Government Cost/benefit Analyses: The Impetus and the Impotence
MAY 01, 1980 by JOHN SEMMENS
Mr. Semmens is an economist for the Arizona Department of Transportation and is studying for an advanced degree in business administration at Arizona State University.
As the activities of government grow to consume more and more of the available scarce resources, the question of the productivity of the various government activities be comes more and more important. By various measures, the government’s consumption of resources is enormous. The proposed federal budget for 1981 exceeds $600 billion in ex penditures. The funded national debt approaches a trillion dollars. The unfunded federal obligations run to an additional eight trillion dollars. Government, at all levels, consumes about 40 per cent of the national income.
The continuous stream of exposure of government induced waste raises serious questions about the utility of government activities. Senator Proxmire’s “Golden Fleece” award, highlighting atrocious episodes of government waste, is merely the tip of a vast iceberg of misallocation of scarce resources. In addition to this direct evidence of the frivolous expenditure of money, there are the costs of fraud, as demonstrated by the General Services Administration’s purchasing practices. The costs of government are not always revealed in government outlays. An imposing burden of expense for a series of government regulations is inflicted upon private business and individual consumers.
The outpouring of negative publicity has inspired demands that the government put its house in order. On the basis of financial return on investment, most, if not all, of the government’s activities are losing propositions. It is asserted, though, that there are vast social benefits that are generated as a result of government programs and that these benefits are not reflected in financial return calculations. Criticisms of costly government programs were routinely brushed aside with the broom of social benefits. Who can place a price on a human life? Who can evaluate the economic value of social justice?
Outrage over the egregious abuse of the non-quantifiable social benefits arguments has produced pressure to justify the effectiveness of government social programs. This pressure has often manifested itself in the demand that a cost/benefit analysis be performed in order to evaluate the benefits received in exchange for government expendi tures. A number of well-known economists have been prominent in the clamor for cost/benefit studies as a means of assessing the value of government programs.
The inclusion of requirements for cost/benefit analyses in recent legislation may be seen by some as a forward step in the effort to control government waste. However, a careful examination of the rules and procedures laid down for the cost/ benefit analyses reveals that hopes for relief from this sector are not well founded.
It is largely a case of “monkey see, monkey do.” The government procedures for cost/benefit analyses are all too often crude distortions of the financial budgeting practices of private industry. These distortions stem, in part, from simple incompetence—the inability to apply sound economic or financial principles to the evaluation of a public investment. The distortions also receive some impetus from conscious intent to mold the methodology and the data to produce the desired answers. An examination of the cost/ benefit methodology prescribed by the Federal Railroad Administration (FRA) in compliance with requirements of the Local Rail Assistance Act of 1978 will serve as a demonstration of the futility of reliance upon government directed cost/benefit analyses as a check on public spending.
Income Gains vs. Income Transfers
Advocates of government subsidies are sensitive to the apt criticism that subsidies are merely income transfers. The whole tax financed nature of government expenditures would seem to lay a strong prima facie case for the mere transfer hypothesis.
The FRA’s guidelines on cost/ benefit analyses are careful to require a distinction between impacts of local rail assistance which are “real” and those which are “just transfers.” This is all very nice as far as directives go, but how the distinction is to be made is unclear. The description of “real” losses from a prospective rail abandonment read as follows: “If the employee remains unemployed for a period, his/her production is lost for that period. This is a real loss.”
On the surface, the description of “real” losses may appear straight forward. The problem with this approach is its simplistic view of “real” losses. If we can show that Mr. Jones or Ms. Smith will be unemployed if a rail abandonment occurs then their production is lost. What goes unacknowledged is the prospective employment impacts of alternative uses of the funds which are proposed for use in subsidizing rail service. This is the classic case of the seen versus the unseen that was illustrated by Frederic Bastiat in the early 19th century.
The rail line in question may provide service for several shippers, a number of whom might go out of business if required to bear the full cost of providing rail transportation. The closing of these businesses would result in some incidence of unemployment of specific individuals. The production of these individuals would be lost. Therefore, the subsidy is supposedly justified. Unnoticed in this neat little scenario is what the incidence of unemployment is among industries that must be taxed in order to provide funds for the rail subsidy. While it might even be admitted that government taxation can cause marginal businesses to go bankrupt, the actual tax dollars extracted cannot be directly traced to a specific government expenditure. Thus, the two incidents of unemployment are segregated from one another in terms of the FRA’s superficial consideration of “real” losses.
An Untenable Position
One possible explanation of this simplistic segregation of integral components of the taxation and subsidization transfer program is that the focus is only on local impacts. Such an explanation would recognize the absence of any net social gain in a broader context, but still justify the program on the grounds of promoting purely local benefits. An attempt to explain the purpose of the program in these terms was, however, characterized as “unacceptable.” The disruption costs of individual unemployment, business relocation and the like are “real,” “net” social losses, according to the FRA.
The FRA’s position is, of course, untenable. In an advanced economy, where money serves as a medium of exchange and a representation of various prices and values throughout the society, there is no need to identify specific dollars in order to draw a conclusion that a transaction is a transfer of resources rather than a generator of resources. Furthermore, the weight of the evidence indicates that the net effect of the transfer program involving local rail assistance is negative.
To begin with, the subsidy program is a consumer rather than a creator of wealth. Enterprises which produce positive returns on investment are taxed at progressive rates in order to fund services which are not self-supporting. This has the consequence of reducing the net financial return on the total capital investment of the entire society. The sustained effect of such transfers is to reduce the real wealth of the community in all subsequent investment periods. In short, the succeeding generations of humans will be less well off as a result of the tax and subsidy transfer.
That the social returns would compensate for the financial losses is, at best, implausible. If there were significant community impacts, it would seem logical that the affected communities would be willing to fund the subsidy program. Just the opposite is the case. Not only are communities not usually willing to fund the subsidies, they olden have the programs forced on them by the federal government. No tax rebates for non- participation are awarded. Any eligible community that does not grab for its share of the federal money loses out. The local clamor for federal assistance arises from the fact that programs are designed to make the federal grant a free good. As any economist could tell you, at a price of zero, demand will be virtually unlimited. The fact that the grant programs have to be structured in this manner is impressive evidence that the social benefits are exceeded by the financial costs of the program.
A large part of the rationale behind the FRA’s cost/benefit analyses guidelines is tied in with the concept of “consumer surplus.” Consumer surplus is the benefit enjoyed by consumers who do not have to pay their full reservation price for a good or service. At any specific price for a good or service, there will always be customers who would have been willing and able to pay more. Since the price paid serves as a proxy for the value of a good or service to the purchaser, the acquisition of a good or service at less than the highest price a purchaser would be willing to pay leaves such a purchaser with a so-called consumer surplus.
The phenomenon of consumer surplus is used by the FRA to justify subsidies for services whose revenues are inadequate to cover costs: “If the railroad could practice perfect price discrimination, then it may generate an economic profit; if so, then there would exist a net economic gain in keeping the service rather than abandoning it. This gain would exist even if institutional and logistic considerations made it impossible for the railroad to practice price discrimination. Although the railroad may suffer a loss, it would be more than offset by the surplus value enjoyed by the shippers.”
There is no explanation of how we know that the economic losses to the railroad are “more than offset” by the gains of the shippers. Even if it is true that in a given case the price of rail service is held below levels that represent the %rue” value of the service, it does not follow that the gains to the shippers must be larger than the income losses to the railroad. The relevant basis for comparison is not perfect price discrimination. Rather, it is some market clearing price (or prices) that must be compared to the regulation- depressed price used in the FRA approach. Under such conditions it is not at all obvious that the consumer gain from subsidy exceeds its cost.
The whole presumption that consumer surplus exceeds the cost of subsidy is contingent upon our knowledge of what the demand curve would look like given a schedule of prices. In real life, we do not know such things. All we can know for sure is the quantity sold at a given price at a given moment in time. The convenient curves used for academic exposition are hypothetical models of static conditions. Inconveniently, the market is a dynamic process in which the price system acts as a signaling device to guide production decisions. Using the suppression of the price system in railroading via the mechanism of government regulation as the justification for ignoring the signals employs convoluted logic. The most obvious remedy would be the removal of railroad price controls.
On the Other Hand
An unrestricted price system would be the most effective way of dealing with the problem of imbalances in consumer surpluses. Instead, the rigged prices under regulation pave the way for rigged computations of consumer surpluses. As elaborate as the discussion of the issue of consumer surplus is in the FRA guidelines, absolutely no con sideration is given for the loss of consumer surpluses experienced when funds are extracted for the financing of rail subsidies. The opportunity cost calculation used to allow for alternative uses of the subsidy considers only the financial returns of the cost of capital.
In the business world, the cost of capital is usually computed on the basis of what the firm would have to pay in the line of interest on borrowed money or earnings to equity of owner-supplied capital. Since the firm cannot capture consumer surpluses it can legitimately ignore them. Consequently, business fi nance has developed no methods for assessing the impact of consumer surplus in the cost of capital.
The FRA has borrowed the traditional business finance computation, but has distorted the outcome by applying consumer surplus considerations to only one side of the equation. Consistently employed, such a methodology would produce a false picture of the comparative returns of subsidized programs versus taxpaying programs. Followed to its logical conclusion, the FRA’s approach would lead to the absurd proposition that social benefit could be maximized only when all economic activity is subsidized. Apprised of this situation, the FRA has promised to re-examine its premises. In the meantime, the procedures resulting in inflated social benefits as a consequence of rail subsidy will continue as before.
Avoidable Dislocation Costs
Confronted with the ineluctable fact that economic progress entails the birth, life, and death of individual business entities, the FRA has postulated that some inefficiency re sults from the market’s lack of consideration for certain sunk costs. Resources may completely lose their value due to rail line abandonment, says the FRA. For example, a coal mine relying upon rail service could become worthless if abandonment occurred. The example is an inversion of the typical cause and effect relationship, i.e.: because the mine can no longer produce at a profit given the costs of its inputs, including the cost of transportation, it can no longer justify the cost of that transportation. Shifting these costs to the general taxpayer is hardly a rational answer to the scarce resource allocation problem.
The FRA’s contention is that a needless shifting of production facilities will lead to avoidable disruption costs. The FRA says preventing factories, farms or commercial enterprises from losing value as a result of the loss of service is less costly than having to re-establish these activities in other locations. Such an outlook, universally applied, would prevent the relocation of any business activity. It should not be surprising to learn that a bill to inhibit business relocation has been introduced in Congress.
Once again, though, this line of argument contradicts the evidence. If the cost of relocating or closing down were greater than the cost of sustaining rail service, the affected firms should be willing to pay higher rail rates. Generally, this is not the case, although there is some truth to the contention that regulation prevents the payment of higher rates.
Pressed to the wall with the evidence of shipper unwillingness to pay higher rates or to fork over a subsidy, the FRA argues that maybe the long run efficiency calls for relo cation, but the short run costs may be too high. The lag time between a shut down operation in one locale and a new start up in another represents a dead loss due to idle re sources. Unfortunate for this line of argument is the phenomenon of long range business planning. There may not be a hiatus between a closing at one location and an opening at another. In fact, as far as lag time is concerned, it would seem that the lapse of time between the collection of taxes and the disbursement of subsidies with all the red tape that typically accompanies government programs would make a case for disruption in other parts of the economy preceding the rescue of a local rail line from abandonment. So, as far as short run costs are concerned, the local rail assistance program would appear to increase, rather then decrease, the net social losses.
Untenable Working Assumptions
The FRA’s favorite cost/benefit analysis (at the time of this writing) was one in which every factor, save one, was held constant. The state performing the analysis was able to announce triumphantly that the proposed rail project was justified because the cost of switching to trucks exceeded the cost of the subsidy. On the route to reaching this conclusion the following assumptions were made: (1) the demand for transportation is not affected by its price; (2) revenues are a function of costs, i.e.: if we can forecast operating expenses we can add a mark-up for profit and have an estimate of future revenues; (3) public expenditures produce values in excess of their cost, but private expenditures do not; (4) taxes have no impact on private decision making; and (5) the “real” cost of capital is in the 3-5 per cent per year range.
To any reasonable person, these simplifying assumptions are clearly untenable. There is no real world instance in which such assumptions could even approach the bounds of reasonableness. The explanation for finding such an absurd methodology for assessing the cost/benefit relationships of government funded projects combines a mixture of academic naivete, bureaucratic cynicism, and plain incompetence.
The draft of cost/benefit instructions passed out by the FRA is replete with the classic equilibrium models one experiences in introductory economics courses. Neatly drawn demand curves intersecting marginal cost curves yield tidy triangles of values to be calculated. Persons attempting to compute a cost/benefit ratio are admonished to begin by calculating what prices would be under perfectly competitive conditions.
The use of the tools of academic exposition in such a simplistic fashion cannot be expected to lead to superior, or even adequate, investment decisions. Two unfortunate components of the economic environment in the real world are the ruling principles of uncertainty and change. Demand curves are not known. Data collected to estimate demand curves may give estimates of what demand was at some point in time. However, such information is ancient history in terms of the status of demand in current and future markets. To structure the decision aiding cost/benefit procedures along these lines is an act of sophomoric economic naivete that, at best, will result in self-delusion.
A worse consequence than self-delusion is the element of public deception involved in the whole process. The procedure has been made inordinately complicated. Unfamiliar terminology makes it difficult to determine what the FRA is talking about. Where most investment analysis procedures refer to streams of cash flow, the FRA uses the terms “benefits” and “disbenefits.” Presumably, this changed terminology is necessary because of the non-cash flow benefits or costs resulting from the public investment, but no glossary of unfamiliar terminology accompanies the guidelines.
An Unreal World
The creation of the concept of “transportation efficiency” is a procedure that is virtually guaranteed to produce benefits so large that almost any subsidy could be justified. It is here that the assumption of the insensitivity of demand for transportation to the price of transportation comes into play. Starting with the not unreasonable proposition that the demand for transportation may be relatively inelastic, an FRA endorsed simplifying assumption of an elasticity of zero, opens the door to the outrageous conclusion that all the freight not moved by rail will be moved in some alternative manner. With the quantity of freight moved taken as a constant, ridiculously high savings from rail subsidies can be “demonstrated,” particularly in the movement of heavy raw materials. Unmentioned in such a comparison are the problems encountered when the cost of alternative transportation exceeds the total value of the delivered freight. In the real world no firm could afford to view the quantity of freight to be moved as an unalterable given, especially if the cost of a second-best transportation option exceeds the recoverable revenue from the delivered price of the commodities shipped.
The simplifying assumption usually is justified on the grounds of conservatism. That is, even though the numbers produced may not reflect actual decisions, they at least reflect a lower limit of the costs entailed in a course of action. In the case of rail abandonment it is simply assumed that the cost of not transporting the potential freight would be greater. The example given is that of a body of ore whose value would go to zero if it were not transported.
The fatal flaw in this whole approach is that it cannot take cognizance of the market’s response to the various costs involved in producing and transporting goods. Left to its own devices, the price system of the free market would determine the least costly alternative. The calculations based upon FRA approved methodologies will produce distorted values which can serve no useful allocation role. Participants in the program are unofficially counselled to adopt procedures and assumptions which will give their projects the best chance of being funded. The purpose of the program is to shovel money out to the participating governmental units and railroads. The better the cost/benefit figures look, the easier will be the disbursement of funds.
Lest we attribute more sinister intent than is warranted to the apparent machinations of the railroad bureaucracy, it must be pointed out that a large measure of the blame for the distorted methodology is likely the result of pure incompetence. What else could explain an estimate of future rail revenues based upon a simple cost-plus formula? The genesis of the entire rail subsidy program could be found in the phenomenon of rail costs rising fast er than rail revenues on specific branch lines. If it were possible to maintain a constant relationship between rail revenues and costs by use of the cost-plus formula, the whole FRA program would have no rationale for its existence.
Ignoring the Risk
The selection of a so-called real rate of interest for the purpose of considering the opportunity cost of rail subsidies is another convincing piece of evidence for the incompetence hypothesis. The FRA suggests that current interest rates include a premium for inflation and that unless the whole future streams of benefits and disbenefits are to be adjusted for inflation, a real cost of capital rate of between 3 and 5 per cent should be used. Aside from the fact that there can be no assurance that both revenues and expenses will be equally affected by inflation, the element of risk is completely overlooked.
The notion that government investments can ignore the element of risk has a curious origin. In the world of finance, government debt instruments are considered to be essentially risk-free in terms of safety of principal and interest payments. The fact that a government bend may be secure for the holder of the bond says absolutely nothing about the risks of the uses to which the funds will be put. The investment decision in the rail subsidy program is no different from any other use of funds—public or private. Funds committed to the program will be converted into irretrievable capital goods or paid out in operating expenses. The risk that funds thus committed may not earn a return sufficient to justify such a conversion is no less than for any other use of the funds. Neglect of this risk factor, because government can borrow money more cheaply because the debt instruments are risk-free to the holders, is clearly a demonstration of ineptitude.
A final example of incompetence is the failure to consider the indirect benefits of alternative uses of scarce resources when computing the opportunity cost of rail subsidies. The entire rail assistance program is contingent upon the concept that benefits not reflected in the price system can be preserved via subsidy. This is the rationale for rejecting the market’s indication, as exemplified in the net loss position of a particular rail branch line, that resources can be put to better use elsewhere. Thus, the FRA guidelines lay out elaborate procedures for calculating consumer surpluses and indirect community benefits that are to be preserved by subsidizing the rail service. The opportunity costs of uses of scarce resources for other than rail subsidy include only the prospective financial returns of those uses. The consumer surpluses and indirect benefits of any use other than rail subsidies are completely ignored.
Apprised of this imbalance in their methodology, the FRA’s only response was that they just hadn’t thought of it. Since it is the contention of the FRA that the program is increasing the net social benefit to the nation, what possible excuse can there be for the failure to even think of the ramifications of the diversion of scarce resources from other uses? Personal experience inclines one toward the hypothesis that incompetence is the most likely explanation.
Those who look to the enactment of mandatory cost/benefit analyses as a check on the excesses of government had better not place much reliance upon it. The “state of the art” in government circles would seem to indicate that cost/benefit analyses will not perform as some are wont to claim.
The fundamental issue centers on the appropriate mechanism for the efficient and equitable allocation of scarce resources. Thus far, no mechanism superior to the market has been discovered. The assertion that the price system is ineffective in instances such as local rail service because of government rate regulations cannot serve as an adequate reason for ignoring market mechanisms. The proposition that we must resort to a less effective allocation device because of a prior government intervention is preposterous. Rail rate regulation is not an irreversible fact of nature. Treating it as if it were can only make things worse.
The assertion that there are externalities that are not reflected in the price system may be true, but it cannot serve as a blanket endorsement of nonprice allocation schemes. The rising suspicion that government interventions may be reducing the net general welfare has produced well-meant concern. Advocates of continued and expanded interventions are pressed to prove that these interventions are producing net benefits. The bureaucracy has been badgered to come up with cost/ benefit evaluations of government programs that can prove the worth of these interventions. One should, then, not be surprised that such methodologies as are adopted will generally “prove” the worth of government intervention.
The most unfortunate consequence of the cost/benefit charade is the ultimate cost of the inevitable misallocation of resources. The only assured result of subsidy programs is a lower return on investment. Resources will be diverted to less productive ends. Over time, the effects of compounding this lower return will mean much lower standards of living and much lower social benefits as well. This is hardly the result advertised or anticipated by proponents of government intervention for the purpose of promoting the social welfare.