The intoxicating drug of prediction
Responsible journalists and professional commentators, editors and media leaders must reject this unbridled passion for polling.
According to a public opinion poll presented by Channel 2 (conducted by the Maagar Mohot research institute ), the union between Likud and Yisrael Beiteinu can be expected to garner 42 Knesset seats. At that very moment, another poll was presented by Channel 10 (conducted by the Midgam Project ), indicating that the combined list will capture only 35 seats. The significant gap between the surveys, with regard to other parties as well, is surprising. After all, the two polls were both supposed to have sought the opinion of a representative sampling, both were undertaken on the same day, were about the same political parties and asked the same question (If the elections were held today, who would you vote for? ).
Variations among opinion polls and pollsters is nothing new. Right after Yair Lapid announced his decision to establish a political party, opinion polls were published, some projecting he would snag 22 seats and others, only seven. Is that surprising? Not to anyone who knows the polling industry in Israel. Yes, it is an industry in every sense of the word. But it lacks a few important elements. It has not one iota of professional ethics, no permanent standard as to who can be a pollster, and it is wildly competitive - obedient only to the blind voracity of the media yearning for a survey. Any survey and any pollster will do.
In the series of studies I conducted over 13 election campaigns in Israel, from 1969 to the present, I examined the coverage of election-related opinion polls by the print and electronic media in Israel. The figures show that surveys are appearing with ever-increasing frequency in headlines, news articles and commentary. In the most recent campaigns, the passion for opinion polls has reached obsessive proportions: The space allocated to such polls in newspapers has grown by thousands of percentage points, as it has on TV news broadcasts. Opinion polls occupy a respectable place among the top stories on TV news shows.
A comparison of the content of opinion poll coverage over the years shows that reports on such surveys are very selective, sometimes suffering from a disregard for the boundaries of the poll's representative sample. Pollsters do not report the numbers of those who refused to answer the questions, which is believed to be as much as 50 percent of those sampled. In many cases, commentators talk about "changes" and "fluctuations" in public opinion based on minuscule differences - a few percentage points that do not statistically justify the conclusion that a significant change has occurred.
Evidence of pollsters' failures can be seen in the great variations among their forecasts: Polls carried out on the same day, asking the same question of a sampling meant to represent the same population lead to contradictory findings (sometimes hugely so ). Notably, most journalists and commentators are aware of the limitations, faults and failures of opinion polls, but chose to ignore them. They are enthralled by the ratings that opinion polls on the elections grab, poisoned by the intoxicating drug of predictions (most of which will turn out to be incorrect ).
Most Western countries have binding standards governing opinion polls, which professional polling organizations have adopted, including the American Association for Public Opinion Research and the World Association for Public Opinion Research. These standards, which present clear directives on how to report opinion polls, including those about elections, have not been adopted by pollsters in Israel. A few years back, the heads of communications departments at Israeli universities tried to institute a professional and ethical covenant with regard to polling in Israel. To this end, such covenants from other countries were translated and adapted for Israel. The result was presented to the heads of leading opinion poll firms but, unfortunately, consensus could not be reached, the covenant was rejected and the polling industry in Israel remained as wild as ever.
Public opinion polls are not without value. On the contrary, in the right hands and for the right goals they are an important and reliable research tool. But when they are used for the amusement of prediction, for wild speculation and baseless interpretations, opinion pollitis is dangerously toxic. Journalists and commentators infected with pollitis blame the public for the proliferation of public opinion polls, thus attempting to shirk their professional responsibility. But responsible journalists and professional commentators, editors and media leaders must reject this unbridled passion for polling.
The writer is a professor of communications at the University of Haifa.