How Local Jails Became America’s Most Dangerous Institutions

Designed for brief detention, local jails now warehouse mental illness, addiction, and poverty, yet they lack the capacity to keep people alive.

They were arrested, booked, and placed in a cell. They never stood before a judge. They never heard a verdict. Within days, or sometimes only hours, they were dead.

This sequence appears again and again in local reporting across the United States. In Mississippi, people jailed “for their own safety” while awaiting psychiatric beds died in county custody, many by suicide, according to reporting by Mississippi Today and ProPublica. In Kentucky, the Lexington Herald-Leader found that hundreds of people have died in local jails since 2020, most of them awaiting trial. In Georgia, the U.S. Department of Justice concluded that conditions in the Fulton County Jail were so dangerous they violated the Constitution.

These deaths tend not to produce national reckonings. They happen in county jails: facilities most Americans never see, think about, or easily distinguish from prisons. They are framed as tragic anomalies, administrative errors, or the unfortunate result of illness and addiction. But taken together, they tell a different story. 

Local jails have quietly become some of the most dangerous institutions in American life – not because they are malfunctioning, but because they are being asked to perform work they were never built to do.

Investigations into Local Jails Across the Country Reveal Unsettling Patterns

While conditions vary across jurisdictions, investigations from Mississippi to Washington State reveal patterns that are difficult to dismiss as isolated failures. Jails now operate as de facto mental health facilities, effectively the nation’s largest providers of care for the mentally ill. Policy analysis from The Bail Project similarly concludes that jails are structurally incapable of functioning as mental health providers, even as they increasingly serve that role by default.

This reality has only recently begun to receive sustained scrutiny, including in Anyplace But Here: The Uncomfortable Convergence Between Mental Illness and the Criminal Justice System by former sheriff Tony Thompson. Thompson describes how facilities designed for brief detention have become warehouses for mental illness, addiction, and poverty, all while lacking the staffing, medical capacity, and institutional stability required to keep people alive.

The first step is not violence; it is depersonalization – the stripping away of individuality that makes neglect feel normal.

One reason this danger can persist, even when it is widely understood, is that jails operate as what sociologist Erving Goffman called “total institutions” – environments that reorganize social life through control, routine, and isolation from the outside world. Vincent Schiraldi, a fellow at the Pinkerton Foundation and formerly chief executive of corrections, probation, and youth justice systems in New York City, Washington, D.C., and Maryland, describes this as a process that changes how harm becomes thinkable. “In total institutions, we rob people of their humanity,” he told The Bail Project. The first step is not violence; it is depersonalization – the stripping away of individuality that makes neglect feel normal. 

Jails Don’t Merely Contain Dangerous Conditions; They Produce Them

That framework also helps explain why jail crises cannot be reduced to a few “bad guards” or a few rogue administrators. Schiraldi – who spent much of his career as an advocate before running systems from the inside – said he once believed staff were “scumbags,” the stock villains of prison movies. But the reality, he found, was more unsettling: “They viewed themselves as the good guys. And yet, simultaneously, they were tolerating stuff they would never tolerate for a second in the community.” In other words, ordinary people – people who coach youth sports, go to church, raise families – can become participants in extraordinary harm, not always through direct cruelty, but through acquiescence.

Schiraldi’s explanation is not about individual depravity so much as social psychology under pressure. “It’s the Milgram study. It’s the Stanford prison experiment,” he said, invoking the classic findings that authority and group consensus can pull people toward behavior they would otherwise reject. “Under the cloak of authority – or authority-feeling consensus – people will do bad stuff.” 

In jail, that “authority-feeling consensus” can become its own code, stronger than policy manuals or constitutional standards. New hires step into a world where the practical lesson is not what the rules say, but what the culture expects: when to intervene, when to use force, when to ignore. And when staffing collapses, oversight weakens, and people arrive already in crisis, that culture does not merely shape conditions – it shapes whether people survive.

[T]hese facilities do not simply contain danger. They produce dangerous conditions through chronic understaffing, inadequate medical care, and the normalization of neglect.

Taken together, Schiraldi’s point is a grim extension of the larger story jail-death investigations keep telling: these facilities do not simply contain danger. They produce dangerous conditions through chronic understaffing, inadequate medical care, and the normalization of neglect.

The scope of the problem is difficult to quantify precisely, an irony that is itself revealing. Federal law requires states to report deaths in custody to the Bureau of Justice Statistics. But compliance is inconsistent, a fact documented by the Vera Institute of Justice in its report “The Hidden Deaths in American Jails.” Some states underreport deaths. Others submit data years late. In Mississippi, Mississippi Today and The Marshall Project found that dozens of jail deaths never appeared in official counts at all.

Independent efforts have filled some of the gaps. Vera’s analysis shows that jail deaths increased sharply during the COVID-19 pandemic and have remained elevated even as jail populations declined. In Kentucky, the Herald-Leader reported that roughly 70% of people who died in jail had not been convicted of a crime. In Washington State, a Columbia Legal Services report titled, “Gone But Not Forgotten” found that more than 70% of jail deaths occurred within the first two weeks of incarceration.

What disappears when deaths go uncounted is not just accountability, but pattern recognition. Without reliable data, each death can be treated as an isolated tragedy rather than evidence of systemic failure. The opacity protects institutions from scrutiny and allows the same conditions to persist.

Fulton County Jail in Atlanta, GA

The Most Dangerous Moments in Jails

To understand why jails are so lethal, it helps to understand how they differ from prisons. Prisons hold people who have been convicted and sentenced, often for years. Jails are meant to be temporary. They house people awaiting trial, serving short sentences, or detained for administrative reasons. Their populations turn over rapidly, explaining why, in 2024, while there were 657,500 people in jails throughout the country, 7.9 million people entered and exited jails that year. Intake is constant. Medical histories are incomplete. People arrive in crisis: detoxing, psychotic, suicidal, injured, or chronically ill.

The most dangerous moment in jail is not month three or year one. It is the first few days.

The most dangerous moment in jail is not month three or year one. It is the first few days. That first window is when many clients assisted by The Bail Project describe the most acute instability: withdrawal without monitoring, missed psychiatric medication, or placement in isolation before any meaningful medical follow-up occurs.

This pattern appears repeatedly in investigations and audits. The Washington State report documented deaths from withdrawal and suicide shortly after booking. A California state audit found that its jails performed insufficient safety checks. In Minnesota, the Department of Corrections ordered Hennepin County to reduce its jail population after finding that missed welfare checks contributed directly to deaths related to medical emergencies, substance withdrawal, and suicide. New York City’s Department of Correction reports that incarcerated people missed nearly 15,000 medical appointments in December alone.

Yet intake in most jails remains an administrative process rather than a medical one. Screening is rushed. Staff are overextended. Follow-up is inconsistent. Safety checks that are supposed to occur every 15 or 30 minutes are missed for hours at a time. Clients released with Bail Project support pretrial frequently recount requests for help that go unanswered or watching others in visible medical distress wait extended periods before assistance arrives.

In prisons, systems are built, however imperfectly, for long-term management. In jails, everything happens at once, and nothing stays still long enough for problems to be absorbed. The system depends on vigilance. What it gets instead is triage.

The Effects of Understaffing Jails

Nearly every investigation into jail deaths eventually arrives at the same root cause: staffing.

Local jails across the country are chronically understaffed, a finding echoed in audits from Wisconsin, Texas, Florida, Oklahoma, and North Carolina. In Bexar County, Texas, officials acknowledged that fewer officers were overseeing increasingly volatile populations. In Leon County, Florida, a consultant warned that jail design combined with staffing shortages created potentially dangerous situations for both detainees and staff.

When staffing collapses, every safeguard collapses with it. Wellness checks are skipped. Medical complaints are dismissed. Emergencies are missed.

When staffing collapses, every safeguard collapses with it. Wellness checks are skipped. Medical complaints are dismissed. Emergencies are missed. Violence increases, not because people in custody are uniquely violent, but because supervision breaks down.

A Department of Justice investigation in South Carolina found that people in jail faced a substantial risk of being stabbed, raped, or beaten due to inadequate staffing and supervision. In Oklahoma County, repeated health inspections failed after inspectors found cells dangerously overcrowded and staff unable to monitor detainees properly.

But understaffing does not occur in a vacuum. Jails are overwhelmed because they are filled beyond necessity, housing people who could safely await trial in their communities. 

Research outlined in The Bail Project’s Inside Bail Reform: Six Core Components of Safe, Fair, and Effective Pretrial Policy identifies proven alternatives – such as pretrial screening, court reminders, and voluntary support services – that reduce detention without increasing failure-to-appear rates, underscoring that population size is a policy choice, not an inevitability.

No staffing model can make mass pretrial detention safe. The Bail Project’s work reflects this reality: the most reliable way to reduce harm inside jails is not to manage overcrowding more efficiently, but to prevent unnecessary detention altogether.

Jails Are America’s Largest Mental Health Institutions

Nowhere is this clearer than in the treatment of mental illness. In some cases, people arrested during psychiatric crises are booked into jail before any hospital bed is located. The Bail Project has assisted clients who were placed in solitary confinement “for safety” while awaiting transfer, only to deteriorate further inside.

Jails have become America’s largest mental health institutions by default, a reality documented by the The Bail Project and reinforced by local reporting nationwide. People experiencing psychosis, suicidal ideation, or severe addiction are routinely incarcerated because there is nowhere else to put them. Psychiatric beds are scarce. Community treatment is underfunded. Police become first responders to crises they are not trained to resolve.

This is not a story of individual malice. It is a story of systems that cannot function as designed under the pressures placed upon them. Officers are tasked with roles that combine security guard, nurse, social worker, and crisis counselor. Failure is inevitable. Harm flows downward.

Once inside jail, care is often fragmented or nonexistent.

Once inside jail, care is often fragmented or nonexistent. Medications are delayed or interrupted. Suicide watch is implemented without adequate monitoring. Isolation replaces treatment. In Mississippi, people jailed while awaiting mental health treatment died in custody, despite being detained explicitly for their own protection. In Fulton County, the U.S. Department of Justice found detainees suffering from untreated illness, malnutrition, and mental health crises in conditions so severe they violated constitutional standards.

Additional Jails Aren’t the Solution

When deaths pile up, the most common political response is a familiar one: build a new jail.

Maine officials commissioned a report suggesting a new jail could address safety failures. Washington, D.C.’s auditor declared the city desperate for a new facility after finding that the jail’s death rate was three times the national average. 

But buildings do not determine who enters a jail, how long they stay, or whether they receive care. New jails do not fix bail systems that detain people for poverty. They do not create clinicians where none exist. They do not resolve staffing shortages driven by low pay and high burnout. Capital projects absorb blame that properly belongs to policy. They offer visible action without addressing intake, oversight, or accountability.

The people most at risk in jails are not those convicted of serious crimes. They are people charged with misdemeanors, people awaiting hearings, and people too poor to post bail.

The people most at risk in jails are not those convicted of serious crimes. They are people charged with misdemeanors, people awaiting hearings, and people too poor to post bail. Many face bond amounts in the hundreds or low thousands of dollars – sums that are relatively minor in policy debates but insurmountable for someone living paycheck to paycheck.

This pattern appears in state after state, and client after client at The Bail Project. 

Oklahoma County Detention Center in Oklahoma City, OK

Reducing Pretrial Detention in Jails Is a Matter of Survival

In Kentucky, most jail deaths involved people awaiting trial. In Duval County, Florida, local reporting linked recent deaths to long-standing problems that disproportionately affected people held pretrial. In Milwaukee County, an external audit found dangerous suicide-watch practices affecting people who had not been convicted.

They are legally innocent. Yet pretrial detention routinely functions as punishment for poverty, separating people from housing, employment, and medical care before any determination of guilt. The state’s obligation to them is not punishment; it is custody with care. Instead, pretrial detention functions as punishment in practice. Time stretches. Court dates are delayed. Conditions worsen.  Deaths are often treated as administrative matters rather than the constitutional failures they represent.

In this context, freedom is not an abstraction. It is a concrete intervention against the conditions that make local jails so deadly.

This is why efforts to reduce pretrial detention matter not only as questions of fairness, but of survival. Data from charitable bail funds like The Bail Project consistently show high court-appearance rates and improved stability for people released pretrial – particularly those with medical, mental health, or substance-use needs – challenging the assumption that detention is necessary for compliance or safety. In this context, freedom is not an abstraction. It is a concrete intervention against the conditions that make local jails so deadly.

Accountability would require more than outrage after the fact. It would begin with mandatory, public reporting of all deaths in custody, as urged by the Vera Institute and civil rights groups. It would require independent medical oversight rather than voluntary inspections. It would demand staffing minimums tied to population caps, along with the political will to jail fewer people when those caps are exceeded.

Fewer people in jails means fewer deaths. That is not a radical claim; it is an empirical one.

Most of all, it would require a rethinking of pretrial detention itself. Fewer people in jails means fewer deaths. That is not a radical claim; it is an empirical one.

Bail Reform Means Fewer People Cycling through Dangerous Local Jail Environments

As policy reporting on bail reform has documented, jurisdictions that reduce reliance on cash bail while investing in court reminders and voluntary pretrial supports have not seen the catastrophic public-safety outcomes often predicted. What they have seen is fewer people cycling through unstable jail environments.

America has spent decades debating prisons: sentencing, mass incarceration, and reform. Meanwhile, its jails have quietly become sites of routine, preventable death. Temporary facilities have taken on permanent risk. Administrative detention has acquired mortal consequence. What happens in jails happens quickly, out of sight, and often without record. But it happens in our name.