Rational Discounting for Regulatory Analysis

Rational Discounting for Regulatory Analysis W Kip Viscusit This Article examines the economic basis for what is termed "rationaldiscounting," which entails full recognition of policy effects over time and exponential discounting at a riskless rate of return. Policies often cannot be ranked unambiguously in terms of their present or future orientation.Both failure to discount and preferential intergenerationaldiscountinggenerate inconsistencies and economic anomalies.Office of Management and Budget (OMB) discounting guidelines now stipulate more reasonable discount rates than earlierguidelines,but err in permitting open-ended preferentialrates for intergenerationaleffects. This Article presents a methodology for monetizing the value of statisticallife for people of different ages and at different points in time. Review of regulatory analyses indicates increased consistency of discounting practices.However, an examination of two policies with intergenerationaleffects, stratosphericozone regulation and nuclear waste storage at Yucca Mountain, reveals failures to adopt a rational discounting approach. The influence of behavioralanomalies such as hyperbolic discounting may make full recognition of intertemporal effects in benefit-cost analysis more consequentialthan the use ofpreferentialdiscount rates. I. INTRODUCTION Intergenerational discounting should be no different than withingeneration discounting. The policy position I will advocate in this Article is that distant benefits and costs should be recognized fully in the policy analysis process, but that they should be weighted based on the same discount rate methodology that is applied to effects on the current generation. The impetus for a preferential rate may stem in part from the dramatic mathematics of exponential discounting. Let the discount rate be r and policy benefits and costs at time t be weighted by the discount factor 1 / (1 + r).' Suppose the value of r is 3 percent. Then benefits a year from now will have a weight of 0.97, benefits two years from now will have a weight of 0.94, and so on. By the time one reaches twenty years in the future, which might well be the latency period for cancer risks from some environmental exposures, the discount factor is 0.55, or benefits and costs are weighted at just over half of their within-period value. Likewise, the discount factor becomes t University Distinguished Professor of Law, Economics, and Management, Vanderbilt University. James Wawrzyniak provided excellent research assistance. 1 This formula can be found in a variety of basic texts. See, for example, Howell E. Jackson, et al, Analytical Methods for Lawyers 244 n 7 (Foundation Press 2003). The University of Chicago Law Review [74:209 0.23 after fifty years, 0.05 after one hundred years, and 1.45 x 1013 for effects one thousand years in the future. For the very distant future, all but the most consequential benefits and costs will drop out of the analysis.2 The discount weight pattern is a straightforward consequence of valuing all policy effects using a consistent discounting approach and need not be a cause for alarm. Thoughtful commentators who advocate a preferential discount rate for future generations have framed the issue in a manner that creates a bias toward thinking of what lower rate should be applied to effects on future generations.3 Thus, the question that is posed is whether society should use the same discount rate for all policy benefits and costs, or whether a lower rate should be used in the future. Indeed, the main policy issue in their view is how much lower the discount rate should be for effects on future generations. Rather than framing the intergenerational discounting question in terms of preferential lower rates, I would like to frame the policy evaluation question in a more fundamental way. Should effects on future generations even be considered in the policy evaluation process? Why not set their values equal to zero? Notwithstanding the possibility of constructing hypothetical social welfare functions in which the welfare of future generations matters, the current generation's policy choice task is much simpler. How do we make choices now to maximize our own discounted well-being? The well-being of future generations may enter our utility functions, or it might not. Some people may care about future generations in an altruistic manner, but perhaps not a great deal. Per capita income levels and living standards have risen over time, and if the past is any guide, future generations will be more affluent and better off economically than we are, just as we have had a higher standard of living than past generations. The current citizenry consequently may feel quite justified in taking a within-generation perspective and might not be too moved by the plight of their more affluent, distant descendants. The degree to which personal self-interest may have profound consequences for future generations is reflected in the public's attitude toward climate change policies. Efforts to combat global warming through gas taxes will necessarily have a deferred impact on global 2 See Frank Ackerman and Lisa Heinzerling, Pricingthe Priceless: Cost-BenefitAnalysis of Environmental Protection,150 U Pa L Rev 1553,1571 (2002) ("At a discount rate of five percent, for example, the death of a billion people 500 years from now becomes less serious than the death of one person today."). 3 See, for example, Richard L. Revesz, Environmental Regulation, Cost-Benefit Analysis, and the Discountingof Human Lives, 99 Colum L Rev 941, 1015-16 (1999) (advocating for intergenerational preferences). 2007] Rational Discountingfor RegulatoryAnalysis climate change, which is a long-term environmental problem. To what extent will older age groups be willing to pay more for gasoline so that gasoline will be less harmful to the environment? My analysis with Joni Hersch of the Eurobarometer survey data found that concern with this environmental amenity declined steadily with age, which reflects the degree to which there is a strong component of self-interest governing the public's willingness to pay for environmental benefits over time. More specifically, in terms of the additional percent amount that respondents were willing to pay for gasoline, the average response was a high value of 2.8 percent among those age fifteen through thirtyfour, 2.3 percent for those thirty-five through forty-four, 2.1 percent for those forty-five through fifty-four, 1.6 percent for those fifty-five through sixty-four, and 1.0 percent for those sixty-five and over.' This dramatic dropoff in valuation led the authors to conclude that there is a generational divide in support for environmental policies! If people are self-interested in the extreme, they might place no value whatsoever on the well-being of future generations. From the standpoint of their policy assessments, concern about what discount rate should be used to value effects (...truncated)