Why talk about this now, with the next election more than a year and a half away? Partly because now is the time when targets get picked and candidates get recruited; competitive races don't usually just pop up out of sheer will but require a lot of groundwork. But more importantly, the Census Bureau finally graced us last month with demographic information for the nation's congressional districts. Although the most recent Census has been in the books for several years now, things got slowed down by the redistricting process (which, of course, relies on the Census' initial population figures); they had to wait until the new district lines were finalized to be able to calculate new district data.
With access to that data, finally, I initially planned to write a piece about the various superlatives in congressional districts (whitest districts! poorest districts! best educated districts! and so on). That's interesting trivia, of course, but by itself doesn't tell us much about how we can reshape the House in 2014 and in future years, so I also decided to pinpoint Republicans in the districts with the demographic categories that seemed most hostile to them (say, for example, the five congressional districts with the highest percentage of African-American residents that are still represented by Republicans).
Rather than put up dozens and dozens of tables, though, that left me wondering: which variables actually matter the most? Which particular demographic categories are most strongly related with whether a district tends to elect a Democrat or Republican? That way, we could focus on only a few most important categories. So, with that in mind, I calculated correlations for each of the categories in the Census' release, factored against the percentage the Democratic candidate for the House got in each district. Some of the results are predictable, but others were a total surprise. Here's a chart of the characteristics that had the strongest positive and negative relationships with Democratic share of the House vote:
|Renter %||0.59||White %||-0.58|
|25-34 y.o.%||0.43||Veteran %||-0.49|
|Black %||0.38||Some College %||-0.32|
|"Some Other" race %||0.38||Male %||-0.31|
|Asian %||0.30||10-14 y.o. %||-0.27|
Some of the other things that you'd think might matter turn out not to matter much at all. For instance, the correlation coefficient on median household income is only 0.02, meaning no relationship in any direction. (With correlations, 1 or -1 means a perfectly corresponding relationship within the data, while 0 means nothing but random noise.) It's tempting to think of the Republicans as the "party of the 1 percent" and to think of all the Democrats representing blue-collar districts in the cities, but stop and think about the number of affluent suburban districts that elect Democrats, or the number of abjectly poor areas in the Appalachians that elect Republicans.
Follow over the fold for full discussion on why these factors might matter, and the promised lists of Republicans vulnerable according to these criteria ...
So what is it about the percentage of renters (or, more technically, the ratio of renter-occupied housing units to total occupied housing units) that makes it a decent predictor of Democratic fortunes? In large part, it's because it points to urban areas; more people tend to rent in urban areas than rural areas. And it takes only a quick glance at last year's election results to see that Democratic votes are most heavily concentrated in the nation's metropolitan areas.
Although the Census Bureau will, at some point, calculate what portion of a congressional district is rural and what portion is urban, they didn't include that in their initial release. (You can see the entire data set here in a Google Doc that I assembled.) However, I have run the correlation between urban/rural percentage and Dem performance in years before, and it wasn't as strong a relationship as "renters" is now. (It had never occurred to me to try a "renters" category before.) Either that urban/rural divide has gotten more precise (as a result of better gerrymandering and/or more self-sorting of Democrats to the cities and GOPers to rural areas and exurbs), or else "renters" incorporates an extra element that makes it even more powerful.
My theory is that while it includes urbanites of all stripes, it also encompasses the rural poor, while avoiding the heavily-Republican rural middle-class (who, because of the relative costs, seem more likely to own property than the urban middle-class). You can see that especially with the most-renter-heavy district that's occupied by a Republican, California's 21st district in the Central Valley, which is represented by David Valadao. This district is a mostly-rural area bookended by Fresno and Bakersfield that's heavily Hispanic, mostly agricultural workers, and one of the nation's poorest districts.
So, the following three charts of districts are: first, the five districts with the highest percentage of renters (heavily focused on New York City, unsurprisingly), and second, the five districts with the lowest percentage of renters (affluent suburbs with low density and mostly single-family homes, particularly in the midwest). And third, the five most-renter-heavy districts with Republicans. These are all districts with large Hispanic populations, though only CA-21 and TX-17 are largely rural; they also include inner-city districts in Miami (FL-27) and Houston (TX-07).
The #2 district, Kenny Marchant's TX-24, may be the most interesting one: it occupies the suburbs between Dallas and Fort Worth around DFW airport. Marchant has never seen much of a challenge before and Dems are only beginning to build a bench here, but this is one of the most rapidly Hispanic-trending parts of the country. Redistricting (and the creation of the adjacent, Hispanic-majority TX-33) may have helped Marchant rest easier for a few years, but even his newly-configured district is rapidly changing.
The second strongest factor, and the strongest negative correlation with Democratic House vote, seems a lot less surprising: the percentage of white residents in the district. It's interesting that "white" is a stronger predictor than any other specific race; considering that African-American voters went around 95 percent Democratic according to exit polls last year, you'd think that might be a stronger predictor.
However, we're talking about districts as a whole, and there are a lot of black residents in rural districts in the deep south where there's a Republican Representative. The same is true of Hispanics, who put up large percentages in Republican-dominated places like west Texas. The absolutely whitest districts, percentage-wise, tend to be in pretty thoroughly Republican places, like Appalachia and the rural midwest. (You'll notice one exception shows up below: upper New England, one of the whitest parts of the country, nevertheless distinctly prefers Democrats.)
Looking at the chart of Republicans in the least-white districts, I would have expected the top one to be Gary Miller in CA-31; that's a Hispanic-majority district in San Bernardino where Miller won a fluky victory thanks to a bizarre outcome in the state's new top two primary, where two Republicans wound up advancing thanks to a heavily split Dem field. However, he's second behind an unexpected neighbor: Ed Royce in CA-39 in north Orange County. This isn't a heavily Hispanic area, but the sum total of Asians plus Hispanics in this area is apparently large enough to push Royce into first place. Democrats haven't been very competitive in this area in the past (it is, after all, Orange County), but this part of the OC is rapidly changing and indeed the Dems are starting to come alive here (case in point, Sharon Quirk-Silva's surprise election to the state Assembly in 2012).
At #3 is Rob Woodall, who represents Gwinnett County in Atlanta's suburbs; this used to be one of the most right-wing areas in the country (it's Newt Gingrich's former district) but it's also one of the most rapidly changing areas in the country, based on a mix of new black, Hispanic, and Asian residents, and former residents white-flighting out of there to newer exurbs further north. Right behind him are two northern Louisiana representatives in districts with large black populations, but not quite large enough to elect a Democrat. (It's hypothetically possible to draw an ugly-looking second black-majority district in rural Louisiana—in fact, one used to exist in the 1990s—but one doesn't seem to be forthcoming.)
The factor with the second strongest positive relationship with Democratic performance is the percentage of 25-34 year olds. That's not too surprising; younger people tend to be more liberal in general, and the Millennial generation is, at least for now, a particularly Dem-friendly generation. However, you may be wondering why 25-34 year olds perform better than 20-24 year olds. Nobody's more idealistic than a college student, right?
Well, again, it has more to do with the districts as a whole than individual people. My guess is that 25-34 year olds tend to be out of school and newly-employed, and likelier to be in cities or close-in suburbs, so their numbers boost, but also get commingled with, the overall Democratic lean of urban areas. By contrast, 20-24 year olds are likelier to be in college towns, which, while they're very liberal pockets, are in many cases surrounded by and balanced out by rural areas. (A larger proportion of 20-24 year olds may also still be living with their parents, rather than having moved to the big city yet.)
In fact, here's the full list of what age groups correlate most and least with Democratic performance:
|Under 5||5-9||10-14||15-19||20-24||25-34||35-44||45-54||55-59||60-64||Over 65|
So what's with the large bulge in the relationship between Republican performance and 'tweens (10-14 year olds)? While it's tempting to point out the similarities between Republicanism and childishness, there's potentially a better demographic explanation, as the children themselves obviously aren't voting. Parents with children are more likely to have sought out more elbow-room in the suburbs or exurbs. You can also see that in the bulge of Republican performance among people on the older side of middle age. Note that I'm not attempting to assign any causation to the idea that having children, or owning a house, somehow "make" you more conservative, simply observing that families with children are more likely to be in the suburbs and suburban areas are more likely to elect Republicans than, say, urban areas.
Also worth noting: as much as we like to stereotype Republicans as being the party of angry old people—and, indeed, exit polling did show senior citizens as one of the GOP's strongest blocs—districts that are particularly heavy on the over-65 set aren't as strongly Republican as certain other districts. Plenty of seniors still remember the impact of the New Deal, or were unionists when labor was stronger, or both.
At any rate, you can see from the charts below that the districts heaviest on the 25-34 year olds are in major cities; in fact, these are fairly affluent, largely-white districts within major cities (Manhattan, north Chicago, San Francisco). The ones with fewest 25-34 year olds are, with the exception of a Long Island suburban district, all retirement areas in Florida, the four districts that would be on top if you ran a list of highest percentages of persons over 65.
The Republican-held districts with most persons 25-34 are mostly in Sun Belt suburbs and yield some familiar names: Kenny Marchant in the Dallas suburbs, Gary Miller in San Bernardino. There's one other name that stands out, though: Lee Terry in Omaha's NE-02. That's not a city that you think of as a real youth destination, but it does underscore a pattern that's especially pronounced in Terry's case among House GOPers: very competitive races in presidential years, easy races in off-years with lower youth turnout.
The second most significant factor that has a negative relationship with Democratic performance is the percentage of residents who are veterans. This makes some instinctive sense, since veterans are likelier to have the pro-military stances that one associates with Republicans. But there are also demographic explanations for that; percentage of veterans is for the most part a strong proxy for percentage of male residents, which, if you refer back to the very first table, is a strong predictor in its own right. It's also a proxy for the elderly, as WWII, Korea, and increasingly, Vietnam veterans are all in the over 65 category.
As you can see, the lowest percentages of veterans tend to be mostly non-white districts in major cities. Although it's a measure of veteran status, not active-duty status, the highest concentrations still tend to be around large military installations anyway, where both recent veterans tend to hang around and older veterans also tend to retire (Virginia Beach, Colorado Springs, Pensacola, Bremerton).
The districts with the fewest vets that are represented by Republicans, again, tend to be urban districts with mostly non-white populations (several in Miami, several in California). Staten Island's NY-11, held by Mike Grimm, is a bit of an exception; it's mostly white, though it's certainly urban.
Finally, let's look at one more category: the third-most-significant category with a negative relationship with Democratic performance. It's also the strongest correlation of any category of educational achievement. And it's kind of an odd one: the "some college" category. In other word, someone who took at least some college classes but never received a degree of any sort. That's one I don't have an easy explanation for; it may have to do with what the "some college" types aren't.
In other words, the persons with the lowest educational achievement tend to also be persons who are poor and/or persons of color, which are groups that tend to vote Democratic. And high educational achievement tends to correlate with Democratic voting. It's natural to think "rich doctors and lawyers, party of the 1 percent, bla bla bla," but in fact probably the largest group of persons with graduate degrees in the country is public school teachers with M.Eds, a pretty solidly Dem constituency. And, at any rate, in the last few decades we've seen well-educated suburbanites in general increasingly move toward the Democrats, perhaps motivated by the GOP's "Mayberry machiavellian" and generally anti-science, anti-progress tendencies.
That tends to leave persons of middling educational attainment leaning more toward the Republicans, creating kind of a U-shaped curve (which isn't limited to this kind of demographic analysis; it tends to show up in exit polling too):
|Less than 9th||9th-12th gr.||HS diploma||Some college||Associates||Bachelors||Graduate|
Of the districts with the lowest percentage of "Some college" residents, they're all in or near northeastern cities; four are wealthy districts with some of the highest percentages of persons with bachelors and graduate degrees, while the other (NY-07) is heavily skewed toward the poorly educated end of the spectrum. Interestingly, the "some college" category is at its strongest in the rural west; it's a heavily Republican bunch of districts, with Peter DeFazio's OR-04 the only exception (a district that's light-blue only because of the presence of, ironically, two large universities). In fact, you have to work your way through the 22 most strongly "some college" districts before you find one that isn't in the west (and that one's in Texas). I'm not sure what it is about the rural west that encourages starting and leaving college, but it's definitely a trend.
The list of Republicans in the district with the fewest "some college" residents is kind of a mishmash: there are two districts in affluent northeastern suburbs that are more skewed toward graduate degrees. Jim Gerlach in PA-06 is used to seeing strong challenges, while Rodney Frelinghuysen in NJ-11 certainly should, especially since his district had to become bluer in redistricting to accommodate Republican plans to wipe out one Democratic district. There are also two rural districts in Pennsylvania's Appalachian-flavored middle that lean heavily toward high school grads, and then there's FL-27 in Miami, which seems rather polarized between the well-educated and the poorly educated without much in between.
|Rank||District||Rep.||Some college %|
|Rank||District||Rep.||Some college %|
|Rank||District||Rep.||Some college %|
You may have noticed some of the same districts keep showing up on the lists of Republican "outlier" districts: two of them are CA-31 and CA-21, which Dems lost in 2012 mostly because of recruiting failures and which are at the top of everyone's target list this year. But there are some other recurring ones, like CA-39 in Orange County, TX-24 in the Dallas suburbs, and FL-27 in Miami, that have never been serious Democratic targets before. They might not be targets this cycle either, but soon should be, given the accelerating pace of demographic change.