| Abstrakt: |
Background: The segregation of racial/ethnic minorities into distinct neighborhoods of cities has been a defining quality of American life for nearly a century. While ethnoracial diversity has altered the structure of segregation in many cities, and overall levels of segregation have visibly declined in others, U.S. communities and the neighborhoods that compose them remain highly stratified by race (Logan and Stults 2011). The impacts of this racialized residential context are both pernicious and pervasive: concentrating poverty and related social ills (Massey and Denton 1993), reducing home values and wealth accumulation in minority neighborhoods (Flippen 2001), eroding the quality of local public services, and magnifying the impacts of racial differences in educational attainment, family stability, vulnerability to crime, health and mortality (Cutler and Glaeser 1997). Given these substantial individual and collective costs, researchers have long sought to document emerging patterns of and trends in residential segregation. As residential context is so strongly linked to formative social institutions - the family, the housing unit, the school, the voting booth - and because primary relationships - friends and romantic partners - are often delineated by the residential setting, its influence on individual and social outcomes is particularly powerful. Yet, it is also true that Americans spend a great deal of time outside of the residential context; indeed the typical American worker spends less than half of their non-sleeping hours at home. Moreover, as Americans continue to work longer hours and make more distant commutes, time away from home has increased. To the extent that exposure to diverse populations differs in residential and non-residential contexts, the implications for understanding group positioning, racial contact, and racial attitudes are potentially large. Our purpose in this paper is to provide a first look at how racial segregation differs in the residential and workplace settings for all U.S. metropolitan areas.1 Our goal is not to diminish the relevance of residential segregation, but rather to highlight that a singular focus on racial segregation in the residential context overlooks the possibility that individuals experience other forms of segregation - or integration - in their daily lives. To do so, we use data from the 1990 and 2000 decennial censuses, and from the 2006-2010 American Community Survey, along with special tabulations from the Census Transportation Planning Package which summarizes the racial composition of neighborhoods where people work. We use these data to calculate segregation indices at home and work, exploring broad trends in both residential and workplace segregation and examining geographic variation in the relative difference between the two. Theoretically, we anticipate several factors to influence levels of workplace segregation in metropolitan areas. First, since workplaces often employ people with different skill sets and from different positions on the socioeconomic ladder, and because the incidence of racial discrimination is likely to be lower in work settings than in residential ones, we anticipate that levels of racial/ethnic segregation will be lower at work than at home. Second, we anticipate racial segregation at work and home to be closely related, as people are likely to find work near their home (or move to neighborhoods near their work), and because social networks, which are often residentially determined, may draw neighbors into the same workplaces. And, third, basic assimilation theory implies that areas with larger immigrant populations - who may live and work in ethnic enclaves - will have higher levels of workplace segregation, particularly for Hispanics and Asians for who the foreign population is proportionally large. Data and Methods: Primary data for this project come from the US Department of Transportation's Census Transportation Planning Package (CTTP), is a set of special tabulations from the Census Bureau that summarize census tables at places of work, rather than at places of residence. We use all of the available data, include the 1990 and 2000 CTTP files, drawn from the long-form of the decennial censuses in 1990 and 2000, respectively, and the 2006-2010 CTTP file which is derived from the 2006-2010 American Community Surveys. We limit our analytic sample to core-based statistical areas, which we have defined commonly over the period using the 2010 OMB definitions. Segregation, in this analysis, is measured along two main dimensions: evenness and exposure. Pairwise evenness - e.g., that between whites and blacks - is assessed using the index of dissimilarity (D); multigroup segregation is measured using Thiel's H; and all exposure measures, including own group isolation, is based on the P* suite of measures. In all of our analyses we consider -as is conventional in segregation research - census tracts to be the lower level of aggregation and CBSAs to be the higher level. Census tracts are residential-based geographic units that the Census Bureau strives to capture about 4,000 persons and whose borders are determined by local committees of data users to follow visible features or legal boundaries. We define "places of work" in multiple ways to capture the populations likely to be present during standard work hours. In our first approach, we simply use the racial counts of workers in the census tracts of work (work tracts). The second approach allows for the possibility that work tracts also include nonworkers - the unemployed, those out of the labor force, and the elderly. Lastly, assuming that children remain in the residential tracts during the workday, we consider a definition of work tracts that includes workers, nonworkers, and children. Theoretically, these groups form the universe of people likely to be present in a given tract during a workday. However, several assumptions implied by this categorization are likely to be false: 1) not all workers are to be working in work tracts at the same time of the day; and 2) not all nonworkers and children are strictly home-based - many nonworking adults have responsibilities that require them to leave their residential neighborhood and children are likely to be in schools that may be located outside of their census tracts of residence. Nevertheless, these definitions span the possible populations likely to be present in a given tract during the workday. Preliminary Results: To date, we have compiled and prepared the 2000 and 2006-10 place of work (CTTP) and residential (Census) files, and have constructed all relevant segregation measures for the 2000 period. We have summarized these measures in Tables 1 and 2. Table 1 shows weighted mean evenness scores in residential and workplace contexts, as measured by the dissimilarity index (D) and the multigroup H index. Consistent with existing segregation research, residential dissimilarity scores of minority groups from whites show moderate to high average levels in 2000, with dissimilarity between blacks and whites (.65) being considerably higher than Hispanicwhite (.52) or Asian-white dissimilarity (.45). Most importantly, Table 1 shows clearly that regardless of how workplace segregation is defined - i.e., whether it encompasses just workers or is broadened to include nonworking adults and children - racial segregation in places of work is substantially lower than in places of residence. When gauged by the working population, black-white dissimilarity in the workplace is less than half (43.1%) its level in residential places. Hispanic-white and Asian-white dissimilarity in workplaces are similarly low, although their difference in residential dissimilarity is somewhat less striking. When segregation is assessed in a more comprehensive way - via the multigroup H index - a similar story is told, with workplace segregation being markedly lower than residential segregation. Parallel measures of racial contact, as measured via P*, are shown in Table 2. The rows of this table indicate the focal group and the values down the columns indicate the focal group's exposure to the four major racial/ethnic groups in the U.S. The italicized values along the diagonal represent group isolation, and indicate the proportion of the typical focal group member's residential- or work-tract that is occupied by others of the same race. In line with previous residential segregation work, the contact scores in the upper portion of the Table 2 show substantial variation in group isolation and exposure, with whites having high rates of isolation and relatively-low levels of exposure to other racial groups. The average residential tract for blacks and Hispanics is just about majority same race, and both groups have moderate levels of exposure to whites. The typical Asian neighborhood is majority white and has relatively equal shares of Asians and Hispanics as co-residents. Two important contrasts are revealed by the place of work exposure scores. First, racial exposure appears to be less homphilic in the workplace than in the residential space. For example, isolation scores - the percent of the work tract that shares the same race - are lower in places of work for all groups. For blacks and Latinos this translate into considerably greater exposure to whites: the typical black person lives in a residential neighborhood that is about one-third white, but works in a tract in which about two-third of other workers are white. The same pattern holds for Asians, but the difference in white exposure between workplace and residential tracts is comparatively small. The second observation from Table 2 is that while minorities tend to work in settings with greater exposure to other groups - especially whites - this same integrative dynamic is considerably less evident for white workers. Indeed, the racial compositions of the typical white worker's residential and workplace settings are nearly equivalent, with other whites making up 82% of residential neighbors and 78% of workplace neighbors. Thus, to the extent that places of work provide enhanced opportunities for racially-diverse social interaction, it mainly does so by diversifying the social contexts of minority workers and continuing to largely shield whites from exposure to nonwhites. Ongoing Research: Over the course of the next few months, we will use our assembled data to consider patterns of temporal change in residential and workplace segregation and will assess cross-metropolitan variation in both the levels of workplace segregation and their magnitudes relative to segregation in the residential domain (i.e., the ratio of workplace to residential segregation). Ongoing work is also directed at understanding sources of variation in workplace segregation across places, with a specific interest in testing theoretical arguments related to the geography of work and structure of labor markets, the arrangement of racial/ethnic populations, and the role of demographic change (e.g., immigration). Our results will inform not only empirical research on racial segregation, but will also broaden our understanding of racial/ethnic incorporation in different social contexts. [ABSTRACT FROM AUTHOR] |
