The 66.6% Problem: What New York’s Bail Experiment Revealed
In the debate over bail reform, there is one number that deserves far more attention than it gets: 66.6%.
That was the re-offense rate identified in a study examining repeat violent felony offenders released under New York’s revised bail system.
Not technical violations.
Not missed court dates.
New alleged criminal conduct.
When policymakers removed judicial discretion for large categories of offenses, the theory was that detention before trial should be rare and narrowly applied. The reality, however, exposed something far less comfortable: a significant number of repeat violent offenders re-entered the system after being released.
What Changed
New York’s bail reforms dramatically limited when judges could require bail or detain individuals pretrial. The framework prioritized release in most non-homicide cases, including many involving repeat offenders.
The guiding philosophy was simple: detention should not hinge on ability to pay.
But in stripping away discretion, the law also removed a key safeguard — the ability of courts to weigh an individual’s prior violent history in determining risk.
What the Data Showed
The study examining repeat violent felony offenders found that two out of every three re-offended after release.
That is not a marginal policy miss.
When a repeat violent offender is released and re-offends, the cost is not abstract. It is borne by victims, families, neighborhoods, and law enforcement.
Public policy does not operate in a vacuum. Every decision reallocates risk. The question is always: who carries it?
Under broad mandatory-release frameworks, the risk shifts outward — from defendant to community.
The Risk Transfer Problem
Bail debates are often framed as a binary choice between fairness and detention. That framing is incomplete.
Pretrial systems exist to balance competing interests:
- The presumption of innocence
- Court appearance
- Public safety
When discretion is eliminated and risk assessment is artificially narrowed, the balance tilts — not toward fairness, but toward unmanaged exposure.
The 66.6% re-offense rate among repeat violent felons is not an indictment of reform in principle. It is a warning about reform without guardrails.
The Larger Lesson
Most jurisdictions considering bail reform are not debating first-time, low-level offenders. They are wrestling with how to handle individuals with documented histories of violence.
The New York data suggest that blanket restrictions on judicial discretion — especially for repeat violent offenders — carry measurable consequences.
Policy is not about slogans. It is about outcomes.
If two-thirds of repeat violent felons released under a particular framework re-offend, that is not a talking point. It is a design flaw that demands correction.
Reform is not inherently reckless. But reform that ignores recidivism data is.
Communities deserve a pretrial system that accounts for both liberty and safety — not one that assumes risk disappears simply because it is no longer acknowledged.
Footnotes & Sources
- The 66.6% re-offense figure is derived from an analysis of repeat violent felony offenders released under New York’s 2019 bail reforms. The study examined rearrest outcomes following mandatory release provisions enacted as part of Chapter 59 of the Laws of 2019 (New York State Bail Reform).
- See New York State Division of Criminal Justice Services (DCJS), Pretrial Release and Rearrest Outcomes Following 2020 Bail Reform, and related statistical reports examining rearrest rates for individuals released under the revised bail statute.
- New York Criminal Procedure Law §§ 510.10 and 530.20 (as amended 2019), limiting judicial discretion in setting bail for designated offenses.
- For broader context on recidivism among released defendants, see New York State Office of Court Administration (OCA) and DCJS published pretrial outcome data (2019–2022 updates).
