I don't know who the Republicans are going to nominate for president in 2012. I do, however, want to take another shot at knocking down one piece of conventional wisdom as it applies to the 2012 contest: that Republicans habitually nominate whoever is next in line, which would presumably mean that either Mitt Romney or Sarah Palin will be nominated. It just isn't true, or at least it isn't true in any sense that helps us understand this election cycle.
Let's go to the history.
First of all, there's a clear dividing line when it comes to the nomination process; the McGovern-Fraser reforms (which took place on the Democratic side, but affected Republicans as well, as did campaign finance rules changes around the same time) are a clean break; we're only interested in the modern era, which began in 1972.
Second, we can ignore all those cases in which a sitting president was renominated. For the GOP, that eliminates 1972, 1976, 1984, 1992. and 2004. I'd recommend also eliminating those cases in which a sitting VP was nominated to succeed a retiring president, further knocking out 1988. It's true that the latter case(s) could be competitive, and the former have been on occasion, but I think it's fair to say they don't really tell us much about 2012.
So, for the Republicans, what does that leave? 1980, 1996, 2000, and 2008. That's it.
In 1980, the Republicans nominated a third-time candidate who had nearly defeated a sitting president four years back, and who had been the clear leader of the dominant conservative wing of the party for about fifteen years (oh, and he was also a two-term governor of the nation's largest state). In 1996, the Republicans nominated a third-time candidate who had been the clear runner-up to a sitting VP in 1988, on the national ballot as the VP nominee in 1976, and, by the way, was also GOP leader in the Senate.
So it's no surprise that Republicans nominated Ronald Reagan in 1980 and Bob Dole in 1996, and it tells us virtually nothing about the 2012 field, because there really is no living Republican who can come close to the resume that Reagan, or even Dole, brought to the show.
The problem is that this leaves only two cycles in which Republicans had no obvious nominee. In 2008, after quite a bit of uncertainty, they eventually nominated a previous candidate, the runner-up from 2000 who had no other obvious objective credentials. In 2000, they had a lot less trouble settling on a first-time candidate who was governor of a big state and the son of a GOP president.
At best, then, Republicans are one for two in nominating the "next in line" when the next in line isn't obvious. In 2000, those with a claim at being next in line were former VP Dan Quayle, former VP nominee Jack Kemp, and 1996 candidates Steve Forbes, Lamar Alexander, and Pat Buchanan. First-timer George W. Bush clobbered those from that group that tried. In 2008, the same group, essentially, was still available, along with Dick Cheney. None of them, for various reasons, wound up making the race, leaving it wide open, and there it remained until it was finally settled well into the primary season.
So the 2012 cycle will be only the third truly open nomination without a serious heavyweight in modern GOP history. Now, it may well be that Republicans are simply inclined to follow hierarchies (although don't ask Lisa Murkowski about that!), and that they saw Bush in 2000 and McCain in 2008 as more "next in line" than the others. But that doesn't really tell us that Romney or Palin will (necessarily) be helped by that tendency; it implies only that whoever does benefit will seem, after the fact, to have been the logical next in line.
For more, see some good posts from last year by Ed Kilgore and Josh Putnam. Kilgore's post overlaps with what I said above, although my accounting is a bit different; Putnam critiques Kilgore, and says it's frontrunners who win. Both are correct! What I'd say is that sometimes it's fairly obvious who the frontrunner is (or will be), and sometimes it isn't; when it isn't, it seems to me that the way to find the frontrunner is to look at objective factors such as endorsements, money raised, and polling, rather than trying to puzzle out who is supposedly entitled to it. But not just polling, especially not this early; as I said the other day, if Republican leaders rally around a candidate, name recognition and thus good polling numbers will rapidly follow.