Tuesday

As a congressional historian, I follow races for Congress, especially for the Senate, with great interest. Excluding Texas, which is in flux because of redistricting, there's a very good chance that more Senate seats could switch parties than House seats, something that (I believe) has never before occurred in American history. I continue to think that Melissa Bean will oust Phil Crane in Illinois, and wouldn't be surprised to see Connecticut's Rob Simmons become the only other Republican incumbent to lose, although Heather Wilson in NM and Max Burns in GA also could fall. Potentially strong Dem challengers in Arizona, Colorado, and Nevada have seen their candidacies fizzle. Meanwhile, right now no Democratic House member is trailing in polls, although Utah's Jim Matheson and Oregon's David Wu have been slipping noticeably, the latter after revelations that he had been punished for a sexual assault while at Stanford nearly three decades ago.

The race for control of the Senate is potentially more interesting. Right now, six seats seem likely to change hands: Illinois, Colorado, and Alaska from Republican to Democratic; Georgia, SC, and Louisiana from Democratic to Republican. Of these six, the only one that is a missed opportunity is LA, where the Democrats seem to have been overconfident that they could prevail in a runoff and allowed Republican David Vitter to amass too large a lead. (It now appears as if Vitter might win on Tuesday without even needing the runoff.)

Beyond this list, only four seats--Florida, North Carolina, South Dakota, and Kentucky--seem like possibilities to change hands. Potential Republican efforts in Wisconsin and Washington have not gained steam, and Oklahoma appears likely to elect (arguably) the most conservative Senate delegation since the institution of direct election for senators by choosing Republican Tom Coburn to join Jim Inhofe. Three historical patterns seem relevant to predictions on these seats: (1) that generally close contests tend to break toward one party; (2) strong Dem candidates in the South generally can run at least 5 points ahead of their national ticket; and (3) there's always at least one Senate upset. (2) suggests that Dems Betty Castor in Florida and Erskine Bowles in NC should prevail; Kerry figures to get at least 45% in NC and, at worst, close to 50% in Florida. I didn't think the race in SD would be as close as it has been, but still find it hard to believe that the state will oust Tom Daschle, one of the most talented politicians of the last quarter-century. Daschle began his career, by the way, by capturing a House election by less than 200 votes, so he knows how to win close races. He also has some important endorsement: from the unified leadership of the state's Indian tribes, and from the state's largest paper, the Sioux Falls Argus-Leader, which enthusiastically endorsed Bush earlier in the campaign. Finally, in Kentucky, Jim Bunning has done just about everything he could to lose this race (his latest was claiming that the WTC attacks occurred on November 11th), and Democrat Dan Mongiardo seems to have all the momentum. He certainly is the target of this year's most vicious smear in a congressional race, as two prominent Kentucky Republicans publicly termed him"limp-wristed," a"switch-hitter" and not a"man."

If Mongiardo and Daschle both prevail, the resulting Senate would be split 50-50, casting all eyes on Rhode Island's Lincoln Chafee, who has already announced that he won't vote for Bush and would be a candidate to mimic Jim Jeffords and declare himself an independent.

I think the presidential election is too close to call, but if forced to choose, would lean toward Bush, for two below-the-surface reasons. The first involves the anti-gay backlash. The attacks against Mongiardo, who is straight, were no accident: Kentucky is one of the states with an anti-gay marriage constitutional amendment on the ballot, and the most recent poll shows the measure with the approval of 76% of the voters. Bush, obviously, will carry Kentucky in any case, but in one state, a surge in Christian right turnout associated with an anti-gay marriage amendment could make a major difference: Michigan, which polls have shown surprisingly close (Bush is actually ahead in the most recent Zogby poll), and a state that Kerry absolutely needs to prevail.

The second hidden issue is Ralph Nader. He's clearly not going to get much of a vote in 2004. But if--as appears likely, at least right now--Ohio and Florida split between Bush and Kerry, the race will be decided by Wisconsin, Minnesota, and Iowa, three states where Nader polled very well in 2000 and where, if he gets 2% in 2004, he could tilt the election to Bush. This gets at one of the stranger issues of this year's election for me--the trend of these three states toward the Republicans, which began in 2000 and has continued this year.