|
Problems of Cost-Benefit Analysis
The major problema of Cost-Benefit Analysis are in the fields of enumeration, evaluation, the choice of a rate of time preference, and of project selection. Most of the problems are concerned with the evaluation of the costs and benefits.
Enumeration
Ideally all the costs and benefits associated with a project under study should be considered, however there are problems in doing so. The first is that it is often difficult to include all costs and benefits without any double counting, and the second is that some of the costs and benefits may be unforeseen, or be intangible. Some intangibles auch as deaths and injuries may be simple enough to enumerate, but others such as visual intrusion will be very difficult. In most transport cost-benefit analyses only direct user costs and benefits have in the past been considered. This has lead to widespread criticism of cost-benefit analysis for ignoring such things as the effects of transport projects on the environment and on groups of travellers not considered, such as cyclists and pedestrians.
Evaluation
Most of the benefits in a cost-benefit analysis are external to the body for whom the analysis is made. These external benefits are due to increases in consumers' surpluses. Their evaluation is based upon two assumptions. The first is, that along the small element of the demand curve over which the change takes place, the marginal utility of money is constant. To the extent that it is not the calculated value of the benefit will be in error. The second assumption is that utilities are comparable. This is the simplifying assumption mentioned in Chapter 3 by which questions of interpersonal comparisons are ignored. However it is in itself a comparison of the utility different people attach to money. The result is that cost-benefit analysis is based upon the subjective judgment that everyone attaches the same utility to the same monetary unit.
As mentioned in Chapter 3 evaluation is based upon market determined prices, since they are the only objective approximation to marginal social values. If the project is so large that it alters prices within the economy, the prices prevailing at the initial date will not reflect values at the end date. In this situation market prices must be used with caution. Theoretically a general equilibrium model would give a better indication of the effect of the project but it is impracticable in real situations since it requires full knowledge of all supply and demand curves.
'Second best'
For the market to produce an 'efficient' allocation of resources, prices should reflect marginal social values. If they do the ratio of price to opportunity cost should be the same in all sectors of the economy. There are, however, cases in any modern western economy where this condition does not hold. An example of this is given by wages. They will probably rise when there is a labour shortage but will not fall when there is a surplus, so preventing the market from achieving an 'efficient' allocation of labour.
In this situation of prices not reflecting costs the 'optimal' distribution of resources is known as the 'second best' solution. However in an article on welfare aspects of cost-benefit analysis, Krutilla (1961) found that there were even arguments that denied the existence of a 'second best' solution[1] , but he carried on to offer several solutions to this problem. Two were criteria for considering whether changes were, or were not, improvements. One was put forward by McKean (1958). It considers that an Improvement is likely to be achieved if production is increased where price exceeds cost. The other is that of Eckstein (1958) following Little (1950), that in most public utility cases prices are a good approximation to marginal costs, and will suffice for decisions, but in some situations (when assumptions are not met) other measures of costs and benefits must be used.
Krutilla concluded that there is no theoretically correct solution and that although academics should continue to say so, the practising economist in government will "..... be grateful for ..... even a perforated rationale to justify recommendations "in the public interest" " (Krutilla 1961).
On a more practical level Foster & Beesley (1963), in their study on the Victoria line, were critical of the relative prices between road and rail transport and considered that road transport was under priced in relation to rail and so produced a non-optimum allocation of resources between the two. They stated that "..... if the Government should decide ..... that these are the "right" relative prices, then it would be logical if it wanted them reflected in investment policy. This would imply that investment would be biased towards heavy urban road investment even if it should be cheaper to build or improve railways. However, the Government might take up an intermediate position declaring that the prices were 'right' because a change would be politically undesirable or impossible, but that investments should be made as if the prices were corrected so that both reflected the real costs involved ..... The last possibility is that the Government acts to correct the present relative prices, in which case investment would proceed on the assumption of the corrected prices" (Foster & Beesley 1963, p58).
Finally, when considering 'second-best' situations caused by market imperfections "Only those divergencies which are immediate, palpable and considerable ..... deserve our attention" (Prest & Turvey 1965).
Intangibles
Having found a suitable measure for the enumeration of intangibles such as deaths or visual intrusion a method must be found to give them some value. The basic problem is that, strictly, monetary values are only valid if they are determined in an exchange situation. This poses an extreme problem when evaluating intangibles because they are not traded in exchange for money. Several methods of evaluation have been suggested, however as they all produce a result which is subjective, none have the same degree of objectivity as market determined prices.
As an example one method put forward is that for determining the value of life. The value which decision makers place on life can be imputed from decisions which involve the implementation of schemes, that involve the saving of lives. Once all other benefits have been accounted for any excess expenditure can be attributed to the saving of lives. If the number of lives saved can be determined the 'value' of each life saved can then be implied.
There are several drawbacks to this procedure. If it is used to determine the value of all intangibles in every decision, the only values available will be those used in the first decision that was made in this way. Second no more than one intangible can be involved in the decision which is used to imput its value without entering into other problems.
Value of time
One of the major intangibles in transport cost-benefit analyses is time saving. In fact in most studies it has been found to be the major benefit. Although the enumeration of time savings are simple enough their evaluation is difficult.
Savings of time made during a person's working hours are generally valued at that person's hourly wage rate. This is on the basis that the employer, and so the community, receives this amount of benefit. It is however only valid if the wage rate accurately reflects the marginal value of the time saved and that the time saved is actually used in a productive manner. The value of 'leisure' time presents more problems. There is no direct way in which to value it, so some indirect way has to be found. These methods will imply a value from any situation in which it is assumed travel time savings are being traded off against some other commodity which does have a price. Examples are time savings due to a faster but more expensive route or mode of transport, or due to living in a more expensive house which is closer to place of work. One of the many problems with these methods is that time saving is not the only consideration when making these types of choice.
At present, probably due to the lack of any concrete evidence to the contrary, the value of time savings are assumed to be proportional to their length, that is, a time saving is valued at the same rate regardless of how long it is, so that the net benefit of sixty people making a one minute saving is considered to be the same as that of one person making an hour's saving. There is also the possibility that time-savings of the same 'type' may have 'values' which vary due to the time of day or the purpose for which the trip is made.
Time preference
The effect of the choice of a rate of time preference is to determine the relative importance between investment in the present and future consumption, and also between consumption now and in the future. In this respect benefits can be considered as consumption. The higher the rate of time preference the less important is future consumption as compared with consumption at the present. For example a discount rate of two per cent would value a benefit of one pound incurred in 114 years time at 10 pence whereas at 10% a one pound benefit would be valued at approximately 10 pence if it were incurred only 23 years hence. This means that the higher the rate of return the shorter the time over which the project need be considered since any costs or benefits incurred after the end date can be considered to be of minimal importance since they have such a low present value.
At present with the prevailing high rates of interest, the period of time over which it is meaningful to consider a project is in the region of 20 to 30 years. However, the use of such high discount rates can be considered as a very short sighted approach to decision making. This is a reflection of the real problem of the selection of a rate of time preference which stems from the possibility of a divergence between the social time preference, social opportunity cost and financial rates of return.
A financial rate of return reflects the collective 'private' time preferences of the owners of capital, whether they are individuals, institutions or government. There is no reason to suppose that it should be the same as the social time preference of the community. In fact "some writers (Pigou (1932), Eckstein (1958) and Marglin (1963) believe that social time preference attaches more weight to the future than private time preference, and that it is the former which is relevant for determining the allocation of society's current resources between investment and consumption". (Prest & Turvey 1965).
However as Prest and Turvey mention there are two difficulties in using a social time preference rate of return. One is to determine the rate[2] and the other is that if the public and private sectors both act in the same field (i.e. state coal and private oil), non-optimal resource allocation will result.
The rate used in most calculations is the government borrowing rate which is taken as a reflection of the social opportunity cost rate. The problem of choosing a rate usually receives little attention and as Prest & Turvey (1965) conclude "The truth of the matter is that, whatever one does, one is trying to unscramble an omelette, and no-one has yet invented a uniquely superior way of doing this".
Project Selection
Project selection is a simple enough process provided a method of selection (such as choice of the projects with highest benefit-cost ratio until the budget is exhausted) can be agreed, and that there are no mutually exclusive schemes. If there are, "one project may have a lower benefit-cost ratio, yet will be preferable if the extra benefits exceed the extra costs". (Prest & Turvey 1965, p704) The real problems of mutually exclusive schemes occur when projects are interdependent. That is if the size of the costs or benefits of one scheme depend upon whether or not another scheme is implemented. In this case both schemes should be considered separately and together as three mutually exclusive projects. In an urban context a transport plan can be considered to consist of a very large number of interdependent schemes which may or may not be implemented. To be rigorous all the combinations of possible schemes should be considered, within any budget constraint that may exist. This would produce an impossibly large number of schemes to be considered, since each would separately require the running of the transport model to produce the estimated travel patterns on which the cost-benefit analysis is based.
Accuracy
Finally there are considerations of how accurate and value free cost-benefit analysis is as a tool of decision making. Firstly, unless constraints are imposed cost-benefit analysis can not be used to evaluate policies other than on the basis of what is the most efficient use of resources. In its ' ideal' form it merely tries to reproduce the effects of market forces in areas in which they do not act. It is not a suitable method of analysing schemes which are created to satisfy policies which have non-economic ends. That is, ends which involve the production of a result which is intangible, such that they cannot be measured in physical terms, or if they can be measured, can not be valued satisfactorily in monetary terms since they are not traded in a market.
The use of cost-benefit analysis in the consideration of programmes within any field will tend to discourage the introduction of any policies other than that of economic efficiency which is implicit in cost-benefl analysis.
There are several areas in cost-benefit analysis where there is room for considerable variation. Just two of these are in the values attached to intangibles and the choice of the rate of return. Nose of these choices are value free and they may well affect the outcome of the analysis. The extent to which the results of an analysis can vary is illustrated in an example given by Prest & Turvey (1965). It is about calculations on the cross-Florida barge canal[3] in which the Corps of Engineers produced a benefit-cost ratio of 1.20 and consultants produced a ratio of 0.13. Prest and Turvey comment that "To what extent the divergence is due to the facts that the Corps likes to build canals and that the consultants were retained by the railroads, and to what extent it is due to the intrinsic impossibility of making accurate estimates is left entirely to the reader to decide".
Conclusion
Firstly cost-benefit analysis is only a method of measuring how 'efficiently' resources are allocated. It cannot consider 'non-economic' ends such as the full effects on the environment or the effects on future land development.
Secondly and probably most importantly, it is based upon assumptions of the comparability of utilities and the acceptability of the Kaldor-Hicks criterion for umprovement. That is, that an improvement will be made if there is a net gain even if this is at the expense of some people suffering an uncompensated loss. So from the very start cost-benefit analysis cannot be considered as being 'objective'.
Thirdly, enumeration, evaluation and the choice of a rate of time preference all contain difficulties. All three contain subjective elements, and margins of doubt and error, so the results are liable to be non-conclusive.
Cost-benefit analysis is therefore a subjective and error prone method for assessing how schemes measure up to 'efficiency' criteria. However as Prest & Turvey (1965) conclude in their review; "The case for using cost-benefit is strengthened, not weakened, if its limitations are openly recognised and indeed emphasised".
|
Quick Reply
