The old college ties

By Mustang Bobby 

The election is five weeks from yesterday. Nate Silver at FiveThirtyEight has run a series of state electoral college scenarios wherein the vote ends up in a tie:

As of Monday's FiveThirtyEight forecast, there were 21 states that Barack Obama was projected to have at least an 85 percent chance of winning on Nov. 6. The list includes three important states, Ohio, Wisconsin and New Hampshire, where Mr. Obama's polling has improved by an especially clear margin since the Democratic convention. It did not include several others, however, where he is favored, but less definitively so, such as Virginia, Iowa, Nevada, Colorado and Florida.

If Mr. Obama's overall standing holds in its current position, he should have no trouble winning some of those states, perhaps along with others like North Carolina. But suppose there is a deterioration in his polls between now and Nov. 6 — or that the polls have overestimated his standing across the board. And so Mr. Obama wins the states where he has at least an 85 percent chance of victory in the forecast, but no others. Then we'd be left with the following map:

If you add up the electoral votes in that case, they come out to Obama 269, Romney 269: an exact Electoral College tie. The election would then be thrown to the House of Representatives, which would cast votes based on the provisions of the 12th Amendment.

Mitt Romney would probably win such an election, because Republicans will probably control a majority of state delegations in the incoming House of Representatives. (The forecast model draws lots and awards the election to Mr. Romney 80 percent of the time in such cases.)

But it would surely come with tremendous controversy, especially if the 269-269 tie came in an election in which Mr. Obama had won the popular vote.

Fortunately, such an outcome remains quite unlikely. Of the 25,001 simulations that we ran on Monday, a 269-269 tie came up in 152 model runs, or about 0.6 percent of the time.

That seems like the worst-case scenario, but I think we all need this kind of reminder just in case folks start putting the bubbly on ice and start looking for hotels in D.C. for the inauguration. (I'm talking to you, Romney folks.)

Frankly, I'd rather be looking at a narrow win forecast now so that people on the Democratic team don't get complacent and cocky; that's how Dewey lost in 1948 (not that I was there or anything).  And there are other signs that some states are beginning to become lost causes for Romney, including New Hampshire, which is as close to being his home state as any, and I'm even picking up rumors that the Romney campaign is pulling ads from Ohio.

The chances for anything other than an Obama win are slipping away, but it ain't over yet. Five weeks is a long time, but I think we're beginning to see the first little signs of the end game.

(Cross-posted at Bark Bark Woof Woof.)