Search Archives:

Custom Search

Wednesday, May 19, 2010

Primaries Don't Tell Us Much

Normally, primary midterm elections would be about as newsworthy as the weather; everyone would report on it one day and then it stops being news. But this year is different. The media has set up a narrative about this election cycle -- gonna be a big year for Republicans -- and anything that happens from here until November will serve that narrative. Part of this line lies in some simple math; it's a bad year for incumbents, Democrats are in the majority, therefore Democrats will bear the brunt of a big anti-incumbent sweep.

And last night we saw incumbents do very poorly, for the most part. In Arkansas, Sen. Blanche Lincoln is forced into a runoff for her party's nomination and, in Pennsylvania, long-time Sen. Arlen Specter lost to Rep. Joe Sestak. For their part, the New York Times is sticking with their story:

The results were sobering for both parties, amounting to a rejection of candidates selected and backed by leaders in Washington who found themselves out of step with their electorates.

Republicans and Democrats alike are now left to learn the lessons from the frustration being expressed by voters, and to unify behind nominees who face daunting general election campaigns.

But using Blanche Lincoln as representative of all incumbent Democrats is crazy and doing the same with Specter is even more so. Specter switched parties just last year -- admitting that he stood no chance of winning as a Republican. Democrats were cheered by the math -- Specter helped seal the 60-seat majority at the time -- but weren't so happy with or trusting of the man. Sestak, on the other hand, is a real Democrat and, as a former two-star admiral, the highest-ranking former military officer in congress. Sestak is by no means a lightweight. Yes, the White House backed Specter, but what else could they really do? He'd jumped the Republican ship and, if Obama had left him twisting in the wind, that'd be a real disincentive for other Republicans to do the same. In any case, as Specter's chances became weaker, White House support for his candidacy did the same. That election is a unique case and treating it as representative of anything other than itself is a mistake.

As for Lincoln, her determination to be a speedbump on the road to healthcare reform cost her big. She is not the favorite of national Democrats and certainly not of Democratic voters. Like Specter, Lincoln has establishment support, but only because of the advantage an incumbent holds. And using this election as evidence of an anti-incumbent mood is even farther off-base -- under almost any other state's rules, Lincoln probably would've won. The last count I could find was 45%-43%. In Arkansas, a majority -- 50%+ -- is needed to avoid a runoff. It's a good law and one I wouldn't mind seeing nationalized (imagine a world without Bush v. Gore), but the "Voters Reject Lincoln" headlines are just plain wrong, no matter how you look at it.

Yes, the teabaggers have rejected Utah Sen. Bob Bennett and Rand Paul won over the GOP establishment choice in Kentucky, but the Tea Party isn't everyone -- no matter how hard the media pushes that meme. In all of the cases, looking at party primaries for clues about the general election seems silly. In the Republican primaries, the craziest candidates seem to be winning -- which probably isn't excellent news for the GOP. And the Democratic primaries are a mixed bag -- just like they usually are.

If you want some sort of a glimpse at how the general is going to go, you need to look at the one actual election that was held last night. In Pennsylvania, voters chose Democrat Mark Critz over Republican Tim Burns for the late John Murtha's seat. RNC chairman Micheal Steele had predicted a win in that district and laughed off the idea that a Democrat had any chance at all. "What are you talking about?" Steele told the Washington Post's Chris Cillizza when asked about the possibility of a loss. "Oh, come on. We're going to win."

Now granted, there was no incumbent in that race, so we can't use it to measure any anti-incumbent mood. But that doesn't mean it's meaningless. Pennsylvania's 12th is exactly the sort of swing district that Republicans will need to take in order to win big in November (it went to McCain in 2008). And they didn't. Whizbang statistician Nate Silver explains:

[I]f the Republican Tim Burns were to win by about 5 points, that would give us one indication that the cycle was shaping up to be more like 2004, from which the Republicans emerged with 232 House seats. And if Democrat Mark Critz were to win by, say, 8 points, that would give us one indication that the cycle was shaping up to be more like 2006 or 2008.

Critz won by eight points. While it's clear that Silver was throwing out a rough estimate on the fly, the weather's looking slightly better for Democrats than Republicans here. Silver says we can read something from this race, but "not necessarily very much." Personally, I agree, although I'd argue that even that "not very much" is more than any party primaries will tell us. Kentucky and Pennsylvania both have closed primaries, meaning only people registered with the respective parties vote -- i.e., only Democrats vote in the Democratic primary and only Republicans vote in the Republican primary. Using these to read the minds of the voters in the general election is more than a little crazy. And, if you had to declare a winner in Arkansas -- the open primary of the night -- you'd have to give it to Lincoln, the incumbent.

I guess in the end what I'm saying is that a lot of people are going to be saying a lot about what all the primaries last night meant and that you should probably not put too much faith in that analysis. We can see that Democrats seem to like real Democrats, Republicans like real crazy people, and voters in one swing district went blue. Add it all up and I have to give the night to Democrats. Just not by much.


Get updates via Twitter

1 comment:

vet said...

In the 24-hour news cycle, stories are prioritised according to a very simple algorithm:

1. Is anyone paying us to cover it? If so, it's news.

2. Do our viewers care? That is, do they already know about it, or will shortly hear about it from some other source? If not, it's probably not news, although it might still qualify under rule (1).

3. Can we talk about it without knowing any more than we already do, plus what one reporting team can learn in 2 hours or less? If not, it's probably not news, unless it scores exceptionally highly on rules 1 and 2.

4. Is it going to give us something to talk about, feeding in "new" news on a reliable schedule, over a long period? That is to say: is it a story that will tie together a long series of predictable events that satisfy criteria 2 and 3 above? If so, it's big news. If it meets criterion 1 as well, it's da bomb.

US national elections - of any kind - are pretty much the perfect story. Every election from now on is going to be huge news, until your country either comes to its senses on campaign finance, or redefines what constitutes "the media".