It is easy to see that an economic welfare measure that is free of this 

 defect is highly desirable. Hicks (1943) provided two such measures (actually 

 four; however, only two of the measures of consumer surplus that Hicks suggested 

 are widely used today). These two measures distinguish sharply between the level 

 of consumer welfare before and after an economic change. 



One of these measures is called the compensating variation. This is the 

 amount of money income that must be taken away from a consumer, after a change 

 in some economic variable, in order to make him exactly as well off as he was 

 before the change took place. The equivalent variation is the amount of money 

 income that must be taken from a consumer in order to make him exactly as well 

 off as he is after the change takes place (under the hypothetical presumption 

 that the change has not occurred). If the change is a fall in price, the 

 compensating variation must be positive (the consumer must lose income if he is 

 to be as well off as he was before the decline in price). 



The equivalent variation must be negative (Just et al . 1982) in the case 

 of a fall in price (the two measures will always have opposite algebraic signs). 

 These two estimates of consumer surplus are now known by the generic term of 

 willingness-to-pay approaches to consumer benefits estimates. Natural resource 

 economists have recently developed willingness-to-pay techniques for estimating 

 the benefits conferred by outdoor recreation sites and other natural resource 

 amenity values. 



The Hicksian measures are free of the internal defect (non-transitivity 

 for multiple price-income changes) defect of traditional Marshall ian consumer 

 surplus. But while it is flawed as a theoretical measure of net benefits 

 conferred, Marshall ian consumer surplus is based on market behavior and consumer 

 responses to recorded prices. Since economists have reason to believe that the 

 Marshallian consumer surplus measure provides a very good approximation (Just 

 et al . 1982) to the Hicksian consumer surplus values, the distinction is a moot 

 one for many practical purposes. 



Moreover, Willig (1976) has forcefully demonstrated that it is the income 

 effect of a change in price, combined with the non-zero price elasticity of the 

 demand curve, that generates the discrepancy between the Marshallian measure of 

 consumer surplus and the Hicksian measure. The relatively small size of the 

 income effect of the typical change in price implies that the consumer surplus 

 measures will not differ markedly. Willig (1976; see also Just et al . 1982) has 

 also shown that the (change in the) Marshallian consumer surplus calculated from 

 an empirically estimated demand curve for some good or service can be combined 

 with data on the income elasticity for the commodity to yield a more refined 

 approximation to the Hicksian (equivalent variation or compensating variation) 

 consumer surplus measures. However, use of the simple income elasticity 

 adjustment formulas discovered by Willig requires reasonably accurate empirical 

 estimates of the income elasticity of the demand (curve). Unfortunately, natural 

 resource economists have experienced remarkable difficulty in estimating the 

 income effect for travel cost (quasi-) demand curves (Walsh et al . 1987). 



It is very useful to understand the conceptual distinction between these 

 two welfare measures in many applied fields, and to have some grasp of the 

 protracted and bitter controversy that once surrounded the Marshallian consumer 



