Exiting and Entering Poor Neighborhoods: Latinos, Blacks, and Anglos Compared

Scott J. South, University at Albany, State University of New York
Kyle D. Crowder, Western Washington University
Erick Chavez, University at Albany, State University of New York

Energized in large measure by Wilson's influential treatise on The Truly Disadvantaged, scholars from a variety of social science disciplines have devoted substantial attention to studying the emergence of an urban underclass. A vast literature on both the causes of growth in the underclass population and the consequences of this growth for inner-city social dislocations has accumulated rapidly over recent decades. The two-pronged thrust of Wilson's theory is to describe the reasons for the growth of underclass neighborhoods, which he attributes largely to the out-migration of middle-class blacks, and the many inimical consequences of this trend, which he ascribes primarily to "concentration effects." Although the conceptual utility and empirical validity of Wilson's claims in both of these areas continue to be vigorously debated, there can be little doubt that his contributions in this area have reinvigorated the study of urban poverty, neighborhood effects on individual behavior, and inner-city population change. Wilson's writings on the urban underclass focus predominantly-though not exclusively-on the deteriorating economic and social conditions among inner-city African Americans. A key issue in this area is the extent which his and related conceptualizations of the underclass provide a useful device for understanding the neighborhood conditions and life chances of non-black minority groups, particularly Latinos. Although this debate is largely a conceptual argument, hinging on somewhat arbitrary definitions and often inconsistent uses of the term "underclass," the potential application of the concept to Latinos raises crucial-and heretofore unexplored-empirical issues regarding the migration of Hispanics between metropolitan neighborhoods of varying socioeconomic status. A common and widely accepted criterion for underclass designation is that members of underclass groups are persistently mired in distressed, inner-city neighborhoods, with little opportunity to leave them for more advantaged local communities. In those rare instances when members of the underclass are able to establish residence in middle-class neighborhoods, their tenure there is thought to be quite tenuous; they are more likely than other groups to move back into poor neighborhoods. Thus, underclass populations suffer at unusually high rates the social dislocations that attend residence in disadvantaged neighborhoods. Prior studies have examined patterns and determinants of inter-neighborhood residential mobility among blacks and non-hispanic whites, but until now the constraints of existing data have prohibited a thorough and systematic evaluation of the analogous migration behaviors among Latinos. This paper uses longitudinal data from the Panel Study of Income Dynamics (PSID), and particularly the recently-developed "Latino-sample" of the PSID, to compare and contrast the probabilities of moving between poor and nonpoor neighborhoods among Latino (N=2,676), black (N=3,017), and non-hispanic white (N=4,218) heads of household. We focus on the period 1990-1995, the only years that data were collected from the PSID-Latino sample. Attaching 1990 census data on the tract-level poverty rate for the PSID respondents at each annual interview allows us track prospectively who moves between poor and nonpoor neighborhoods. We examine not only differences in the rates at which Latinos, non-hispanic blacks, and non-hispanic whites (i.e., anglos) move between poor and nonpoor neighborhoods, but we also attend to possible differences in the effects of socioeconomic and contextual predictors on these probabilities. Multinomial and binary logistic regression analyses are used to examine the effects of race/ethnicity and other mobility determinants on the likelihood of moving between poor and nonpoor census tracts. Preliminary findings indicate that members of all three major U.S. Latino groups (Mexicans, Puerto Ricans, and Cubans) are significantly less likely than non-hispanic whites to move from a poor to a nonpoor neighborhood, and that Puerto Ricans (but not Mexicans or Cubans) are significantly less likely than blacks to move in this direction. When conventional determinants of mobility are controlled (e.g., age, marital status, homeownership, duration of residence, household crowding, education, income), the difference in out-mobility rates between Cubans and anglos disappears. The effects of education and personal income on the likelihood of "escaping" poor neighborhoods are stronger for blacks, Mexicans, and Puerto Ricans than for anglos, findings that are broadly consistent with the weak version of the place stratification model of locational attainment. Net of other factors, Cubans demonstrate high rates of moving out of neighborhoods with large black populations. Regarding mobility from nonpoor to poor neighborhoods, we find that members of all three Latino groups are significantly more likely than anglos, but significantly less likely than blacks, to experience this type of move, even in the face of controls for conventional determinants of residential mobility. Indeed, only among blacks is the probability of moving from a nonpoor to a poor tract greater than the probability of moving from a poor to a nonpoor tract. For the most part, however, the effects of socioeconomic status on the likelihood of moving from a nonpoor to a poor neighborhood do not vary by race or Hispanic origin. We conclude with a discussion of the implications of these findings for the applicability of the concept of "underclass" to Latinos, for theoretical models of locational attainment, and for race-and ethnic-specific patterns of intrametropolitan population redistribution more generally.