Smart money is on the state polls, showing Obama ahead
MITT ROMNEY and President Barack Obama remain roughly tied in national polls, while state polls suggest a lead for Obama in the electoral college – the mechanism through which the president is actually elected.
Many people take this to mean there is a fairly good chance of a split outcome between the electoral college and the popular vote. But the story may not be so simple.
For both the state and national polls to be accurate, and the maths to work, something needs to give. If Obama is performing well in swing states, but is only tied in the popular vote nationally, he must be underperforming in non-competitive states.
But polls of non-competitive states don’t always co-operate with the story. Take the polls that were released last Tuesday.
Obama trailed by “only” eight points in a poll of Georgia. That is somewhat worse than how he fared in 2008, when he lost Georgia by five points. But it is only a little bit worse, whereas the national polls suggest a larger decline for Obama in the popular vote.
Or take the poll of Texas that had Obama behind by 16 points, only a four-point decline from 2008.
These high-population, Republican-supporting, red states are just the sort of places where Obama would need to lose a lot of ground to increase the likelihood of his winning the electoral college while losing the popular vote.
Is it that he’s underperforming in deeply blue, or Democrat, states rather than deeply red ones?
There have been some numbers that match this assumption: Obama got some mediocre polling in Oregon on Tuesday, for instance. But he also had polls showing him ahead by 23 points in California and 31 points in Massachusetts.
Yes, I am cherry-picking a bit, but the discrepancy seems to hold if you look at the data in a more comprehensive way. Nearly every forecast that evaluates the election based on state polls seems to hint at a very slight lead for Obama in both the popular vote and the electoral college.
What if turnout next Tuesday looks more as it did in 2004, a better Republican year? That does not help to break the discord between state and national polls: in this case, Obama would lead by two percentage points.
Or we can weigh the states by their turnout in 2010, a very good Republican year. That does not help, either: based on this method, Obama leads by 2.1 percentage points.
But perhaps it is the national polls which tell the right story of the race; perhaps the state polls systematically overrate Obama’s standing.
It is certainly possible. (It keeps me up late at night.) If the polls in states like Ohio and Wisconsin are wrong, then pollsters are not going to be happy as the count unfolds.
That said, my decision to cast my lot mostly with the state polls is not arbitrary. In recent years, they have been a slightly more unbiased indicator of how the election will play out.
Bias, in a statistical sense, means missing consistently in
one direction – for example, overrating either candidate’s performance across a number of examples. It is to be distinguished from the term “accuracy”, which refers to how close you come to the outcome in either direction.
If my forecasts overestimate Obama’s vote share by 10 percentage points in Nevada, but underestimate it by 10 percentage points in Iowa, the forecasts will not have been biased, since the misses were in opposite directions: they will just have been bad.
In recent elections – since state polling data have become more robust – state polls have been more free of bias than national polls. On average since 1992, the state polls have had a bias in one direction or another of 1.1 percentage points, while national polls have had a 2.1-point bias.
We are approaching the point where Romney may need the state polls to be systematically biased against him in order for him to win the electoral college. That certainly could turn out to be the case: if Romney wins the popular vote by more than about two percentage points, for example, he will be very likely to cobble together a winning electoral map.
But the historical evidence slightly favours the state polls, in my view, when they seem to contradict the national ones. If the state polls are right, then Obama is not just the favourite in the electoral college but probably also in the popular vote.
Nate Silver is an American statistician, baseball data expert, psephologist and writer for the New York Times