Rebuttal: The Vera Institute's Earned-Time Survey Is Advocacy, Not Analysis
The Vera Institute's summary of its national survey on earned-time and good-time policies presents itself as research, but the document's own language gives the game away. It is published under Vera's "Ending Mass Incarceration" initiative. Its overview flatly declares the case for offense-based exclusions "untrue" without argument. Its key takeaway opens with the phrase "Fortunately for safety" — an editorial cheer, not a finding. Readers deserve to know that this is a policy brief written to advance a predetermined conclusion: that sentence-reduction credits should be extended broadly, including to many people convicted of violent crimes.
The recidivism claim rests on a measurement sleight of hand
Vera's central empirical assertion is that people convicted of violent offenses "have some of the lowest recidivism rates." This is true only under a carefully chosen definition. It is accurate that same-offense recidivism for homicide is rare — people convicted of murder seldom commit another murder after release, and Vera is entitled to that data point. But it is not the measure that matters to the public. The Bureau of Justice Statistics' nine-year follow-up of state prisoners released in 2005 — the largest longitudinal recidivism study available — found that a substantial majority of released violent offenders were rearrested for some new offense within nine years of release. Lower than the rearrest rate for property offenders, yes. "Low" by any ordinary meaning of the word, no.
When an organization selects the narrowest favorable metric (same-category reoffending), the shortest favorable window, and presents it as though it settled the question, that is advocacy. A serious analysis would have disclosed the full BJS picture and let readers weigh it. Vera's summary does not.
The Arizona example proves less than Vera implies
The summary leans on Arizona's mid-1990s truth-in-sentencing law as a cautionary tale, claiming that eliminating early-release incentives increased rule violations and recidivism. Even taking that finding at face value, it does not support what Vera is actually proposing. There is a wide policy space between eliminating incentives entirely and extending generous credits to people convicted of murder, rape, and armed robbery. Evidence that the former had costs is not evidence that the latter is safe. Arizona's experience argues, at most, for modest good-time credits as a prison-management tool — which more than 40 states already provide. It says nothing about whether the most serious violent offenders should serve substantially less than the sentence a judge and jury imposed.
Offense distinctions are a safeguard, not a "narrative"
To its credit, the report catalogs offense-based exclusions in detail — that is the survey's whole subject. The problem is its posture toward them. Vera treats crime-type exclusions as a defect to be engineered away, a relic of outdated "narratives" about who is capable of change. But these exclusions exist because legislatures in both red and blue states made a considered judgment: that murder, rape, armed robbery, and aggravated assault are categorically different from low-level, nonviolent offenses, and that sentences for the gravest crimes should mean what they say.
That judgment rests on purposes of sentencing the report never engages: proportional punishment, deterrence, and incapacitation. Good behavior in prison is a baseline expectation of custody, not a bargaining chip that converts a twenty-year sentence for a violent crime into twelve. A report that dismisses this position as mere narrative, rather than answering it, has not rebutted it.
What the summary omits
Just as telling as what the report says is what it leaves out. There is no discussion of victims — no acknowledgment that survivors of violent crime rely on the finality of a sentence, and that early release re-opens wounds and, in some cases, real danger. There is no cost-benefit accounting of the harms caused when higher-risk individuals return to communities sooner. There is no engagement with why truth-in-sentencing swept through more than 40 states in the 1980s and 1990s with broad bipartisan support: the public had watched sentences that meant a fraction of their stated term and demanded honesty in sentencing. A survey genuinely interested in "what works" would weigh these considerations. This one does not mention them.
A better standard for policymakers
None of this requires opposing earned time altogether. Modest good-time credits for genuinely nonviolent offenders, tied to completion of verified programming, are defensible and widely used across the political spectrum — the federal system's own good-time allowance is capped at roughly 15 percent of the sentence for exactly this reason. The dispute is over Vera's push to extend expansive credits to serious violent offenses, and on that question the summary offers selective metrics, an overstretched case study, and silence on victims and public safety.
Policymakers should treat this document as what it is: a brief from an organization whose stated mission is ending mass incarceration, not a neutral assessment of the evidence. Before expanding early release for violent crimes, legislators should demand full recidivism data using rearrest for any offense over long follow-up windows, independent evaluation rather than reform-aligned sources, and a seat at the table for crime victims. Serious crimes were given serious sentences for a reason. Honoring those sentences is not an outdated narrative — it is a promise the justice system made to the public.
