The Obama Coalition, the Enthusiasm Gap, and the Upcoming Midterms

I’ve stopped reading Fivethirtyeight as much as I used to since they moved over to the New York Times. The analysis is still good, but something seems missing. Perhaps the stringent NYT policies have neutered Nate and the gang a little more than I was anticipating. He rarely editorializes any more, which I suppose is a good thing for fans of objectivity. But it’s hard to run a successful election blog if you never talk about the likely policies implemented by those who win them. Maybe I’m just biased, but I seem to remember Nate taking a more proactive stance when it came time to call a teabagger a teabagger before he got into bed with the Gray Lady.

As a result of this real or perceived change, his analysis has become a lot less interesting. Predicting Republican gains of 52 seats in the House and 7 in the Senate doesn’t mean anything in the abstract. I used to be addicted to the Fix, Chris Cilliza’s blog over at the Washington Post. Then, during the 2006 election season I was forced to endure post after post about campaign insider bullshit and “balance” from both sides. Even after it became pretty clear Democrats were headed for big gains, we still heard from one Republican strategist after another.

This is all a roundabout way of saying that the analysis for this election, and I’m beginning to think, all midterm elections, is destined to be rubbish. And if it’s all rubbish, then my rubbish will fit right in with the rest of it, even if it tends to be more bullish than bearish on the Democratic prospects. So get ready for some serious unfounded speculation.

Obama won in 2008 with a very solid coalition that included a lot of first time voters, minorities, young voters, etc. As you’ve no doubt heard, this very coalition is the kind that does not hold up in midterms. Only old white people vote in midterms, and those Sids and Berthas are ready to deliver the election to the GOP. Not only that, but the tea party has them so riled up that they can’t wait to roll out of the home, get on the free shuttle to the elementary school, and cast their vote for Barry Goldwater. Why, he’ll show those Russkies whatfor!* Polling has pretty consistently shown an enthusiasm gap that, coupled with the already more conservative electorate expected, will drive the GOP on to unprecedented heights. I’m not going to argue that, in years past, this would almost definitely have been the case.

*Getting sick of the old jokes yet?

That said,  in his analysis today, Nate Silver says that a 2% bias in polling would give Democrats a 50% chance of holding the House. Basically, if the national polling average shows is biased towards Republicans by only 2%, Democrats would be a coin flip to hold on to both Houses of Congress. Is it likely? No, but it’s not unheard of. As Silver points out, in 1998 the polling was off by 5%.

Basically, the best chance Democrats have right now comes from professional polling companies being relatively terrible at their jobs. It’s not very likely, is it? Well, no. But I have a few things that give me a little ray of hope:

• The Cell Phone conundrum: I suppose if you follow polling with any degree of seriousness, you probably already have an opinion on this. Basically, it’s much harder to call cell phones for a poll than land lines. Since cell phone-only households tend to be younger and poorer than tradition phone homes, theoretically this would skew any sample towards the Republican candidate. Most pollsters compensate for this problem by simply weighting certain demographics more heavily than others. At some point, polling goes from scientific (getting a good sample, weighting questions properly, analyzing trends) to artistic (applying “likely voter” screens, weighting samples). So, if a pollster doesn’t feel like he’s reached enough 18-24 year olds, he’ll take the ones he did get and make their opinions count for more until he has what he believes to be a representative sample. You can probably already see where the problems might arise. So if pollsters across the board are underweighting the cell-phone only users, you could see bias arise. The other problem is actually reaching enough people from the correct demographic to get a representative sample to weight at all. Are these problems likely? Not especially, perhaps. But remember, it’s only 2%.
• The Organizing for America machine is still in operation, albeit running at less than full capacity. Just because there’s an enthusiasm gap doesn’t mean people won’t be reached by GOTV operations and driven to their local polling place. It was a major advantage for Democrats in 2008. Will it be enough to overcome the enthusiasm gap? Probably not, but it’s only 2%. There’s a reason slimy Republican lawyers have crawled out of their homes in Gringott’s and on the planet Dagobah to try to drive down participation with voter “fraud” schemes.
• Does anyone really know how moderates and independents are going to vote and in what numbers? Surveys have consistently shown that even when Republicans are winning, the party has very negative ratings. Presumably, the independents who show up will tend to be right-leaning. But what if there are left-leaning independents who are being left out of these polls who are really scared of the teahadist agenda? Democrats only need to get them to the polls.
• Finally, and this is really cold-hard baseless speculation here, nobody really knows how the Obama coalition will actually hold up in a midterm. Don’t forgot, Obama won in places no Democrats could have been expected to win in just 4 years earlier. I always harp on about 2008 being a realigning election. If that’s the case, then Democrats are pulling from a very different, much larger base all over the country. Just because the old Democratic coalition was weak in midterms, it does not necessarily follow that the new one will be. Historically, yes, these groups tend to show up less frequently in midterms. But maybe the pollsters this year have made assumptions about this new coalition that are no longer accurate. Is it especially likely? No, of course not. But again, it’s only 2%.

I suppose we’ll know on election day how accurate the polling was.  But as I’ve said before, the Democrats have a very low bar to clear next week. And it’s not outlandish for them to clear it comfortably. The current projections are around 25% to hold the House, and that’s probably accurate. How unlikely is 25%, really? Well, pull out a deck of cards and pick one. If it’s a diamond, the Democrats hold the House. And if any of the assumptions pollsters have been making are off by only 2%, all you have to do is pick a red card from the deck. The only difference is everyone is predicting, nay, expecting, that you’ll pick a club, and if you don’t, everyone wins a dollar while the clubs and spades have a thoroughly entertaining civil war.

On another note, if you’d like to read a little historical perspective on the tea Party and the Know-Nothings from the mid 19th Century, why not check out this Op-Ed from the Boston Herald? Full disclosure, it was written by my father. Don’t worry though. You won’t find the same rambling nonsense from him that you do here.

Advertisement