In the days since I posted my case for a blowout Republican majority in this fall's elections, a number of people have helpfully sent data and other tips for constructing a comprehensive target list of Democrat-held seats we might be able to pick off without warning in a year like this one.
One started with a copy-paste of the Cook PVI (Partisan Voting Index) scores from Wikipedia for each district. The PVI is a crude metric -- unlike this Thursday's British elections, incumbency matters a great deal in Congress, keeping a minority party incumbent in office long after his constituents have started voting the other way at the Presidential level. Ask Gene Taylor, Chet Edwards, Joseph Cao and Scott Brown about the predictive value of the Cook PVI.
Nonetheless, even a cursory glance at the PVI numbers exposes, at a 30,000 foot level, a massive Republican underperformance in the House that was there even when we were in the majority, and indeed, may show vestiges of our weakness from our 40 years in the wilderness. For this analysis, I'm not interested in individual districts, but macro-level trends. Suppose every incumbent resigned tomorrow and we held special elections in every district whose overall results would mirror partisan preferences in each district. In the long-term, after all, we are headed for a total cycling out of current incumbents, to be replaced by representatives more in tune with the views of their districts. What would the results be?
PVI scores show that it would be a Republican blowout: In 234 districts, Republicans perform above average, compared to just 192 districts for Democrats, and 9 that are tied. Split these evenly, and you've got a 239-196 Republican House. This is a shade above the biggest Republican majority in their 12 years in power -- and that's when Republicans perform as expected.
We talk about the Republicans taking over vestigial Southern Democratic House seats in 1994 as though political cognitive dissonance were a thing of the past, but in reality it persists to this day, and in both chambers. Democrats have been simply better at electing Democrats in Republican-leaning districts than Republicans are at returning the favor. How else to explain that there are 30 or so Republican states -- and so, 60 Republican-leaning Senate seats -- in a tied electoral college but the Senate is 59-41 against us?
I wondered if there were something wrong with the Cook data, whether it had come to be out of balance since 2008. But in fact, a quick tally of total PVI scores on both sides show that Democrats have more overall strength in Dem PVI seats than Republicans do in their greater number of seats, by a total of 2,744 to 2,393 PVI points. It turns out that Republicans are more evenly distributed, with the hulk of net Democratic votes crammed into a smaller number of urban seats.
Democratic overperfomance in the House grows more striking in swing districts. Let's take a look at seats within 5 points of the national average, and which party they're represented by:
| PVI |
Dem |
GOP |
| D+5 |
11 |
0 |
| D+4 |
7 |
1 |
| D+3 |
11 |
1 |
| D+2 |
7 |
1 |
| D+1 |
7 |
2 |
| TIE |
4 |
5 |
| R+1 |
8 |
2 |
| R+2 |
2 |
2 |
| R+3 |
4 |
7 |
| R+4 |
5 |
5 |
| R+5 |
8 |
6 |
Now, let's group these together into marginally Democrat and Republican seats by PVI:
| PVI |
Dem |
GOP |
| D +1 to +5 |
43 |
5 |
| TIE |
4 |
5 |
| R +1 to +5 |
27 |
22 |
These pretty much speak for themselves. We get crushed by a net 38 seats -- 88 to 12 percent in percentage terms -- in seats that lean Democrats. And in lean Republican seats? Democrats beat us there too, by a smaller 27-22 margin. Somehow, we manage to miraculously win the tie seats, giving hope that victories anywhere in political swing districts are attainable. Overall, Democrats hold 74 "swing" districts to Republicans 32, a net of 42 from a quarter of the whole House.
Sure, you might say this is expected after two good Democratic election cycles. And I can buy that: these numbers show that big changes in the electorate reflect easily in the overall House tally, lending credence to the potential for a big pendulum swing in 2010.
Yet this doesn't solve the fundamental question that during our high water marks after 1994 and to a lesser extent 2002/04, we weren't able to raid lean Democratic seats to nearly this extent. And it does raise the upside question of whether doing so might be possible in the future by boldly targeting more seats.
Another way to visualize the upside potential is to consider the fact that while Republicans hold just eight net Dem PVI seats, Democrats hold 69 of "our" seats. (Those eight seats, in case you're wondering, are NJ-2, OH-12, PA-15, WA-8, PA-6, IL-10, DE-AL, and of course, LA-2. Democrats have a serious chance at picking off the last two, even in 2010.)
Meanwhile, nearly a third of House Democrats hail from districts that were won by Bush and/or McCain. In MS-4 and TX-17, Gene Taylor and Chet Edwards hold the 18th and 19th most demographically Republican seats in the country. The only remotely comparable example is Joseph Cao, in the 28th most Democratic seat. Taylor and Edwards getting re-elected is the direct equivalent of a Republican winning Jan Schakowsky's district on the north side of Chicago. And nor are these two outliers: Democrats routinely get elected in R+10 PVI seats or better. Democratic performance in Republican seats between a +10 and +15 PVI is better than Republican performance in seats between a +1 and +5 Democratic PVI.
This is why winning back the House alone is not enough. We could get the needed 40 seats by beating every Democrat in an R+5 seat or better. Getting to my outlandish speculation of 70 seats would mean taking out every Democrat in a Republican-leaning seat (that's 69 seats) plus one tie district. And that's before any net takeovers of Democrat PVI seats, which we ought to be winning in spades in a year like this.
There's a reason why American Congressional elections aren't nice and clean as this analysis would suggest. Old bulls like John Spratt in SC-5 don't go easy. And for the longest time, we didn't challenge these Democrats. For the most part, we are this year. Retirements will also be our best means of forcing change on these districts, and those can come all too slowly.
Still, a few conclusions suggest themselves:
- The inexorable tide, all things being equal, is for a more Republican House.
- Democrats have been able to defy this trend by 1) having more popular Southern holdovers, 2) seeking out and destroying moderate Northeast and Midwest Republicans in a way Republicans haven't been able to do down South, and 3) under Rahm Emanuel's leadership, boldly targeting more takeovers deep in enemy territory, like ID-1 (Walt Minnick) and NC-11 (Heath Shuler).
- Picking off the "easy" seats should be a gimme. If we can't beat Chet Edwards this year, we're just going to have to wait till he dies or retires. Guys like him will be hardened targets. Watch those swing Democrat seats as they are the soft underbelly of the Democratic majority. There is no reason they should have a 9-to-1 edge in those seats. Getting to even in those districts would give us half the seats we need for a takeover off of just over 10% of the House.