Americans and Polls

Please take a moment to answer these few simple questions, which will be used in the drafting of a constitutional statement that may be presented to King George III. Your feedback will be helpful to the persons framing this document, who wish to include a variety of opinions and viewpoints in the text.

PART ONE: Do you hold these truths to be self-evident, yes or no?

1. All men are created equal 2. All men are endowed with certain unalienable rights 3. Governments are instituted to secure these rights by popular consent 4. A government which fails to secure these rights should be altered or deposed 5. It’s okay to start a new government along safer and happier lines

Let’s take a moment to discuss your views on specific points. In the group of persons cited above, “all men”, would you include:

a) White males with property only b) White males c) White males and white women d) Everybody except slaves

Tell us how you feel about the following rights on a scale of one to five, one being ‘negligible’ and five being ‘unalienable’.

a) Life b) Liberty c) The Pursuit of Happiness d) Free Checking

Our cousins across the Atlantic want to know why Americans continue to rate George W. Bush so favorably in the polls. In general one can discount peculiar tastes as a matter of culture. The French, for example, eat frogs; in Spain the women cultivate moustaches. But cultural degeneration can’t explain our fondness for Incurious George. Recently I was asked this question by a friend of mine from England (actually I hate his guts but I continue to answer his calls because when I hang up the phone I can turn to my guests and say, “Oh, that was just a friend of mine from England”), and I rather glibly told him I didn’t have any idea, nobody I knew liked Bush. Which is true. But this sort of answer evades the point. Of course nobody I know likes Bush; I hang out with a bunch of black Floridian Jews, most of them ex-cons. So who are all these other people the pollsters are talking to? What I needed was a poll of my own.

Opportunity knocked, and I answered the door in a filmy negligee. Through no fault of my own, I was enlisted to drive across country. The next time I get enlisted it may be to fight in Iraq, so I agreed, and conducted my poll along the way.

Las Vegas, Nevada, a gas station attendant glances furtively around in case the FBI is listening, then mutters: “Don’t even get me started”. Provo, Utah, a server at the Aloha Shave Ice: “I’m taking pills for my nerves.” An aged one-legged antiques dealer in Nebraska: “Bush isn’t in charge anyway, it’s Cheney, and he’s a pirate.” From West to East, I could find nobody who supported the President, and I went out of my way to ask real hombres in pickup trucks and big belt buckles. There was a guy at the Quik-Lube in Mentor, Ohio who said Bush was okay, he guessed– but he thought I was an undercover Fed because of my sunglasses and plaid shorts, and so may not have been entirely forthcoming. Be that as it may, and even assuming a margin of error of 60% or so, Bush still came in a far second to Benito Mussolini. So rest assured, my European brethren. It’s not us Americans. It’s the polls.

The course of our nation is now guided by polls. If this had been true a couple of centuries ago, the United States of America would never have gotten past the questionnaire stage – you can’t build a nation by building consensus. If Moses had done poll research before he got the tablets engraved for Shavuot, the Ten Commandments might have included “Thou Shalt Not Leave The Seat Up” . Polls are fascinating things, not least because anybody believes them. Just as one doesn’t base a military campaign on Michael Jackson’s advice just because he dresses like a naval officer, one shouldn’t base national policy on poll results just because they’re dressed up like public opinion.

Let us be clear: there are different kinds of data gathered on the public, all of them wrong; but some are at least defensible, which public opinion polls (which is what is generally meant by ‘polls’ in national discourse) are not. The United States Census is a statistical poll, but not an opinion poll- unless people are lying, which they often do when a federal employee shows up at their house unannounced. When the Census Lady arrived at my door, I told her I was African American. She just waited, surveying my golden curls and alabaster complexion. I tried Samoan. There was a look of infinite sadness in her eyes as she wrote down ‘white’ and went on to the next question. I successfully lied about my age.

Other types of statistical poll data are gathered from commercial or institutional sources. How many billion hamburgers do Americans eat every year? How many cases of morbid obesity squeeze through the hospital doors in the same period? These questions are answered simply by gathering the numbers from the source. The numbers are always jiggered to reflect what the industry or institution wants the world to know, but you can get a general idea that won’t be far off. The key here is that statistical polls, unlike opinion polls, are aiming at data that can be proven. If I wanted to, I could go out there and count the dead smokers myself. (I’m not offering, it was just an example.) But how many of them believed smoking was harmful to their health? That’s an opinion, and the poll results will be wrong– if only because it’s hard to hear what somebody with no larynx has to say. Stochastic data, which are what clever people call statistics, aren’t of much use; witness the celebrated example of a person with one foot in a bucket of ice and the other foot in a bucket of boiling water who is, on average, warm. Now take the statistical form and apply it to opinion polls, which is what the government relies upon to take the public pulse: you can see why the patient is dying.

The preface to the Declaration of Independence suggests in no uncertain terms that “Governments are instituted among Men, deriving their just powers from the consent of the governed” for the purpose of securing certain unalienable rights. “The consent of the governed” is ordinarily secured by electing representatives whose job it is to ruin everything according to the will of their constituency. The problem with government by poll is that it short-circuits the democratic process. A poll is a kind of ballot, with the matter to be voted upon chosen not by elected officials pretending to reflect the concerns of their constituency, but rather by the entity paying for the poll. There are a great many important questions that don’t get asked in polls; the spectrum of debate is limited by the spectrum of poll questions. And although polls are generally random, for economy’s sake they represent a much smaller sample of people than census-type data, typically fewer than 2,000 respondents out of 200 million eligible voters (hey! A statistic), or 1/1000 of the demographic- so there is a much higher risk of bias in the results.

If, for example, the pollsters conduct their survey by telephone, then people without telephones are excluded from the poll. They may also live in caves, but that’s beside the point. If the poll is conducted during early evening hours (as they usually are, so the pollsters can catch the soaps during the day) the sample is further reduced to people at home when the phone rings. So people with two jobs won’t get asked. I think I caught several of them on my route. Nearly half of American households (wow! Another statistic) have caller I.D.- chances are, these folks won’t pick up the phone for what looks like a telemarketing call, so they’re out of the sample. And unless the pollsters use computer-generated ‘random-digit dial’, they can only call listed numbers. I’m guessing they talk to a lot of old people.

Internet polls are worse, and they’re gaining popularity. The problem here is obvious: not only is the Internet poll restricted to Internet users, it’s also self-selecting. Only people who favor the ideology of the site conducting the poll will participate. Plus only a real wiener would bother. You’ll get better results from a stoned gibbon with an Ouija board.

But even assuming a wide-awake pollster could get a truly representative sample of eligible voters, the fun has only just begun. Let’s have a look at some of the questions these precocious pollsters are pitching. Here’s the Harris Poll, one of the biggest names in the field, asking that pesky question that has our European pals in a lather:

“How would you rate the overall job President George W. Bush is doing as president: excellent, pretty good, only fair, or poor?”

Got that? Excellent, pretty good, only fair, or poor. A,B,C,D, to use the dreaded school grading system. Bush eked out a C average in school. C+ is average. So that breakdown works well enough, although “only fair” is a rather suggestive category- why not just “fair”? Doesn’t ‘only fair’ sound worse than ‘fair’? Still, we’ll set that aside for now, preferably somewhere dark and moist, and pretend the categories are reasonably unbiased. A,B,C,D. But Harris doesn’t break the participant responses down according to these four categories. Instead, they are grouped as follows: “excellent/pretty good”, “only fair/poor”, and “not sure/refused”. Strewth! All of a sudden the poll is skewed worse than Dick Cheney’s snarl. Excellent and pretty good are not the same thing, but they poll the same. What about ‘fair’ and ‘poor’- the cataclysmic difference between a C and a D in school, but they poll the same. And there’s a whole category for non-answers. Suddenly the poll favors Bush- because the average milquetoast respondent who would say Bush is doing “okay, I guess” (C) doesn’t want to say “only fair” because that sounds more negative (C-), so they say “pretty good” (C+), which doesn’t mean Bush is doing great- he’s just getting by. But “pretty good” and “excellent” rate the same, so Bush gets bumped up from a C+ to an A+. Which is typical of his luck- a special Bell curve just for him. Ask not for whom the Bell polls. It polls for you.

And then there are the actual questions. Here’s a Fox News/ Opinion Dynamics poll question:

“On the issue of abortion, would you say you are more pro-life or more pro-choice?”

Here’s the same question, from Newsweek via Princeton Survey Research Associates:

“Which side of the political debate on abortion do you sympathize with more: the right-to-life movement that believes abortion is the taking of human life and should be outlawed, OR, the pro-choice movement that believes a woman has the right to choose what happens to her body, including the right to decide to have an abortion?”

In the first example from Fox, 47% of respondents were pro-choice, 41% pro-life- almost a dead heat, in statistical terms- they don’t cite the margin of error. In the second example from Newsweek, which carefully defines the terms, 51% of respondents were pro-choice in sympathy, and 39% were pro-life; but the terms ‘pro-choice’ and ‘pro-life’ are themselves leading: they suggest value judgements. Who, after all, is anti-life? If the question is posed in an impersonal, practical way, not based on sympathy, the numbers shift dramatically, as seen here from ABC News/Washington Post:

When should abortion be legal? Put it this way, 59% say in all or most cases it should be legal, 39% say it should be illegal. Which is not the result Fox was looking for. Damn liberal media. But for a policy maker, this version of the question is the one to look at- otherwise, legislation of vital interest to the nation is based upon the wrong end of the issue: if you ask people what they personally feel about abortions, the pro- and anti- camps are almost evenly divided. If you ask people if it should be legal to have one, the majority says yes. It’s like euthanizing dentists- most tooth-bearing adults would agree this is a good idea. But if you actually suggested doing it, most of these same people would disagree- you never know, the next periapical abscess could be your own.

In the end, even the most impartially worded poll, based upon the replies of the most perfectly representative cross-section of the population, still ain’t worth a hill of goose dooty once the poll data are recontextualized by that many-tongued hydra, the Media. For example, here is a genuine statistical fact: 49% of Americans are below average, while 49% are above average. Here is the same statistic stated in two different ways:

Nearly half of all Americans are above average.

Less than half of Americans are above average.

Here we see the difference between an American newspaper and a European newspaper, assuming the European one is in English- otherwise, who knows what they’re saying? You see the problem: it is an unshakeable statistical truth that if 50% is average (C+), the 49% below that is below average. This is why statisticians only hang around with other statisticians. But hand that fact to two different journalists (‘journalist’ comes from the Latin word ‘urinal’, meaning ‘hogwash’) and you get two different pieces of news- one good, one bad. Your Congressman reads one of these pieces of news– and bases his policy decisions on it. Or rather an aide reads one, mentions it to your Congressman over morning drinks, and your Congressman does whatever his corporate sponsors tell him to do; but he spins his actions according to the spin he got from the Media on the data. And this is what happens to a statistical fact– now imagine what happens to poll questions, which are spun by the Media only after they are pre-spun by the way the poll question is worded and re-spun by the way the answers are tabulated. You pitch that kind of spin on a baseball, the batter will never see it– but it may hit the second baseman in the head.

Let us return to our original issue, odious as it may be: the popularity of King George (the Harvard man, not the Hanoverian). Although his poll numbers of late have been deflating like a cheesecloth balloon, he still remains fairly popular (B+). Even adjusting for the layers of spin applied to the already biased polls, which might reduce his popularity rating by half, what accounts for the remaining stalwarts who still think he’s pretty good, all things considered? The rest of the world except Tony Blair agrees: the administration of Bush II is the worst in modern history. How can even 30% of Americans support such a disaster? Here I must educate our European comrades on one final variable in American polling: the character of the average American.

Ask the French citizen what he or she thinks on any issue, and the Gallic tendency is to give a sneering reply. This comes of reading too much Andr? Gide while eating gastropods. Germans will tend to express rigidly conformist, fatalistic opinions, because they lost two World Wars in one century and they think it’s a trick question to see if they’re planning another one. Now ask an American. Americans are famous for their solicitousness to a stranger in need- if you doubt this, just ask an American for directions to the nearest gas station. Half the time he’ll drive you there. And Americans don’t like to speak badly of another individual around someone they don’t know (although we will really cut loose among friends). Americans leave the criticism unspoken, hinted at by tone of voice.

So that pollster calls in the mellow after-dinner hour, and when Joe American answers the phone, he doesn’t want to talk merde about anybody. What does he think of George W. Bush’s job performance? “Well. . . okay, I guess. Good enough. Bush is doing the best he can. . .” The unspoken part, “. . .For a goddamn fool,” doesn’t show up in the final poll results.

Ben Tripp is a screenwriter. His last story for CounterPunch was the enormously popular Jesus v. George.

Tripp can be reached at: CounterPunch Special Report: 9/11 One Year After

Bill Christison A Year Later: It’s Happening Here

Alexander Cockburn The Tenth Crusade

Susan Davis Mr. Ashcroft’s Neighborhood

Bruce Jackson When War Came Home

David Krieger Looking Back on September 11

Mike Leon Bush and War

Peter Linebaugh Levellers and 9/11

William McDougal September 11 One Year On: That’s Entertainment!

Riad Z. Abdelkarim and Jason Erb How American Muslims Really Responded to 9/11

Jeffrey St. Clair The Trouble with Normal

Tom Stephens Rise Up…Dump Bush

[CDATA[ $('input[type="radio"]
[CDATA[ $('input[type="radio"]
[CDATA[ $('input[type="radio"]
[CDATA[ $('input[type="radio"]