Vol. 3, No. 3
Depending on statistics:
With increasing reliance on statistics to provide information about trends and to forecast future indicators, the old issue of "lies, damn lies, and statistics" emerges as a genuine consideration. When should you trust statistics?
As Dan Ackman writes in Forbes.com this month, the media continues to sound the alarm about a failing U.S. economy, citing manufacturing slowdowns, declining consumer confidence, increased unemployment, and of course layoffs. All of these indicators can be demonstrated with sound data, right? Not so fast, as Ackman points out. Or--consider the source.
The National Association of Purchasing Managers, for example, creates some of the most widely reported economic data. But Ackman says that both the NAPM and the Conference Board, another purveyor of data, tend to pay little attention to the reliability of the data they provide. Even official government statistics are subject to sampling error, he points out. The Bureau of Labor Statistics, for example, reported that unemployment had gone from 4.0% to 4.2%. But total unemployment remained constant at 136.0 million. The rise in unemployment was attributable completely to the size of the labor force, which grew by 466,000 to 142 million. Participation in the labor force, according to Gary Steinberg, a spokesman for the BLS, can be affected by discouraged workers now seeking work, graduating students seeking first-time employment, and retirees who return to work--none of which is necessarily bad news.
"Consumer confidence" is another often-used indicator of the health of the economy. The report is based on The Conference Board's survey reveals how consumers are feeling about the economy at present and their expectations about the economy six months from now. One might ask, as Ackman does in the Forbes.com article, why a survey is used to predict, rather than simply looking at hard data that reflects actual consumer spending.
As Ackman points out, "Because confidence is subjective, it is more susceptible to change than reality. In this way, confidence is akin to the stock market, whose twists and turns provide a new story and a new opportunity for prognostication every day, perhaps even every hour."
To read the text of Ackman's article, go to http://www.Forbes.com. In the meantime, be wary of definitive interpretations of data that may by specious either in its origin, its history of reliability, or its sampling methods.
Copyright 2001 PQ Systems.
Please direct questions or problems regarding this web site to the Webmaster.