Random drug testing in prisons: Does a little testing go a long way?

Research Summary: We investigated whether higher rates of random drug testing lower substance use among people who are incarcerated and improve prison safety. To answer this question, we estimated linear panel two-way fixed effects models using naturally varying monthly rates of random drug testing across all Pennsylvania state prisons over a 45-month period during 2016–2019. Overall, we find that the testing rate is not related to any of the variables we consider, including our key measure of interest, positive drug tests. Policy Implications: Our analyses of historical monthly data find that higher levels of randomized testing did not yield more positive drug tests among incarcerated persons. Further and importantly, we did not find that lower testing rates was associated with negative consequences on other outcomes, suggesting that marginal increases in testing rates do not generate commensurate benefits. In this sense, our findings provide preliminary evidence that “a little testing goes a long way.” Replication and expansion of our study is required for stronger conclusions.