Missouri, the state that was once considered the nation’s ultimate bellwether, looks as though it is likely to be out of reach for President Obama this year, unless there is a significant shift toward him in the final 100 days of the campaign.
A Mason-Dixon poll of the state, released on Saturday, gave Mitt Romney a nine-point lead there. Mr. Romney’s nine-point lead matches his advantage from another poll of the state, conducted by the firm We Ask America, which was released earlier this week.
The forecast model now estimates that Mr. Romney has an 88 percent chance of winning Missouri in November. And Missouri has fallen off the list of tipping point states, meaning that it is very unlikely to be a decisive state in determining the winner of the Electoral College. The cases where Mr. Obama wins Missouri are probably those where he is headed toward some sort of near-landslide in the national race, like because of an unexpected rebound in the economy.
Democrats have made gains in some parts of the country in recent years — most notably in the Mountain West and in Virginia, which has increasingly begun to behave as a Northern state. But they have receded in the poorer states of the upland South, like Missouri, Arkansas, Kentucky, Tennessee and West Virginia.
With the exception of a few urban centers like Knoxville, Tenn., these regions have always been culturally conservative. But they also once had a reputation for being economically moderate or even populist, tending to support elements of the welfare state.
In the past four presidential elections, political partisanship has become more uni-dimensional, with a higher percentage of voters tending to align with the same party on both economic and social issues — even though there is often little intrinsic relation between them. In this part of the country, the trade seems to have worked against Democrats — or at least it has worked against the fairly liberal Democrats who have been the party’s presidential nominees in the past four election cycles.
Over all, Mr. Romney’s chances of winning the Electoral College improved somewhat in Saturday’s forecast, to 32.6 percent from 30.9 percent on Friday, according to the model.
One reason for Mr. Romney’s improvement — in addition to his strong Missouri poll — was that we went back and double-checked our polling database for cases in which polling firms had reported both registered voter and likely voter results within the same survey release.
A comparison between these numbers forms the basis of the forecast model’s likely voter adjustment, which shifts numbers from polls of registered voters toward Republican candidates since Republicans typically perform better on surveys that estimate how likely each voter is to actually participate in the election.
The model shifts registered voter polls toward the Republican candidate by 1.5 percentage points as a default, since this is the average change from previous presidential election years. But the adjustment can increase or decrease as additional data becomes available. Based on the data so far this year, the model now estimates that the shift will be slightly larger than average this year, equal to about 2.3 percentage points instead.