By Mark Mellman - 09/07/11 12:06 AM EDT
Jobs are the issue of the moment in Washington, and properly so. President Obama is delivering a major address on the subject, as jobs and the economy dominate voters’ ranking of the nation’s most important problems and, according to Gallup, near-record numbers worry about being laid off themselves. Our own polling for the Alliance for American Manufacturing found more than twice as many voters saying they wanted Congress and the president to focus on creating jobs rather than on reducing the deficit.
But while jobs rightly command our attention, those who divine the president’s reelection prospects by consulting the unemployment rate commit a serious error. While the economy plays a crucial role in presidential elections, not all economic indicators are created equal and unemployment is one of the less politically telling measures.
The 7.2 percent used in the formula was recorded at the time Ronald Reagan was reelected. Had someone wanted to make a case against his reelection then, they could have rightly said that no president since Roosevelt had won reelection with unemployment over 5.3 percent, and it would have been equally meaningless. And if one had wanted to argue that Nixon was doomed in 1972, one could have accurately argued that no president since Roosevelt had been reelected with unemployment higher than 4.8 percent.
“No one has ever” statements are analytically dubious on several grounds, including in this instance, the small number of cases. We haven’t had many presidents running for reelection since Franklin Roosevelt (9); we haven’t even had that many presidential elections since Franklin Roosevelt (16). Generalizing from a small number of cases is always dangerous. As illustrated above, simply adding to the number of cases is likely to cause records to fall.
Moreover, the unemployment rate fails to differentiate big winners from narrow ones. When Reagan was reelected with the highest unemployment since Roosevelt, he won 49 states and by the biggest margin since, well, Roosevelt. By the formula, he should have barely hung on. Go further. Do the thought experiment — if unemployment had been, say, 7.5 percent instead of 7.2, would Reagan have been reelected? Perhaps with something less than a 23-point margin, but it’s hard to imagine even another 1 or 2 percent in unemployment eliminating all of his lead.
It is particularly difficult to imagine because presidents presiding over much lower levels of unemployment fared worse, not better, than Reagan. Truman defeated Dewey by just over four points when unemployment stood at just about half of what it was under Reagan.
Conversely, Carter lost to Reagan by 10 points when unemployment was just three-tenths of a point higher than it was as Reagan waltzed to reelection under the “Morning in America” banner.
In fact, a graph plotting unemployment rates against victory margins would look like a Jackson Pollock painting — with splotches of color randomly distributed around the chart, clearly demonstrating the absence of a consistent relationship.
Sophisticated statistical models confirmed the finding — neither unemployment rates, nor changes in those rates, are correlated with votes.
None of this is to suggest that economic factors do not exert a powerful influence on politics — they do. Unemployment is just not the relevant measure.
The less discussed, and more esoteric, change in real disposable income is the more important statistic. Not that the country is faring particularly well on that measure either. But the latest forecasts predict that real disposable income will be growing by 1.5 to 2 percent, come 2012 — and that’s a level of growth consistent with reelection.
Mellman is president of The Mellman Group and has worked for Democratic candidates and causes since 1982. Current clients include the Majority Leader of the Senate and the Democratic Whip in the House.