Nate Silver has compiled a truly spectacular list of every economic variable that might possibly affect a presidential election and then ranked them by how effectively they actually predict presidential elections. (Since 1948, anyway.) The top ten are below, but click the link for the full list of 43 indicators and a bunch of explanations of what it all means.
The descriptors in the list are a little confusing, but as near as I can tell they’re almost all changes, not absolute levels. The exceptions are the various indexes (like the #1 indicator), unemployment, inflation, and a few others. But #6, for example, which is labeled “Real gross domestic product,” is actually the change in real GDP, which makes sense. It’s the growth rate that usually matters in these things.
The top indicators mostly aren’t too surprising. I wouldn’t have guessed that the ISM manufacturing index was so great, but change in payroll, change in unemployment, and change in GDP all make a lot of sense. This is one reason that I think President Obama has a good chance to win next year despite presiding over a lousy economy. It’s quite possible that GDP will be growing and that unemployment, though high, will be improving too. Combine that with the fact that (a) incumbents usually get reelected and (b) Republicans seem to have taken up permanent residence in crazy town, and he has a pretty good shot at winning even if unemployment is still over 8%.
