Permanent and transitory income in models of housing demand

Abstract

Permanent income has for some time been recognized as the appropriate income variable for models of housing demand. This paper examines a recently developed model wherein permanent and transitory income are considered to be the fitted and residual components, respectively, of a regression of actual income on several household characteristics. An important caveat for these models is pointed out. One logical remedial strategy is shown to result in underidentification for the coefficient on permanent income. However, the technique uncovers a tangential result which summarizes why an alternative class of models can be expected to underestimate income elasticities.