Is Romney gaining strength in South Carolina as this Huffington Post headline proclaims? Two recent polls indicate that he is.
But both are likely wrong.
The first poll is an interactive voice response automated poll by New Frontier Strategy (NFS) that was conducted January 11-12. It has Romney in a comfortable lead over Newt Gingrich, 31.7% to 23%. Perhaps the most interesting result, however, is that Ron Paul is a distant fourth, with only 9% support. (The margin of error is +/- 3.44%).
Note that Romney’s margin over Gingrich, and his overall level of support, is larger than recent polls in South Carolina have indicated, and Paul’s support is much lower. The reason, I think, becomes apparent when we look at the NFS’s sample’s demographics. Most noticeably, less than five percent of their sample includes individuals 40 years old or younger – precisely the age group from which Paul draws most of his support. At the same time, those age 60 or older constitute a whopping 55% of the sample! Note that Romney has done quite well among older voters. For comparison purposes, exit polls from the South Carolina Republican primary in 2008 indicate only 35% of voters were 60 or older, and 33% were under 45. Based on this, I suspect this poll is heavily under sampling Paul voters, and oversampling Romney’s.
The second poll – and the one the Huffington Post article discusses under the headline “South Carolina Primary: Mitt Romney Opens Up Big Lead In The Palmetto State” – has Romney comfortably ahead with 37% of the vote, with Paul and Santorum tied for second at 16%. In contrast to the NFS poll, this one has Gingrich way down at 12%. Once again, however, there is good reason not to take this poll seriously, despite the Huffington Post’s headline: the online poll included 995 South Carolina registered voters, of which 398 were Republicans and 380 Democrats. The comparable totals from 2008, based on exit polls, were 80% Republicans and 2% Democrats. Even though South Carolina is an open primary, it is highly unlikely that Democrats are going to turn out in numbers approaching 40% of the electorate – a proportion that would almost equal the Republican level of participation. Once again, I suspect this poll is overstating Romney’s support and likely understating Gingrich’s.
Bottom line? Tim Tebow’s chances of winning today are better than the likelihood that these polls are accurate.