Daniel sent us this one — he's asking about the entire five-thousand-year history of imprisonment, which countries lock up the most and fewest people per capita, and where you'd want to live if you wanted to minimize your odds of seeing the inside of a cell. Oh, and which nations are actively moving away from incarceration. Just a light Tuesday question.
I love this. It's two questions that sound simple and then you realize the answer touches basically all of human civilization. The history of what we even think punishment is, and then the modern data on who actually uses prisons and who doesn't.
The data is genuinely surprising. Most people assume the US has the highest incarceration rate in the world. That was true for decades. It's not true anymore.
El Salvador surpassed the US in 2023, and there are several countries that don't publish data at all — China, North Korea, Eritrea — where the real numbers could be much higher. So the global picture is messier than the headlines suggest.
Let's start at the beginning. When did humans first decide that locking someone up was a reasonable response to wrongdoing?
The earliest evidence we have is from Mesopotamia, around 2100 BCE. The Code of Ur-Nammu — this is one of the oldest known legal codes, predating Hammurabi by about three centuries — it mentions detention specifically for debt and theft. If you couldn't pay what you owed, you could be held until you worked it off or your family paid.
The first prisons were essentially debt collection mechanisms.
And Egypt had something similar — workhouses under the Pharaohs where people were confined and forced to labor. But here's the key thing that most people get wrong: for the vast majority of human history, prisons were not the punishment. They were the waiting room before the punishment.
Meaning you sat in a cell until someone decided whether to fine you, exile you, beat you, or kill you.
The dungeon was a holding cell, not a sentence. The Romans built the Mamertine Prison around 640 BCE — it's often called the first dedicated state prison — but it was almost entirely for pretrial detention and execution. People didn't get sentenced to "ten years in the Mamertine." They waited there until they were executed or released.
Which makes a certain grim sense. If your main punishments are fines, flogging, exile, and death, what do you need a long-term prison for?
That remained true through the medieval period. Castles had dungeons, but they were for holding prisoners of war, political hostages, or people awaiting trial. The idea of "you stole a sheep, therefore you will spend three years in a cell" — that just wasn't how anyone thought about justice.
I want to pause on that for a second, because it's such a foreign concept to the modern mind. If someone steals a sheep today, we immediately think "jail time." But in the medieval mind, why would you feed and house a thief for three years? That's just a drain on resources.
The medieval economy couldn't support a system where the state took on the burden of housing, feeding, and guarding large numbers of people who weren't producing anything. If someone stole a sheep, the response was restitution — often multiple times the value of the sheep — combined with some form of physical punishment, or exile if they couldn't pay. You only locked someone up if you needed them to stick around for trial, or if they were a valuable hostage. The idea of a prison as a place where you send people for years at state expense would have seemed like madness to a medieval lord.
The eighteenth century. And really, one guy's book tour changed everything. John Howard — he was a British sheriff who started inspecting prisons in 1773 and was so horrified by what he saw that he wrote a book in 1777 called "The State of the Prisons." He documented disease, extortion by jailers, people dying of gaol fever — which was typhus — before they even got to trial. His book sparked a reform movement across Europe and North America.
I think it's worth understanding just how bad these places were. Howard described prisoners being held in underground cells with no light, no ventilation, standing water on the floor. Jailers weren't paid a salary — they made their money by charging prisoners for food, for blankets, for the privilege of not being chained to the wall. If you couldn't pay, you starved. And because pretrial detention and convicted prisoners were all thrown in together, you'd have someone awaiting trial for stealing a loaf of bread sharing a cell with a murderer, and both of them would catch typhus from the same flea-infested straw.
What made Howard's book so impactful was that he didn't just describe the conditions — he personally visited hundreds of prisons across England, Scotland, Wales, and continental Europe. He measured cell dimensions, he documented ventilation, he counted the dead. It was essentially an investigative journalism project a century before that term existed. He died in 1790 of gaol fever himself, contracted while inspecting prisons in what's now Ukraine.
Out of that reform movement came the idea of the penitentiary.
The word "penitentiary" comes from "penitent" — the idea was that if you isolated someone in a cell with a Bible and nothing else, they would reflect on their sins and become penitent. Reform through solitude. The first real experiment with this was the Walnut Street Jail in Philadelphia in 1790.
Which is where we get the Pennsylvania system — total isolation, zero human contact except for the chaplain and the occasional guard.
It broke people. The rates of mental illness and suicide were catastrophic. Charles Dickens visited Eastern State Penitentiary in 1842 and wrote that the solitary system was worse than any physical torture — he called it "a slow and daily tampering with the mysteries of the brain." But at the time, this was considered progressive. It was the humane alternative to public flogging and execution.
Because the reformers were trying to be kinder. "Instead of beating this man or hanging him, we'll give him quiet time to think about what he did.
That's the uncomfortable thread through prison history — a lot of the worst outcomes came from well-intentioned reforms. The Auburn system that competed with Pennsylvania's model had prisoners working together in silence during the day and isolated at night. It was considered more practical because the prison could sell the labor. By 1850, imprisonment had become the default punishment across the Western world. Corporal punishment and execution were increasingly seen as barbaric.
There's an irony there that's hard to miss. The reformers wanted to end the spectacle of public violence — the floggings, the hangings in the town square — and they succeeded. But they replaced it with a hidden violence that was arguably more psychologically destructive, just less visible.
The suffering moved behind walls. And because it was invisible, it was easier for the public to ignore. Dickens was one of the few people who actually went inside and reported what he saw. Most people just knew that criminals were being "reformed" somewhere, and that was enough.
In the span of about seventy years, we went from "prisons are a holding cell" to "prisons are the punishment.
Then the twentieth century happened. And this is where the numbers get staggering. In 1970, the US prison population was about two hundred thousand. By 2008, it was two point three million. That's an eleven hundred and fifty percent increase in less than forty years.
That's not because Americans suddenly became eleven times more criminal.
Not even close. Crime rates actually peaked in the early 1990s and then declined for decades while incarceration kept climbing. The driver was policy. The War on Drugs declared in 1971, mandatory minimum sentencing laws, three-strikes laws, truth-in-sentencing laws that required people to serve eighty-five percent of their sentence before parole. The 1986 Anti-Drug Abuse Act established a hundred-to-one sentencing disparity between crack and cocaine powder — same drug, different form, but if you were caught with five grams of crack you got the same mandatory minimum as someone with five hundred grams of powder cocaine.
That disparity fell almost entirely on black Americans.
The US Sentencing Commission found that eighty-five percent of crack defendants were black, even though the majority of crack users were white. That single law is a case study in how policy, not crime, drives incarceration rates.
Let me give people a concrete example of how this hundred-to-one ratio played out in real life. Take two first-time offenders. One is a stockbroker caught with five grams of powder cocaine at a party. Under federal sentencing guidelines, he's looking at probation, maybe a few months. Another person — same age, no prior record — is caught with five grams of crack cocaine. That's an automatic five-year mandatory minimum. Five years, no parole. Same drug, same quantity, same lack of criminal history. The only difference is the form the cocaine is in — and, overwhelmingly, the race of the person holding it.
That's not an edge case — that's exactly how the law was structured. Congress eventually reduced the disparity to eighteen-to-one in 2010 with the Fair Sentencing Act, and then made it retroactive, but for twenty-four years, that hundred-to-one ratio was federal law. Hundreds of thousands of people were sentenced under it.
That's the history — five thousand years compressed into about nine minutes. Let's pull up the global map. Where do things stand right now?
The key metric is incarceration rate per hundred thousand population. Not total prisoners — that just tells you which countries are big. Per capita tells you the likelihood that a random person in that country is behind bars.
The top of the list?
As of 2024 data from the World Prison Brief, the United States is at six hundred and twenty-nine per hundred thousand. That's still the highest among OECD nations — roughly five times the OECD average of about one hundred and twenty. But El Salvador is now at six hundred and five, and during its state of emergency crackdown on gangs between 2022 and 2025, it actually spiked to over a thousand.
El Salvador briefly had the highest incarceration rate ever recorded.
In a functioning state that reports data, yes. They added about sixty thousand prisoners in eighteen months. The government suspended basic due process rights — people were arrested based on appearance, neighborhood, vague accusations. The homicide rate plummeted, and President Bukele's approval rating is astronomical, so there's a real tension there between civil liberties and public safety that the country is actively wrestling with.
How does that actually work on the ground? If you suspend due process, what does an arrest look like?
During the state of emergency, police were given broad authority to detain anyone they suspected of gang affiliation. In practice, that meant young men in certain neighborhoods were rounded up en masse. Tattoos became de facto evidence — if you had certain tattoos associated with MS-13 or Barrio 18, you were arrested. The problem is that tattoos are also just part of youth culture in many communities, and there are documented cases of people with non-gang tattoos being swept up. The government built a massive new prison — the Terrorism Confinement Center — designed to hold forty thousand people. It's essentially a warehouse for human beings, with rows of cages and minimal access to legal representation.
The rest of the top five?
Rwanda, Turkmenistan, and Cuba round out the typical top five, though the exact order fluctuates. Rwanda's high rate is partly a legacy of the genocide — they're still processing cases from 1994 through community courts called gacaca, and the prison system is dealing with that historical overload.
Then there's the data caveat.
China officially reports about one hundred and eighteen per hundred thousand, but the World Prison Brief and human rights organizations estimate the real number could be two to three times higher. Political prisoners, re-education camps in Xinjiang — none of that shows up in official incarceration statistics. North Korea and Eritrea are complete black boxes. So when we say "the US is number one," we mean among countries that actually publish reliable data.
Which is an important asterisk. Now flip it — who's at the bottom?
Iceland has the lowest reported rate at thirty-three per hundred thousand. Japan is at thirty-eight, Norway at forty-nine, Finland at fifty-one, Switzerland at seventy. These are countries where incarceration is a last resort.
Japan is an interesting case because it gets held up as a low-incarceration success story, but the experience of being imprisoned there is famously harsh.
Japan's conviction rate is over ninety-nine percent — prosecutors simply don't bring cases they can't win. The detention system allows for up to twenty-three days of interrogation without charge, and confessions are extracted at a rate that makes outside observers uncomfortable. The prisons are clean and orderly, but the isolation is extreme. Prisoners can be punished for making eye contact with guards. So low incarceration doesn't automatically mean humane incarceration.
That's the musical equivalent of beige wallpaper with razor blades underneath.
Now contrast that with Norway. Norway's incarceration rate is forty-nine per hundred thousand, and their recidivism rate within two years is twenty percent. The US recidivism rate, according to the Bureau of Justice Statistics 2023 report, is seventy-six percent.
Three out of four people released from US prisons are rearrested within two years, versus one in five in Norway.
Norway spends more per prisoner — a lot more. Halden Prison looks like a Scandinavian design hotel. Bastøy Prison is an open facility on an island with no fences. Prisoners have keys to their own rooms, they cook their own meals, they have jobs, they have access to education and therapy. The philosophy is that the punishment is the loss of liberty — everything else should approximate normal life so that when people get out, they know how to function in society.
The counterargument writes itself: "So you're rewarding criminals with a nicer life than some law-abiding citizens have?
The Norwegian response is: what outcome do you want? If you want retribution, lock people in harsh conditions and accept that they'll come out more damaged and more likely to reoffend. If you want less crime, design a system that produces less crime. The data is pretty clear on which approach works.
I remember reading about a specific case — a man who'd committed a violent crime and was sent to Bastøy. He'd never cooked a meal in his life, never held a job, never managed a budget. He came from a background of total chaos. And the prison staff taught him how to do all of those things — how to plan meals, how to show up on time, how to resolve conflicts without violence. He was released, got a job, and didn't reoffend. And the question the Norwegian system asks is: would you rather have that man as your neighbor after five years, or would you rather have him after five years in a concrete box where he learned nothing except how to survive in a concrete box?
That's the choice. And most societies are still choosing the concrete box, because the alternative feels like being soft on crime, even when the evidence says it produces more crime in the long run.
Let's talk about the countries that are actively moving away from incarceration. Finland is the classic case study.
Finland's story is remarkable because it wasn't gradual cultural evolution — it was a deliberate policy choice. In 1950, Finland's incarceration rate was one hundred and eighty-seven per hundred thousand — comparable to the US at the time, actually. By 2024, it was fifty-one. That's a seventy-three percent reduction.
What did they actually do?
They systematically reformed their criminal code to reduce sentences for property crimes, they expanded community service as an alternative to prison, they invested heavily in probation and social services, and they closed institutions. They went from having dozens of small, often harsh prisons to a smaller number of facilities focused on rehabilitation. And the key thing — crime did not go up.
Which is the fear that paralyzes reform everywhere. "If we release people or don't lock them up in the first place, crime will spike.
In Finland, it didn't. Neither did it in Portugal, which is the other major success story. Portugal decriminalized all drugs in 2001 — not just cannabis, all drugs. Possession for personal use became an administrative offense, not a criminal one. Instead of arresting people, they refer them to "dissuasion commissions" made up of a lawyer, a doctor, and a social worker. They assess whether the person needs treatment, harm reduction services, or just a fine.
I want to be precise about what decriminalization means here, because it's often misrepresented. Portugal did not legalize drugs. If you're caught with heroin, the police still confiscate it. You still get a citation. But instead of going to a criminal court, you go to this dissuasion commission. And they're not there to punish you — they're there to figure out if you have a problem and what kind of help you need. It's the difference between treating someone as a criminal and treating them as a patient.
And the commission has real teeth — they can impose fines, they can suspend professional licenses, they can mandate treatment. But the key is that no one gets a criminal record for personal drug possession. That means no prison time, no barrier to employment, no loss of voting rights, no cascade of consequences that often makes drug problems worse by cutting people off from legitimate society.
Drug-related deaths dropped by thirty percent. HIV infection rates among drug users plummeted. And the prison population dropped by seventeen percent by 2010, and continued falling through 2025. Portugal closed its fifth prison in a decade just this month — May 2026.
Meanwhile, the Netherlands had a different kind of success story that then reversed. They closed nineteen prisons between 2013 and 2022 because crime rates had fallen so much they didn't have enough prisoners to fill them.
Which is the kind of problem most countries would love to have. They actually started importing prisoners from Belgium and Norway to keep some facilities open and preserve jobs. But then organized crime — particularly drug trafficking through Rotterdam — started rising, and they had to reopen some prisons in 2024. It shows how dynamic this is. Incarceration policy isn't something you set and forget.
Does importing prisoners actually work as a strategy? How does that even function legally?
It's done through bilateral agreements. Belgium had overcrowding, the Netherlands had empty cells, so Belgium paid the Netherlands to house Belgian prisoners in Dutch facilities, staffed by Dutch guards but operating under Belgian legal standards. It's unusual but not unprecedented — Norway has rented prison space in the Netherlands as well. The practical challenge is that prisoners are far from their families and legal representation, which raises human rights concerns. But from a purely logistical standpoint, it kept Dutch prison workers employed and gave Belgium a pressure release valve.
If you're an individual trying to game this — you're a law-abiding person who wants to know where a minor offense is least likely to land you in a cell — where do you go?
Iceland, Japan, Finland, Norway, Switzerland. But you have to be careful about what "offense" means. Japan criminalizes defamation with potential jail time. Iceland basically doesn't. Finland treats drunk driving very seriously — you can get jail time for a first offense at a blood alcohol level that would be a fine in other countries. So "least likely to imprison you" depends partly on what kind of laws you're likely to break.
Which is the fundamental problem with cross-country comparisons. An "imprisonable offense" isn't the same thing everywhere.
And that's why the single biggest predictor of a country's incarceration rate isn't its crime rate — it's its policy choices. The US and Canada have similar violent crime rates, but the US incarcerates at roughly five times the rate. The difference isn't that Americans are five times more criminal. It's that the US has mandatory minimums, three-strikes laws, cash bail systems that keep people in jail pretrial because they can't afford bond, and a War on Drugs infrastructure that Canada never built to the same degree.
That cash bail point is worth highlighting, because it's one of those things that sounds technical but has enormous human consequences. In the US, roughly half a million people are in jail on any given day who have not been convicted of anything. They're awaiting trial. And the only reason they're in a cell is that they can't afford to pay bail. A wealthy person charged with the same offense goes home. So you have this system where freedom is literally priced — and the price falls hardest on the poor.
Pretrial detention is devastating even if you're eventually acquitted. People lose their jobs, they lose housing, they lose custody of their children. Studies show that people held pretrial are more likely to plead guilty — even if they're innocent — just to get out, because the prosecutor offers time served and they can go home. The bail system creates a perverse incentive where innocence is less important than solvency.
What are the common traits among the countries that are successfully reducing incarceration?
Three things keep showing up. First, drug decriminalization or legalization — treating substance use as a public health issue rather than a criminal justice issue. Second, investment in mental health and addiction treatment — most prison populations have disproportionately high rates of mental illness and substance use disorders, and if you don't treat those, you're just cycling people through. Third, judicial discretion over mandatory minimums — giving judges the ability to consider circumstances rather than being bound by rigid sentencing formulas.
All three of those are politically difficult in different ways.
Drug decriminalization sounds to many voters like "being soft on drugs." Mental health investment is expensive and the payoff is measured in decades, not election cycles. And judicial discretion can look like "lenient judges letting criminals off easy" when a high-profile case goes wrong.
The politics of fear are just more immediate than the politics of data.
And that's why the countries that have succeeded tend to have either a strong technocratic tradition — like the Nordic countries, where policy is driven by expert consensus more than political theater — or a crisis that forces change, like Portugal's heroin epidemic in the 1990s that was so bad that decriminalization became the "we've tried everything else" option.
Let's talk about El Salvador for a moment, because it's the counterexample that's being celebrated in a lot of circles right now. Homicide rates collapsed. The president is wildly popular. And they did it by locking up an enormous percentage of the population very quickly with minimal due process.
This is the tension at the heart of incarceration policy. If you ask people in El Salvador whether they feel safer, the answer is overwhelmingly yes. The murder rate went from being one of the highest in the world to lower than many US cities. Gangs that controlled entire neighborhoods for decades were dismantled. The cost was tens of thousands of people — including some innocent people — being imprisoned, often in massively overcrowded conditions, with no clear path to trial.
Is the question just "how much injustice are you willing to tolerate for how much safety?
I think that's exactly the question, and different societies answer it differently. The Nordic model answers: "Very little injustice, and we'll accept somewhat more crime as the trade-off." El Salvador answered: "A lot of injustice, because the crime level was intolerable." And both populations are, by and large, satisfied with their choice.
Which makes universal prescriptions difficult. You can't just tell a country with a homicide rate of fifty per hundred thousand "be more like Norway.
Norway's homicide rate is around zero point five per hundred thousand. The social conditions are completely different. The trust in institutions is completely different. The homogeneity and social cohesion are different. That doesn't mean lessons can't transfer, but they transfer as principles — invest in rehabilitation, treat addiction as health care, use prison sparingly — not as a copy-paste policy.
If someone listening wants to follow this issue, where should they look?
The Prison Policy Initiative does excellent data-driven work on US incarceration. The Howard League for Penal Reform in the UK — named after that same John Howard from 1777 — is one of the oldest prison reform organizations in the world. And the International Centre for Prison Studies hosts the World Prison Brief, which is the definitive global dataset.
For people who want to understand the mechanism — how does a country actually reduce its prison population — Finland's reform history is well-documented in English, and Portugal's decriminalization model has been studied extensively. The evidence is there. The question is whether the political will exists.
That brings us to the forward-looking question. We're now in an era where AI-powered policing and predictive sentencing tools are spreading. COMPAS in the US, HART in the UK — these are algorithms that assess a defendant's risk of reoffending and influence sentencing and bail decisions. The question is whether they'll reduce incarceration by diverting low-risk people, or increase it by giving a veneer of objectivity to harsher sentences.
Because an algorithm can't be accused of being soft on crime. It's just math.
And if the math is trained on historical data that reflects biased policing and sentencing patterns, it reproduces those patterns. There's a real risk that predictive tools lock in the current incarceration disparities rather than correcting them.
There was a ProPublica investigation in 2016 that looked at COMPAS, the algorithm used in Florida and several other states. They found that the algorithm was twice as likely to falsely flag black defendants as high risk compared to white defendants, and twice as likely to incorrectly label white defendants as low risk. So the algorithm wasn't predicting future crime — it was predicting future policing patterns, which are themselves racially skewed. And the judges using these scores thought they were getting objective science.
That's the seduction of the number. It feels neutral. It comes out of a computer. But the training data is soaked in the same biases that produced the incarceration crisis in the first place. If you train an algorithm on a system that arrested black people for crack at a hundred times the rate of white people for powder cocaine, the algorithm learns that being black is a risk factor. It doesn't know about the sentencing disparity. It just sees the pattern and replicates it.
The UN's Nelson Mandela Rules from 2015 set global standards for prison conditions — things like adequate healthcare, separation of pretrial and convicted prisoners, limits on solitary confinement — but enforcement is essentially voluntary. The next decade may see more international pressure on high-incarceration states, but it's not clear what leverage exists.
That's the structural problem. There's no World Prison Police. There's no international body that can sanction a country for locking up too many people in inhumane conditions. The Nelson Mandela Rules are an important normative standard — they say "this is what a civilized prison system looks like" — but compliance depends entirely on domestic political will. And in countries where being tough on crime is a winning electoral strategy, that will is often absent.
The five-thousand-year arc bends from "prison is a waiting room for punishment" to "prison is the punishment" to "prison is a political tool for managing fear." And now we're adding algorithms to the mix.
The countries that are bending the arc in a different direction — toward less incarceration, more treatment, more community-based justice — are not doing it because they're morally superior. They're doing it because they looked at the data and decided the current system was too expensive and produced worse outcomes. It's a pragmatic argument as much as an ethical one.
Which is maybe the most useful way to frame it. You don't have to be a bleeding heart to look at a seventy-six percent recidivism rate and think "this isn't working.
And now: Hilbert's daily fun fact.
Hilbert: In the 1860s, a naturalist in Chad documented that horseshoe crab blood contains amebocytes that clot on contact with bacterial endotoxins — a discovery that, a century later, would make the blue blood of these four-hundred-million-year-old arthropods the medical industry's standard for detecting contamination in vaccines and implants, at a value of roughly fifteen thousand dollars per liter.
Fifteen thousand dollars a liter. For crab blood.
I have so many questions and I'm not sure I want any of the answers.
That'll haunt me.
This has been My Weird Prompts. Our producer is Hilbert Flumingtop. If this episode made you think differently about freedom and punishment, share it with one person — that's the best way to help the show. Rate us if algorithms are your thing. We're at myweirdprompts.com. I'm Corn.
I'm Herman Poppleberry. See you next time.
