Democrats will nominate for president the candidate they like the best. Republicans grumpily will settle upon the one they dislike the least. That's why I think journalists who wrote Mitt Romney's political obituary after the New Hampshire primary may be as wildly premature as they were when they wrote Hillary Clinton's before it.
The conventional wisdom is that if he doesn't win Tuesday in Michigan, where Dad was governor, then Mitt Romney is dead politically. Mitt seems to think so, because he's cut back on advertising for the South Carolina primary a week later. But Mr. Romney could be resurrected, as John McCain has been.
Gov. Romney hasn't caught on with most Republicans because they suspect his recent conversion to social conservatism is more a matter of convenience than conviction, and because some evangelicals are concerned about his Mormon faith.
But if Mr. Romney loses in Michigan, he'll lose either to Mr. McCain or to former Arkansas Gov. Mike Huckabee. The concerns conservative Republicans have about those two are greater than the qualms they have about Mr. Romney.
If Mr. McCain wins in Michigan, he'll be dubbed the front runner. But that will be meaningless if he loses the next week in South Carolina, where Mr. Huckabee has been leading in most polls.
Journalists put too much emphasis on momentum, which has been absent in the first two contests. The winners in Iowa were losers in New Hampshire. This is mostly because a lot of voters are deciding late, and don't care very much what voters in other states have done.
Journalists who predicted a landslide for Sen. Obama in New Hampshire look to Democratic consultant Jerry Skurnick's "two electorates" theory to explain why they were so wrong. Voters who don't follow politics are much less informed than they used to be, Mr. Skurnick said, so polls can change dramatically when they do inform themselves, which is usually shortly before they cast their ballots.
If Mr. Skurnick is right, only fools will try to predict the outcome in Michigan. This is especially so because Democrats and independents can vote in the Republican primary if they choose, and there is no meaningful Democratic contest. Mr. McCain won in 2000 with crossover votes. A supporter of Sen. Obama, state Rep. Lamar Lemmons of Detroit, is urging Democrats to cross over to vote for Mr. Huckabee.
If Mr. Romney prevails in Michigan, he'll have as much right to claim front runner status as anyone. But even if he loses, he can claim a victory of sorts if exit polls indicate he won a plurality among Republicans. In most future contests, only Republicans will be allowed to vote in the Republican primary.
If Mr. McCain loses in Michigan and South Carolina, the Comeback Kid's comeback will be short-lived. Paradoxically, if Mr. Huckabee wins both contests, it may hasten his doom.
It's fairly easy to estimate the vote total for Mr. Huckabee: Take the number of evangelical Christians who typically vote in Republican primaries. Add 10 percent to it, because Mr. Huckabee attracts to the polls some people who don't normally vote in primaries. Then divide by two, because a lot of evangelicals care as much about economic and national security issues, on which they find Mr. Huckabee less than persuasive, as they do about social issues. The 46 percent of the vote among evangelicals Mr. Huckabee got in Iowa probably is his high-water mark.
Still, in a crowded field this solid base would be enough to keep Mr. Huckabee at or near the top. But it won't be enough once the field narrows, and Republicans leery of Mr. Huckabee will rally around the last man standing to oppose him.
Who could be Rudy Giuliani. If Michigan and South Carolina split, or both are won by Mr. Huckabee, the Florida primary Jan. 29 will be critical, and the former New York mayor is still leading in polls there.
Out of sight, out of mind might end up being a good early strategy for Mr. Giuliani. Once conservatives are reminded of what they don't like about Mr. McCain and Mr. Huckabee, Rudy could look better to them.
The Democratic contest will be settled Feb. 5, because their race has effectively narrowed to two candidates. But if more than two GOP candidates are viable after Florida, the mega-primary likely will produce mixed results, because no candidate has enough money to compete in all the primaries that day, so each will cherry pick. This means that for the first time since 1976, the GOP candidate could be chosen at the national convention.
In a brokered convention, the candidate who's disliked least has the best chance. I'm not sure who that is.