There is a disconnect between how we think about racial segregation and how we measure it. When people think about segregation, they typically think about the people that live next door to them, in close proximity. For example, when sociologists surveyed black and white homeowners in the 1990s about the neighborhoods they would prefer to live in, they drew a diagram of a street and asked which house they'd prefer, depending on the race of their neighbors. This makes sense; on a personal level, segregation is best understood by who we share property lines with. The people that we see on the street, for better or worse, are a part of our lives. We personalize the problem because that’s how we live.
On a personal level, segregation is best understood by who we share property lines with.
But social scientists measure residential segregation by first dividing cities into smaller areas, census tracks or city wards, and then analyzing the distribution of minority populations across those areas. The basic idea is to see how black and white people are distributed across neighborhoods. These traditional measures have been telling us a well-accepted story: Cities became more segregated over time as blacks left the South and migrated to Northern cities in the early decades of the last century. This was aided by direct investment in the tools of segregation — like interstate highways and federally backed mortgages — and municipal districts that carved out public goods, schools and police and parks in cities that blacks could not afford or were actively kept out of. Segregation in major cities reached its peak in the 1970s and has declined since then as suburban areas diversify and urban areas gentrify.
Analyzing the reality of segregation is a vitally important undertaking. Through my and my colleague’s research, we now know that we have been separating by race more and more over time. And how do we heal a racial divide when we are less likely to interact as neighbors? Personal contact reduces both racial prejudice and bias. Thus, changing racial attitudes requires interaction among the races.
The open question is how and why we grew apart. Was segregation enabled and supported by policies that allowed white people to move out of cities, or was it created by policies that allowed whites to create communities where they did not have to have any interactions or political decision-making with blacks?
In almost every other aspect, social scientists measure interpersonal contact. When it came to where we live, we were stuck with census tracts and percentages. But who knows the boundaries of their census tract? Along with my college John Parman, we worked out a new way to analyze segregation and bridge the gap between the way that people think about segregation and how we measure it. We developed a segregation measure that compares the number of black households and white neighbors in a community to the number of households one would expect in a perfectly integrated neighborhood (where households are randomly placed in the neighborhood) and a completely segregated one (where all white households are concentrated in one part of the neighborhood).
This measure, which we developed using the census, has intuitive appeal and another distinct advantage: Since it does not require boundaries such as census tracts or wards, we could apply this measure to rural areas as well as urban ones. For all of the talk of segregation and policies designed around it, segregation in rural areas has not been well-measured and we know very little about it.
Since the measure requires the complete census, the project looks at how segregation evolved from the end of Reconstruction to the last years of the Great Depression. In looking at segregation from 1880 and 1940, a new and more nuanced picture of American segregation takes shape.
The most important finding, however, was that segregation in cities is only half of the story. Segregation in rural areas changed over time as well.
First, we were able to confirm that cities became more segregated from the end of the 19th century to the middle of the 20th. Second, we found that while traditional measures imply that Southern cities are more integrated than Northern cities, we found exactly the opposite. More important, we think we know why. While whites and blacks were likely to live in the same census tracts in Southern cities, they were less likely to live next to one another.
The most important finding, however, was that segregation in cities is only half of the story. Segregation in rural areas changed over time as well. In fact, segregation in rural areas increased just as dramatically as in urban areas. This means that blacks leaving the rural and urban South were migrating to increasingly segregated Northern cities, but the areas they were leaving behind were becoming increasingly segregated as well.
From policing and mass incarceration to school funding and discrimination, we are less likely to live near people who experience vastly different realities along these dimensions because of race. In fact, we have created racially homogenous echo chambers that increase racial prejudices and racial inequalities and decrease racial understanding. As Americans continue to try to bridge the racial divide, our ability to do so is limited by policies and preferences that keep us apart.