The Opinion Makers: An Insider Exposes the Truth Behind the Polls Hardcover – September 1, 2008
Enter your mobile number or email address below and we'll send you a link to download the free Kindle App. Then you can start reading Kindle books on your smartphone, tablet, or computer - no Kindle device required.
To get the free app, enter your mobile phone number.
Frequently bought together
Customers who bought this item also bought
From Publishers Weekly
Copyright © Reed Business Information, a division of Reed Elsevier Inc. All rights reserved.
"The next time your phone rings with questions from a pollster, beware. David Moore rings an alarm bell that democracy is endangered by the way the news media use public opinion polls. In chapter and verse, he exposes how false and misleading polling practices actually create public opinion and this, in turn, influences what government does. The Opinion Makers demonstrates what James Madison said two hundred years ago-a misinformed public becomes a threat to democracy."—Ben H. Bagdikian, author of The New Media Monopoly
"The account of how news stories drive polls should make us stop and ask whether the close relationship between the newsroom and polling operations is perhaps a bit too close. A must read."—W. Lance Bennett, director, Center for Communication and Civic Engagement University of Washington, Seattle
"We all know that the corporate press conducts its own opinion polls and keeps headlining the results as if such stuff were news. What we don't know is just how sloppy-and misleading-most of that work really is. In this important book, veteran pollster David Moore uses many harrowing examples from the recent past to meticulously note the many defects in such polling."—Mark Crispin Miller, Professor of Media, Culture, and Communication at
New York University, and author of Fooled Again: The Real Case for Electoral Reform
There was a problem filtering reviews right now. Please try again later.
The author, a former professor of political science and a former vice president of one of the nation's largest and most well known polling operations explains, in great detail, the problems with polling as it is done today. In addition, he explains why the media is so cozy with the polling groups and how this tends to cause massive distortion in the numbers. Finally, along the way, the history of "scientific" polling is discussed from it's formation to the current day.
The author writes in an easy to understand style that takes a fairly complex issue and digests it into material that any reader will be able to read. Far from dull, the author has a biting edge to his writing and is passionate about the subject. In addition, the author manages to keep the information neutral, so he isn't pointing fingers at any one political side; he manages to have enough on everyone to go around. This should be read by every person who plans to vote and follows the polls. You will never look at the information in the same way again.
I used to derive much comfort in the very low + or - 3% error margin of polls. But, Moore's book brilliantly illustrates why the error margin is a delusion. This is the case for several critical qualitative reasons. The first one is what he calls "forced choice." Pollsters want to generate definite opinions from the public. But, the public is often ignorant or undecided. Yet, the choice is binomial (yes/no) are you in favor of something or not. As an example, a poll may tell us 62% are in favor of something with 35% against it and only 3% undecided (the few who dare say they did not know anything about the issue). Occasionally, Gallup has redone such polls asking the public first if they knew something about the relevant issue. Using this second method, on the exact same issue they would get this kind of results: in favor (19%), against (25%), and unsure (56%). That's a completely different result then the first poll.
The virtual national primaries are an extension of the "forced choice" fallacy. He thinks national polls of Presidential candidates taken a full year before an election are meaningless. Typically, only a small percent of the public express they are unsure because it is not even mentioned as a possible answer. If asked first if they had a defined choice about Presidential candidates, we'd see that the majority of the public one year before an election is undecided. This is how pollsters got surprised and stated that in early 2008 Huckabee in the Republican Party came out of nowhere. That's because unlike what polls were stating the vast majority of voters were undecided and a significant % went to Huckabee.
Lack of recency can be another major flaw. Moore refers to the most spectacular failure of political polls when they all predicted the victory of Dewey over Truman in 1948. Moore refers to the other major gaffe when they predicted an Obama victory over Hillary in the 2008 New Hampshire primary. Moore stated that in both cases the pollsters stopped polling too early. In 1948, they stopped polling a full two weeks before the election while Truman was picking rapid momentum among the undecided (that were not well tracked because of the forced choice format). In 2008, they stopped two days before the vote. During those two days, Hillary showed emotions on TV and rose among the women voters. Moore indicated that if not for the lack of recency, the mentioned polls in 1948 and 2008 would have gotten it right. If polls are taken just before an election most people have truly made their minds and that's the only time when "forced choice" format works.
Moore mentions many other interesting flaws of polls. Most polls on issues are conducted with lead questions framing the issue. Those lead questions dictate the direction of the ultimate poll answer. Regarding polls on the opening of the ANWR to oil drilling. If framed as an environmental issue the majority is against it. If framed as an energy independence issue the majority is for it. And, if asked what ANWR stands for the majority does not even know. So, what do those polls tell you? Nothing, but the Media does influence our opinion by conveying such polls as the legitimate opinion of the entire nation. Wording can have a key influence on outcome. Minute difference in the wording can influence the direction of the poll. Sequence of questions asked and even sequence of possible answers influence outcome too. On a phone survey, respondents will choose answer B more often than A simply because they remember answer B better. They ran experiments in reversing the order of answer A and B and got opposite outcome.
Change in technology is causing havoc with the polls. People don't respond to polls on phone landlines as much as they used to because of answering machines and cell phones. Reaching cell phone users is difficult. Internet polling is emerging but it has many flaws. Even if you weight results for ethnicity, gender, age, and income, internet users are different than the general population. They are typically more opinioned, informed, and conservative. Studies found that Zogby and Harris internet polls were the worst. Their error in % vs the actual outcome was twice the one of traditional phone polls (even their own).
Pollsters are currently writing some of the most interesting books on social science and politics. Along those lines I can strongly recommend Microtrends: The Small Forces Behind Tomorrow's Big Changes and The Way We'll Be: The Zogby Report on the Transformation of the American Dream .
But Hillary Clinton proved the polls wrong and won in New Hampshire. The fight for the Democratic nomination was far from over. Months of fierce battling lay ahead before the final states voted and Obama outlasted Clinton to win, barely.
All the New Hampshire polling was way off. What went wrong? After all, the New Hampshire primary has been the first in the nation for a while now. Surely polling scientists should be able to work their prediction magic in this small, well-studied state. At least be able to pick the winner, if not the winning margin. Or call the election too close to call. But the polls were not even close to the voting results.
David Moore tells us why polls cannot predict well. Several reasons lie behind that. Pollsters pressure people into picking, even when they are undecided. Events after polling can have a big impact, like Clinton's emotional coffee shop comments. People's answers are influenced by how questions are posed. All these things affect the predictive power of polling.
But others have also written about polling accuracy. Moore goes beyond that to talk about more troubling problems. He thinks that politicians and the media use polling to make public opinion, as much as to measure it. He says that pollsters tell Americans what they think, rather than the other way around. That, according to Moore, damages democracy.
Moore's crisp and persuasive writing gives his views power. Though polling and statistics can be boring, The Opinion Makers is not. This past presidential election proves that the problems with polling are not going away. They are getting worse.
We should heed Moore's warning.