In emails with RCP this week, Bartels explained that statistical estimates drawn from his and other research, when applied to the pace of real disposable income growth (that is, income growth adjusted for price changes), can produce predictive correlations that suggest outcomes a year from now.
The distillation of a rollicking presidential contest to a mathematical calculation that forecasts Obama’s electoral fortunes against an as-yet-unknown challenger — as measured by the popular vote — seems, in a word, provocative.
“Obama’s expected popular vote margin would be 5.17+3.49 x (2012 income growth),” wrote Bartels, the co-director of Vanderbilt’s Center for the Study of Democratic Institutions who holds the school’s Shayne Chair in public policy and social science. “This implies that he is likely to be re-elected even if real incomes are stagnant in 2012, and even more likely to win if there is some real income growth in the next 12 months.”
Bartels, along with some of his co-authors and political science colleagues, points to a collection of findings from years of empirical research: First, voters choose presidents with the economy in mind, period. Second, the economic indicator that carries predictive relevancy, according to studies of past elections (even during the economically painful 1930s) is not growth or employment, but personal incomes. Third, voters do not choose presidents while contemplating the state of the world over a span of history or as they imagine it to be in the future. With short memories and shorter attention spans, they myopically make decisions about the incumbent political party based on their personal pocketbooks in the four to six months before they vote. Fourth, incumbent presidents possess limited powers to lift personal incomes — the influential indicator — during election years. And fifth, they frequently try.