Abstract
As scientific instruments used to reveal something about how the world works, the telescope and the microscope each serve distinct and complementary functions. So, too, with the tools we use to understand patterns of racial residential segregation in the United States. The Census and survey data used in the accompanying article give us a useful telescope with which to scan the horizon and view the extent to which race shapes neighborhood patterns. But microscopes let us narrow our view and provide for a closer inspection of the behaviors and attitudes that make up those patterns. The insights from four different types of data uncover some of the micro-processes that may help explain why segregation, though on the decline, is still far from extinct.
Discrimination as a moving target.
Looking for a place to live involves a number of steps: finding an available house or apartment, contacting and meeting with the owner/agent, viewing the unit, submitting an application, securing approval, and signing a lease or executing a contract. Most of the information we have about whether race shapes how one is treated in this process—including the national Housing Audit Studies reviewed in the accompanying article—looks at one specific step in this process: the face-to-face encounter between housing provider and renter/buyer. But in the world of cell phones, caller ID, and voicemail, and in the context of a widespread ability to discern race based solely on a person’s voice, Douglas Massey and Garvey Lundy have shown, with research published in Urban Affairs Review, that racial discrimination can happen even before homeseeker and housing provider ever meet.
In their rigorous audit study of the Philadelphia rental market, researchers posing as prospective tenants called the same housing providers to inquire about a unit—there were men and women, and they differed in terms of whether their accents conveyed their status as white middle class, black middle class, or black poor. Though all presented themselves as looking for the same kind of apartment and having the same kinds of financial abilities, even the most basic courtesy of having one’s phone call returned showed unequal treatment. At the two extremes, 87 percent of white middle class men were called back by the housing provider, while just 63 percent of poor black women were given that courtesy. And the differences didn’t stop there: African Americans—especially poor African American women—were disadvantaged in terms of whether they were told on the phone that the unit was available, the quoted application fee, and whether issues of credit worthiness came up. This study and others confirm that discrimination is, in Massey’s words, a “moving target.” Discrimination “with a smile” may have been replaced by discrimination “with a dial tone.”
Is it race or is it class?
The attitudinal data Reynolds Farley reports in this issue is a valuable tool for framing the landscape of racial attitudes related to residential segregation, but it’s also a relatively blunt instrument. One shortcoming is its singular focus on a neighborhood’s racial composition. Indeed, some have argued that any reservations expressed—especially by white respondents—about sharing neighborhoods with African Americans may not be about race at all. To the extent that people associate neighborhoods with more African Americans as being of lower quality (lower housing value appreciation, worse upkeep, higher crime rates, poorer quality schools, etc.), one might argue that it’s not that people want to avoid living with black people, it’s just that they don’t want lower quality neighborhoods. Because the “neighborhood cards” approach is silent on all other neighborhood characteristics, it’s impossible to disentangle whether it is race or class driving a reluctance to live in integrated neighborhoods.
Along with my colleagues, I tackled this challenge in a recent experiment embedded in a random sample survey of residents in the Chicago and Detroit metropolitan areas. In this household-based survey, respondents were shown several videos of actual neighborhoods that varied in terms of their social class characteristics (size of home, upkeep of property, etc.). Actors portrayed residents in these short videos, allowing us to experimentally manipulate the racial composition of the neighborhood while holding constant its observable social class characteristics.
While showing that social class does matter—pretty much everybody wants to live in a “nicer” neighborhood—our results also show that even if a neighborhood is identical on all other dimensions, the presence of black residents makes the area less desirable for white homeseekers. African Americans, for their part, are also influenced by race, but less so and in a different direction: for them, all-white neighborhoods were least desirable and racially-mixed were most desired. To further make the case that race—and not just social class—matters, white respondents who held negative stereotypes of black people were significantly more likely to be affected by the neighborhood’s racial composition. Despite optimism about the racial attitude changes documented by the survey data, this experiment demonstrates people may be class conscious, but they’re not color blind.
Taking it to the real world.
The neighborhood cards measure has also been criticized because it just isn’t the real world. People don’t move into hypothetical neighborhoods, they make actual housing decisions in real cities and neighborhoods. My co-author Michael Bader and I asked a random sample of Detroit-area residents to look at a colorful map of the Detroit metropolitan area that showed 33 different neighborhoods and communities ranging from all-white suburbs to all-black Detroit neighborhoods. They were then asked to identify communities in which they would “seriously consider” living and any where they would “never consider” living. There was some agreement between black and white respondents: for example, neither group was attracted to older, lower-income suburbs with little recent economic growth. But there were still striking racial differences; both the race of the resident and the racial composition of the community affected whether or not a particular area was considered desirable. And, controlling for community characteristics like housing values and location, we found that the “whiteness” of a community predicted whether white people would seriously consider moving there, while racial composition didn’t affect African American considerations. Specifically, a 10 percent increase in the proportion of white residents in a community was accompanied by a 53 percent increase in the odds that a white person would “seriously consider” it as a place to live. For African Americans, the percent white or percent black made no difference in whether they’d “seriously consider” or “never consider” a community.
The pattern for African Americans is especially important in light of the frequent assertion (by scholars and policymakers alike) that black people prefer “50/50” neighborhoods. Using a tool that measures preferences in the real world reveals far more flexibility on the part of African Americans—and also reinforces the conclusion that race still matters for whites. This emerges again when the microscope is turned on actual behaviors.
From neighborhood change to changing neighborhoods.
Indexes of dissimilarity are a great window onto what neighborhoods look like—from a distance—and how they have or have not changed. But they come up short for understanding who is actually changing neighborhoods and what makes its residents move. People move to new neighborhoods for lots of different reasons—they move because they want a bigger house or need to be closer to work or because the racial composition of the neighborhood is changing. The latter, in the context of moves made by whites, is often referred to as “white flight,” and it’s been examined extensively through the lens of Census data. Now, another growing body of research—made possible by longitudinal data—allows us to answer specific questions about the moves people make and assess the extent to which neighborhood racial composition, independent of a variety of other individual and neighborhood characteristics, leads people to make housing decisions that, taken together, contribute to persistent patterns of segregation.
For example, we can assess not whether a neighborhood has experienced racial change, but whether individuals change neighborhoods in the face of a large or growing minority population. We can also determine who is more or less likely to do so. Sociologists Kyle Crowder and Scott South find that, quite apart from individual characteristics that predict mobility (such as age, presence of children, and homeownership), and above and beyond other neighborhood characteristics (like average income levels and stability), the greater the percentage of minority residents in a community, the greater the likelihood that white families will move to a new neighborhood. Interestingly, in an earlier article, Crowder found no effect depending on which racial/ethnic group is present (African Americans, Asians, or Latinos); instead, white people are particularly likely to flee multi-ethnic neighborhoods. But whites living in neighborhoods experiencing recent growth in the black population—though not the Asian or Latino population—are especially likely to leave.
In analyses using the same longitudinal dataset, social demographer Lincoln Quillian further demonstrates in Social Science Research that stubborn patterns of segregation are maintained to a great extent not only by the decisions of whites to leave particular neighborhoods, but also by their choice of destination (generally, neighborhoods with few black residents).
Crowder also highlights how context matters in whether housing choices help perpetuate segregation. The abundant availability of alternative, all-white communities and the presence of new housing developments both increase the likelihood that white residents will flee. Findings like this merit close scrutiny, as they may provide clues as to why some cities have shown promising signs of integration, while others remain stubbornly segregated.
Implicit, if not explicit, in proclamations of a “post-racial era” is the notion that people have become color blind. In some respects, the attitudinal data and some concrete achievements support this contention. Slow but steady declines in segregation, for instance, might be understood as marking the beginning of the end of “chocolate cities and vanilla suburbs.” But other studies suggest a less sanguine conclusion. Landlords are not color blind when choosing tenants; residents are not color blind when evaluating neighborhoods; and when people—especially whites—make moves, they’re not color blind. These more microscopic investigations of the patterns and processes related to residential segregation and how race plays a role, suggest possible answers to the pressing question: Why are levels of segregation not lower still?