The Effect of Length of Run on Measured Demand Elasticities

Abstract

for use, its effect on the demand for stocks is not as widely recognized. The following analysis considers the effect of length of run on both of these components of the total demand for a commodity and then brings these results together to evaluate their joint effect on measured demand elasticities. Demand for use is shown to be more elastic in the long run when the relationship between commodities is either one of substitutability or complementarity. Also, conditions are presented under which the demand for a commodity for storage is more elastic in the short run. Finally, the condition necessary for the total elasticity of demand to have a U shape with respect to length of adjustment period is specified. IT IS generally accepted that the demand for most commodities is likely to vary for different lengths of adjustment period. This implies variation in the elasticity of demand associated with time.1 Although the elasticity of demand is often considered to vary directly with time, empirical results obtained by Tomek and Pasour provide an example in which demand appears to be more elastic for a shorter length of run. Tomek estimated the price elasticity of demand for fresh apples at the farm level using annual data for the period 1947-1962. This estimate was -.46.2 Pasour, using data for a similar period of time, divided the marketing year into three periods and computed seasonal elasticities at the farm level. Although a satisfactory relationship was not obtained in one of the periods, the coefficients of price elasticity of demand in the other two periods were -.75 and -1.33, both apparently higher than the estimate obtained by Tomek.3 Shepherd, in considering the effect of time, states that there are two opposing forces affecting the elasticity of demand.4 On the one hand, short-time elasticities (e.g., day, week, or month) are likely to be greater than longer-time elasticities (e.g., year) since a larger proportion of the * Published with the approval of the director of research as Paper No. 1971 of the Journal Series. The authors wish to acknowledge helpful comments from Paul R. Johnson and T. D. Wallace. The authors alone, however, are responsible for any errors.