Election forecast

The Baehr Essentials

We are four years on from the near—tied election of 2000.  But at least in terms of Presidential politics, things have not changed very much.  It is, in fact, quite possible that every state will vote the same way in 2004 as it did in 2000.  If this occurs, Bush would win 278—260, a slightly larger margin than in 2000 (271—266), since Congressional redistricting has added seven net Electoral College votes to the 30 red states.

As of Sunday evening, I believe there are only three states that are likely to vote differently than they did in 2000, and every one of them could potentially go to the party that won that state in 2000.  New Hampshire's 4 Electoral College votes are tipping ever so slightly to Senator Kerry, though Bush has been closing fast in New Hampshire the last few days, necessitating a Kerry visit today. New Mexico's five Electoral College votes are leaning to President Bush, as are Iowa's seven, though it is Kerry who seems to have the late momentum in Iowa (as he does in Wisconsin with ten Electoral College votes). If only these three states switch sides, then Bush will win 286—252.  One other blue state that seems to be slightly in the Kerry corner may also turn red — Minnesota, with ten Electoral College votes. So, taking a flyer on Minnesota, my election prediction is Bush 296, Kerry 242.

I believe Bush will win because, of the 30 red states only three are in play on the last weekend — Florida (27), Ohio (20), and New Hampshire (4). In Ohio, Bush has the late momentum, and is ahead or tied in every recent poll.  A week ago, I thought Kerry had crept ahead in Ohio. In Florida, Bush is ahead in most polls, and I think the dynamics of the state are more favorable to him (rapid economic growth, and a popular Governor who happens to be his brother).  I also think Bush will do better among Jewish voters in Southeast Florida than he did in 2000.  He will need this because he will undoubtedly not do as well as he did last time among Muslim voters, and may not do as well with Cubans, either. Republican Senate candidate Mel Martinez may help Bush with non—Cuban Hispanics in Central Florida. 

If Kerry loses Ohio and Florida, it will not be for lack of effort. The Democrats and their 527 support groups have assembled a very tough ground game. They have registered many new voters, and their vote pulling operation is better than it has ever been. But Republicans showed in the 2002 Senate races that their final 72 hour effort has gotten much better too.

Bush could lose either Ohio or Florida. Neither state is locked up. A 2 or 3 point lead is not a safe lead in either case. A loss of Florida would be more costly. If Bush is at 286 with Florida and Ohio, and loses either state, Kerry wins. If Bush wins Minnesota or Wisconsin and is at 296 with Florida and Ohio, he could lose either state and still win. Losing Florida would produce a 269—269 tie, and Bush would win the Presidency in the House of Representatives. But the 296 base, of course, is not assured. Bush could lose Iowa for instance (I think his lead in New Mexico is safer). That is why the Bush campaign is working to win Hawaii's 4 Electoral College votes and New Hampshire's 4, as insurance for Iowa.

If I had to give percentage odds, I think that Bush has roughly a 2/3 chance to win, and Kerry roughly 1/3. The three possible outcomes are a very narrow Kerry win, a very narrow Bush win, or a somewhat more decisive Bush win. I think all three are equal possibilities. The best outcome for Kerry would be to win Florida, New Hampshire and Ohio, and hold all 20 Gore states. He would then win the Presidency 311—227. I think it unlikely he would win both Florida and Ohio, so I am doubtful of Kerry hitting 300 Electoral College votes even in a winning scenario.  I also do not think Kerry will hold all the Gore states.

If Bush were to win decisively — say by 5% in the popular vote, as a few national polls suggest (there are more polls at the moment that show the national race tied, or give a 1—2% lead for Bush), then he could get to over 300 Electoral College votes. If Bush wins by 5%, I think he takes Iowa, Wisconsin, Minnesota, New Mexico, New Hampshire, and possibly Michigan (17) and Pennsylvania (21) as well. A win by over 5% and Oregon (7) might be in play, as well as Maine's 2nd district.  I do not see Bush winning New Jersey (15), Maine (3 for the state and the first district), Washington (11) or Hawaii even with a 5% national win. Bush could be in 350 Electoral College vote territory with a 5% win. But I think the most likely outcome is the center of the continuum — a narrow Bush win.

Several Bush states from 2000, are likely to stay in his column, but not by huge margins. These include Missouri (11), Nevada (5), West Virginia (5), Virginia (13), Arkansas (6), Colorado (9), and Arizona (10). Just as Minnesota is becoming more GOP friendly, Virginia is becoming more competitive for the Democrats. I do not expect Bush blowouts in these states. 

The reason the Democrats will lose, I believe, is that they have too few targets, and as a result are dependent on winning Ohio or Florida, or potentially needing to win both. The GOP, on the other hand, has forced the Democrats to defend Michigan and Pennsylvania, and the three states in the upper Midwest (Iowa, Wisconsin, and Minnesota), and New Mexico. Oregon has been probably the biggest disappointment for the Bush team.

As for Congress, I have not followed individual House races closely. But the polls for some swing Districts suggests that Republicans will pick up a few seats due to redistricting in Texas, and may lose a few elsewhere. I expect that the Republicans will wind up with about 230, very close to where they are now.

The Senate looks better for the GOP. There are five open Democrat—held seats in the South.  The Republicans are poised to win all of them. South Carolina (DeMint) and Georgia (Isakson) are done.  Richard Burr has a few point lead in North Carolina and should win. The  Martinez Castor race in Florida is very close, but I think Martinez wins. David Vitter is very close to 50% in Louisiana which would mean he could win on Tuesday without a runoff. In a runoff, his chances are about even perhaps better, if Kerry were to win the Presidency .

I also think John Thune will narrowly defeat Tom Daschle in South Dakota, though I do not think Daschle is doomed. He also could win narrowly, but in a Presidential election year, the GOP tide in the state will help Thune, and the GOP's congressional candidate as well.   

The Democrats I think will win a few Republican—held seats in the West. In Alaska, former Governor Tony Knowles has held a consistent lead over Lisa Murkowski all year. I think he wins, though Murkowski has a shot, given a big Bush win in the state. I also think Democrat Ken Salazar is primed to beat Pete Coors in Colorado, though this is close, and could go either way (like Florida, or South Dakota). I think Tom Coburn will win the open GOP held seat in Oklahoma over Brad Carson.  And of course Barack Obama will win the Illinois seat overwhelmingly.  If the GOP has a red letter day, and wins Alaska and Colorado, they could pick up 5 net seats, taking them to 56. I think a more likely result is plus 3 to 54. A worst case would be remaining at 51, losing Florida, Louisiana, and South Dakota, which I think is unlikely.