Patterns of White Avoidance, Residential Mobility, and Integration in Diverse Cities and Communities

Daniel Lichter, Cornell University
Domenico Parisi, Mississippi State University

Our goal is to shift attention from the usual concerns about residential integration and attainment among ethnoracial minorities to emerging patterns of spatial integration among whites in America’s growing multiracial communities and neighborhoods. Our fundamental research questions are two-fold: (1) Do U.S. whites avoid or move into racial diverse cities and suburbs; and (2) Are emerging patterns of white avoidance and exposure segmented by socioeconomic status (i.e., education, income, and family structure)? To accomplish these goals, we use proprietary household data from the Panel Survey of Income Dynamics linked with population data from the U.S. decennial censuses and American Community Survey. We provide, for the first time, national estimates of the residential mobility of whites into and out of racially-diverse metropolitan areas, cities and suburbs, and blocks. By nesting blocks within places (i.e., cities and suburbs), we can map and model white mobility from places undergoing rapid racial change at multiple levels of geography. We examine whether whites are moving to predominately white blocks, even as surrounding suburban places and the broader metropolitan area have undergone significant increases in racial diversity over the study period. We provide an empirical benchmark of diversity using the entropy measure (E), which is a multiracial measure of diversity that takes into account the presence of more than one racial or ethnic group in the city. Unlike previous studies, we identify streams of movers, focusing in particular on both the origin and destination of white movers, broadly conceived. That is, we identify the racial composition of the origin and destination metropolitan areas, cities and suburbs, and blocks of both inter- and intra-metropolitan white movers. Previous studies have typically excluded the residential attainment of new in-migrants from other metropolitan areas.

            Theresidential attainment literature overwhelmingly focuses on the residentialmobility of blacks or other racial and ethnic minority populations intopredominately white or middle-class neighborhoods (as measured by medianincome, housing values, or low crime rates).  In this paper, we shift attentionfrom the usual concerns about residential integration and attainment amongracial minorities to emerging patterns of spatial integration and segregationamong whites in America’s growing multiracial communities and neighborhoods(Frey 2013).

We asked two questions: (1)Do U.S. whites avoid or move into racially diverse cities and suburbs; and (2) Iswhite exposure to diversity expressed within places—at the block level?  Weargue that white residential mobility involves two distinct but interconnected processesof racial sorting which operate at different spatial scales.  Residentialintegration and segregation is revealed in macro processes at the place level andin micro processes at the neighborhood level.  Specifically, racial sorting involves(1) residential mobility between places (i.e., between origins anddestinations with different racial compositions), and (2) neighborhood residentialattainment within destination places (i.e., moving into segregatedneighborhoods).

Data and Methods.   Toaccomplish our goals, we use newly available proprietary household data fromthe Panel Survey of Income Dynamics (PSID) linked with population data at themetropolitan, place, and block level.  We concatenate 20 years of person-yearrecords from multiple waves of the PSID to spatial data from the U.S. decennialcensuses and American Community Survey.  The PSID began in 1968 with anationally representative sample of about 18,000 individuals living in 5,000 households.Measuring the Dependent Variables. Weoperationalize racial diversity at the place level using the five-group entropyscore (Fowler, Lee, and Matthews 2016; Lee et al. 2014; Parisi et al. 2015). Thefive pan-ethnic groups are non-Hispanic white (hereafter white), non-Hispanicblack (black), non-Hispanic Asian and Pacific Islanders (Asian), Hispanic, andother. Indian/Alaskan Natives are included in the other category, as are peoplewith racial statuses not identified by census categories.  Following Hall etal. (2016), we also identify three types of blocks based on racial composition.Whites may move to blocks that are predominantly white, i.e., where 90 percentor more of the population is white. Whites may move to white/other race blocksthat are majority white or, alternatively, majority minority. Finally, whitescan move to blocks where the majority population is black. Specifically,majority-black places are where whites and blacks exceed 10 percent of thepopulation and where no other racial group accounts for more than 10 percent ofthe population living on the block.  PreliminaryAnalyses.  Wehave completed much of the preliminary analyses which are listed in Tablesbelow, along with two illustrative tables.  (1) Patterns of ResidentialMobility within and Between Places. This analysis is reported inTables 2-4. Table 2 provides patterns of mobility for whites moving but stayingwithin a suburban place. Table 3 provides patterns of mobility for whitesbetween places with different levels of diversity. Table 4 providespatterns of mobility for whites moving to a block within another place.  (2)Determinants of White Flight from Diversity.  We estimate two preliminarysets of models, which sets the stage of the subsequent analyses (in Tables 7-9).Table 5 provides the determinants of block out-mobility, regardless of whetherthe destination to another block is within the origin place or in anotherdestination place. Table 6 reports the analysis for whites who moved to anotherblock within their origin place or to a block within another destinationplace.  (3) Determinants of Block Destination Selection amongWithin-Place Movers. This analysis is reported in Tables 7-9. 

   

Presented in Session 1220: Internal Migration and Urbanization