A Critical Look at Rising Homeownership Rates

George Masnick, Nancy McArdle, Eric Belsky

W99-2: In spite of warnings to exercise caution when using the Current Population Survey to track trends between 1993 and later years because of major changes since 1994 in the way in which CPS data were sampled, collected and weighted to produce national estimates, housing analysts have pretty much taken recent homeownership trends on face value. These trends show a large increase in the homeownership rate, especially between 1994 and 1996, that most observers have accepted as a logical outcome of a good economy, favorable mortgage interest rates, and aggressive public policy initiatives to increase homeownership. A closer look at this recent upward trend in homeownership reveals some anomalies that are difficult to explain. First is the substantial decline in the number of renter households that took place between 1994 and 1996, without an equally large change in the number of vacant-for-rent units on the market. Second is the observation that the prior decline in homeownership took place unevenly across households stratified by age of head and family composition, while the upswing has been evenly spread across all age groups and broad family types. Third is the fact that the geographic distribution of the changes in homeownership since 1994 is not always consistent with where increases in homeownership might have been expected to occur because of economic trends or public policy initiatives. While we believe that homeownership rates have certainly increased in recent years, we conclude that several changes in CPS methodology between 1993 and 1996 very well could have exaggerated this change…