Mean Reversion Toward Stable Voting Intentions in Poll Aggregates...
Must-Read: Kevin Drum: Chart of the Day: Voting Intentions Are Probably Set In Stone By Now: "This chart is a follow-up to my post last night about response rates in polls...
...from the paper that started the whole thing, published earlier this year by Andrew Gelman and three other researchers.... The 2012 campaign.... The red line is what the polls looked like in real time. The black line is what they look like when you control for different response rates.... During the last six weeks of the campaign, Obama's support was never more than a couple of points away from 52 percent.... In reality, there's probably no more than two or three points of change in actual voting intentions during the last month of the campaign. And in the last week? Practically none at all.
Andrew Gelman et al.: The Mythical Swing Voter: "Most surveys conducted during the 2012 U.S. presidential campaign showed large swings in support...
...Using a combination of traditional cross-sectional surveys, a unique panel survey (in terms of scale, frequency, and source), and a high response rate panel, we find that daily sample composition varied more in response to campaign events than did vote intentions.... Demographic post-stratification, similar to that used in most academic and media polls, is inadequate, but the addition of attitudinal variables (party identification, ideological self-placement, and past vote) appears to make selection ignorable in our data. We conclude that vote swings in 2012 were mostly sample artifacts and that real swings were quite small. While this account is at odds with most contemporaneous analyses, it better corresponds with our understanding of partisan polarization in modern American politics...