But what if the economy picks up and, say, we have pretty good economic growth by the middle of next year? Don’t the election forecasting models say that if GDP growth or, even better, growth in per capita disposable income is strong, then Bush more or less has a lock on the election?
Just like Al Gore did in 2000, I guess, a very recent election that most of the economy-driven election forecasting models missed egregiously. And let’s not forget 1992, when such models also performed poorly, because, by the middle of 1992 the economy was, according to most indicators, picking up a good deal of steam. So George H.W. Bush should have won. But he didn’t.
There’s a lesson here. Sure the economy in 1992 was picking up and, perhaps reasonably, the forecasting models assumed this would be a big deal to voters–but, both then and now, these are simple models based on a very small number of elections that simply can’t capture unusual, but critical, aspects of an upcoming election. (DR strongly recommends Jay Greene’s critique of election forecasting models that was published back in 1993 in The American Prospect). In the 1992 case, the unusual aspect was the persistence of relatively high unemployment, even as growth picked up. That made voters cranky who were expecting better and they took it out on the incumbent.
What’s happening today? Well, growth may be picking up and it could be pretty good by the middle of next year. Like 1992. But most projections expect unemployment to remain high through next year, even with the growth pickup. Like 1992. And right now 48 percent of voters in the NBC/Wall Street Journal poll say that unemployment is the most important economic issue facing the country–far outdistancing any other particular economic issue. That’s also just like 1992, when 46 percent of Americans said unemployment was the most important economic issue.
So, throw those forecasting models in the circular file. And. with any luck, we’ll have an election result just like 1992, too.