It’s not unusual to hear pundits or political observers say that national polls are less important than state polls, since we don’t elect presidents by popular vote. There is a little truth to that; you can build a fairly accurate picture of the presidential race by aggregating state-by-state polling. But in general, national polls are fairly accurate; on election day this year, the final results will almost certainly reflect the overall trend of nationwide polling.
Even still, this idea that we should look to state conditions rather than national trends has carried over to other realms of political prognostication. With regards to the economy, for example, Chris Cillizza argues that we should pay the most attention to economic growth in swing states:
But the state-by-state unemployment numbers are a reminder that the 2012 election — like all presidential contests — is a national election in name only. That is, although the U.S. unemployment rate matters as a broad thematic, the rates in the eight to 12 swing states may well be more telling indicators of whether Obama can sell voters on his plans for the economy. That handful of states is where the election will be decided; the unemployment rates in places such as California (10.7 percent in June) and North Dakota (2.9 percent) are, essentially, meaningless.
As with the line on national polling, this sounds like it should be correct. But research runs in the other direction. In general, the national economy drives how voters evaluate the incumbent. A 2008 study by Stephen Ansolabehere, Marc Meredith and Erik Snowberg found that, compared to state unemployment, the national unemployment rate had five times the impact on presidential approval.
Of course, there are important nuances. In a more recent study, which sought to understand individual perceptions of the overall economy, the researchers found that voter information about the larger picture is determined by the economic condition of people similar to themselves. If white men in Virginia feel that white men are doing poorly overall, then Virginia’s low unemployment rate will make less of a difference in how they vote. Even still, national economic conditions are the driving variable in this evaluation.
And national economic conditions make Obama the slight favorite. The economy is measurably better than where it was when he entered office, and is still growing, albeit slightly. What’s more, it’s unusual for an incumbent president to lose reelection with positive economic growth. Indeed, most models have Obama as a slight favorite, which reflects itself in current polls showing a small but consistent lead for President Obama. The economy isn’t great, but voters are still unwilling to place complete blame on Obama. If things get worse, however, that can quickly change.