Contemporary Economic Analysis (Routledge Revivals)

Papers Presented at the Conference of the Association of University Teachers of Economics 1978

David Currie, Will Peters

• Publisher: Routledge
• Year: 2016
• ISBN:
  9781317215097
• Edition: 1
• Topic: Business
• Subtopic: Business General
• Index: Business

First published in 1980, this book collects 17 lectures presented at the annual conference of the Association of University Teachers of Economics covering a wide range issues and debates. They include new theoretical points, criticisms of existing theory, the reporting of empirical studies and their implications, and refinements of methodological techniques. Among the topics covered are government deficits and capital accumulation; macroeconomic issues of management policy and foreign trade; empirical studies of foreign exchange markets, and supply and demand of hours of work; public sector and welfare economics; risk and uncertainty; and monopoly, competition and markets.

Public Sector Economics

7.
Where did Prescriptive Welfare Economics go Wrong?

D.J. Mayston

“It is a disease of the soul to be enamoured of things impossible”.

(DIOGENES LAETIUS, Bias, Book 1)

1 Introduction¹

At the time an earlier draft (Mayston, 1976a) of this paper was being written, I was located slightly closer to Philadelphia – the site of numerous sudden deaths attributed to the mysterious “Legionaires Disease”. At times one feels that large parts of prescriptive welfare economics have been struck down by a similar mysterious disease. If one surveys the battlefield of welfare economics in its encounter with real world problems, one is struck by the large number of dead analytical bodies one finds strewn through the literature, incapacitated by internal inconsistencies before ever meeting the enemy – the real world. In the literature on prescriptive welfare economics, and the associated theory of social choice, there are indeed a large number of papers devoted to demonstrating the non-existence of consistent social decision rules, of the kind policy-makers might consider to be desirable.

This is true, for instance, of compensation tests and many of the papers touching on the so-called New Welfare Economics. As discussed at length elsewhere (Chipman and Moore; 1973, Graaff, 1963), these became bedevilled by internal contradictions in the form of the paradoxes noted by Scitovsky (1941) and others. The logical problems produced by the New Welfare Economics led Chipman and Moore recently to conclude that:

When all is said and done, the New Welfare Economics has succeeded in replacing the Utilitarian smoke-screen by a still thicker and and more terrifying smoke-screen of its own.

(1973, p. 51)

Similarly, non-existence and logical paradoxes characterise Arrow’s Impossibility Theorem (1963) and much of the later work on the theory of social choice (see Mayston, 1974). In addition it is true of “deadweight loss” or “excess burden” measures of welfare loss, and similar consumer measures of welfare gain. As Silberberg (1972) and others (Burns; 1973, Mayston, 1974) have shown, these measures typically suffer from a dependence upon the arbitrary feature of the path of integration chosen for their evaluation. Hence, again we can produce inconsistencies in the welfare measure by choosing a different path of integration to any path initially chosen, thereby changing the social evaluation of the size (and possibly also the direction) of the welfare gain or loss.

One way out of this battlefield of problems, with inconsistencies strewn to the left of us and impossibility theorems to the right of us, was always thought to be use of a Bergson-Samuelson social welfare function. This conventionally takes the form:

where F is assumed to possess finite, continuous, first and second derivatives with respect to each uⁱ and u^j for all i j=1,…, m.² x = (x₁,…, x_r) here corresponds to a “social state” made up of r characteristics each at level x_j. (for j=1, r) and characterising all of the features of the state of society which are considered relevant. In many economic contexts we might then have that r = mn where m is the number of individuals and n the number of distinct commodities. We could then write in this case

for some i and k, where

denotes the level of consumption by individual i of commodity k in social state x. uⁱ on (1) is the ordinal utility index of individual i. Throughout we will assume that individual preferences are independent, in the sense that the preference ordering of individual i, which uⁱ represents, is determined independently of the preference ordering of any other individual j.

It was argued by Graaff (1963), and others, that a welfare function of the form (1) would exist, given the assumption of simply ordinality of individual preferences. If any subset of the uⁱ’s were transformed in a positive monotonie way, the form of F is simply adjusted to maintain the same social choices.³,⁴ Moreover, it was argued, particularly by Samuelson (1967), that the Bergson-Samuelson social welfare function (B-S swf) was not susceptible to Arrow’s Impossibility Theorem. Arrow’s theorem involves the condition of Unrestricted Domain which admitted all logically possible combinations of individual preferences. In contrast the Bergson-Samuelson social welfare function seeks only to operate for just a single given combination of individual preferences, represented by the given vector of ordinal individual utility functions (uⁱ,…, u^m). Hence, Samuelson argued, Arrow’s theorem does not disprove the existence of a Bergson-Samuelson social welfare function.

Recently, however, the view that there does exist such a welfare function under the assumption of simply ordinality of individual preferences has been disputed. In particular, R.P. Parks, (1976), Kemp and Ng,(1976), Robert Pollak, (1976). have produced impossibility theorems specifically on the existence of the Bergson-Samuelson social welfare function. In their paper (1976), Murray Kemp and Kwang Ng conclude:

… for any particular set of individual preferences chosen from a very broad class and for any particular set of social alternatives satisfying rather weak conditions, it is impossible to find a ‘reasonable’ Bergson-Samuelson swf based on individual orderings. (1976, p. 59).

The use of a Bergson-Samuelson social welfare function of the form (1) has become something of a corner-stone of many welfare analyses for resolving distribution questions (see, for instances, Bator, 1957). In addition, it has been widely used in applied areas, such as the op...

Table of contents

• Cover
• Title
• Copyright
• Original Title
• Original Copyright
• CONTENTS
• List of Contributors
• Editors’ Introduction
• THE HARRY JOHNSON MEMORIAL LECTURE
• THE FRANK PAISH LECTURE
• MACROECONOMIC ISSUES
• EMPIRICAL STUDIES
• PUBLIC SECTOR ECONOMICS
• RISK AND UNCERTAINTY
• MONOPOLY, COMPETITION AND MARKETS