Political scientists John Sides and Lyn Vavreck have an interesting post up at YouGov that pertains to one of the arguments I made in my South Carolina analysis yesterday. (Sides and Vavreck are two of the best around when it comes to election analysis.) Drawing on national polling from the first week of January, they conclude that, “It simply is not the case that a vote for someone other than Romney is a vote against Romney. As the field narrows, a Romney nomination becomes more inevitable, not less.” The reason why, they argue, is that while it may be true that a plurality of likely Republican voters prefer someone else to Romney, he is the preferred second choice of most of those anti-Romney voters. As a result, he will gain support as the field is winnowed and those voters turn to their second choice.
I have not been paying attention to the national polling data as yet, so am not in a position to react to their findings. However, their data is consistent with some of the polling data I discussed in an earlier blog post regarding the state of the race in South Carolina. Drawing mostly on a Public Policy Polling (PPP) survey from Jan. 11-13, I noted that fully 58% of those surveyed wanted someone other than Romney to win in South Carolina. However, unlike what Sides and Vavreck found in their national polls, Gingrich leads as the top second-choice candidate of 20% of South Carolina respondents, although Mitt is tied for second with Santorum and “someone else/not sure” with 16-17%.
Nonetheless – and consistent with the Sides/Vavreck argument regarding national polling – in head-to-head matchups South Carolina voters chose Romney over every other candidate, with Gingrich and Santorum losing to him by 48-37 (15% not sure) and 48-39 (13% not sure), respectively.
I have been arguing that as the Republican conservative/Tea Party field winnows, support will begin to coalesce around the non-Mitt candidate, but at least some of the South Carolina data, consistent with the Sides/Vavreck national polling, suggests otherwise – it shows that in a two-person race, Romney would still win, at least in South Carolina – even though Gingrich leads among second-choice candidates.
So, how do we reconcile these apparently different claims – does Romney benefit if the field winnows or not? The key point to remember is that in polling data showing Romney gaining strength in a two-person race – the assumption is that it is a two-person race! That is, there’s no Ron Paul running along with a Tea Party/conservative favorite from among Gingrich, Perry or Santorum. That’s unlikely to happen anytime soon even if two of the three card-carrying Tea Party/conservative members do quit the race. Indeed, it’s likely that Ron Paul will carry the fight all the way to the convention. In short, my scenario is based on my belief that we are going to see a three-person race in the near future. The question then becomes: how does that affect the campaign dynamics? I can’t speak to national polling – it may be that Romney would still lead. But in South Carolina the PPP poll does provide some information that might help tease out an answer. (Keep in mind that this poll was conducted almost five days before last night’s debate, so the numbers are old.) The polling crosstabs indicates that Mitt does slightly better in a two-person race than in a three-person that includes Paul. Let’s play the polling numbers to see why (keeping in mind this is one poll in one state taken almost a week ago. But it will serve to illustrate my broader point.)
Let’s assume that Perry and Santorum drop out, and their South Carolina supporters move straight to their second choices. In this scenario Newt would pick up almost 9% more support, while Mitt would gain 7%, and Paul would gain 2.5% So, this is consistent with the scenario I laid out that argues that if two of the three conservatives drop out, the remaining one – Newt – benefits the most. How much likely varies from state to state (and it may not hold nationally – I don’t have any data as yet.)
However, the dynamics are different if Paul drops out as well, because a plurality of Paul’s supporters (at least in South Carolina!) support Mitt; he gets 38% of Paul’s voters compared to 28% who would in theory switch to Newt if Paul dropped out. If you combine them with the Huntsman voters (remember, he’s still in the race in this poll), you can see how in a straight matchup between Gingrich and Romney – with no Paul in the race – Romney still comes out ahead in South Carolina. Presumably these are the same dynamics that Sides/Vavreck are finding at the national level.
Again, keep in mind all the caveats regarding the data here. My larger point is that in assessing who benefits if the conservative/Tea Party three-some is winnowed to one, it likely matters if the analysis presumes Paul is still in the race as well (although I readily confess I haven’t done any national-level assessment of this presumption as yet). Once we account for the second place preferences of all the candidates – including Paul and Huntsman – one can understand why a plurality of South Carolina voters oppose Mitt, and yet he wins in a hypothetical one-on-one matchup with any other candidate in that state.
Of course, that hypothetical two-person race is not going to happen on Saturday in South Carolina. And if Paul stays in the race – it may not happen for many weeks, thus rendering the two-person matchup question moot – at least for now. And, if Mitt does begin losing to the anti-Mitt in a three-person race, who knows what impact that will have? I don’t.