Playing Around with SUSA
by Matt Stoller, Fri Nov 24, 2006 at 07:55:48 PM EST
So I've been playing around with the SUSA tracking polls for each Senator. It's a fun gauge, though kind of kooky. I don't notice any clear patterns with the 2006 elections. While Santorum and Burns were both loathed by their voters, only Santorum was blown out. Burns, with disapprovals in the 60s, nearly managed to eke out a victory. Talent was pretty well-liked until the end, when his disapprovals went up into the 50s and his approval crashed into the high 30s/low 40s. Even so, he nearly won. Mike DeWine by contrast had lukewarm approval ratings until September, when he became quite unpopular. He was blown out. Both George Allen and Lincoln Chafee stayed reasonably popular the whole time, except for a brief drop in late October (oops), and then bounced back. Chafee lost badly, Allen lost in a squeaker.
In terms of winners, Democrats didn't have any trouble holding on. Bill Nelson and Bob Menendez exhibited the same pattern of relatively high and stable unfavorable as Talent and DeWine, but without the electoral consequences. Jon Kyl's fav/unfav chart looks worse than George Allen's, but he had no trouble handling Pederson. Maria Cantwell had pretty high unfaves, but she had no problem staying easily aloft. Joe Lieberman was the most popular Senator who had a reelection campaign of any note, with a consistently high approval in the low 60s and low disapproval in the high 20s until the spring primary campaign. He dripped down until November, and could have been beaten, but his position in the electorate was a very strong one. It wasn't as strong as Robert Byrd, but it was strong nonetheless.
I don't have any interesting conclusions here. It looks to me like favorable/unfavorable ratings can give you a very loose sense of whether someone might be vulnerable, though I really have no clear idea of whether that's true and tend to be skeptical on the utility of this poll as a predictive tool of anything but trends. In 2006, if you were above 60 in favorables or below 30 in unfavorables a year out, you were in pretty good shape to be reelected. Otherwise it looks fairly unpredictable. You can be loathed as Burns was and nearly win reelection; you can be well-liked as Allen was and lose. You can be hated like Menendez and stomp all over your opponent, disliked like Kyl and win, only kind of sort of appreciated like Cantwell and crush your opponent, or held in neutral esteem like DeWine and be destroyed. It all depends on the state, the opponent, the dynamics of the environment, and the tenor of the campaign. Since we don't know the opponent, the environment, or the tenor of the campaign, it's way too early to know what to expect in 2008.
My recommendations are pretty obvious. If you're facing someone like Lieberman with high approvals and low disapprovals, start early. He bled a lot but not enough. If you're facing someone with disapprovals in the 40s, good for you. Unless it's New Jersey and you're a Republican, in which case you should give up now and save yourself a whole lot of trouble. Then again, even these recommendations are probably kind of dumb, since some states like the Dakotas tend to have Senators with very high approval ratings, but it's possible to 'like' a Senator and vote against him/her. Midwesterners like everything.
When I'm looking at a race, I assume that partisans will come home to roost, which usually bumps a challenger's numbers up substantially. And then I try to understand the main issues in a state, and give a structural advantage to the dominant party candidate (red candidates get extra points in red states, and vice versa), and throw in demographic changes (Northern Virginia, for instance). This year it was obviously a nationalized election, so the main issues in every state were the same - Iraq and Bush. In 2008 we could see a nationalized election yet again, much as 1978 and 1980 were both nationalized around reactionary themes. We could get our Reagan, though I doubt it very strongly. Hillary Clinton is not it; she had no coattails this cycle. This leaves Obama and Edwards, but neither one has a coherent sense of how to ask for sacrifice, so I imagine both will be easily steamrolled in either the primary or general. I suppose the Republican candidate could nationalize the election - McCain has that potential. Or maybe we'll get another 2000, another limp set of candidates and a Senate map of localized campaigns.
Who knows? I don't. I imgagine we have to concede that Iraq is not going away, and that candidates will have to take positions on the investigations that are obviously coming, as well as policy changes on torture, wiretapping, taxes, corporate profits, and war profiteering. If those work well in your state, then start up your engines.