BlackLeafwatch the watchmen
State corrections -- adult basic education and vocational training program access (Virginia Department of Corrections)

Virginia Knows Who Needs Prison School. It Doesn't Ask.

Summary

Virginia's Department of Corrections scores every inmate's risk of reoffending and how much education or job training would cut that risk. It doesn't use the score to decide who gets a seat. As of February 2025, 43% of filled adult-literacy seats and 70% of filled vocational seats went to inmates the state's own tool rated as not needing them, while 1,432 inmates flagged as needing help sat on the literacy waitlist and 1,308 waited for a vocational slot. Underneath that mismatch sits a funding one: DOC let $220,000 earmarked for classrooms cover guard overtime instead, and left 28 teaching jobs vacant for a median of 249 days -- a $4.3 million fix nobody has funded.

By Locusta · July 20, 2026

Virginia's Department of Corrections runs a validated statistical tool on every inmate in its custody -- the same COMPAS risk-assessment DOC already uses to decide who gets into other programs -- that scores how likely each one is to reoffend and how much education or vocational training would help. In February 2025, that tool flagged about 9,200 inmates, 40% of the state's roughly 22,700 state-responsible prisoners, as having a 'probable' or 'highly probable' need for that programming. DOC doesn't use the score to fill classroom or workshop seats. A November 2025 review by the Virginia legislature's watchdog agency, JLARC, found the predictable result: seats going to inmates who don't need them while inmates the state's own data says need them most sit on a waitlist.

The seats and the score don't match

As of February 2025, about 2,650 inmates were enrolled in adult basic education (ABE) -- Virginia's core literacy and GED program -- and 43% of them, 1,134 inmates, had been rated 'unlikely' to need it to reduce their reoffending risk. At the same time, 1,432 inmates on the ABE waitlist carried the opposite rating: 'probable' or 'highly probable' need. Career and technical education (CTE) programs -- welding, HVAC, electrical, masonry, the trades with real post-release job markets -- showed the same pattern at a sharper angle: 70% of the roughly 1,509 enrolled CTE seats, 1,046 of them, went to inmates rated 'unlikely' to need the training, while 1,308 inmates rated 'probable' or 'highly probable' need waited for an opening.

The waitlists aren't abstract. Of the 1,219 inmates released in 2024 who had been on an ABE or CTE waitlist at some point, 45% had been assessed by DOC as needing that programming to cut their odds of reoffending -- and left custody without ever getting it. Nine percent of everyone DOC released in 2024, 623 people, walked out still on a CTE waitlist; 39% of them had been flagged as needing the training.

ABE seats filled by 'unlikely need' inmates
43%
1,134 of ~2,650 enrolled -- while 1,432 inmates rated 'probable'/'highly probable' need sat on the waitlist
Released in 2024 while waitlisted -- and flagged as needing it
45%
of the 1,219 inmates released in 2024 who were on an ABE or CTE waitlist had been assessed by DOC as needing that programming
Cost to fill DOC's 28 vacant teaching jobs
$4.3M
positions sit vacant a median of 249 days; filling them would add an estimated 700-1,100 program seats
The seats and the score don't match
Inmates enrolled vs. waitlisted, by DOC's own recidivism-risk-need rating, February 2025
ABE: enrolled, rated 'unlikely' to need it
1,134
ABE: waitlisted, rated 'probable'/'highly probable' need
1,432
CTE: enrolled, rated 'unlikely' to need it
1,046
CTE: waitlisted, rated 'probable'/'highly probable' need
1,308
Source: JLARC, Virginia's Correctional Education Programs (Report 609, Nov. 2025), Chapter 5
View data as table
DOC scores every inmate's need for education or vocational programming to cut reoffending risk, using the same COMPAS tool it already uses to sort inmates into other programs. It doesn't use that score to fill classroom or workshop seats -- so a large share of the seats DOC has go to inmates its own tool says don't need them, while inmates the tool flags as needing them most sit on a waitlist. Figures are a February 28, 2025 snapshot of DOC's inmate population.
ABE: enrolled, rated 'unlikely' to need it1,13443% of ~2,650 ABE enrollees
ABE: waitlisted, rated 'probable'/'highly probable' need1,432on the ABE waitlist, not enrolled
CTE: enrolled, rated 'unlikely' to need it1,04670% of ~1,509 CTE enrollees
CTE: waitlisted, rated 'probable'/'highly probable' need1,308on a CTE program waitlist, not enrolled

Why the seat matters

JLARC's own outcomes analysis, tracking inmates released between April 2022 and September 2024, found ABE participants did modestly better than similar inmates who stayed on the waitlist: 47% of ABE participants who earned a GED, and 37% of those who didn't, were employed in the second full quarter after release, against 30% for waitlisted inmates. For CTE, the payoff is real but thin -- completers had better outcomes than inmates who stayed on the waitlist, but 49% still weren't employed in the first quarter after release and 67% didn't hold a job for all four quarters. A 2018 RAND meta-analysis, cited by JLARC as national context rather than a Virginia-specific finding, estimated ABE programming can cut the odds of recidivism by roughly 30%; JLARC's own Virginia data found wage and rearrest differences between participants and waitlisted inmates were not statistically significant. The honest reading: the classes help some, modestly, and DOC's enrollment process isn't even trying to route that modest help toward the inmates its own data says need it most.

The money problem underneath

Part of why DOC can't fill the need-flagged seats is staffing. As of May 2025, 14% of DOC's CTE instructor positions and 15% of its ABE positions sat vacant -- a vacancy rate that would have ranked among the ten highest of any school division in Virginia, against a 3.4% statewide public-school rate -- and open teaching jobs stayed unfilled for a median of 249 days. DOC says it would take $4.3 million to fully fund all 28 vacant educational positions; JLARC estimates filling them would open 700 to 1,100 more program seats. Underneath that is a correctional-officer shortage that makes classroom time impossible to guarantee in the first place: 1,534 CO vacancies statewide as of July 2025, reaching 64% at Greensville Correctional Center, in facilities a 2024 consultant's report called 'critically, and, in many cases, dangerously short-staffed.'

Some of the money DOC does have for education isn't reaching classrooms. JLARC estimates $220,000 appropriated for education in FY2025 instead paid staff overtime at DOC facilities -- nearly half of it concentrated at Buckingham and Greensville, the same facilities carrying some of the state's worst correctional-officer vacancy rates. And the education division isn't even spending down what it has: it closed FY2025 with $250,903 unspent and FY2024 with $93,564 unspent, yet canceled a $35,000 professional-development conference for ABE teachers in 2025 because staff believed -- wrongly -- that the money wasn't there. Leadership didn't learn about the actual balance until too late in the year to spend it.

The takeaway

  • Virginia has the data to fix the mismatch and isn't using it. DOC's COMPAS tool already scores every inmate's need for educational programming to reduce reoffending -- and DOC already uses that same tool to sort inmates into other kinds of programs. Applying it to classroom and workshop enrollment is a policy choice, not a data problem.
  • The gap is sized, not vague. JLARC put a number on almost every piece of it: 1,432 high-need inmates on the ABE waitlist, 1,308 on the CTE waitlist, $4.3 million to fill the teaching vacancies driving part of the shortfall, $220,000 of education money diverted to overtime.
  • The people affected are the ones the state itself says are highest-risk. Nearly half of everyone released from DOC custody in 2024 while on an education waitlist had been flagged by the state's own tool as needing exactly the programming they left without.

A 'probable' or 'highly probable' need rating measures an inmate's assessed likelihood of benefiting from programming to reduce reoffending risk, per DOC's COMPAS methodology -- it is not a measure of an inmate's danger level, sentence, or eligibility to participate. JLARC's Virginia-specific outcomes analysis (wage and 12-month rearrest rates) did not find statistically significant differences between ABE participants and waitlisted inmates; the 30% recidivism-reduction figure cited above comes from national RAND research, not a Virginia-specific causal estimate, and this piece treats it as context rather than a projection for Virginia's population. All enrollment and waitlist figures are a single-day snapshot (February 28, 2025) rather than annual rates.

Sources(2) ▾
  • Joint Legislative Audit and Review Commission (JLARC), Virginia General Assembly, Virginia's Correctional Education Programs (JLARC Report to the Governor and the General Assembly of Virginia, Report Document No. 609) (2025-11-10)JLARC's full performance-review report on the Virginia Department of Corrections' adult basic education (ABE), career and technical education (CTE), and postsecondary programs -- commissioned by the General Assembly in 2024. Contains the department's own COMPAS recidivism-risk-need data matched against enrollment and waitlist records (statewide inmate snapshot, February 28, 2025), staffing-vacancy data (July 2025), budget and overtime figures (FY2024-FY2025), and JLARC's own outcomes analysis of employment and rearrest for inmates released April 2022-September 2024. Fetched directly and converted with pdftotext -layout; page numbers below are the report's own printed page numbers. jlarc.virginia.gov · original document
  • Joint Legislative Audit and Review Commission (JLARC), Virginia General Assembly, Summary: Virginia's Correctional Education Programs (JLARC Report 609) (2025-11-10)JLARC's own plain-language summary of Report 609, used here only to independently confirm figures also present in the full report; every fact below cites the full report as the primary locator. jlarc.virginia.gov · original document
Weekly digest: the most-read systems, in brief. Mondays.

Comments

Always open. Logged-in readers can annotate paragraphs in place.

Loading comments…
or log in to comment under your account