The Amplifier: How Five Unelected Officials Turned Georgia’s Prisons Into a Constitutional Crisis

On October 1, 2024, the United States Department of Justice finished an eight-year investigation of Georgia’s prisons and put its conclusion in one sentence: “The State fails to protect incarcerated people from violence and harm by other incarcerated people in violation of the Eighth Amendment.” ¹ The same report found correctional officer “vacancy rates around 50% systemwide and over 70% at ten of the largest facilities.” Georgia reported 310 homicides in its custody to the federal government across fiscal years 2020 through 2024, more than California, Florida and Texas combined. ² On September 4, 2026, the state held 52,430 people. ³

Everyone knows who wrote the sentences. The General Assembly did, in public, on the record, in votes that can be looked up. GPS has documented what those votes cost in Georgia’s $40 Billion Mistake and what they did to the promise of release in The Illusion of Parole. Those two articles are the foundation of this one.

This article is about the part nobody voted for.

Between the sentence a judge pronounces and the day a person walks out of the gate, there’s a second set of rules. They decide how much of that sentence is served inside a prison and how much is served outside it, under supervision. They aren’t statutes. They are made by five people the public doesn’t elect, and for twenty-eight years those five people have used them to move the line in one direction only: more of every sentence behind the wall, less of it in the community. Not once. Three times, each time stacking on top of the last, and once by building a mechanism that moves the line automatically whenever the legislature raises a sentence, with no vote by anyone at all.

That is the amplifier. It is why a set of laws the Southern Center for Human Rights called a policy disaster in 2006 has produced, twenty years on, a prison system the federal government says is unconstitutional, and why the people leaving those prisons come home more likely to hurt someone than when they went in.

How a sentence works

A twenty-year sentence was never twenty years in a cell. It is twenty years under the authority of the state. The judge often splits it on the day it’s pronounced, so many years in prison and the rest on probation. For the prison part, Georgia law sets the point at which the state may start moving a person from custody to supervision: a person serving a felony sentence “shall only be eligible for consideration for parole after the expiration of nine months of his or her sentence or one-third of the time of the sentences, whichever is greater.” ^{4 5}

That one-third line is the legislature’s. Past it, the sentence doesn’t end. It continues outside, under a parole officer, with conditions, and it can be revoked. Parole isn’t a discount on the sentence. It’s the second half of it, designed to be served where a person can hold a job and be watched at a fraction of what a prison bed costs.

Deciding when that move happens is the Parole Board’s entire job. Its own rule says so: for anyone serving less than life, “the Board applies the Parole Decision Guidelines to the particular circumstances of the inmate’s case and notifies the inmate that he or she is denied parole throughout his or her confinement or notifies the inmate of a tentative parole month for some future time.” ⁶

Society decided that people should come home under supervision before their sentence ends. The Board is paid to make that happen. What follows is the record of how it stopped.

Who holds the dial

The Georgia Constitution creates the Board in a single paragraph, and the Board describes its own membership the same way: “The members are appointed by the Governor to a seven year term subject to confirmation by the State Senate.” ⁷ No election. No term that a voter can end.

The guidelines are a grid. Down one side, the severity of the crime. Across the top, a score. In each cell, a number of months, or a percentage of the sentence. The Board writes the grid, and the Board rewrites it when it likes. The rule that carries it has been amended thirteen times since 2001. ⁸

Layer one, 1998: a policy laid on top of the statutes

In 1994 and 1995 the legislature did its part. Mandatory minimums for seven serious violent felonies. Life for a second one. Parole abolished for the offences the statute named. The $40 Billion Mistake covers those laws and the $82,211,036 in federal Truth in Sentencing grants that rewarded them. ⁹

The legislature stopped at seven crimes. The Board didn’t.

On December 9, 1997, the Board adopted a policy of its own: people convicted of twenty further violent felonies would serve at least ninety percent of their sentence before a first parole review. Robbery. Aggravated assault. Residential burglary. Voluntary manslaughter. Statutory rape. Cruelty to children. None of those twenty had been touched by the 1994 law. The Board’s stated aim, in its own words at the time, was to “make Georgia the toughest state in the nation for time-served for violent felonies.” ¹⁰

By fiscal year 2004 that policy had governed 12,790 parole decisions. Thirty-seven percent of those people served every day of their sentence. Forty-eight percent served ninety. ¹¹

It was never a rule. The Board adopted it at a meeting and applied it for eight years without putting it through the process Georgia law requires. A federal judge held in 2002 that applying it to people whose crimes predated it was unconstitutional. Then a Fulton County judge went further. The Board recorded that ruling itself, in its annual report for fiscal year 2005:

“The new policy came about after a Fulton County Judge ruled that ‘the Board’s 90% policy in 1997 was implemented improperly and as a result is ineffective and has no force or effect.'” ¹²

Twelve thousand seven hundred and ninety decisions under a policy with no force or effect. That was layer one.

Layer two, 2006: the percentage that multiplies the sentence

The Board’s answer to the court was not to stop. It was to make the policy durable. On December 9, 2005, eight years to the day after the original policy, the Board filed a rule creating an eighth Crime Severity Level. Levels one through seven recommend months. Level eight recommends a share of the sentence: sixty-five, seventy-five or ninety percent, by score. Twenty-three offences were placed in it. ¹³

The Board will say, correctly, that for twenty of those offences the new rule was softer than the flat ninety percent it replaced. For the highest-scoring band it was the same, and for the rest it was lower.

What it also did was change the unit. A months band doesn’t care how long the sentence is. A percentage does. From January 2006, for every offence in Level eight, the Board’s floor became a fraction of whatever number the legislature chose to write into the statute. The Board had built a multiplier and pointed it at the General Assembly.

Six months later the General Assembly supplied the other factor. House Bill 1059, signed on April 26, 2006 and effective July 1, raised mandatory minimums across the sex offences by two and a half to ten times, and raised the maximums with them. ¹⁴ Rape went from a minimum of ten years to twenty-five, and a maximum of twenty to life. Aggravated child molestation, the same. Incest went from one to twenty years to ten to thirty.

Take incest, because it sits in both instruments. Before 2006, a person convicted of it faced a sentence of one to twenty years, and the Board’s grid put the time to serve at thirty-four to fifty-two months. After July 2006, the same crime carries ten to thirty years, and Level eight requires sixty-five to ninety percent of it. A ten-year sentence, the new minimum, now means six and a half to nine years inside. The legislature raised the sentence. The Board’s rule turned that raise into time actually served, automatically, because the rule was written as a percentage. Neither body voted on the combined result. Nobody had to.

Layer three, 2008: the ratchet

Two years later the Board rebuilt the whole grid and, for the first time, wired it to the courts. It filed the change on December 10, 2007 and explained it in its next annual report:

“These new Guidelines represent, for the first time, a linkage with the statewide-average length of prison sentences imposed by Superior Court Judges.” ¹⁵

The rule says how. The minimum recommendation for each severity level “represents one-third, or more, of the Statewide-average prison sentence for all crimes assigned that Crime Severity Level.” That sentence is still in the rule today.

Read it as a machine. When judges hand down longer sentences for a category of crime, the Board’s floor for that category rises with them. When the legislature raises a maximum and judges use it, the floor rises again. No Board vote. No legislative vote. No public notice. A number that governs how long a human being stays in a Georgia prison now moves on its own, and the only people who could stop it are the five who built it.

The Board also said what it had checked before doing this. The amendments “were preceded by a three-year study and analysis of risk factors utilized in granting clemency to offenders, past clemency practices, and the effects of the new guidelines on prison capacity.”

The Board studied what its ratchet would do to prison capacity. It has never published the study. And in the same week the ratchet took effect, the Board’s public web page dropped the one sentence that told people what raised their risk score. GPS recovered that sentence from a web archive. It isn’t on the Board’s site now.

The rule that looks like a limit

One rule reads like a brake on the grid. Rule 475-3-.06 says that if the guidelines recommend denying parole for the whole sentence, “the inmate will be considered for parole at the expiration of one-third of the sentence or sentences.” ⁶

Read it again. Considered. Not released. The statute already makes a person eligible for consideration at one third. The rule promises that the Board will look at the file on a date the legislature had already fixed, and nothing more. The Board can look and say no, and the guidelines rule says in terms that it may. The guidelines system, in the rule’s own words, “does not create a liberty interest,” and the Board “may make an independent decision to deny parole or establish a Tentative Parole Month at any time prior to sentence expiration.” ¹⁶

So the rule doesn’t limit the grid. It guarantees a review, and a review costs the Board nothing. That’s why it has sat untouched since November 2004 while the guidelines rule beside it was amended thirteen times: there was nothing in it the Board needed to change. The Board’s own website says what happens at that review. “Few inmates are granted parole at their initial eligibility date.” ⁵

Nothing outside the Board limits the grid. The legislature set the eligibility date. Everything after it belongs to five people.

What the amplifier does to the communities it claims to protect

The case for all of this was public safety. The evidence runs the other way, and it comes from Georgia’s own data.

In 2013 the Quarterly Journal of Economics published Ilyana Kuziemko’s study of 78,393 Georgia prisoners, built around the Board’s 1998 ninety percent policy. Removing the prospect of earlier release didn’t make prisons calmer. Disciplinary infractions rose about four percent a month, a fifteen percent relative increase. Program completion fell twelve percent and enrollment thirteen. And the people released under the policy came back at a higher rate: a 2.4 to 2.5 percentage point increase in three-year recidivism, eight percent relative. ¹⁷

Put those numbers next to the Department of Justice’s. Prisons where half the officer posts are empty and the men inside have been told that nothing they do changes their release date. The federal government calls the result an Eighth Amendment violation. The study calls it predictable. Both are describing the same building.

A person who serves ninety percent of a sentence in a prison like that doesn’t disappear. He comes home to a Georgia neighbourhood, later, angrier, and measurably more likely to offend again. The amplifier doesn’t protect the public from him. It manufactures him and then delivers him.

Nobody has checked

Georgia has a State Auditor. GPS indexed every report the office has published since 2002, 2,216 of them. ¹⁸ Not one title contains the word parole or pardon, and not one names the Board of Pardons and Paroles as the agency examined. The last audit with the Department of Corrections as its subject was published on December 19, 2018.

Twenty-eight years of a five-member board rewriting how much of a sentence Georgians serve, and no arm of the state has ever looked at whether it worked.

Take the dial back

The legislature can be voted out. The Board cannot. That is the whole point of the design, and it’s why the amplifier belongs to the Board and not to the people who pay for its consequences.

A General Assembly that wants to know why its prisons are unconstitutional, why they cost what they cost, and why the people leaving them are more dangerous than the people who entered, can start with one rule. Read 475-3-.05. Ask who approved the linkage to average sentences, what the three-year capacity study found, and why the one rule that looks like a limit guarantees a review and nothing else. Then decide whether five appointees should keep a dial the legislature never knew it had handed over.

---

Call to Action: What You Can Do

Awareness without action changes nothing. Here’s how you can help push for accountability and real reform:

Join the GPS Advocacy Network: Sign up at https://gps.press/become-an-advocate/ and we’ll advocate on your behalf every week. GPS identifies your state legislators, crafts personalized letters on the most pressing prison issues, and sends them directly to the representatives who represent you. You receive a copy of every letter. It takes two minutes to sign up, and we handle the rest.

Send a 60-Second Message: Pick an issue, get a ready-to-edit message with the verified facts already in it, and email your state House representative and senator directly from your own inbox at https://gps.press/send-a-message/. No signup, nothing stored. It takes about a minute.

Tell My Story: Are you or a loved one affected by Georgia’s prison system? GPS publishes first-person accounts from incarcerated people and their families. Submit your story at https://gps.press/category/tellmystory/ and help the world understand what’s really happening behind the walls.

Contact Your Representatives: Your state legislators control GDC’s budget, oversight, and the laws that created these failures. Find your Georgia legislators at https://gps.press/find-your-legislator/ or call Governor Kemp at (404) 656-1776 or the GDC Commissioner at (478) 992-5246.

Demand Media Coverage: Contact newsrooms at the AJC, local TV stations, and national criminal justice outlets. More coverage means more pressure.

Amplify on Social Media: Share this article and tag @GovKemp, @GDC_Georgia, and your local representatives. Use #GAPrisons, #PrisonReform, #GeorgiaPrisonerSpeak.

Message the Board of Corrections: The citizen board that oversees the GDC publishes no contact details for its members. GPS built the missing front door: pick a board member and a policy message and send it from your own email at https://gps.press/write-the-board/.

File Public Records Requests: Georgia’s Open Records Act gives every citizen the right to request incident reports, death records, staffing data, medical logs, and financial documents. Start with the GPS Action Toolkits at https://gps.press/toolkits/. The Open Records Toolkit walks you through what to ask for, gives you ready-to-file request templates, and tells you what to do when the state says no. File directly with GDC at https://georgiadcor.govqa.us/WEBAPP/_rs/SupportHome.aspx.

Attend Public Meetings: The Georgia Board of Corrections and legislative committees hold public meetings. Your presence is noticed.

Contact the Department of Justice: File civil rights complaints at https://civilrights.justice.gov. Federal oversight has forced abusive systems to change before.

Support Organizations Doing This Work: Donate to or volunteer with Georgia-based prison reform groups fighting for change on the ground.

Vote: Research candidates’ positions on criminal justice. Primary elections often determine outcomes in Georgia.

Contact GPS: If you have information about conditions inside Georgia’s prisons, reach us securely at GPS.press.

---

Part of Something Bigger

This article is part of the GPS Reform Agenda, two active campaigns to transform Georgia’s criminal justice system.

Vision 2027 THIS SERIES

Three model bills for the 2027 Georgia legislature. The legislature doesn’t need new laws. It needs to enforce two dormant statutes it already passed.

End the Warehouse

Transform Georgia’s prisons from punishment to rehabilitation. Two tracks: litigation to reduce overcrowding + evidence-based programs that work.

Read the full GPS Reform Agenda

---

Further Reading

Georgia’s $40 Billion Mistake: How Bad Science and Federal Bribes Created a Constitutional Crisis

The legislature’s side of the ledger: the 1994 to 1998 laws, the federal money that rewarded them, and the peer-reviewed research showing what they did.

The Illusion of Parole

How four overlapping policies between 1995 and 2006 turned parole in Georgia into a promise the state rarely keeps.

The Board That Stopped Saying Yes

Georgia’s lifer parole collapse, told entirely in the Parole Board’s own published numbers.

Georgia’s Shadow Sentencing System

The decisions that add years to a sentence after a judge has finished speaking.

The Seven-Year Promise

Four decades of a parole system that keeps rewriting what eligibility means.

---

GPS Intelligence System

The GPS Intelligence System maintains living research profiles that aggregate data, news, settlements, and analysis on Georgia’s prisons and the issues defining them. The profiles below provide deeper context for the issues raised in this article:

Oversight and Investigations

A running record of who is supposed to be watching Georgia’s prison system, and what they have and haven’t examined.

End the Warehouse

The evidence base on overcrowding and capacity, which is what a release instrument ultimately governs.

---

Explore the Data

GPS makes GDC statistics accessible to the public through several resources:

• GPS Statistics Portal: Interactive dashboards translating complex GDC reports into accessible formats, updated within days of official releases.
• GPS Lighthouse AI: Ask questions about Georgia’s prison system and get answers drawn from GPS’s investigative archive and data analysis.
• GPS llms.txt: A single machine-readable index of every GPS data resource, published using the open llms.txt standard. Point any AI tool (ChatGPT, Claude, Gemini, Perplexity) at this URL and the model can navigate to facility profiles, intelligence briefs, mortality records, statistics, and the full investigative archive, with no other configuration needed. It is the fastest way to ground an AI conversation in verified GPS data.
• GPS Public MCP Server: For AI tools that support the Model Context Protocol (Claude, ChatGPT, and others), connect directly to GPS’s data: 12 read-only tools covering population, facilities, mortality, parole, the Georgia Code, settlements, and more. Every answer arrives with its source link and citation built in.

For a walkthrough of how to put these resources to work with AI, see How to Use GPS Data with AI Tools: a step-by-step guide for researchers, advocates, families, and journalists analyzing Georgia prison conditions, statistics, and policy with tools like ChatGPT, Claude, and Gemini.

Contact GPS at media@gps.press for access to underlying datasets used in this analysis.

---

About Georgia Prisoners’ Speak (GPS)

Georgia Prisoners’ Speak (GPS) is a nonprofit investigative newsroom built in partnership with incarcerated reporters, families, advocates, and data analysts. Operating independently from the Georgia Department of Corrections, GPS documents the truth the state refuses to acknowledge: extreme violence, fatal medical neglect, gang-controlled dorms, collapsed staffing, fraudulent reporting practices, and unconstitutional conditions across Georgia’s prisons.

Through confidential reporting channels, secure communication, evidence verification, public-records requests, legislative research, and professional investigative standards, GPS provides the transparency the system lacks. Our mission is to expose abuses, protect incarcerated people, support families, and push Georgia toward meaningful reform based on human rights, evidence, and public accountability.

Every article is part of a larger fight: to end the silence, reveal the truth, and demand justice.

---

Footnotes

1. DOJ findings report on the investigation of Georgia prisons October 1 2024, https://www.justice.gov/d9/2024-09/findings_report_-_investigation_of_georgia_prisons.pdf [↩]
2. Bureau of Justice Assistance Death in Custody Reporting Act transparency tables, https://bja.ojp.gov/program/dcra/reported-data [↩]
3. GDC Friday Report for 4 September 2026, https://gdc.georgia.gov/document/standing-statistical-reports/fridayreport20260904pdf/download [↩]
4. O.C.G.A. section 42-9-45 as carried in the GPS Georgia Code corpus, https://gps.press/ocga/42-9-45/ [↩]
5. State Board of Pardons and Paroles page on the parole process, https://pap.georgia.gov/parole-consideration/parole-process-georgia [↩][↩]
6. Georgia Secretary of State rules archive Rule 475-3-.06, https://rules.sos.ga.gov/GAC/475-3-.06 [↩][↩]
7. State Board of Pardons and Paroles press release on the chairman 23 June 2020, https://pap.georgia.gov/press-releases/2020-06-23/barnard-serve-again-chairman [↩]
8. Georgia Secretary of State rules archive Rule 475-3-.05 history, https://rules.sos.ga.gov/GAC/475-3-.05 [↩]
9. Georgia Department of Corrections Standing Report on Truth in Sentencing, https://gdc.georgia.gov/document/standing-special-analyses/standing-report-truth-sentencing/download [↩]
10. Prison Legal News on the Georgia Parole Board 90 percent policy, https://www.prisonlegalnews.org/news/2003/apr/15/georgia-parole-boards-quot90-policyquot-ruled-ex-post-facto/ [↩]
11. Georgia State Board of Pardons and Paroles FY2004 Annual Report page 26, https://pap.georgia.gov/document/document/annual-report-2004/download [↩]
12. Georgia State Board of Pardons and Paroles FY2005 Annual Report page 10, https://pap.georgia.gov/document/document/annual-report-2005/download [↩]
13. Crime Severity Levels chart for crimes committed on or after 1 January 2006 archived copy, https://web.archive.org/web/20110205200940id_/http://www.pap.state.ga.us/opencms/export/sites/default/resources/Crime_Severity_Levels_01-2006.pdf [↩]
14. Georgia General Assembly HB 1059 status page archived copy, https://web.archive.org/web/20080709022424id_/http://www.legis.ga.gov/legis/2005_06/sum/hb1059.htm [↩]
15. Georgia State Board of Pardons and Paroles FY2008 Annual Report page 18, https://pap.georgia.gov/document/document/annual-report-2008/download [↩]
16. Georgia Secretary of State rules archive Rule 475-3-.05, https://rules.sos.ga.gov/GAC/475-3-.05 [↩]
17. Kuziemko How Should Inmates Be Released From Prison Quarterly Journal of Economics 2013, https://doi.org/10.1093/qje/qjs052 [↩]
18. Georgia Department of Audits and Accounts report search, https://www.audits.ga.gov/ReportSearch [↩]