In my last post I noted the historically high turnout (at least since 1990) – close to 32% of registered Republicans- in the Delaware Republican primary Senate race that was won by the Tea Party-backed candidate Christine O’Donnell. She beat the favored Congressman Mike Castle by a decisive 53.1% to 46.5% margin. While acknowledging that O’Donnell faced an uphill battle in the general election, I suggested that the high turnout was still another indication that the Tea Party had successfully tapped into voter anger during the current electoral cycle – anger that did not bode well for Democrats in the midterm elections.
In response, a political science colleague sends me the following turnout data that puts O’Donnell’s support in a slightly different perspective:
The take home point is that O’Donnell’s victory is due to the votes of about 5% of Delaware’s registered electorate. It’s a useful reminder that we should not overstate O’Donnell’s – or the Tea Party’s – overall level of support among all voters. And it drives home two points that I’ve made before. First, the current high level of partisan polarization in Congress is not the product of a polarized electorate as much as it is due to moderate voters having to choose between more ideologically extreme primary winners. Unless Castle runs an independent campaign, much like Republican Senator Lisa Murkowski is doing in Alaska after her defeat in the primary there, Delaware voters who wish to cast a protest vote in the general election against the party in power must do so by voting for O’Donnell – even if they don’t agree with all her policies. The alternative is not to vote at all. Now, I am not suggesting that O’Donnell will win the Delaware Senate race in November (although I confess to having no first-hand knowledge of Delaware political dynamics). But to the extent that the Delaware Senate race becomes a referendum on Obama and the Democrats, she will be the beneficiary – even if moderate protest voters don’t necessarily share her views. Note also that O’Donnell is frantically backing away from her most controversial past statements in an effort to portray herself as a more centrist candidate. We will know come November whether she is successful, but most pundits don’t think much of her chances.
My second point is that even during “wave” elections, such as the Republican takeover in 1994, most voters’ underlying partisan preferences and voting habits don’t shift from one party to the other. Remember that in the more competitive Senate races, since 1982 incumbent senators have nonetheless won reelection at about an 83% clip. (The House rate is even higher.) In this election cycle there may be 10 Senate races that will truly be competitive. However, in competitive races, a small shift in underlying voter sentiment is often enough to swing a race one way or the other. And, in the aggregate, a shift in a few percentage points in voters’ support for either party can be enough to produce a change in congressional control. For example, the Democratic takeover in the 2006 midterms reflected only a 4% Democrat gain in the aggregate national vote. This small shift in voter sentiment was enough to end the Republican majority.
In short, changes in the partisan makeover within Congress often are driven by a change in voting habits of only a small sliver of the electorate. A movement that has the support of only about 20% of voters, as most polls suggest is the case with the Tea Party, can nonetheless play a significant role in congressional elections, particularly if the 20% is more motivated to vote. The significance of O’Donnell’s win, then, has less to do with her chances of winning the Senate seat in Delaware, and more, I think, with what the turnout in her primary victory says about the national mood, and the relative chances of Republicans and Democrats come November.