AMERICAN schools could be wasting millions of dollars on student drug-testing programmes that have little or no effect. But anti-smoking initiatives can make a difference - at least with younger pupils - the American Educational Research Association conference was told.

Around half of American students are believed to have tried drugs, and the US Supreme Court has upheld schools’ right to carry out random tests.

However, researchers from Michigan University, who analysed five years of data covering tens of thousands of students and hundreds of schools, have now raised serious doubts about this strategy. The proportion of pupils who admitted taking drugs regularly was virtually the same in schools that undertook testing as in those that did not. Smaller surveys of 3,000 male athletes and 8,000 experienced cannabis-users also supported this conclusion.

In schools with no drug tests, 16 per cent of eighth-graders (13-year-olds) used drugs, compared to 15 per cent in schools that do test.

Around 18 per cent of schools undertook testing and usually selected students they suspected were taking drugs.

Ryoko Yamaguchi told delegates that, given the financial costs of testing and the potential for causing student resentment, schools should reconsider their policies. A single test for cannabis, tobacco, cocaine, heroin, opiates, amphetamines, barbiturates and tranquilisers can cost up to $30 (pound;18.84). One for steroids that might be used by student athletes costs $100 (pound;62.82).

The same Michigan University team found that anti-smoking programmes were effective with 13-year-olds but did not deter older students. They looked at 2,780 regular smokers in 100 middle and high schools.

“The relationship between student illicit drug use and school drug-testing polices”, and “Effects of school-based tobacco cessation services on adolescent tobacco use”, by Ryoko Yamaguchi, Patrick M O’Malley, and Lloyd D Johnston, Michigan University. Contact ryokoy@umich.edu