Did pollsters underestimate the strength of Rep. Ron Paul's New Hampshire support because they didn't include enough younger voters or independents in their samples?
Yes, argues Stefan Hankin, a Washington, D.C. based pollster in a piece on the Campaign and Elections website.
"The miss on the Paul numbers is a bit troubling. When looking at the polling firms that release their demographics, it is clear that younger voters were under sampled. In the last Suffolk Poll, under 35s were at 7.6 percent; PPP had under 30s at 10 percent. Moreover, both these firms were also off on the sampling of independent voters. PPP had independents at 37 percent and Suffolk had them at 40 percent."
Young voters are traditionally harder than their elders for pollsters to accurately gauge before an election. They often don't follow through on voting intentions to the same degree as seniors and more likely to make up their minds later than older voters.
Those factors might have played a role Tuesday in New Hampshire. Hankin makes the point that while the New Hampshire results were fairly predictable, South Carolina should be less so.
"Given Romney's wide lead in New Hampshire, the under sampling of independents and younger voters didn't throw the polls off wildly. But in South Carolina, where the candidates are packed much tighter, a similar missampling could be much more consequential."