DEMAND AND SUPPLY OF HOUSING IN THE U.S. 1929–1978

Annual data for the period 1929–1978 are used to estimate a complete model of demand and supply of housing services and consumption goods in the U.S. by maximum likelihood methods. The demand functions are derived by maximizing a utility function characterized by weak inter‐temporal separability. Utility in each period is assumed to be of a generalized CES type with housing services and consumption goods as arguments. The estimating demand relation is transformed into a relatively simple form by focusing on the relative demand of housing to consumption goods. It is found that the intra‐temporal elasticity of substitution between housing and consumption goods is unitary. The maximum likelihood estimates of the other parameters of the elementary utility function are also presented. Finally, it is noted that by estimating the structure of individual preferences a basis is established for the calculation of long‐run efficiency gains of a change in the tax treatment of housing.