Intra-Metropolitan Mobility, Residential Location and Homeownership Choice Among Minority and White Households: Estimates of a Nested Multinomial Logit Model

Recent academic and policy analyses have sought to explicate the persistently depressed levels of black and Latino homeownership. While prior research has focused largely on racial disparitiesin household endowments (see, for example, Bostic and Surette (2001), Gabriel and Painter(2001), Painter, Gabriel, and Myers (2001), Wachter and Megbolugbe (1992), Gyourko and Linneman (1996), and Coulson (1999)), few studies have jointly modeled the structure and determinants of the household mobility, residential location, and homeownership decisions. The intra-metropolitan mobility and residential location choices of minority and white householdsmay vary considerably, owing in part to the different endowments, constraints, and locational preferences of those groups. An improved understanding of the linkages between those decisions and housing tenure choice may yield new insights and better-informed policies to enhance minority homeownership. This paper estimates a three-level nested multinomial logit model of household intra-metropolitanmobility, residential location, and homeownership choice. In so doing, the study applies individual level 1990 Census data to test relevant economic, demographic, and neighborhood hypotheses. The model is then simulated to assess the effects of changes in household endowments, neighborhood racial composition and other amenities on the intra-metropolitan mobility, residential location, and tenure choices of minority and white households. Research findings indicate significant variability in intra-metropolitan mobility, residentiallocation, and tenure choice among white and minority households. The inclusive values of the three-level nested logit model are statistically significant, indicating the appropriateness of the tiered specification of household mobility, residential location, and homeownership decisions. Simulated shocks to household endowments and neighborhood characteristics reveal varied effects across the racial groups and locations. For example, attribution of white endowment characteristics to black households serves to appreciably raise black homeownership rates in virtually all Los Angeles area counties—so as to close the white-black gap in homeownership bya full 17 percentage points. In the context of this shock, black rates of homeownership move up to 41 percent in the Los Angeles metropolitan area (compared to 53 percent for whites), reflecting strong homeownership gains in the relatively higher income counties of Los Angeles, Orange, and Ventura. A similar shock to the incomes of Latinos serves to elevate their area-widehomeownership rates to 47%, whereas little homeownership change derives to Asian householdsvia such an income shock, given their already high levels of economic endowment. Other simulated effects of changes in neighborhood characteristics, including shocks to house prices, rents, amenities, and minority population representation, are evidenced with respect to theirimpacts on residential location and homeownership choice. For example, a simulated increase inminority population shares in the Inland Empire serves to perceptibly enhance the dispersion of black and Latino populations (particularly renters) to suburban areas, but provides less immediate support as regards the minority homeownership goal.