Home A Republican wave? A big Dem poll skew? What the hell happened?

A Republican wave? A big Dem poll skew? What the hell happened?


Democrat vs. Republican on whiteConventional wisdom had Republicans big favorites to win a Senate majority, build on its strong House majority and pick up several governorships.

The wisdom didn’t have it being as convincing as it was.

Was it a Republican wave, or were the polls just all that far off?

Virginia’s surprisingly toocloseforcomfort U.S. Senate race can give us some clues. Democratic incumbent Mark Warner was presumed to be such a heavy favorite that the race was barely polled, outside a couple of Virginia colleges that undertook regular polls maybe as much as an academic exercise as anything else.

Warner had a 25-point lead in the summer, and maintained a safe double-digit lead up until one final poll released on Oct. 31 showed a narrowed margin at seven points.

Even with that poll seeming to indicate late movement toward Gillespie, it was one poll, in a climate with Democrats in several perceived battleground states running in the range of one, two, three points behind their Republican counterparts.

The numbers had partisan Democrats thinking that a late wave in their direction could help their side pull the big upset, maintain control of the Senate, and set the tone for the 2016 road to the White House to their liking.

If there was a late wave, it was breaking against them, as it turned out. Not only was the double-digit lead for Warner about to become a razor-thin one-point (or less) win on Election Night, but that slew of one- to three-point races turned into GOP routs.

A quick analysis by Nate Silver at FiveThirtyEight.com showed an average Democratic bias in the pre-election polls at +6, and that analysis didn’t include in its sample the Warner-Gillespie race, which itself was a +9 (from the RealClearPolitics pre-election poll average).

So we saw something: was it movement, a Republican wave, or just simply shoddy polling?

Republican partisans might want to try to claim both. Terming it a Republican wave helps frame a narrative that voters don’t like the direction of the country under two-term Democratic President Barack Obama, and sent a message through their votes in the 2014 midterms that can serve as a warning for the bigger-stakes prize in 2016.

But Republicans also like the idea that the polls were just skewed, because reporters and egghead professors and statisticians are responsible for polls, and of course they’re biased against Republicans.

Dismissing the egghead pollsters being biased or just bad at their own math thesis, we have to look back at the wave idea, and go with the premise that the wave was somewhat late-breaking, enough not to be picked up on in the final-week polls.

What would have been at the root of a late-breaking GOP wave, then? You can’t just say, Voters don’t like Obama, even as exit polls show the president’s unfavorable ratings as a factor; the president didn’t suddenly run up high unfavorable numbers in the last week of the 2014 campaign.

Was Ebola the October surprise in 2014 that hurricane sandy was to 2012? That seems at least plausible, given the intense media focus on the two cases of Ebola contracted in the U.S., that made those two cases seem a pandemic.

There hasn’t been bad economic news, a terrorism attack, any losses on a foreign battlefield.

The two sides will be fighting over how to frame the post-election narrative, but don’t necessarily buy what they’re selling at face value, because it’s likely that they don’t know, either.

– Column by Chris Graham



Have a guest column, letter to the editor, story idea or a news tip? Email editor Chris Graham at [email protected]. Subscribe to AFP podcasts on Apple PodcastsSpotifyPandora and YouTube.