Rasmusen Weblog

Every four years I write my prediction for the presidential election. Usually I do so much sooner in the campaign but this year has been unbelievably close with polls conducted on the same day showing vastly different results.

Here is my track record: In 1992 I predicted Clinton over Bush 331-207; he won 370-168. In 1996, I predicted Clinton over Dole 312-226; he won 379-159. In 2000, I predicted Bush over Gore 290-248; he won 271-266. You will thus notice an overestimation of the Republican's chances. I don't claim not to be biased but that bodes ill for Bush.

A candidate needs 270 electoral votes to win.

Let's look at the states that are solidly for Bush: Alabama, Alaska, Arizona, Georgia, Idaho, Indiana, Kansas, Kentucky, Louisiana, Mississippi, Montana, Nebraska, North Carolina (Edwards no help here), North Dakota, Oklahoma, South Carolina, South Dakota, Tennessee, Texas, Utah, Virginia and Wyoming. That's 191.

Kerry's solid states are: California, Connecticut, Delaware, District of Columbia, Illinois, Maryland, Massachusetts, New York, Rhode Island, Vermont and Washington. That's 160.

States that I think are likely to go to Bush are: Arkansas (Bush leads in almost all polls, he won in 2000), Colorado (he leads in all polls, has gone Democrat only once since 1980 (1992) - and it doesn't seem the proposal to split the electoral votes is likely to pass), Missouri (he leads in all polls, he won in 2000, culturally conservative state), Nevada, and West Virginia. That's 36, which gives Bush 227.

States that I think are likely to go to Kerry are: Hawaii (usually an extremely Democrat state, Bush is doing unusually well at the polls, perhaps because Hawaii has a popular Republican (Jewish female) governor), Maine (without splitting the electoral vote, which can happen in Maine), Michigan (Kerry is ahead in most polls, union state with Democrat governor, hasn't voted Republican since 1988 - but could be close), New Jersey (you hear some talk how it is tighter than expected but I think Kerry will win big here), Oregon (Bush had high hopes here but isn't close), Pennsylvania (Kerry wins pretty much every poll, went for Gore in 2000). That's 68, which gives Kerry 228.

Now for the really close states:

New Mexico: The most Hispanic state in the nation, Clinton won it twice, Gore won by less than a thousand, but Bush ahead in most polls. But I am giving it to Kerry. Kerry now at 233.

Minnesota: Supposed to be real close, but hasn't voted Republican for President since Nixon (of course, Mondale was on the ballot twice). I think Minnesota will trend Republican (and it elected Norm Coleman over Mondale two years ago), but this year will more likely than not stay Democrat. Kerry now at 243.

Iowa: Kerry has done slightly better in the polls and Iowa is one of the most dovish states. It went for Dukakis in 1988 (Dukakis only won 10) and it went for Clinton over Dole in 1996, although you might have thought Mr. Ethanol could have taken that state). Therefore, I put Iowa in Kerry's camp. Kerry now at 250.

Florida: It will be close, but I think Bush will take the state. He has done slightly better in the polls; Jeb is popular and won handily 2 years ago, even though a lot of people thought the Republicans would pay for the 2000 fiasco. W got a lot of credit for timely assistance after the hurricanes. I think the Jews will be a little less supportive of the Democrat ticket this time (no Lieberman, Bush clearly the more pro-Israel candidate). The new generation of Cubans is less anti-communist and therefore more likely to support Democrats, but the Republican candidate for senator, Mel Martinez is Cuban, which might help Bush keep those younger Cubans. Bush now at 254.

New Hampshire: Bush did win this in 2000, but this state is in Kerry's backyard and Kerry leads in a majority of the polls. Therefore, I give it to Kerry. Kerry now at 254.

Wisconsin: Kerry is 8 points ahead in the latest Zogby poll although Bush was ahead in most polls a week ago. Wisconsin has seen movement toward the Republicans but I am not sure enough to give this state to Bush. Therefore I give it to Kerry. He is now at 264.

Ohio: Obviously, under my analysis, it is now down to Ohio which is really too close to call. The Realclearpolitics.com poll consensus is a dead tie. However, Bush has done better in the most recent polls (probably outdated by the time you read this). Bush won in 2000, not by much. The state has two Republican senators and a Republican governor and in recent years has only voted for Democrats when they are from the South (Carter, Clinton). When I am feeling optimistic, I give the state and thus the election to Bush; when I am feeling pessimistic, I give the state and thus the election to Kerry. I am going to give it to Bush, partly because I have given most of the really close states (Michigan, Pennsylvania, Wisconsin, New Hampshire, Iowa, New Mexico) to Kerry. If Bush loses Ohio, he can still win by winning either Michigan or Pennsylvania, or Wisconsin and Iowa or Wisconsin, New Hampshire and New Mexico. If Kerry loses here, he has lost (unless he wins Florida). A real possibility, however, is that Bush loses Ohio but wins Wisconsin and either New Mexico or New Hampshire, in which case we will have a 269-269 tie.

What happens in a tie: The correct question is what happens if no candidate gets a majority of electoral votes (cast by the electors - there is always the possibility of a "faithless elector", especially in a tie). In that case, the House of Representatives, voting by state, with each state getting one vote, decides among the top three vote-getters (thus, if there is one faithless elector who votes for McCain, for example, the House could choose him). How a state votes is dependent on how its representatives (the ones chosen in this election) vote. Right now the Republicans control a significant majority of the house delegations and we can assume that probably won't change. Thus, Bush would probably win if the House decided. The Senate would choose the Vice-President based on a majority vote of the new Senate - which could go either way. Theoretically, they could choose Edwards.

Obviously there are a lot of factors that I couldn't take into account (the effects of voter fraud (or intimidation), any event that occurs between now and election day, the inaccuracy of polls, weather, and most importantly, post-election litigation.