Discounting for Time and Risk in Energy Policy

Robert C. Lind, Kenneth J. Arrow, Gordon R. Corey, Partha Dasgupta, Amartya K. Sen, Thomas Stauffer, Joseph E. Stiglitz, J.A. Stockfisch

• Publisher: RFF Press
• Year: 2013
• ISBN:
  9781135985257
• Edition: 1
• Topic: Economics
• Subtopic: Environmental Economics
• Index: Economics

This is acollection of theoretical papers, including contributions by Partha Dasgupta and three Nobel prize-winning economists: Kenneth Arrow, Amartya Sen, and Joseph Stiglitz. Originally published in 1982.

1

Robert C. Lind*

Introduction

Not since the late 1950s and 1960s, when the economic evaluation of federal water projects centered on the selection of discount rates to be used in policy analysis, has this issue been as important as it is now in connection with national energy policy. Development of our nation’s options for new energy supplies and for conservation requires large investments now that will produce benefits well into the future. To evaluate these options we must weigh the future benefits against the present costs. Therefore, the weight we give to benefits and costs at different times is critical and is determined by the discount rate that we use. The higher the discount rate, the lower we value future benefits and costs as compared with present ones.

Energy research and development projects costing billions of dollars may not begin to produce benefits for twenty to thirty years, and these benefits may then accrue over one or more centuries. To compare these benefits and costs over time, the standard analytical procedure is to discount them to their present values. Their sum, the net present value of the project, is the critical measure of a project’s economic value. In investments that will produce large benefits or costs well into the future, the discount rate used in the calculation of present values strongly affects these totals.

If the discount rate is as high as 10 percent, the present values of costs and benefits in the future become insignificant compared with those of the present. For example, the present value of $1,000 of benefits fifty years hence is worth $8.52 if discounted at 10 percent, $87.20 if discounted at 5 percent, and $371.53 if discounted at 2 percent. Therefore, if we use a discount rate of 10 percent rather than 2 percent, we reduce the present value of benefits and costs fifty years hence by more than a factor of 40. The implication of this basic arithmetic is nowhere more apparent than in the evaluations of the U.S. liquid metal fast breeder reactor program. The net present value of the benefits from breeder development as estimated by the Energy Research and Development Administration (1975) falls from $46.8 billion to $16 billion when the discount rate is raised from 7.5 percent to 10 percent.¹ Not only are the benefits and costs of many energy policy options highly sensitive to the discount rate, but whether the net benefits are positive or negative depends critically on the discount rate; small variations in that rate will often tip the balance.

Even the sensitivity of the present value of the net benefits of many energy policy options would not pose such a serious analytical problem if there were agreement that the appropriate discount rate lay within relatively narrow bounds. Unfortunately, there is no such agreement, and rates of 2 percent and 10 percent both lie within the range of rates that have been proposed and defended for evaluating energy policy decisions. Furthermore, different studies of alternative energy technologies use widely varying discount rates. In evaluation of the breeder reactor alone, Manne and Richels (1978) use 5 percent and 10 percent; Stauffer, Wycoff, and Palmer (1975, p. 6), 6 percent; the Energy Research and Development Administration, 7.5 percent and 10 percent; and Chow (1975) recommends a procedure for determining the appropriate rate that in one particular case results in a rate of 10.07 percent.

To put the discount rate issue in perspective, it is useful to compare today’s analytical debates, accepted wisdom, and positions on energy policy with those twenty years ago on water resource policy. In many ways the present debate is a replay or at least a continuation of the previous one, but there are some new elements and fundamental differences that are peculiar to energy.

The basic similarity is that energy projects, like water projects, require initial investments that produce benefits well into the future. Therefore, with energy projects, as with water projects, the discount rate is critical to three major dimensions of energy policy evaluation. First, it is a major determinant of whether a project is economically efficient, that is, whether the net present value of benefits is positive. This is a critical test of whether we should invest in a project. A second, closely related point is that the discount rate is a major determinant of the relative value of competing projects. Because funds are limited, we may have to choose among competing energy technologies, and this choice will depend on which have the highest net present value. Third, the discount rate is a major determinant of the optimal timing of projects. For example, should we proceed now to develop a technology such as the breeder, or should we postpone its development?

With energy projects, as with water projects, sometimes even a small change in the discount rate will make a significant difference in the evaluation of a project’s economic potential. Sometimes the evaluation will not be as sensitive to the discount rate. There are times when an analyst needs to be able to specify the rate of discount with precision, and other times when, if the discount rate lies within a fairly broad range of values, the economic evaluation of the investment decision remains unchanged.

One technique that has long been used to resolve policy disputes arising over the selection of a discount rate is to test the sensitivity of a particular policy to the discount rate. However, in many other cases, either it may not be practical to test for sensitivity (because of the cost of energy model runs, for example), or the benefits of the policy may be found to be highly sensitive to the discount rate that is used. In these cases, greater resolution of the discount rate issue is required if benefit–cost studies and energy models that incorporate a discount rate are to be useful in choosing among alternative energy policies. Therefore, the energy modeling community, like the water policy community before it, has been one of the groups most interested in resolving the discount rate issue and, in particular, in developing a defensible discount rate that can be used in economic and energy models and in energy policy analysis.

In many economic and energy models, a discount rate enters in two ways. First, analysts use a discount rate to calculate the net present value of national economic benefits for alternative energy policies or investments. To compute the social value of these benefits from a national perspective, one must discount using the appropriate rate for such a calculation, that is, the “social rate” of discount. Second, these models sometimes use a discount rate in order to simulate private sector investment behavior. The models predict which investments will be made in the private sector, evaluating private investment alternatives by means of a discount rate equal to the required rate of return on investment in the private sector. One of the big open questions in the controversy over discount rates is whether the social rate should be the same as the required rate of return on private investment.

This volume will primarily address the question of the appropriate rate of discount for evaluating the present value of benefits from a national or social perspective. That is, it focuses on the question of the social rate of discount. There are several reasons for examining this rate as opposed to the appropriate rates for use by utilities or oil companies. First, the federal government will play a major role in making energy policy and will probably be one of the major supporters of energy research and development. In making public policy decisions on the development of the breeder reactor, the mandating of coal conversion, the setting of energy efficiency standards, and the regulation of utilities, many individuals in government will want to know the net present value of alternative policies from a national or social viewpoint. The social rate of discount is relevant to this calculation.

A second reason for considering the social rate of discount is that, despite its importance for energy policy, there is no agreement on how this rate should be determined nor on what the correct rate should be. By contrast, firms and individuals will make their decisions on the basis of the rates of return and rates of interest prevailing in the market place. When compared with determination of the social rate, determination of rates that should be used by private firms in making their own decisions, while subject to some controversy, is relatively straightforward. However, determination of a defensible rate of discount for use in computation of the present value of net national economic benefits has caused considerable difficulty for energy modelers and policy analysts. Hence, it is the focus of this volume.

Before exploring various dimensions and issues associated with the social rate of discount, it is useful to consider how the discount rate has been handled in examples of energy models and benefit–cost studies of energy options. The modeling group for the study by the National Research Council’s Committee on Nuclear and Alternative Energy Systems (CONAES) used an average pretax rate of 13 percent as the required rate of return on private investment, and they used 6 percent to discount national economic benefits (National Academy of Sciences, 1978). The 6-percent rate is considered to represent the after-tax rate of return on private investment. From the CONAES project followed the Energy Modeling Forum (EMF), sponsored by the Electric Power Research Institute (EPRI). In five of the first six energy models that were developed under its aegis, one or two discount rates enter as important parameters.² Because of practical limitations on the number of scenarios that could be considered in the studies based on these models, sensitivity testing was not generally performed to determine whether the discount rate assumptions that these models incorporate have a significant effect on the policy-relevant outputs of the models. However, the perceived dependence of the model results on the assumed discount rates coupled with the absence of a clear-cut choice of values for these parameters has frequently been cited as a major potential limitation by energy modelers (Hogan and coauthors, 1979).

To see the variety of discount rate assumptions that are built into EMF models alone, consider that the report of the first EMF study, Energy and the Economy (EMF, 1977), follows the CONAES practice of using discount rates of 13 percent and 6 percent to represent the required rate of return on private investment and the social rate of discount. The second EMF study report, Coal in Transition (EMF, 1978), incorporates only one rate, namely, 13 percent, as the required private sector rate of return. The third EMF study report, Electric Load Forecasting (EMF, 1979), uses different discount rates for different regions. A 1980 EMF working paper, World Oil Study Design (EMF, 1980b) uses a uniform rate of discount of 5 percent, and U.S. Oil and Gas Supply (EMF, 1980a) incorporates a required rate of return for the oil and gas industry of 8 percent. However, in this last case, the results were tested for sensitivity, and changing the discount rate from 8 percent to 16 percent reduced the 1995 projection of conventional crude oil production from 9.5 million barrels a day to 4 million barrels a day.

The report of another study that does limited testing of the results for changes in the discount rate is “A Decision Analysis of the U.S. Breeder Reactor Program” by Manne and Richels (1978). They compute the present value of national economic benefits from alternative breeder reactor programs using 5 percent and 10 percent as discount rates. They state: “Since both breeder and reprocessing benefits accrue almost entirely post-2000, their present value is quite sensitive to the choice of a public discount rate. (This comment holds not only for the breeder but also for fusion, solar electric, and virtually any other phenomenon of the 21st century!)”

Because the choice of the discount rate can influence strongly which public policies can be supported by benefit–cost analysis and which cannot, it is a matter of concern to politicians as well as policy analysts. The choice of the discount rate for evaluating public choices is itself a public policy decision that in most cases will be politically determined. While philosophers, economists, and financial analysts may debate the appropriateness of one rate as opposed to another for public policy decisions, and while their arguments may well be influential, the final choice will often be determined politically. It will depend not only on the merits of the supporting economic arguments but also on the policy implications of one choice versus another and on the political strength of forces in support of those implications.

For example, under the Nixon administration, the Office of Management and Budget (OMB) in March 1972 directed most federal agencies to apply a 10 percent real rate of discount when calculating the present value of the costs and benefits of federal projects (OMB, 1972). Previously there had been a wide range of practices with regard to discounting and discount rates used by federal agencies, as reported by the comptroller general to the Joint Economic Committee in 1968, and some agencies did not use discounting at all. The comptroller general’s report states:

A divergence of opinion on discounting is reflected in agency practices. Some agencies use the Treasury cost of borrowing money as the discount rate while others use a rate based on the return on investment in the private sector of the economy. Still others employ different criteria to determine the agencies discount rate and agency borrowing cost. The agencies included in our survey used discount rates which vary over an extremely wide range—from about 3 to 12 percent.

Within each school of thought there are important differences of opinion. Of those agencies which tie the discount rate to Treasury borrowing costs, one uses the estimated cost of new money to the Treasury, another uses the average cost of money to the Treasury, others use the cost prescribed by Senate Document 97 (about 3.2 percent at the time of our review).

One of those agencies which would tie the discount rate to the rate of return in the private sector uses a rate of return on a safe investment and a slightly higher rate (3 percent and 5 percent) while another uses for some programs the rate representative of average capital returns in the private sector (presumed to be 12 percent) and for some programs permits the analyst to determine the rate on the basis of his judgment as to the nature of the program and the kind of analysis considered most meaningful.

Views are equally disparate in those agencies which report that discounting was not used in their analyses of individual programs for support of their fiscal year 1969 budget decisions. At one extreme is the view that decisions on programs should be made on the basis of first year costs and benefits—a procedure which implies a discount rate of 100 percent, since the future is ignored. At the other extreme is the view that decisions should be based on total undiscounted costs and benefits—a procedure which implies a discount rate of zero, since costs and benefits applicable to, say, the 20th year are treated as being as important as current costs and benefits (U.S. Congress, 1968, pp. 11–12).

The directive in OMB Circular A–94 raised the discount rate used by most agencies to 10 percent and made it consistent across agencies. One exception, however, was the rate used to evaluate water resources projects. Congress resisted the move to raise the discount rate to 10 percent for water projects by refusing to abandon use of the formula set forth in Senate Document 97 (U.S. Congress, 1962). Raising the discount rate for water projects would have had adverse implications for public works projects; that is, many fewer would have met the benefit–cost test. Finally, Congress itself wrote into law the formula that determines the discount rate to be used in evaluating water projects in Section 80–A of the Water Resources Development Act of 1974 (Public Law 93–251). The prescribed discount rate for use in evaluating water projects has been consistently below the 10 percent rate for other agencies including energy agencies.

Those supporting the 10 percent rate, which was believed to approximate the marginal real rate of return on capital in the private sector, argued that to achieve greater overall economic efficiency, the rate of return in the public sector should be the same as in the private sector. Perhaps a more important argument for this change, however, was that the administration in power favored reduced government spending and fewer government projects. With a higher rate of discount, fewer government projects would pass the test of economic efficiency. Consequently, it was easier for the administration to make the case that they should not be funded.

It is instructive to look at how the political interests of different groups would be affected by a high versus a low discount rate in evaluation of both water and energy policies. Those groups favoring regional development, particularly in the West and South, supported large-scale, government-supported water projects and generally argued for lower discount rates. Their counterparts today are groups favoring an expanded role for government in implementing a national energy policy and groups favoring expanded government funding for research and development projects for new energy-producing and energy-conserving technologies; for example, the breeder reactor, solar and fusion power, and alternative propulsion systems. Similarly, now as in the 1950s and 1960s, those who favor cuts in government spending and less direct government involvement in the economy tend to support a higher rate.

Two other groups have a stake in the choice of the discount rate that is used in evaluating public investments and public policies. They are the environmentalists and the private, investor-owned utilities. The situation of the environmentalists with respect to the discount rate, as they have come to realize, is complicated by the fact that in some cases a higher rate leads to policy conclusions that are consistent with their goals, and in some cases it does not. With regard to water resources, for example, environmentalists saw the requirement of a high rate of discount for use in evaluating water projects as a way of slowing development and preserving natural areas. The result was that the environmentalist groups found themselves in coalition with fiscal conservatives in support of a high rate of discount and fewer federal water projects.

However, a high discount rate cuts the other way in energy and minerals policy. The economic case for rapid development and exploitation of our mineral and fossil fuel resources is enhanced by the use of a high discount rate. This is because the higher the discount rate, the lower the value that these resources will have if left for future development. Another example in which a higher discount rate goes against the environmentalists’ position is the case of nuclear waste or any other case in which environmental pollutants may pose a potential threat to the environment well into the future. To the extent that these long-term costs are quantified and incorporated into a benefit–cost analysis, the higher the rate of discount, the less important these costs will appear.

In the case of nuclear waste, where the costs of contamination might accrue several thousand years from now, any positive rate of discount, when ...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Original Copyright Page
• Contents
• Foreword
• Acknowledgments
• 1 Introduction, Robert C. Lind
• 2 A Primer on the Major Issues Relating to the Discount Rate for Evaluating National Energy Options, Robert C. Lind
• 3 A Reader's Guide to the Papers in this Volume, Robert C. Lind
• 4 The Rate of Discount on Public Investments with Imperfect Capital Markets, Kenneth J. Arrow
• 5 The Rate of Discount for Benefit–Cost Analysis and the Theory of the Second Best, Joseph E. Stiglitz
• 6 Risk Measurement of Public Projects, Robert Wilson
• 7 Measuring the Social Rate of Return on Private Investment, J. A. Stockfisch
• 8 Resource Depletion, Research and Development, and the Social Rate of Return, Partha Dasgupta
• 9 Approaches to the Choice of Discount Rates for Social Benefit–Cost Analysis, Amartya K. Sen
• 10 Plant Investment Decision Making in the Electric Power Industry, Gordon R. Corey
• 11 The Social Efficiency of Electric Utility Decision Criteria, Thomas Stauffer
• 12 The Rate of Discount and the Application of Social Benefit–Cost Analysis in the Context of Energy Policy Decisions, Robert C. Lind
• Index