Voters are going to the polls Tuesday in West Virginia to cast ballots in a special election for governor pitting Earl Ray Tomblin, a Democrat who was appointed governor after Joe Manchin III was elected to the Senate, against Bill Maloney, a Republican.
The lack of national interest in the race is aptly demonstrated by the lack of polling on it. Just one firm, Public Policy Polling, has surveyed the contest. They had initially shown Mr. Tomblin with a lead, but it has shrunken with every poll since and was down to just 1 percentage point in a survey that the firm conducted from Sept. 30 to Oct. 2.
It’s a shame that we have only one polling firm in the field, because West Virginia is a tricky state to poll and precisely the kind of place where having multiple polls to pick from would allow us to make a more robust forecast.
In 2008, polls generally underestimated John McCain’s eventual 13-point margin of victory in the state. The Public Policy Polling survey was the exception and was right on the nose, calling the 13-point margin exactly. But other polls released in the closing days of the campaign had Barack Obama trailing by as few as 5 points, and one bizarre poll a month in advance of the election even showed Mr. Obama 8 points ahead in the state.
But in 2010, the polls underestimated the performance of Mr. Manchin, a Democrat, in what turned out to be a 10-point victory over John R. Raese, a Republican. Polls immediately before the election had Mr. Manchin with a smaller, 4-point lead on average, while one survey conducted just two weeks before the election had instead shown Mr. Raese with a 7-point lead.
It’s not surprising that West Virginia has proven to be a tricky state for pollsters. Although traditionally very Democratic and although still very heavily Democratic in its voter registration, it has turned into a red state at the presidential level and would be thought of as conservative by most definitions, especially on cultural issues. The state is very homogeneous demographically, meanwhile — very white and very working-class — so some of the shortcuts that pollsters take if they don’t collect a sufficiently random sample initially, like demographic weighting, will not bail them out here.
In short, anything from a double-digit victory for Mr. Tomblin to a double-digit victory for Mr. Maloney should not be considered terribly surprising. Or maybe the race will be as close as Public Policy Polling envisions it and we’ll be waiting for the last box of ballots to come in from Wheeling. If you wanted to make a case for Mr. Tomblin, it would be that a 1-point lead going his way is better than a 1-point lead going the other way and that his approval ratings aren’t bad.
If you are a believer in “momentum,” on the other hand, that would point in Mr. Maloney’s direction. (The evidence suggests that what people think of as momentum in pre-election polling is vastly overrated and that the polls reverse course as often as they continue in one direction — but the trendline toward Mr. Maloney has been unusually clear in this case.) Mainly, though, my “forecast” for this contest is high uncertainty, as polls have missed badly in both directions within the past two election cycles.
Regardless of the outcome, I do not think the results are likely to provide us with much in the way of useful information about the national landscape.
There’s no doubt that Mr. Tomblin is getting no favors from President Obama, who has never been popular in West Virginia to begin with and who is exceedingly unpopular there now, with an approval rating of about 30 percent.
But in contrast to special elections to the House, which do have some predictive power for future federal races, races for governor have been found not to be good predictors.
On top of that, there is no good way to benchmark the results. What results would we expect, say, if Mr. Obama had a 55 percent approval rating nationally instead of 42 percent? Well, it’s hard to know, and if you can’t answer that question, you should probably avoid drawing too many implications from the race. West Virginia has elected its share of Republican governors before. They elected one, Cecil H. Underwood, in 1996, when Bill Clinton was on the ballot and carried the presidential race there by 13 points. But this is the same state that helped to end Republicans’ hopes of taking over the Senate last year by electing Mr. Manchin, and Mr. Obama was not really any more popular then.