The Washington Times

At the risk of being insufferable, I said George W. Bush would win “just over 50 percent of the popular vote to John Kerry’s 48 percent.” As of late Sunday’s tallies, Mr. Bush actually finished with 51.0 percent of the popular vote and Mr. Kerry with 48.0 percent. I thus underestimated Mr. Bush’s majority victory by a few tenths of a percentage point. Sorry about that.

I also said Mr. Bush would win 286 electoral votes, including Florida’s, Ohio’s, New Mexico’s and Iowa’s. So he did.

I called a net gain of one GOP seat in the Senate, but not until a runoff in Louisiana. That was a bum prediction. The GOP picked up four seats (and won in the first round in Louisiana). My error here was in seeing this election as one in which Mr. Bush was essentially running for his own political life, not to benefit those down the ticket. I had in mind a contrast with Mr. Bush’s successful efforts on behalf of GOP Senate candidates in 2002. Since I thought Mr. Bush would win narrowly, I figured the GOP might accordingly come off the tiniest bit ahead in the Senate.

I simply did not understand in advance the way large Bush majorities in red states would help carry Republicans into the Senate. What’s striking in the big Bush states is, in general, how much larger GOP Senate victory margins were than the final polling suggested. The state polls are sketchier than national polls, for obvious reasons. But there was a clear pattern to the outcome. Karl Rove’s architecture for the Bush victory served his party in the Senate, too.

As for the House, where I predicted no change and the GOP gained four seats, I suffered a lapse of judgment over the effect of GOP-imposed redistricting in Texas. Republican gains there were predictable, and I should have predicted them, rather than concluding on the basis of nothing that the narrowness of the election elsewhere would produce a general offset.

But note well: The main thing that mitigates in favor of change in the composition of the House is exogenous shock, such as redistricting. Incumbents’ seats are safe, and when seats are open due to the retirement of the incumbent, they are most likely to remain in the hands of the party that held them. We have achieved, if that’s the word, Politburo-like levels of retention in the House. There is an elaborate dance going on here, one in which the majority GOP seeks to buy the acquiescence of Democrats in their shrinking minority status by offering Democratic incumbents easy districts to run in.

So much for my predictions. Let us now take a moment to appreciate the progress of our knowledge about politics that has been possible thanks to the results of the 2004 election. Most of it takes the form of our ability to reject as dubious certain contentions about the 2004 election that were propounded with certitude, especially, I’m sorry to say, by Democrats, many of whom are suffering the double misery of losing and of being all wrong in their expectations.

I know something about this pain from the days of the “Republican Revolution” a decade ago. One’s political preferences tend to subvert one’s analytical judgment, and the political world makes this easier by offering up innumerable truisms that are actually little more than broad generalizations based on very little data. Here are some that have bitten the dust in 2004, and good riddance.

Presidents who took office having lost the popular vote don’t get reelected. Kind of depends on the president and the circumstances, no? If an incumbent’s job approval rating is hovering just under 50 percent, it’s because Americans have decided to fire him. No, that’s not determinative, though it’s probably fair to say that such an incumbent will have his work cut out for him.

The undecided vote breaks against the incumbent. The idea is that if people haven’t made their peace with the devil they know before they get to the voting booth, they will risk the challenger. Not bad, unless the ambivalence about the incumbent is matched by ambivalence about the challenger. And in any case, the undecided vote, in breaking one way or the other, can hardly be expected to do so unanimously, or even overwhelmingly.

High turnout favors Democrats. Not if Republicans are also turning out voters in high numbers. One 2004 lesson is that the days in which the GOP hopes for rain on election day are over.

If the Redskins lose at home in the last game before the election, the incumbent White House party is out. Unless you’re expecting the incumbent party to get tossed out almost all the time, you’re going to need a Redskins team that wins at home. Maybe in 2008.