August 22, 2015 push poll
(1990’s | journalese (politics) | “dirty trick”)
Push polls originated in political campaigns. They are presented as impartial, conducted by an organization not technically connected with a candidate (although if a respondent pushes back hard enough, the caller may be forced to reveal his true employer). The point is not neutral assessment of public opinion, but rather influencing it directly, often by making a dubious, if not outright false, imputation about an opponent — sometimes the mechanism relies on laudatory comments about oneself, but the intent is the same. It usually takes the form, “If you knew so-and-so about Candidate X, would it make you more or less likely to vote for her?” The phrase established itself in the mid-1990’s, with the first hit in LexisNexis due to David Broder in October 1994. By the 1996 election, the expression had common currency in political reporting, and many commentators no longer bothered to define it, which had been the rule only a few months earlier.
There’s nothing new about libeling political opponents, but the problem is the means. To be effective over the long term, opinion polls have to be fair, designed to avoid favoring one group over another. A survey has the presumption of fairness; therefore, it’s worse to use a poll to perpetuate slander than to use other means. The main point, as stated by Matthew Reichbach in the New Mexico Independent (September 22, 2009): “a push poll is not a poll at all. Its a fraud, an attempt to disseminate information under the guise of a legitimate survey.” It is a fraud in that it presents itself in a misleading way, but the “information” conveyed may be fraudulent, too. Reputable pollsters hate them and are forever calling for an end to push polls.
In 1996, the derivation of “push poll” was generally explained as a simple elaboration on the act of pushing voters away from a particular candidate. That’s folk etymology, but there is probably some underlying truth. The term comes out of pollsters’ jargon, by evolution or corruption. “Push poll” is actually a descendant of “push question,” described in 1982 by William Safire as “designed to squeeze respondents to come up with the answer the sponsors want.” It’s a variety of loaded question native to the survey business, and not necessarily unethical, although it does lend itself to unethical use.
A push poll may contain more than one push question, heaven knows. In 2001, I was on the receiving end of a Bloomberg-for-mayor push poll which consisted almost entirely of pro-Bloomberg statements masquerading as questions. I finally said something like “Aw, c’mon,” and the (apparently young and definitely inexperienced) questioner agreed that the bias was pretty obvious. I kept going, answering each question gamely in the most anti-Bloomberg manner I could muster, but the whole thing was a farce. Did Bloomberg pay for it? Who knows? Yes or no, the whole process showed nothing but contempt for our intelligence.
And now, the scope of push polls has broadened, and you don’t have to be a politician any more. Anyone trying to influence public opinion — a corporation (Walmart seeking to open a store in Chicago), a social movement (a group trying to promote, or scuttle, gun-control legislation), etc. — can initiate them. I’m pleased to note that the phrase still carries strong opprobrium, and it is thrown around in grim accusation or indignant denial, never in approbation. It may be true that one man’s push poll is another man’s opposition research, and political professionals once defended push questions as legitimate, if they raised a verifiable point about the rival candidate. No one defends them any more, at least not in public.