Tuesday, October 23, 2012

The 2012 Presidential Election: My Prediction

There are exactly two weeks to go before the polls open in the 2012 presidential election.  The primaries are history, the conventions are long since over, the debates are done, and there's nothing left to do but go vote.

Presidential elections are not decided by a popular vote.  Instead, they are decided by an electoral college vote.  Each state gets a set number of electoral votes, based on its population.  In all but two states, electoral votes are given in a "winner-take-all" system; in other words, all the electoral votes in the state go to the candidate who wins the most popular votes in the state.  Maine and Nebraska apportion their electoral votes into districts, with the popular vote in each district determining who wins that district's electoral vote.

In this year's election, most of the states are already more or less decided.  It's not that those states have already voted, or that those states don't still have individual undecided voters, it's just that the polls in those states show one candidate leading by such a significant margin that the chances of that state going to the other person are slim to none.

These states, then, are basically already locked up for one candidate or the other.  There are 40 of them, and when you add up their electoral votes, you find that Romney currently has 206 probable electoral votes, and Obama has 201 (270 are needed to win.)

That leaves 10 states remaining, where the polls are still very close (within 5 percentage points) and which neither candidate has yet "locked up."  These are the states, then, that will ultimately decide the election.

Those states are: Colorado, Iowa, Ohio, Pennsylvania, New Hampshire, Michigan, Wisconsin, Virginia, Florida, and Nevada.

Let's look at each of those states individually:

Wisconsin: Obama currently has an average lead of 2.8 points.  That's an average of all state polls taken over the last few weeks.  However, Romney has not led in a single Wisconsin poll since mid-August, when he had a 1-point lead in two polls.  Prior to that, you have to go back to June before you find another Wisconsin poll Romney led in.  So despite the average lead being relatively small, Wisconsin has consistently been in Obama's corner throughout the election season.  Advantage: Obama.

Virginia: When you average all the Virginia polls over the last few weeks, the candidates are in a dead heat.  However, in all the polls taken since Romney's prominent victory in the first debate, Romney has led (albeit by small amounts).  Obama had a comfortable lead in Virginia throughout September, showing as much as an 8-point lead in two different polls during that month.  But clearly the momentum in Virginia is on Romney's side now.  Advantage: Unclear, but leaning Romney.

Florida: Romney currently has an average lead of 1.8 points.  More than a dozen polls have been taken in Florida this month alone, and Romney has led in all but three.  Like Virginia, Obama had a consistent lead throughout September, but Romney has the momentum there now.  Advantage: Romney.

Nevada: Obama has an average lead of 3 points.  Like Wisconsin, Romney's surge in the national polls has not been enough to overcome the comfortable margin Obama has enjoyed there throughout the campaign.  Not as many polls are taken in Nevada as some of the other states, but you have to go all the way back to April to find a Nevada poll won by Romney.  Advantage: Obama.

Colorado: Obama had a consistent lead in Colorado in September, but it has wilted in October.  This month, the candidates have gone back and forth, with Romney currently holding a fractional edge (0.2%) over the last three weeks.  Advantage: Unclear.

Iowa: Obama has tended to lead in Iowa throughout the campaign season, and currently has an average lead of 2 points in the polls of the past week.  As recently as last Wednesday a poll showed him with an 8-point lead.  But other polls have showed the candidates tied.  Even with Romney's post-debate surge, he has won only one of six polls taken since that time - and that was by just 1 point.  Advantage: Obama.

Michigan: Obama's current average lead in Michigan is 5 points, making Michigan just barely a toss-up state at this point.  Romney has not won any polls in Michigan since mid-August.  Advantage: Obama.

Ohio: Over the last ten days, Obama's average lead in Ohio polls is 1.9 points.  Since the first week of September, 31 polls have been taken in Ohio.  Obama has won, or been tied, in all but three of them.  The only three polls he lost were in the first few days after Romney's debate victory in early October, and Romney had only a 1-point lead in each of those three polls.  In the last week alone, Obama has had a lead as high as 5 points in one poll.  Advantage: Obama.

New Hampshire: In the last two weeks, Obama has an average lead of 1 point.  The last poll taken, however was just released today and it shows Obama with a 9-point lead.  However, as recently as just ten days ago, another poll had Romney leading by 4 points.  Obama has tended to lead in New Hampshire throughout the campaign season, but the polls have been unsteady since the debate.  Advantage: Unclear, but leaning Obama.

Pennsylvania: Like Michigan, Pennsylvania is only barely a toss-up at this point, with Obama leading on average by 4.8 points.  His lead in Pennsylvania was simply too large for Romney's post-debate surge to have completely erased.  Romney has not won a single poll in Pennsylvania since February, before the primaries were even finished.  Advantage: Obama.

So what does all this mean?  Well, it means that Obama appears to have weathered Romney's surge since the first debate, and seems on pace to win the election.  Of the ten states left whose electoral votes are still up in the air, the only one that seems very likely to go to Romney is Florida.  We saw that Virginia, Colorado, and New Hampshire are "unclear."  Let's give them all to Romney.  Even then, Obama still has 277 electoral votes and wins the election.  In fact, we could even throw in Nevada or Iowa, and Obama would still have 271!  So Romney could take 6 of the 10 remaining states, and it would still not be enough to give him the election.  And remember, of the 10 toss-up states, only one is leaning heavily towards Romney.

I see only one way for Romney to win at this point, and that would be if he could manage to steal Ohio. Ohio has polled almost exclusively in favor of Obama all season long, but it has become very close there over the last month.  Obama has maintained a lead, but only just barely.  If Romney won Ohio, plus Virginia and Florida, he would need to win only one more of the remaining seven states - even tiny New Hampshire would be enough to tip him over 270 in that case.  However, it is notable that even if Romney won the windfall states of Ohio, Virginia, and Florida, Obama could still win if he took the remaining 7 toss-ups.

My prediction: Romney will win Florida, Virginia, and Colorado, but lose the rest, giving Obama a 281-257 electoral college victory.  As an alternative, New Hampshire might go to Romney, and Virginia might go to Obama, which would make the final tally 290-248 in favor of Obama.

Romney's only hope lies in winning Ohio, Virginia, and Florida, plus one more of the toss-ups.  I think this is an extremely unlikely scenario, based on the polling evidence.    


Tim said...

This is a reasonable prediction. I live in Michigan, which is dominated by the Democratic leaning Detroit-Ann Arbor corridor in the winner take all situation. The rest of the state is very Republican. Thus, a occassional Rebublican Governor, and a constantRepublican State Legislature is the norm. The fact that Mitt is the son of a former governor doesn't even help him here. Also, the amount of support Obama has given the auto industry has pretty much locked up the state for him.
Ohio is a real possibility for Romney. I wonder, does one feel like their vote even counts if they don't live in FL, OH, or VA?

Trent N. said...

Solid analysis Scott and I agree with everything you said. The BIG question is: will the college kids and the minorities show up in record numbers like they did 4 years ago? That in my opinion will determine this election.

Trent N

Scott said...

Tim: Yeah, I think it's time for serious consideration of getting rid of the electoral college. Even without the importance of the Ohio's and Florida's, the fact is that when you vote for a candidate that ends up losing your state, your vote basically doesn't matter. I live in Kentucky, and Romney is going to win big here, so it really doesn't matter whether I vote at all. My vote won't ultimately count towards Obama's electoral tally.

If we elected the president on a popular vote, every vote would count, even if your vote is the minority in your particular state.

Thanks for the comment!

Scott said...

Trent: Yes, voter turnout could ultimately decide the election. I think one of the reasons we see some of these polls with outrageous numbers (like the NH poll yesterday with Obama leading by 9, when Romney led by 4 just a few days earlier) may result from polling people who aren't likely to vote. Obama's victory is definitely in the hands of voter turnout among the young and minorities.

Scott said...

One thing I wanted to add to this post is this: If Obama wins Ohio, Virginia, or Florida, it's pretty much game over. Turn off the TV and go to bed because he's going to win the election. Romney can't lose any of those states and still win, in my opinion. Bad news for Romney is that he's behind in two of those three states, and Obama's lead in OH is bigger than Romney's lead in FL.

Cassie said...

Unfortunately the excitement from four years ago in terms of record voter turnout is not happening this year. This is my opinion of course but I think a lot of people just want the whole process to be over. Two weeks from now should be very interesting. I'm not a big fan of the electoral college.

Scott said...

The Republican Party Platform this year explicitly condemns bills aimed at getting rid of the electoral college system.

I agree that voter turnout won't be as good this year, but I haven't seen anything to cause me to think that it will be worse for one party than the other.