Data journalism has had a field day with national politics, turning crosstab mining and precinct shifts into an entire subgenre of human intellectual enterprise. It’s fun, until it’s exhausting, which is pretty soon.

But there’s much less to go on about local Chicago politics, which means I’m far from exhausted. To be sure, there’s been some good work: WBEZ’s dot maps of the 2011 and 2015 mayoral elections, for example; everything Scott Kennedy does at Illinois Election Data; more from WBEZ’s data Tumblr; and this long, but worth it, memo from the Illinois Campaign for Political Reform.

But I haven’t been able to find much that satisfactorily answers my biggest question about recent Chicago politics, which is: How different is the Emanuel coalition from the Daley one? That is, to what extent has Rahm Emanuel simply inherited the voters who kept Richard Daley in office for over two decades, and to what extent has he forged a different kind of voter base?

To skip to the answer: It’s, uh, pretty different.


Emanuel’s 2011 winning map looks very little like Daley’s map from his last election in 2007. Sure, both of them won the north lakefront—but that’s about where the similarities end. Daley’s coalition added to that a dominant performance on the Northwest and Southwest sides, as well as the Far South Side pockets of East Side, Beverly, Mt. Greenwood, and Hegewisch. Emanuel, by contrast, lost the Southwest Side heavily, split the Northwest Side, and lost all the Far South Side pockets that Daley won.

For anyone with a passing familiarity with Chicago demographics, this is clearly not so much about geography as about race. To put it bluntly: Daley won on a white-Hispanic coalition, while Emanuel won on a white-black coalition.

One other difference: The depth of the blue. Daley, in true Daley fashion, absolutely dominated in his base areas, frequently racking up 80 to 90 percent of the vote, and even in his weaker districts—that is, majority-black neighborhoods—got over 50 percent more often than not. Emanuel, by contrast, outright lost large areas of the city—almost always majority-Hispanic neighborhoods—and won much more modest victories in his stronger districts.

And then in 2015? Emanuel’s white-black coalition turned into a, um, white coalition. The black areas of the South and West sides that had narrowly supported him in 2011 turned reddish, along with the more integrated neighborhoods of the north lakefront, including Rogers Park and Edgewater. With the support of only one of the three racial groups that make up the overwhelming majority of Chicagoans, Emanuel was held to under 50% of the total vote, forcing a runoff.

The chart below is a different way of visualizing the same data. It’s actually two charts: For each of them, the vertical axis indicates the share of votes in each “precinct”—represented by the dots—won by Daley in 2007. The horizontal axis on the first chart indicates the share of votes in those “precincts” won by Emanuel in 2011, and the horizontal axis on the second chart indicates the share of votes in those “precincts” won by Emanuel in 2015. The dots are also colored to indicate the majority race in each “precinct.”

(Why do I keep putting “precinct” in quotes? Because these dots aren’t really precincts. Between 2007 and 2015, Chicago redrew its ward boundaries, and so precinct boundaries changed as well. To come up with units I could compare over all three elections, I laid a grid of little squares over the city, and used math to convert the votes and demographics of real precincts to values for those squares, which I could then use to make apples-to-apples comparisons, like the one below. If you look closely, you’ll see the grid in the maps above.)

Screen Shot 2017-04-11 at 10.55.36 PM

There’s a lot going on here, but a few things to notice:

  1. It’s pretty nuts how dominant Daley was. The bulk of both Latino and white “precincts” gave more than 80% of their votes to him in 2007, and—as we could see from the maps—even most black “precincts” gave him more than 50%.
  2. The racial coalitions we could see above are very apparent here: On the vertical (Daley) axis, Latino and white “precincts” line up close to the top, indicating strong support, with black “precincts” clearly clustered below. On the horizontal axis of the first chart (Emanuel 2011), white and black “precincts” cluster much farther to the right than Latino “precincts,” indicating greater support.
  3. But maybe the most interesting thing is those majority-black “precincts.” Not much changes from Emanuel’s 2011 performance to his 2015 performance—except that those black “precincts” flip from being lined up with the white vote just to the right of the 50% line, to being more lined up with the Latino vote just to the left of the 50% line. In other words, they flipped from being somewhat pro-Emanuel to being somewhat anti-Emanuel.

Here’s one more way of looking at the racial angle:

Screen Shot 2017-04-12 at 11.58.58 PM

This is the same thing, boiled down to the essence, and broken out so that the majority-Asian American “precincts” and those without a racial majority—which get lost in the noise in the scatterplots—are intelligible.

Finally, class intersects with race in an interesting way here. Grouping “precincts” into three buckets based on the proportion of workers in professional-class jobs according to the Census, it becomes clear that while Daley could rely on overwhelming white support across the class spectrum, the same is not true for Emanuel, who gets a significantly higher proportion of the vote from white “precincts” with relatively large concentrations of professional-class workers.

Screen Shot 2017-04-12 at 8.49.00 PM

Latino-majority “precincts” show a somewhat similar, but not as extreme pattern:

Screen Shot 2017-04-13 at 12.18.10 AM

And majority-black “precincts,” by contrast, show remarkable consistency across class for all three elections:

Screen Shot 2017-04-13 at 12.20.10 AM

There are obviously an endless number of additional ways to cut the data. I’m happy to take suggestions and requests! These numbers are interesting not just as trivia, but as a glimpse into the shifting alliances and interests competing for political power in the city. They suggest that quite a bit really has changed since the Daley era.