Introduction

In Chapters 4 and 5 we examined the various restrictive assumptions that would be required if consumer surplus techniques were to satisfy the conditions of an ordinal welfare metric. The following question then arises. Suppose that preferences do not satisfy the required restrictions. Do consumer surplus procedures nevertheless provide “reasonably good” approximations to a true welfare indicator? Basically, two approaches to this approximation problem have emerged in the literature. The first assumes that we know only the prices and quantities that hold in alternative situations, but not information about the shape of preferences or the consumer demand functions. It is based on the use of the Paasche and Laspeyres index numbers, and as we shall see, it has a close affinity to a consumer surplus formulation discussed by Harberger (1971). The second approach assumes that we know the shape of consumer demand functions, and it attempts to formulate conditions under which the traditional Marshallian surplus measure can be used either to rank alternatives or to approximate the true compensating or equivalent variations. Given the procedures that we have fully discussed in previous chapters, this second approach is really superfluous, because we can obtain exact (not approximate) representations of the CV and EV measures only if we are armed with knowledge of the demand functions. Nevertheless, it is important that we have a perspective on the attempts to use the Marshallian surplus.