Recidivism & Reentry
Where this research is used
This topic brings 63 research collections together into one overview. GPS puts it to work in three ways:
- Intelligence topics. GPS's investigative pages on Georgia's prisons build on this research. This one draws on it most directly:
- GPS articles. We use this research when we write our own articles. 52 articles connected to this topic are listed below.
- Lighthouse. Lighthouse, GPS's public AI assistant, uses this research and our intelligence topics to answer people's questions.
Key Findings
Critical data points synthesized across multiple research collections. Tap a card for the full finding, GPS commentary on why it matters, and its source.
Georgia (GDC, felony reconviction within 3 years, report dated July 9, 2025): FY2022 release cohort 30.7% for all inmate facilities, the highest in GDC's published 2011–2022 series.
By the state's own narrow measure, which counts only reconvictions within three years, the latest cohort sits above the official 25–27% rate Georgia cites, so the low-recidivism picture is not stable even on its own terms.
Our reading of the finding above. It is not part of the source.Three-year felony reconviction for transition-center releases in the FY2022 cohort: 11.4 percent, versus 33.7 percent for state prisons (GDC, July 9, 2025). Selection effects are substantial.
GDC's own internal data shows vocational program completers recidivate at approximately 13.64% — roughly half the general population rate.
With 14,000–16,000 people released annually, the 2,344 transition center beds can serve fewer than 15% of annual releases at any given time.
The risk of death in the first two weeks post-release is 12.7 times higher than for the general population.
Because Georgia's official recidivism figure excludes people who die, deaths in the dangerous first weeks after release do not register in the state's main measure of how people fare after prison.
Our reading of the finding above. It is not part of the source.Vocational education contracts totaled just $172,000 in FY 2025, against a total GDC budget of $1.48 billion.
What Georgia Counts, and What It Leaves Out
Georgia reports one of the lowest recidivism rates of any state. That number is real, and it is also narrow. The state's official figure — a three-year felony reconviction rate of roughly 25 to 27 percent — measures exactly one thing: whether a person released from Georgia Department of Corrections custody is convicted of a new felony within three years of release (#2082). Cliff Hogan, the department's Director of Data Management and Research, described the methodology to state legislators in 2021: recidivism is tracked as a felony conviction within three years, and the resulting rate — about 26 percent at the time — "does not include arrests without convictions or technical violations that cause return to prison" (#2968, #7582).
The exclusions stack up. Georgia counts reconvictions, not rearrests (#2093). It uses a three-year window, so anyone who returns to prison in year four, five, or later is invisible (#2092). It removes from the dataset anyone who dies during the measurement period, even though post-release mortality is dramatically elevated (#2096). Only nine states report all three standard recidivism measures — rearrest, reconviction, and reincarceration — and Georgia is not among them (#5290, #6143). The result is a figure that is internally consistent and externally misleading: Georgia's 25 to 31 percent and the federal Bureau of Justice Statistics' 83 percent nine-year rearrest figure are not the same measurement, and both are accurate descriptions of different things (#7409).
The definitional gap is not a technicality. State auditors and outside researchers have consistently found that return-to-prison measures differ by jurisdiction in ways that make raw comparison meaningless: Georgia counts only felony reconvictions, California counts any conviction, and the Bureau of Justice Statistics counts arrests (#13676). When technical violations, arrests that never produced a conviction, and longer measurement windows are folded in, the estimated return-to-incarceration rate in Georgia is closer to 50 percent — roughly double the official number (#2084). GPS's research library has reached the same estimate independently (#6148 equivalent, #5289).
There is a further accounting problem hiding inside the recidivism number: people who die during the three-year window are deleted rather than counted as an outcome. Kate Boccia of the National Incarceration Association told the 2024 Senate Study Committee that her son died in 2021 of a fentanyl overdose during the typical three-year recidivism period, and that individuals who die within that window never appear in recidivism datasets (#7568, #2961). Given that overdose is the leading cause of death among people recently released from incarceration (#2144), the state's headline rate is computed on a population that excludes some of its worst outcomes.
The Trend Line: Georgia's Reconviction Rate Is Rising, Not Falling
For a decade before the 2010s reforms, Georgia's recidivism rate sat near 30 percent and did not move. The Georgia Council on Criminal Justice Reform acknowledged that the rate had remained "virtually unchanged for 10 years" despite a doubling of corrections spending (#7372), and Justice Reinvestment Initiative reporting described the same stagnation: approximately 30 percent of people released from prison were reconvicted of a felony within three years, a figure flat for at least a decade (#14329, #19136).
The reforms that followed — H.B. 1176 in 2012, the 2014 creation of the Georgia Prisoner Reentry Initiative, and a $57 million reinvestment of projected savings into accountability courts, substance abuse treatment, and community supervision (#2114, #14366) — were followed by a genuine reduction in the prison population of about 6 percent, achieved without an increase in crime (#2112, #2115). But the recidivism trend has since reversed. Georgia Department of Corrections data show three-year felony reconviction rates of roughly 26 to 28 percent from 2011 through 2019, dipping to 23.9 percent in 2018 and 25.3 percent in 2019, then climbing to 29.9 percent for the 2021 cohort and 31.1 percent for the 2022 cohort (#7368). A more recent published series, dated July 9, 2025, puts the same trajectory more starkly: 23.3 percent for the FY2019 release cohort, 24.3 percent for FY2020, 27.1 percent for FY2021, and 30.7 percent for FY2022 — the highest figure in the department's published 2011–2022 series (#13671, #13672, #13673, #13666, #13674).
The facility-level breakdown for that FY2022 cohort is where the story becomes legible. State prisons posted a 33.7 percent three-year felony reconviction rate; private prisons 32.1 percent; county correctional institutions 30.7 percent; and transition centers 11.4 percent (#13667, #13668, #13669, #13712). Transition center releases did roughly a third as well as state prison releases. GDC itself cautions that selection effects here are substantial — transition centers accept people who are close to parole and medically eligible, not a random sample (#13712, #2118) — but the size of the gap is difficult to attribute entirely to selection.
Georgia's rate still looks low next to national baselines, and that comparison is doing a lot of rhetorical work. Nationally, 68 percent of people released from state prisons in 2005 were rearrested within three years, 79 percent within six, and 83 percent within nine; the 401,288 releasees in that cohort accumulated roughly 2.0 million arrests, an average of five per person, with 60 percent of those arrests occurring in years four through nine (#7365, #7366, #7367, #5288). The 2012 cohort, tracked across 34 states, showed 62 percent arrested within three years, 71 percent within five, and 46 percent returned to prison (#13662, #13663, #13664). National recidivism has been declining — 77 percent of 2005 releases arrested within five years versus 70 percent of 2012 releases (#2835, #6437) — while Georgia's reconviction rate, on the state's own reported series, has been rising.
The Reentry Cliff: Release, Housing, Health, and the First Two Weeks
Between 12,000 and 16,000 people leave Georgia prisons each year (#2081, #5233). In calendar year 2025 the state recorded 13,724 total releases; 7,486 — 54.55 percent — maxed out their sentences rather than leaving on parole, and only 4,283 (31.21 percent) were released on parole (#3312, #3313, #3314). Over the decade from 2014 through 2023, GDC released 167,185 people; 45.9 percent left because their sentence expired and 37.9 percent through a parole certificate (#7606). The distinction matters: people who max out leave without supervision, without a parole officer, and — in most cases — without the transitional programming that parole-eligible people can access.
The material circumstances of release are documented in Georgia's own reporting. GDC provides, on release, a $25 prepaid Visa card, whatever balance remains in the person's trust account, a shirt, pants and cheap shoes, and a bus ticket if no one picks them up. There is no housing referral, no job placement, no identification, no phone, and no transitional support (#5248). A quarter of the prison population is not from Georgia, which makes securing basic documentation such as a birth certificate a substantial undertaking at exactly the moment it is most needed (#2891).
Housing is the binding constraint. Georgia operates 12 Reentry and Cognitive Programming centers with a total of 2,344 transitional beds, including 346 beds for women across two facilities (#2085, #13683, #2117). Against 14,000 to 16,000 annual releases, those beds can serve fewer than 15 percent of releases at any given time (#2116); the beds represent about 4.7 percent of a prison population of roughly 50,000 (#5251). The Reentry Partnership Housing program, run through the parole board, provides up to three to six months of transitional housing for people on active supervision (#2119), at $600 per month for up to three months, capped at $1,800 per parolee (#9995), and explicitly does not fund housing for children or family — a structural barrier for the majority of incarcerated women with dependents (#5138). Each year several hundred people remain in state prison solely because they have no approved residence to be paroled to (#10087, #10868).
Health outcomes in the reentry period are the most documented failure in the system. The risk of death in the first two weeks after release is 12.7 times that of the general population; over the first two years it remains 3.5 times higher (#2087, #2097, #5883, #20529). Drug overdose risk in those first two weeks is 129 times higher (#2141, #5884), and opioid-specific overdose risk is 40 times higher (#2142). Overdose is the leading cause of death among people recently released from incarceration, followed by cardiovascular disease, homicide, and suicide (#2144, #2170). Insurance coverage does not close the gap: 78 percent of men and 66 percent of women are uninsured two to three months after release, and 68 percent of men and 58 percent of women remain uninsured eight to ten months out (#2129, #2130, #2131, #2132). Georgia has not expanded Medicaid; its Pathways to Coverage work-requirement program reaches a narrow population, and most released people with mental illness are uninsured (#6932).
What Works — and How Little of It Georgia Runs at Scale
The intervention evidence is not ambiguous. The RAND Corporation's meta-analysis of correctional education covering 1980 through 2013 found that participants had 43 percent lower odds of recidivating than non-participants — a 13-percentage-point absolute reduction, from a 43 percent three-year recidivism rate to 30 percent — and 13 percent higher odds of post-release employment (#7373, #1215, #13629, #1217). Every $1 invested in prison education returns $4 to $5 in avoided reincarceration costs (#4847, #7375, #6147), and savings from reduced reincarceration run $8,700 to $9,700 per participant over three years (#1219). Even partial participation reduces recidivism relative to no participation (#1251).
Cognitive-behavioral programming shows a similar, if smaller, consistent effect. Landenberger and Lipsey's meta-analysis of 58 experimental and quasi-experimental studies found mean recidivism 25 percent lower in treatment groups than controls (#7383, #13636); representative programs show 20 to 30 percent reductions, with the largest effects among higher-risk participants and high-fidelity implementation (#4832, #13637, #7384). Notably, brand-name curricula — Reasoning and Rehabilitation, Moral Reconation Therapy, Thinking for a Change — do not outperform generic cognitive-behavioral programming (#7385). Georgia has run versions of all three through its parole system since the late 1990s, including a statewide Reasoning and Rehabilitation implementation whose early evaluations reported an 18 percent reduction in recidivism for class graduates against controls (#11128, #11109, #19224).
Georgia's own program data is the most striking and the most fragile. People who complete vocational programs in Georgia prisons have a recidivism rate of approximately 13.64 percent — roughly half the general rate of 26 percent (#6149, #3229, #3358, #5253). The state Senate Study Committee heard that vocational programming in particular reduces recidivism by half compared with the general population (#2937, #7560). Transition center releases, as noted, reconvict at 11.4 percent against 33.7 percent for state prisons (#13670). The problem is scale and selection. GDC's vocational education contracts totaled $172,000 in fiscal 2025 against a total department budget of $1.48 billion — about $3.44 per incarcerated person per year (#2086, #2111, #5252). Georgia State University shut down its prison education program in March 2024; it had cost $180,000 a year and served 60 students (#5257). The Department of Justice found in 2024 that educational and vocational programming in Georgia prisons had been "slashed rather than expanded" and that participation was "effectively impossible" in facilities where conditions were chaotic and violent (#2125, #5256).
Outside evaluations warrant caution before importing headline numbers. The InnerChange Freedom Initiative in Texas reported large reductions among graduates, but when dropouts and non-completers were included the participant group did no better — and on some measures worse — than the comparison group (#7386, #7410). Bard Prison Initiative reports under 4 percent recidivism (#5237, #4950) and the Prison Entrepreneurship Program about 7 percent against 23 percent for Texas male releasees (#13645), but both are highly selective programs admitting a small fraction of applicants. GRIP's widely cited 1.18 percent figure is drawn almost entirely from people serving life sentences who were paroled by a board, whose baseline recidivism GRIP itself states is below 3 percent (#13652, #13651, #13653). The strongest claim the evidence sustains is that education, vocational training, cognitive-behavioral treatment, and transitional programming reduce recidivism modestly to substantially — and that the effects are largest for the highest-risk people, who are least often selected (#7710, #4969).
Parole Supervision: High Completion Rates, Shrinking Releases, Rising Revocations
Georgia's parole board reports completion rates that consistently beat the national average. In fiscal year 2025, 73 percent of people on parole successfully completed supervision, against a national average estimated at approximately 60 percent (#8557, #6404, #20183). The same figure — 72 or 73 percent — recurs across fiscal years 2018 through 2025 (#8431, #8831, #9177, #11830, #12027, #11413, #11238), and in earlier years the board reported 66 percent in FY2009 against a 45 percent national rate (#13138) and 60 percent in FY2005 against 46 percent (#10634, #10635). Board leadership has described completion rate as a core performance measure and has attributed the margin to employment accountability, drug treatment, and cognitive programming (#10606, #12179).
What the completion rate does not capture is who is being released at all. Parole releases fell 42 percent over five years, from 9,455 granted out of 24,738 cases (38 percent) in 2019 to 5,443 granted out of 19,328 cases (28 percent) in 2024 (#3316, #5231, #20181). The parole rate dropped from 56.7 percent in 1995 to 34.3 percent in 2024, while average time served rose from 1.6 years to 4.1 years (#5703). Fewer people leave under supervision, and more leave by expiration — which means the population on which the 73 percent completion rate is computed is a shrinking and increasingly selected one. A high success rate among a small, screened group is not evidence that reentry is working for the 7,486 people who maxed out in 2025.
Revocations have declined sharply in absolute terms — from 2,720 in FY2019 to 2,199 in FY2020, 2,373 in FY2021, 1,825 in FY2022, 1,552 in FY2023, 1,437 in FY2024, and 1,273 in FY2025 (#2699, #2654, #8581, #8668). The board attributes part of this to intermediate sanctions — treatment referrals, administrative hearings, electronic and GPS monitoring, day reporting centers, parole detention centers — used as alternatives to revocation for people who are not a public safety threat (#8666, #11991). But the composition of who returns matters. According to the National Incarceration Association, only about 12 percent of the population returning to prison for parole violations are new offenses; the rest are technical violations (#2958). The board itself reported that fewer than 1 percent of FY2024 revocations were based on technical violations (#2654), a figure that sits awkwardly against the 12 percent new-offense estimate and against the systematic exclusion of technical violations from the state's recidivism measure (#2093).
The research the parole board has cited for two decades points at the same four levers: continuity of employment, drug testing results, attendance in treatment programs, and residence stability (#13193). Georgia's own 2001 study of 6,327 parolees quantified them: each day a parolee remained employed reduced arrest probability by about 1 percent, so a single month of employment produced a 30 percent reduction (#13536, #13537); each month of program attendance reduced arrest likelihood by 2 percent, a 24 percent reduction over a year (#13534, #13535); each residential move raised arrest likelihood by 25 percent (#13533); and a jump in positive drug screens from one-third to one-half pushed the odds of arrest nearly twentyfold (#13538). Outside research on family contact points the same direction: a 2011 Minnesota study of 16,420 prisoners found any visit reduced felony reconviction by 13 percent and parole violations by 25 percent (#5347), and a 2008 Florida study of 7,000 prisoners found each additional visit reduced two-year recidivism by 3.8 percent (#5346).
The Fiscal Case, the Budget Line, and What Gets Cut
Recidivism in Georgia is expensive in ways the state does not itemize. The estimated annual cost of recidivism runs $395 million to $790 million, based on 12,500 to 25,000 people returning to prison each year at $31,612 each (#2164). Total annual spending on incarceration plus recidivism is estimated at $1.9 billion to $2.3 billion (#2165). Reducing the reconviction rate by 10 percentage points would mean roughly 1,200 fewer reconvictions and approximately $40 million a year in avoided incarceration costs — more than fifteen times what GDC currently spends on rehabilitation programming (#6145, #5235). The parole board's own earlier estimate was that a one-percent drop in recidivism translates to $6 million in taxpayer savings (#13476).
The budget documents tell a consistent story about priorities. There are no dedicated line items for comprehensive reentry programming, transition planning, or post-release support services visible in publicly available GDC budget documents (#2111). Transition Centers received $50,067,051 in total public funds for FY2027 (#5573, #4312), and the state has added beds — $4.7 million in amended FY2024 and $10 million in FY2025 to fund 400 additional transition center beds (#3047, #3052) — while the same Senate Appropriations Committee substitute eliminated all funding for Metro Reentry programming, reducing it from the Governor's proposed $39,786 to $0 (#5553). Technical and vocational education expansion received a $1.2 million increase in amended FY2025 (#3069). That is the whole of the state's visible reentry investment against a $1.48 billion department budget.
The federal oversight record is thin on programming. Of the Department of Justice's 82 remedial measures following its 2024 findings, only three address programming at all — substance abuse, graduated housing with programs, and programs for community return. There are no mandates for education, vocational training, cognitive-behavioral programming, GED, or college (#5286). The Department of Justice concluded that people "leave prison worse than when they came in" (#2089).
The comparison states are not exotic. Michigan's Vocational Villages — three facilities, approximately 600 students across 13 trade programs, open to people within 12 to 24 months of release (#4722) — produced a 15.6 percent recidivism rate among 2019 graduates against 22.1 percent statewide, a 6.5 percentage point reduction (#4687, #7755, #7714), with 12.6 percent of participants returning to prison between 2016 and July 2023, roughly half the return rate for all 2020 releases (#4688). Michigan's overall recidivism rate was under 23 percent in 2024, and the state estimates it saves about $49,000 per person per year (#4691, #4672). Norway reports an 18 percent reconviction rate within two years and 25 percent within five, down from 60 to 70 percent before its reforms (#7650, #6200). Georgia's own vocational completers already perform at 13.64 percent (#6149) — the model works here, in small numbers, with almost no money attached to it. GPS has advocated raising vocational education funding from $172,000 to at least $15 million, about 1 percent of the GDC budget (#2174).
Related Topics
Explore related areas of research.
Contributing Collections
Research collections that contribute data to this topic.
Sources
100 cited sources across all contributing collections.