The evaluation finds a substantial impact of random drug testing on self-reported recent use of the tested drugs. It does not find an effect on participation in the activities that trigger random drug testing eligibility. I am not surprised by the former as prices matter. What does perhaps surprise me is the lack of a detectable response on the participation margin, though the point estimates are negative.
The key caveats are the self-reported nature of the substance use outcome measures and the use (typical for IES evaluations for political and operational reasons) of volunteer schools.
The most entertaining part of the evaluation was the TWG meeting where I discovered that I was the only one of the technical advisors, some of whom study anti-drug programs for a living, who knew the meaning of "420". Any room where I am the hippest person is a room in serious trouble.