Student drug testing not effective in reducing drug use

ANN ARBOR, Mich.—Drug testing of students in schools does not deter drug use, University of Michigan researchers have concluded, based on a large, multi-year national sample of the nation's high schools and middle schools.

The findings were reported recently in the Journal of School Health. The research challenges the premise that has been central to the rationale for schools adopting a drug testing policy. The contention that testing is a deterrent to drug use also was an important consideration in a recent split decision by the United States Supreme Court that upheld the constitutionality of drug testing of students as a condition of participating in extracurricular activities.

The authors looked at the combined data from surveys done in 1998, 1999, 2000 and 2001, representing information from 722 secondary schools from across the nation, including 497 high schools and 225 middle schools. Questionnaires were administered to one school administrator (usually the principal) in each of these schools to determine, among other things, the drug testing policies of the school.

Self-administered questionnaires were also given to students in one grade in each school (8th, 10th, or 12th grade), which determined whether and what drugs they might be using.

The survey represents the only large or nationally representative samples of schools that have ever been used to assess the effectiveness of drug testing policy.

The Effects of Testing
At each grade level studied—8, 10, and 12—the investigators found virtually identical rates of drug use in the schools that have drug testing and the schools that do not. For example, in 12th grade, 36percent of those in non-testing schools reported having used marijuana in the twelve months prior to the survey, versus 37 percent in the schools that did test. The measures of drug use examined were the prevalence and frequency of use of marijuana in the prior twelve months and the prevalence and frequency of use of illicit drugs other than marijuana over the same period. (Prevalence refers to the percent reporting any use in the 12-month period.)

Additional analyses focusing on specific groups of students also were conducted. In those high schools that tested athletes, use by male athletes of marijuana (or of any other illicit drugs) was not significantly different from use among male athletes in the great majority of high schools that do not test their athletes. (There were not sufficient numbers of female athletes available to conduct parallel analyses for them.)

The investigators even looked separately at established marijuana users, defined as those students reporting use of marijuana on twenty or more occasions in their lifetime, to see if any testing effects might be visible among heavier users. But even in this group, the rates of use of marijuana and other drugs in the prior year were virtually identical in schools that conduct drug tests compared with those that do not. Some 94percent of both groups indicated using marijuana in the prior twelve months.

Subsequent to the submission of the journal article, the authors conducted additional analyses in which data from the 2002 national sample were added to those already reported for 1998-2001. This added 169 schools, bringing the overall school sample to 891. All of the findings from the earlier samples continued to hold in the enlarged and updated sample. The investigators also looked specifically at the possibility that random drug testing, in which the entire student body is subject to being tested, might yield discernable effects. Only seven schools out of the 891 surveyed reported having such a policy—all of them high schools. No statistically significant difference was found in student use of marijuana or other illicit drugs between these seven schools and the great majority of high schools that did not have random testing. After controlling for the types of students served and the grade level of the students surveyed, there was virtually no difference in drug use rates between those schools that had random testing and those that did not.

The authors—Ryoko Yamaguchi, Lloyd D. Johnston, and Patrick M. O'Malley—are all social scientists at the University of Michigan's Institute for Social Research.

"We think that one reason that so few schools test their students for drugs is that is an expensive undertaking," comments Johnston. "Schools are very pressed for funds, and I would say that the results of our investigation raise a serious question of whether drug testing is a wise investment of their scarce resources. It's also very controversial with a lot of parents and students," he adds.

"The way that drug testing has been carried out in the schools looks very unpromising. I have no doubt that one could design a drug testing program that could deter teen drug use, but at what monetary cost and at what cost in terms of intrusion into the privacy of our young people?"

The researchers found that about 19 percent of American secondary schools have some form of student drug testing. The group of students most commonly tested are those identified "for cause," that is, based on evidence or suspicion that they had been using an illicit drug—14 percent of all secondary schools report this practice. Many fewer schools indicate that they test for any other reason.

Athletes are the next most commonly tested group; but only about 5 percent of schools indicate that they test athletes. (In 1995 the Supreme Court ruled that the testing of athletes in a given school district was constitutional. ) Students who volunteer for testing are tested in just under 4 percent of the schools, as are students that have been on school probation. Students participating in extracurriculars other than athletics—the subject of the recent Supreme Court decision—are tested at present only in about 2 percent of American secondary schools.

Public and private secondary schools are about equally likely to use drug testing, but high schools are considerably more likely than middle schools to do so. The size of the community bears little relationship to whether or not schools test for drugs. Schools in which the majority of students are African American or in which the majority are Hispanic are no more likely to drug test than those that are predominately white. If anything, they may be slightly less likely to test.

The researchers surveyed the school administrators under a grant from the Robert Wood Johnson Foundation, in support of the Youth, Education, and Society (YES) study, an element in the Foundation's Bridging the Gap Initiative. That initiative is intended to examine the impact of policies, programs and practices at the school, community, and state levels on substance use by young Americans. Students from the same schools were surveyed as part of The Monitoring the Future (MTF) study, funded by the National Institute on Drug Abuse.

(Johnston is the principal investigator of both the YES and MTF studies.)

Related links:

Institute for Social Research >

Monitoring the Future >

National Institute on Drug Abuse >

Contact: Diane Swanbrow
Phone: (734) 647-9069
E-mail: Swanbrow@umich.edu