The reaction among pundits to Christine O’Donnell’s surprising win over Representative Michael Castle in yesterday’s Republican Senate primary in Delaware has been interesting, to say the least. Polling done early in this race suggested that O’Donnell, who lost two previous bids for the Republican Senate nomination, was destined to lose a third time. Instead, she won convincingly, 53.1% to 46.5% – a victory that likely owed something to support from the Tea Party movement.
If the pundits are to be believed, however, Republicans have just snatched defeat from the jaws of victory. Castle, a well-known Republican with a moderate voting record (that record placed him as the 11th most liberal member of the Republican House caucus) was viewed by most observers as the stronger candidate, one capable of beating the Democratic nominee Chris Coons in the general election for the Senate seat previously held by Joe Biden. O’Donnell’s win was another sign, according to conventional wisdom, that the Republican Party has been captured by the ideological loony-toon fringe who seem intent on purging the party of any sane individuals that might possess centrist predilections. Karl Rove captured this sentiment well when, in dismissing O’Donnell’s chances of winning in the general election, he cited the “nutty things” she has been saying. Similarly, Politico led their Delaware coverage with this paragraph: “The Republican Party’s hopes for winning back the Senate rest on a perennial candidate with a sketchy employment history who has dissembled about her education, defaulted on her student loan and her mortgage, sued a former employer for mental anguish, railed against the evils of masturbation and questioned whether it would have been OK to lie to prevent Nazis from killing Jews during World War II.”
With those credentials, it’s hard to believe she won! But then again, we have heard this refrain before after Tea Party-backed candidates secured Republican Senate nominations in Nevada and Utah, and when the libertarian candidate won the Republican primary in Kentucky. After each election pundits suggested that the Republican Party, by nominating a fringe candidate, had actually lessened its chances of securing the Senate seat in the general election. For Tea Partiers and their ideological kin, the critics suggested, principle appeared to matter more than politics.
Well, maybe. I understand the logic behind this argument, because it is a logic I’ve pushed myself, in previous blog postings, to explain the growing polarization of Congress. Turnout in primaries tends to be low, and disproportionately weighted toward the more ideologically-extreme voters in each party. In the general election, however, turnout is higher and independents and more moderate partisans are more likely to participate. The Tea Party candidate is thus likely to do less well in the general election.
As the Tea Party keeps defying the odds and wracking up victories, however, one must ask whether it is really a fringe movement, or whether it has successfully tapped into the sentiment of a substantial portion of the electorate – one which is likely to act on those sentiments and vote in the general election too?
Consider the Delaware Republican Senate primary. The Times Jeff Zeleney notes that “The primary drew 57,000 voters, a small slice of the overall electorate.” What he did not report, however, is that this number represents a 32% turnout rate among registered Republicans. I searched the Delaware state records for statewide primary turnout in all elections dating back to 1990 (the earliest year for which I could calculate turnout data) and guess what? Yesterday’s turnout was the highest recorded (at least as far as I could tell) for almost any primary – presidential or midterm election – since then. (Here is the turnout data I was able to cull from the Delaware records for all the primaries, including presidential primaries, dating back to 1990. It’s incomplete, and I had to do some back-of-the-envelope calculations to derive many of the figures, so use these numbers with caution. (Note that I am comparing turnout rates – not the number of voters. ) Moreover, there are surely contextual differences that make comparing turnout across different elections problematic. Nonetheless, the data indicates yesterday’s turnout was extraordinarily high – higher, even, than turnout for primaries held in some presidential election years.
|Year of Primary||Republican Turnout||Democrat Turnout|
|2004||15% (Dem Pres. Primary)|
|1996||23.5% (Presid Primary)||6.5% (Presid. Primary)|
Note that in 2006, when O’Donnell finished a distant third in the Republican Senate primary, statewide turnout was only 8%. When Castle won the U.S. Representative seat in 1998, turnout in the general election was only 39%. (Of course, this represents a far bigger voting pool. I don’t have totals for the primary for that year, or for 2000).
My point is that I think it’s difficult to categorize O’Donnell’s win as simply the triumph of the narrow slice of the Far Right fringe brigade. In fact, the focus on O’Donnell’s views may be misplaced. I suspect this was as much a vote against Castle as it was a vote for O’Donnell. Indeed, all the virtues the pundits cite in Castle’s favor – his moderation, experience (read: age – he is 71) and long history in Delaware politics – are precisely the factors that cost him the election. This doesn’t mean O’Donnell will win the Senate seat come November. The latest polls have her losing by double digits to Coons. Delaware is a solidly blue state, with 47% of registered voters affiliated with the Democrat Party compared to 29% Republican. (There are 24% in the “other” category.)
And yet – how many voters will show up to vote in the general election come November? I do not have much knowledge regarding Delaware politics, but it would not surprise me if turnout was up, particularly among Republicans. Having listened to the same conventional wisdom (of which I was a part) in the run-up to the election to replace Ted Kennedy, and with polls there showing Martha Coakley far ahead of Scott Brown two weeks before the special election, I am far more skeptical now that the punditocracy really has an accurate read on voter sentiment this election cycle. We may get a clearer picture when the first post-election polls begin coming in.
In the meantime, the Delaware result, I think, is relevant to the larger issue I raised in yesterday’s post, when trying to explain the difference in the political science forecast models. Those models that incorporate some measure of voter sentiment, such as the generic ballot results, are predicting a much greater seat gain for Republicans than do the models that are based only on structural factors, such as economic measures. O’Donnell’s victory may – and I stress may – provide some evidence that the models incorporating voter sentiment are likely to be more accurate. Voters are angry and are taking out that anger on anyone associated too closely with the Washington establishment. This sentiment may not be broad enough to propel O’Donnell to victory in November (although I am not counting her out). But it doesn’t bode well for the Democrats more generally.