Skip to main content
The Not-So-Bad-Case Scenario

If the conventional wisdom of the summer was that Democrats were in deep, deep trouble when it came to the midterms of 2010, the emergent theme of the fall seems to be that things might not be so bad. See Paul Krugman or Ezra Klein for an articulate distillation of this; the argument in a nutshell is that, for all the trouble the Democrats might get themselves into between now and next November, the Republican brand is simply too damaged for them to capitalize on this in more than a marginal way.

Indeed, there is little sign that very many Americans have come to view the Republican Party more favorably. Party identification among all adults hasn’t shifted much (to the extent it has, independents have gained a couple of points at both major party’s expense) and in yesterday’s Washington Post poll hit 20 percent for Republicans, their lowest figure since 1983.

But most adults don’t vote in midterm elections — only about 40 percent do. And there’s still plenty of suggestion — both anecdotal and statistical — that those adults who are most inclined to vote next year will tend toward the conservative side of the ledger. Over the past two months, 11 distinct organizations have released ‘generic ballot’ polls asking voters how they’re planning to vote in their House district. Those polls show quite a strong discrepancy depending on the sample frame used:

Between three polls (ABC/Pos, Bloomberg, and YouGov) that sample all adults, Democrats hold a 10-point lead — not much different than the 10.5-point margin by which they won the House aggregate popular vote in 2008. But their advantage is reduced to 7.4 points in polls of registered voters — which would probably correspond to a loss of perhaps 10-15 seats — and just 1.4 points in polls of likely voters, which might correspond to a 30-40 seat loss.

Now, I’m generally not a huge fan of applying a likely voter model more than a year in advance of an election — enthusiasm, which is essentially what a likely voter model is measuring — tends to be more fluid than one’s underlying political preferences. But this does present, in my view, a highly plausible scenario. No, the Republicans will not present a particularly compelling rationale for voting for them (they certainly haven’t so far). But their anti-Democratic message could tend to capture the most angry and motivated voters — people who are not voting “Republican” so much as they’re voting “not Democrat”.

Needless to say, there are a lot of contingencies. Will the Tea Party/Glenn Beck crowd go to the polls in large numbers, or are they so disillusioned that even that feels like selling out to them? Will Democrats be motivated to turn out to support a President who has managed to pass, say, a major stimulus package, a big health care bill, and a financial regulation initiative — or will their expectations have been so high that they’ll prove to be complacent? What role will third party candidates play? (They’re playing a very important role right now in New Jersey and NY-23). And of course, what will the economy and the situation in Afghanistan look like one year hence?

Over the summer, I estimated about a 30 percent chance that Republicans would take over the House. I would now dial that back slightly to 20 or perhaps 25 percent, as key metrics like Presidential approval and the generic ballot appear to have stabilized for the Democrats. The bottom, for now, doesn’t seem to be falling out. But it still could, and even if it doesn’t, the enthusiasm gap remains something for them to worry about.

Nate Silver is the founder and editor in chief of FiveThirtyEight.