Specification Error in the Estimation of the Elasticity of Substitution between Capital and Land in residential Construction.

Previous estimates of the elasticity of substitution between capital and land ignore the role of labor and financing in the production function for housing. In the case of cross-section data between cities, time-series data, or pooled cross-section and time series data, the exclusion of one or both variables may lead to specification error. This study tests the hypothesis that the exclusion of these variables does not bias the estimate of the elasticity of substitution between capital and land. The methodology employs data for U.S. cities from 1973 to 1982 to estimate translog cost functions. Exclusion of financing biases the elasticity estimate upward.