In today’s Telegraph column, Kevin Landrigan asks if Sen. Kelly Ayotte is “in free fall” over her vote opposing expanded background checks for gun purchasers. “Let’s not carried away with this,” he answers.
Landrigan then proceeds to make his case by discrediting Public Policy Polling, the pollster who released a survey this week showing Ayotte’s favorability rating has dropped 15 points since her vote.
Let’s ignore the perjorative slant, (“Fortunately, we have other data, and it doesn’t affirm PPP’s finding, to be sure.”) and focus on Landrigan’s argument, which is misleading and deceptive.
Strike one: Apples and oranges
Landrigan compares the PPP survey to a poll by the UNH Survey Center:
But the notion that Public Policy Polling’s survey found that Ayotte’s favorability dropped 15 points bears further scrutiny. … Fortunately, we have other data, and it doesn’t affirm PPP’s finding, to be sure.
[T]he UNH poll out last week didn’t find any slippage for Ayotte. She was at 50 to 25 percent compared to 51 to 28 percent two months earlier.
Remarkably, Landrigan fails to point out that the UNH poll was completed before Ayotte’s gun safety vote!
Strike two: Party identification and unskewed polls
Landrigan echoes the discredited “skewed poll” refrain we heard from Republicans in the last presidential election, by suggesting the PPP survey oversampled Democrats:
The PPP also gave Democrats a 6 percent advantage in respondents over identified Republicans, while actual party registration in the state has both parties close to a statistic tie.
New York Times polling expert Nate Silver knocks down that argument:
But pollsters, at least if they are following the industry’s standard guidelines, do not choose how many Democrats, Republicans or independent voters to put into their samples — any more than they choose the number of voters for Mr. Obama or Mitt Romney….
[P]arty identification is not a hard-and-fast demographic characteristic like race, age or gender. Instead, it can change in reaction to news and political events from the party conventions to the Sept. 11 attacks. Since changes in public opinion are precisely what polls are trying to measure, it would defeat the purpose of conducting a survey if pollsters insisted that they knew what it was ahead of time.
And Landrigan fails to point out the party identification measured by the UNH survey, which gave Democrats an identical six-point advantage!
Strike three: Never let the facts get in the way of a good story
As if cherry picking results from a previous survey discredits the current one, Landrigan looks back at PPP polling from Ayotte’s 2010 race:
Keep in mind that the Democratic-leaning PPP had 2010 Democratic nominee Paul Hodes within four points of knocking off Ayotte just prior to that election. Ayotte won by 23.
This is wrong. On September 15, 2010, PPP did indeed release a poll showing Ayotte leading Hodes by four points, 47-43. Six weeks later however — “just prior to that election” — a survey from PPP indicated Ayotte had widened her lead to 15 points, 56-41. Those results were similar to the 18-point lead measured by UNH, which gave Ayotte a 54-36 lead.
Landrigan compares the results of the PPP survey taken after Ayotte’s gun safety vote to one taken before — without issuing a caveat. He suggests a six-point Democratic advantage in party identification makes the poll unreliable, but fails to mention survey professionals disagree — or that the UNH survey has the same breakdown. He misstates previous survey results. We rate Landrigan’s column Pants on Fire!