Why Equity Should Be Considered in Cost-Benefit Analysis

Equity and Efficiency

Any productive process can be considered in two separate lights, one, the quantity of production, the other who gets the benefits, or the "distribution". In economic analysis these are generally considered separately under the respective headings of efficiency and equity. Efficiency is concerned with quantity produced for a given input of resources, whereas equity is concerned with the fairness[1] of the resultant distribution of income or benefits.

Consideration of Equity In Economic Analyses

Any analysis which considers the equity of a change must Involve the comparison of the change in benefits (or income) received. If the benefits (or income) are evaluated in monetary terms, assumptions must be made as to the value (or more strictly the utility) each individual places upon the same monetary unit. In other words when considering equity, interpersonal utility comparisons must be made. However there is no objective way to make such comparisons.

Justifications for ignoring equity

This lack of objectivity caused by the consideration of equity in economic analysis is usually overcome by ignoring equity and considering efficiency only. There have been several justifications of this approach.

The change in this distribution caused by any activity subject to analysis will be small, and the changes to individuals' incomes will be small and occur randomly. So when all activities are considered any loss to an individual will be offset by other gains.
If the distribution of income (or benefits) is unsatisfactory corrections will be made in the political field by changing the tax structure so as to alter the incidence of transfer payments.
The existing distribution of income is satisfactory. That is an income distribution which is accepted as being socially desirable will arise through the creation of a tax structure, which involves transfer payments, through the political process. It can be argued that a progressive income tax will offset the decreasing marginal utility of income so that every unit of net income has the same utility.

Pareto Criterion for an Improvement

The only unambiguous criterion for an improvement is that put forward by Pareto, namely that at least one person is made better off while no-one is made worse off. Any other criterion will involve the making of interpersonal comparisons. In the interests of simplicity and practicability simplifying assumptions are made.

Kaldor-Hicks Criterion

In cost-benefit analysis the criterion for an improvement is that put forward by Kaldor and Hicks, namely that there is a net gain (i.e. the gainers could compensate the losers, whether or not they do actually compensate them). This criterion depends upon the assumptions that the marginal utility of money is constant and that the utility different people attach to money is comparable.

Pigovian Social Welfare function

The ignoring of equity and the much simplified criterion for a change to be considered an improvement have led to the wide scale use of what Foster (1966) has called the "Pigovian social welfare function" in cost-benefit analysis. Foster defines a decision (or social welfare) function, SWF, as one which is logical, states (i) to whom the benefits accrue (ii) who bears the costs and (iii) the weights to be given to the various components of (i) and (ii). The "Pigovian SWF" is that where all (i) and (ii) are taken into account and (iii) the monetary values are all given equal weight.

However the reasoning which leads up to the adoption of the 'Pigovian Social Welfare function' in cost benefit analysis is extremely tenuous.

Inadequacies of not making inter-personal utility comparisons

The consequences of not making inter-personal utility comparisons ".....were pointed out in 1932 by Professor Robbins (1932). He maintained that if economics were to have the objectivity of a science, economists may not make inter-personal comparisons and may not, in their capacity as economists, argue for or against any policy or change of policy that would make some people better and others worse off than they were before. Considering that practically every economic change favours some and hurts others, Professor Robbins was, in effect, baring himself and his colleagues from any policy recommendations whatever"[2] .

Inadequacies of ignoring equity

The argument that any losses which result from a change are small and random is open to considerable doubt, they are just as likely to be large and cumulative. For example any redevelopment or road building in a city will probably tend to give most benefits to the richer suburban residents and most costs will be borne by the poorer inner city residents. This is due to two causes. Firstly the poorer sections of the community do not have the political power to ensure they do not suffer dis-benefits from redevelopment or road construction, secondly most of the building will take place in the central areas, and it will be the inner city residents who are disturbed the most.

The arguments that the existing income distribution is satisfactory or will be made satisfactory by the political process is also open to doubt since there is not a uniform distribution of political power. Those in an economically favourable position also tend to hold political power. They will be unlikely to agree to any change in distribution.

Inadequacies of the Kaldor-Hicks Criterion

The Kaldor-Hicks criterion is unsatisfactory since whenever it is used the implicit assumption that money has the same utility for everyone, is made. This is an inter-personal utility comparison and, as such, it is a value judgment. In consequence every time the Kaldor-Hicks criterion is used an implicit value judgment is made. Also, it is not self evident that there is an improve ment if anyone is worse off, which is allowed by the Kaldor-Hicks criterion. However, even if this were overcome by stipulating that everyone who suffers a loss must be compensated, "the notion that .... any unfavourable effects on income distribution can be overcome by making some of the gains compensate some of the losers is rarely applicable in practice." (Prest & Turvey 1965).

Inadequacies of Pigovian income distribution

The use of the Pigovlan social welfare function depends upon the assumption that the resultant income distribution is satisfactory, but as Foster (1966) has pointed out "most writers on cost-benefit analysis have recognised that the income distribution that results from the use of the Pigovian function has no apriori validity, and that it is possible and sometimes politically desirable to modify the function so as to bias project selection in favour of some other income distribution".

Conclusion

Equity has in the past been ignored in decision making processes which are based upon cost-benefit analysis, in an attempt to give objective advice, since considering equity requires that subjective inter-personal comparisons be made. However the arguments used for ignoring equity are based upon the implicit assumptions that utilities are comparable and that the Kaldor-Hicks criterion is 'acceptable' which are themselves subjective judgments. Consequently no 'objective' advice can be given, and any advice which is given will be better if equity is explicitly considered.

[1] The usage of the word equity in economics is somewhat different from the dictionary definition.
equity (ek'witi) (O.F. equilte, L. aequitas - ta tem, from aequus, fair, n Justice, fairness; the application of principles of justice to correct the deficiencies of law.....
(Cassell's New English Dictionary 1949)
In a paper on "Equity" Scitovsky (1964) says:
It is no accident that the English word equity comes from equus. Latin for equal. By equity people mean, if not equality, at least something that approximates it closely enough to satisfy them. The public is satisfied with something short of equality, partly, perhaps, because it is resigned to this being an imperfect world, and partly also because it recognizes the impracticability of perfect equality. The latter is unattainable as long as we need to provide economic incentive to produce the national product.
Here equity is taken to mean the fairness of the distribution of income or benefits, but the decision about what is fair must be made elsewhere.

[2] From a paper on "The State of Welfare Economics" (Scitovsky 1964, pp174-189).