The test which KQ employ to demonstrate the existence of supply restrictions is based on the relative lack of single family houses available to blacks, pp. 270-72. Thev find, in a sample which uses cities as observations, that the extent to which the actual black homeownership rate falls short of the expected rate is negatively correlated with the proportion of central city housing which is single family.1 As KQ acknowledge, theirs is not a complete test of the supplyT restriction hypothesis. This comment provides further tests of that hypothesis. A relative lack of single family houses in the black housing submarket2 causes lower black homeownership because apartments are not generally available for owner occupancy and single family houses are usually owner occupied. However, some single family houses in black areas may not be available for owner occupancy due to difficulties which blacks face in the mortgage market. The procedure followed in this paper is to make a correction for housing stock characteristics. Assuming the correction is made properly, the remaining difference between the black and white homneownership rates is due to the fact that blacks have more difficultv obtaining mortgages on single family houses. Section I presents a linear probability-ofhomneownership function with a correction for housing stock characteristics. Section II presents a model in which the type of tenure (own or rent) and the structure type of a household are jointly dependent variables. Section III examines some implications for policv.
Stored only when both papers are in the corpus — 176 citations exist upstream.