[13 November 2008]
Tracking polls, the poll of polls, exit polls—could Americans get through an election season without pollsters and their media conduits telling them what, at this very moment, collective “Americans” in general are thinking? Now, with the election over, it is only a matter of months before opinion polls begin revealing whether Americans approve or disapprove of President Obama’s first hundred days in office and their views on the economic plan he presents to Congress. But how much is really known about the accuracy of polls? And, if polls do distort public opinion, how do these inaccuracies influence the choices citizens and politicians make?
Perhaps these questions tugged at the back of your mind if you allowed yourself to be inundated by polling data during the recent election. Or, maybe you’ve always taken poll results at face value. Either way, The Opinion Makers: An Insider Exposes the Truth behind Polls by David W. Moore is an important, accessible and quick read that will change forever the way you interpret polls and the credence you give them.
Moore is a former senior editor at the Gallop Poll and a political science professor. He presents his argument like a college lecture, using insightful case studies, impeccable logic and illustrative graphs to drive home his argument that pollsters manufacture public opinion by failing to report on “widespread public apathy, indecision and ignorance”.
Herewith, my lecture notes:
Polling problem #1: Pollsters fail to ask people how strongly they hold to their opinions.
Polling problem #2: Pollsters fail to report on the public’s indecision; instead their questions force people to make a choice.
Polling problem #3: Pollsters assume that everyone has enough background knowledge to answer a question, even on complex policy issues.
All of these things skew poll results.
I never was one for taking lengthy notes.
Moore provides compelling examples of these problems and, most significantly, highlights the danger they pose to the way American democracy operates. To demonstrate problem #1, Moore flashes back to the weeks before the Iraq War, when Gallup reported that Americans favored going to war, 59% to 38%. However, a follow-up question found that only 29% of Americans supported going to war and said they would be upset if the war did not happen; 30% opposed the war and said they would be upset if the war went forward. A full 38% of people said they would not be upset if the government did the exact opposite of their preference.
While the first set of statistics paint a picture of a war-mongering US public, the second poll results show a public willing to let their leaders make a decision. Pollsters rarely ask the follow-up question about the intensity of people’s opinions, and the media rarely reports on it. But could measuring and publicizing how tightly people hold to their expressed views give American politicians a better sense of the public pulse? Could it have made Bush at least tap the brakes during his headlong rush into war?
Polls’ tendency to understate the number of people who are undecided is Moore’s favorite bailiwick. The standard question employed during campaign season is: If the election were held today, would you vote for Candidate X or Candidate Y? “I don’t know” is not presented as an option, and few people take the initiative to tell the pollster that they’re undecided. They have been asked to make a choice, and they do. Even if respondents admit to being undecided, pollsters then often ask which way they “lean”. Moore describes how this phenomenon can make a candidate appear to have more support than they actually do, especially early in campaigns when candidates with greater name recognition are a more popular top-of-mind response for people forced to make a choice. The media begins ignoring lesser-known candidates based on their poor polling results, before the public even gets to know them.
Polls have the potential to enhance democracy by ensuring the people’s voice is heard clearly and regularly. But the media prefers reporting poll results that demonstrate preferences, not ambiguity and indecision, and pollsters provide what is asked. This means that polls have the potential to misrepresent what people truly think and what they really care about.
The good news is that the problems Moore highlights can be addressed by ensuring pollsters ask respondents questions that more accurately capture public apathy, indecision and ignorance. The power is in pollsters’ hands, and in ours. So read this book and then, the next time the media reports on a potentially misleading poll, call or write to let them know that you see through their effort to make it seem that everyone in America was informed, engaged and decided on the issue. Tell them you’ve studied Moore’s book and now expect, well, more.