I wrote earlier this week about some of the challenges in comparing state polls and national polls. Sometimes, apparent differences between the two sets of numbers can result from methodological quirks of the polling firms that are active in each arena, as well as random sampling error.
With that said, we are starting to see a bit of a gap between our Electoral College and popular vote forecasts based on the latest polling data this week — one which potentially favors President Obama.
In general, the polls from nonswing states this week, ranging from New Jersey to North Dakota, were mediocre for Mr. Obama. But his numbers held up better in swing states.
Nowhere was this more apparent than in Ohio, where there were two new polls out on Friday. One of them, from the firm We Ask America, gave Mr. Obama an eight-point lead there. Another, from Magellan Strategies, put Mr. Obama up by two points.
Our model “thinks” the Magellan Strategies poll is a more realistic estimate of the state of play in Ohio. The model now forecasts a three-point victory for Mr. Obama there, which it translates to about a two-in-three chance of his winning the state given the uncertainty in the forecast.
Mr. Obama’s projected three-point lead in Ohio is important for the following reason, however: it’s slightly larger than the 2.4-point advantage that the model now gives Mr. Obama in the national popular vote.
In other words, based on the data so far this year, Ohio has been slightly Democratic-leaning relative to the country as a whole. That reflects a reversal from the usual circumstances. Normally, Ohio — though very close to the national averages — leans Republican by two points or so.
A split between the winners of the popular vote and the Electoral College is still relatively unlikely, in the view of the model. There is only a 3.9 percent chance that Mr. Obama will win the Electoral College but lose the popular vote, it estimates.
However, the model now assigns just a 1.3 percent chance to the reverse happening: Mr. Romney’s winning the Electoral College while losing the popular vote.
Ohio is a big part of the reason. It’s by no means impossible that Mr. Romney could win the Electoral College despite losing Ohio, but it would be an uphill battle.
If Mr. Obama won all the states that John Kerry did in 2004, and Ohio, he would have 264 electoral votes.
Winning New Mexico in addition to these states, where Mr. Obama is heavily favored, would make the Electoral College a 269-269 tie. That is not necessarily a problem for Mr. Romney. The incoming House of Representatives is likely to contain a majority of Republican delegations, which would presumably vote for him under the 12th Amendment.
However, if Mr. Obama claimed any other competitive state that Mr. Kerry lost — Nevada, Iowa, Virginia, Florida or Colorado, for example — he would win the Electoral College outright, regardless of what happened in New Mexico.
Of the states, Nevada is probably the most problematic for Mr. Romney. He has trailed in all the polls there, and it has increasingly begun to behave as a Democratic-leaning state.
Another plausible scenario is that Mr. Romney could win New Hampshire, which Mr. Kerry won in 2004, but lose New Mexico and Nevada. In that case, he’d lose the Electoral College by two votes, with 268 electoral votes to Mr. Obama’s 270.
Alternatively, Mr. Romney could try to win one of the Midwestern states that Mr. Kerry won — most likely, Pennsylvania, Michigan or Wisconsin. But these states have broadly similar demographics to Ohio, and it is unlikely that Mr. Romney would win them conditional upon losing the Buckeye State.
Although state economic statistics have only modest predictive power as compared with the national numbers, Mr. Obama may be benefiting from the fact that Ohio’s unemployment rate has fallen substantially — to 7.3 percent from a peak of 10.6 percent. Had the national unemployment rate declined by a similar margin, Mr. Obama would likely be a clear favorite in the election right now.
Much of the gain in Ohio has been from auto industry jobs, a relative bright spot in the economy, and a potential problem for Mr. Romney since he opposed the federal bailout of the major auto makers.
Mr. Romney could seek to even the score by naming Senator Rob Portman of Ohio as his vice-presidential nominee. Vice presidents do not make as much difference as the conventional wisdom sometimes holds: a net gain of two points by appointing a home-state candidate is a roughly realistic estimate, although it can vary a lot from case to case. Still, gaining two points in Ohio would provide a pretty meaningful push to Mr. Romney given the way that the polls look right now.
Mr. Obama also benefited on Friday from the rise in the stock market, which improved the model’s economic index. The S. & P. 500 closed at 1,386 points on Friday, its highest figure since May.
The stock market is likely to remain volatile in the weeks ahead, as investors seek to weigh the consequences of critical economic news — including next Friday’s jobs report, possible actions by the Federal Reserve to stimulate the economy and efforts by the European Central Bank to defend the euro zone.
My personal read on the stock market is that it seems to be pricing in a fairly high likelihood of stimulative actions by central banks, both in the United States and Europe. If investors do not get what they are hoping for, it could decline significantly.
In the meantime, Mr. Romney’s campaign will need to consider whether it trusts the polls showing him down in Ohio, and if so, what actions it might take to remedy that problem.