ANALYSISApril 13, 2026
Should long-term solitary confinement be banned?
The debate over banning long-term solitary confinement has intensified in 2025, with federal legislation reintroduced in Congress, a major correctional officers' strike in New York resulting in the suspension of the state's 2021 solitary confinement ban, and Virginia's governor vetoing a bill to restrict the practice. On any given day, more than 80,000 to 120,000 people in U.S. prisons and detention facilities are held in solitary confinement.
When isolation lasting months or years reliably produces psychosis, self-harm, and permanent psychiatric damage, does the state have a constitutional — or simply moral — obligation to stop using it, even against the most dangerous prisoners? And if we ban it, what replaces it?
Not familiar with this story? Get context →
Colorado data as proof of concept
C
Colorado reduced solitary use by 90% and saw prison violence drop 40%, recidivism improve, and millions saved. If you believe in outcomes and fiscal discipline, the conservative case against long-term solitary writes itself — this isn't ideology, it's what the numbers say happened.
L
We're in agreement on the Colorado data, which makes it worth pressing on why you'd then resist a federal standard. If the results are that clear, what's the argument for letting other states keep doing what Colorado proved doesn't work?
C
Colorado succeeded without a federal mandate — that's the point. The lesson is 'fund alternatives and let states lead,' not 'Washington should write the rule.' Replicating Colorado's outcome requires Colorado's process, not a four-hour cap imposed from above.
L
States have had 40 years since Marion Federal Penitentiary industrialized this practice. Voluntary replication hasn't happened — Colorado is the exception, not the trend, which is precisely why a federal floor matters.
Suicide rate as mortality signal
C
When 8% of the prison population accounts for nearly half of all Bureau of Prisons suicides, you are not looking at a policy side effect — you are looking at a structural failure. That is a mortality signal so large that no single countervailing study can neutralize it.
L
That 8% suicide statistic is exactly right, and it's worth naming what it means: we have built a system that concentrates death among the people we isolate most completely. The Colorado longitudinal study finding limited psychological effects doesn't touch this — those are affect scores, and the harm is showing up in bodies.
C
Agreed on the suicide data — that's why I said reform advocates are right to rest their case on it rather than claiming research unanimity. Where we differ is the remedy, not the diagnosis.
L
If the diagnosis is this unambiguous and we've agreed on it, the burden shifts entirely to opponents of reform to explain what we're waiting for.
Federal mandate versus state-led reform
C
The End Solitary Confinement Act's four-hour cap applied uniformly to federal prisons, jails, and immigration detention ignores genuine variation in institutional environments and population dangerousness. There is meaningful distance between endorsing the UN Mandela Rules' 15-day threshold and endorsing a mandate written by people who have never managed a prison.
L
You're treating the four-hour cap and the 15-day threshold as if choosing between them is the real debate, but Governor Youngkin just vetoed the 15-day bill — the UN's own standard for cruel treatment. If even that floor is politically unsustainable at the state level, what exactly is state-led reform supposed to deliver?
C
One governor's veto doesn't prove state-led reform is impossible — it proves Virginia's current politics. Colorado didn't need a federal mandate, and neither did the other states that have successfully reduced solitary.
L
Colorado is one state. Virginia, Texas, and Louisiana are also states. The federal floor exists precisely because we don't accept a race to the bottom on constitutional rights depending on zip code.
New York crisis as cautionary example
C
New York's 22-day wildcat strike by 15,000 corrections officers destabilizing 42 prisons is not implementation noise — those are workers expressing real fear, and a poorly resourced reform that endangers officers and inmates alike is a genuine indictment of how reform gets done, not just how fast.
L
You're right that it can't be waved away, but the variable that separates New York from Colorado isn't the principle of limiting isolation — it's sequencing. New York removed the tool before funding the alternatives; Colorado invested first. The lesson is 'mandate the investment alongside the limit,' not 'abandon the limit because one state got the order wrong.'
C
That framing makes the sequencing problem sound like a technical footnote, but a court had to step in on July 2 to reverse the suspension — meaning the reform nearly collapsed entirely. 'Get the order right' is harder than it sounds when institutional actors can force a crisis.
L
Institutional actors forcing a crisis to kill democratic legislation passed by a supermajority is an argument for more robust implementation, not for giving institutional actors a veto over reform.
Unchecked state power over isolation
C
Conservatives should be deeply uncomfortable with any government placing a human being in sensory deprivation for years with minimal judicial review. The Supreme Court recognized the psychological destruction of this practice in 1890 — limited government means limited government especially when the subject is unpopular.
L
The Burke framing is useful here, and I'd push it further: what you're describing isn't just philosophically troubling, it's constitutionally suspect. Madrid v. Gomez in 1995 found isolating mentally ill prisoners at Pelican Bay violated the Eighth Amendment, and the 2025 New York federal order extends that logic. The constitutional trajectory is clear even if the Supreme Court hasn't drawn the final line.
C
The constitutional trajectory argument actually supports my position more than yours — courts can draw those lines through case-by-case review, which is exactly the kind of limited, evidence-grounded constraint that doesn't require a sweeping federal mandate.
L
Case-by-case litigation means the people most harmed wait decades for relief while each new case gets litigated from scratch — that's not a check on power, that's a managed delay.
Conservative's hardest question
New York's HALT Act crisis — in which restricting solitary without adequately funding alternatives contributed to a 22-day strike destabilizing 42 prisons — is a genuine cautionary example that reform advocates cannot simply wave away. If the policy is right but its implementation creates dangerous conditions for officers and inmates alike, the practical case for federal mandates is weakened, and critics who argue that pace and resources matter as much as principle have a point that is genuinely hard to dismiss.
Liberal's hardest question
A 2023 Colorado longitudinal study found limited measurable effect on mental well-being among people in segregation, which complicates the uniform psychological harm narrative and gives opponents a peer-reviewed foothold to contest the research consensus. This cannot be dismissed outright — it demands that reform advocates rely on the weight of evidence rather than claiming unanimity, and it means the suicide and violence data must carry the argument rather than psychological harm studies alone.
Both sides agree: Both sides accept the Colorado data as valid and probative: a 90% reduction in solitary use produced a 40% drop in prison violence and cost savings, meaning neither side defends indefinite isolation as a proven safety tool.
The real conflict: A factual and predictive conflict: the conservative position holds that states and institutions should lead reform at their own pace with evidence-based outer limits, while the liberal position holds that democratic majorities and judicial trajectories already justify a federal mandate — and that without one, institutional resistance will predictably override reform every time.
What nobody has answered: If Colorado's reforms succeeded because the state funded alternatives before removing the tool, and New York's failed because it did not, then the policy question is no longer whether to reform but who bears the cost of adequate resourcing — and neither side has specified where that money comes from or what happens to reform timelines in states that simply refuse to appropriate it.
Sources
- Solitary Watch 2023 report on solitary confinement population figures
- U.S. Senate records: End Solitary Confinement Act, introduced July 28, 2025, by Sen. Markey et al., referred to Senate Judiciary Committee
- New York HALT Act legislative history (2021) and 2025 wildcat strike reporting
- Virginia HB 1244/SB719 veto by Governor Glenn Youngkin
- UN Standard Minimum Rules for the Treatment of Prisoners (Mandela Rules), 2015
- UN Special Rapporteur on Torture reports on solitary confinement
- Colorado Department of Corrections data on solitary confinement reform outcomes
- Bureau of Prisons suicide statistics in solitary confinement
- Federal court order (2025) on mentally ill inmates in New York solitary-like conditions
- Madrid v. Gomez (1995) federal court ruling on Pelican Bay