Not a toss-up

The Public’s prediction

More Americans think Obama will win than Romney. Historically, the public is pretty good at this (via Gallup):

past election predictions.gif

What I found most interesting was the partisan breakdown for this year.  Basically, by large margins, partisans of both parties think their guy will win, but Democrats are nonetheless substantially more confident (86%) than Republicans (71%):

who will win the election, by party.gif

Understanding the likelihood of an Obama victory

Andrew Gelman explains how to understand the probability of an Obama win.  This paragraph is the key:

What I’m saying is that I can simultaneously (a) accept that Obama has a 72 percent chance of winning and (b) say the election is too close to call. What if the weatherman told you there was a 30 percent chance of rain — would you be shocked if it rained that day? No. To put it another way, suppose Mitt Romney pulls out 51 percent of the popular vote and wins the election. That doesn’t mean that Nate Silver skews the polls (as is suggested by this repulsive article at, which, among other things, criticizes Silver for being thin and having a soft voice). Romney winning the election with 51 percent of the vote is well within the margin of error, as Silver clearly indicates. That’s what too close to call is all about.

As for me, I’ve been telling everybody that I predict that Obama will win, but would not at all be surprised if Romney wins (very disappointed, yes; surprised no).

Photo of the day

So many amazing Sandy photos out there.  I think we should mix things up, though.  Really loved this image from N&O day’s best:

A runner makes its way across to the stables at Taunton racecourse on October 30, 2012 in Taunton, England. (Photo by Alan Crowhurst/Getty Images)

NC is tied?

Although you cannot judge from just any one poll.  This Elon poll highlighted in Nate Cohn’s post is done right (I actually read through the whole release).  Called cell phones, etc.

How could North Carolina be close during a tied national election? Demographics. While Obama has bled white supporters nationally, his coalition was less dependent on whites in North Carolina than any other state. Demographic changes are sweeping the state at a fast enough rate to compensate for or potentially overwhelm declining turnout rates or losses among Obama ’08 voters.

While many have questioned whether Obama could still win North Carolina, the polls, demographics, registration numbers, and perhaps even the early voting numbers start to speak for themselves. Romney holds a 2 point lead in an average of North Carolina polls conducted since October 15, and that’s despite two automated surveys showing Romney ahead by more than 6 points (Elon, Civitas, and Groves (D) each call cell phones).

Cohn gives good reasons while NC should still be close.  And based on advertising, the Romney camp clearly does not think they have it in the bag.  It seems hard to imagine Romney winning the national popular vote while struggling so to win NC, but stranger things have happened.  Still, I do think you have to consider the closeness in recent Florida and NC polls a big plus for Obama.


Then again

Nate Cohn says the latest polling data shows a split in polls, not a split in electoral college vs. popular vote:

But polls over the last week give more credence to the argument that there isn’t a split between the popular vote and the Electoral College, but instead between the national polls and the state polls. Over the last week, we’ve learned more about the state of the race in several under-polled states, like California, Massachusetts, Tennessee, Texas, North Dakota, Arkansas, Oklahoma, Georgia, and New York. While Obama was performing worse than ’08 in most surveys, some showed him performing nearly as well while only a few showed Obama performing eight-plus points worse than he did four years ago–the magic number to give Romney a popular vote victory despite Obama’s resilience in the battlegrounds.

This was even true in Appalachia, where many have speculated that Obama might suffer catastrophic losses. Recent polls show a uniform swing in Tennessee, Arkansas, Louisiana, and one poll even shows Obama doing surprisingly well in Oklahoma. If one takes post-debate state polling averages and then categorizes states into demographically similar regions with a historic tendency to move together, one can calculate a national estimate by assuming that each region’s voting preference changes by an amount equal to the movement of the component states with post-debate state polls. This method shows Obama leading by 1.5 to 2 points in the national popular vote, depending on regional choices…

Presuming that those state polls are accurate and Obama doesn’t lose 7 or 8 points in his home state of Illinois, a Romney national popular vote win would require big 9 or 10 point improvements in California and Texas, where large gains are difficult to imagine without declines in Hispanic turnout or support for Obama, and large gains in the highland states of the inland south, where many suspect Obama’s support could collapse to levels unseen since McGovern. The state polls don’t yet support either half of this scenario. If it happens on Election Day, score a win for the national polls.

An electoral college/popular vote split

Nate Silver’s latest model puts the likelihood of Obama winning the electoral college and losing the popular vote at almost 6% (and 2% for the converse).  Charlie Cook’s latest analysis also suggests this as a very real possibility:

The seven jump ball states with a total of 94 electoral votes are Colorado (9), Florida (29), Iowa (6), New Hampshire (4), North Carolina (15), Ohio (18), and Virginia (13).

To win, Obama needs to win states with 17 (or 18 percent) of the 94 electoral votes in the seven Toss Up states, while Romney needs a whopping 79 (or 84 percent) of the 94 electoral votes.

However, the Obama advantage is not as clear cut as this suggests.  In each of these states, Obama and Romney are within 5 percentage points of each other and in most they are within 2 or 3 points of each other.

This race appears to be going to the wire, and the chances of a split popular vote/Electoral College vote are real. Romney looks to be at least an even-money bet for the popular vote, but Obama seems to have the edge in the electoral vote.

As your presumably know, I hate the electoral college.  Before 2000, I thought one of these perverse results would finally result in us getting rid of it. Obviously, it had almost no impact.  Now, maybe I’m completely wrong, but I think that there’s no way the Republicans would simply put up with this crap in a way Democrats have.  I’m thinking this result would have a decent chance of finally ending the electoral college behind the Republican fury.

Now, presumably there’s a downside in that Republicans would spend four years attacking Obama’s legitimacy in a way Democrats never did with GWB.  But honestly, can Republicans cooperate with Obama any less than they have?  And, as much as I would hate to see Obama win this way, with the outcome being protecting Obamacare, I’ll definitely take it.

%d bloggers like this: