TORTURING THE NUMBERS

Published on .

In his May 30 letter to the editor, David Buchanan stated that Steven Frankfurt [in an earlier letter] exhibited a disregard for the facts and that the Partnership for a Drug-Free America campaign to reduce drug abuse does more harm than good.

He supports his argument by stating that between 1981 and 1986 (before the PDFA campaign) regular drug use among high school seniors declined an average of 1.96% per year. He went on to point out that between 1987 and 1992, during the PDFA campaign, regular use of illicit drugs declined an average of 2.06% per year. The difference between 2.06% per year and 1.96% per year was deemed insignificant.

At least two problems are readily apparent with this logic. First, the 1981 to 1986 decrease from 36.9% to 27.1% represents a 26.6% decline in regular drug users, while the 1987 to 1992 decline from 24.7% to 14.4% represents a 41.7% decline in regular drug users. This difference is most certainly significant. Second, it can not be assumed that the rate of decrease per year would/should be the same. Just as every marketer knows that it often becomes increasingly difficult to capture more and more market share, it would seem logical that it would be more difficult to lower the percent of drug users as the base becomes smaller because you may be down to the "hardcore" users.

As an econometrics professor once told me while attending grad school at the University of Washington, "If you torture the numbers enough, they'll confess." Consider: if we ban the eating of tomatoes almost all murders will cease because 99.3% of all murderers have eaten tomatoes.

I won't pretend to have enough information to determine what effect the PDFA campaign has actually had, but as a former undercover officer I must agree with William Gray III (AA, May 2): "It takes consistent, repeated messages as well as the actions of committed individuals and organizations to make social change."

Paul D. Swenson

Director, ESP Marketing

Circuit City Stores

Richmond, Va.

In this article:
Most Popular