WRDS Library [Home] Digital Library Publications Videos Card Catalog

CHAPTER 3

BENEFIT-COST MEASURES AND CONSIDERATIONS

3.1 INTRODUCTION

In the previous chapter we discussed some of the shortcomings of using benefit-cost analysis alone to assess the economic implications of Wyoming water projects. However, we fully recognize these shortcomings do not preclude the consideration of traditional efficiency measures. That is, the relevance of criteria other than efficiency does not preclude the use of efficiency criteria in project evaluations. As such, in this chapter we consider those aspects of water project evaluation which are "traditional" in the sense of bringing together those quantifiable aspects, in terms of a dollar metric, that are relevant for preparing a benefit-cost measure. This discussion will then lead us to the development of a preliminary methodology (Chapters 4 and 5) which consider both efficiency and non-efficiency measures.

Conceptual and empirical procedures for estimating national economic development (NED) benefits and costs associated with agricultural, municipal/industrial, and recreational water projects are well known and abound in the literature (see, e.g., Eckstein [1958]; Maass et al. [1962]; Mishan [1976]; Howe [1971]; Peskin and Seskin [1975]; Merewitz and Sosnick [1971]; Dasgupta and Pearce [1972]; and Young and Howe [1988]). There is little to be added to these writings concerning so-called NED benefits and costs. We do not intend to reproduce the details. Thus, our attention will focus upon methodological issues which are generally relevant for the valuation of a Wyoming water project from a Wyoming perspective.

3.2 BASIC PRINCIPLES AND A WYOMING PERSPECTIVE

Any evaluation of a Wyoming water project that incorporates benefit-cost measures, is anchored by the concept of allocative efficiency. The task is to determine, using a rather narrow band of defined and measured benefits and costs, whether the proposed project brings about a net gain to Wyoming as measured by the metric of dollars. Thus, for purposes of an analysis, the benefits are how much Wyoming residents are willing to pay for the project outputs. Costs are the opportunity costs of the foregone opportunities in using Wyoming resources in an alternative manner. Conceptually this is rather straightforward.

There are several reasons for using a Wyoming rather than a federal (NED) perspective for project evaluation. First, projects evaluated from the federal perspective assume that full employment always exists. This is clearly not always the case in Wyoming. Second, the objectives of the residents of Wyoming might well be different than the federal perspective. The survey results presented elsewhere in this report would appear to support this possibility. Finally, Wyoming is competing with other states for water and will continue to have to do so in the future. The analysis in Appendices A and E support this perspective. Thus, we argue that the rules published by the federal government over the years, which culminated in the Economic and Environmental Principles Guidelines for Water Related Land Resource Implementation Studies (Water Resources Council [1973]), are not appropriate from a Wyoming perspective.

Another reason for a Wyoming water perspective is to enhance Wyoming's competitive position relative to other economic regions. That is, Wyoming is not investing in the west as the federal government has over the years but in Wyoming. In part, this is the central motivation of the Wyoming Water Development Program where money has been explicitly set aside to develop the water resources of the state for economic development purposes. We recognize that this viewpoint explicitly accepts the legislative and administrative structure of the Wyoming Water Development Program. That is, it does not allow for the question that many economists might raise: "How much of the states resources should be devoted to water development and how much to other projects?" It is conceivable that if this were the question then how the array of benefits and costs are established would potentially be different.

The implications of this view are rather straightforward: the Wyoming analyst must consider benefits beyond those traditionally considered for federal projects. In the federal case, only direct benefits to the nation (NED benefits) are considered. For example, consider a project which yields direct and indirect benefits for enterprises in Wyoming and out-of-state. Only the direct and indirect benefits that accrue to the citizens of Wyoming are appropriate if a Wyoming perspective is taken. From the federal perspective, all of the direct and none of the indirect benefits would be considered. The same argument for a Wyoming evaluation would apply "to "ths cost side of the ledger: only the opportunity cost to Wyoming would be included in the analysis. In both cases, double counting should be avoided.

3.3 ESTIMATION OF BENEFITS AND COSTS

There are two central issues regarding the estimation of benefits and costs: (1) what are the types of benefits and costs generated by a project and (2) how does the analyst implement measurement of the benefits and costs in order to achieve a monetary metric? What is appropriately included and what is not included is dependent upon the perspective taken. However, having adopted the Wyoming perspective, our task is rather straightforward. We are only interested in those benefits and costs that are relevant to the citizens of Wyoming. In our discussion, we will limit ourselves, for illustrative purposes, to the benefits and costs from a traditional water containment facility. ¹² We address the issue of secondary effects later in the chapter.

3.3.1 Municipal and Small Industrial Users

Benefits depend upon the user classes. These would include residential, commercial, public and industrial water users. Thus, for any project, the users of the water from the facility would have to be identified. Two methods are traditionally accepted to measure these benefits: (1) the willingness to pay approach as deduced from market measures, ¹³ and (2) the alternative cost approach as deduced from the cost of the next best alternative source.

Costs would essentially be the cost of developing any new distribution system to deliver the water. Typically, it would be assumed in the analysis that the direct project costs would capture such elements as water rights costs (if any) and the associated capital and operating costs.

3.3.2 Large Industrial Users

For a large industrial user, the benefit stream would consist of: (1) payments received by Wyoming under a purchase and/or lease arrangement for utilization of the containment facilities industrial water yield, (2) the direct income to Wyoming residents, (3) any indirect income generated by the industrial user, and (4) any tax payments to the state or local governments. Measurement of these categories of benefits are rather straightforward. For instance, the direct payments can be determined based upon the value of water in the region or nationwide. Appendix A is suggestive of the order of magnitude of this type of estimate.

The direct income estimates can be based upon employment projections. The indirect income benefits consist of two components: (1) increased income attributable to the industrial facility, and (2) the associated multiplier effect on the Wyoming economy. Input-output models, while having many shortcomings, can be used in this regard. If the input- output approach is inadequate then the analyst can turn to measures of induced employment where a similar type facility has been constructed.

Costs associated with an industrial facility would depend upon the agreement between the state and the industrial user. In principle, all of the appropriate costs would be identified as in the case of the municipal costing situation.

3.3.3 Agricultural

A principle set of benefits stemming from agriculture would be the increase in moving the land from dry land farming to irrigated land and/or reducing the potential for water shortages. Either or both of these effects might well lead to more intensive cultivation, higher valued crops and expanded acreage. Benefits are typically market measures and represented as net farm income.

The costs would include any infrastructure necessary to deliver the water, and the operating and maintenance of the infrastructure. Ideally, the analyst would determine whether there was, as a result of the project, a net income loss to agriculture in Wyoming and any other costs stemming from surplus stocks. Again market values would be the principle mechanism to estimate the costs.

3.3.4 Recreation

The benefits stemming from recreation would result from increased opportunities for water based recreation. This might include increased flat water recreation as well as fishing. There exists a wide variety of methods for estimating such benefits. In principle these estimates are based upon the notion of willingness to pay. These methods include the travel cost method and the contingent valuation method. For a discussion of these methods, see Cummings, Brookshire and Schuize [1986] and Mitchell and Carson [1989] for the contingent valuation method and McConnell [1985] and Smith and Desvousges [1986] for the travel cost method. Any calculation of recreation benefits would have to take into account the potential for losses in certain types of recreation activities as a result of the containment facility.

3.4 MEASURING BENEFITS AND COSTS: SOME CONCERNS

3.4.1 Concerns About Secondary Benefits and Costs

The use of secondary benefits and costs has lead to criticism of benefit-cost measures. This criticism stems from the observation that "double counting" might well be the result. However, the issue is not quite so clear when a "Wyoming eyes" perspective is adopted. Central to the argument are assumptions regarding the level of employment. Federal applications assume full employment. As such, secondary benefits and costs are not counted. That is, the state of Wyoming does not have full employment. Further, there are idle resources. As such, the arguments that are set forth vis-a-vis federal project evaluation procedures might or might not apply in a project evaluation in the state of Wyoming.

What then is an appropriate perspective to take? Initially it is important to acknowledge that a project will always generate secondary benefits and costs. A project will effect both input markets and output markets.

For projects in Wyoming, the agricultural and recreation sectors are likely to generate secondary benefits and costs. For example, consider a project which attracts a number of recreational visitors to the project site. The analyst must know whether this is a net increase for the state or simply a "reallocation" of visitors already using recreation facilities in the state. If the visitors are "new" then there might well be a secondary benefit as represented by increased expenditures on fishing tackle. If there is only a reallocation then there would not be any secondary benefits. Thus, our perspective is that the decision will be project specific as to whether to include or exclude secondary effects.

3.4.2 The Use of Prices for Estimating Benefits

For project inputs and outputs which have market prices, such as agricultural products, construction and O&M materials, prices are typically used as the appropriate measure for project benefits and costs. Several issues must be considered in the use of prices.

First, the use of market prices for inputs and outputs implies an important assumption: the scale of the project is such that market prices will be unaffected. ¹⁴ Such conditions will generally prevail when, in the case of agricultural outputs, the increase in the total production of any one particular crop expected as a result of the project is small relative to the total market of the crop in the relevant market area. When very large projects are under consideration, and substantial increases in the production of any one or more crops are anticipated, the use of prevailing prices can lead to overestimates of project benefits. This follows from the simple notion that substantial increases in the quantity of the crop put on the market will likely depress prices received for the crop. Market studies designed to estimate the likely price response to the increase in crop production are typically non-existent.

Second, it is common practice to use current (or five year average) prices to value all future costs and benefits over the life of the project. This practice implies the assumption that current prices can serve as "real," inflation- free prices for all future years. More to the point, this assumes that the future rate of inflation relevant for prices used for the costs of the project is the same as the expected future rate of inflation relevant for the costs of the project. Thus, the inflation rates for benefits and costs will cancel, and current prices for benefits and costs are appropriate for the valuation of future benefits and costs. ¹⁵

The use of real prices for benefits and costs has appeal in that it relieves the researcher from the near-impossible task of estimating future prices for items included as benefits and costs. The problem is, however, that historical data tell us that we should know that this practice has invariably resulted in the persistent overestimation of benefits attributed to water development projects, particularly those designed to serve agricultural purposes. ¹⁶

Still another indication of such overestimates for agricultural benefits is seen in the Farm Parity Ratio (the ratio between prices received by farms and prices paid by farms). The Farm Parity Ratio has fallen consistently over the past 35 years.¹⁷ Relative to 1950, prices received by farmers in 1985 had increased by only about one half of the rate at which prices paid by farmers had increased. Thus, projects assessed in 1950 which used current prices and costs for benefits and costs expected in the 1980s would have substantially overestimated project net benefits.

Overestimates of project benefits are not corrected by adjustments for productivity gains.¹⁸ Typically, farm production is increased in future years based on past trends in productivity gains. This reflects output changes. At issue then are the net returns associated with output. As is demonstrated (footnote 18) terms of trade have persistently moved against agriculture over time. ¹⁹

In pointing out these problems in using "current" prices for valuing project benefits and costs, our intention is to draw attention to possible biases. Practically, there are few palatable alternatives for the current practice.

3.4.3 The Choice of a Planning Horizon

The "planning horizon," or planning period, refers to the length of time over which expected benefits and/or costs are to be included in project analyses (see Young and Howe [1988], pp. 36-38). For many years, it was common for the Bureau of Reclamation to use the expected physical life of the project as the appropriate planning horizon. For dams and reservoirs, the expected physical life was typically around 100 years. Since around 1973, however, the Bureau's planning horizon for project analyses has been more akin to the economic life of the project, on the order of 50 years. The economic life of a project is generally shorter than the physical life of the project, due to such things as anticipated technological changes and/or market obsolescence, population shifts, changes in government support programs, and shifting patterns in international trade in agricultural commodities.

Over the last decade or so, it has become common for researchers to use a 50-year horizon but the rationale is seldom stated. Young and Howe [1988] recommend a 50-year planning horizon, and seemingly base their choice on the observation that with discount rates of 5 percent or more, ". . . at least 90% of total present value is accounted for by year 50" (Young and Howe [1988], p. 37).

While certainly no more compelling than the rationale used by Young and Howe, our justification for a planning horizon on the order of 50 (or fewer) years reflects our concerns with biases in estimates of benefits and costs discussed above, which typically become more pronounced the longer the planning horizon. Clearly, the state should use planning horizons longer than those used in private companies, given the broader range of social goals relevant for water development projects. A 50-year horizon would generally balance the need for longer periods of time against which to amortize the investment costs of water development projects and the concern for the uncertainties of benefit/cost estimates in estimates of future values.

3.4.4 Choosing a Discount Rate

As was observed by Baumol:

. . . few topics in our discipline rival the social rate of discount as a subject exhibiting simultaneously a very considerable degree of knowledge and a very substantial level of ignorance ([1968], p. 788).

In our simplest theories, the choice of a discount rate is straightforward: consumer's rate of time preference would equal the marginal productivity of capital, in which case the market rate of interest is the appropriate rate of discount.

Problems arise from a plethora of sources (for comprehensive overviews see Lind [1990]). Examples are: (1) differences in tax rates applied to consumer rates of interest and those related to returns on private investment give downward biases to consumer rates, (2) new evidence suggests that the "shadow price of capital" may distort measures sought in a social discount rate (Lind [1990]; Lyon [1990]), and (3) recent dramatic changes in the world and U.S. economies, in terms of more integrated international capital markets, make questionable earlier estimates of social discount rates based on the assumption of a closed economy (Lind [1990]; Feldstein [1985]).

It has generally been accepted that "appropriate" estimates for a social discount rate must focus on opportunity costs resulting from foregone consumption. Thus, mandating a focus on displaced consumption and consumer rates of interest. Results from recent research, however, suggest that such focus may be misleading given that consumers may rationally pay and receive wide ranges of different interest rates as a result of a lack of self control (Thaler and Shefrin [1981]; Thaler [1985]; and Lind [1990]).²⁰

Given the morass of problems surrounding the discount issue, the obvious question arises as to what is currently being done in terms of dealing with these problems, and what approach to discounting might be in the best interests of Wyoming. The U.S. Office of Management and Budget (OMB) [1972] has, for the last 17 years mandated the use of 10 percent as a discount rate; this rate is based upon the pre- tax rate of return on private capital (Lyon [1990]). An exception is made, however, for assessments of water projects. For water projects, the "appropriate" rate is taken to be the Treasury's borrowing rate for instruments with maturity in 15+ years.²¹ It is commonly recognized, however, that this rate, as with the 10 percent OMB rate, is a nominal as opposed to a real, discount rate (Lind [1990]; Lyon [1990]). In contrast to the OMB rate, the Government Accounting Office uses discount rates based on the treasury borrowing rate. ²²

In considering the above arguments and recalling the argument that Wyoming's water projects should be viewed from a Wyoming eyes perspective, we find that the discount rate should reflect the rate of return on state borrowing. This rate will properly reflect the choice facing the state; to build the project now or later. We have seen no evidence since we last visited this issue (Watts, Brookshire and Cummings [1989]) to change our recommendation of approximately a four percent real discount rate.

3.4.5 Issues of Uncertainty

There are a wide range of sources for uncertainties surrounding important variables in water project assessments. Substantial uncertainties may also be relevant for the legal and institutional environment relevant for water resources planning in Wyoming (Appendices A and E). Here we discuss some of these major sources of uncertainty and means for bringing such uncertainty to bear on project assessments are described.

At the outset of any assessment of a water project, the practice is to estimate future demands (uses) for (of) the outputs of the project. Examples include: (1) increases in agricultural output and crop prices, (2) floods avoided by the project, (3) recreational uses associated with the project, (4) power output and prices, and (5) municipal water uses. Most often, such estimates of future uses of project outputs are based upon first, historical use patterns and secondly, commitments from basin residents for future water contracts.

The record of success for estimates based upon these considerations is poor. Uses "expected" or "predicted" at the time at which a large number of water reclamation projects were being evaluated have often failed to materialize after the project was in place (Franklin and Hageman [1984]). ²³ Referring specifically to irrigation, the notion that large irrigation developments will give rise to significant growth in a region's employment levels is belied by a large number of studies.²⁴

There are basically three ways by which the uncertainty of projections are treated. The first of these involves the use of "expected values." Thus, if benefits of $500 million may accrue from industrial water uses 10 years in the future, and the probability that such industrial uses will in fact occur is 10 percent, the expected value of the $500 million is $500 million multiplied by 10 percent (and then discounted, of course), or in non-discounted terms, $50 million. An application of this use of expected values is seen in (Watts, Brookshire and Cummings [1989]). The major weakness of the expected value approach is obvious: it is difficult to specify probabilities associated with future water use developments.

A second approach for treating uncertainty is "sensitivity analysis." Critical variables for which values are uncertain are varied (usually, one at a time) in efforts to determine the sensitivity of the benefit-cost measure to changes in the values of the variable being analyzed. Thus, given a $500 million estimate for industrial benefits as in the above example, and considerable uncertainty as to whether or not such benefits will actually accrue to the water development project, the benefit cost measure might be calculated with alternative values assigned to industrial benefits varying, e.g., from zero to $500 million.

The major weakness of this approach is that it does not "treat" uncertainty in the sense of allowing the analyst to arrive at some objective number which would be assigned (in this example) to industrial benefits. One is really simply asking the question: does the value assigned to industrial benefits "matter" in the sense of effecting substantial changes in the benefit-cost measure? If the value assigned to the variable does matter, the analyst can do little more than attach a caveat to the reported benefit-cost measure.

The third approach is referred to as the analyses of "switching values," and is an extension of the sensitivity analysis approach. The analyst attempts to define the value of the variable in question which results in negative net benefits (a benefit-cost ratio less than unity). Continuing the industrial benefits example, if one were using sensitivity analysis, the benefit-cost ratio might be calculated with the following arbitrarily chosen values for industrial benefits of (resulting benefit-cost ratios are in parentheses): $0 (.60); $100 million (.75); $200 million (.90); $300 million (1.2); $400 million (1.4); and $500 million (2.0). In looking for a switching value, one would search for that value for industrial benefits between $200 million and $300 million which would result in a benefit-cost ratio of I-for example, $245 million. The result of this type of analysis is a statement like the following: if one "believes" or accepts industrial benefits at levels greater or equal to $245 million, the project is efficient; if not, the project is not efficient.

The above examples make obvious the fact that there are no ideal or perfect means for treating the uncertainty of projections. As suggested by Young and Howe [1988], however, "the most important point is that imperfect knowledge of the future should not be ignored" (p. 75). One means for easing the weight of uncertainty on the benefit cost measures is to choose a shorter planning horizon as discussed above.

3.5 CLOSING REMARKS

For the purposes of this Phase I work, we have attempted to set out some of the more important considerations which should be considered in any effort to structure "traditional" measures of the economic efficiency of a project. Missing here are a number of extensions which remain as tasks for Phase II efforts, examples of which include: (1) the development of a manual which provides step-by-step guidance of the preparation of benefit cost measures, (2) extensions and applications of example or situational methods for measuring indirect benefits, and (3) modifications of existing farm budgets which one might use in assessments of agricultural projects.

90-12 Table of Contents
Water Resources Publications List
Water Resources Data System Library | Water Resources Data System Homepage