There’s nothing outrageous about PPP’s Kentucky Senate poll

Matthew Hurtt Contributor
Font Size:

I got annoyed Tuesday when folks began using the Twitter hash tag #pppquestions to criticize Public Policy Polling’s recent Kentucky Senate poll. It’s no secret that PPP is a Democratic polling firm, but the reaction from the right-of-center Twitterverse — including the NRSC’s Brad Dayspring and the RNC’s Sean Spicer — bothered me.

The charge, highlighted in a post on The Right Sphere, is that PPP is “push polling” voters in Kentucky regarding the potential Senate match-up between Republican Senate Minority Leader Mitch McConnell and Kentucky’s Democratic secretary of state, Alison Lundergan Grimes.

The American Association for Public Opinion Research (AAPOR) defines push polling as “a form of negative campaigning that is disguised as a political poll. ‘Push polls’ are actually political telemarketing — telephone calls disguised as research that aim to persuade large numbers of voters and affect election outcomes, rather than measure opinions.” Push polls survey far more people than political polls, and their results aren’t recorded.

The issue of “push polling” has made the news a few times lately. ThinkProgress accused a “mysterious conservative group” called SSI Polling of dirty tricks in the special election between Mark Sanford and Elizabeth Colbert Busch. Scott Keyes and Adam Peck write:

A mysterious conservative group has been placing highly-misleading phone calls to South Carolina voters, trying to dissuade them from voting for the Democrat in an upcoming congressional special election.


South Carolina has a reputation for dirty tricks, and next week’s special election between former Gov. Mark Sanford (R) and businesswoman Elizabeth Colbert Busch (D) is no exception. One of the most popular tactics is known as “push polling,” whereby a group calls up voters under the guise of conducting a poll, only to ask questions that leave the voter with a highly-misleading impression about a certain candidate.


ThinkProgress spoke with multiple individuals in South Carolina’s first congressional district who have received push polls from an unknown conservative group that only referred to itself as “SSI Polling.”


April Wolford, a middle-aged woman who has long been active in Democratic politics in the state, was one. At 12:55pm on February 25th, Wolford’s cell phone lit up with “Unavailable” on the caller ID screen. A young man without a discernible accent — “he certainly wasn’t from South Carolina,” she noted — said he was conducting a poll and began with general questions about the race. “But they quickly got slanted,” Wolford noted, “and they didn’t ask a single question about Sanford at all!”

In an update to that post, ThinkProgress confirmed that SSI Polling is indeed a legitimate market research firm that was testing messages in the SC-01 special election.

The questions asked in PPP’s Kentucky poll ranged from the straightforward (“Do you approve or disapprove of Mitch McConnell’s job performance?”) to the arguably skewed (“Mitch McConnell has voted to cut taxes for millionaires like himself, while supporting cuts to Social Security and Medicare for hardworking Kentucky seniors. Does this make you more or less likely to vote for him, or does it not make a difference?”).

PPP asked 556 Kentuckians 14 question via automated telephone poll: five of them straightforward, four of them demographic and five of them “negative.” But this isn’t a case of push polling; it’s an example of “message testing.” The five negative questions were designed to help a potential Democratic challenger see which attacks resonate with voters. One hint that PPP was messaging testing is the small number of people — 556 — it surveyed. Another is that PPP collected and then published the data. Based on the results, a potential Democratic challenger might want to paint McConnell as a “Washington insider” who is “out of touch” with average Kentuckians.

The crosstabs actually reveal some good news for McConnell. He’s picking up one in four Democrats to Grimes’ one in five Republicans, and 42% of Independents over Grimes’ 39%. Both figures are within the margin of error, though.

Stuart Rothenberg complained at some length back in 2007 about the difference between actual message testing (“push questions,” they’re called) and push polling, and why it’s important to understand the difference. At the time, Congress was taking up the issue in a reincarnated form of the Push Poll Disclosure Act of 2005. Rothenberg writes:

Push questions are not the same thing as push polls. Push questions, which are included in a survey of only 500 to 1,000 respondents, are a legitimate part of a public opinion poll that seeks to test effective messages.


My guess is that someone who didn’t know very much about survey research — I’m putting my money on some twentysomething who worked in the media — heard about push questions and incorrectly confused them with advocacy calls, creating the illogical term “push polls.”


Serious polls can include push questions that contain some explosive or even incorrect information, but that doesn’t make them advocacy calls. Testing possible messages is a legitimate survey research function, and as long as the question is asked of a small sample and seeks to get a response to know whether the issue is useful in an election, it really doesn’t matter how negative the message is.

If I wanted to know what messages increased my opponent’s negatives, I’d ask “push questions” that push respondents in one direction or another. “Would you be more or less likely to vote for Michael Vick for dog catcher if you knew he got busted gambling on dog-fighting matches?” Is it negative? Yes. Is it true? Yes. Could it move voters in one direction or another? Yes. This data would be collected by a firm, put together in a report and used by the campaign.

If I wanted to “push poll” voters, I’d hire a firm to send out a robocall to thousands of people in the district. The robocall would not collect data, and the statements would likely be false or extremely misleading. During the 2000 South Carolina Republican presidential primary, for instance, an anonymous robocall alleged that John McCain had fathered an illegitimate black child. No data was recorded, and the information was completely false.

The questions asked by PPP of Kentucky voters, while skewed to increase McConnell’s negatives in a number of instances, are generally true. In the age of never-ending Internet outrage, it’s important to understand what’s worth getting upset over and what isn’t. The PPP questions are a prime example of the latter.

Matthew Hurtt is a sometimes campaign flack who’s worked with polls and message testing. He has no clients in the Kentucky Senate race but supports Republican candidates, generally. Follow him on Twitter: @matthewhurtt