[N]o early polling on the Republican side has ever found an eventual nominee in the single digits at this point. Indeed, the only time in the last several open races that the eventual nominee wasn’t in first place in early polling was in 2007 when the absurd, media-driven candidacy of Rudi Giuliani was still taken seriously. Republicans have no habit of nominating candidates with poor name recognition on their first attempt in the modern primary process. Put another way, if someone polling as poorly as Dukakis in 1987 had been running on the Republican side at any time in the last forty years, he would never have won the nomination.I'm very open to the idea that the parties differ either in process or, in some relevant way, make-up, but I'm going to continue strongly disagree with this one. The overwhelming fact of previous Republican nomination cycles has been that there was an obvious nominee; that is, there was a clear party leader. That was the case in 1972, 1976, 1984, 1992, and 2004, when they renominated a president; it was also the case in 1988, when they nominated a sitting vice-president. That leaves only 1980, 1996, 2000, and 2008. Four cases. In my view, at least, that's not enough to draw any conclusions.
On top of that, each of the other four cases had a stronger leading candidate at this point than the 2012 field has. In 1980 Ronald Reagan was clearly the leader of the party, having run twice before, almost unseating a sitting president for renomination, and generally having been the biggest name in the dominant wing of the party for over a decade. In 1996, Bob Dole wasn't quite of that magnitude, but he was also a three-time candidate as well as a former VP nominee. Weaker still were the president's son in 2000 and guy who lost to the president's son eight years later, but still both were more obvious choices than Mitt Romney is this time around. That's not to say Romney can't win, but only to say that it wouldn't be surprising if something different happened this time -- because the situation is different.
Note too that when Democrats were in those situations, they renominated their president, or in 2000 the sitting vice-president, every time. Indeed, they also nominated the recent VP in 1984 (although not in 1972, but almost, and there are complications involved).
I can add one other thing from Silver's data. No one with single-digit polling numbers has won a GOP nomination, but it's not the case that Republicans won't vote for someone they had never heard of the previous year. George H.W. Bush went from 1% in the polls in 1979 to the clear second choice in 1980; John McCain was at 5% in 1999 and finished a solid second in 2000; Romney (9%) and Huckabee (2%) were nowhere in 2007 but serious contenders the next year. Now, it's true that none of them won, but I would say that Romney and Huck came fairly close -- and it wouldn't be shocking if the 2012 version of Romney turns out to be weaker than the 2008 version of McCain.
I'm not predicting a winner between Romney and Pawlenty, but I think that anyone who believes that the polling so far dooms Pawlenty is dead wrong. It's possible that the polling is telling us something, but that's about it. At any rate, it's certainly important to note that we're not extrapolating, as Larison would have it, from "the last forty years" -- we're extrapolating from exactly four cycles at best, and perhaps as few as one.