Understanding the impacts of demand shocks on the price of owner-occupied housing: A study of the supply elasticity of housing

Date of Award


Degree Type


Degree Name

Doctor of Philosophy (PhD)




James Follain


Mortgage, Interest, Tax deduction, Prices, Housing, Supply elasticity

Subject Categories



Most believe that elimination of the mortgage interest deduction will, at least in the short run, decrease the value of owner-occupied housing. What is unknown is the ability of house prices to adjust to such shocks in the long run. The answer to this question depends critically on the supply elasticity of housing and the speed of adjustment of house prices to their steady state values.

The goals of this study are to offer new evidence regarding the supply elasticity of housing, speed of adjustment of house prices to the steady state, and regional variation in the either of these two components. This is accomplished by using vector autoregressions (VAR's), with pooled metropolitan area data. The pools are currently defined by coastal versus non-coastal, constrained versus unconstrained, and all metropolitan areas in the sample. The variables used in the vector autoregressions are chosen based on a simple model of demand and supply, with housing price, housing starts, one exogenous measure of supply (construction wage), and one exogenous measure of demand (either income, user cost, or population). Impulse response functions generated from the VAR's are used to make inferences regarding supply elasticities.

Currently, two major conclusions may be gleaned from these results. The first is that supply elasticity and speed of adjustment vary significantly based on location of metropolitan area. When using either population or income demand variables in pools of coastal/non-coastal or constrained/unconstrained MSA's, supply elasticity is very inelastic and prices are slow to adjust for coastal and constrained areas, versus very elastic and quicker to adjust for non-coastal and unconstrained areas. The second is that pooling at the national level (all metropolitan areas) produced results that would indicate that supply is very inelastic and prices are slow to adjust. The degree of heterogeneity induced by pooling all metropolitan areas together, however, casts suspicion on this particular result.

Finally, new panel bootstrapping techniques are used to produce standard errors and confidence intervals for impulse response functions. Two standard error bands of confidence indicate a high degree of precision of the impulse response functions.


Surface provides description only. Full text is available to ProQuest subscribers. Ask your Librarian for assistance.