Recently, it has seemed that looking at polling on a state-by-state basis is far more favorable to Obama than looking at national polling. The question is, is this really true? And if so, why?
Yesterday, Steve Singiser set out to answer this question using recent polling data. The answer was yes, swing states have shifted less than national polls compared to the 2008 margins.
But the post left a few questions unanswered because of a lack of available polling data. Is it the red states where Obama is underperforming compared to 2008? Or blue states? Or both?
Fortunately, we have the Daily Kos/SEIU/PPP polling data to peruse (using one methodology, N=10,000 over the last few months for the Obama versus Romney question, covering all 50 states), and we can compare to the final 2008 polling average, not the election results (an apples-to-apples comparison - comparing polls to polls - which makes a big difference for states that are highly partisan).
So first, let's look at blue states (polled >55% Obama in 2008), red states (polled <45% Obama in 2008), and swing states, as well as the national results, for the 2012 DKos polls and the 2008 polling averages:
The DKos poll currently has Obama polling a few points worse than his final average against McCain in 2008. But the swing states (NH, PA, FL, NC, VA, CO, NM, NV, IA, MI, MN, OH, WI) are about the same as 2008. This confirms Steve's findings.
We also see that Obama is, indeed, doing slightly worse in blue states compared to 2008 polls, on average.
But surprisingly, Obama is not doing worse in red states, on average.
There's a little more to it if you dig a little deeper. Here's the graph by region:
The only one of these regions where Obama is doing significantly worse than in 2008 polling is the Northeast.
There is a potential explanation for this: this region has seen the least improvement in unemployment over the past year. Change in unemployment from April 2011 to April 2012 was -1.3 points in the South, -1.2 points in the Midwest, and -1.0 points in the West, but just -0.3 points in Northeast.
Another possible explanation would be that the Northeast is partial to Romney, except change in Obama's vote share from 2011 (vs Generic Republican) to 2012 (vs Romney) is closely related to change in Obama's approval, and the Northeast is no exception.
So, yes, the national polls paint an uglier picture for Obama than if you look at the state polls. And it's because of the Northeast.
_______
Beyond the Margin of Error is a series exploring problems in polling other than random error, which is what the margin of error measures.
Previously:
When Polls Fail, or Why Eilizabeth Warren Will Dash GOP Hopes. Why polls for close races for Governor and Senate are sometimes way off, and how to predict how far off they will be.
Presidential Polls Are Almost Always Right, Even When They're Wrong. How the presidential polls in red and blue states are off, sometimes way off, and how to predict how far off they'll be.