I don’t have a clue. Really. I’m not kidding or trying to be dramatic. I just don’t know. Believe me, I’d tell you if I thought I knew. [IMGCAP(1)]

When you are a political analyst or election handicapper, you want to make a pick. Refusing to do so always seems like a cop-out. I prefer to crawl out on a limb, even if I’m not 100 percent certain that I’m right.

But there is no reason to pick a horse just to pick one. The data, I’m afraid, don’t give either President Bush or Sen. John Kerry (D-Mass.) enough of an advantage to conclude that one or the other has an edge in Tuesday’s voting.

People who predict that Bush will win often argue that voters won’t change presidents in the middle of a war, and note, the conflict in Iraq and against terror makes it likely that the president will win a second term.

Maybe they are right, but the president’s overall performance in office has been sufficiently underwhelming that voters may see change as not only acceptable, but necessary.

Democrats cite 1980 to buttress their case that Kerry will win. Voters are looking for an alternative to Bush, they argue, and in the final days, swing voters and undecideds (who tend not to like Bush and think the country is headed off on the wrong track) will turn to Kerry, much as late-deciding voters opted overwhelmingly for Ronald Reagan in 1980 over Democrat Jimmy Carter — another incumbent president with weak poll numbers, a poor domestic economy and international problems.

Maybe they are right, but Reagan surged in the polls after the Oct. 28 presidential debate, when voters decided he was an acceptable alternative to Carter. The 2004 debates have been history for a couple of weeks, but Kerry still hasn’t closed the deal. If voters haven’t embraced him over the past two weeks, why should I assume they will do it in the next few days?

Here’s what I think: This election is going to boil down to Ohio and Florida. Most of the other states that we are all talking about are a mere sideshow.

Bush may well win Iowa, and Kerry has a good chance at taking New Hampshire. Wisconsin, Oregon, Pennsylvania, New Mexico and Minnesota might change their colors, but Democratic and Republican strategists who should know seem skeptical. They think this is a two-state presidential race.

(Like most of us, insiders too are confused by state polls in Hawaii, Colorado, Arkansas and Michigan showing a close race. Could both Hawaii and Ohio be tossups in the same election? It seems improbable, if not impossible.)

Anyway, if I’m right, two 2000 Bush states will determine the next president. And polling shows both states even.

Normally, “it’s all about turnout” is a safe harbor for political jellyfish who want to avoid looking bad. But, in this case, it looks about right. The makeup of this year’s electorate remains an important unknown.

The national polls are strikingly unhelpful, unless you know which poll to believe. Democracy Corps polls show Kerry ahead, but the survey is conducted by Democrats. Early in the cycle, its numbers matched up pretty well with events and with other polls. More recently, I’ve lost a little faith in the results.

Fox News shows Bush with a big lead, but Fox is, well, Fox. Its results may well be right on target, but a recent Bush 7-point advantage over Kerry seemed squirrely to me.

That leaves a slew of surveys to believe or not believe.

Newsweek, Pew, The Associated Press and Zogby often are at the low end of the Bush vote, but all but The Associated Press have suggested recently that the president has a statistically insignificant edge. CNN/USA Today/Gallup has been showing Bush clearly ahead, but it often seems to be an outlier. The Los Angeles Times has the race even.

That leaves ABC News/Washington Post (which has swung wildly in the past few days from a clear Bush edge to a tossup), CBS/New York Times (which generally has the president up by a couple of points) and NBC News/Wall Street Journal (which has the race roughly even, though Bush is ahead among registered voters).

Generally, if you look at all the polls, Bush appears to be ahead by a couple of points among likely voters. That’s not much though, given the expected high turnout.

Polling in Ohio suggests an even race. Ditto for Florida. Some people say that benefits Kerry, since Democratic registration and get-out-the-vote efforts have been so strong. But the GOP actually outregistered Democrats in the past four years in Florida, adding to my uncertainty.

No, I don’t know who the next president will be. The crystal ball is foggy for all but the most partisan. All I’m sure is that whoever wins will claim much more of a mandate than he really earns. The country is likely to remain divided.

Stuart Rothenberg is editor of the Rothenberg Political Report.