Abstract

Alcohol is well known for its ability to impair human faculties, which creates risks when driving a vehicle or when performing safety- and risk-sensitive tasks in workplaces. The article aims to highlight some shortcomings in the legal-scientific approach for alcohol testing in South Africa. In particular, we investigate the measurement uncertainty of blood alcohol test results, which is critical in adjudicating over-the-limit cases. The South African regulatory framework for alcohol testing in the criminal- and private-law environments is examined from an analytical due-process perspective, considering measurement uncertainty and other well-established scientific principles which are essential at the interface of science and law. Special attention is paid to the National Road Traffic Amendment Bill, which aims to decrease the alcohol limit to zero. We found that the measurement uncertainty concept has not yet been received into the South African legal system, even though it is a well-established scientific principle. We suggest changes to the current alcohol legislation to accommodate the measurement uncertainty principle and the related likelihood ratio, which we believe could assist in quantifying the odds of compliance. In particular, we believe that our suggestions regarding quantification and reporting of measurement uncertainty can assist the courts and tribunals to avoid false-positive errors that may have a devastating effect on innocent subjects.