Brief written July 18, 2026 from GPS Intelligence System data.(A)

Legislative Intelligence Brief: The Fiscal and Constitutional Exposure of the Georgia Department of Corrections

Prepared by Georgia Prisoners' Speak (GPS) for members of the Georgia General Assembly, legislative staff, and committee analysts.

Executive Summary: The Cost of the Status Quo

Georgia is spending record sums to operate a prison system that a federal civil-rights investigation has found unconstitutional, that its own former commanders describe as a coverup, and that generates a steady, recurring drain on the state treasury through liability payouts. The Georgia Department of Corrections (GDC) now commands more than $1.8 billion a year in appropriations — $1,799,204,979 in the Amended FY 2026 cycle and a peak of $1,913,888,054 in actual FY 2025 spending, the highest on record (Governor's Budget Report, Amended FY 2026 and FY 2027). Yet across every measurable outcome — deaths, homicides, sexual violence, litigation exposure — the system is deteriorating.

For legislators, this is not primarily a moral question, though it is that too. It is a fiscal-management and liability-exposure question. The state is buying worse outcomes at higher prices. The daily cost to incarcerate one person reached $86.61 in FY 2024 ($31,612 annually), driven by healthcare contracts and chronic staffing failures, while community supervision costs $2.89 per day (GPS budget analysis; Governor's Budget Report). The health program line alone rose from $325.6 million in FY 2024 to a projected $432.2 million in FY 2027 — a 33 percent increase now consuming roughly 24 percent of the entire agency budget.

This brief is organized around what legislators can act on: settlement exposure, projected future liability, two reform packages with bill-language entry points (Vision 2027 and End the Warehouse), documented sanitation failures in specific district facilities, the mortality tracking gap that hides the true scale of the crisis, and the wrongful-conviction landscape. Every claim traces to the GPS corpus and public records obtained through open-records requests.

The full legislative deep-dive is maintained at https://gps.press/intelligence/legislative/ and the underlying budget analysis at https://gps.press/intelligence/issue/budget-analysis/.

Section 1: Settlement and Verdict Exposure — What the State Has Already Paid

The clearest fiscal signal available to legislators is the ledger of what the state has already paid to resolve civil claims arising from deaths and injuries inside GDC facilities. GPS obtained the Georgia Department of Administrative Services (DOAS) Risk Management settlement ledger through open-records requests. Court-verified records document a pattern of high-dollar payouts clustered at specific facilities — a pattern that should concern any appropriator, because these are recurring, predictable liabilities, not one-time anomalies.

Documented death-linked settlements

The DOAS ledger records payouts ranging from a $10,836 payment in 2008 to a $29,170 resolution in late 2025, but the severe cases dominate the fiscal picture (GPS analysis of DOAS Risk Management ledger; see https://gps.press/intelligence/issue/legal-settlements/):

• $4 million to the estate of David Lamar Henegar for a death at Johnson State Prison in 2021 — the largest single settlement in the ledger.
• $3.25 million for Reginald Jacobs Jr., who died of dehydration in a solitary confinement cell at Calhoun State Prison.
• $3.25 million to the family of Juan Carlos Ramirez Bibiano after a 2023 death at Telfair State Prison.
• $3 million for Thomas Henry Giles, who died in the care of Augusta State Medical Prison.
• $2.99 million paid in 2021 to Khalid Mouton at Georgia State Prison.

The clustering matters for budget forecasting. Telfair State Prison alone tallies at least five settlements since 2012. When a single facility generates repeated multi-million-dollar payouts, that is a signal that the underlying conditions — not isolated incidents — are producing the liability. Each of the deceased individuals named above represents a completed transaction against the state treasury; each represents conditions that GPS mortality data and the DOJ findings indicate persist.

Why this is a recurring line item, not a closed chapter

The settlement pattern is downstream of documented, ongoing conditions: dehydration in solitary (Jacobs), medical failures (Giles at a medical prison), and violence in understaffed facilities. Because GDC has not remediated the underlying staffing collapse (Section 4) or the conditions failures (Section 5), the state should budget for continued and likely growing settlement exposure. The FY 2025 spending peak coincided with the deadliest year in state history (Section 6) — a correlation that presages further liability.

Section 2: Pending Litigation and Projected Future Liability

Beyond completed settlements, the litigation docket signals where future exposure lies. GPS tracks federal filings in the Georgia district courts. The recently terminated cases in the corpus — including multiple Humphreys v. Oliver filings (GAND, 1:25-cv-07012 closed 2026-01-15; 1:25-cv-06100 closed 2025-12-10), Grant v. Ward (GAMD, 5:22-cv-00396 closed 2025-08-28), and Ballard v. Davis (GAMD, 5:25-cv-00046 closed 2025-09-26) — resolved at $0, but the sheer volume of GDC-directed federal litigation against named commissioners and wardens is itself a warning indicator.

The most consequential source of projected liability is not a single case but the federal umbrella: the U.S. Department of Justice's October 2024 CRIPA findings. Court-verified DOJ findings concluded that conditions in Georgia's prisons violate the Eighth Amendment and amount to "deliberate indifference" — findings the DOJ described as "among the most severe violations" it had documented in any prison investigation (DOJ CRIPA findings, October 2024). A CRIPA finding of this severity typically precedes a consent decree or federal litigation carrying court-monitored compliance costs, mandatory staffing and facility investments, and ongoing legal fees — a category of liability that dwarfs individual settlements.

Georgia already has direct experience with the cost of non-compliance. The solitary-confinement litigation over the Special Management Unit at the Georgia Diagnostic and Classification State Prison produced a 2019 federal settlement followed by a 2024 civil contempt order after a court found GDC officials "falsifying compliance records, placing people in strip cells, and stalling 'with no desire or intention to comply'" (court records; https://gps.press/intelligence/issue/solitary-confinement/). Civil contempt findings are a direct fiscal risk: they invite sanctions, expanded monitoring, and attorney-fee awards.

Legislative takeaway: The state faces a two-tier liability structure — recurring individual settlements (Section 1) and a looming systemic federal remedy (DOJ/CRIPA). Appropriators should request from GDC and DOAS a full accounting of (a) settlement reserves, (b) projected CRIPA-compliance costs, and (c) outstanding contempt exposure in the SMU litigation. GPS can provide the underlying settlement ledger and litigation tracking to support that request.

Section 3: Vision 2027 — The Post-Conviction Reform Package and Its District Lift

Vision 2027 is GPS's post-conviction reform agenda, grounded in the fiscal paradox at the heart of Georgia corrections: the state spends more than ever to operate a system the DOJ has found unconstitutional, while the best available research shows this incarceration produces little or no additional public safety (https://gps.press/intelligence/issue/vision-2027/).

The fiscal case for parole and medical-reprieve reform

The single most actionable fiscal lever in Vision 2027 is the collapse of the parole grant rate and the resulting cost of warehousing an aging population. The numbers, drawn from GPS analysis of the Governor's Budget Report and Parole Board data:

• The parole grant rate collapsed from 38 percent in FY 2019 to a record-low 28 percent in FY 2024.
• For people serving life sentences, only 93 of 2,046 cases were granted.
• Georgia's medical reprieve process is among the most restrictive in the nation, requiring an inmate to be "entirely incapacitated" with death expected within 12 months.

Meanwhile, the population most expensive to hold is precisely the population least likely to reoffend. As of April 2026, 27 percent of active prisoners — 12,777 individuals — were 50 or older; more than 1,300 are 70 or older, and 217 are 80 or older (GPS inmate database, over 293,000 records). These older prisoners account for 57.4 percent of all GPS-tracked deaths in custody, require costly specialist care, and recidivate at strikingly low rates — nationally, people released after age 65 commit new crimes at just 3.5 percent, versus a 41 percent general rate. Georgia's own Parole Board reports a 72 percent successful parole completion rate.

GPS has calculated that cost avoidance from parole alone reached $343 million in FY 2024, given the $2.89-per-day community-supervision cost versus $86.61 for incarceration. The aging lifer population — now 2,653 lifers 55 or older, 72 percent of whom are Black — represents a mounting unfunded liability, because GDC is constitutionally mandated to provide medical care until death.

The habeas deadline and a system its own Chief Justice calls broken

Vision 2027's post-conviction plank centers on the four-year statute of limitations on habeas corpus petitions, enacted in 2004 under O.C.G.A. § 9-14-42. The deadline has spent two decades extinguishing claims before underlying injustices come to light. In March 2026, the state's highest judicial officer — the Chief Justice — publicly addressed the problem, and GPS reporting notes the Chief Justice has called the post-conviction system "a mess" and the habeas deadline "broken" (https://gps.press/intelligence/issue/legal-access/; https://gps.press/intelligence/issue/legal-settlements/). When the head of the judiciary characterizes a statutory deadline as broken, the legislature has both political cover and a clear mandate to revisit O.C.G.A. § 9-14-42.

District lift

Vision 2027 reforms deliver district-specific fiscal relief because the aging, low-risk population is dispersed across facilities in every region. Every parole grant to a medically fragile, low-recidivism individual removes an $86.61/day medical liability from the state ledger and returns a constituent to a district's community-supervision system at $2.89/day. This is a rare reform where the fiscal argument and the constituent-service argument align.

Bill-language entry points for Vision 2027: 1. Amend O.C.G.A. § 9-14-42 to extend or create equitable-tolling exceptions to the four-year habeas deadline, consistent with the Chief Justice's stated concerns. 2. Reform the medical-reprieve standard to replace "entirely incapacitated / death within 12 months" with a functional standard tied to care cost and recidivism risk. 3. Establish a geriatric-parole presumption for incarcerated people over a defined age threshold with clean disciplinary records, with fiscal-note credit for the documented $343 million cost-avoidance mechanism.

Section 4: End the Warehouse — Rehabilitation Strategy and ROI from Comparable States

The End the Warehouse plan (https://gps.press/intelligence/issue/end-the-warehouse/) rests on a documented contradiction: rising spending, collapsing outcomes. Department-wide spending rose from roughly $1.53 billion in actual FY 2024 to more than $1.9 billion in actual FY 2025, with the approved FY 2027 plan at about $1.79 billion (Governor's Budget Report; HB 974, FY 2027G) — yet the population has been essentially flat, holding roughly 50,082 people as of late June 2026.

This tracks a national pattern the Prison Policy Initiative's Following the Money 2026 report documented: U.S. prison spending rose 27 percent between 2017 and 2025 even as prison populations shrank 15 percent — driven largely by overtime and emergency pay. Georgia's trajectory mirrors that inversion exactly: the state now pays more per officer while employing fewer officers than before the spending surge began.

The staffing collapse driving the violence and the spending

Georgia carried a 52.5 percent correctional officer vacancy rate systemwide as of 2024, and the Guidehouse consulting team hired by the Governor found at least eight of the state's 34 prisons operating above 70 percent vacancy, with Valdosta State Prison reaching 80 percent by April 2024 (https://gps.press/intelligence/issue/staffing-crisis/). The DOJ confirmed facilities running with as few as one to three officers supervising 1,500 to 1,800 prisoners on nights and weekends.

The state's official numbers understate the crisis. Former GDC assistant warden Russell Zirkle went on the record with GPS, testifying that the state "blends fully-staffed halfway houses into system averages" to conceal 70–80 percent vacancy rates at major prisons. Zirkle — whose career ran officer to sergeant at Valdosta, to captain and chief of security at Autry — named Telfair, Hancock, Smith, Valdosta, Macon, Autry, and Calhoun state prisons as examples (on-record, timestamped). Former GDC officer Tyler Ryals provided corroborating staffing-coverup testimony; Zirkle stated of Ryals: "Everything he says is completely true" (corroborated ×2). Zirkle's 70–80 percent vacancy testimony corroborates the GPS open-records production showing six officers on duty when the riot began at Washington State Prison, and GPS reporting documenting five officers to cover 69 posts when four men were killed at Washington on January 11, 2026.

This is the critical point for appropriators: the roughly $600 million spending surge on staffing has failed to reverse the spiral. More money into the existing model produces overtime and emergency pay, not filled posts or safer facilities. This is the definition of a failing ROI.

ROI from comparable states

End the Warehouse and the companion gang-separation analysis (https see below) document that other states solved comparable problems with lower-cost, structural interventions:

• Texas, Arizona, and California each developed comprehensive gang-management approaches — housing-based separation, intelligence-driven classification, structured renouncement/exit programs, and incentive systems — decades ago, with documented results ("315 Gangs, Zero Strategy: How Georgia Abandoned Its Prisons While Other States Found Solutions," GPS, 2026-03-08).
• Georgia, by contrast, has identified 315 gangs and validated roughly 15,200 people — 31 percent of its incarcerated population — as gang-affiliated (more than double the ~13 percent national average per the National Institute of Justice), yet operates no systematic gang-separation housing policy, no structured exit program, and no dedicated operational strategy for keeping rivals apart.

The gang-separation reform is the flagship ROI case because it costs essentially nothing. As GPS documented in "Separate the Gangs. It Costs Nothing. Georgia Keeps Choosing the Bodies." (2026-05-22), separation is a classification and housing decision, not a capital expenditure. And Georgia already has the legal mandate: O.C.G.A. § 42-5-52, on the books since 1897, requires — "shall provide for the classification and separation of inmates" — the very separation the department declines to perform ("On the Books Since 1897: The Separation Law Georgia Refuses to Enforce," GPS, 2026-05-22). A second statute establishes every person's right to be safe from gang fear and violence. The department honors neither.

The rehabilitation ROI: normalization

The broader End the Warehouse thesis is that enforced idleness is the precondition for the drugs, gangs, despair, and severed family ties documented across GPS's reporting ("Reopen the Doors — Normalization," GPS, 2026-06-05; "Normalization: The Principle That Changes Everything," GPS, 2025-11-20). In a typical state dormitory, only three or four men in a hundred hold a work detail. The FY 2027 budget's rehabilitation investments are near-zero against the operating machine: $4.2 million to add 263 private-prison beds at Coffee and Wheeler, and a $150,000 pilot for peer-led programming at Autry State Prison (Vision 2027 analysis; HB 974). That ratio — millions for beds, a rounding error for programming — is the policy choice this legislature can reverse.

Bill-language entry points for End the Warehouse: 1. Enforce O.C.G.A. § 42-5-52 with a statutory reporting requirement: mandate GDC to report validated-gang population by facility and demonstrate a written separation-housing plan, with the fiscal note credited against violence-driven settlement exposure (Section 1). 2. Require a transparent, unblended staffing report to the General Assembly that separates halfway-house/transitional-center staffing from major-prison staffing — directly addressing the Zirkle-documented blending. 3. Reallocate a defined percentage of new appropriations from bed expansion to programming, work, and education, modeled on the Texas/Arizona/California ROI evidence. 4. Direct overtime/emergency-pay auditing to determine how much of the $600 million staffing surge purchased actual filled posts versus premium pay.

Section 5: Sanitation and Conditions — DPH Inspection Data by Facility

The Georgia Department of Public Health's restaurant-inspection portal publishes prison-kitchen scores, and the distribution offers legislators a rare, independently sourced, verifiable conditions metric (https://gps.press/intelligence/issue/scores-without-sanitation/).

The scoreboard and the failures

GDC's own data confirms that adequate sanitation is achievable: Autry, Baldwin, Calhoun, Central, and Hancock State Prisons each scored 100 on routine visits, and Lee, Rutledge, and Riverbend posted multiple perfect scores over three years. That makes the failures a matter of management, not inevitability.

Three GDC facilities have fallen below the DPH 70-point passing threshold since 2022:

• Johnson State Prison received a 64 on December 11, 2023 — the lowest documented Georgia prison-kitchen score. Inspectors found multiple rats and roaches "with little to no change," rat droppings and urine on bulk oil, flour, and rice-bran packaging, and broken ovens, skillet, kettle, griddle, freezer, and ice machine. A follow-up nine days later produced only a 67. The score climbed to 96 in December 2024 but fell back to 80 by March 2025 — an improvement that "never stuck."
• Pulaski State Prison hit 67 on January 29, 2026 — the kitchen's only handwashing sink nonfunctional with plumbing ripped from the wall, sewage backing up through floor drains (a repeat violation), and hot-holding violations (nacho meat at 65°F, sauce at 123°F). A February 6, 2026 follow-up rebounded to 96 — the same failure-correction-failure pattern.
• Smith State Prison has been a chronic underperformer: 68 in May 2022, back to 72 (a C) in February 2026, with rodent activity noted in every inspection reviewed from 2022 through 2025 and cockroaches added at the bakery and tray-making station in February 2026.

The measurement gap

GPS's core finding — reflected in the topic title "Scores Without Sanitation" — is that even passing scores don't reflect what incarcerated people actually eat from, and that the food-safety numbers coexist with a $0.54-per-meal / roughly $1.61–$1.65-per-day food budget, against an American Correctional Association recommended benchmark of roughly $3.66 per meal (https://gps.press/intelligence/issue/prison-nutrition-georgia/). The Marshall Project independently reported Georgia's prison food cost at roughly 60 cents per meal in May 2026. The FY 2027 food line ($30.7 million) is a slight real-terms decline from FY 2024 ($30.9 million).

Bill-language entry point: Mandate quarterly public reporting of DPH prison-kitchen scores to the relevant oversight committee, with an automatic corrective-action requirement and legislative notification when any facility scores below 70. This is a low-cost transparency measure using data GDC and DPH already generate.

Section 6: Mortality Trend and the Tracking Gap

Georgia publishes no official prison-homicide count. GPS maintains the state's most complete independent mortality archive (https://gps.press/georgia-prison-deaths/), and the gap between GPS figures and GDC's public reporting is itself a governance failure legislators should address.

The trend (canonical GPS figures — always a floor)

GPS has tracked at least 1,847 deaths in GDC custody since 2020 (GPS-tracked mortality data, canonical count). Because GDC reports cause of death roughly two months late, every figure below is a floor, not a ceiling:

• 2020: at least 294
• 2021: at least 257
• 2022: at least 255
• 2023: at least 262
• 2024: at least 333 — the deadliest year on record, a 27 percent increase over the prior year
• 2025: at least 301
• 2026 year-to-date: at least 145

The 2024 toll included more than 100 homicides by GPS's count — far exceeding the state's own public figures and dwarfing the 8–9 annual homicides of just a few years earlier. The DOJ counted 142 homicides between 2018 and 2023, a 95.8 percent increase in the second three-year period, climbing from 7 in 2018 to a then-record 35 in 2023. The DOJ-funded Safe Inside initiative found the prison death rate rose 47 percent between 2019 and 2024 (2.8 to 4.1 per 100,000), with assaults on inmates up 54 percent and assaults on staff up 77 percent.

The reporting gap and misclassification

The DOJ found that GDC "inaccurately reports these deaths both internally and externally, and in a manner that underreports the extent of violence and homicide in its prisons." GPS's own investigative documentation makes the mechanism concrete. At Ware State Prison, GPS obtained 231 death-investigation reports, GBI autopsies, and toxicology findings covering 84 deaths since 2020, documenting at least 19 homicides in under six years — the state named a suspected attacker in 11 of those deaths, but GPS could find no public record that any accused was charged, indicted, or tried ("At Least Nineteen: The Murders the State Didn't Prosecute," GPS, 2026-07-06). At Macon State Prison, GPS documented the case of Glen Christian Krauch, tortured for three weeks and hidden under a bunk while roughly 168 mandatory counts were reportedly recorded as complete ("Three Weeks Under a Bunk," GPS, 2026-02-05).

Recent deaths concentrate at the same facilities the data flags

The mortality record from the last 180 days confirms the clustering. Recent deaths include multiple losses at Valdosta State Prison (Ramon Ortiz, Steven Bryant, Antavious Bailey, Kevin James Flamer, age 26; Jeremiah Alan Brown, age 40), Washington State Prison (Courtney Davis, Isreal Moses Jones, Deshawn Poole), Ware State Prison (Kojack Junior Thomas, age 27; Justin Dean Pulley, age 48; Johnathan Cleo Hardman-Simmons, age 36), Augusta State Medical Prison, Pulaski State Prison (two Pulaski County in-custody deaths within two weeks, per WGXA and 41NBC reporting, July 2026), and the Special Management Unit (Antony Ramon Penick, age 32). These are the same facilities Zirkle named for 70–80 percent vacancy and where settlements have clustered — the liability, staffing, and mortality data all point to the same set of institutions.

Bill-language entry point: Establish a statutory in-custody death reporting mandate — modeled on federal Death in Custody Reporting Act requirements — requiring GDC to report every death within a fixed window, with manner-of-death classification, to a designated legislative body and to publish it. The GPS/GDC gap exists because Georgia has no enforceable, timely, public reporting requirement. The tracking data GPS maintains at https://gps.press/facility/gdc-hidden-deaths/ documents the scale of what current reporting omits.

Section 7: The Wrongful-Conviction Landscape

The post-conviction plank of this brief connects directly to wrongful-conviction risk. Two structural features of Georgia's system elevate that risk and the associated liability.

First, the habeas deadline forecloses discovery of wrongful convictions. As documented in Section 3, the four-year O.C.G.A. § 9-14-42 deadline "extinguishes claims that might otherwise have uncovered unconstitutional convictions" (https://gps.press/intelligence/issue/legal-access/). When new forensic evidence, recanted testimony, or Brady violations surface after year four, the door is closed regardless of the merits — the precise scenario the Chief Justice's "broken" characterization addresses.

Second, the wrongful-conviction landscape carries direct state fiscal exposure. Recent reporting confirms the compensation mechanism is active: a judge ruled that a couple imprisoned for a baby's death is eligible for a state payout (Atlanta Journal-Constitution, 2026-07-10). And the system's basic reliability controls are failing at the front and back ends: a Clayton inmate was released by mistake, joining a "long list of similar Georgia cases" (Atlanta Journal-Constitution, 2026-07-07). A system that cannot reliably track who is in custody or who should be released is a system that generates both wrongful-detention and wrongful-release liability.

GPS does not yet have a comprehensive public dataset on GBI crime-lab forensic-science error rates specific to Georgia in this corpus, and we will not manufacture one. What the corpus does establish is that GBI autopsy and toxicology findings are central evidence in the death-investigation record (the Ware State Prison files GPS obtained included GBI autopsies), and that forensic determinations can cut both ways — GPS's derived, high-confidence records note that official autopsy findings at Washington State Prison refuted a stabbing-victim hypothesis in late 2025, illustrating that forensic conclusions materially change the classification of a death. That underscores the importance of independent, timely, and reviewable forensic reporting.

Bill-language entry points: 1. Amend O.C.G.A. § 9-14-42 to create a discovery-based or actual-innocence exception to the four-year habeas deadline (also listed under Vision 2027). 2. Require an audit of GDC's release and custody-tracking systems in response to the documented pattern of mistaken releases, with a fiscal note reflecting wrongful-detention and wrongful-release liability. 3. Strengthen the wrongful-conviction compensation statute's procedural pathway so eligibility determinations do not require separate protracted litigation.

Section 8: District-Relevant Facility Reference

Every legislator represents constituents connected to the GDC system — as employees, as families of the incarcerated, and as taxpayers funding the $1.8 billion budget. The following facilities appear most frequently in the fiscal-exposure, staffing, mortality, and conditions data in this brief. Committee staff can pull facility-specific pages for any prison in a member's region.

Highest-liability / highest-mortality clusters (data-flagged in this brief):

• Valdosta State Prison — 80% officer vacancy (April 2024); multiple recent deaths; named by Zirkle. https://gps.press/facility/valdosta-state-prison/
• Ware State Prison — at least 19 documented homicides since 2020, none prosecuted; multiple recent deaths; ~1,447 held against ~500 design capacity. https://gps.press/facility/ware-state-prison/
• Washington State Prison — four killed January 11, 2026 with five officers for 69 posts; multiple recent deaths; ~72% vacancy. https://gps.press/facility/washington-state-prison/
• Smith State Prison — seven prisoners killed in 2024 (most of any facility); warden indictment (racketeering/bribery); chronic DPH kitchen failures; named by Zirkle. https://gps.press/facility/smith-state-prison/
• Telfair State Prison — at least five settlements since 2012 including the $3.25M Ramirez Bibiano payout; named by Zirkle. https://gps.press/facility/telfair-state-prison/
• Calhoun State Prison — $3.25M Jacobs dehydration settlement; named by Zirkle. https://gps.press/facility/calhoun-state-prison/
• Macon State Prison — Krauch torture case; named by Zirkle. https://gps.press/facility/macon-state-prison/
• Hancock State Prison — gang-violence deaths; named by Zirkle. https://gps.press/facility/hancock-state-prison/
• Autry State Prison — named by Zirkle; site of the $150,000 peer-programming pilot; DPH score of 100. https://gps.press/facility/autry-state-prison/
• Augusta State Medical Prison — $3M Giles settlement; multiple recent deaths. https://gps.press/facility/augusta-state-medical-prison/
• Johnson State Prison — $4M Henegar settlement; lowest documented DPH kitchen score (64). https://gps.press/facility/johnson-state-prison/
• Pulaski State Prison — two in-custody deaths within two weeks (July 2026); DPH kitchen score of 67. https://gps.press/facility/pulaski-state-prison/
• Georgia Diagnostic and Classification State Prison — houses the Special Management Unit under federal contempt order; staff-misconduct allegation cluster. https://gps.press/facility/georgia-diagnostic-and-classification-prison/
• Special Management Unit — subject of 2024 civil contempt order; recent death. https://gps.press/facility/special-management-unit/
• Georgia State Prison — $2.99M Mouton settlement. https://gps.press/facility/georgia-state-prison/

County and regional facilities in members' districts can be located through the machine-readable facility dataset at https://gps.press/facilities-data/, and every facility page follows the pattern https://gps.press/facility/{name}/. GPS maintains 114 facility profiles.

Section 9: Consolidated Policy Recommendations with Bill-Language Entry Points

The following recommendations are ordered by fiscal leverage — the highest cost-avoidance and lowest implementation cost first.

1. Enforce the 1897 separation statute (near-zero cost, direct violence/liability reduction). Add a reporting-and-plan mandate to O.C.G.A. § 42-5-52 requiring GDC to (a) report validated-gang population by facility and (b) file a written separation-housing plan with the oversight committee. The evidence base — Texas, Arizona, California — is documented; the statutory duty already exists; the cost is administrative.

2. Mandate unblended, timely staffing reports. Require GDC to report correctional-officer vacancy by facility with transitional centers and halfway houses reported separately from major prisons, directly addressing the blending documented on the record by former assistant warden Russell Zirkle.

3. Enact a statutory in-custody death reporting requirement. Close the GPS/GDC mortality gap with a fixed-window reporting mandate including manner-of-death classification, published and reported to a legislative body. This addresses the DOJ finding that GDC underreports violence and homicide.

4. Reform parole and medical reprieve for the aging, low-risk population. Establish a geriatric-parole presumption and replace the "entirely incapacitated / death within 12 months" medical-reprieve standard with a functional standard. Fiscal note: credited against the documented $343 million FY 2024 parole cost-avoidance figure and the mounting aging-lifer medical liability.

5. Amend the habeas deadline (O.C.G.A. § 9-14-42). Create a discovery-based / actual-innocence exception, consistent with the Chief Justice's March 2026 statements. This reduces wrongful-conviction compensation exposure and addresses a system the judiciary's own head calls broken.

6. Require quarterly public reporting of DPH prison-kitchen scores with automatic corrective action and legislative notification below 70. Uses existing DPH/GDC data.

7. Audit release and custody-tracking systems in response to the documented pattern of mistaken releases.

8. Reallocate appropriations from bed expansion to programming. Redirect a defined percentage of new funds from private-prison bed additions (the FY 2027 $4.2 million / 263 beds at Coffee and Wheeler) toward work, education, and rehabilitation, modeled on comparable-state ROI.

Call to Action for Legislators and Staff

Georgia Prisoners' Speak maintains the underlying datasets behind every figure in this brief — the DOAS settlement ledger, the mortality archive, DPH inspection scores, the 293,000-record inmate database, on-record former-commander testimony, and open-records productions including the Washington State Prison staffing document. We make these available to legislators and legislative staff for fiscal notes, committee memos, and bill drafting.

We ask members and staff to take three concrete steps:

1. Schedule a briefing with GPS. We will walk your committee or personal office through the settlement exposure, the CRIPA liability outlook, and the facility-level data for your district. Reports and intake are available at https://gps.press/submit-a-report/ and https://gps.press/tellmystory/.

2. Request specific data. Ask us for (a) the full DOAS settlement ledger, (b) facility-level mortality and DPH scores for the prisons in your region (https://gps.press/facilities-data/), and (c) the unredacted, source-attributed staffing testimony from former GDC commanders. Constituents can identify their legislators at https://gps.press/find-your-legislator/.

3. Sponsor a bill. The eight entry points in Section 9 are drafted around existing statutes — O.C.G.A. § 42-5-52 and O.C.G.A. § 9-14-42 — and existing data streams, so they can move quickly and carry defensible fiscal notes. The lowest-cost, highest-leverage starting point is enforcement of the 1897 separation statute the state already has on its books.

The full legislative intelligence hub is at https://gps.press/intelligence/legislative/, with deep-dive briefings on the budget (https://gps.press/intelligence/issue/budget-analysis/), Vision 2027 (https://gps.press/intelligence/issue/vision-2027/), End the Warehouse (https://gps.press/intelligence/issue/end-the-warehouse/), settlements (https://gps.press/intelligence/issue/legal-settlements/), and oversight (https://gps.press/intelligence/issue/oversight-investigations/).

The state is already paying for this crisis — in multi-million-dollar settlements, in a $1.8 billion budget that buys worse outcomes each year, and in a federal liability exposure that is only growing. The question before this legislature is whether it will keep paying for failure or invest in the low-cost, evidence-based reforms that comparable states have already proven work. GPS is ready to help you build the case.