The Virginia GOP's relationship with its pollster appears unclear after he wrongly predicted a 34 percent margin of victory for former U.S. House Majority Leader Eric Cantor.
"I have my pollster and I like him," Virginia GOP Communications Director Garren Shipley says of pollster John McLaughlin.
But Shipley declines to be more specific on the party's status with McLaughlin, or comment on media reports that the national party is encouraging Republicans to drop him.
"I don't really have anything I can say about that," Shipley says.
Cantor, a seven-term incumbent, went into the June 10 primary armed with numbers from McLaughlin & Associates, which showed a decisive victory against tea party-aligned Randolph-Macon College professor Dave Brat.
"It's a fiasco," says Norm Leahy, a board member at conservative Virginia politics blog Bearing Drift. "There's no way to sugarcoat it."
The defeat is one in a series of wrong predictions from the McLaughlin team in Virginia, including an October 2012 poll that showed Democratic U.S. Senate candidate Tim Kaine losing to Republican candidate George Allen. Kaine went to Washington.
McLaughlin offered an explanation for Cantor's defeat in a poll released last Thursday, showing that the 10-point loss margin came from non-Republicans infiltrating the open primary.
"It is now clear that Eric Cantor's national standing gave the race a lot of local interest among many more voters than just past Republican primary voters," McLaughlin says in his analysis, "including politically interested independents and Democrats as well."
That explanation doesn't add up, Leahy says.
"[McLaughlin] did an enormous disservice to his client," Leahy says. "If he doesn't model that sort of behavior, that's an even greater black mark on what he's done."
Leahy says one mitigating factor is that no one else saw Brat winning either.
"Not even [Brat] did," he says. "It's sort of like swimming in the ocean at night. You can't see the wave until it's on top of you."