Solutions

Brief written September 27, 2026 from GPS Intelligence System data.(K)

Every other page in the GPS intelligence system documents what is wrong with Georgia's prisons. This page documents what would fix it — each solution costed, each obstacle named, and each grounded in reforms that have already been measured in peer states or demanded by Georgia's own law.

The starting facts are not in dispute. Georgia runs an roughly $1.8 billion-a-year corrections system holding 52,134 people across 88 tracked facilities, with 2,798 more state prisoners backed up in county jails (GDC population snapshot, September 25, 2026). GPS has independently tracked at least 1,896 deaths in GDC custody since January 1, 2020 — a floor, because GDC reports deaths roughly two months late and the U.S. Department of Justice found the state "inaccurately reports these deaths both internally and externally, and in a manner that underreports the extent of violence and homicide" (DOJ CRIPA findings, October 2024). The same DOJ investigation concluded that GDC violates the Eighth Amendment, that violence "function[s] as the norm," and that leadership "has lost control of its facilities."

None of that is inevitable. It is the product of specific, reversible choices. Here is the path forward, ordered roughly by cost-to-impact — cheapest and highest-leverage first.

The Flagship Fix: Separate the Gangs

The single cheapest, fastest way to cut the stabbings, extortion, and deaths inside Georgia's prisons is not a new law, a new building, or a new contract. It is a housing-assignment policy: house rival gangs apart from each other, and house gang-affiliated prisoners apart from the unaffiliated "civilian" population that the gangs recruit, tax, and prey upon. GPS laid out the full case in two reports: "Separate the Gangs. It Costs Nothing. Georgia Keeps Choosing the Bodies." and "On the Books Since 1897: The Separation Law Georgia Refuses to Enforce."

The law Georgia refuses to use

Georgia has carried a statutory command to do exactly this for more than a century. Under O.C.G.A. § 42-5-52, language tracing to 1897, "the department shall provide for the classification and separation of inmates with respect to age, first offenders, habitual criminals and incorrigibles, diseased inmates, mentally diseased inmates, and those having contagious, infectious, and incurable diseases." The word is shall — a duty, not a courtesy. The state is not missing a tool. It is ignoring one it has held since 1897.

The refusal is visible in GDC's own numbers. In September 2026, GDC produced to GPS — in response to an open-records request — a spreadsheet of active offenders by facility and supervision level. It shows four nominally medium-security prisons packed with men classified Close, the level GDC reserves for people with "assaultive criminal or institutional history" who need "constant supervision by an armed correctional officer" (Georgia Board of Corrections Rule 125-3-1-.02). At three of the four, the Close-security share went up between October 2025 and September 2026. All four house three men in cells built for one — while Commissioner Tyrone Oliver has told state senators that single cells reduce violence (GPS, "Three to a Cell," September 2026).

What the refusal costs

The violence is organized and it is current. On April 1, 2026, coordinated gang violence locked down at least twelve prisons, with life flights to two facilities and stabbings at five. On May 21, 2026, gang violence hit Ware, Dooly (nine hospitalized), Central (six), and Augusta State Medical Prison, and a man was stabbed inside the lockdown at Washington State Prison — a facility already locked down since four men were killed there on January 11, while running roughly 72% officer vacancy. On September 3, 2026, a tactical squad searched F2 dorm at Wilcox State Prison from about 7 a.m. to noon; when the men were let out to shower, the attack began, and the county coroner was called by about 2 p.m. (three separate inmate accounts sent to GPS; WGXA reported the coroner's call). On September 16, a gang fight at Telfair State Prison put three men in local hospitals, and GDC locked down every state prison again — while its own press page that week carried only two deputy warden appointments.

A statewide lockdown is not a strategy; it is a confession that the state cannot say who is safe in which dorm. The dead have names, and the GPS death registry records them:

• Maurice Antonio Brown died September 3, 2026, at Wilcox State Prison; the registry lists the manner as homicide. He was pronounced dead where he fell, two hours after the squad left the dorm.
• Marcus Walker died January 10, 2026, at Wilcox State Prison, age 20; the registry lists the manner as homicide. State death records describe manual strangulation in the cell he shared with another man — at a medium-security prison where GDC's own count put 694 of 1,824 men at Close security.
• Corey Jose July died June 17, 2025, at Hancock State Prison, age 33; the registry lists the manner as homicide. Surveillance footage showed him chased across the yard and cornered at a locked gate; a Hancock County jury convicted his attacker of malice murder on August 12, 2026.
• Zeary Davante Davis died September 26, 2024, at Dooly State Prison, age 31; the registry lists the manner as homicide. He was stabbed to death on his birthday, three days after a warning letter about gang control of housing units at Dooly was sent to Commissioner Oliver; news records show the prisoners' own volunteer fire crew reached him before the shift supervisor did, and that no officer had keys to the three gates between the dorm and outside medical help.
• Squaier Lamar Jiles died September 16, 2024, at Ware State Prison, age 35; the registry lists the manner as homicide. Coroner and GBI records obtained through an open-records request show he bled to death from sharp-force injuries.
• Christopher Eli Gresham died September 30, 2021, at Ware State Prison, age 39; the registry lists the manner as homicide. The state's own autopsy records indicate he was stabbed by at least three other prisoners.

Ware State Prison alone has produced at least nineteen prisoner-on-prisoner homicides since 2020, according to coroner and state crime-lab records GPS obtained through an open-records request. In eleven of those deaths, the state's own death investigations name a suspected attacker — and GPS found no public record that any of the accused was ever charged ("At Least Nineteen: The Murders the State Didn't Prosecute"). The DOJ documented 142 homicides systemwide from 2018 through 2023, climbing from 7 in 2018 to 35 in 2023 — a 95.8% increase across the period's second half.

The proof it works

This is not a theory. When Texas implemented wholesale placement of confirmed gang affiliates in restrictive housing between 1990 and 1999, gang-related incidents dropped dramatically — while an earlier strategy of segregating only gang leaders had failed (research by Ralph and Marquart; GPS, Gang Separation as Violence Reduction Strategy). Arizona's Security Threat Group program, evaluated by Arizona State University researchers, found assaults and other violations fell by over 50% among segregated gang members, with a 30% system-wide reduction in total rule violations after gang segregation was implemented. The lesson runs in both directions: Arizona members who formally renounced showed only non-significant improvement, which means renouncement programming must be paired with separation, not substituted for it. Texas's nine-month GRAD exit pathway (2,668 completions as of 2012) offers that pair — though GPS flags honestly that no rigorous recidivism or before/after violence evaluation of GRAD exists, and the National Institute of Justice notes no gang-exit program has yet been rigorously shown to remove people from prison gangs permanently. Separation does not require belief. It requires a wall between people trying to kill each other.

Cost, feasibility, and the obstacle

Cost: near zero. GDC already classifies every prisoner; its SCRIBE system has tracked restrictive-housing populations since March 2019 and its Warden's Dashboard has displayed days-in-restrictive-housing since FY2020. The data exists. Separation is a bed-assignment decision, not a construction project — and it would save money by reducing lockdowns, life flights, and the homicide caseload.

Obstacle: GDC leadership, and the politics of admission. Enforcing § 42-5-52 means admitting the current classification practice is lethal. The Department's public posture has been denial — it publicly dismissed the DOJ's documented findings as reflecting a "fundamental misunderstanding," and its response to each gang event is a system-wide lockdown rather than a housing redesign. The fix therefore has to be compelled: by the General Assembly enforcing its own statute, by appropriations language requiring published close-security counts by facility, or by a federal court under the CRIPA action the DOJ findings explicitly contemplate.

Push it forward: ask your state senator and representative one question — why is the Department of Corrections not enforcing O.C.G.A. § 42-5-52? Find them at gps.press/find-your-legislator. The full violence evidence base is at gps.press/intelligence/issue/violence/.

Vision 2027: The Legislative Package

GPS's Vision 2027 is the post-conviction justice reform framework: the synthesis of more than a dozen GPS research collections covering the economics of incarceration, the legal architecture of accountability, and the statutory levers that already exist in Georgia law. It is organized around seven planks. For each: what it does, the lift it creates, and who blocks it.

1. Enforce and codify gang separation and classification transparency. Codifies § 42-5-52's duty into a modern classification statute with published quarterly counts by facility and security level. Lift: the flagship fix above, made durable across administrations. Obstacle: GDC leadership, which has never published a Tier or close-security population count in 110 non-monthly reports GPS reviewed.

2. Create an independent corrections oversight office. Modeled on New Jersey's Office of the Corrections Ombudsperson — fully independent, reporting to the Governor, with subpoena power, unannounced-inspection authority, and confidential communications — which oversees nine prisons and about 13,000 people for $2,806,000 a year with 26 staff (FY2025), and was strengthened by New Jersey's Dignity Act in 2020. Washington's Office of Corrections Ombuds was created by statute in 2018 (RCW 43.06C), runs about $2.5 million a year, and its statutory purpose explicitly includes reducing DOC litigation exposure. Scaled to Georgia's 52,000-person system, peer-state costs run $75 to $191 per incarcerated person per year (Connecticut $75; Minnesota $147; Washington $191) — roughly 0.2% of GDC's budget. Design caveat from the evidence base: Britain's inspectorate (HMIP) "cannot enforce" its findings, so Georgia's office must pair inspection with an enforcement or mandatory public-response mandate. Obstacle: neither GDC nor the Governor's office has supported an independent office; enabling legislation is required.

3. Death and critical-injury reporting transparency. Requires public reporting of every in-custody death within days, not months, with cause and manner of death published as they are certified; an independent death-review process; and enforcement of the Board of Corrections' own 1984 rule — still in force — that "the warden shall notify" families when a prisoner is severely injured (GPS, "The Warden Shall Notify"). Today, 690 of the at-least-1,896 GPS-tracked deaths since 2020 carry no determined cause; GDC has recorded deaths as suicide months before the medical examiner certified a manner (records at Georgia Diagnostic and Classification State Prison); and some Washington State Prison deaths were certified natural with no autopsy while homicides and overdoses received autopsies. Lift: the state can no longer hide how people die. Obstacle: GDC's information control — it publishes nothing about lockdowns or deaths on its own initiative.

4. Restore meaningful post-conviction court access. Georgia's four-year habeas deadline has no innocence exception, and its law libraries guarantee two hours a week on paper. The Chief Justice of Georgia's own Supreme Court has publicly called the post-conviction system "a mess" and asked lawmakers to fix it (GPS, "We Did a Lot of the Breaking"). GPS has documented people convicted on forensic science the FBI has since disowned, with the courthouse door then closed ("Nobody Told Them"). Lift: merits, not procedural defaults, decide outcomes. Obstacle: inertia and district-attorney opposition.

5. Establish a statewide conviction integrity commission. Costed at $2.0–2.5 million — 0.11–0.14% of the corrections budget, less than the cost of incarcerating 80 people a year. Conviction Integrity Units nationally secured 62 exonerations in 2024; Philadelphia's CIU has exonerated people collectively imprisoned over 1,160 years since 2018; North Carolina's state-funded Prisoner Legal Services has saved taxpayers over $12 million by correcting illegal sentences. Obstacle: prosecutorial resistance and the election of prosecutors who deprioritize this work.

6. Restore parole and use geriatric release. GPS's "The Board That Stopped Saying Yes" documents the lifer parole collapse in the Board's own numbers. The fiscal lift is enormous: GDC's own cost figures put incarceration at $80.31 per person per day against $3.13 for parole supervision (FY2025), and the parole board itself claims $343 million in cost avoidance in FY2024. For older prisoners the safety case is settled: the federal CARES Act released 11,000+ elderly and medically vulnerable people, of whom 0.15% were arrested for new crimes; federal compassionate-release recidivism is 3.5% versus 41% for the general population; the U.S. Sentencing Commission found offenders over 50 recidivate at 21.3% versus 53.4% for those under 50; and GDC's own Aging-Inmate Population Project puts medical costs at $8,500/year for prisoners 65+ versus $950 for those under 65 — a 9x multiplier on a population that included 5,753 people aged 60+ in mid-2026. Virginia saved $6.6 million releasing just 62 eligible elderly prisoners. Obstacle: parole-board risk aversion and restrictive eligibility criteria — 45 states have compassionate-release laws and barely use them.

7. Fund rehabilitation and end extraction. GPS advocates raising vocational education funding from $172,000 to at least $15 million — 1% of the GDC budget — with a goal of 5% within three years (against a RAND-documented return of $4–$5 per dollar). The same plank repeals Georgia's pay-to-stay statute (O.C.G.A. § 42-5-59 mandates deducting "the cost of the inmate's keep" from inmate earnings), ends Georgia's membership in the seven-state club that pays nothing for most prison work, caps the deductions that can reach 80% of inmate pay, and stops the Securus phone contract's 59.6% commission — roughly $8 million a year flowing back to the state off 13-cents-a-minute family calls. Fine-and-fee reforms elsewhere lifted $2.6 billion in uncollectable debt across eight reforms in six states, often at no revenue loss, because the debt was never going to be collected. Obstacle: these revenue streams are baked into agency budgets and county expectations.

End the Warehouse: Rehabilitation That Pays for Itself

GPS's End the Warehouse campaign starts from a diagnosis: Georgia strips prisons of work, school, purpose, and family contact, and into that vacuum rushes the gang, the drugs, and the despair. The alternative has a name — normalization: life inside should resemble life outside as closely as security allows, because the point is to return people to that outside life intact. The punishment is the loss of liberty, and nothing beyond it.

This is not idealism; it is the best-documented return on investment in corrections:

• Pennsylvania's "Little Scandinavia" unit at SCI Chester — 64 beds, about $300,000–$310,000 in renovation (~$4,844 per bed), a 1:8 officer-to-resident ratio versus 1:128 in the rest of the facility — has recorded one physical altercation since opening in 2022 and near-zero violence in 2024, even as Pennsylvania's statewide prison violence rose about 21.6%. Pennsylvania announced expansion to three more facilities in March 2025. Caveat, and it travels with the numbers: results are early and mostly non-causal; per-inmate daily cost runs about 1.5x double-celling.
• South Carolina's Restoring Promise units produced, in a randomized controlled trial, a 73% reduction in the odds of violent incident write-ups and an 83% reduction in restrictive-housing stays in the first year; 94.6% of surveyed young adults said they felt safe.
• Michigan's Vocational Village graduates in 2019 recidivated at 15.6% versus 22.1% statewide — 6.5 percentage points lower; from 2016 to mid-2023 only 12.6% of participants returned to prison. Michigan's overall recidivism fell under 23% in 2024, second-lowest in state history, saving about $49,000 per person per year.
• Correctional education overall: RAND's meta-analysis found participants have 43% lower odds of recidivating, 13% higher post-release employment, and every dollar invested saves $4–5 in reincarceration costs ($8,700–$9,700 per participant over three years, against program costs of $1,400–$1,744 per inmate). The Washington State Institute for Public Policy ranks correctional education the highest-return adult corrections program there is.
• Solitary reduction frees staff and money: solitary requires roughly twice as many guards. North Dakota cut solitary use 74.28% between 2016 and 2020 (monthly sanction rates fell 99% at one prison); Oregon's resource-team model cut disciplinary infractions 55.7% and assaults 73.9% among participants, and use-of-force in its Behavioral Health Unit dropped nearly 86% from 2016 to 2021; California's Ashker settlement cut Pelican Bay's long-term isolation population from 513 to 2 — a 99.6% reduction, with no reported violence surge — and cut the statewide SHU population 65%. Colorado banned solitary in two mental-health prisons and saw assaults, forced cell entries, and heavy restraints decline 40%.
• Treatment: when Rhode Island implemented all three FDA-approved medications for opioid use disorder during and after incarceration, post-release overdose deaths fell 75%. Maine now treats nearly 50% of its incarcerated population with MAT and saw resident-on-resident assaults fall 40%, staff use-of-force fall 69%, and self-inflicted injuries fall 84% at its state prison.

Decarceration does not raise crime — the single best-documented finding in the GPS research library. New York more than halved its prison population from 1999 to 2023 while violent crime fell 34% (faster than the national 28%), and closed more than a dozen prisons. New York City cut serious crime 58% while incarceration fell 55% (1996–2014; 31,120 fewer residents behind bars). California downsized 23% from 2006 to 2012 with violent crime falling at the national rate. Five states — Alaska, Connecticut, New Jersey, New York, Vermont — have cut prison populations over 50% from peak. The Netherlands cut its prison rate about 46% from 2005 to 2016 and closed roughly half its prisons as crime fell. Norway runs 18% two-year reconviction (down from 60–70% pre-reform) at $127,671 per inmate per year — higher per year, but per-spell costs are lower because sentences are short. Caveat, and it must travel with those numbers: Nordic and Western European outcomes reflect much shorter sentences, smaller populations, and stronger welfare baselines. They prove measurable alternatives exist — not that they transplant one-to-one to Georgia.

Georgia's own proof it can do this: Governor Deal's justice reinvestment (2012–2015) cut the prison population 6%, averted $264 million in incarceration costs, and reinvested $57 million in recidivism reduction. Texas's reinvestment avoided $3 billion-plus in projected prison costs and cut parole revocations 46%; Louisiana cut its prison population 9% by the end of 2018 and shifted $30 million to community services.

Cost and obstacle. The money is not the constraint — it is the allocation. Between FY2025 actual and FY2027 approved, the State Prisons program was cut $187.5 million (16.78%) while private prisons grew $20.9 million and security-and-technology line items totaled roughly $27.1 million against $1,225,705 in rehabilitation investment — a corrected ratio of about 22:1 surveillance to rehabilitation. Georgia also pays CoreCivic and GEO Group roughly $140 million a year for about 7,800 prisoners, and the state's own auditors found private beds cost about 10% more per day than state beds when controlling for risk. The obstacle is appropriations choices and a department whose leadership has publicly dismissed documented conditions as "propaganda."

A Fix for Every Documented Failure

Medical care

The problem, in the federal government's words, is "abhorrent," "life-threatening" care under a privatized model: Centurion holds a $2.4 billion no-bid contract after 23 years of university-run care, and DOJ found Georgia was treating only about 10% of the people it knew had hepatitis C — a disease curable since 2014 (GPS's "The Cure Exists" documents a 41-year-old who died of it). Stephen Prochaska died January 21, 2025, at Augusta State Medical Prison, age 70; the GPS registry lists the manner as suicide — at the system's designated medical facility, in a mental-health caseload where GDC's own classification data records 1,243 people "poorly controlled" and 45 in "active mental health crisis."

The fixes: re-bid the contract with enforceable staffing ratios, liquidated damages, and an independent clinical monitor; a statewide hepatitis C eradication program (the cure is cheaper than the cirrhosis); universal MOUD on the Rhode Island model; and geriatric release to move the 9x-cost, near-zero-risk elderly population off the medical budget. Obstacle: the Health program grew $37.3 million to $427.2 million in FY2027 with no performance conditions attached — money without accountability is the current policy. More at gps.press/intelligence/issue/medical-neglect/.

Staffing

Georgia has run at roughly 50% correctional-officer vacancy for years — 5,991 budgeted positions, 2,985 vacant as of January 2024 — over 70% at ten of the largest facilities (DOJ). Former GDC commanders Russell Zirkle and Tyler Ryals have testified that state averages are "skewed" by blending fully-staffed halfway houses into system figures, hiding 70–80% vacancy at major prisons including Telfair, Hancock, Smith, and Valdosta — corroborated by GPS's open-records production showing six officers on duty when the Ware riot began.

The fix is proven: Pennsylvania cut its CO vacancy rate from 10.5% to 4.8% in two years by creating a dedicated Recruitment and Retention Division (late 2022), holding 750+ job fairs in 2024, and raising trainee pay to $46,986. The cautionary tale is also proven: Alabama raised trainee pay to ~$57,000 and cut resignations 28% — but hiring collapsed 50% anyway and officer staffing fell 55% over nine years, at a turnover cost of $78,402 per officer and over $11 million a year. Pay alone does not fix a workplace where officers fear for their lives; Georgia must pair recruitment with violence reduction (separation), solitary reduction (to free the double-staffing drain), and honest vacancy reporting. Commissioner Oliver told lawmakers, via GPB, "We don't have a problem with recruiting. We have a problem with the retention piece" — the peer-state evidence says it is both, and that both are fixable. More at gps.press/intelligence/issue/staffing-crisis/.

Conditions and sanitation

Georgia's prison kitchens mostly score A's on scheduled inspections — and then there is Johnson State Prison at 64 (rats and roaches "with little to no change," bulk food bags gnawed through with droppings and urine, five ovens and a tilting skillet and a kettle and a griddle and a freezer and an ice machine all broken), Pulaski State Prison declining 83 → 73 → 78 → 67 (its only designated handwashing sink ripped from the wall, sewage backing through floor drains as a repeat violation, hot-holding failures marked repeat), and Smith State Prison at 68. The system-wide A's prove adequacy is achievable; the failures prove nobody is made to fix anything. The fixes are cheap: unannounced inspections with mandatory repair deadlines and re-inspection, a dedicated maintenance line item that cannot be swept, and publication of full inspection data (GPS already publishes what it obtains). GPS does not yet have system-wide data on housing-unit temperatures; what the record does show is a physical plant in documented collapse and a $600+ million infusion that critics — GPS among them — say funded operational triage rather than structural repair. More at gps.press/intelligence/issue/conditions/.

Death reporting

Covered as Vision 2027 plank 3 because it is that central: Georgia reports deaths roughly two months late, underreports homicides per DOJ, leaves 690 deaths without a determined cause, and left families trading rumors in Facebook groups while the Department posted deputy-warden announcements. A statutory timeline, a public cause-of-death database, mandatory autopsy standards, and family notification with teeth would end the darkness at trivial cost. The mortality record is at gps.press/georgia-prison-deaths/ and the analysis at gps.press/intelligence/issue/deaths-in-custody/.

Oversight

Georgia is one of the largest prison systems in America with no external ombudsman, no corrections inspector general, no oversight commission — the only entity that reviews GDC is GDC, whose 273 consecutive passed PREA audits sit beside a DOJ finding that sexual assault is "rampant." Roughly 19 states plus D.C. have independent oversight; the New Jersey and Washington models above are costed, statutory, and proven. Full analysis at gps.press/intelligence/issue/oversight-investigations/.

The cost of doing nothing

Georgia already pays for failure — quietly, through the state risk-management ledger GPS obtained by open records: at least $50,633,556 across 261+ settled claims from 2008 to 2025, a floor, not a ceiling. The largest entries are deaths, and four of the people the evidence discusses are in GPS's registry:

• David Lamar Henegar died October 16, 2021, at Johnson State Prison, age 44; the registry lists the manner as homicide. His estate received $4,000,000.
• Reginald Jerome Jacobs Jr died February 6, 2023, at Calhoun State Prison, age 24; the registry lists the cause as deliberate indifference. His estate received $3,250,000.
• Juan Carlos Ramirez died July 20, 2023, at Telfair State Prison, age 27; the registry lists the cause as Other. The state paid $3,249,873 in connection with the death at Telfair.
• Thomas Henry Giles died October 28, 2020, at Augusta State Medical Prison, age 31; the registry lists the manner as homicide. His case produced a $3,000,000 payout.

Add Khalid Mouton's $2,998,994 (Georgia State Prison) and Christopher Heath's $1,450,000 (Smith State Prison), and remember the comparison point: California's Ashker plaintiffs' attorney fees alone exceeded $4.55 million plus years of monitoring — several times the annual cost of a New Jersey-style oversight office that exists to catch problems before they become seven-figure settlements. The full ledger analysis is at gps.press/intelligence/issue/legal-settlements/.

Handle the Numbers Honestly

GPS's comparative research library is explicit about its confidence tags, and this page keeps them: recidivism definitions (rearrest, reconviction, reincarceration) and follow-up windows vary across every source, so raw cross-state comparisons mislead. Little Scandinavia and the Amend programs report early, mostly non-causal results — their own researchers stress this. Nordic outcomes reflect shorter sentences and stronger welfare baselines. Alabama proves pay alone does not fix staffing. Texas's GRAD and Arizona's STG step-down lack rigorous modern outcome evaluations, and GPS has flagged the open-records requests needed to close those gaps. What is not in doubt: gang separation cut violence where it was genuinely tried; oversight offices cost less than a single settlement; education returns $4–$5 per dollar; elderly release is near-zero risk; and decarceration at scale has never produced the crime wave its opponents predict.

What You Can Do Now

Sign on. Add your name to Vision 2027 and the End the Warehouse campaign — the two frameworks that carry every solution on this page.

Contact your legislators. Use gps.press/find-your-legislator and ask three questions: Will you demand enforcement of O.C.G.A. § 42-5-52 and publish close-security counts by facility? Will you co-sponsor an independent corrections oversight office with subpoena power? Will you fund vocational education at even 1% of the GDC budget? Legislators count calls; the gang-separation ask costs the state nothing and is the hardest to refuse.

Share the case. The gang-separation reporting — "Separate the Gangs. It Costs Nothing. Georgia Keeps Choosing the Bodies." and "On the Books Since 1897" — is the single most shareable fix on this page, because it requires no new money and no new law. Point people to the evidence: violence, deaths in custody, the budget analysis, medical neglect, solitary confinement, and the settlements ledger.

Add to the record. If you have information from inside a facility, submit a report; families can use the Tell My Story intake. Every verified account sharpens the case that the fixes on this page are overdue.

Georgia wrote a separation statute in 1897, watched other states prove every one of these reforms works, and kept choosing the bodies anyway. The tools exist. The only thing left to build is the pressure to use them — and that part is yours.