“Redness,” Unemployment, and the Election

“Redder” is better, generally speaking. For many reasons, including economic health. Using Bush’s average margin of loss or victory in 2004 and 2008 as an index of “redness” (and disregarding the anomalous 2008 race), here is the relationship between unemployment and a State’s degree of “redness”:

Derived from Bureau of Labor Statistics, Unemployment Rates for States (preliminary September estimates, issued 10/19/12), and official tabulations of popular votes by State. The correlation, though not strong, is statistically significant (less than 1-percent probability of occurring by chance).

The “outlier” on the left is the District of Columbia. DC, despite its predominantly black population, does not have an exceedingly high unemployment rate because the federal government and its contractors are havens of patronage and reverse discrimination. In any event, the omission of DC would strengthen the correlation, and would yield a more pronounced negative relationship between “redness” and unemployment: y = -0.0386x + 7.6566; R² = 0.1434.

I have seen some “news” stories which suggest that lower unemployment in swing States will help Obama. Such speculation strikes me as wishful thinking by left-biased media. In fact, of the  four States that seem to have swung to Romney — Florida, Missouri, North Carolina, and Virginia — the first three experienced better-than-average improvements in unemployment from a year earlier. A possible reason for this apparent anomaly is that voters know that there has been little change in the real rate of unemployment. Further, they also know that unless Obama is kicked out, things will not get better very soon, if ever.