No, it's not the unemployment rate.
OK, the debate's over. President Obama did a poor job, Mitt Romney did well, and facts and arguable misstatements got lost amid the unavoidable focus on performance skills.
Where does that leave us?
If you go to one of the academics who really studies elections, then, well, you'd best count to ten, maybe to 10,000, and look far beyond the debate to see how the election will play out. It is about the economy, stupid, and a few matters related to it.
"I form my expectations from five different structural factors that empirical research indicates are the main influences, election in and election out," said John Mark Hansen, a political scientist and former dean of social sciences at the University of Chicago. "They are the economy, condition of foreign affairs, the policy positions of the two nominees relative to the electorate, the baseline partisanship of the electorate, and incumbency."
What matters in foreign affairs, he said, is expensive shooting wars, like Vietnam and Korea, which means it's a minor influence this time.
He wonders whether Romney's embrace of more extreme positions in the primaries and his pick of Rep. Paul Ryan for running mate will offset the general impression that he's not as conservative as he says he is. Hansen has been assessing his chances as a relative moderate, especially as he tacks pretty actively back toward the center, including during the first debate.
A few demographic trends (rising Latino population, younger voters' partisanship) are working in the Democrats' favor, but not necessarily dramatically, at least not yet.
That leaves the two major structural conditions affecting this race: the economy and incumbency.
In 2012, as in 2004, he believes the incumbent has to be deemed the favorite, largely because the empirical literature indicates that incumbency is a huge advantage. But what's eating into that advantage and making this a close call -- even closer than in 2004 -- is the condition of the economy.
What matters politically in the economy, he argues, is the delta -- the change, not the level -- and what matters most of all is personal income growth.
"The recovery has been so unsteady, though, that it's difficult to make a forecast of income growth," he said. For the purposes of his elections course, he just looked up existing figures: "The most recent figures for per capita personal income growth (adjusted for inflation) are for the second quarter of 2012 (2012:II), which ended in June. If you look at the one-year change in personal income, 2011:II to 2012:II, it is 0.6 percent. If that's the reality, then it's really bad for Obama. Personal income growth in 2008 was 0.7 percent."
"But I don't think personal income growth for the election year (2011:III to 2012:III) is going to be that low. If you look at the annualized growth from 2011:III to 2012:II, it's 1.7 percent; if you compare 2011:IV to 2012:II, it's 2.7 percent; if you compare 2012:I to 2012:II, it's 2.3 percent."
So if one were to make a completely mechanical assessment of the likely outcome based on the historical relationship between the vote and personal income growth, the growth level that gets an incumbent party to 50 percent is 1.95 percent. In 2004, it was 2.7 percent, which is actually a bit below the postwar average for election years, he noted.
"That's why I'd like to know how 2012:III comes in. The higher the number, the better it is for Obama -- not, I want to emphasize, because of the number itself but because of what it represents: the lived experience of voters in this economy during this election year."
So amid the punditry and harrumphing about Wednesday's face-off between an aggressive Mitt Romney and a decidedly diffident President Obama, keep your eye -- or data sets -- on personal income growth.
Yeah, kind of boring, but perhaps predictive.
We want to hear what you think about this article. Submit a letter to the editor or write to email@example.com.