State and National Polling: Is There a Disparity?

State and National Polling: Is There a Disparity?

By Sean Trende - August 2, 2012

The RCP Average shows President Obama leading Mitt Romney by 3 points nationally. This margin is down considerably from his 7.2 percent win in 2008.

This has some people scratching their heads at what they see as a disparity in the state polls. Obama leads in almost all of the states he won in 2008, with the exceptions of Indiana and North Carolina (and the latter is very close).

Indeed, the president’s leads in Virginia and Florida are down only a few points from where they were four years ago, and his lead in Ohio is larger than it was then. The conclusion that some have drawn is that Obama’s national lead is understated.

To test this theory, I first took the state RCP Averages, modified to eliminate polling taken before the Republican primary ended in May. If there were no averages for a state, I took the most recent polls. If no recent polling was available, I averaged the 2008 and 2004 outcomes for the state (since Obama’s national lead is roughly between those of 2008 and 2004). I then weighted the data by 2008 turnout. This essentially reverse-engineers a national poll from the state polling.

The weighted state poll average gives Obama a 2.4 percent lead, almost identical to what we see in the national polls. In other words, there isn’t much of a national poll/state poll disparity. Even if we just use the 2008 results for states where we don’t have RCP Averages or recent polling, the president’s lead is 3.4 percent, and that is assuming 2008 turnout and results in 16 of these states.

I think there are two important things to keep in mind here. First, states do change relative to the country as a whole all the time. In other words, a state’s Partisan Voting Index is not always stable. From 1996 to 2004, West Virginia went from being a state that leaned Democrat by two-to-three points to one that leaned Republican by four-to-five points. In 1988, New Hampshire leaned Republican by almost 10 points, while Florida leaned Republican by eight points. Just four years later, New Hampshire’s lean was down to two points Republican, while Florida’s was four. By 1996, New Hampshire had a two-point Democratic lean.

It wouldn’t surprise me all that much if Ohio moved a few notches toward the Democrats, or if some of the other states (Michigan, perhaps) lurched toward Republicans this cycle. It’s a turbulent time in our politics, and in many ways stability should be more surprising than change.

I think there’s also something deeper, and it’s a possibility that I alluded to a few weeks ago. I don’t think there’s any evidence that Obama’s spending has moved surveys in his direction -- even those CBS/NYT/Quinnipiac polls from yesterday showed movement toward Obama only in Pennsylvania, relative to the next most recent Quinnipiac poll.

But I do think there’s a real possibility that Obama’s advertising blitz has stopped his slide in these states. Obama is running significantly behind his 2008 margin in states like Michigan, Iowa, Missouri, Wisconsin, Nevada and New Mexico. In fact, his slide in these states is roughly commensurate with his slide in the national polls from 2008 to 2012 (a few points more, actually). These are states where his campaign had not engaged with Romney. There are also a few blue states where he has weakened, such as Connecticut. This possibly reflects Romney’s standing among those in the finance industry.

But Obama is holding steady in Ohio, North Carolina, Florida and a few other states where he has really unloaded on Romney. In other words, his improvement is relative. He’s also holding steady in places like Arkansas, where he was probably reduced to the Democratic base in 2008 (and thus had no farther to fall).

Whether these states’ stability represents fundamental shifts in their electorates, responses to short-term stimuli from the Obama campaigns ads, or a visceral dislike of Romney remains to be seen. But the RCP state polling averages are, overall, consistent with the RCP national polling average. 

Sean Trende is Senior Elections Analyst for RealClearPolitics. He can be reached at

Latest On Twitter