Nate Silver’s investigation of the relationship of the unemployment rate to presidential re-election prospects will probably become a staple of poly sci courses, with its rigorous analysis of the data and prudent conclusions. It might also be good for Journalism majors to chew on it, especially aspiring political writers, since it provides a good example of why the rag of record hired Silver — to give some data-driven heft to their reportage.
Credit Silver with doing a lot of good work and providing intelligent analysis, such as the following:

…Historically, the relationship between the unemployment rate and a president’s performance at the next election is complicated and tenuous…An article in today’s Times notes, for example, that “no American president since Franklin Delano Roosevelt has won a second term in office when the unemployment rate on Election Day topped 7.2 percent.” The 7.2 percent figure refers to Ronald Reagan, who resoundingly won a second term when the unemployment rate was at that number in November 1984.
This type of data may be of limited utility for predictive purposes, however. Reagan won re-election by 18 points in 1984, suggesting that he had quite a bit of slack. An unemployment rate of 7.5 percent would presumably have been good enough to win him another term, as might have one of 8.0 percent, 8.5 percent or even higher.

Silver then speculates that FDR’s experience may be relevant since he won re-election with serious double-digit unemployment (16.6 in ’36 and 14.6 in ’40), but it was headed downward.
Silver argues that “the rate of change over a president’s term — is probably the more worthwhile approach. But it too is not always reliable.” He cites the examples of Nixon, W and Ike, all of whom who got re-elected with rising joblessness, although Ford, Carter and Papa Bush got defeated. Silver crunches the numbers for the last century and demonstrates that there is no positive correlation between “the unemployment rate to the incumbent party’s performance in the popular vote” and only a “weak” correlation between “the change in the unemployment rate over the course of a presidential term” and incumbents’ reelection prospects.
Silver concedes that there may be a predictive formula relating joblessness and other economic statistics to reelection prospects that works, but he cites a host of complicating factors and notes,

Some political scientists prefer other economic indicators to the unemployment rate, and there is evidence that measures like growth in real disposable income do a better job of predicting election results. Here too, however, we ought to be cautious. There are literally thousands of plausible models that one might build, using different economic indicators measured in different ways and over different time periods, taken alone or in combination with one another, and applied to different subsets of elections that are deemed to be relevant.

In the comments following Silver’s post, various responders suggest factoring in underemployment, “the change in unemployment during the final 12 to 24 months of a presidency,” gas prices, The CPI etc. It’s possible that there may be some formula that does a credible job of predicting electoral outcomes. Perhaps another stat wizard, like Alan I. Abramowitz could ferret it out.
As Silver acknowledges, however, the common sense argument for reducing unemployment is strong enough, even without statistical verification. For Dems, job-creation must remain a critical priority, the daunting difficulties of doing so cited by Andrew Levison in his recent TDS Strategy Memo notwithstanding. As Levison shows, Keynesian job-creation remains a tough sell with a significant segment of the public. Same goes for encouraging the private sector to invest in jobs when consumer demand is limp. Fresh ideas to break these two logjams are urgently needed.