and involun tary harms, suggests that an adjustment by a factor of two is appropri  [*982]  ate.   n205 As to the second adjustment, the study by Tolley, Kenkel, and Fabian finds that avoiding deaths from cancer is valued twice as much as avoiding instantaneous deaths.   n206

There is a question about how to combine the results of these two studies. It is not completely clear that the correction from the Tolley, Kenkel, and Fabian study is based only on the dread nature of the harm, and is not also affected by different degrees of voluntariness of the harm. If the carcinogenic and non-carcinogenic harms compared by these authors shared the same level of voluntariness, then it would be reasonable to multiply the two factors of two, and conclude that an adjustment by a factor of four is necessary to account for the differences in voluntariness and dread.

In contrast, if the carcinogenic harm considered in their estimate is less voluntary than the non-carcinogenic harm, such a correction would be excessive. It is clear that the difference in valuations comes in part from the morbidity that precedes carcinogenic deaths - one component of the dread nature of cancer.   n207 Moreover, nothing in the survey on which this study relied for the remainder of the correction focused the attention of the respondents on differences in the level of voluntariness.   n208 Thus, it seems unlikely that this issue would have played a large role in the valuations.   n209

While further research on these matters is clearly needed, to a first approximation it is reasonable in light of the designs of the two studies to treat the two factors as multiplicative. Thus, other things being equal, the value of avoiding a death from an involuntary, carcinogenic risk should be estimated as four times as large as the value of avoiding an instantaneous workplace fatality. This upward adjustment thus cancels the two downward adjustments resulting from the fewer number of life-years lost and the discounting for the latency period.

Moreover, other upward adjustments are necessary as well.   n210 First, as indicated above, the median salary for all wage earners is about 23% higher than the median salary for operators, fabricators and laborers, the U.S. Census category most likely to contain the subjects of willingness-to- pay studies in the context of industrial accidents.   n211 Thus, the valuation  [*983]  of lives threatened by environmental carcinogens should be the subject of an upward adjustment of another 23%.

Second, economic growth must be accounted for. As a result, based on the 1982-1996 period, the discount rate used in making the down ward adjustment necessary to account for the fact that the life-years would be lost in the future should be reduced by about 1%.   n212 Thus, accounting for economic growth leads to an upward adjustment of the valuation of life of 22%.   n213

As indicated above, the OMB approach is to take the valuations of life from workplace settings and discount them for the length of the latency period at a rate of 7%.   n214 While this approach does not reduce the valuation to reflect the smaller number of life-years saved,   n215 using a 7% discount rate instead of a 3% rate over a twenty-year latency period leads to a downward adjustment of the valuation by a factor of about four, rather than by a factor of about two.   n216 One would arrive at the same downward adjustment by a factor of four, however, if one took account of the smaller number of life-years saved and discounted at a 3% rate.

Moreover, the OMB approach neglects to perform any of the necessary upward adjustments. Thus, over a twenty-year latency period the approach may undervalue human life by a factor of about six.   n217 For contaminants with longer latency periods, the undervaluation would be even greater.   n218

Finally, this estimate of the undervaluation that results from the OMB approach is probably a lower bound. The true figure may well be higher because the calculation is based only on those differences between instantaneous deaths from workplace accidents and deaths from environ mental carcinogens that can be quantified on the basis of plausible empirical studies. The preceding discussion has identified two additional possible sources of undervaluation, but the quantification of the impact of these sources is not possible as a result of the lack of relevant empirical analysis. First, and probably most importantly, the population exposed to  [*984]  workplace accidents has a comparatively low willingness-to-pay to avoid death, as a result of a disproportionate tolerance for risk.   n219 Second, to the extent that, for a given level of resources available for consumption, the utility of being alive at a particular age falls with increasing age, the estimates in the literature of the rate at which individuals discount their future consumption would be higher than warranted.   n220

H. Recasting the Debate

 

 It is now worth highlighting that this Article's approach to discounting in an intragenerational setting does not pose significant ethical issues that are distinct from those raised by cost-benefit analysis in general or the valuation of human life in particular.   n221 In principle, one could directly ascertain, through willingness-to-pay studies, the value of lives threatened by latent harms. Because practical problems stand in the way of obtaining such valuations, a second-best measure, constructed in part by means of discounting future utilities, must be used instead.   n222 The use of such a proxy, however, does not give rise to ethical issues other than those that might exist if the measurement were done directly.

The reason for discounting in the case of latent harms is not that a regulator or some other outsider determines that life in the future is less valuable than life in the present.   n223 Instead, discounting simply reflects the fact that the individual who is valuing her own life derives less utility from living a year in the future than in the present.   n224 Discounting is therefore necessary to provide an accurate value of the utility that the individual loses in the present as a result of a premature death that might occur in the future.

At the same time, however, discounting is only one of many necessary adjustments that need to be made when valuations in the context of industrial accidents are used as the starting point to construct a value of human life for the purpose of regulating environmental carcinogens. It has no greater call for legitimacy than any of the other adjustments analyzed in Part I.E. As the various empirical estimates show, it is not even dominant in terms of magnitude.   n225 Thus, the failure of the regulatory process to make other adjustments, principally as a result of OMB's approach to the matter, leads to a substantial undervaluation of human life.   n226

 [*985]  The preceding discussion views discounting in this intrapersonal situation raised by the presence of latent harms as an essentially technocratic procedure, which must be undertaken in conjunction with other adjustments of the value of life from instantaneous industrial accidents, in order to obtain a second-best estimate of the value of a human life threatened by latent environmental contaminants. This characterization of the problem may give rise to two types of concerns. Neither, however, calls for a reevaluation of the ethical status of discounting in the case of latent harms.

First, one might worry that an individual's decisions today do not sufficiently protect the person that the individual might become in several decades. This perspective views the individual as a succession of "multiple selves."   n227 Its concern is that the individual's current self would make decisions that undervalued the interests of the individual's future self by choosing a discount rate that was too high. This formulation gives rise to a typical externality problem and converts a technocratic intrapersonal problem into an ethically-laden quasi-interpersonal one.

The objection, however, would not be confined to the role that discounting plays as a step toward a second-best valuation of human life threatened by latent harms. Precisely the same objection could be lodged against an attempt to measure this value directly through willing ness-to-pay studies. One would worry in this context that the wage premiums demanded by an individual would be too low because the future costs would be borne not by her current self but by a future self. The complaint would thus not be attributable to the specific role played by discounting but, more generally, to the process of valuing life itself. Thus, as a formal matter, the objection does not disprove my claim that discounting in an intragenerational setting poses no significant ethical issues that are distinct from those raised by cost-benefit analysis in general or the valuation of human life in particular.   n228

Moreover, such a criticism of revealed preference approaches to the valuation of threats to human life would not be confined to latent harms. Take, for example, an instantaneous industrial accident in which an individual faces probabilities of both death and serious morbidity. The individual's current self might not have sufficient empathy towards a future self confined to a wheelchair, and might therefore demand too low a wage premium.

 [*986]  More broadly, most decisions that we make have future consequences. Every time that we borrow money, we reduce the resources that will be available to us in the future. Similarly, every current expenditure affects the amount that will be available for future expenditures. To find an externality in each decision with future consequences as a result of the presence of multiple selves would open the door to government regulation of essentially every financial decision that we make. Such an approach would therefore constitute a serious affront to individual autonomy.

Interfering with individual preferences in this manner might be appropriate in the face of fairly egregious myopia. For example, in the somewhat analogous context of social welfare policy, Bruce Ackerman and Anne Alstott note:

 

 The aim of liberal policy is not to second-guess [individuals'] choices by supposing that everybody 'ought' to save a lot for retirement if they are to maximize their happiness over their life times. Its mission is more modest but more fundamental. It is to protect elderly citizens against the worst consequences of their earlier psychological myopia. The watchword is not utility maximization but the assurance of dignified existence in old age.   n229

 

 It would be unwarranted, however, to attack this Article's approach to the problem of latent harms by deploying the machinery of "multiple selves" analysis. Recall that the approach advocated here is to use the after-tax return on riskless investments - a rate that currently stands at between 2 and 3%.   n230 If this rate were to be trumped as insufficiently protective of the future, one would need to trump every decision to borrow money at market rates of interest. Then, governmental regulation of individual choices in the face of any decision with future consequences would become the norm, rather than a relatively rare club to be wielded only in the face of egregious lack of foresight.

A different type of objection might be raised to the claim that, in the context of latent harms, discounting is a technocratic exercise that does not give rise to difficult ethical choices. Different individuals have different discount rates, but the social decision of how to control latent environmental harms needs to be based on a single rate. Thus, in choosing the rate on which to base social policy, one needs to make some type of interpersonal comparison. Such comparisons, which are highly value laden, are inevitable, even if they are made implicitly by using a common rule of thumb such as basing the policy on the median discount rate.

Because environmental quality is a public good, once the govern ment acts, individuals will enjoy a uniform level of quality regardless of their individual discount rates. Thus, individuals with low discount rates would be exposed to more latent harms than they would have preferred,  [*987]  and individuals with high discount rates will be exposed to harms that are lower than they would have preferred (and consequently, perhaps, would have to face too high a current financial sacrifice to fund the policy).

This objection, again, is not particular to the role played by discounting future utilities in the case of latent harms, but can be raised more generally against both cost-benefit analysis and the valuation of human lives. Under cost-benefit analysis, public policy is chosen on the basis of the aggregate valuations of the benefits. Thus, individuals with particularly high valuations have to accept a policy that is laxer than they would have preferred, whereas individuals with a particularly low valuation face the opposite problem. Similarly, in the case of public policy decisions taken to prevent even instantaneous deaths, individuals who value their lives particularly highly (perhaps because they are unusually wealthy or have a particularly low tolerance for risk) will face a policy that is laxer than they would have preferred.

In summary, to the extent that the valuation procedures discussed in Part I give rise to ethical objections, these objections should be leveled either against cost-benefit analysis generally or against the valuation of life in particular.   n231 If these two techniques survive ethical scrutiny, no substantial independent ethical argument should be raised against the role played by discounting in an intragenerational setting. More generally, it is not defensible to argue that the value assigned by the regulatory process to a human life should be independent of when an individual's life-years are lost, regardless of how the timing affects the individual's own valuation.

II. Harms to Future Generations

 

 As indicated at the outset of this Article, discounting at a rate of re turn comparable to that earned by financial investments turns the utilities of generations living a few hundred years from now into a negligible present discounted value.   n232 Under such conditions, practically no current expenditure for the benefit of relatively distant generations could be jus tified within a cost-benefit framework. Because many of the consequences of climate change will not manifest themselves for a long time,   n233 the consequences of discounting at the rate of return of financial instruments may well be to make any plausible expenditure to address climate change fail a cost-benefit test.

 [*988]  The emphasis of many economists on the use of constant discounting models stands in stark contrast to the approach of international environmental law, which has given its unqualified endorsement to an alter native concept to guide intergenerational allocations: the principle of sustainable development. Indeed, the concept of sustainable development figures prominently in the most important agreements concerning international environmental law,   n234 including the Stockholm Declaration,   n235 the Rio Declaration,   n236 and the Framework Convention on Climate Change.   n237

Section A shows that models of discounting harms to future generations cannot be justified merely through appeals to logic. Section B re views the empirical literature concerning how individuals would discount benefits to future generations. The results reveal a strong intuition against the use of constant discounting models. Section C analyzes the serious shortcomings of discounting models when they are used in an intergenerational context. Section D discusses the role of opportunity costs; even if future utilities are not discounted, expenditures for environ mental projects might nonetheless be postponed if other investments can yield higher returns. Section E analyzes the principle of sustainable development and shows why it too suffers from serious shortcomings. Finally, Section F presents the outlines of an attractive theory of intergener ational obligations with respect to the environment.

A. Discounting and Appeals to Logic

 

 Some proponents of discounting the benefits to future generations justify their position through appeals to logic, invoking a set of absurd consequences that would inexorably follow if discounting was not per formed. Their arguments in this regard are unpersuasive.

1. No Environmental Projects Will Be Undertaken Unless One Discounts at a Market Rate. - Some commentators argue that unless environmental benefits are discounted at the rate of return on other investments, environmental expenditures would always be deferred into the future and ultimately would never be undertaken. For example, Susan Putnam and John Graham state:

 

 If a smaller discount rate were to be applied to health than to money, it would always make sense to postpone adoption of public health programs that invest money now for deferred health improvements. In short, society would continually delay  [*989]  risk reduction into the future and impose the burdens on future generations.   n238

 

 Similarly, according to Emmett Keeler and Shan Cretin:

 

 The discounting of costs but not benefits ... has a paralyzing effect on a decisionmaker.... For any attractive program, there is always a superior delayed program which should be funded first. The result is that no program with a finite starting date can be selected.   n239

 

 The idea behind this position is that, instead of undertaking the environ mental program, one could invest the funds in an alternative project, watch the investment grow, and then address the environmental problem at some time in the future. At this future time, moreover, one would engage in the same calculus and decide to postpone the environmental expenditure once more.

Environmentalists have traditionally favored low discount rates be cause the costs of environmental protection generally must be borne well before the benefits begin to accrue.   n240 Thus, a low discount rate makes a given expenditure seem more desirable. The argument that no environmental programs would be undertaken absent discounting at a market rate turns this view on its head: lack of discounting becomes environ mentally undesirable.

There are several responses to the justification of the discounting of environmental benefits by an appeal to a seemingly logical claim that any alternative would lead to the indefinite postponement of environmental expenditures. To begin, regardless of whether one discounted the environmental benefits at the market rate, it would always be desirable to undertake environmental investments that yielded a market rate of re turn. So, the claim has to be somewhat more modest: that only environ mental investments yielding at least a market rate of return would be undertaken. Other environmental projects, in contrast, would be delayed forever because they would always look more attractive in the future, after the funds that would have been allocated to these projects earned a higher rate of return elsewhere.   n241

There is then a seemingly inescapable logic to discounting environmental benefits at the rate of return earned by other investments. If one used a lower discount rate for environmental benefits, environmental remediation projects could pass a cost-benefit inquiry even though the resources would be best spent elsewhere. The use of a discount rate  [*990]  equal to the rate of return on other projects ensures that only desirable projects pass a cost-benefit test.   n242

Even with this reformulation, however, the appeal to logic assumes implicitly that the costs and benefits of the environmental program will remain unchanged over time;   n243 it is because of this invariance that delaying expenditures in order to invest at the market rate of return seems attractive. This assumption, however, is inconsistent with the structure of many environmental problems.

For example, in the case of the remediation of hazardous waste sites under the Superfund program, the damages caused by the contamination are likely to increase significantly over time if the problem is left unattended.   n244 If addressed early, a cleanup can take place before the hazardous waste has seeped down to an aquifer, affecting the quality of the groundwater. At this stage, the cost of remediation is comparatively modest and the damage from the contamination (and therefore the benefit of undertaking a remediation) is comparatively modest as well.

A few years or decades later, however, the pollutants may have worked their way down to the aquifer.   n245 Then, the damage may be far higher, since the pollutants could have destroyed important sources of drinking water. In turn, the costs of remediation would be far higher as well.   n246

Alternatively, certain environmental problems may become irreversible. Once that occurs, any finite expenditure on abatement, no matter how high, will fail to remedy the problem. The costs of abatement will effectively have increased to infinity.

Thus, in deciding whether to undertake an environmental project now, one cannot merely perform a static calculation of the magnitude of costs and damages on a particular date. One needs also to look at the problem dynamically and determine how the costs and damages would vary over time if the problem were left unattended.

Consider the following simple example. We could remove some soil from the site and incinerate it now at a cost of $ 110,   n247 and the damage from the current contamination is $ 100, reflecting a small increase in the cancer risk of certain residents in neighboring areas. If one looked at these figures statically, one would decide, on cost-benefit grounds, not to  [*991]  undertake the cleanup. If the problem is left unattended, however, in 10 years the remediation cost would be $ 500, as a result of the need to pump and treat groundwater, and damage from the contamination would be $ 600. At that point, the cleanup would be justifiable on cost-benefit grounds. For any plausible discount rate, however, it would be better to spend the $ 110 upfront to remove and incinerate the contaminated soil, thereby addressing the current $ 100 damage problem as well as prevent ing it from becoming a $ 600 damage problem in the future.

Thus, the situation described above presents three policy options: remediate now, remediate later, or do not remediate. It is desirable to remediate now not only when the current damage is greater than the current cost of addressing this damage, but also when the future damage is greater than the future cost of addressing it, and the increase in costs in the intervening period is greater than the rate of return on other investments.   n248

These features concerning the structure of environmental benefits and costs are no less an issue for climate change than they are for Superfund problems.   n249 Certain climate change problems may be irreversible,   n250 and in such cases delaying investment in the environmental project is not an option. More generally, to make intelligent policy choices one needs to know, for example, not only the costs and damages at the time that carbon dioxide loadings in the atmosphere are doubled relative to some baseline, but also how the damage changes over time and the extent to which this damage can be reduced by means of particular policy measures.   n251

In addition, in the case of climate change, there is the possibility of catastrophic consequences.   n252 In the face of such consequences, risk aversion would justify undertaking projects even if their expected return was lower than that of other projects.   n253

Moreover, the view that before addressing environmental programs we should exhaust higher-yielding investments in other areas overlooks important difficulties concerning the transfer of resources across  [*992]  projects.   n254 Say, for example, that initially the greatest returns to a given investment would be to improve the educational system of particularly poor developing countries.   n255 Over the first twenty years, resources in vested in this manner earn a greater return than if they had been placed in an environmental project. Moreover, over this period, the costs of environmental remediation are increasing at a rate lower than the return on the educational investment.

After twenty years, however, the calculus changes. The costs of the environmental project, though less than the resulting benefits, begin to rise at a rate higher than the rate of return to education in the developing country. At that point, it is desirable to take the proceeds of the educational investment and transfer them to the environmental investment.

There is good reason to be skeptical about the feasibility of this transfer. Part of the returns from the educational investment may have been consumed by its beneficiaries, and may therefore no longer be available to fund the environmental project. Other resources may be sunk in long- term investments, such as infrastructure, from which they could not feasibly be extricated.

The transfer of even liquid investments may raise problems. The developing countries (or whatever interest group benefits from the initial allocation) might object to having the resources transferred to address a problem that they attribute to developed countries. Absent their con sent, there might be no clear mechanism for effecting the transfer. Of course, one could attempt to deal with this problem ex ante by contracting between the provider of the funds and the temporary recipient. Nonetheless, there are likely to be difficulties enforcing the rights under such a contract.

In summary, the resort to logic must fail. Perhaps the argument could be further recast to state that environmental expenditures should not be undertaken if other projects have a higher return, if the costs and damages associated with leaving the environmental problem unattended do not rise too fast, if the potential for catastrophic environmental consequences in the absence of immediate measures is sufficiently low, and if the difficulties of transferring resources across projects are not insur mountable. Then, of course, the claim made by supporters of discounting would have lost all their bite and would have become essentially tautological.

2. Failure to Discount Would Lead to the Impoverishment of the Current Generation. - A different argument maintains that not discounting the  [*993]  value of benefits to future generations makes it desirable for us to impoverish ourselves down to subsistence levels for the benefit of future generations. As Tyler Cowen and Derek Parfit describe the argument (to which they do not subscribe):

 

 We clearly need a discount rate for theoretical reasons. Other wise any small increase in benefits that extends far into the future might demand any amount of sacrifice in the present, be cause in time the benefits would outweigh the cost.   n256

 

 The logic is not limited to our generation. In turn, subsequent generations face the same incentive, and they become impoverished as well. Thus, "failure to discount would leave all generations at a subsistence level of existence, because benefits would be postponed perpetually for the future."   n257

There are two serious problems with the argument. First, it assumes implicitly that the objective of the decisionmaker is to maximize a social welfare function that adds up the interests of all generations. Then, deferring consumption now makes additional resources available for the future, when more people are around to derive utility from them. The question of whether it is appropriate to determine our obligations to future generations by reference to an aggregate social welfare function can not be resolved as a matter of logic. Instead, it must be defended by means of an ethical theory.   n258 The argument that all generations will be impoverished unless we discount environmental benefits assumes away the hard ethical choice,   n259 and then notes that an absurd conclusion would follow absent discounting.

Moreover, the argument for discounting as a way to avoid impoverishment takes a truncated and fundamentally misleading view of the manner in which one generation affects the welfare of subsequent generations. One component, to be sure, is through its consumption of renewable and nonrenewable resources. Thus, one way in which we could attempt to impoverish ourselves is by foregoing the consumption of such resources.

 [*994]  But to a large extent the standard of living of future generations will depend on current investments in areas such as technological knowledge, educational attainment, and productive capacity.   n260 Would our generation make those investments if it was wholly deprived of the resulting benefit? The answer, presumably, must be negative - that the level of effort that we bring to the business of making investments with long-term consequences is a function of the benefits that we can realize from those investments.

As a result, a requirement that we impoverish ourselves to leave more resources for future generations could actually decrease, rather than in crease, the resources available in the future. One might respond by saying that our generation has an obligation to provide the level of invest ment that it would have provided under a regime in which it could at least share in the fruits of its labors. That may well be a plausible argument, but it derives from an ethical judgment. Thus, the appeal to logic fails here as well.

In summary, the failure to discount does not inexorably lead to the impoverishment of all generations; it does so only if one makes two ethical judgments: that the appropriate social welfare function adds up the utilities of all generations, and that the current generation has an ethical obligation to invest in a stock of activities affecting long-term well-being even if it cannot keep any of the resulting benefits.

B. Intuitions About Discounting

 

 Before proceeding further, it is worth reviewing some empirical studies seeking to determine how individuals think about long-term discounting issues. A caveat is appropriate at the outset. If individuals in the cur rent generation indicate that they would discount the benefits of future generations, one should not automatically conclude that the decision reflects an honest ethical judgment. Instead, the judgment of these individuals might be compromised by self-interest. On the other hand, it would be relevant if members of the current generation, despite their self-interest to the contrary, were prepared to make social decisions protective of future generations. Their generosity might be indicative of an ethical in tuition that the benefits accruing to future generations should not be discounted very much, or perhaps not at all.

Most of the empirical studies in this area use a similar methodology. Typical of the approach is the questionnaire prepared by Maureen Cropper, Sema Aydede and Paul Portney, which states:

 

 Without new programs, 100 people will die this year from pollution and 200 people will die 50 years from now. The government has to choose between programs that cost the same, but  [*995]  there is only enough money for one .... Which program would you choose?   n261

 

 In their surveys, the authors varied the number of lives that would be saved in the future (but kept constant at 100 the number of lives saved in the present). They also varied, between 5 years and 100 years, the time at which the future lives would be saved.   n262 From the responses, they computed the discount rates that the respondents assigned to future consequences. The mean of the respondents' discount rates was 8.6%, 6.8%, and 3.4%, for time horizons of 25, 50, and 100 years, respectively.   n263 A similar study, conducted in Sweden, calculated discount rates of about 25%, 12%, and 8%, for time horizons of 20, 50, and 100 years, respectively.   n264

More strikingly, another Swedish study sought to compare the seriousness of a leakage of spent nuclear fuel at times ranging between one thousand and almost two million years into the future. Almost one third of the respondents did not discount the future consequences at all. Among those who did, the mean discount rate attached to an accident in the year 10,000 was less than one-hundredth of one percent - practically zero.   n265

The studies reveal an essentially unanimous opposition to the core component of the traditional discounting model: that future consequences should be discounted at a constant rate and that the rate of discounting should be set by reference to the rate of return on particular investments.   n266 Instead, the studies show a consistent pattern under  [*996]  which the discount rate falls as the time horizon gets longer.   n267 More over, the discount rate with respect to very long time horizons is well under the rate of return on investments in financial markets.   n268

C. Discounting in a Global Utilitarian Calculus

 

 Thus, at this point the argument has established that the propriety of discounting the benefits to future generations cannot be resolved by appeals to logic. Moreover, empirical studies reveal a moral intuition op posed, over the long-term, to constant discounting at a rate of return comparable to that generated by financial markets. It is now time to focus directly on the propriety of discounting.

Most economic formulations of discounting in an intergenerational context posit a social welfare function that aggregates the utilities of individuals in the different generations.   n269 For each time period, the utility is multiplied by a rate of pure time preference, which is a measure of the difference in importance attached to current utility as compared to utility in the future.   n270 This rate could be zero (the utilities of current and future generations have the equal importance) or positive (the utilities of earlier generations are privileged).   n271 The goal of the decisionmaker is to maximize the aggregate utility function.   n272

 [*997]  In this framework, the discount rate that maximizes aggregate utility can be written as follows:

d = [rho] + [THETA] g where d is the discount rate, [rho] is the rate of pure time preference, [THETA] is the absolute value of the elasticity of marginal utility (a measure of the relative effect of a change in income on utility), and g is the growth rate of per capita consumption.   n273

The pure rate of time preference, [rho] , reflects the fact that if the social welfare function gives less weight to the utilities of later generations, then those utilities must be discounted in order to make them comparable to the utility of the current generation. The term composed of the product of [THETA] and g has a less direct genesis. Most economic models of discounting assume that individuals in the future will enjoy higher rates of consumption than individuals in the present: more specifically, the level of consumption will increase at a rate of g.   n274 The models also assume that individuals exhibit a declining marginal utility of consumption - that is, that a unit of consumption has a greater effect on the utility of an individual with a lower level of consumption than on one with a higher level of consumption.   n275

As a result, if later generations will enjoy a higher level of consumption as a result of economic growth, social welfare can be increased by allocating some additional resources to earlier generations. The [THETA] g term represents the amount of discounting that must be performed, in order to maximize social welfare, on account of the higher levels of consumption of later generations.

The following subsections deal specifically with each of the two components of the discount rate.

1. Pure Rate of Time Preference. - Exemplifying the position of many economists, Victor Fuchs and Richard Zeckhauser take a strong position in favor of discounting at the rate of return on financial instruments. They maintain:

 

 Most policy planning discussions assume full altruism - future citizens are given equal weight with present citizens - and discount solely for the time value of money. Given this ethical premise, the value of life years to future generations should be discounted at the time-value-of-money rate.   n276

 [*998] 

 

 Terming this approach "full altruism" is somewhat contrived. In fact, it privileges the interests of the current generation to a very large extent.

Recall that, at a time-value-of-money rate of 5%, this approach equates the loss of one life today with the loss of a billion lives in 500 years.   n277 Stated somewhat differently, assume that the population of the world remains constant at about 6 billion people over the next 500 years. Under a model of time discounting, what would be the maximum current expenditure that could be justified in order to prevent the death of every living individual in 500 years? Placing a value of life of $ 5 million, in constant dollars, the maximum current amount that we could justify spending now to avert the destruction of the human race in 500 years would be $ 30 million. (At the OMB rate of 7%, this amount would be only about $ 10!) More conventional definitions of altruism would presumably call for a different result.

Indeed, the discount factors are simply the weights used to compare the value attached to the utilities of individuals in different generations. A pure rate of time preference of zero is equivalent to giving the utility of persons living at different points in time the same weight in the social welfare calculus.   n278 Any positive rate simply reflects the preferences of a social welfare evaluator to depreciate the utilities of future generations.   n279

The ethically compromised status of discounting for time preference at a constant rate can perhaps be best illustrated by the following exam ple. Consider an exceedingly simple economy with 100 units of re sources. Two individuals, with identical utility functions, live in this economy: one from year 1 to year 50 and the other from year 51 to year 100. There is no possibility for productive activity; thus, the individuals will be able to derive utility only from the existing 100 units of resources.   n280

In the absence of discounting for time preference, each individual would be allocated 50 units of resources. In the face of a positive rate of time preference, however, even a relatively modest one, the first individual would get the bulk of the resources. It would be difficult to construct an attractive ethical theory that privileged the first individual in this manner merely because she lived fifty years earlier than the second individual.

The possible justifications for discounting for time preference at a positive rate are not compelling. First, one might posit that if discounting for time is appropriate intragenerationally, it should be acceptable in  [*999]  tergenerationally as well. There is a fundamental difference, however, between the two situations.

Intragenerational discounting affects the timing with which a partic ular individual decides to expend a fixed amount of resources. It is merely a reflection of the individual's preferences and, as discussed in Part I.H, does not raise any significant ethical questions.   n281 In contrast, intergenerational discounting affects the quantity of resources available to each individual.

In an intergenerational context, one must initially decide how to al locate resources to individuals in different generations - a societal decision with ethical underpinnings. Then, each individual must decide how to time the consumption of resources across her lifetime - a personal decision with no ethical ramifications,   n282 other than a weak concern about excessive myopia.   n283

Some economic models that purport to analyze intergenerational problems construct their utility function by reference to an individual who lives forever.   n284 Models of this type collapse the intergenerational and intragenerational aspects of the optimization across generations.   n285 Thus, they overlook an important dimension of the problem. One simply cannot avoid making ethical judgments about intergenerational transfers by mechanically importing to this endeavor the intragenerational framework.   n286

 [*1000]  The second possible justification is that time discounting does not show lesser regard for future generations because even though it under values the interests of a particular generation relative to an earlier generation, it overvalues its interests relative to a later one. According to this claim, each generation is treated in a comparable way: somewhat worse than its predecessors and somewhat better than its successors.

The claim is not an affirmative argument for discounting. Instead, its ambition is far narrower: it merely responds to one possible argument against discounting. It does not carry the day, however, even in this limited respect. Absent economic growth, as would be the case for example in economies with high levels of consumption, constant discounting for time preference would lead to the progressive impoverishment of subsequent generations. Given the choice between consuming resources in the present and leaving them for future generations one would choose the former because the utilities derived from these resources by later generations would be heavily discounted.

It is true that if discounting actually threatened to impoverish future generations additional resources would be allocated to these generations as a result of the declining marginal utility of consumption, which would make the poorer generations value a unit of consumption more. This phenomenon, which is a feature of growth discounting at a negative rate of growth, could mitigate some of the harshness that would otherwise result. The existence of such a safety valve, however, is hardly a ringing endorsement of discounting for the pure rate of time preference.

Yet another argument for discounting for time preference focuses on the greater affinity that the current generation feels for itself and for the generations that immediately follow it. As Kenneth Arrow and several co-authors note, the rate of time preference "may represent discounting for empathetic distance (because we may feel greater affinity for generations closer to us)."   n287 By its terms, the statement purports to make a descriptive claim rather than a normative judgment: it does not explain why a social welfare function that reflects such judgments is ethically defensible.   n288

Moreover, this argument for discounting is suspect even as a descriptive claim, as the empirical evidence discussed in Part II.B shows quite  [*1001]  clearly.   n289 It is plausible that we would like to favor ourselves over future generations, and that with respect to future generations we would like to privilege the generations of our children and grandchildren, and per haps even great-grandchildren, over subsequent generations.   n290 But discounting at a constant rate implies that our decreasing regard for subsequent generations continues forever. For example, it seems unlikely that, on this account, we would value the loss of one billion lives 1000 years no more than the loss of one life 500 years from now, as would be the case if we used a discount rate of 5%.   n291

Other commentators justify discounting by reference to the probability that some catastrophe in the future will result in the destruction of human civilization.   n292 The point then is that if we are not sure that a future generation will exist, we should allocate more resources to earlier generations, which are more likely to be around to enjoy the re sources. This argument could well justify discounting at a constant rate, but it is very unlikely that the rate would be more than infinitesimal.   n293

Also embedded in the claim is an ethical issue. To some extent, the survival of humanity is imperiled by actions of our generation, and of a few generations immediately preceding ours. The consequences of nu clear war are one such example. Over the long run, climate change itself may result in a catastrophic scenario.   n294 If we are contributing to the probability of humanity's extinction, should we then invoke this possible outcome as an argument to allocate more resources to ourselves? A quite plausible principle is that the current generation should not benefit in this manner from its externalizing behavior.

Finally, time discounting is sometimes justified on the grounds that over time some kind of countermeasures or cures for environmental problems may be devised.   n295 If, indeed, there were a scientific basis to support such an assumption, a welfarist framework would call for reducing the harm by the probability that ultimately the harm will not in fact accrue. To the extent that the harm was potentially a catastrophic one, however, risk aversion would mitigate that reduction.   n296 More funda mentally, it would be an exceedingly unusual coincidence if the  [*1002]  probability that an environmental problem would self-correct just happened to equal the interest rate on financial instruments for every problem and for every length of time.   n297 Thus, in its general formulation, this argument for discounting must be rejected as devoid of any factual basis.   n298

In summary, the arguments for discounting as a result of the pure time preference are not compelling.   n299 The confusion surrounding the issue stems, at least in part, from equating intragenerational discounting, which ought not to be considered particularly controversial,   n300 with intergenerational discounting,   n301 which raises a different set of issues.   n302  [*1003]  To conclude, it is worth noting that even though discounting for time preference is a relatively standard technique in economics, there is a long and respectable tradition, traced to an article published in 1926 by Frank Ramsey, that rejects such discounting in intergenerational contexts.   n303

2. Growth in Levels of Consumption Over Time. - It is time to turn to the question of discounting as a result of the growth in levels of consumption over time. Recall that the argument in favor of such discounting rests on the predicted additional wealth of future generations and the decreasing marginal utility of consumption.   n304 Given these conditions, growth discounting leads to the maximization of the social welfare function.   n305

Before evaluating the argument for such discounting, it is worth pausing to consider the magnitude of what is at stake. As explained above, the discount rate for growth that maximizes social welfare is the product of g, the growth rate of per capita consumption, and [THETA] , the abso lute value of the elasticity of marginal utility. Arrow and his co-authors indicate that most empirical estimates of this elasticity place it in the range between one and two; thus they use the mid-point, 1.5, in some of their calculations.   n306 With respect to long-term per capita growth, the central estimate of the Intergovernmental Panel on Climate Change placed it at 1.6%.   n307 Thus, the rate of discount for growth would be 2.4%. This amount is far from inconsequential. It implies, for example, that we would be indifferent between saving one life now and 10.7 lives in 100 years, or between saving one life now and 141,247 lives in 500 years.

This type of discounting gives rise to two important concerns. First, to the extent that subsequent generations are wealthier, they will value the benefits of environmental protection more highly. The standard economic models calculate the environmental damage on the basis of the valuation of the current generation: economic growth implies that later generations will have higher valuations.   n308 Standard estimates of the benefits of climate change measures include a reduction in the loss of lives.   n309 As shown above, the elasticity of this valuation with respect to levels of consumption is approximately one.   n310 Thus, this valuation  [*1004]  should be expected to rise at the rate of economic growth.   n311 Similarly, valuations of environmental amenities and natural resources are closely linked to levels of income,   n312 and will rise with rising income.   n313 If the valuation of all the components of the damage of climate change in creased at the rate of economic growth, this factor would either completely cancel out any discounting as a result of greater wealth (when [THETA] is equal to one), or greatly reduce the extent of such discounting (when [THETA] is somewhat greater than one).

More fundamentally, the growth discounting account assumes implicitly that the benefits of environmental activities are distributed in the same manner as the costs. Then, because the benefits accrue to individuals who are wealthier than those who bear the costs, the beneficiaries have a lower marginal utility of consumption, and discounting is necessary to maximize social welfare. This implicit assumption is highly questionable. Most studies of the impact of climate change show that the damages will be suffered disproportionately by individuals in poor developing countries: Bangladesh, for example, is likely to be particularly affected by sea level rises.   n314 In contrast, the contribution to the global warming problem lies to a large extent with the developed countries, and financial responsibility for mitigation measures will be borne primarily by these countries.   n315

Currently, the United States and Bangladesh have per capita gross national products (GNP) of $ 26,980 and $ 240, respectively.   n316 The figures differ by a factor of about 112. It is quite unlikely that in 100 years or so Bangladesh and the United States will have the same per capita GNP. Thus, to the extent that the United States is paying for the environmental measures and Bangladesh is benefiting from them, the kind of growth discounting contemplated in the standard economic models is clearly inapposite. In order to maximize the social welfare function, a lower factor would have to be used to reflect the fact that even when the  [*1005]  benefits of climate change measures begin to accrue, Bangladesh will be poorer than the United States.

It is quite possible that even in a hundred years Bangladesh's per capita GNP, in constant dollars, will be lower than the per capita GNP in the United States is now. Then, in order to maximize the social welfare function, one would have to apply a negative discount rate. Such a rate would justify spending more now than the benefits in the future because the benefits in the future would accrue to individuals with lower levels of consumption, and hence higher marginal utilities of consumption.

One might object to this line of argument on the grounds that citizens of the United States have no obligation to improve the lot of Bangladesh. Such a position is certainly debatable, but it resides outside the domain of utilitarianism, where the concept of discounting future utilities has its intellectual home. In the example described above, where in constant dollars the per capita GNP in Bangladesh in 100 years is lower than the current per capita GNP in the United States, a negative discount rate does maximize the social welfare function and is the policy that should be chosen on utilitarian grounds.

This discussion points to an obvious anomaly. If we are prepared to be serious about utilitarianism in the intergenerational context, why do we not take it seriously in the intragenerational context? Doing so would imply a large increase in the aid from developed to developing countries, where the marginal utility of consumption is far higher as a result of the much lower per capita GNP.

One can, to be sure, construct a plausible ethical theory under which greater current foreign aid is not compelled but mitigation measures for climate change are. The depressed economic status of developing countries might not be the direct consequence of any actions by the developed countries, although the issue is not uncontroversial. In contrast, any damages that might affect developing countries as a result of climate changes are caused to a large degree by energy consumption patterns in the developed countries.   n317 So, the developed countries might have an obligation to mitigate a problem that they caused and yet not have a similar obligation to reduce a level of inequality that they did not cause.

It is difficult, however, to reconcile such an ethical theory with welfarist approaches. Whether the lower level of per capita GNP in developing countries is caused by climate change or not, it still results in a higher marginal utility of consumption. If the purpose is to transfer re sources to where they will produce the greatest increase in utility, the cause of the inequality simply does not matter. Moreover, the selective rejection of utilitarianism to justify the current low levels of foreign aid  [*1006]  would call into question its selective invocation to justify discounting in some fashion the benefits to future generations of environmental measures.   n318

Alternatively, one might argue that utilitarianism calls for maximizing only the aggregate social welfare function of the relevant polity. With respect to the analysis of foreign aid, the relevant polity might be each individual nation. Foreign aid would then be justified only to the extent that donors in a wealthy country derive utility from helping recipients in a poorer country, not on the basis of the utility derived by the recipients.

In the context of climate change, given the global nature of the problem, it would be paradoxical to decide on a nation's obligations merely by reference to that nation's aggregate social welfare function. Indeed, the standard economic formulation of discounting aggregates across a global social welfare function and no commentator that I am aware of argues for a more constrained view. Perhaps one could construct a defensible theory under which the relevant polity changed with the nature of the problem, but it could not be derived solely from utilitarian principles and would have to be grounded on some nonconsequentialist ethical norm.

Growth discounting also inappropriately merges the decision con cerning the desirability of a project with distributional considerations. Under cost-benefit analysis, projects are undertaken based on the aggregate willingness-to-pay of the beneficiaries. Because the government un dertakes large numbers of projects and regulatory initiatives, the losers with respect to one governmental intervention may well become winners with respect to another. It therefore does not make sense to suffer social welfare losses with respect to an individual project simply to obtain a more desirable distribution of resources.

After aggregating all projects, however, the set of policies that maximizes net social welfare across the population as a whole might impose significant net costs on a subset of the population. To the extent that such inequities persist, the government can effect redistribution intragenerationally through the income tax system. Such an approach generally gives rise to fewer distortions and is therefore more desirable than compromising the social welfare consequences of individual projects.   n319

In contrast, under growth discounting, the amount invested in an environmental project will be less than that justified by reference to the  [*1007]  aggregate willingness-to-pay of the beneficiaries. Thus, the efficiency of each individual project would be compromised in order to effect redistribution.

It is true, of course, that intergenerational redistribution is more dif ficult to achieve than its intragenerational counterpart. For example, if we allocate more to the current generation in order to improve the aggregate social welfare but feel that such a policy imposes net costs on future generations, there is no easy means to compensate future genera tions. In theory, we could tax ourselves to create a trust fund that future generations could tap into at predetermined times, but there is a high likelihood that the money would become an attractive target in the future for our generation, or for intervening generations. Thus, the durability of the arrangement over the long-term could not be assured.

A different problem would arise if social welfare were to be maximized by allocating resources to future generations in a manner that imposed unacceptably high net costs on the current generation - the phenomenon that underlies the growth discounting approach. There is no obviously desirable mechanism by which we could tax future generations in order to compensate ourselves.   n320 While we could consume suboptimally high levels of renewable and nonrenewable resources, such consumption imperils social welfare in a way that is avoided by redistribution through the tax system. A better alternative is to finance measures that benefit the current generation through long-term debt, the burden of which would eventually fall on future generations.

These difficulties suggest that the benefits of intragenerational redistribution through the tax system will not be fully available intergenerationally. Nonetheless, these difficulties do not necessarily call for conflating the resource allocation and distribution inquiries, as growth discounting does. Instead, one needs to ascertain, as one typically does in the intragenerational context, whether bifurcating the inquiry and per forming the redistribution through a different mechanism would reduce undesirable distortions.

D. Role of Opportunity Costs

 

 My argument should not be read to imply that discounting has no role to play in the intergenerational context. For example, consider a harm that could be averted either now or in the future. In this scenario, assume that if the problem were addressed in the future, funds could be invested now in other projects and then transferred at a later time to avert the harm. The most that it would be worth paying to avert the future harm now is the present discounted value, at the rate of return generated by these alternative projects, of the amount that would be needed if the problem were addressed in the future. Regardless of the nature of  [*1008]  our obligation to future generations, it makes no sense to spend more when we can achieve the same result for less.

A similar result could attach even to an irreversible environmental problem. Consider an environmental harm that can be remedied only through a current expenditure: if the problem is not addressed now, it cannot be successfully addressed in the future. Even if the objective were to transfer resources to a future generation, it might nonetheless be preferable to leave the problem unattended if alternative investments would yield a higher rate of return. Then, the future generation would have to face the environmental harm but would enjoy, for example, the fruits of greater investments in technological innovation.   n321

The substitutability of environmental and non-environmental benefits can be seen most clearly from the vantage point of a utilitarian perspective. The utilitarian objective is to deploy society's resources in whatever way increases aggregate utility by the largest amount, not to pre vent specific environmental harms. Suppose that aggregate utility would increase by transferring current resources to a future generation. If a given investment of resources would yield a larger return in a non-environmental project, the utilitarian calculus would favor this investment over an environmental investment yielding a lower return.

One might conclude at first glance that my disagreement with advocates of discounting the utilities of future generations is only semantic. It might appear, indeed, that taking account of opportunity costs in deciding whether to undertake environmental projects for the benefit of future generations leads to the same results as discounting the utilities of those generations.

Indeed, consider the following two procedures. Under the first procedure, one undertakes any project for which the current cost (in fore gone utility for the current generation) is greater than the present discounted value of the utilities of the future generation that the project is intended to benefit. Under the second procedure, one does not discount the utilities of future generations, but undertakes the project only if the rate of return of the investment is greater than the rate of return of alter native investments (otherwise, even if resources are worth transferring into the future, the alternative investments will be preferable).

As is almost self-evident, these two procedures will yield the same results in certain cases. These procedures, however, are conceptually different and can yield different results in other cases.

Most importantly, discounting the utilities of future generations is a means for determining our obligations to those generations. It is the objective function of a specific ethical theory. In contrast, paying attention to opportunity costs does not imply the choice of any particular theory. It  [*1009]  is simply a way of ensuring that society furthers its chosen theory, whatever that theory may be, in the most cost-effective way possible.

For example, suppose that a societal goal is in fact to prevent certain types of irreversible environmental harms, as may be the case under formulations of the principle of sustainable development.   n322 We would still defer expenditures for environmental projects if alternative uses of the funds could have a higher rate of return over a given period. But at the point at which such a harm was about to become irreversible, we would undertake the environmental expenditure to prevent this outcome regardless of the rate of return on other projects. Moreover, in deciding how long to delay the expenditure, one would have to consider whether funds invested in other projects could easily be transferred at a later time to the environmental project.   n323 In contrast, if the social objective were to maximize a discounted social welfare function, the expenditure would never be undertaken if the present discounted value of the benefits was lower than the costs.

Similarly, under a corrective justice approach, countries responsible for environmental degradation would have an obligation to mitigate the adverse effects of such degradation. It would nonetheless be appropriate to delay expenditures if alternative interim investments were to yield a higher rate of return. But, at some point, the mitigation would have to be tackled. In contrast, the approach of discounting the utilities of future generations could provide a different prescription altogether.

E. Intergenerational Obligations and Sustainable Development

 

 There is virtual agreement that the central function of the principle of sustainable development is to guide intergenerational allocations.   n324 Because this principle is strongly endorsed in international environmental law agreements,   n325 it is important to ascertain the extent  [*1010]  to which it sets forth an attractive theory of intergenerational obligations.   n326

Before turning to this task, however, one must at least attempt to convert what is still quite an amorphous concept, which suffers from the lack of a uniform definition,   n327 into a tool that can actually guide decisions. The starting point to most discussions in this area is the language in Our Common Future, the 1987 report of the World Commission on Environment and Development (often referred to as the Brundtland Report, after its chair, the then Prime Minister of Norway).   n328 This report defines sustainable development as development that "meets the needs of the present without compromising the ability of future generations to meet their own needs."   n329 This statement, however, leaves open wide room for disagreement.

Perhaps the two most influential perspectives on what obligations to future generations are encompassed by the principle of sustainable development are those of Edith Brown Weiss and Robert Solow, which are rooted in the traditions of international law and of economics, respectively.   n330

Weiss equates sustainable development with intergenerational equity, which she defines by reference to three principles.   n331 First, the principle of conservation of options requires each generation to preserve the natural and cultural resource bases so that the options available to future generations are not unduly restricted. Second, the principle of conservation of quality requires each generation to prevent a worsening of the planet's environmental quality. Third, the principle of conservation of  [*1011]  access requires each generation to provide its members with equitable rights of access to the legacy of past generations, and to conserve this access for the benefit of future generations.   n332

In contrast, according to Solow, sustainability requires that each future generation have the means to be as well off as its predecessors. He gives content to this principle by proposing a modification to the traditional measure of a nation's economic activity. From Net National Product (NNP) - Gross Domestic Product (GDP) minus the depreciation of fixed capital assets - he would subtract the value of expended nonrenewable resources and environmental assets like clean air and water.   n333 Solow argues that each generation must use its nonrenewable and environmental resources in a way that does not detract from the ability of future generations to have a similar standard of living.   n334 He admits that certain unique and irreplaceable resources, like certain national parks, should be preserved for their own sake,   n335 but maintains that the con sumption of non-unique natural and environmental resources ought to be permissible as long as they are replaced by other resources such as equipment or technological knowledge.

The two formulations share important characteristics. First, they de fine the primary obligation to future generations in terms of a constraint that specifies how much must be left to a subsequent generation.   n336 Second, Weiss and Solow would both allow some level of destruction of most natural resources, as long as future generations are compensated in an other way, such as by technological development.   n337 Third, they both regard certain natural resources as irreplaceable and would require that such resources be protected for subsequent generations.   n338

 [*1012]  In essence, then, under both formulations, every generation must provide the subsequent generation with the means to do at least as well as it did. So, for example, sustainable development would be consistent with the current generation seeking to maximize its own utility, as long as this maximization is subject to a constraint resulting from the need to leave sufficient resources to future generations.

There are, of course, daunting challenges ahead in providing further specificity to the principle. For example, additional work needs to be done to determine how to value the increase in knowledge or the negative long-term environmental effects of economic activity.   n339

Also, throughout history, there has been a progressive increase in standards of living. Should the constraint defining one generation's obligation to its successors thus provide for a progressive increase in well- being, so that this pattern may continue? On what basis would that in crease be determined? What would be the ethical underpinnings for such a requirement?

Moreover, the link between sustainable development and population policy is not well articulated.   n340 The population in any generation is a function of decisions of prior generations.   n341 For example, one might argue that if the current generation's actions were to lead to an increase in population, it would have an obligation to provide additional resources so as not imperil the level of well-being of an average person in the next generation.   n342

 [*1013]  Many commentators also believe that the concept of sustainable de velopment contains a precautionary principle, which prescribes that sci entific uncertainties be resolved in favor of environmental controls.   n343 As discussed above, there is some possibility that catastrophic events would materialize in the future if the climate change problem is left unat tended.   n344 The precautionary principle would presumably call for avoid ing such consequences. In fact, given that technological advances may greatly contribute to the wealth of future generations, it may be that the precautionary principle will do most of the work in justifying climate change expenditures.

Left unanswered in the academic discussions concerning the precautionary principle, however, are important questions about its scope. For example, what probability of a catastrophic event is sufficiently high to trigger the operation of the principle? Similarly, what is a sufficiently harmful consequence?   n345 Spending the resources needed to avoid a low- probability, catastrophic outcome might interfere with the ability to make resources available to subsequent generations. How should this tradeoff be resolved?

This background on the scope of the principle of sustainable development is sufficient to permit an evaluation of the extent to which the principle can form the basis for a desirable theory of intergenerational obligations with respect to environmental matters. At a very general level, the principle appropriately underscores that the current generation, which has control of vast decisionmaking authority concerning the resources that will be available in the future, should not simply ignore the interests of future generations.

Beyond this level of generality, however, the principle suffers from severe shortcomings. Most importantly, in practice it is likely to impose too limited an obligation on the current generation. Say, for example, that the current generation, for a comparative small sacrifice, can prevent a very large harm to a subsequent generation. Perhaps an expenditure of only $ 1 at the present would lead to averting harm of several hundred billion dollars in 100 years. Even if the future benefit were discounted at a high level, the present discounted value of the benefit would greatly exceed the corresponding cost.

 [*1014]  The principle of sustainable development, however, would not re quire this expenditure if the subsequent generation would, despite the harm, be better off than the current one. Thus, if the next hundred years can be expected to bring sufficiently rapid technological progress, the environmental expenditure would not need to be undertaken. In fact, because the rate of technological progress is currently so high, the principle of sustainable development could in fact remove from the current generation any obligation to undertake environmental measures for the benefit of future generations.

Conversely, while this issue is of less direct practical importance, the principle of sustainable development could, in theory, demand excessive sacrifice from the current generation. Say, for example, that absent some intervention, the generation living 100 years from now would be $ 1 poorer than the current generation, and that for an expenditure of several hundred billion we could confer upon that generation an extra $ 1. The principle of sustainable development would require the expenditure, despite the obvious waste in resources.   n346

These shortcomings of the principle of sustainable development serve to underscore the relative attractiveness of utilitarian approaches. Consistent with such approaches, in an intragenerational context, the social decisionmaker would seek first to undertake all projects that have desirable cost-benefit ratios. Then, if the resulting distribution of re sources was unattractive, the social decisionmaker would require redistribution. In a utilitarian framework, redistribution is justified as a result of the fact that poorer individuals have a higher marginal utility of consumption; total utility is therefore increased by redistributing from rich to poor.   n347

The costs of effecting redistribution (whether in the form of transaction costs or perverse incentives) play an important role in determining how much redistribution is socially desirable. Indeed, sufficiently high costs could dominate the benefits that would come from transferring re sources from wealthier individuals, with a lower marginal utility of consumption, to their poorer counterparts.

In an intergenerational context, the inquiry could be essentially the same: pick projects with good cost-benefit ratios and redistribute as guided by reference to the relative marginal utilities of consumption and by the costs of effecting redistribution. In contrast, the principle of sustainable development requires expenditures with unattractive cost-benefit ratios, fails to require expenditures with attractive cost-benefit ratios, and is oblivious to the costs of effecting redistribution.

 [*1015] 

F. Toward a Theory of Intergenerational Obligations

 

 The articulation of a complete theory of intergenerational obligations with respect to environmental matters is beyond the scope of this Article. Nonetheless, the preceding discussion can be crystallized into a set of principles setting forth the backbone for such a theory.

First, the mechanical importation of discounting for time preference at the rate used intragenerationally is wholly unjustified: how one individual decides to time her expenditure of a fixed set of resources over her lifetime is a fundamentally different question from how society allocates a given set of resources among individuals in different generations.   n348 Intergenerationally, discounting for time preference unjustifiably underval ues the interests of future generations.

Second, discounting for economic growth is also fraught with problems. Most importantly, the formula used in the standard economic models ignores the fact that the primary contributors to international environmental measures are far wealthier than the primary beneficiaries of such measures. In fact, even in the future, when the benefits of measures undertaken now actually accrue, these beneficiaries are likely to be poorer than the contributors to such measures are now. Under these circumstances, any positive discounting for economic growth would be inappropriate. To the contrary, given the decreasing marginal utility of consumption, a utilitarian framework would call for environmentally protective measures even if the current costs are somewhat greater than the future benefits.   n349

Third, a theory of intergenerational obligation must play close attention to opportunity costs. Even though it is inappropriate to discount the utility functions of future generations, it does not make sense to under take environmental expenditures for the benefit of future generations if the investment can yield higher benefits elsewhere, and if no ethical obligations are compromised by delaying expenditures.

Fourth, consistent with the principle of sustainable development,   n350 an attractive theory of intergenerational obligations should seek to pre vent catastrophic harms and the destruction of unique natural resources. Admittedly, however, the dividing line between the use of everyday re newable and nonrenewable natural resources, and the destruction of unique resources may be hard to draw in particular circumstances.

Fifth, proper attention needs to be given to distributional issues. As in the intragenerational context, one should not compromise the efficiency of a particular environmental policy in the name of distributional concerns, but one should be prepared to redistribute if the aggregate effects of such policies lead to unattractive distributional outcomes. In  [*1016]  the intergenerational context, the mechanisms for redistribution are more cumbersome,   n351 but the issue nonetheless merits attention.

Sixth, an attractive theory of intergenerational obligations is likely to contain a corrective justice component. Within a traditional utilitarian framework, one cannot explain the moral intuition that industrialized nations have a responsibility to mitigate the adverse effects of climate change, but not to effect massive current redistributions of wealth to poorer countries.   n352 To the extent that the current pattern of expenditures and concern on the part of industrialized countries derives from a moral intuition concerning differential levels of responsibility for the two situations,   n353 this intuition should be an element of a theory of intergenerational obligations.

Conclusion

 

 This Article shows that the lack of a proper understanding of discounting has led to bad regulatory decisions in the case of latent harms and to an undesirable skewing of the debate in the case of harms to fu ture generations.

If two individuals of the same age are exposed to a latent harm from an environmental carcinogen and to a risk of instantaneous death, respectively, the person exposed to the carcinogen stands to lose fewer life- years and to lose them later in life. Discounting is an appropriate technique for taking account of the latter factor. The use of discounting, however, will lead to misleadingly low valuations of life unless it is coupled with significant upward adjustments to account for the dread and involuntary nature of environmental carcinogens, as well as for higher income levels of the victims. Unfortunately, the regulatory regime has failed to recognize the need for such adjustments.

With respect to harms to future generations, the Article shows that the use of discounting is ethically unjustified. It privileges the interests of the current generation without a defensible foundation.

The misguided approach to discounting in the two contexts may be attributable in part to a fairly generalized failure to take proper account of the differences between the cases of latent harms and harms to future generations. For the former, discounting raises no significant ethical objections that are independent of those that could be raised against cost- benefit analysis in general and the valuation of human lives in particular. For the latter, in contrast, discounting gives rise to daunting ethical issues.

This Article aims to effect two important public policy changes. With respect to latent harms, it seeks to provide an impetus for correcting the substantial undervaluation of environmental benefits that comes from  [*1017]  the regulatory system's approach of mechanically taking valuations of life from the workplace setting and discounting them at an artificially high rate, without performing any of the necessary upward adjustments. With respect to harms to future generations, it seeks to move the debate away from discounting and towards more attractive alternatives.

 

 

FOOTNOTES:

n1. Exec. Order No. 12,866, 3 C.F.R. 1993, p.638, reprinted in 5 U.S.C. 601 (1994). This order replaced a similar Executive Order, promulgated by President Reagan. See Exec. Order No. 12,291, 3 C.F.R. 1981, p.127, formerly in 5 U.S.C. 601. Given its legal status, however, it cannot displace contrary statutory provisions.

For discussion of the practice of OMB review, see Environmental Policy Under Reagan's Executive Order: The Role of Benefit-Cost Analysis (V. Kerry Smith ed., 1984); Thomas O. McGarity, Reinventing Rationality: The Role of Regulatory Analysis in the Federal Bureaucracy (1991); Richard H. Pildes & Cass R. Sunstein, Reinventing the Regulatory State, 62 U. Chi. L. Rev. 1 (1995).

n2. Currently, a bill sponsored by Senator Carl M. Levin, Democrat of Michigan, which enjoys bipartisan co-sponsorship, is pending before the Senate. S. 746, 106th Cong. (1999). It mandates the preparation of a cost-benefit analysis for major rules. See id.

 623(b)(2). The bill does not preclude an agency from promulgating regulations that fail a cost-benefit test but imposes seemingly tough hurdles to such regulations. See id.

 623(d)(2). Legislative efforts to require that essentially all important regulations satisfy a cost-benefit test, began in earnest with the 104th Congress "Contract with America." See Cass R. Sunstein, Congress, Constitutional Moments, and the Cost-Benefit State, 48 Stan. L. Rev. 247 (1996); see infra text accompanying notes 56-58 (views of Senator Leahy on S. 343). The House passed a bill during the Congress' second month, Sunstein, supra, at 275-76, but a companion bill in the Senate failed to move forward when cloture was defeated, id. at 277-82.

n3. For example, Richard Morgenstern explains: "The value of fatality risk reduction figures prominently in assessment of environmental benefits. In the case of air pollution, the reduced risk of death often accounts for the largest single component of the dollar value of environmental benefits." Richard D. Morgenstern, Conducting an Economic Analysis: Rationale, Issues, and Requirements, in Economic Analyses at EPA: Assessing Regulatory Impact 25, 41-42 (Richard D. Morgenstern, ed., 1997); see James K. Hammitt, Stratospheric-Ozone Depletion, in id. at 131, 151-52 (value of averted skin cancer mortality comprises 98% of the benefits of the regulations implementing the Montreal Protocol). More generally, for all health-and-safety regulations, one recent estimate is that "about 60 percent of the total benefits results from reduction in the risk of death, disease, and injury." Robert W. Hahn, Regulatory Reform: What Do the Government's Numbers Tell Us?, in Risks, Costs, and Lives Saved: Getting Better Results from Regulation 208, 219 (Robert W. Hahn ed., 1996).

Moreover, even in cases in which there are other benefits, EPA's calculation of the magnitude of the benefits focuses on human health effects. See Lisa Heinzerling, Reductionist Regulatory Reform, 8 Fordham Envtl. L.J. 459, 461-62 (1997). For examples, see id. at 495 (asbestos ban); Ronnie Levin, Lead in Drinking Water, in Economic Analyses at EPA, supra, at 205, 227 (corrosion control). The same failure to quantify benefits other than those related to human health effects and mortality are also present with regard to agricultural pesticides, worker protection and primary air quality standards for ozone depletion. Louis P. True Jr., Agricultural Pesticides and Worker Protection 303, 318. However misguided such a policy might be, it magnifies the importance of the discounting issues analyzed in this Article.

n4. See Thomas O. McGarity & Sidney A. Shapiro, OSHA's Critics and Regulatory Reform, 31 Wake Forest L. Rev. 587, 629 (1996) (discussing occupational safety).

n5. Compare Emmett B. Keeler & Shan Cretin, Discounting of Life-Saving and Other Nonmonetary Effects, 29 Mgmt. Sci. 300, 303-05 (1983) (favoring discounting), I. Steven Udvarhelyi et al., Cost-Effectiveness and Cost-Benefit Analyses in the Medical Literature, 116 Annals Internal Med. 238, 239 (1992) (same), and Milton C. Weinstein & William B. Stason, Foundations of Cost-Effectiveness Analysis for Health and Medical Practices, 296 New Engl. J. Med. 716, 719-20 (1977) (same) with Alan L. Hillman & Myoung S. Kim, Economic Decision Making in Healthcare: A Standard Approach to Discounting Health Outcomes, 7 PharmacoEconomics 198, 198 (1995) (rejecting automatic discounting but arguing for "thoughtful adjustments" to reflect period of latency) and Michael Parsonage & Henry Neuburger, Discounting and Health Benefits, 1 Health Econ. 71 (1992) (opposing discounting).

For discussion of different methods for discounting the benefits of medical interventions, see Magnus Johannesson, On the Discounting of Gained Life-Years in Cost- Effectiveness Analysis, 8 Int'l J. Tech. Assessment in Health Care 359 (1992).

n6. See, e.g., U.S. Office of Management and Budget, Regulatory Program of the United States Government, April 1, 1991-March 31, 1992, at 147-48 (1991); Susan W. Putnam & John D. Graham, Chemicals Versus Microbials in Drinking Water: A Decision Sciences Perspective, J. Am. Water Works Ass'n, March 1993, at 57, 60; W. Kip Viscusi, Equivalent Frames of Reference for Judging Risk Regulation Policies, 3 N.Y.U. Envtl. L. J. 431, 436 (1995); infra notes 28-55 (discussing Corrosion Proof Fittings case).

n7. See Michael B. Gerrard, Demons and Angels in Hazardous Waste Regulation: Are Justice, Efficiency, and Democracy Reconcilable?, 92 Nw. L. Rev. 706, 743 (1998) ("[The] protection of future generations is not merely a matter for accountants. The Constitution was adopted in part to 'secure the Blessings of Liberty to ourselves and our Posterity.'"); Lisa Heinzerling, Regulatory Costs of Mythic Proportions, 107 Yale L.J. 1981, 2044 (1998) ("the decision to discount lives saved in the future involves a choice about values, as to which reasonable people may disagree"); A. Dan Tarlock, Now, Think Again About Adaptation, 9 Ariz. J. Int'l & Comp. L. 169, 173 (1992) ("Speculation about discount rates becomes a disguised debate about our ethical duties toward future generations.").

n8. See Gerrard, supra note 7, at 742-43 ("If a human life is considered to be worth $ 8 million, and a ten percent discount rate is chosen, then the present value of saving a life one hundred years from now is only $ 581.... Neither I nor anyone else uses this kind of argument...."); McGarity & Shapiro, supra note 4, at 629 ("The practice of discounting future benefits to present value ... biases cost-benefit analysis against future generations. A high discount rate clearly biases the analysis against future benefits, even though 'it is not clear why the later-born should have to pay interest to induce their predecessors not to exhaust [depletable resources.]'").

n9. The government of the United Kingdom, for example, has rejected the concept of discounting in connection with the health benefits of medical interventions. See Hillman & Kim, supra note 5, at 198.

n10. See What Price Posterity?, Economist, Mar. 23, 1991, at 73.

n11. See John K. Horowitz & Richard T. Carson, Discounting Statistical Lives, 3 J. Risk & Uncertainty 403, 412 n.2 (1990).

n12. See Al Gore, Earth in the Balance: Ecology and the Human Spirit 190-91 (1992). Gore takes a negative view toward discounting:

 

 The accepted formulas of conventional economic analysis contain short-sighted and arguably illogical assumptions about what is valuable in the future as opposed to the present; specifically, the standard 'discount rate' that assesses cost and benefit flows resulting from the use or development of natural resources routinely assumes that all resources belong to the present generation.... In the words of Herman Daly, "There is something fundamentally wrong in treating the earth as if it were a business in liquidation."

 

 Id.

n13. See, e.g., Peter S. Burton, Intertemporal Preferences and Intergenerational Equity Considerations in Optimal Resource Harvesting, 24 J. Envtl. Econ. & Mgmt. 119, 119 (1993) ("Standard discounting practices confuse two issues: (1) intertemporal discount rates of members of the society and (2) intergenerational equity considerations."); Harold P. Green, Legal Aspects of Intergenerational Equity Issues, in Equity Issues in Radioactive Waste Management 189, 192 (Roger E. Kasperson ed. 1983) (noting that most of the statutes governing conservation of land and water resources and wildlife preservation "do not distinguish between benefits accruing in the short-term future to members of the current generation and longer-term benefits to future generations"); Heinzerling, supra note 7, at 2043-56 (not distinguishing the analysis of carcinogenic risks to the current generation and of risks to future generations); Magnus Johannesson & Per-Olov Johansson, The Discounting of Lives Saved in Future Generations: Some Empirical Results, 5 Health Econ. 329, 329 (1996); Putnam & Graham, supra note 6, at 60 (equating delays in the adoption of public health problems with burdens on future generations).

n14. See infra text accompanying notes 28-55 (providing more detailed analysis of the proceedings).

n15. See Exec. Order 12,866, supra note 1, 2(b), 6(b) (responsibilities of OMB's Office of Information and Regulatory Affairs (OIRA)).

n16. See infra text accompanying notes 32-38.

n17. See Corrosion Proof Fittings v. EPA, 947 F.2d 1201, 1218-19 (5th Cir. 1991).

n18. See Heinzerling, supra note 7.

n19. See id. at 1984-85. Heinzerling does not ultimately take a position on the propriety of discounting. See id. at 2055-56 ("More case-by-case attention needs to be given to the question of whether the future benefits of health and environmental regulation should be discounted at all, and if so, at what rate."). In passing, however, she makes arguments that reveal a deep animosity toward discounting. See id. at 2043-54. The legal literature contains one other sustained discussion on the discounting of environmental benefits. See Daniel A. Farber & Paul A. Hemmersbaugh, The Shadow of the Future: Discount Rates, Later Generations, and the Environment, 46 Vand. L. Rev. 267 (1993). The authors urge that, both intra- and intergenerationally, benefits should be discounted at the long-term real rate of return on riskless investments, which they take to be "in the neighborhood of one percent." See id. at 280, 303-04.

n20. See Christopher D. Stone, Beyond Rio: "Insuring" Against Global Warming, 86 Am. J. Int'l L. 445, 476 (1992) ("Any variations in policy that might be implied from defensible attitudes toward risk may well be swamped by the implications of defensible discount rates, and, indeed, of how one resolves the philosophical conundrums of valuing the welfare of future generations."); Tarlock, supra note 7, at 173 ("The selection of the [discount] rate determines the strategy.").

n21. Derek Parfit, Reasons and Persons 357 (1984). For other examples, see Gerrard, supra note 7, at 742-43 ("If a human life is considered to be worth $ 8 million and a ten percent discount rate is chosen, then the present value of saving a life one hundred years from now is only $ 581."); McGarity & Shapiro, supra note 4, at 629 ("At a discount rate of 10%, a dollar's worth of benefits fifty years from now is worth slightly less than a penny today.").

n22. Clifford S. Russell, "Discounting Human Life" (Or, the Anatomy of a Moral- Economic Issue), Resources, Winter 1986, at 8, 8; see Frank S. Arnold, Economic Analysis of Environmental Policy and Regulation 193 (1995) ("When the delay between the present and the time the benefits of a regulatory action are enjoyed is very large, say hundreds of years, using virtually any positive discount rate will render the present value of the benefits almost nil."); Robert C. Lind, Reassessing the Government's Discount Rate Policy in Light of New Theory and Data in an Economy with a High Degree of Capital Mobility, 18 J. Envtl. Econ. & Mgmt. S-8, S-20 (1990). ("The basic arithmetic of exponential growth applied in a cost-benefit analysis implies that, regardless of how small the cost today of preventing an environmental catastrophe that will eventually wipe out the entire economy, it would not be worth this cost to the present generation if the benefits in the future are sufficiently distant.").

n23. See supra text accompanying note 13.

n24. See, e.g., Environmental Policy Under Reagan's Executive Order, supra note 1; McGarity, supra note 1, at 29-59, 174-76, 239-61; Pildes & Sunstein, supra note 1; Sunstein, supra note 2.

n25. See supra text accompanying notes 1-4.

n26. See infra Part I.G.

n27. A similar set of issues arises where current expenditures can prevent future harms to individuals now alive, even though the harm is not a latent disease. The analysis in Part I is therefore relevant to this situation as well.

n28. 947 F.2d 1201 (5th Cir. 1991); see Russell, supra note 22, at 9 (noting that before this proceeding, "'discounting of human lives' had not yet become an issue in the public debate"). For discussion of the case, see Rita L. Wecker, Case Comment: A "Hard Look" at a Soft Analysis, Corrosion Proof Fittings v. Environmental Protection Agency, 4 B.U. Pub. Int. L.J. 145 (1994).

n29. 51 Fed. Reg. 3738 (1986).

n30. See supra text accompanying notes 1-4.

n31. See Letter of Robert P. Bedell, Deputy Administrator, Office of Information and Regulatory Affairs to A. James Barnes, Acting Deputy Administrator, Environmental Protection Agency (March 27, 1985), reprinted in Peter S. Menell & Richard B. Stewart, Environmental Law and Policy 104 (1994).

n32. See id.

n33. See infra text accompanying note 182.

n34. See Letter of Robert P. Bedell, supra note 31, at 104.

n35. Subcomm. on Oversight and Investigations of the House Comm. on Energy and Commerce, EPA's Asbestos Regulations: Report on a Case Study on OMB Interference in Agency Rulemaking, reprinted in Menell & Stewart, supra note 31, at 111. The Barnes comment does not deal specifically with the problem of latent harms, but it reflects a general antipathy to discounting the valuations of human life.

n36. Some members of Congress took a strident position against discounting. For example, Representative Bob Eckhardt noted that "'it was difficult to say whether that kind of approach was more callous or more foolish'" and Representative James Florio called OMB's approach "ghoulish[ ]." See Russell, supra note 22, at 9.

n37. See Subcomm. on Oversight and Investigations, supra note 35, reprinted in Menell & Stewart, supra note 31, at 109.

n38. See id.

n39. See id. at 110; Sidney A. Shapiro & Thomas O. McGarity, Not So Paradoxical: The Rationale for Technology-Based Regulation, 1991 Duke L.J. 729, 735 ("In cases of toxic substance exposure, where the onset of disease can be delayed by as much as thirty years, [discounting] effectively ignores the risk altogether.").

n40. Subcomm. on Oversight and Investigations, supra note 35, reprinted in Menell & Stewart, supra note 31, at 111.

n41. Id.

n42. 51 Fed. Reg. 3738, 3757-59 (1986).

n43. See id. at 3748; 54 Fed. Reg. 29,460, 29,487 (1989).

n44. 54 Fed. Reg. 29,460, 29,483 (1989).

n45. See id. at 29,485.

n46. See id.

n47. Id. at 29,487.

n48. Id.

n49. Id.

n50. 947 F.2d 1201 (5th Cir. 1991).

n51. Id. at 1218. Lisa Heinzerling criticizes the Fifth Circuit's position: "One worries about 'preserving an apples-to-apples comparison,' however, only if one is dealing only with apples. In the asbestos case, the costs were dollars and the benefits were lives. These costs and benefits are the same only if dollars and lives are the same." Heinzerling, supra note 7, at 2053. Both positions overlook an aspect of the problem. The Fifth Circuit misses the fact that the intertemporal choices of individuals do not necessarily reflect discounting at the rates used by financial markets (though in fact empirical studies show no statistically significant differences). See infra Part I.F.1. In turn, Heinzerling's rhetorical device fails to acknowledge that the cost-benefit calculus in the case required the valuation of the life, and that the question whether this amount should be discounted is one that depends on how individuals compare the utilities derived from living in the present to the utilities derived from living in the future. See infra text accompanying notes 223-224.

n52. See Corrosion Proof Fittings, 947 F.2d at 1218-19, 1229-30. The court's analysis revealed confusion. It relied primarily on the following example:

 

 Suppose two workers will be exposed to asbestos in 1995, with worker X subjected to a tiny amount of asbestos that will have no adverse health effects, and worker Y exposed to massive amounts of asbestos that quickly will lead to an asbestos- related disease. Under the EPA's approach, which takes into account only the time of the exposure rather than the time at which any injury manifests itself, both examples would be treated the same.

 

 Id. at 1218. In fact, if worker X would never get cancer, the regulation would have no benefit with respect to this worker. With zero benefits, there would be nothing to discount. What the court might have meant is that if workers X and Y had been exposed to asbestos at the same time, and worker Y was injured before worker X, the EPA would treat both cases in the same way (and presumably the Fifth Circuit would have wanted to treat them differently).

n53. Id.

n54. For related discussion, see infra text accompanying notes 155-157.

n55. See Corrosion Proof Fittings, 947 F.2d at 1218 n.19. For further discussion of discount rates, see infra Part I.F.2.

n56. S. 343, 104th Cong. (1995).

n57. See id. at 623 ("[no] final rule ... shall be promulgated unless the agency finds that ... the potential benefits from the rule ... justify the potential costs of the rule"); id. at 621-622 (dealing with the preparation of cost-benefit analyses); see generally supra text accompanying notes 1-4 (discussing regulatory reform).

n58. S. Rep. No. 104-90, at 153 (1995) (supplemental views of Senator Leahy).

n59. The only two sustained treatments of the question of discounting in the legal academic literature were those of Farber & Hemmersbaugh, supra note 19, and Heinzerling, supra note 7. See supra note 19 (discussing their positions). While the economics literature has focused on isolated nuances, it has not taken a broad look at the problem or connected the various strands that are necessary to a sophisticated analysis of the public policy choices.

n60. See W. Kip Viscusi, The Valuation of Risks to Life and Health: Guidelines for Policy Analysis, in Benefits Assessment: The State of the Art 193, 193 (Judith D. Bentkover et al. eds., 1986) [hereinafter Viscusi, Valuation]. For a more recent survey, see W. Kip Viscusi, The Value of Risks to Life and Health, 31 J. Econ. Literature 1912 (1993) [hereinafter Viscusi, Value]. The technique is generally traced to Thomas C. Schelling, The Life You Save May Be Your Own, in Problems in Public Expenditure Analysis 127 (Samuel B. Chase, Jr. ed., 1968), and E.J. Mishan, Evaluation of Life and Limb: A Theoretical Approach, 79 J. Pol. Econ. 687, 695-705 (1971).

Before the ascendancy of willingness-to-pay studies, the human capital approach was prevalent. This approach valued life in terms of lost earnings. See Viscusi, Valuation, supra, at 198. The technique is subject to the obvious criticism that earnings provide that "individual well-being goes far beyond its financial implications." Id.; accord W.B. Arthur, The Economics of Risks to Life, 71 Am. Econ. Rev. 54, 54 (1981); Lewis A. Kornhauser, The Value of Life, 38 Clev. St. L. Rev. 209, 212 (1990).

n61. See Viscusi, Valuation, supra note 60, at 200.

n62. See id. at 199-200.

n63. Such workers might also face a higher probability of nonfatal risks. Some studies estimate the portion of the wage differential that is attributable to such non-fatal risks. The residual wage differential is then attributed to fatal risks. See Viscusi, Value, supra note 60, at 1919. Some studies, however, do not separate the wage differential into these two components. See id.

n64. For criticism of the approach, see McGarity, supra note 1, at 147-48; Steven Kelman, Cost-Benefit Analysis and Environmental, Safety, and Health Regulation: Ethical and Philosophical Considerations, in Cost-Benefit Analysis and Environmental Regulations: Politics, Ethics, and Methods 137, 143-45 (Daniel Swartzman et al. eds., 1982); J. Paul Leigh, Compensating Wages, Value of a Statistical Life, and Inter-industry Differentials, 28 J. Envtl. Econ. & Mgmt. 83, 94-95 (1995); McGarity & Shapiro, supra note 4, at 628-29.

An alternative methodology consists of surveying individuals and asking them how much they would be willing to pay for a particular risk reduction. See Viscusi, Valuation, supra note 60, at 204-05. The disadvantage of this contingent valuation method is that the responses are to hypothetical situations and have no economic consequences. See V. Kerry Smith & William H. Desvousges, An Empirical Analysis of the Economic Value of Risk Changes, 95 J. Pol. Econ. 89, 93-94 (1987).

n65. See Maureen L. Cropper & Frances G. Sussman, Valuing Future Risks to Life, 19 J. Envtl. Econ. & Mgmt. 160, 160 (1990) ("The empirical literature on valuing risks to life has focused almost exclusively on valuing mortality risks that occur today - the risk of accidental death a worker faces during the coming year or the risk of dying this month in an auto accident."); Horowitz & Carson, supra note 11, at 405 ("Virtually all the empirical work on the value of risk reductions has considered risks that occur entirely in the present...."); Shapiro & McGarity, supra note 39, at 734 ("most wage premium studies ... are based on safety hazards, not health risks"). Of course, to the extent that there is a probability of a non-fatal accident, the resulting morbidity risk could also be measured using a willingness-to-pay approach.

n66. See Leigh, supra note 64, at 86-87; Viscusi, Valuation, supra note 60, at 200. Of course, in some cases, industrial accidents result in long-term disability rather than death.

n67. One ongoing attempt to derive a willingness-to-pay valuation of human lives threatened by carcinogens is reflected in John R. Lott, Jr. & Richard L. Manning, Have Changing Liability Rules Compensated Workers Twice for Occupational Hazards?: Earnings Premiums and Cancer Risks (June 28, 1998) (manuscript on file with the Columbia Law Review). For a contingent valuation study inquiring how individuals value risk reductions from hazardous waste sites, see Smith & Desvousges, supra note 64.

n68. Both the Occupational Safety and Health Administration (OSHA) and EPA were established in 1970. See Sidney A. Shapiro & Thomas O. McGarity, Reorienting OSHA: Regulatory Alternatives and Legislative Reform, 6 Yale J. on Reg. 1, 1 n.1, 2 n.9 (1989).

n69. See Cropper & Sussman, supra note 65, at 166 n.8. Moreover, certain risks may be poorly understood even by experts. See Smith & Desvousges, supra note 64, at 108-09.

n70. See Sherwin Rosen, The Quantity and Quality of Life: A Conceptual Framework, in George Tolley et al., Valuing Health for Policy: An Economic Approach 221 (1994).

n71. One commentator estimates that "the average age of the workplace accident fatality is about 41" whereas "the average age of the workplace cancer victim is likely to be 55, 65, or even higher." John M. Mendeloff, The Dilemma of Toxic Substance Regulation: How Overregulation Causes Underregulation at OSHA 48 (1988).

n72. Additional complications are introduced when the length of the person's life is uncertain. See Rosen, supra note 70, at 236-45. No important insights are lost, however, as a result of this simplification. In practice, of course, an individual who would have died of cancer at the end of the latency period may die earlier of other causes. See Lester B. Lave, The Strategy of Social Regulation: Decision Frameworks for Policy 43 (1981).

n73. See Maureen L. Cropper & Paul R. Portney, Discounting and the Evaluation of Lifesaving Programs, 3 J. Risk & Uncertainty 369, 376 (1990).

n74. A more complicated situation arises when an individual is exposed to a carcinogen over a long period of time and the harm resulting from the exposure is cumulative.

n75. See Cropper & Sussman, supra note 65, at 172-73.

n76. See W. Kip Viscusi, Discounting Health Effects for Medical Decisions, in Valuing Health Care: Costs, Benefits, and Effectiveness of Pharmaceuticals and Other Medical Technologies 125, 129 (Frank A. Sloan ed., 1995). In contrast, a nominal rate is used to discount current dollars. The real rate is the nominal rate minus the rate of inflation.

n77. See Edith Stokey & Richard Zeckhauser, A Primer for Policy Analysis 161-65 (1978).

n78. See infra Part I.F.1.

n79. See Cropper & Sussman, supra note 65, at 165-66.

n80. See supra text accompanying note 71 (hypothesizing that the worker exposed to the risk of instantaneous death is forty-years old).

n81. See Cropper & Portney, supra note 73, at 378 n.12.

n82. See Cropper & Sussman, supra note 65, at 172 ("This fact ... is often ignored in risk-benefit analyses.").

n83. See infra Part I.G.

n84. See Robert F. Bordley, Making Social Trade-Offs Among Lives, Disabilities, and Cost, 9 J. Risk & Uncertainty 135, 138 (1994).

n85. See Cropper & Portney, supra note 73, at 371-72; Rosen, supra note 70, at 222-23.

n86. A similar issue arises in the literature on QALYs, or quality-adjusted life years, which are a means for adjusting the utility that an individual gets in a period by the quality of her health in that period. So, for example, an individual derives greater utility from a year in which her health is excellent than in one in which she is disabled. See Richard Zeckhauser & Donald Shepard, Where Now for Saving Lives?, Law & Contemp. Probs., Autumn 1976, at 5, 12-13. In the context of QALYs, separability implies that the utility that a person derives from the quality of her life in a particular year is independent of the qualities of her life in past years. See John Broome, QALYs, 50 J. Pub. Econ. 149, 151-52 (1993).

n87. Broome, supra note 86, at 151-52. Broome applies this label to a separability model in the context of QALYs. See supra note 86.

n88. See Bordley, supra note 84, at 138.

n89. See infra Part I.F.1.

n90. See infra Part I.E.3.

n91. See supra text accompanying notes 65-69.

n92. See Bordley, supra note 84, at 138; Michael J. Moore & W. Kip Viscusi, Discounting Environmental Health Risks: New Evidence and Policy Implications, 18 J. Envtl. Econ. & Mgmt. S-51, S-54 (1990); Rosen, supra note 70, at 224.

n93. Donald S. Shepard & Richard J. Zeckhauser, Survival Versus Consumption, 30 Mgmt. Sci. 423, 424 (1984).

n94. Id. at 424; see also Joseph Lipscomb, Time Preference for Health in Cost- Effectiveness Analysis, 27 Med. Care S233, S237 (1989) (asking whether individuals evaluate multiperiod health outcomes "in accordance with constant-rate discounting").

n95. See W. Kip Viscusi & Michael J. Moore, Rates of Time Preference and Valuations of the Duration of Life, 38 J. Pub. Econ. 297, 297-98 (1989) ("Although money is readily transferable across time, health status is not."). Part I.F.1, infra, explains more generally why discounting health risks is analytically different from discounting financial flows.

n96. There have been attempts to estimate the rate at which individuals discount their utilities, but they have been conducted on the basis of constant discounting models. See Moore & Viscusi, supra note 92, at S-54. There also are empirical estimates of how discount rates depend on the period over which the discounting is performed, but these studies are intergenerational, or at the very least interpersonal. See infra Part II.B.

n97. See Donald A. Redelmeier & Daniel N. Heller, Time Preference in Medical Decision Making and Cost-Effectiveness Analysis, 13 Med. Decision Making 212, 216 (1993); id. at 214-15 (finding that rates for temporally proximate events were larger than for more distant events); infra Part II.B (same finding in intergenerational models).

n98. See supra text accompanying notes 36-38.

n99. See Cropper & Portney, supra note 73, at 377.

n100. See Shepard & Zeckhauser, supra note 93, at 437 n.18; Viscusi, supra note 76, at 130. But see Glenn Blomquist, Value of Life Saving: Implications of Consumption Activity, 87 J. Pol. Econ. 540, 555 (1979) (finding lower elasticity).

n101. See supra text accompanying notes 78-80.

n102. See Viscusi, supra note 76, at 130; Richard Zeckhauser, Procedures for Valuing Lives, 23 Pub. Pol'y 419, 437 (1975).

n103. See William D. Nordhaus, To Slow or Not to Slow: The Economics of the Greenhouse Effect, 101 Econ. J. 920, 925-26 (1991); Viscusi, supra note 76, at 130.

n104. Farber & Hemmersbaugh, supra note 19, state that "the discount rate even for economic benefits cannot significantly exceed the expected long-term rate of economic growth; otherwise, we would discount even the destruction of most future Gross Domestic Product to a low present value over periods of only decades." Id. at 296. The authors appear to be making a pragmatic argument for keeping the effective discount rate low. There is, however, no plausible normative argument for linking the two rates in this manner.

n105. U.S. Census Bureau, Historical Income Tables - Persons, Table P-44 (visited June 22, 1998) <http://www.census.gov/hhes/income/histinc/p44.html>.

n106. Over the longer run, the rate has been higher. See William R. Cline, The Economics of Global Warming 251 (1992) (estimating that "real per capita income in the United States has grown at about 1.7 percent annually over the past century").

n107. See Donald S. Shepard & Richard J. Zeckhauser, Life-Cycle Consumption and Willingness to Pay for Increased Survival, in The Value of Life and Safety 95, 120-27 (M.W. Jones-Lee ed., 1982) [hereinafter Shepard & Zeckhauser, Life-Cycle Consumption]; Shepard & Zeckhauser, supra note 93, at 432-36.

n108. See Zeckhauser, supra note 102, at 437.

n109. Id. at 438.

n110. In general, one's credit suitability for loans is evaluated on the basis of one's present income. There are some exceptions, however, such as student loans to finance post-secondary education.

n111. See Shepard & Zeckhauser, Life-Cycle Consumption, supra note 107, at 107-15. There is potentially a logical inconsistency in believing that individuals cannot process the fact that they will have higher incomes in the future in order to value their lives accordingly, but positing that individuals will borrow money in the expectation of higher income in the future.

n112. See id. at 125.

n113. See id.

n114. See id. at 121.

n115. See Shepard & Zeckhauser, supra note 93, at 434.

n116. See id. at 435 (noting that "the real world lies somewhere in between" the two models).

n117. See supra text accompanying note 114.

n118. See Shepard & Zeckhauser, supra note 93, at 433.

n119. See Viscusi, Value, supra note 60, at 1942-43 ("the population of exposed workers ... generally have lower incomes than the individuals being protected by broadly based risk regulation").

n120. EPA should, however, vary its valuations of life on the basis of the age profile of the affected population, to account for the different numbers of life-years at stake in various regulatory programs.

n121. For discussion of environmental justice, see Vicki Been, Coming to the Nuisance or Going to the Barrios? A Longitudinal Analysis of Environmental Justice Claims, 24 Ecology L.Q. 1 (1997); Vicki Been, Locally Undesirable Land Uses in Minority Neighborhoods: Disproportionate Siting or Market Dynamics?, 103 Yale L.J. 1383 (1994); Robert D. Bullard, Anatomy of Environmental Racism and the Environmental Justice Movement, in Confronting Environmental Racism: Voices from the Grassroots 15 (Robert D. Bullard ed., 1993); Richard J. Lazarus, Pursuing "Environmental Justice": The Distributional Effects of Environmental Protection, 87 Nw. U. L. Rev. 787 (1993).

n122. An ethical objection to such particularization would be an attack on cost-benefit analysis in general and to the use of a willingness-to-pay methodology for valuing lives in particular. See Guido Calabresi & Philip Bobbitt, Tragic Choices: The Conflicts Society Confronts in the Allocation of Tragically Scarce Resources 32 (1978) (referring to "the external costs - moralisms and the affront to values, for example - of market determinations that say or imply that the value of a life or of some precious activity integral to life is reducible to a money figure"). Nonetheless, using differential valuations of life based on income levels is likely to prove objectionable to some supporters of cost-benefit analysis, and to magnify the objections adduced by opponents of this approach.

n123. See U.S. Census Bureau, Historical Income Tables - Persons: Table P-36: Occupation of Longest Job - Workers (Both Sexes Combined) by Median and Mean Earnings (visited June 22, 1998) <http://www.census.gov/hhes/income/ histinc/p36.html>.

n124. See id.

n125. See id.

n126. See Paul Slovic, Perception of Risk, 236 Science 280 (1987).

n127. Zeckhauser, supra note 102, at 445 n.27.

n128. Cass Sunstein cogently explains that "the question whether a risk is run voluntarily or not is often not a categorical one but instead a matter of degree." Cass R. Sunstein, Bad Deaths, 14 J. Risk & Uncertainty 259, 272 (1997). Sunstein would place risks on a voluntariness/involuntariness continuum based on three factors: whether the worker has adequate information about the risk; whether the worker is compensated for the risk; and whether the compensation package does not appear unfair, even if voluntarily chosen by the parties, as a result of background inequality between the employer and employee. See id.; see also Shapiro & McGarity, supra note 39, at 734 ("Unfortunately, low-paid workers in hazardous industries where there are no (or weak) unions may act more out of desperation than choice.").

n129. See Maureen L. Cropper & Uma Subramanian, Public Choice Between Lifesaving Programs 6 (World Bank Policy Research Working Paper 1497, 1995). Of course, if an individual is exposed to a toxic air pollutant, she could move somewhere else. Sunstein would nonetheless classify the risk as involuntary because the individuals are not in a contractual relationship with the producer of the risk and cannot avoid the risk except at great cost, in this case by moving to another area. See Sunstein, supra note 128, at 271.

Moreover, in many cases, individuals may lack sufficient information about environmental risks to make informed choices. Even if they had such information, risks that are uniformly distributed throughout the country could obviously not be avoided by moving elsewhere. For further discussion of the difference between voluntary and involuntary risks, see Richard H. Pildes & Cass R. Sunstein, Democrats and Technocrats, Journees d'Etudes Juridiques Jean Dabin (forthcoming 2000) (manuscript on file with the Columbia Law Review).

n130. See supra text accompanying notes 60-62.

n131. Even studies of how the price of a house in an area with high concentrations of this pollutant compares to the price of an otherwise similar house in an area with better air quality do not capture the value of involuntary risk. While such hedonic price studies are a commonly used revealed preference tool for economic valuations, see Ronald G. Cummings et al., General Methods for Benefits Assessment, in Benefits Assessment, supra note 60, at 171-76, the participants in these housing markets are individuals attempting to decide where to live. They are making a choice about whether to live in one area rather than another. As a result, it would be a stretch to regard their "choice" as involuntary. Rather, the involuntary label is better used for individuals who have lived in an area for a long time, have strong personal ties to the area, and lack the resources to move.

n132. An extensive list of such references is provided in Heinzerling, supra note 7, at 1983 n.1, 2. The genesis for these studies is a table prepared in the 1980s by John Morrall, an OMB official. See John F. Morrall III, A Review of the Record, Regulation, Nov./Dec. 1986, at 25, 30 tbl.4. Heinzerling notes, however, that the regulations with numbers at the high end were never promulgated. Moreover, she argues that the remaining differences would be less stark if Morrall had not discounted the benefits of environmental regulation or reduced the estimates of risk prepared by the agencies. See Heinzerling, supra note 7, at 1984-85.

n133. There has been strong criticism to valuations based on survey responses. See Richard B. Stewart, Liability for Natural Resource Injury: Beyond Tort, in Analyzing Superfund: Economics, Science, and Law 219, 234-38 (Richard L. Revesz & Richard B. Stewart eds., 1995). Nonetheless, a panel of distinguished economists, co-chaired by Nobel Prize winners Kenneth Arrow and Robert Solow, which had been empaneled by the National Oceanic and Atmospheric Administration (NOAA), gave qualified endorsement to the use of contingent valuation techniques. See 58 Fed. Reg. 4601, 4610 (1993). Clearly, revealed preference valuations would be preferable, but, as indicated above, such valuations cannot be used for involuntary harms. See supra text accompanying notes 130-131.

n134. See Cropper & Subramanian, supra note 129, at 2.

n135. See id. at 16-18.

n136. See id. at 3-7.

n137. The remaining characteristics were the extent to which the affected population was to blame for the risk, the seriousness of the risk, and whether the risks affected respondents personally. In addition to these four risk characteristics, the respondents were also asked to assess four program characteristics: the efficacy of the program, the appropriateness of government intervention, the fairness of the funding mechanism, and the time before the program begins to save lives. See id. at 39.

n138. See id. at 40.

n139. A labeling program, designating food to be free of pesticide, could work effectively if the claims were in fact truthful and adequate information was conveyed to prospective buyers. But social coordination would be necessary to set up the labeling program and to police its integrity.

n140. See Cropper & Subramanian, supra note 129, at 40.

n141. See id. at 41.

n142. See id. at 48.

n143. See id. at 24, 41.

n144. See id. at 4-5.

n145. See McGarity, supra note 1, at 146-49; Kelman, supra note 64, at 144; Viscusi, Value, supra note 60, at 1928.

This effect is discussed even though it has not been the focus of empirical study, see supra text accompanying notes 99-100, because it flows in part from the difference between the voluntary nature of workplace harms and the involuntary nature of environmental harms.

n146. Some self-selection can take place with respect to reasonably local risks, such as those that result from proximity to hazardous waste sites. With respect to more regional risks, such as regional air pollution, however, such self-selection is far more difficult.

n147. See Sunstein, supra note 128, at 259.

n148. See Tammy O. Teng et al., Five-Hundred Life-Saving Interventions and Their Cost-Effectiveness, 15 Risk Analysis 369 (1995).

n149. See id. at 370.

n150. See id. at 371.

n151. See id.

n152. See supra text accompanying notes 131-133.

n153. George Tolley et al., State-of-the-Art Health Values, in Tolley et al., supra note 70, at 323, 339-44.

n154. See id. at 339-40.

n155. See supra text accompanying notes 64-68.

n156. But cf. Sunstein, supra note 128, at 269 (an extended period before death can contain benefits, since it allows grief and adjustment).

n157. See Tolley et al., supra note 153, at 329-32, 340; supra note 133 and accompanying text.

n158. See Tolley et al., supra note 153, at 340.

n159. See id. at 340-41; see also Michael W. Jones-Lee et al., The Value of Safety: Results of a National Sample Survey, Econ. J., March 1985, at 49, 58-60. For a more recent study finding a higher willingness-to-pay to avoid carcinogenic harms, see Ian Savage, An Empirical Investigation into the Effect of Psychological Perceptions on the Willingness-to- Pay to Reduce Risk, 6 J. Risk & Uncertainty 75, 77, 85 (1993).

n160. For intuitions supporting a higher valuation for dreaded harms, see Mendeloff, supra note 71, at 48; Shapiro & McGarity, supra note 39, at 734 n.29.

n161. See Lave, supra note 72, at 44 ("Discounting future health effects at the standard rate makes sense only if there is a fixed transformation rate between dollars and health."); John Mendeloff, Measuring Elusive Benefits: On the Value of Health, 8 J. Health Pol., Pol'y & Law 554, 568 (1983) ("discount rate for health effects should largely be based upon individuals' time preferences"); supra note 51 and accompanying text; infra Part I.F.1. But see Victor R. Fuchs & Richard Zeckhauser, Valuing Health - A "Priceless" Commodity, 77 Am. Econ. Rev. 263, 264 (1987) (suggesting that life years should be discounted in the same manner as cash flows).

n162. See Farber & Hemmersbaugh, supra note 19, at 287.

n163. Viscusi, supra note 76, at 131-32.

n164. See John A. Cairns, Valuing Future Benefits, 3 Health Econ. 221, 221 (1994) ("Little is known about individual time preferences with respect to future health, and in particular whether they differ from preferences with respect to future wealth."); Putnam & Graham, supra note 6, at 60 ("Instead of choosing a standard discount rate ... the rate should be based on the ... preferences of citizens.").

n165. See Moore & Viscusi, supra note 92, at S-61 ("One should also be cognizant of the ultimate objective of our study, which is to ascertain whether systematic differences exist between rates of time preference for health and financial rates of return.").

n166. See id. at S-52-S-55.

n167. See id. at S-53.

n168. See id. at S-57. These studies follow a revealed preference approach, which consists of observing the prices at which market transactions take place. See supra text accompanying notes 130-131.

n169. See Moore & Viscusi, supra note 92, at S-59, S-61.

n170. Id. at S-59; see also supra text accompanying note 55; supra note 76 (discussing difference between real and nominal rates).

n171. Moore & Viscusi, supra note 92, at S-61; see also id. at S-52.

It is worth thinking about how the regulatory system ought to react if, contrary to the findings by Moore and Viscusi, one found that individuals discounted health risks at a very high rate, even when they were well informed about these risks. In such situations, it might be appropriate for the government to act in a paternalistic fashion and make social policy on the basis of a lower discount rate. The rationale would be somewhat analogous to the rationale for the usury laws, which prohibit lending at an overly high interest rate.

The utility of an individual with an unusually high discount rate would increase if she were allowed to borrow at a rate up to her discount rate in order to transfer consumption from the future to the present. The usury laws, however, prevent her from doing so because of concern that she might later experience excessive regret. Similarly, in deciding how stringently to regulate future environmental risks, the government could be skeptical of discount rates for health risks that are high compared to the rates at which money gets transferred through the financial markets.

Empirical findings of high discount rates would at the very least be troubling and raise difficult questions as to how social policymakers should react. The Moore and Viscusi studies, showing an equivalence between the rates at which individuals discount health risks and the rates at which the market discounts flows of money, make it unnecessary to face this issue.

n172. See id. at S-61. The earlier studies are Michael J. Moore & W. Kip Viscusi, Models for Estimating Discount Rates for Long-Term Health Risks Using Labor Market Data, 3 J. Risk & Uncertainty 381 (1990); Michael J. Moore & W. Kip Viscusi, The Quantity- Adjusted Value of Life, 26 Econ. Inquiry 369 (1988); Viscusi & Moore, supra note 95.

n173. See Moore & Viscusi, supra note 92, at S-61.

n174. Id.

n175. See Viscusi & Moore, supra note 95, at 314.

n176. The issue is not entirely free of doubt. For example, a more recent study by Viscusi and a different co-author, using a similar methodology, found real discount rates ranging from 11-17%, in the context of automobile safety. See Mark K. Dreyfus & W. Kip Viscusi, Rates of Time Preference and Consumer Valuations of Automobile Safety and Fuel Efficiency, 38 J.L. & Econ. 79, 84, 99 (1995). The authors note that the riskless rate of interest, which they estimate in the 2-5% range, is outside the confidence limit of their estimates. See id. at 99. They note, however, that in many cases consumers face interest rates that are far higher than the riskless rate, and that their estimated discount rate was not statistically different, at a 95% confidence interval, from the real rates for the financing of automobile purchases (8.5% and 11.0% for new and used cars, respectively). See id. at 99-100.

Individuals also exhibit inordinately high discount rates with respect to purchases having an effect on energy conservation. Thus, they have not been willing to pay much of a premium on the purchase of products such as air conditioning or heating units in return for lower energy costs in the future. See Jeffrey A. Dubin, Will Mandatory Conservation Promote Energy Efficiency in the Selection of Household Appliance Stocks?, 7 Energy J. 99, 109-13 (1986); Jerry A. Hausman, Individual Discount Rates and the Purchase and Utilization of Energy-Using Durables, 10 Bell J. Econ. 33, 50-52 (1979); Douglas A. Houston, Implicit Discount Rates and the Purchase of Untried, Energy-Saving Durable Goods, 10 J. Consumer Res. 236, 236-37 (1983).

These studies, which are discussed in Dreyfus & Viscusi, supra, at 83-84, affect only financial flows and do not raise the question of how to discount future health risks. The problem here may well be that consumers lack clear information on energy savings benefits or cannot properly process this information if they have it, see Wesley A. Magat & W. Kip Viscusi, Informational Approaches to Regulation 5 (1992), or that they violate some of the postulates of rational theory, see George Loewenstein & Richard H. Thaler, Intertemporal Choice, 3 J. Econ. Persp. 181, 182-83, 192 (1989).

n177. See supra text accompanying note 167.

n178. See Rosen, supra note 70, at 224; supra text accompanying notes 99-100.

n179. In fact, the situation may be even more complicated. Children, for example, may increase one's utility. See Richard A. Epstein, Justice Across Generations, 67 Tex. L. Rev. 1465, 1472 (1989). Then, for a given level of consumption, after one has children one's utility might be higher than before.

n180. See supra text accompanying notes 36-38.

n181. See Circular No. A-94, 57 Fed. Reg. 53,519 (1992).

n182. See Robert C. Lind, Discounting for Time and Risk in Energy Policy 5-6 (1982). For criticisms, see Daniel A. Farber, Risk Regulation in Perspective: Reserve Mining Revisited, 21 Envtl. L. 1321, 1349-50 (1991); Farber & Hemmersbaugh, supra note 19, at 278 & n.43; Viscusi, supra note 76, at 129.

n183. See 57 Fed. Reg. at 53,522-23.

n184. Id. at 53,523.

n185. See id. at 53,520, 53,523.

n186. See id. at 53,528 (3.8%); 61 Fed. Reg. 6397, 6397 (1996) (3.0%); 63 Fed. Reg. 3932, 3933 (1998) (3.8%).

n187. For clear analyses, see Arnold, supra note 22, at 177-97; Lind, supra note 22. For an excellent primer on discounting, see Lind, supra note 182, at 21-94.

n188. Arnold, supra note 22, at 180.

n189. See id. at 181.

n190. Because income taxes are due on nominal interest, the tax adjustment must be performed first. See id. at 192 n.10.

n191. See id. at 192.

n192. See id. at 192; Viscusi, supra note 76, at 129, 134.

In 1998, the yield on 30-year Treasury bonds stood at 5.57%, the lowest since auctions on these bonds began in 1977. See Guy Dixon & Candace Cumberbatch, Bond Price Hit New Highs, Lifted by Concerns About Japan and Signals of a U.S. Slowdown, Wall St. J., July 7, 1998, at C19. An individual facing a 28% federal marginal tax rate would have an after-tax return of 4.0%. Subtracting the change in the consumer price index for the twelve-month period ending in May 1998, which is 1.7%, see U.S. Bureau of Labor Statistics, Consumer Price Index Summary (visited July 8, 1998) <http://stats.bls.gov/news.release/cpi.nws.html>, would result in a discount rate of 2.3%.

n193. See Arnold, supra note 22, at 181.

n194. See id. at 184-85.

n195. See id. at 190.

n196. See supra text accompanying notes 183-184.

n197. In the case of environmental regulation, the government is not making the investment, but is instead requiring private parties to make it. The same analysis is applicable, however. See Arnold, supra note 22, at 189-91.

n198. See id. at 180-84; Lind, supra note 22, at S-10, S-11.

The Department of Energy continues to engage in this inquiry:

 

Because the proposed appliance efficiency standards will primarily affect private, rather than public, investment, the Department continues to believe that using the average real rate of return on private investment as the basis for the social discount rate is most appropriate. If the primary impact of the standards were on Federal or other public expenditures, DOE agrees that real interest rates on long term government securities would likely be a better basis.

 

 60 Fed. Reg. 37,388, 37,394 (1995).

n199. See Arnold, supra note 22, at 184-85; Lind, supra note 22, at S-8, S-9.

n200. See Arnold, supra note 22, at 184-85, 190-91; Lind, supra note 22, at S-8, S-9.

n201. See Ronald G. Cummings, Legal and Administrative Uses of Economic Paradigms: A Critique, 31 Nat. Resources J. 463, 471 (1991); Randolph M. Lyon, Federal Discount Rate Policy, The Shadow Price of Capital, and Challenges for Reforms, 18 J. Envtl. Econ. & Mgmt. S-29, S-30 (1990). For an interesting survey of the different choices of discount rates in federal agencies, see Edward R. Morrison, Comment, Judicial Review of Discount Rates Used in Regulatory Cost-Benefit Analysis, 65 U. Chi. L. Rev. 1333, 1336-37, 1364-69 (1998).

n202. 59 Fed. Reg. 45,872, 45,895-97 (1994).

n203. See 43 C.F.R. 11.84(e)(2) (1998) (Department of the Interior).  Ohio v. Department of the Interior, 880 F.2d 432, 464-65 (D.C. Cir. 1989), upheld the Department of the Interior's choice of a 10% discount rate for natural resources damages, following OMB's pre-1992 policy, see supra text accompanying note 182.

n204. See supra text accompanying notes 73-75.

n205. See supra text accompanying notes 133-143.

n206. See supra text accompanying notes 153-160.

n207. See supra text accompanying note 156.

n208. See Jones-Lee et al., supra note 159, at 55-57.

n209. In contrast, in the Cropper and Subramanian study, the respondents were asked to evaluate the ease with which each of the risks could be avoided. See supra text accompanying notes 136-138.

n210. The upward adjustment resulting from the unrepresentativeness of the risk preferences of the population exposed to workplace risks cannot be estimated as a result of the paucity of the empirical data, though logic compels the conclusion that such workers will have a lower-than-average willingness-to-pay to avoid risk. See supra Part I.E.2.b.

n211. See supra text accompanying notes 123-125.

n212. See supra text accompanying notes 103-106.

n213. For a twenty year lag, a discount rate of 2% reduces the valuation to 67% of the undiscounted amount, as compared to a reduction to 55% of the undiscounted amount for a 3% discount rate.

n214. See supra text accompanying notes 183-184.

n215. See supra text accompanying notes 80-83.

n216. The OMB approach, however, avoids the pitfall of using V[in'60,60'] as the basis for estimating V[in'40,40']. Such a procedure might lead to undervaluation because of changes over time in the income and saving levels of individuals. See supra Part I.E.1.b.

n217. The adjustments for the dread nature of the harm, the involuntary nature of the harm, the salary differential, and the impact of economic growth are 2, 2, 1.23, and 1.22, respectively. See supra text accompanying notes 204-213. The calculation assumes that all the factors are multiplicative. See supra text accompanying notes 206-210. This assumption should be the focus of empirical study.

n218. See B.T. Westerfield, Asbestos-Related Lung Disease, 85 Southern Med. J. 616 (1992). Some of the adverse consequences of exposure to asbestos have latency periods of 30 and 40 years. See id. at 618.

n219. See supra Part I.E.2.b.

n220. See supra text accompanying notes 177-179.

n221. For discussion of the differences with the intergenerational setting, see infra text accompanying notes 281-283.

n222. See supra text accompanying notes 65-69.

n223. See supra text accompanying note 35 (discussing Barnes's testimony).

n224. See supra Part I.F.1.

n225. See supra Part I.G.

n226. See supra text accompanying notes 214-218.

n227. For applications of this concept in the legal literature, see Bruce Ackerman & Anne Alstott, The Stakeholder Society (forthcoming 1999) (manuscript at 131-42, on file with the Columbia Law Review); Christine Jolls, Contracts as Bilateral Commitments: A New Perspective on Contract Modification, 26 J. Legal Stud. 203, 210, 219-24 (1997); Christine Jolls et al., A Behavioral Approach to Law and Economics, 50 Stan. L. Rev. 1471, 1538-41 (1998); Deborah M. Weiss, Paternalistic Pension Policy: Psychological Evidence and Economic Theory, 58 U. Chi. L. Rev. 1275, 1285-86, 1300-06 (1991).

n228. Intergenerationally, the situation is different because the individual making the decision is different from the individual affected by the decision. See infra text accompanying notes 281-283.

n229. Ackerman & Alstott, supra note 227, at 141.

n230. See supra note 192.

n231. For critiques of cost-benefit analysis, see Steven Kelman, Cost-Benefit Analysis: An Ethical Critique, Regulation, Jan./Feb. 1981, at 33; Duncan Kennedy, Cost-Benefit Analysis of Entitlement Problems: A Critique, 33 Stan. L. Rev. 387 (1981). For critiques of the techniques for valuing human lives, see sources cited supra note 64.

n232. See supra text accompanying notes 21-22.

n233. See William D. Nordhaus, Managing the Global Commons: The Economics of Climate Change 4 (1994) ("A complete analysis of the economics of climate change must recognize the extraordinarily long time lags involved in the reaction of the climate and economy to greenhouse gas emissions.").

n234. For a comprehensive list, see 1 Philippe Sands, Principles of International Environmental Law 198-213 (1995).

n235. Stockholm Declaration of the United Nations Conference on the Human Environment, June 16, 1972, 11 I.L.M. 1461.

n236. United Nations Conference on Environment and Development: Rio Declaration on Environment and Development, June 13, 1992, 31 I.L.M. 874.

n237. United Nations Conference on Environment and Development: Framework Convention on Climate Change, May 9, 1992, 31 I.L.M. 849.

n238. Putnam & Graham, supra note 6, at 60.

n239. Keeler & Cretin, supra note 5, at 303; see also id. at 304 ("Delaying any program ... increases its benefit to cost ratio.").

n240. See Arnold, supra note 22, at 178.

n241. See Nordhaus, supra note 233, at 125 ("If investments in equipment or human capital yield 10 percent annually, it would be inefficient to make investments that yielded only 3 percent."); id. at 135.

n242. See id. at 125.

n243. See Hillman & Kim, supra note 5, at 200-02; Michael W. Jones-Lee & Graham Loomes, Discounting and Safety, 47 Oxford Econ. Papers 501, 511 (1995); Lipscomb, supra note 94, at S237.

n244. See Lewis A. Kornhauser & Richard L. Revesz, Evaluating the Effects of Alternative Superfund Liability Rules, in Analyzing Superfund, supra note 133, at 115, 118.

n245. See id.

n246. In some cases, in contrast, environmental remediation costs may fall over time as a result of technological innovation.

n247. Even if the cost were less than $ 100, a static evaluation would counsel against investing in remediation if the funds could be invested in an alternative project with a sufficient return.

n248. In practice, the problem is more complicated because the increase in costs and damages is likely to be continuous but the structure of the analysis remains the same.

n249. See William D. Nordhaus, Economic Approaches to Greenhouse Warming, in Global Warming: Economic Policy Responses 33, 58 (Rudiger Dornbusch & James M. Poterba eds., 1991) ("we are likely to be increasingly averse to climate change as the change becomes larger").

n250. See Robert C. Lind, Intergenerational Equity, Discounting, and the Role of Cost- Benefit Analysis in Evaluating Global Climate Policy, 23 Energy Pol'y 379, 382 (1995); David W. Pearce et al., The Social Costs of Climate Change: Greenhouse Damage and the Benefits of Control, in Climate Change 1995: Economic and Social Dimensions of Climate Change 179, 184-86 (James P. Bruce et al. eds., 1996) [hereinafter Climate Change 1995].

n251. See Pearce et al., supra note 250, at 214.

n252. See Lind, supra note 250, at 384.

n253. See James K. Hammitt, Outcome and Value Uncertainties in Global-Change Policy, 30 Climatic Change 125, 130 (1995).

n254. See K. J. Arrow et al., Intertemporal Equity, Discounting, and Economic Efficiency, in Climate Change 1995, supra note 250, at 125, 132 ("society cannot set aside investments over the next three centuries, earmarking the proceeds for the eventual compensation of those adversely affected by global warming"); Farber & Hemmersbaugh, supra note 19, at 297 (same); Lind, supra note 250, at 381-82 (questioning society's ability to make transfers across several generations).

n255. See Nordhaus, supra note 249, at 57.

n256. Tyler Cowen & Derek Parfit, Against the Social Discount Rate, in Justice Between Age Groups and Generations 144, 148 (Peter Laslett & James S. Fishkin eds., 1992); see Farber & Hemmersbaugh, supra note 19, at 291; James C. Wood, Intergenerational Equity and Climate Change, 8 Geo. Int'l Envtl. L. Rev. 293, 321 (1996).

n257. David W. Pearce & R. Kerry Turner, Economics of Natural Resources and the Environment 223-24 (1990); see Morrall, supra note 132, at 28 (without discounting "all rules yielding continuous benefits are worth any amount of immediate costs").

n258. For further discussion, see infra Part II.C.

n259. For example, Tyler Cowen and Derek Parfit note:

 

No generation can be morally required to make more than certain kinds of sacrifice for the sake of future generations. And this is part of a more general view, which has nothing to do with time. On this view, no one is required to make great sacrifices merely to benefit others.

 

Cowen & Parfit, supra note 256, at 149.

n260. See Robert Solow, An Almost Practical Step Toward Sustainability, 19 Resources Pol'y 162, 168 (1993).

n261. Maureen L. Cropper et al., Rates of Preference for Saving Lives, 80 Am. Econ. Rev. Papers & Proc. 469, 469 (1992) [hereinafter Cropper et al., Rates of Time Preference]. For an earlier version of the study, see Maureen L. Cropper et al., Discounting Human Lives, 3 Am. J. Agric. Econ. 1410 (1991).

n262. See Cropper et al., Rates of Time Preference, supra note 261, at 469.

n263. See id. at 471 tbl.1. For studies using shorter time frames, see Cairns, supra note 164, at 222; John A. Cairns & Marjon M. van der Pol, Saving Future Lives: A Comparison of Three Discounting Models, 6 Health Econ. 341, 343 (1997); Horowitz & Carson, supra note 11, at 408; Jan Abel Olsen, Time Preferences for Health Gains: An Empirical Investigation, 2 Health Econ. 257, 259 (1993).

n264. See Johannesson & Johansson, supra note 13, at 331. For an evaluation of the extent to which the framing of the question affects the results, see Magnus Johannesson & Per-Olov Johansson, Saving Lives in the Present Versus Saving Lives in the Future - Is There a Framing Effect, 15 J. Risk & Uncertainty 167, 169 (1997) [hereinafter Johannesson & Johansson, Risk & Uncertainty].

n265. See Cropper & Portney, supra note 73, at 375. The study is Ola Svenson & Gunnar Karlsson, Decision-Making, Time Horizons, and Risk in the Very Long-Term Perspective, 9 Risk Analysis 385 (1989).

n266. See supra text accompanying notes 92-97. As three prominent commentators recently explained:

 

If one discounts present world GNP over 200 years at 5 percent per annum, it is worth only a few hundred thousand dollars, the price of a good apartment. On the basis of such valuations, it is clearly irrational to be concerned about global warming, nuclear waste, species extinction, and other long-term phenomena. Yet we are worried about these issues, and are actively considering devoting very substantial resources to them. There appears to be a part of our concern about the future that is not captured by discounted utilitarianism.

 

Andrea Beltratti et al., Sustainable Growth and the Green Golden Rule, in The Economics of Sustainable Development 147, 149 (Ian Goldin & L. Alan Winters eds., 1995).

n267. See Cairns, supra note 164, at 224-25 ("the further in the future the benefit the lower the rate at which most individuals discount it"); Cairns & van der Pol, supra note 263, at 342 (referring to "increasing evidence ... that individuals do not appear to apply a constant discounting model"); Cropper et al., Rates of Time Preference, supra note 261, at 471 ("Discount rates are much higher for short horizons than for long horizons."); Johannesson & Johansson, Risk & Uncertainty, supra note 264, at 174 ("estimated discount rates decrease[ ] with the time horizon"); Olsen, supra note 263, at 262 ("The longer the time horizon, the lower are the implied [discount] rates."). One study found a similar result in an intragenerational context. See Loewenstein & Thaler, supra note 176, at 184 ("discount rates declined sharply with the length of time to be waited").

n268. See supra text accompanying notes 264-265 (discussing Svenson & Karlsson study).

In arguing in favor of a constant discounting model, William Nordhaus states that "it would be unrealistic to make decisions based on the premise that there is, in fact, no time preference given that many social decisions are, in fact, tilted in favor of present generations." Nordhaus, supra note 233, at 123. It is therefore worth emphasizing that the studies discussed in this section reveal a strong moral intuition against such discounting.

n269. See Arrow et al., supra note 254, at 137-38; Cropper & Sussman, supra note 65, at 162; Fuchs & Zeckhauser, supra note 161, at 265; Jones-Lee & Loomes, supra note 243, at 501; Lind, supra note 250, at 385-86.

n270. See Arrow et al., supra note 254, at 130, 134.

n271. In theory, the rate could also be negative, which would imply the privileging of the utilities of later generations.

n272. See Arrow et al., supra note 254, at 134-35; Lind, supra note 250, at 385. If one adds utilities over an infinite time period, the social welfare function will be ill-defined; to avoid this problem, some discounting would be required. See Arrow et al., supra note 254, at 136; Jones-Lee & Loomes, supra note 243, at 507, n.10. As Kenneth Arrow and his co- authors explain, however, "because even a very small positive discount rate ... would resolve the mathematical issue, this objection has little practical moment." Arrow et al., supra note 254, at 136.

n273. Arrow et al., supra note 254, at 130; see Nordhaus, supra note 233, at 123-24; David Pearce et al., Sustainable Development: Economics and Environment in the Third World 30 (1990). For the derivation of the relationship, see Arrow et al., supra note 254, at 134-35.

n274. See Arrow et al., supra note 254, at 130; Lind, supra note 250, at 384.

n275. See Cline, supra note 106, at 249; Arrow et al., supra note 254, at 134.

n276. Fuchs & Zeckhauser, supra note 161, at 265 (emphasis added).

n277. See supra text accompanying notes 21-22.

n278. See Robert C. Lind, Intertemporal Equity, Discounting, and Economic Efficiency in Water Policy Evaluation, 37 Climatic Change 41, 52 (1997).

n279. See Thomas C. Schelling, Intergenerational Discounting, 23 Energy Pol'y 395, 396 (1995) ("To be less interested in the welfare of East Africans than former Yugoslavians is less like 'discounting' than, perhaps, 'depreciating.' When we count future welfare less than our own we are depreciating generations that are distant in time, in familiarity, in culture, in kinship, and along other dimensions.").

n280. As a result, the issue of growth discounting is not presented by the example.

n281. Of course, taking a "multiple selves" analysis to its logical conclusions, see supra text accompanying notes 227-230, would turn any intragenerational problem into an intergenerational problem.

n282. See Cowen & Parfit, supra note 256, at 155 ("Pure time preference within a single life does not imply pure time preference across different lives."). As Joseph Lipscomb notes in the medical context, with respect to future generations, "discounting represents a global political decision about the relative weights current decision makers should attach to future population cohorts." Lipscomb, supra note 94, at S246. He adds that this discount rate "need have no relationship to how a given population member (or a statistically representative member) values current versus future gains in health status." Id.

n283. See supra text accompanying notes 227-230 (discussing "multiple selves").

n284. See Richard Dubourg & David Pearce, Paradigms for Environmental Choice: Sustainability versus Optimality, in Models of Sustainable Development 21, 24 (Sylvie Faucheux et al. eds., 1996) ("For maximizing a single utility function ... over infinite time cannot help but suggest that we are dealing with a single generation which exists forever, or even a single individual."); Lind, supra note 250, at 385 (discussing why other approaches are preferable). For example, Kenneth Arrow and his co-authors acknowledge that the rate of time preference "is sometimes said to represent discounting for impatience or myopia." Arrow et al., supra note 254, at 131. These are precisely the sorts of psychological characteristics that justify intragenerational discounting.

n285. The problem is fairly pervasive. For example, Kenneth Arrow and his co-authors note that discounting for time preference reflects that "one cares less about tomorrow's consumer than today's, or about one's own welfare tomorrow than today." Arrow et al., supra note 254, at 130. This formulation conflates the intergenerational and intragenerational problems.

n286. See Lipscomb, supra note 94, at 238 (constant discounting "is basically a political judgment about intergenerational equity").

n287. Arrow et al., supra note 254, at 131; Parfit, supra note 21, at 485.

n288. An even narrower view of the role of future generations in the utilitarian calculus is that of Maureen Cropper and Frances Sussman. They explain their approach:

 

Each generation receives utility from its own consumption and that of its immediate descendants. Because this is true of all generations, the current generation necessarily takes into account the utilities of all future generations in making its consumption and bequest plans.

 

Cropper & Sussman, supra note 65, at 170.

This approach has been criticized as unduly privileging the position of the current generation. See Zeckhauser, supra note 102, at 440-41 ("There is the significant issue ... whether ... this sort of altruism does not substantially underrepresent the impacts that will be truly felt.").

n289. See supra text accompanying notes 261-268.

n290. See Arrow et al., supra note 254, at 137.

n291. See supra text accompanying notes 21-22; Schelling, supra note 279, at 396.

n292. See Arrow et al., supra note 254, at 136; Jones-Lee & Loomes, supra note 243, at 502 n.4; George Tolley & Robert Fabian, Future Directions for Health Value Research, in Tolley et al., supra note 70, at 300, 311.

n293. See Arrow et al., supra note 254, at 136 ("Some have argued that the discount rate should be adjusted for the probability of extinction. Plausible estimates of this effect would add very little to the discount rate.").

n294. See supra text accompanying notes 252-253.

n295. See Parfit, supra note 21, at 482; Jones-Lee & Loomes, supra note 243, at 502 n.4; John F. Morrall III, Cotton Dust: An Economist's View, in The Scientific Basis of Health and Safety Regulation 93, 107-08 (Robert W. Crandall & Lester B. Lave eds., 1981).

n296. See supra text accompanying notes 252-253.

n297. See Heinzerling, supra note 7, at 2044-45.

n298. It is conceivable that in some instance one could make a particularized, factually grounded case for a probabilistic reduction of harms.

n299. John Rawls makes the following case against a pure time preference:

 

There is no reason for the parties [in the original position] to give any weight to mere position in time. They have to choose a rate of saving for each level of civilization. If they make a distinction between earlier and more remote periods because, say, future states of affairs seem less important now, the present state of affairs will seem less important in the future. Although any decision has to be made now, there is no ground for their using today's discount of the future rather than the future's discount of today. The situation is symmetrical and one choice is as arbitrary as the other. Since the persons in the original position take up the standpoint of each period, being subject to the veil of ignorance, this symmetry is clear to them and they will not consent to a principle that weighs nearer periods more or less heavily.

 

John Rawls, A Theory of Justice 294 (1971); see also id. at 284-98 (setting forth a theory of intergenerational justice). For commentary, see John Broome, Counting the Cost of Global Warming 31, 96-98 (1992); B.M. Barry, Justice Between Generations, in Law, Morality, and Society: Essays in Honour of H.L.A. Hart 268, 276-81 (P.M.S. Hacker & J. Raz eds., 1977).

n300. See supra Part I.H.

n301. Some prominent economists are at the very least ambivalent about discounting for pure time preference. For example, Robert Solow notes:

 

You may wonder why I allow discounting at all. I wonder, too: no generation 'should' be favored over any other. The usual scholarly excuse - which relies on the idea that there is a very small fixed probability that civilization will end during any little interval of time - sounds far-fetched. We can think of intergenerational discounting as a concession to human weakness or as a technical assumption of convenience (which it is).

 

Solow, supra note 260, at 165; see also Cline, supra note 106, at 249 ("Impatience or 'myopia' may be a legitimate basis for a single individual's preferring consumption earlier rather than later in his lifetime, but from society's standpoint it is hardly a justifiable basis for making intergenerational comparisons"); Lind, supra note 22, at S-20 (intergenerational discounting "would seem a highly questionable if not immoral public policy"); Robert M. Solow, Intergenerational Equity and Exhaustible Resources, 41 Rev. Econ. Stud. 29, 40 (1973) (expressing doubts as to whether time discounting is appropriate). Kenneth Arrow and his co-authors do not analyze explicitly what the rate of time preference should be, but assume at times that it would be zero. See Arrow et al., supra note 254, at 131.

n302. In practice, the distinction is not as crisp because generations are not successive, but overlapping. The conceptual distinction, however, remains important. For models of overlapping generations, see Burton, supra note 13; Cropper & Sussman, supra note 65, at 169-72. When generations overlap, the current generation tends to convey benefits on the next generation even when it is motivated only by its self-interest. See Barry, supra note 299, at 268 (as a result of the overlap "prudent provision for the welfare of all those currently alive therefore entails some considerable regard for the future").

n303. See Geoffrey H. Heal, Discounting and Climate Change: An Editorial Comment, 37 Climate Change 335, 335 (1997).

n304. See supra text accompanying notes 273-275.

n305. See Arrow et al., supra note 254, at 134-35.

n306. See id. at 131-32, 141 n.10.

n307. See id. at 132.

n308. See supra Part I.E.1.a.

n309. See Cline, supra note 106, at 116-19; Pearce et al., supra note 250, at 195, 198.

n310. See supra text accompanying note 100.

n311. See Heinzerling, supra note 7, at 2051.

n312. See Cline, supra note 106, at 101-06 (discussing species loss and damage to forests).

n313. See Arnold, supra note 22, at 177; Heinzerling, supra note 7, at 2051.

n314. See Cline, supra note 106, at 110-12.

n315. See Schelling, supra note 279, at 399.

The 1990 Amendments to the Montreal Protocol on Substances that Deplete the Ozone Layer marked the first time that a developing country's adherence to the provisions of an international environmental treaty was linked to the receipt of financial resources. See 1 Sands, supra note 234, at 269. As Philippe Sands points out, these amendments "introduced a radical and innovative change which has had profound consequences on the negotiation of subsequent global environmental treaties." Id. This change is evident in the provisions of the 1992 Climate Change Convention which requires developed countries to provide financial assistance and technological assistance to developing countries. See id. at 740-41.

n316. See World Bank, GNP Per Capita (visited July 24, 1998) <http://www.worldbank.org/depweb/gnp/gnpaas01.htm>.

n317. The differences in the patterns of per capita energy consumption between developed and developing countries are stark. See International Energy Agency, Climate Change Policy Initiatives 28 tbl.3 (1992). Over time, this share of the responsibility might decrease as developing countries industrialize.

n318. Perhaps, however, there is a concern that direct foreign aid would not be spent wisely by the recipient, or could create undesirable incentives. If these problems were sufficiently serious, long-term environmental investments could be the most desirable way of providing foreign assistance.

n319. See Louis Kaplow, The Optimal Supply of Public Goods and the Distortionary Cost of Taxation, 49 Nat'l Tax J. 513, 516-19 (1996); Louis Kaplow & Steven Shavell, Property Rules Versus Liability Rules: An Economic Analysis, 109 Harv. L. Rev. 713, 744-45 (1996). For discussion of the distributional consequences of environmental policy, see Richard L. Revesz, Foundations of Environmental Law and Policy 102-03 (1997).

n320. See Farber & Hemmersbaugh, supra note 19, at 300.

n321. The substitutability of these future benefits is discussed below in the context of the principle of sustainable development. See infra Part II.E.

n322. See infra text accompanying notes 343-344.

n323. See supra text accompanying notes 254-256.

n324. See, e.g., Gary D. Meyers & Simone C. Muller, The Ethical Implications, Political Ramifications and Practical Limitations of Adopting Sustainable Development as National and International Policy, 4 Buff. Envtl. L.J. 1, 10 (1996) ("The core idea of sustainability, then, is the concept that current decisions should not impair the prospects for maintaining or improving future living standards."); Edith Brown Weiss, Intergenerational Equity: A Legal Framework for Global Environmental Change, in Environmental Change and International Law: New Challenges and Dimensions 385, 385 (Edith Brown Weiss ed., 1991) ("Sustainable development rests on a commitment to equity with future generations.").

For a strong critique of the concept of sustainable development, see Wilfred Beckerman, Through Green-Colored Glasses: Environmentalism Reconsidered 143-60 (1996).

n325. See Dubourg & Pearce, supra note 284, at 27 ("Sustainability has become a common policy objective of many government institutions, international agencies, and non-governmental organisations."); supra text accompanying notes 234-237.

n326. Some commentators link the attractiveness of sustainable development with criticisms of discounting approaches: "There appears to be a part of our concern about the future that is not captured by discounted utilitarianism. Perhaps as much as anything it is this that is driving an interest in formalising the concept of sustainability." Beltratti et al., supra note 266, at 149.

n327. See David Hodas, The Climate Change Convention and Evolving Legal Models of Sustainable Development, 13 Pace Envtl. L. Rev. 75, 77 (1995); Averil Rothrock, Oregon's Goal Five: Is Ecologically Sustainable Development Reflected?, 31 Willamette L. Rev. 449, 451 (1995); Mary Pat Williams Silveira, International Legal Instruments and Sustainable Development: Principles, Requirements, and Restructuring, 31 Willamette L. Rev. 239, 243 (1995); Christopher D. Stone, Deciphering "Sustainable Development", 69 Chi.-Kent L. Rev. 977, 978 (1994). For general discussion of the principle of sustainable development, see 1 Sands, supra note 234, at 198-208.

One commentator has suggested that over 70 definitions of the term exist. See Susan L. Smith, Ecologically Sustainable Development: Integrating Economics, Ecology, and Law, 31 Willamette L. Rev. 261, 276 (1995); see also John Peezey, World Bank Environment Paper Number 2: Sustainable Development Concepts: An Economic Analysis, app. A (1992) (presenting an extensive list of definitions).

n328. World Commission on Environment and Development, Our Common Future (1987).

n329. Id. at 43.

n330. See Weiss, supra note 324, at 401-05; Solow, supra note 260, at 162.

n331. The following two paragraphs are adapted from Revesz, supra note 319, at 307-08.

n332. See Weiss, supra note 324, at 401-05; Edith Brown Weiss, In Fairness to Future Generations: International Law, Common Patrimony, and Intergenerational Equity 40-45 (1988).

n333. See Solow, supra note 260, at 162-63.

n334. See id. at 167-68.

n335. See id. at 168.

n336. Compare Weiss, supra note 324, at 404 ("The principle of conservation of quality requires that we leave the quality of the natural and cultural environments in no worse condition than we received it."), with Solow, supra note 260, at 167 ("If sustainability means anything more than a vague emotional commitment, it must require that something be conserved for the very long run.").

n337. Compare Weiss, supra note 324, at 404 ("We may exhaust more reserves of a natural resource and cause modest levels of pollution, but pass on a higher level of income, capital, and knowledge sufficient to enable future generations to develop substitutes for the depleted resource and methods for abating or removing pollutants."), with Solow, supra note 260, at 168 ("Most routine natural resources are desirable for what they do, not for what they are. It is their capacity to provide usable goods and services that we value. Once that principle is accepted, we are in the everyday world of substitutions and trade-offs.").

n338. Compare Weiss, supra note 324, at 403 (we "must proceed extremely cautiously" with respect to the possible destruction of a "unique natural resource"), with Solow, supra note 260, at 168 ("It makes perfectly good sense to insist that certain unique and irreplaceable assets should be preserved for their own sake").

n339. See Solow, supra note 260, at 163 ("So far ... the proper adjustments needed to measure the stocks and flows of our natural resources and environmental assets are not being made in the published national accounts.").

n340. See Michael Jacobs, The Green Economy: Environment, Sustainable Development and the Politics of the Future 84 (1991) ("The final objection which might be made to our definition of sustainability is that it ignores population growth."); Michael Redclift, Sustainable Development: Exploring the Contradictions 29 (1987) ("The concept of 'sustainability' makes little sense ... unless we also consider the impact of rapid population growth on the physical resource base."); Nafis Sadik, Population, Environment, and Sustainable Development, in In the Aftermath of the Earth Summit 21, 23 (Andreas Gettkant ed., 1993) ("The universal acceptance of the strong links between sustainable development and the preservation of the environment does not extend to the links between these two and the population policy."). But see President's Council on Sustainable Development, Population and Consumption Task Force Report 13-32 (1997) (discussing how population growth is linked to sustainability).

The link to population does not play a role in the discussions by Weiss, supra note 324, at 401-05, and Solow, supra note 260.

n341. For an exploration of the ethical consequences of this link, see Parfit, supra note 21, at 351-441; Broome, supra note 86, at 161-62.

n342. See Jacobs, supra note 340, at 84 ("It could be argued that what sustainability demands is not simply a constant level of environmental capacity but a constant per capita or per person level."); Richard Baldwin, Does Sustainability Require Growth?, in The Economics of Sustainable Development, supra note 266, at 51, 52 ("The simple fact is that current population growth rates, if they were maintained, would lead to an unsustainable world population.").

n343. See Gregory D. Fullem, The Precautionary Principle: Environmental Protection in the Face of Scientific Uncertainty, 31 Willamette L. Rev. 495, 500-01 (1995); Alexandre Kiss, The Rights and Interests of Future Generations and the Precautionary Principle, in The Precautionary Principle and International Law: The Challenge of Implementation 19, 27 (David Freestone & Ellen Hey eds., 1996); Bernard A. Weintraub, Science, International Environmental Regulation, and the Precautionary Principle: Setting Standards and Defining Terms, 1 N.Y.U. Envtl. L.J. 173, 177-78 (1992). For a discussion of the status of the precautionary principle in international environmental law, see 1 Sands, supra note 234, at 208-13.

n344. See supra text accompanying notes 252-253.

n345. These issues are explored briefly in Revesz, supra note 319, at 330-31.

n346. In this context, the principle of sustainable development has the same features as the maximin principle.

n347. See supra text accompanying notes 314-316.

n348. See supra text accompanying notes 281-286.

n349. Other objections to growth discounting are discussed at supra text accompanying notes 308-311.

n350. See supra Part II.E.

n351. See supra text accompanying notes 319-320.

n352. See supra text accompanying notes 316-317.

n353. See supra text accompanying note 317.