Many studies of gentrification (for example, the Federal Reserve Bank of Philadelphia study we wrote about last week) begin by dividing neighborhoods into one of two categories: gentrifiable and non-gentrifiable. Usually, to qualify as “gentrifiable,” a neighborhood must rank relatively low on the socioeconomic ladder: one standard used by at least a few different reports is having a median household income that’s below average compared to the metropolitan area as a whole.

In the common understanding of gentrification, this isn’t a terribly unreasonable line. But that frame can give us a very distorted vision of neighborhood change and issues related to the shifting geography of economic privilege.

On the one hand, it becomes easy to conflate “below average” with “poor.” Some of the studies using this standard for “gentrifiable” have reported extremely high rates of gentrification in major cities, which might be understood to mean that low-income neighborhoods are rapidly filling up with the relatively well-to-do. But of course a neighborhood at the 49th percentile of median income regionally—or, for that matter, the 40th or even 30th—is probably very, very different than one at the 10th percentile. Studies that focus on truly low-income neighborhoods, including our own “Lost in Place,” have shown that gentrification is generally rare in communities with very high concentrations of poverty. If we’re worried about the plight of especially low-income people and neighborhoods, this is an important nuance.

From the other end, by writing half of all neighborhoods out of the story from the very beginning, these studies may imply that economic change in relatively higher-income communities is either less prevalent, or less important. It’s not clear that either is true—and at the very least, this question deserves to be a part of the conversation.

The Onion is on it.
The Onion is on it.


Socioeconomic shifts in neighborhoods whose residents are disproportionately low-income or people of color are particularly notable, because they involve changing some of the most highly visible social boundaries in American life. But that doesn’t mean that other kinds of changes, like a predominantly white, middle-class neighborhood becoming a predominantly white, wealthy neighborhood, don’t also matter. For one, this kind of economic segregation can put housing even further out of reach for working class and low-income households, requiring deeper, less efficient subsidies, or increasing housing prices so much that vouchers won’t cover the cost of rent. Moreover, growing high-income segregation has grave implications for cities, and society, as a whole.

Previous research, in fact, has demonstrated that much of the increase in income segregation over the past few decades has been driven by the “secession of the rich”: the phenomenon of the very wealthy separating themselves from everyone else. Increasing segregation of high-income households can reduce the resources available to support common, community-wide services—a problem that’s particularly bad when these elite neighborhoods happen to be separated from others by municipal or other government bodies, which can collect super-sized tax revenues to provide extremely high-quality “public” services, but only to their rich constituents.

New Trier Township High School in Winnetka, Illinois.

And looking at income growth across the whole distribution of neighborhoods can help tell us about how income segregation is growing—or not. After all, for income segregation to be declining, low-income neighborhoods have to see faster income growth than more affluent ones. If the opposite is happening—or even if gains are equally distributed—then we’re not making progress on income segregation.

So what would a study of neighborhood income change find if it didn’t define some places as “ungentrifiable”? To get some idea, we looked at five metropolitan areas: Chicago, Detroit, Miami, Philadelphia, and Portland, OR. Using 2000 Census data, we divvied up Census tracts (areas roughly the size of a neighborhood) into buckets based on their median incomes in 2000—below $25,000, from $25,000 to $50,000, and so on in increments of $25,000. We then looked at how fast median income grew in each of those buckets between 2000 and 2010.

The results are too complex to be summarized in a phrase—but of the five cities, only Miami appears to show a clear pattern of lower-income neighborhoods seeing faster income growth than middle- or upper-income neighborhoods. In Chicago, very low income neighborhoods—those with median incomes of below $25,000 in 2000—grew very fast, but many of those tracts appear to be in communities where public housing projects were dismantled in the 2000s, removing large numbers of low-income people and, as a result, making the median income higher. Above $25,000, Chicago appears to follow a “rich-get-richer” pattern—at least up to a median income of $125,000.


Portland shows the fastest income growth at the extremes of wealthy and poor, though its fast-growing upper-income bucket ($75,000 to $100,000 median income) grew faster, at 26.6 percent, than its lower-income bucket (under $25,000 median income), at 23.8 percent. The middle-income buckets grew more slowly. Even so, Portland’s incomes are notable for being evenly distributed: unlike the other metros in our sample, none of Portland’s neighborhoods had a median income of more than $125,000.

Philadelphia and Detroit both show a fairly clear pattern of “rich-get-richer,” with Detroit’s being particularly stark: The median neighborhood with an income below $50,000 grew not at all or actually became poorer, while those with incomes above $100,000 grew most quickly. In Philadelphia, very poor neighborhoods (those with 2000 incomes under $25,000) grew slowly, at 12.4 percent, while other neighborhoods grew faster: neighborhoods with incomes between $75,000 and $100,000, for example, grew median incomes by 27.1 percent.

So what to take out of all of this? In part, it isn’t surprising: We know that income segregation is growing nationally, and as we said, for that to be true, wealthy neighborhoods have to be growing faster than lower-income neighborhoods on average. As a caveat, by averaging out rates of income growth across many neighborhoods, we’re missing huge amounts of variation. There are some higher-income neighborhoods that are becoming less affluent, and there are some low-income neighborhoods that have seen very rapid income growth.

But looking at income growth across all kinds of communities gives some much-needed context to the conversation about neighborhood change. It’s not that change in low-income neighborhoods is less important: indeed, we ought to be concerned about the welfare of people who are most vulnerable, and what happens in their neighborhoods. But social equity involves the distribution of resources across all neighborhoods, and so it matters if high- and moderate-income communities are becoming even more exclusive.

The numbers we’ve come up with here are far from a full picture, of course—more like a first attempt to explore the question. We’re looking forward to learning more in the future.