Race and Sex Discrimination in Housing: The Evidence from Probabilities of Homeownership

In recent contributions, Kain and Quigley [4; 5] investigated the relationship between race and homeownership. Using a 1967 cross-section of St. Louis households, they concluded that blacks were significantly less likely to be homeowners than whites. This conclusion held not only for their entire sample but also for the subset of households that had moved recently. This is important since if discrimination is disappearing, one would expect this fact to show up in data using the latter subset much sooner than in the data for the entire sample. A subsequent paper by Roistacher and Goodman [10] estimated the parameters of Kain and Quigley's model using 1971 data from the Panel Study of Income Dynamics1 for the 24 largest SMSA's. They found that for the entire sample the difference between probabilities of black and white homeownership was almost twice as great as that estimated by Kain and Quigley, but that when the sample was restricted to the subset of families that