First, I will briefly explain the technique for valuing human lives that is generally employed as the starting point for the determination of the benefits of environmental regulation. This technique involves ascertaining the wage premiums demanded by workers employed in risky occupations who face a probability of death from industrial accidents. Second, I will explain why valuations based on such wage premiums need to be adjusted upward before they can be properly used in the context of environmental regulation. Third, I will show why certain downward adjustments of the value of life that have been advocated in certain academic and public policy circles are in fact inconsistent with the technique of cost-benefit analysis and should not be performed. Fourth, I will explain why the discount rate used by the Office of Management and Budget (OMB) in its administration of Executive Order 12,866 is a great deal higher than rates supported by economic theory, and show that, as a result, certain environmental benefits are severely undervalued. Fifth, I will discuss how the Executive Order and various legislative proposals couple cost-benefit analysis with procedural devices designed to thwart regulation, rather than to make regulation more rational.
The issues discussed in Parts I, II, and IV of this testimony are discussed in more detail in an article that I recently published, which is attached as Appendix I.
I. Valuations of Human Life in Workplace Context
The primary benefit of many important environmental statutes is the human lives that are saved. Thus, properly valuing human lives must be an important part of any cost-benefit inquiry.
Since the 1970s, willingness-to-pay studies have become the standard economic technique for placing a value on human life. By far the most common method for performing such valuations focuses on the choices that workers make in accepting risky jobs. The approach begins by defining 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. Presumably, a rational worker would not accept the riskier job unless she obtained sufficient compensation for the additional risk. The wage differential between the riskier and the less risky jobs is the compensation that the worker therefore is assumed to demand for the additional probability of death that she faces as a result of having taken the riskier job. The wage differential divided by the additional probability of death is then considered to be the value of life.
II. Why Certain Upward Adjustments Are Necessary
The value of life figures obtained from studies of risky occupations need to be adjusted upward to obtain a meaningful valuation of the benefits of environmental regulation. These adjustments must account for the generally involuntary nature of most environmental harms, for the differences between the median income of the workers who are the subjects of these studies and of the population as a whole, and for the dreaded nature of certain environmental contaminants (principally carcinogens). Not performing these necessary adjustment can result in an undervaluation of life by as much as a factor of six (or even more in certain contexts).
A. Involuntary Nature of the Harm
1. Valuations of Voluntary Versus Involuntary Risks
There is an extensive literature suggesting that individuals assign greater value to avoiding risks that are thrust upon them involuntarily than to risks that they incur voluntarily. The risk assumed by individuals who subject themselves to possible of industrial accidents is generally thought of as a risk assumed voluntarily. In contrast, the risk of exposure to environmental contaminants like air pollutants, is generally thought of as involuntary. As a result, if one takes the willingness-to-pay to avoid voluntary harms and imports that figure into the context of environmental regulation, there will be a systematic undervaluation of the benefits of regulation.
Determining the extent of the undervaluation, however, is complicated. In general, the economics profession 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 differentials individuals demand to accept a risk of death are a prominent example of a revealed preference technique. In contrast, because involuntary risks are by definition not based upon informed market transactions, revealed preference techniques are not available to assess the value of involuntary harms.
Thus, in order to estimate how the valuations of involuntary and voluntary risks differ, one has to ask individuals directly the relative value that they attach to avoiding the two types of harms. The most comprehensive study of this type conducted a nationwide telephone survey of 1,000 households, asking interviewees to compare, among other pairs of risks, radon control in homes and a pesticide ban on fruits. The respondents also were asked to assess, on a ten point scale, the ease with which the respective risks could be avoided.
The respondents' answers revealed that they considered the radon risk more voluntary in that it could be avoided with greater ease. When the 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. The median respondent viewed saving 100 lives by means of the pesticide ban as equivalent to saving 213 lives through radon control. Thus, the median respondent implicitly found the involuntary risk to be twice as harmful.
2. Unrepresentativeness of the Population Exposed to
Workplace Risks
Valuations of life in workplace settings are inaccurate as a measure of the value of life for environmental programs for a second reason. In a competitive marketplace, individuals who take relatively risky jobs by definition have the lowest willingness-to-pay to avoid the risk. Other things being equal, employers will pay the least possible amount to fill the jobs, so individuals with higher valuations will not be hired.
As a result, the willingness-to-pay valuations derived from the study of risky jobs are the valuations of a relatively small subgroup of the population with a disproportionate tolerance for risk. In contrast, most environmental risks affect a far broader sector of society. Thus, the valuations of the individual with the median valuation of risk (not an individual with an unusually low valuation) would be the appropriate metric to use in the valuation of life for cost-benefit analysis of environmental regulation. As a result, an appropriate correction needs to be made when extrapolating from the workplace to the environmental arena. Unfortunately, at this time there is no empirical literature that sheds light on the magnitude of such a correction. But if cost-benefit analysis becomes part of the decisionmaking process under the Clean Air Act, careful attention will need to be paid to this issue.
B. Impact of Income on the Valuations of Life
It is generally recognized that willingness-to-pay valuations of life, such as those obtained in the workplace setting, are a function of the income of the subjects of the study. Economists have estimated, for example, that a 10% increase in income leads to a 10% increase in the value of life. As a result, there are at least two problems with using the valuations from workplace studies for cost-benefit analyses of environmental regulation.
1. Distribution of Income Across Occupations
First, individuals who take risky jobs generally have lower-than-average income. Thus, there is a problem in extrapolating from the willingness-to-pay studies conducted in high-risk occupations to the broader population affected by environmental carcinogens.
The U.S. Census provides median and mean earnings for all workers and for various occupational categories. The category including operators, fabricators, and laborers is probably the best 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 for this category of workers were $16,883 and $19,981, respectively. In contrast, the corresponding figures for the population as a whole were $20,716 and $27,366, respectively. Thus, the median earnings of the population as a whole are 22.7% higher than the median for workers in risky occupations, and the mean earnings of the population as a whole are 37.0% higher. Adjustments of this magnitude therefore need to be performed to the valuations of life from the workplace setting.
2. Increases in Income Over Time
A second problem arises in connection with environmental risks that have a latency period, so that the death will not occur immediately but only after the passage of some time. Empirical studies show that individuals value their lives as a function of their current income, and not on the basis of projections of future income. But for latent harms, the valuation that individuals would have at the time of their death is what matters.
Over the last several decades, median and mean incomes in constant dollars have been rising at a compound rate of about 1% per year. Thus, for contaminants with a 20 year latency period (as is the case with some carcinogens regulated under section 112 of the Clean Air Act) an upward adjustment of about 22% would have to be made to the valuation of life from the workplace setting to make the figure appropriate for environmental regulation.
C. Nature of Carcinogenic Deaths
Particularly with respect to carcinogens regulated under section 112, an upward adjustment of the value of life to account for the dreaded nature of the harm also needs to be performed. Indeed, there is 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.
In addition to the loss of the life itself, two other components need to be valued in the case of carcinogenic harms: the very painful and often extended period of morbidity that precedes the death and the dread aspects of carcinogenic deaths. The leading empirical study in this area found that the valuation of life in the case of carcinogenic exposure is about twice as high as the corresponding valuation in the case of an instantaneous death from an unforeseen accident.
III. Why Certain Downward Adjustment Are Inappropriate
Some policy analysts have suggested that downward adjustments of the value of life obtained in workplace studies must be performed to account for the fact that the beneficiaries of certain environmental programs are older individuals, and that these individuals are often not in good health. The question of how cost-benefit analysis should account for the particular features of the population benefited by environmental regulation is very complex. In any event, however, the particular downward adjustments that have been advocated are inconsistent with a proper understanding of economic theory. A. Treatment of Older Individuals
One important pitfall to be avoided concerns the manner in which cost-benefit analysis deals with programs designed to benefit older individuals, particularly individuals in their seventies and above. Some academics and policy analysts argue that, in computing the benefit of an environmental program designed to save the lives of such individuals, their remaining life expectancy should be multiplied by the value of a life year. In turn, they assert that values of life years should be computed by assuming that workers who take risky jobs (whose median age is about 40) value each of their remaining years the same amount, and that their valuation of life can therefore be broken down into a value for life years. So, for example, making just a minor simplification, if the value of life derived from a worker with a 40-year life expectancy is $6,000,000 (and the value of a life year is therefore $150,000) the value of the life of an elderly individual with only a 4-year life expectancy would be only $600,000.
This methodology is seriously flawed. It assumes that the value of a life year is independent of the number of life years an individual has left to live. But this approach overlooks the critical role that scarcity plays in determining economic value. Just as individuals value diamonds more than water (because diamonds are scarcer), so too they are likely to value life years more highly when they have fewer life years left. Thus, there is no principled basis for taking the valuation of life year given by a forty-year old and assuming that a seventy-year old would have the same valuation. Instead, the latter's valuation should be expected to be considerably higher.
B. Quality Adjustments
Another inappropriate approach consists in using an approach generally referred to as quality-adjusted life years (QALYs) in performing cost-benefit analyses. The idea behind QALYs is that the lives of sick individuals--asthmatics for example--should be assigned a lower value than the lives of healthy individuals (for comparable life expectancies). For example, the life of an asthmatic might be assigned only half the value of the life of a healthy individual. In the case of the individuals with a four-year life expectancy discussed above, the value would then be reduced to $300,000--one twentieth the value of the life of a healthy individual with a 40-year life expectancy.
The QALY technique, as generally employed, is incompatible with cost-benefit analysis. Indeed, the measure of benefits in cost-benefit analysis is derived from the aggregation of the individual preferences of all the individuals affected by a policy. Specifically, each individual has a willingness-to-pay to avoid being subjected to some risk, and it is the aggregation of the individual willingnesses to pay that determines what the benefit of the policy would be.
The QALY technique, in contrast, does not seek to determine what individuals in poor health would be willing to pay to avoid a premature death. Instead, it relies heavily on the assessment of third parties, sometimes healthy individuals and medical professionals, of how undesirable a life in poor physical condition is relative to a healthy life.
Such an inquiry suffers from two fundamental flaws. First, it does not construct the valuation by reference to the views of the affected individuals themselves, when it is the preferences of the affected individuals that form the fundamental units on which cost-benefit analysis is based. Second, how much more miserable one might be in one state rather than another is not responsive to the question of how one's willingness to pay to avoid a premature death varies in the two circumstances. Thus, the rankings provided by the QALY technique typically have no connection to willingnesses to pay, and therefore cannot properly be incorporated into cost-benefit analyses.
IV. Choice of Discount Rate
For many environmental contaminants, such as carcinogens regulated under section 112 of the Clean Air Act, the harm does not occur contemporaneously with the exposure: there is instead a period of latency. It has been the practice of the Office of Management and Budget (OMB), in its review of agency regulations under Executive Order 12,866, to apply a discount rate in the case of latent harms to reflect the fact that the benefit of regulation would not accrue until the future.
OMB currently uses a discount rate of 7%. There is a strong consensus in the economics profession that this rate is too high, and that an appropriate rate would be between 2 and 3%. The 7% rate used by OMB is set by reference to the pre-tax rate of return on private investments. This rate would be the appropriate one to use if the United States had a closed economy, so that investments for pollution control displaced investments in other activities and, as a result, the government lost tax revenues.
Increasing globalization, however, has led to the integration of capital markets and to the opening of the U.S. economy to foreign investment. 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. This rate is generally taken to be the after-tax rate of return, adjusted for inflation, on relatively risk-free financial instruments such as government bonds--a rate that currently stands in the 2-3% range.
The flawed OMB approach leads to a substantial undervaluation of the benefits of environmental regulation. Consider the difference caused by using a 7% discount rate as opposed to a 2.5% discount rate (the mid-point of the plausible range). For contaminants with a 20 year latency period, the OMB approach undervalues the environmental benefits by a factor of 2.36 (the environmental benefits are 136% higher than OMB calculates them to be, so that a benefit that OMB determines to be $100,000,000 is in fact $236,000,000). For a 30-year latency period OMB's approach leads to an undervaluation by a factor of 3.63 (the environmental benefits are 263% higher than OMB calculates them to be, so that a benefit that OMB determines to be $100,000,000 is in fact $363,000,000).
If cost-benefit analysis were to become more prevalent as a result of congressional action, this longstanding problem should be corrected. It is noteworthy, moreover, that both the General Accounting Office (GAO) and the Congressional Budget Office (CBO) correctly employ the 2-3% rate range rather than the flawed 7% rate used by OMB.
V. Procedural Issues
Over the last two decades, cost-benefit analysis has acquired an understandably bad reputation. In its administration of Executive Order 12,866 (and its predecessor, Executive Order 12,291), OMB has attached to the use of cost-benefit analysis procedures that, at least in part, have turned it into an anti-regulatory tool, rather than into a tool to make regulation more rational. Similarly, several of the regulatory reform bills that have been introduced since 1995 contained procedural mechanisms designed to thwart rather than to improve regulation. Let me draw your attention to these pitfalls so that the mistakes of the past (and of the present) can be avoided.
First, the OMB mechanism and the various regulatory reform bills use cost-benefit analysis only in the context of the adoption of a new regulation. In contrast, satisfying a cost-benefit test is not required for the repeal of an existing regulation, the failure to adopt a new regulation, or the failure to make an existing regulation more stringent. The concern for the maximization of social welfare that is implicit in cost-benefit analysis would call for the use of the technique in each of these contexts. The possible losses in social welfare flowing from the repeal of a regulation, the failure to adopt a regulation, or the failure to make a regulation more stringent can be as detrimental--in fact, even more detrimental--than the social welfare losses caused by the adoption of regulations that do not pass a cost-benefit test. There is simply no plausible justification in economic theory (or for that matter in logic) for caring about social welfare losses in one context but not in others.
Second, at times there has been no disclosure (and at other times only limited disclosure) of the communications between interested parties and OMB concerning the cost-benefit analyses of environmental regulations. It is a core requirement of administrative law, embodied in section 4 of the Administrative Procedure Act, that any submissions to an agency in connection with the promulgation of regulations must be made part of the public record and available for public inspection. The obvious purpose of this requirement is to foster openness and to make judicial review more effective. These goals are seriously compromised when communications concerning cost-benefit are either wholly or partially shielded from the public, as has been the case until now in connection with OMB's administration of Executive Orders 12,866 and 12,291.
Third, some regulatory reform bills allowed for judicial review of the cost-benefit analysis prior to the promulgation of the regulation. This type of challenge runs counter to another well entrenched principle of administrative law--limiting judicial review to "final agency action." This principle, which is codified in section 10(c) of the Administrative Procedure Act, bars piecemeal challenges. The approach of those bills would lead to piecemeal challenges with respect to a single regulation, seriously impairing the efficient use of judicial resources. More importantly, such an approach would paralyze the regulatory process during the pendency of any challenge to a cost-benefit analysis and would have the clear effect of delaying regulation that passes the cost-benefit test and therefore improves social welfare.
Fourth, some regulatory reform bills contained a petition process under which any individual or firm could ask the agency to repeal a regulation that did not satisfy the cost-benefit test. The agency, in turn, was required to respond promptly to such petitions and a denial of the request was reviewable by the courts. Administering a petition process of this sort would have the effect of paralyzing the Environmental Protection Agency and would seriously undermine the goal of rationalizing the regulatory process.
Conclusion
If cost-benefit analysis were to play a larger role under the Clean Air Act, great care must be taken to ensure that it is used fairly, as a tool to make regulation more rational, and not as a tool that is biased against regulation. On the substantive front, it is important to ensure that human lives are not undervalued as a result of the use of various techniques that appear to have currency in some academic sectors. On the procedural front, cost-benefit analysis must be woven into the existing structure of administrative law, without either hiding it from public scrutiny nor subjecting it to legal challenges under rules that are different from those generally provided in the Administrative Procedure Act.
Once again, I am grateful for the invitation to testify at this hearing. I am pleased to answer any questions that you might have.