Capital Punishment and Deterrence: Some Further Thoughts (Journal of Political Economy, 1977)
Classic econometric analysis arguing that executions have a deterrent effect on murder rates under certain model assumptions.
All topicsDeath penalty · 2 peer-reviewed studies for, 7 against
The best empirical case for the death penalty is deterrence: some studies report execution-linked reductions in homicide under particular econometric designs, including influential early work and later panel-data analyses (Journal of Political Economy, 1977; Economic Inquiry, 2006). But the modern literature does not converge on a stable, policy-relevant estimate. Reviews and re-analyses stress that results are highly sensitive to model specification, measurement choices, and assumptions about how would-be offenders perceive execution risk, which is typically low and uneven across jurisdictions (The Economists' Voice, 2006). Formal treatments of model uncertainty find any deterrent signal to be weak and fragile rather than robust (American Law and Economics Review, 2009). Meanwhile, potential harms are clearer: some research finds patterns consistent with “brutalization,” where executions correlate with increases in certain killings, at least in specific settings (Criminology, 1998). Beyond deterrence, capital punishment has an unavoidable irreversibility problem: estimates of wrongful conviction among death-sentenced defendants imply a nontrivial risk of executing an innocent person (PNAS, 2014). Overall, given uncertain benefits, plausible adverse effects, and irreversible error costs, the evidence-weighted policy conclusion leans against the death penalty, especially absent exceptionally strong procedural safeguards and demonstrably reliable deterrent impacts.
Ranked by citation count.
Classic econometric analysis arguing that executions have a deterrent effect on murder rates under certain model assumptions.
Analysis of execution data from North Carolina and California reveals that current lethal injection protocols may not reliably cause death through intended mechanisms, with potentially conscious inmates experiencing pancuronium-induced asphyxiation rather than peaceful anesthesia.
Classic econometric analysis arguing that executions have a deterrent effect on murder rates under certain model assumptions.
Finds estimates consistent with a deterrent effect of executions using state-level postmoratorium panel data, though results depend on specification choices.
Analysis of execution data from North Carolina and California reveals that current lethal injection protocols may not reliably cause death through intended mechanisms, with potentially conscious inmates experiencing pancuronium-induced asphyxiation rather than peaceful anesthesia.
Finds evidence consistent with “brutalization,” where executions are followed by increases in certain categories of killings in the studied context.
Reviews the empirical literature and argues results are too fragile and inconsistent to support a reliable deterrent effect.
Study of 50 correctional officers conducting executions in Louisiana found high trauma exposure, conflicted feelings, and reliance on dissociation and diffusion of responsibility to manage psychological distress from carrying out executions.
Applies model-uncertainty methods and concludes evidence for deterrence exists but is weak and highly sensitive to modeling choices.
Estimates a substantial risk of wrongful conviction among death-sentenced defendants, highlighting irreversible-error costs in capital punishment systems.
An empirical study of 3,284 mock jurors found that those self-reporting racial biases were 18.3%-18.4% more likely to sentence Black defendants to death compared to White defendants with identical factors. The study demonstrates systemic racial bias in capital sentencing decisions.
Some studies, including early econometric work and later state-level panel analyses, report a deterrent effect. Reviews and model-uncertainty analyses find that these results are highly sensitive to modeling choices and that any deterrent signal is weak and fragile.
Research estimating false conviction rates among death-sentenced defendants finds a substantial risk of wrongful conviction. Because executions cannot be reversed, this is a central concern in the evidence against capital punishment.
A study of 3,284 mock jurors found that those who self-reported racial biases were about 18% more likely to sentence Black defendants to death than White defendants with identical case factors.
An analysis of execution data from North Carolina and California found that lethal injection protocols may not reliably work as intended. Some inmates may have been conscious and experienced asphyxiation rather than peaceful anesthesia.