Push polls
A campaign contact disguised as an opinion survey, whose questions exist to plant a damaging claim about an opponent rather than to measure anything.
A survey has standing that a campaign call does not. People answer a pollster who would hang up on a canvasser, and they take the questions as neutral. The push poll borrows that standing: it calls thousands of voters, asks a few real-sounding questions, and then asks whether they would be "more or less likely" to support a candidate "if you knew" some allegation, true, false, or unprovable. Nobody tabulates the answers. The allegation has been delivered to every household called, under the cover of research.
The device is the loaded question industrialized. The American Association for Public Opinion Research condemned it in 1997 as "telemarketing masquerading as research" and distinguished it from legitimate message testing, which uses small samples and reports its results. Exploits Authority Bias and the Misinformation Effect.
Historical examples
Sabato and Simpson's 1996 study documented the spread of push polling through congressional and state races in the mid-1990s: calls to tens of thousands of voters asking whether allegations about an opponent's record, family, or finances would change their vote. Several state legislatures responded with disclosure laws requiring such calls to identify their sponsor.
Sabato, Larry J., and Glenn R. Simpson. Dirty Little Secrets: The Persistence of Corruption in American Politics. New York: Times Books, 1996.
Voters in the 2000 South Carolina Republican primary reported calls asking whether they would be more or less likely to support Senator John McCain "if you knew he had fathered an illegitimate black child." The McCains had adopted a daughter from Bangladesh. The calls were never traced to a sponsor, which is the design; the episode is the most cited instance of the device.
Banks, Richard L. "The Anatomy of a Smear Campaign." The Boston Globe, 21 March 2004; American Association for Public Opinion Research, "AAPOR Statement on Push Polls," revised 2007.
In advertising and commerce
"Sugging," selling under the guise of research, is the commercial form: a "survey" that exists to deliver a pitch or a competitor's alleged failing, and that the market-research profession's codes prohibit. The prohibition exists because the technique works and damages the standing of real surveys.
ESOMAR/ICC. International Code on Market, Opinion and Social Research and Data Analytics. Amsterdam: ESOMAR, 2016, on the separation of research from selling.
Write a genuine survey question about a candidate. Now rewrite it as a push-poll question. What changed, and who benefits?
In the wild
Short examples of the technique at work, each with what to notice. The same examples are questions in the quiz, where the answer is not given away.
"Would you be more or less likely to vote for the councillor if you knew she had been sued by her own business partners?" The caller won't say who sponsored the survey.
What to notice. A campaign message delivered as a survey question to thousands, with no results to publish.
The night before the vote, 20,000 households received a two-minute "poll" whose last three questions all described the challenger's alleged tax problems.
What to notice. Scale, timing, and one-directional questions mark a push poll rather than research.
"Quick survey: are you aware that Brand X's cleaner has been linked to skin irritation? Would you consider switching to ours?"
What to notice. Selling under the guise of research; the "survey" delivers a competitor's alleged failing and a pitch.
Sources
- Sabato, Larry J., and Glenn R. Simpson. Dirty Little Secrets. New York: Times Books, 1996. ISBN 978-0812924510
- American Association for Public Opinion Research. "AAPOR Statement on Push Polls." 1997, revised 2007.
- Traugott, Michael W., and Paul J. Lavrakas. The Voter's Guide to Election Polls. 4th ed. Lanham: Rowman & Littlefield, 2008. ISBN 978-0742547162