Monday, August 29, 2005

Matt Taibbi: Problems With Our Polling Culture

48% for Taco Bell; 52% for Invading Iraq
By Matt Taibbi
Originally published in New York Press; repuplished at AlterNet

Did you know that polling is illegal in some countries? In Russia, published polls are not allowed before an election; the same is true in Nicaragua. In Belarus, polls are illegal in general -- but then again, so is everything else. Still, how interesting!

I think we take our survey freedoms for granted. Nothing else can explain the appallingly low quality of our polling. Polling in this country has degenerated almost entirely into a tool for describing consumer behavior, where the goal of almost every well-funded survey is to make a numerical determination about the strength of X product vs. Y product in the general marketplace.

The brand names might be Taco Bell and Jack in the Box, they might be Democrats and Republicans; the methodology is, to a degree at once damning and hilarious, exactly the same. Take a look at the press releases for two of the top two polls conducted by Zogby last week:


1. Coke Is It: Americans Choose Coca Cola over Pepsi by 47% to 28%; 'Real Thing' Leads Every Demographic; 'Choice of a New Generation' Unpopular With Younger Consumers -- New Zogby Consumer Profile Finding


2. No Bounce: Bush Job Approval Unchanged by War Speech; Question on Impeachment Shows Polarization of Nation; Americans Tired of Divisiveness in Congress -- Want Bi-Partisan Solutions -- New Zogby Poll

The degree to which polling methodology reflects the bias of the interested (and usually commissioning) parties is seldom noted when the polls are cited by reporters. For instance, pre-election polls are almost always presented in their, final, less embarrassing, airbrushed form -- e.g., 51 percent for Bush, 49 percent for Kerry -- when the actual numbers are more like 26-24 percent, if you include nonvoters.

Respondents, when quizzed, about, say, their favorite fast-food restaurant, are never asked the obvious cross-reference questions. Thus you never see press releases that read like this: "74 percent of Americans who cannot climb two flights of stairs without gasping for breath said that McDonald's was their favorite fast-food destination, while a surprising 47 percent of respondents who 'expect to be dead within weeks' said that the Wendy's Big Classic was their 'favorite sandwich.'"

Our prominent polling agencies almost never take it upon themselves to actually pose a new question. Instead, they almost always content themselves with recording the answers to a question that in some very public way has already been asked -- usually in the form of a choice presented by the media. Do you prefer Friends to Seinfeld? Is Michael Jackson guilty or innocent? Are you for or against the invasion of Iraq?

To Read the Entire Essay

No comments: