The article argues states should impose repeat‑offender sentencing enhancements keyed to prior felony counts (or severity) rather than rely on predictive reoffending tools. It claims criminal history predicts future offending across crime types and that persistent offenders don’t necessarily age out in their 30s.
— This reframes the risk‑assessment debate toward simple, auditable rules over opaque algorithms, with implications for fairness, effectiveness, and public safety.
Steve Stewart-Williams
2026.05.09
85% relevant
The article engages directly with empirical sentencing research about how sentence lengths correlate with crime type and recidivism risk (Glaeser & Sacerdote). It highlights that while some sentencing aligns with risk‑based deterrence (longer sentences where apprehension is low or recidivism risk is high), extraneous biases (race, sex, vengeance) still distort outcomes — precisely the tension captured by the existing idea that sentencing often follows conviction narratives rather than formalized risk metrics.
Rafael A. Mangual
2026.04.17
80% relevant
The article highlights Jennifer Doleac’s argument for calibrated, evidence‑based reforms that privilege concrete outcomes (recidivism rates, targeted interventions like reminders or DNA databases) over broad, ideology‑driven leniency; this maps to the idea that sentencing and reform should rely on convictions and observed behavior rather than opaque risk‑score instruments or doctrinal frames.
2025.10.03
70% relevant
The 'Lock Up Repeat Offenders' item urges incapacitating the small cohort of high‑propensity offenders and cites extensive prior records, aligning with using criminal history rather than predictive scores to guide sentencing enhancements.
Jakob Dupuis
2025.10.02
100% relevant
The author proposes enhancements that apply the higher felony class’s maximum term based strictly on an offender’s prior convictions, rejecting predictive analytics.