The One-Percent Solution

In his continuing effort to outdo the Republicans as a drug warrior, Sen. Joseph Biden (D–Del.) released a Senate Judiciary Committee report in May purporting to show that 1 in 100 Americans is a "hard-core cocaine addict."

The startling finding, which Biden touted as two-and-a-half times the Bush administration's estimate, made a big splash in the news media. A closer look at the report, however, reveals a number of unjustified assumptions.

First, the committee's survey applies the label hard-core cocaine addict to several different groups, using inconsistent criteria. The groups include cocaine users who entered drug treatment during 1989; respondents who reported using cocaine once a week or more frequently in the 1988 National Household Survey on Drug Abuse; and arrestees with traces of cocaine in their urine. The report adds 55,000 homeless addicts, based on an unsupported estimate of 2.1 million homeless; it further assumes that a homeless person is five times as likely as someone living in a household to abuse cocaine.

Second, the study uses shaky methods to measure the number of cocaine-using arrestees, who represent most of the newly discovered "addicts." It employs the results of urinalysis tests from the nation's 20 largest cities to estimate the number of cocaine-using arrestees in 44 other cities that were not surveyed. It applies arbitrary proportions to smaller towns and rural and suburban areas. Moreover, the study assumes that 80 percent of cocaine-using arrestees are addicts.

Third, the state-by-state tallies of addicts offered by the committee report are spurious. They are based on the assumption that the 862,000 people identified as weekly cocaine users in the National Household Survey on Drug Abuse were geographically distributed in precisely the same way as the users who had been arrested or who had entered treatment.

The report goes through the motions of eliminating overlap among the four sets of figures on which the study relies—each of which is derived by a different method and based on a different definition of addiction. But because they require wild conjecture, these efforts would be inadequate even if the various numbers referred to the same phenomenon.

So was the figure of 2.2 million cocaine addicts pulled out of thin air? Not exactly, but it might as well have been.