October, 2003. The place: the cavernous underground hallways that connect the major administrative buildings in DC. It's here, that the pollsters par excellence hunch over their bratwurst and engage in conversations that threaten the very balance of power of the republic. Everywhere you look, wild packs of cardigan wearing demographics experts are loitering around, peppering their conversation with inane slogans (e.g., "preliminary evaluation to date reveals it's not the tax plan. It's the proclivity for slobbering, stupid," "we gotta saturate the airwaves and go for the throat" and "trust me babe.")
Election cycle must be kicking in.
After a while, you actually start to feel sorry for the pollsters. Professional mandate requires them to lock their own opinions away in the dark recesses of their subconscious. They exist to give life to your ideas. They compile your opinions, and suddenly your opinions exist. They are ruthlessly dedicated to the other-other's thoughts, other's ideas. Their own "selves" are constantly undermined by the requirements of the job. It's a socially shrunken existence. And this time of year you can't throw a stone without hitting one of these creatures. (note to reader: if you throw it, throw it hard)
In any regard, this seems an appropriate time to ask, do polls have genuine meaning? Should they merit our consideration? Are they biased? (After all, pollsters are paid by a client with an agenda. Does this provide some room for subjectivity to sneak in?
"Of course polls can be biased," admits Jon McHenry, a pollster with the Northern Virginia polling company, Ayres, McHenry & Associates, Inc. "It¹s a lot easier to write questions that present only the point of view of your own side than it is to put careful consideration into a balanced question that includes the best arguments of both sides." Mc Henry tacks on an addendum about locking away personal biases in order to provide clients with objective survey material. Either way, it's still the pollsters who give meaning to the political landscape. They distill incredibly complex issues into easily digestible numbers that spare us the burden of closely examining the world around us. In the extreme Clitnonesque example, it's the poll numbers that decide the political decisions that have the greatest impact on our lives. The representative democracy wags the politician. The politician is reduced to virtual leader. (Could this explain our enduring affection for President Clinton? That he was such a willing and jovial receptacle for our opinions?)
It begs the question, just how accurate are these polls? My left brained rational friends assure me that through a complex system of formulations and calculus, polls become amazingly accurate barometers of public opinion. Just one thing: polls feature the opinions of almost every variety of citizens, with one glaring exception-people who tend not to be home in the evenings. Pollster can only call between certain hours (typically between 5-9 p.m., Mon-Thurs.). There is also a growing segment of our society that flatly refuses to respond to telephone polls. For those who do respond, there is really no guarantee that they will actually vote. Therein lies the rub: polls are, at bottom, based on odds and projections. Or, as Economist A. Marshall put it, "Statistics are [merely] the straw out of which...all economists have to make the bricks."
An example: until a recent Supreme Court ruling, the Census Bureau had always counted the U.S. population by mailing questionnaires. Despite their best efforts, the bureau missed some three or four MILLION people in the 1990 census. Not to worry. In the future, the census bureau will account for such oversights by inventing people. I believe the technical term for inventing people is "limited statistical sampling," or using sample groups to project larger trends. These projections would then be the basis for, among other things, distributing $180 billion in federal funds for minorities and the needy. In other words, the pollsters
project out these millions of made up people and we reincorporate the numbers as reality. It's virtual democracy.
So, what is the lesson in all this? Mark Twain summed it up succinctly: "Statistics are like ladies of the night. Once you get them down you can do anything with them." I ran that quote by several top notch demographic experts and they can't completely rule out its veracity. But given the public clamor for public polls and the purpose they serve in giving order to our surroundings, they are likely here to stay.
Meanwhile, our representative democracy becomes increasingly less so.