Push Polls

PUBLICLY CONDEMNED BY professional polling agencies and associations, a push poll is not a poll at all, but a campaign technique designed to deride the opponent. Though it follows the format of a poll, those conducting a push poll do not care about the respondent's answer. Like a rhetorical question, it only asks in order to pass on information and ideas. For instance, a multi-question poll may open with a block of information and a leading question to which, as phrased, there is only one reasonable answer; a follow-up question will then indict the target of the poll, the first question having warmed up and conditioned the respondents to get them thinking about the issue.

In the guise of seeking a voter's opinion, a push poll seeks to influence and even deceive a voter. Often, a large block of information will be presented, with only one question asked, “Will this affect the likelihood of your voting for Candidate Smith?” The question reinforces the idea that the respondent's opinion is important; respondents may extemporize beyond the simple yes or no, but what matters to the pollster is that they listened to that block of information. Consequently, often many more voters are contacted in push polls, since one common objective of a push poll is to smear the opposition candidate. It may have concerned candidate Smith's voting record, for example, perhaps honestly (as a way to notify voters of candidates who have supported positions unpopular in the voter's neighborhood) and perhaps deceptively (a senator voting in support of civil unions may be described as voting against traditional heterosexual marriage).

But the information passed on by a poll need not even be true. During the 2000 South Carolina presidential primary, a polling agency assumed to be affiliated with the Bush campaign called voters and asked them if their vote would be affected by the revelation that John McCain had fathered an illegitimate child with a black woman. He had not done so, nor did the poll actually state that he had, but if treated as a fair and honest poll, it sounds as though it is making that claim. Given the human tendency to remember information better than its source, the poll must have been very effective at spreading and creating this rumor, and McCain lost the primary and the election.