There is no more highly contentious issue among those of us who spend their time thinking about the future of cities than gentrification.
Its progress has been massively uneven, as change has tended to concentrate in large, knowledge-based cities like New York, L.A., Chicago, D.C., Boston, and San Francisco—with much less movement in older Rustbelt or sprawling Sunbelt cities, according to a Cleveland Fed study I wrote about last fall. Another recent study found that, for every single neighborhood that's gentrified since 1970, 10 have remained poor and another 12 have slipped into poverty.
When we talk about why some places gentrify and others don't, there's often a pressing, underlying question at stake: To what degree is gentrification bound up with and shaped by race?
This is the subject of a path-breaking new study by Harvard doctoral student Jackelyn Hwang and urban sociologist Robert Sampson published in the August issue of the American Sociological Review.
Hwang and Sampson organize their research around a perplexing paradox. For all of these dollars flowing back into downtowns, how has the inequality of neighborhoods by race and class remained so persistent?
Looking at the fate of Chicago neighborhoods that were primed for redevelopment in 1995, Hwang and Sampson found that by the late 2000s, racial composition did in fact have a significant effect on a neighborhood's chance of improvement and ultimate gentrification. The neighborhoods that saw the most improvement met a minimum threshold proportion of white residents—about 35 percent—and a maximum threshold of black residents—about 40 percent.
The study used the case of Chicago to investigate how gentrification reflects existing patterns of race and class-based inequality. It is organized around a set of a Chicago neighborhoods that were deemed “gentrifiable” in 1995—those with a median income below the citywide median in 1960. Building off the work of researchers Daniel Hammel and Elvin Wyly, they examined a subset of 99 tracts that they could match up with other data sets. This included 26 that Hammel and Wyly had deemed already at the core of gentrification in 1995, 16 more that were on the fringes of this process, and another 57 that Hwang and Sampson identified as adjacent and therefore ripe for change.
To identify the level of gentrification, Hwang and Sampson turned to the changed physical condition of the neighborhood using Google Street View. Their research coded for signs of new building projects, upkeep of old buildings, and visible signs of order and disorder to determine a quantifiable level of gentrification for each study tract. The maps below, from the study, show the spread of reinvestment across the city. Notice how much is concentrated in the northern part of the city, away from the traditional centers of black neighborhoods on the South Side.
The research then connected these indicators of gentrification with a number of social and demographic characteristics from the census, city records of new investment, and qualitative responses about community perceptions from the 1995 survey, the Project on Human Development in Chicago neighborhoods.
Based on this, they found that, looking beyond the original amount of gentrification in each tract, their initial racial compositions accounted for a substantial 29 percent more of the variation in reinvestment levels.
Overall, they found a clear threshold effect for race. Across the board, the likelihood of gentrification declined as the proportion of black residents in a neighborhood went up. The difference in reinvestment levels between neighborhoods of 35 and 45 percent black residents was more than twice the gap in extent of gentrification between neighborhoods of 5 and 15 percent black residents.
The graphs below, also from the study, show the opposite patterns of gentrification when isolating for the black and white composition of the neighborhood.
It's not necessarily explicit racism that creates these patterns, according to the study. Rather, the process of gentrification is basically one of sorting, where gentrifiers' more implicit biases play a major role in determining where they want to live.
Two major implications of this research stand out.
First, any benefits of gentrification remain highly bounded, showing little signs of spilling over or trickling down to adjacent neighborhoods, particularly those with more minority residents. Hwang and Sampson's study limited itself just to tracts that were geographically within spitting distance of ones that had already seen signs of renewal. But the benefits of upgrading couldn't always cross those racial borders. As the authors conclude, “Such a pattern perpetuates, and perhaps worsens, urban inequality.”
Second, it is not just those who are displaced that gentrification hurts. There is an even larger category of left-behinds, those whose neighborhoods are essentially impervious to upgrading. These are the areas where poverty, race, and class have their most persistent effects, reinforcing and compounding concentrated disadvantage.
It is this very persistence of segregation and class—even in the face of huge reinvestment schemes—that helps explain the newly divided cities and metropolises of our time.