Death and Deterrence Redux: Science, Law and Causal Reasoning on Capital Punishment

A recent cohort of studies report deterrent effects of capital punishment that substantially exceed almost all previous estimates of lives saved by execution. Some of the new studies go further to claim that pardons, commutations, and exonerations cause murders to increase, as does trial delay. This putative life-life tradeoff is the basis for claims by legal academics and advocates of a moral imperative to aggressively prosecute capital crimes, brushing off evidentiary doubts as unreasonable cautions that place potential beneficiaries at risk of severe harm. Challenges to this new deterrence literature find that the evidence is too unstable and unreliable to support policy choices on capital punishment. This article identifies numerous technical and conceptual errors in the new deterrence studies that further erode their reliability: inappropriate methods of statistical analysis, failures to consider several factors such as drug epidemics that drive murder rates, missing data on key variables in key states, the tyranny of a few outlier states and years, weak to non-existent tests of concurrent effects of incarceration, inadequate instruments to disentangle statistical confounding of murder rates with death sentences and other punishments, failure to consider the general performance of the criminal justice system as a competing deterrent, artifactual results from truncated time frames, and the absence of any direct test of the components of contemporary theoretical constructions of deterrence. Re-analysis of one of the data sets shows that even simple adjustments to the data produce contradictory results, while alternate statistical methods produce contrary estimates. But the central mistake in this enterprise is one of causal reasoning: the attempt to draw causal inferences from a flawed and limited set of observational data, the absence of direct tests of the moving parts of the deterrence story, and the failure to address important competing influences on murder. There is no reliable, scientifically sound evidence that pits execution against a robust set of competing explanations to identify whether it exerts a deterrent effect that is uniquely and sufficiently powerful to overwhelm the recurring epidemic cycles of murder. This and other rebukes remind us to invoke tough, neutral social science standards and commonsense causal reasoning before expanding the use of execution with its attendant risks and costs