Efficiency versus Equity in Economic Policy Analysis

Abstract

In this paper we outline a new approach to welfare economics. The purpose of welfare economics is to provide an ordering of alter native economic policies. The most desirable economic policy is the policy yielding the highest level of social welfare. This principle can be used to evaluate a specific policy change or to select the optimal policy from a set of alternatives. Dupuit (1969) originated the appraisal of alternative economic policies in terms of their impact on consumer welfare. He proposed to measure individual welfare on the basis of preferences revealed by consumer behavior. The prices faced by the consumer and the corresponding quantities demanded were used to obtain estimates of consumer's sur plus.1 Hicks (1942) introduced measures of consumer's surplus based on compensating and equivalent variations in income or total expenditure. The intuition underlying Hicks's approach to welfare economics is straightforward. Lev els of welfare before and after a change in economic policy can be ordered by compar ing the required levels of total expenditure. For Hicks's compensating variation the dif ference in total expenditure is evaluated at prices prevailing after the change in policy. For the equivalent variation the difference is evaluated at prices before the policy change. Chipman and Moore (1980) have shown that a necessary and sufficient condition for Hicks's compensating variation to provide an appropriate ordering of economic policies is that individual preferences are homothetic. Homothetic preferences are inconsistent with well established regularities in the behavior of individual consumers, such as those reviewed by Houthakker (1957). Chipman and Moore recommend Hicks's equivalent variation. The individual expenditure function intro duced by McKenzie (1957) provides the sim plest approach for implementing Hicks's measures of welfare. The expenditure func tion gives the minimum level of total expend iture required to attain a stipulated level of utility as a function of the prices faced by the consumer. This level of expenditure can be derived from the indirect utility function, which gives the maximum attainable utility level as a function of prices and total expend iture. Jorgenson, Lau, and Stoker (1980, 1981, 1982) have developed methods for construct ing indirect utility functions and individual expenditure functions for a population of consumers. Exact measures of compensating variations based on these methods were intro duced by Jorgenson, Lau, and Stoker (1980). The corresponding exact measures of equiv alent variations were introduced by Jorgen son, Lau, and Stoker (1981). We describe these measures in Section 2 below. The approach to welfare measurement originated by Dupuit is limited to individual welfare. Under the Pareto principle a change in economic policy can be recommended if all consuming units are at least as well off under the policy change and at least one consuming unit is better off. This principle provides a partial ordering of economic poli cies. To obtain a complete ordering we re quire the concept of a social welfare function originated by Bergson (1938) and discussed by Samuelson (1982-1983). A social welfare function gives the level of social welfare as a function of the distribution of individual welfare over the population of consumers. Social welfare functions incor porate the effects of changes in economic policy on the welfare of individual con sumers. A requirement often imposed on so cial welfare functions is that they must obey the Pareto principle. Social welfare functions also include the impacts of policy changes on