Journal of Real Estate Finance and Economics, 18:1, 25±42 (1999)
# 1999 Kluwer Academic Publishers, Boston. Manufactured in The Netherlands.
The Long-Run Elasticity of New Housing Supply
in the United States: Empirical Evidence for 1950
DIXIE M. BLACKLEY
Department of Economics, Le Moyne College, Syracuse, NY 13214-1399,
The long-run price elasticity for alternative speci®cations of new housing supply is estimated using U.S. annual
data for 1950 through 1994. The basic model expresses residential construction as a linear function of new
housing price and the prices of construction inputs. Long-run elasticities range from 1.6 to 3.7, suggesting that
new housing supply is price elastic. Residential construction responds to both the real interest and expected
in¯ation rates, but other construction cost variables perform poorly. However, the results are sensitive to the time-
series processes underlying the variables. A modi®ed model that expresses residential construction as a function
of changes in input prices, rather than their levels, produces a long-run elasticity of about 0.8 and a signi®cant
inverse relationship between new housing supply and the construction wage rate.
Key Words: U.S. housing supply, residential construction
The supply of housing in the United States plays a critical role in determining housing
market outcomes, yet for a variety of reasons it has been the focus of relatively little
empirical analysis. Studies focusing on household behavior in the housing market are
relatively plentiful and address a wide range of issues, including the demand for housing
services, the demand for housing or neighborhood amenities, the demand for new or
existing homes, tenure choice, locational choice, and mortgage choice. Limiting his survey
to the market for housing service, Olsen (1987, p. 1015) reports that ``empirical studies of
the supply of housing service are as scarce as studies of its demand are abundant.''
Housing demand is affected by several factors, including changes in tax policy, the age
composition of the population, and the rate of household formation. However, the net
impact on housing prices and availability cannot necessarily be determined from demand
analysis alone. For example, the net effect of scaling back or eliminating the deductibility
of mortgage interest depends on both demand and supply responses. Similarly, Mankiw
and Weil's (1989) much-discussed prediction that real housing prices will fall by 47% by
the year 2007 re¯ect, in part, the underlying housing supply elasticity. Mankiw and Weil
®nd that changes in housing demand have a substantial impact on housing prices and
explain this result by suggesting that housing demand and supply are both highly inelastic.
Critics argue that housing supply is much more elastic than Mankiw and Weil assert (e.g.,