The recent geography of gentrification in Chicago

There is a widespread sense that some parts of Chicago are gentrifying at a rapid pace. Rising rents in many neighborhoods have been cited as support for this view.

I downloaded and mapped data from the recently released 2011/2015 American Community Survey (ACS) in order to examine the geography of recent gentrification in Chicago.

To interpret these maps, one needs to understand that some academic work on gentrification portrays it as a kind of wave, moving regularly block by block or tract by tract, in which neighborhoods are completely transformed as their old inhabitants are displaced by wealthier newcomers.¹ Some of the many activists who oppose gentrification seem to accept this view as well, although they’re more likely to talk about neighborhoods than census tracts.²

Recent ACS data suggest that gentrification in Chicago in the last few years often hasn’t quite worked that way. Its geography has been much more complicated.

Here’s a map showing the change in real per capita income by census tract between the 2006/2010 ACS and the 2011/2015 ACS:³

Here’s a similar map for the region:

Someone wedded to the idea of gentrification as a wave could certainly find evidence for it in the dozen or so tracts in Logan Square, West Town, Humboldt Park, North Center, Uptown, and western Lake View where per capita income was up by 25% or more. Gentrification on the North Side does indeed seem to be moving west and north and cutting into the few remaining pockets of poverty. But, in fact, there are tracts with similar income rises scattered throughout the urban area. Many (maybe most) of those in the city include areas where high-rise housing projects were removed in the years between the 2006/2010 and 2011/2015 surveys, for example, the tracts covering the part of the city where Cabrini-Green Homes once stood. A change in building stock was associated with an increase in wealth in several other areas as well, for example in the West Loop, where new expensive housing has mostly replaced parking lots and industrial buildings. Whatever one thinks of this process, it would be simplistic to describe it as a wave of rich people displacing the poor.

What most struck me about these maps is how widespread an increase in per capita income has been in the Chicago area. All the tracts colored red or pink showed a positive change; that is, they did better than average. Such tracts are commoner on the North Side than the South Side, and perhaps a little commoner close to the Loop than away from it. But the spatial differences do not make a dramatically clear pattern. Gentrification as measured by income change between 2006/2010 and 2011/2015 seems to have been widespread. It was only to a limited degree strongly focused on a few areas. There is additional evidence for this view, for example, in the continued influx of white people into numerous North Side and near-downtown neighborhoods. The fact that the city of Chicago, unlike the Chicago Metropolitan Statistical Area, the state of Illinois, or the United States as a whole, emerged from the Great Recession as wealthy as it was before the recession began provides further support for the idea that a modest amount of gentrification has been widespread in Chicago.⁴

Some of the areas of rising income were near the northern and northwestern edges of the city. It’s possible that middle-income people have become more willing to live in traditionally working-class neighborhoods like Albany Park and West Ridge (West Rogers Park) than in the past. Newcomers may have raised the per capita income of parts of these areas a little bit, but these were already reasonably healthy places, and most newcomers weren’t really wealthy. Calling population turnover there “gentrification” without some qualification seems like a distortion.

But it was more complicated than that. A surprising number of tracts experienced a decline in real per capita income. Some of these are in relatively poor areas, but many are in parts of the city and suburbs that are generally considered quite prosperous. The latter changes may be due to declining income among older people who make up a large part of the population in some areas, for example, along the North Side Lakefront. There was also a well-documented influx of renters in condo buildings whose departing owners have found it more profitable to rent than to sell. Rents in condo buildings aren’t usually low, and renters in these buildings aren’t likely to be really poor, but in most cases they’re probably poorer (and younger) than their landlords.

The details on the map need to be viewed skeptically, since ACS data have very high margins of error, easily enough to move a substantial portion of tracts one or two categories up or down. The complicated patchwork of income changes may not be due to the existence of a complicated microgeography as much as to noise introduced by the necessarily approximate data, at least in part. The difference between, say, a 10% change and a 25% change portrayed on the maps may not be completely meaningful.

There’s another reason for caution. The 2006/2010 period included both the last year of the early 21st-century boom and the low point of the Great Recession, while the 2011/2015 period was generally one of recovery from that low point. Dealing with ACS data is not like dealing with traditional census data, which aims to report conditions on a single day. During a five-year period over the course of which there were substantial changes, even a small spatial difference in different years’ sampling would result in inaccuracies in maps like these that show changes. I don’t know the extent to which the Census Bureau corrected for this.

Let me add that, even if there is reason to suspect that the figures are only approximations, broadly speaking they seem about right. Here’s a 2011/2015 map of per capita income in Chicago and vicinity:

Here’s a comparable map for the region:

These maps look pretty accurate.

Income is of course not the only way to measure gentrification. Changing education levels and a change in the distribution of occupations are two additional areas where census data are available.

Here’s a map of Chicago and vicinity showing the change between 2006/2010 and 2011/2015 in the percentage of people 25 and older with a college degree:

Here’s a comparable map for the region:

Here too there are surprisingly complicated patterns that do not suggest wholesale recent replacements of minimally educated people by well-educated ones in a spatially simple wave. A major impediment to interpreting these maps is that there has generally been an increase in the proportion of the population with a college degree, even among relatively poor people. Some of the tracts where there are have been the largest increases in the percent of those with college degrees are in fact in problem-ridden areas on the South and West Sides. In many cases the percent went from very low to just low. You really need a general map showing percentage of the population 25 and over with a college degree to interpret these maps. Here is such a map for Chicago and vicinity in 2011/2015:

And here’s a regional map:

Generally speaking, the maps of change in education level, at least for areas that started with a moderate level of educational attainment, have a message that’s similar to that of the income maps: There seems to have been continued slow increase in education levels in many North and Northwest Side neighborhoods (especially those away from the long-prosperous Lakefront), as well as around the Loop, but, over the five years covered by the two ACS surveys, changes were so spotty that they are hard to interpret in a simple way. There does not seem to have been a wave of gentrification displacing all before it. However, with education as with income it’s pretty clear that the city did better than the metropolitan area, the state, and the nation.⁵ There is every reason to think that there’s been continued slow gentrification as measured by educational level, just as there has been by income; it’s just that it hasn’t been spatially simple.

Maps that show changes over more years reveal much simpler geographies. Here’s a set of maps much like those above that shows changes between 2000 (with data from the old long form) and 2011/2015 (with data from the ACS). There were many adjustments in census tract boundaries in these years, and I’ve used the analysis worked up by the Longitudinal Tract Database project at Brown University to put 2000 data into 2010 boundaries. As the authors⁶ of this wonderful source of data would be the first to admit, reworking data compiled for one set of boundaries into a different set of boundaries inevitably results in some distortions.

Despite this caveat, these maps all show a much more lucid geography than the 2006/2010 to 2011/2015 map set, no doubt in part because, with more years, patterns got clearer. It also seems to be the case that changes were more intense and more geographically coherent in the first years of the 21st century than afterward. To sum these up: There was a pretty substantial increase in income and educational attainment in many North Side neighborhoods (especially away from the Lake) in the first dozen or so years of the 21st century. The areas around the Loop (including those on the South Side) also did well. Some poor neighborhoods got poorer, while others (Bronzeville, for example) may have really turned around. Chicago in general did better than its suburbs. Nearly all the dramatic increases in income and educational attainment in the Chicago area occurred within the city’s boundaries.

As for those dramatically rising rents, it may just be that many people are having to pay a larger percentage of income for housing than they used to. (Condo owners, the value of whose properties is in many cases lower than before the Great Recession, may be paying less.) Whether the current construction boom will change this, time will tell.