Most of the political prognosticators are focusing on whether Obama can be re-elected if the unemployment rate is still near 9 percent by election day next year, but most of the fancy quantitative political science models suggest that this is the wrong variable (or perhaps the dependent variable if you are into multiple regressions). Most of the models find that the most important economic factor is income growth. If incomes are falling, or growing very slowly, incumbents or the incumbent party usually loses. I’ll put in my caveat here that I’m a skeptic of these kind of quantitative political science...