Outgoing House Majority Leader Eric CantorEric Ivan CantorWhite House says bills are bipartisan even if GOP doesn't vote for them Trump the X-factor in Virginia governor race Conservative House Republican welcomes Clark as chief of US Chamber MORE's (R-Va.) campaign manager is pinning the blame on Democrats for his shocking loss.

Longtime Cantor adviser Ray Allen, in his first interview since Cantor was stunned by little-known professor Dave Brat (R), told The Hill that he believed Cantor was a victim of meddling from Democrats who crossed over in the primary to vote against him.


"We had probably 15,000 card-carrying Democrats come into this primary. There's just no way to anticipate something like that," Allen tells The Hill.

Some Republican allies of Cantor are quietly furious that he and his team had no idea what was coming, and blame inaccurate polling and a weak campaign for not recognizing he was in trouble. The majority leader was so confident that he spent much of Election Day in Washington, D.C., working on unrelated issues.

Allen defended his pollster, John McLaughlin, who less than two weeks before the race conducted a poll that found Cantor up by a 34-point margin. Brat ended up winning by 11 percentage points, a 45-point swing.

Calling McLaughlin "one of the absolute best in the business," Allen said there was no way they could have seen this coming from normal polling methods.

"My guess is now looking at the precinct-by-precinct analysis, probably half this electorate wasn't Republican," he said. "There's not a pollster in this country who polls voters from the other party in primary."

His analysis mirrors McLaughlin's.

"Where the voter surge was, it was Democrats who united against Cantor, who don't like him. And the mistake for our campaign above all the others, we should have polled Democratic primary voters and see if they were going to come into the Republican primary," McLaughlin said. "That's really the difference here. The reason the poll was wrong is no one anticipated Democrats were going to come in and play and they did, and that was the surge in turnout. That's why the poll was wrong, we were polling off the Republican sample."

Turnout was indeed much higher in this primary than it was in 2012. But outside observers are highly skeptical that Allen's analysis is correct.

George Mason University professor Michael McDonald, a vote modeling expert, has crunched the precinct-level data on what happened and said it was highly unlikely that enough Democrats turned out to swing the election, noting that turnout increased more in heavily Republican precincts than heavily Democratic precincts this year, much less cover the 45-point difference between the poll and the actual result.

"It sounds like a political answer, not a serious analytical guide into what actually happened. It sounds like rhetoric, not a serious analysis of what happened. And the pollster has a big stake on this too, McLaughlin's reputation is on the line and his performance has not been good in recent elections," said McDonald. "To somehow claim Democrats are somehow responsible for the miss on the poll — maybe it accounts for a couple of percentage points, but it's a drop in the bucket with what was really going on with that poll."

Allen scoffed at similar analysis that has come out rebutting his and McLaughlin's theory.

"I don't need the New York Times analysis," Allen scoffed when asked about precinct-level analysis that proved his claims dubious. "I'm down here, I know what happened. It's not the whole story but it's a big hunk of the story."

Allen said it had been a trying few days, and that he'd finally slept Wednesday night, meaning he wasn't in "physical pain" anymore.

"I've been with Eric for 20 years, since his very first House of Delegates race. It's just tough," he said.

Allen took the bullet when asked why the campaign didn't poll closer to the election.

"I don't know. In hindsight maybe we should have gone in again. But that was my decision, not John's."