Senate Judiciary hears data limits, definitions as S-12 (record sealing) moves forward

Montpelier, Vt. — Members of the Senate Judiciary committee on Feb. 18 heard researchers explain limits of Vermont’s criminal-justice data and the state’s official recidivism definition as the panel continued consideration of S-12, a bill that would affect sealing/expungement of records and researcher access to those records.

The presentation came from Monica Weber, executive director of the Scribe Research Group, and Dr. Robin Joliet, the group’s director of research. Weber said the Scribe Research Group compiled data across law enforcement, the judiciary and corrections to support policymaking and that federal Bureau of Justice Statistics funding supported much of the work. “I’m Monica Weber. I’m the executive director of the Scribe Research Group,” she told the committee. Joliet and Weber outlined how different datasets are created and where administrative records leave gaps.

The researchers told senators that Vermont’s formal, statutory recidivism definition adopted in 2011 — limited to people “sentenced to incarceration for one year or more who return to prison within three years for a conviction or for violation of supervision” — captures a small portion of people with criminal cases and can change with policy, not just behavior. “This is an entirely political creation so that you can compare across jurisdictions,” Weber said, explaining why many researchers use broader definitions when evaluating programs.

Why it matters: Committee members are weighing S-12’s provisions on who can access and petition to seal records. Lawmakers asked whether changing the statutory recidivism measure or expanding researcher access would alter practical outcomes for people with records, what the bill would actually measure and whether the committee could reliably use current state dashboards for policy decisions.

Key facts and data cited - The Scribe Research Group’s Act 40 work used multiple state systems and a 2018 start point for comparable statewide reporting; researchers said NIBRS (National Incident Based Reporting System) reporting before about 2018 is uneven for some jurisdictions and they therefore rely on 2018–present for many comparisons. - The statutory recidivism definition the researchers described was adopted in 2011 as part of earlier justice-reinvestment efforts and uses a one-year incarceration threshold and a three-year follow-up window. - In the report cohort cited by the researchers, 902 people in the recidivist cohort returned within the three-year window; researchers said about 71% of those returns occurred within the first year after release. - Researchers told the committee that only a minority of people sentenced in criminal court receive an incarceration sentence (researchers cited roughly 26 percent in the years referenced) and that an even smaller subset receives sentences of one year or more, which is the statutory group counted as recidivists under the current definition.

What the researchers said about data systems and caveats Weber and Joliet described a “criminal-justice data pyramid” showing who and what is excluded from administrative records. They said many victim experiences never enter police or court records; police calls-for-service catalogs a wide range of noncriminal contacts; NIBRS and vendor systems (commonly referred to in testimony as Valcour and Spillman) each have differences in coverage and quality; and fingerprint‑supported criminal histories are the most reliable match for tracking an individual across systems.

The researchers noted specific limitations the committee should consider: Burlington and other early adopters of the Valcour system produced incomplete NIBRS reporting during transition years (roughly 2014–2017), so statewide trend comparisons use 2018 as a reliable baseline. They also said out‑of‑state criminal histories matter: in one analysis roughly 40% of a cohort had out‑of‑state history that affected risk scores and eligibility for programs.

Discussion and next steps Committee members discussed S-12’s timelines for eligibility to seal records (references in committee conversation to 3, 7 and other-year windows), how court processing delays intersect with statutory waiting periods, and whether the measure the state publishes for recidivism is useful for policy evaluation. Senators asked for follow-up information on returns to prison that are violations of supervision and on the department practices that feed the state dashboard. The committee scheduled additional testimony and follow-up: the researchers requested that provisions preserving researcher access to criminal-history records remain in S-12; senators planned to hear Marshall Paul on Thursday and follow up with Judge Zonne on petitionless sealing questions the judge had discussed previously.

No vote was taken on S-12 during the session. Committee members discussed the aim of scheduling a vote by Friday, Feb. 25, or the week following, but set no formal action in the transcript.

Committee context and caveats The researchers emphasized that changes in law or policy (for example, decriminalizing possession of marijuana or changing return‑to‑prison policies for supervision violations) can reduce or raise the published recidivism rate without reflecting individual behavioral change. They recommended that program evaluations use conviction‑based measures over a three‑year follow-up when possible, and that population‑level policy effects be measured over longer periods (five to 10 years) given Vermont’s policy volatility.

The session closed with the committee asking staff to collect the clarifying materials the presenters referenced (Act 40 materials, a conditions‑of‑release analysis, the CDC protective‑factors guidance) and to arrange the scheduled testimony for additional clarification.