Seven weeks out from the 2014 general election, the first really significant differences of opinion have broken out among the non-partisan experts who make a living forecasting election results.
FiveThirtyEight, the New York Times‘ Upshot, and the Washington Post have all shifted their Senate forecasts in a Democratic direction. The Post’s “Election Lab” now has Democrats favored to hold onto the Senate, with the Times and 538 approaching a coin toss.
Outside the data-oriented forecasters, though, confidence in a GOP Senate victory is actually going up. Roll Call‘s Stu Rothenberg is now predicting “a sizable Republican wave,” mainly because of the president’s low approval ratings, which he considers supremely important.
But the only war of words that has broken out to date was between two of the data wizards, FiveThirtyEight’s Nate Silver and Princeton Election Consortium’s Sam Wang. The latter, who is currently forecasting the probability of Democratic control of the Senate at 81%, refuses to use anything other than polling data, and criticizes those like Silver who use “state fundamentals”–historical precedents about a state’s performance in similar elections. Silver argues polls-only projections like Wang’s fatally lose accuracy when there’s too little polling on a particular race.
There remains plenty of time for all the experts to reach a consensus by November 4, but at the moment, it looks like reputations and methodologies as well as private wagers will be at stake when the votes come in.