Abstract
This paper, in the neoclassical tradition, uses house purchase data to estimate income and (indirectly) price elasticities of demand. Grouping of data is a familiar means of 'washing out' transitory variations in measured income. Errors in the independent variable would be expected to bias downward the elasticity estimate, but recent work demonstrates that incorrect grouping may yield overestimates. Of the aggregation methods used, the 'correct' method yields results very close to those from the ungrouped data. A conventional demand analysis then follows which points to income inelasticity of demand. If price is proxied by location, then demand seems superficially price elastic. However, there are suggestions, with caveats, that income elasticity may vary with income and further work on specification seems called for.
Get full access to this article
View all access options for this article.