Ezra Klein asks:
…we’ve still got four months until the election. Four months in which we’re going to be covering the campaign also nonstop. So help me out here: What’s going to matter between now and then?
“Going to matter” is a high bar. I’ll frame my answer as “worth watching because some previous research suggests they could matter.” Here are three items.
Economic trends. There is pretty good evidence that economic trends during the year before the election matter more than earlier trends. Chris Achen and Larry Bartels have argued that the second and third quarters of the election year appear to matter most—at least if you focus on changes in personal income as the economic indicator. So changes in July, August, and September are important. Note, however, that this doesn’t mean we should hyperventilate about every monthly jobs report. Most voters aren’t following the numbers that closely. It’s more likely that a gradual accrual of good or bad economic news will filter down via news coverage.
Campaign advertising. As I’ve noted before, campaign ads can have an effect, even in presidential races. However, three caveats are important here, which speak to how one should follow the ads. First, the effect of ads seems to emerge when one side is outspending the other by a significant margin. How much of a margin is hard to say; let’s take 2-1 as a rough estimate, which corresponds to the apparently consequential imbalance in Bush and Gore ads in battleground states right before the 2000 election. I’m not sure either Romney or Obama will muster that kind of advantage, even with the independent spending taken into account. TBD.
Second, the effect of ads seems to dissipate quickly, even within a week (see point #3). So you may not need to think about the effects of ads for another 3+ months. In fact, let’s shout that: FOR ANOTHER 3+ MONTHS. This notion that you have to advertise early to “define” the candidate or the opposition is folklore. Maybe there is some truth to that, but the truly rigorous studies have not identified such an effect, but have identified rapid decay.
Third, whether any effect of ads actually affects the outcome is a real question. It may be that the net effect of ads only slightly widens the winner’s margin of victory, without actually making the difference between winning and losing.
The ground game. To paraphrase a comment I heard from Sasha Issenberg, the problem with horse-race journalism is that it’s watching the wrong part of the horse. It spends too much time pondering the (poorly understood) effects of messaging and not enough on the nitty-gritty activities of the campaigns on the ground. Most importantly, we know from social science that get-out-the-vote drives can really work. So if the election comes down to mobilizing supporters, the comments of Cory Booker on the Sunday morning shows will be far less important than whether Obama and Romney can use the accumulated data they have about voters to identify promising targets and then contact them in the most effective way.
Alas, this is the toughest part of the campaign to watch, since much of it is invisible except in occasional reporting like Issenberg’s or this from Alec MacGillis. But if the election is close, field operations could be more important than the ads.
Let me close with this admonition: if you want to follow races where the campaign really might matter, you should be watching the Senate, House, and Senate and state legislative races. (Why? See #1 and #2 here.) And big things are at stake in these races—whether Republicans can consolidate their 2010 gains, whether they can take the Senate, and so on. Political observers (and political scientists too) pay too much attention to presidential elections. Don’t forget that a lot of interesting action is elsewhere.