The Demographics of Swing State Project

Now that the open thread about what Congressional districts the Swing State Project readership hails from has died down, I thought I’d pull together some data to try and make some generalizations about what type of places we inhabit and who represents us.

This is a question that has always nagged me in the past when reading blogosphere calls to action. In other words, when all of us in the blogosphere get off our butts and call our representatives and complain, are we preaching to the choir? It’s been documented that the liberal blogosphere is more metropolitan, more affluent, more educated, whiter, and maler than the population at large. Does that mean that we are concentrated in urban, heavily Democratic districts, to the extent that most of us are represented by progressive reps who are already voting the way we want to, regardless of our demands? Or, based on the fact that many of the most educated and affluent districts are suburban swing districts, where our input might actually have some impact on a representative facing competing demands and potentially competitive elections?

One more Daily Kos demographic post by DrSteveB from May 2007 (so reflecting Congress’ current composition) sheds a little light on this: “Is Your Congressperson a Dem or Rep?” 63% of the respondents (sample of 2,610) said that they are represented by a Democrat, and 47% indicated that they would not support a primary challenge to their representative (most likely indicating satisfaction with the progressiveness of their representative, although it may also indicate resignation to their representative’s conservativeness as being acceptable given the district’s lean). 37% of the respondents were represented by a Republican. Compare this with the overall composition of the House, which is 54% Democrat and 46% Republican. Daily Kos is disproportionately represented by Democratic representatives.

To my surprise, this almost exactly matched the results from the much smaller sample here at Swing State Project. I found a sample of 81, using comments but not the Frappr map (not many Frappr participants actually cited their district, and those that did were often the same people who participated in the comments). Where commenters (usually college students) mentioned living in multiple districts without saying where they were registered, I assigned them to their ‘home’ (i.e. parents’) districts.

Here’s how we at SSP break down:

But by knowing specifically which districts each respondent lives in, we can go a lot further than the Daily Kos survey did. For instance, we can check out what ideological caucuses our representatives are members of. Look at the first line: 16 of the 81 SSPers are represented by members of the Progressive Caucus, or 20% of us. In reality, Progressives are 68 of the 435 in the House, or 16% of the House members.

We’re disproportionately represented not just by Progressives but even more so by New Dems. Interestingly, we’re also disproportionately represented by Main Street Partnership members (maybe not surprising, since they tend to be concentrated in affluent and educated suburban districts). We’re under-represented among Blue Dog and RSC constituencies (again not surprising, since these tend to be the rural and less-educated districts). (Don’t look for these numbers to add up to 100, as many members belong to more than one caucus.)

We can also take a look at the ranked liberalness or conservativeness of our representatives. For this, I’ll use National Journal composite scores from 2007 (since they’re already an attempt to scale reps on their liberalness from 0 to 100). On average, our reps are more liberal than average, but, oddly, we’re under-represented by representatives who are in the top decile for liberalness. That may have something to do with the fact that we’re particularly over-represented by New Dems, while not being over-represented by CBC members, many of whom are among the House’s most liberal members.

Liberalness

Our median NJ score: 63.65 (Overall median is 50.75)

Our range: 95 (Al “Ooops, I’d better veer left because of my primary” Wynn in MD-04) to 7.7 (Virginia Foxx in NC-05)

6 of 81 (7.4%) are in top decile for liberalness (score of 87.3 or more)

29 of 81 (35.8%) are in top quartile for liberalness (76.8 or more)

14 of 81 (17.3%) are in bottom quartile for liberalness (21.3 or less)

5 of 81 (6.1%) are in bottom decile for liberalness (14.5 or less)

SSPers also tend to inhabit districts with a lean that is predisposed toward the Democrats at the presidential level. Only 29 out of the 81 of us live in districts with a Cook PVI rating that is Republican-leaning, and only 15 out of 81 live in districts with a rating of R+6 or more (which is where I’d start to say “that’s pretty red”).

PVI

Our median PVI: D+5 (Overall median is R+1)

Our range: D+43 (NY-15) to R+16 (TX-07)

12 of 81 (14.8%) are in top decile for PVI (D+22 or more)

29 of 81 (35.8%) are in top quartile for PVI (D+10 or more)

10 of 81 (12.3%) are in bottom quartile for PVI (R+10 or more)

5 of 81 (6.2%) are in bottom decile for PVI (R+15 or more)

Let’s look at a few other demographic indicators. Overall, SSPers are an extremely metropolitan bunch (it’s hard to break down ‘urban’ vs. ‘suburban’ because a lot of districts contain elements of both, and the census bureau uses a binary system where someone is either ‘urban’ or ‘rural,’ although I’ve observed that many districts that are 5-10% ‘rural’ tend to be what you’d think of as stereotypical suburban districts). The number, for each district, represents the census bureau’s count of people living in a ‘rural’ environment.

Ruralness

Our median ruralness: 5% (Overall median is 15.7%)

Our range: 64% (VA-05) to 0% (12-way tie)

5 of 81 (6.2%) are in top decile for ruralness (50.6% or more)

10 of 81 (12.3%) are in top quartile for ruralness (35.8% or more)

33 of 81 (40.7%) are in bottom quartile for ruralness (1.5% or less)

12 of 81 (14.8%) are in bottom decile for ruralness (0%)

SSPers tend to come from affluent districts. That, of course, doesn’t mean that they themselves are affluent, just that they live among people with high per capita incomes. (Especially considering that we seem to have a large number of college students and post-collegiate activists here.) These are using 2000 census numbers for each district’s per capita income, so bear in mind that these numbers have gone up even more (at least in some parts of the country).

Per capita income

Our median PCI: $23,208 (Overall median is $20,529)

Our range: $47,498 (CA-30) to $14,021 (CA-38)

21 of 81 (25.9%) are in top decile for PCI ($28,560 or more)

36 of 81 (44.4%) are in top quartile for PCI ($24,527 or more)

7 of 81 (8.6%) are in bottom quartile for PCI ($17,820 or less)

2 of 81 (2.5%) are in bottom decile for PCI ($15,277 or less)

And the area where SSPers seem most out of whack with the nation, even more so than per capita income, is education. Look at the numbers, which are each districts’ percentage of persons 25 or older with at least 4-year college degrees.

Education

Our median education: 30.5% (Overall median is 22.6%)

Our range: 53.8% (VA-08) to 12.5% (CA-38)

22 of 81 (27.2%) are in top decile for education (36.5% or more)

45 of 81 (55.6%) are in top quartile for education (28.9% or more)

6 of 81 (7.4%) are in bottom quartile for education (17.5% or less)

1 of 81 (1.2%) are in bottom decile for education (14.1% or less)

Taken as a whole, we can see that Swing State Project members (or at least the ones who responded to the question) are disproportionately represented by Democrats, and by Progressives or New Dems in particular. We’re coming from districts that are disproportionately urban, affluent, and educated. And when we get in touch with our representatives, many of us are getting in touch with someone who already shares our values.

(I’m probably as good a case in point as anyone. I’m in WA-07, which is Seattle. We’re represented by Jim McDermott, who’s in the Progressive Caucus and in the top quartile for liberalness. The district is in the top decile for PVI, educational attainment, and per capita income, and the bottom quartile for ruralness.)

For those who are interested in the full data set (and I know you’re out there), go to Google Docs for the database.