New estimates of the elasticity of substitution of land for capital

We reconcile conflicting evidence on the magnitude of the elasticity of substitution of land for capital, which is a key determinant of the relationship between the price of land and the density of land use. We first compare the performance of classic estimation approaches with a new estimation procedure using a series of Monte Carlo experiments. We then apply the approaches to various real-world data sets drawn from Berlin, Chicago, and Pittsburgh. Our results indicate that many existing estimates are likely to be biased downward, and the true elasticity is likely to be closer to one than widely believed in the literature. The results suggest that a Cobb-Douglas form is a reasonable approximation to of the production function for housing.