Junk Science at Sentencing

Junk science used in criminal trials has contributed to hundreds of wrongful convictions. But the problem is much worse than that. Junk science does not only harm criminal defendants who go to trial, but also the overwhelming majority of defendants—over ninety-five percent—who plead guilty, skip trial, and proceed straight to sentencing. Scientific, technical, and other specialized evidence (“STS evidence”) is used regularly, and with increasing frequency, at sentencing. Despite this, Federal Rule of Evidence 702 and its state equivalents—which help filter unreliable STS evidence at trials—do not apply at the critical sentencing stage. In fact, at sentencing, no meaningful admissibility standard guards against junk science deciding punishment. Over ninety-five percent of defendants, therefore, do not get the basic protection against faulty STS evidence that trial defendants get. This may result in harsher sentences based on junk science that has been admitted and considered without any screening or vetting. This Article offers the first in-depth exploration of STS evidence at sentencing. It links two bodies of literature: the first analyzing the negative effects of junk science on the reliability of trials and the second arguing for extending procedural protections to sentencing. This Article builds upon these literatures by proposing an implementable mechanism for evaluating STS evidence at sentencing while retaining special protections for criminal defendants. The Article recommends that Federal Rule of Evidence 702 (or its state analog) apply at sentencing to determine the admissibility of STS evidence offered in support of harsher punishment, but not to such evidence offered by defendants as mitigation.