Historians tell us that we are indebted to the Babylonians for coming up with a workable symbol for the concept of zero sometime about 300 B.C.
In India over the next few centuries, they refined the Babylonian idea of nothing to the math concept we know today. (Imagine being an Indian mathematician in those times, coming home exhausted from your labors, and your spouse asks what you worked on all day and you answer, “Nothing.”)
Zero is valuable to us in two ways. First, as a concept, it allows us to deal in a logical, procedural manner with the idea of nothingness.
More important from an economics standpoint, zero is extremely valuable as what mathematicians call a positional, or place device. By putting the other digits in their place, zero allows us to quickly tell the difference, for example, between 1 and 100, or 1,000.
Without zeros, numbers would be huge clusters of symbols, filling pages just to record a four-digit number – like a phone bill, only more so. The notational efficiency of zero allowed efficient processing of truly big numbers and thus paved the way for economic concepts such as the federal budget and the national debt.
As important and as valuable as the concept of zero is, we are still, after 2,300 years or so, trying to deal with some of its implications. One aspect of zero is even affecting our presidential campaign.
A considerable amount of effort and money is expended on opinion surveys and polls, and resources in even greater amounts are used to analyze, comment on, and, in some cases, spin the results. The quality of these polls, however, is deeply suspect, because their accuracy is being sabotaged – by zeros.
Political opinion polls came into their own after World War II. They were well known in the 1920s and 1930s, but their credibility was given a setback by a spectacular failure. In the infamous Literary Digest poll of 1936, because of a poorly designed sample, prospective voters strongly expressed their preference for Alf Landon as president. It was not to be.
In the private sector, opinion surveys supported product development and advertising from the 1920s on, but polling emerged as a major force in politics with the 1960 presidential race. At that same time, there was an underlying math problem emerging in opinion surveys – in both corporate and political polls. Each year, people were less willing to share their opinions with pollsters.
It didn’t matter whether the questions dealt with candidates or candy bars, the non-responders, or zeros, were a relentlessly growing portion of those contacted.
Pollsters and statisticians applied increasingly sophisticated analyses to the data, which probably helped, but there was still growing unease about all those zeros and what they meant to the validity of the poll results.
Current polls with the greatest visibility – those tracking voters preferences for President Bush and Sen. John Kerry – have been remarkably unchanging, and seemingly unaffected by news or other outside events. This has prompted some to believe that the people who talk with pollsters are mostly those who are already strongly committed to one candidate or the other. What the rest – the huge number of zeros, or non-responders – think about the candidates, there seems to be no way to tell until the votes are counted.
Corporations are in pretty much the same boat. They use opinion surveys to help them develop and launch new products, but because of the zeros, they don’t really know how valid the data is until the consumer votes, in actual dollar purchases, are tallied up.
And as if zeros weren’t enough to worry about, now hyperzeros are a growing problem for researchers and statisticians. Hyperzeros, or “placebo responders,” are those people in a study who actually get better even though they are given an inert pill, or placebo, instead of real medicine.
The proportion of hyperzeros in pharmaceutical studies is large and, what is so puzzling for analysts, growing. In some experiments on patients with depression, the placebo responders outnumbered those who felt better after taking the real medication.
The proliferation of zeros and hyperzeros has serious implications for our economy. Both presidential candidates have proposed expensive health care programs (Bush’s $90 billion and Kerry’s $650 billion over 10 years, according to economist Kenneth Thorpe) that are very dependent on containing drug costs.
If the growing number of hyperzeros in pharmaceutical studies pushes up drug costs, because it takes longer to sort out the real thing from the placebo, our economy will suffer. And that’s not nothing.
James McCusker is a Bothell economist, educator and consultant. He also writes “Business 101,” which appears monthly in The Snohomish County Business Journal.
Talk to us
> Give us your news tips.
> Send us a letter to the editor.
> More Herald contact information.