Daniel's question this week starts with a border that's running in the wrong direction. Most fences, walls, checkpoints — the whole apparatus of border control — exists to keep people out. North Korea's exists to keep people in. And every person who gets out isn't just a lost citizen, they're a leak. They carry the inside of the country with them, and the regime knows it. So Daniel wants to know what the actual borders look like — the DMZ, the Chinese frontier, the Russian frontier, the sea — and what the neighbors do when someone makes it across. How many try the DMZ versus the land routes, and whether anyone's made it out by boat.
The short answer on the DMZ is that almost nobody tries it anymore, and the numbers bear that out. In the first half of this year, South Korea's Unification Ministry recorded sixty-three defectors arriving. Every single one came through a third country. Zero direct crossings from the North. That's not a rounding error — that's the DMZ being functionally closed as an escape route.
Sixty-three in six months. I remember when the annual figure was three thousand.
The two thousands were a different era. Twenty-two nine in twenty twenty, then sixty-three, sixty-seven, a bounce to one ninety-six in twenty twenty-three, two thirty-six in twenty twenty-four, two twenty-four last year. Cumulative total now sits around thirty-four thousand six hundred in the South. But the composition has shifted completely. Twenty-five thousand of them are women. And the ones getting out now, a pastor who works with defectors put it bluntly: only the rich and well-connected can defect anymore.
Which is its own kind of filter. You're no longer getting a representative sample of what life in North Korea is like. You're getting the people who could afford to pay a broker, bribe a border guard, or wait out a pandemic.
And the pandemic is the hinge point here. Kim Jong Un sealed the China border in early twenty twenty, and that changed everything. Before that, crossing into China was the main route. Maybe a thousand people a year crossed the Tumen River. After the sealing, that collapsed to roughly twenty people in four years, of whom four made it to South Korea. The sea became almost the only option.
Which is a sentence that should stop you. The sea. The Sea of Japan, the Yellow Sea — these are not gentle crossings. And yet that's the route people are taking because the land options got welded shut.
The land options are China and Russia. China is the one we know the most about, and the picture is grim. China labels undocumented North Koreans as economic migrants, not refugees, and sends them back under a bilateral protocol from nineteen eighty-six. Human Rights Watch documented at least four hundred six forced returns since twenty twenty-four, and over a thousand since twenty twenty. The returns are not gentle. The returnees go to political prison camps, labor camps, interrogation.
China's party to the Refugee Convention and the Convention Against Torture. Both of those say you don't send people back to a place where they'll be persecuted. And China just... does it anyway. With quotas.
That's the detail that stuck with me. Reuters reported in September twenty twenty-four that Chinese border police have quotas to identify and expel undocumented migrants. They've got patrol boats, deportation centers. It's an industrial process. A database center in Seoul, the NKDB, released a report in March arguing the forced repatriation system amounts to crimes against humanity. They've got eight thousand two hundred forty-five recorded cases and ninety-six survivor testimonies.
Ninety-six testimonies. That means there are ninety-six people alive to describe what happens after you get sent back. What do they say?
One repatriated man described a nineteen-month ordeal. Interrogation, then a labor camp in Ryanggang Province. Forced to work from four in the morning until ten at night. He said they ate rats and maggots to stay alive. His parents sold their house to bribe officials for his release. His only wish, he said, was to have one night of deep, peaceful sleep.
That's the border working as designed. Not the fence, not the guards — the process after you're caught. The punishment is part of the deterrent. And it's not just North Korea doing the punishing. China's the one handing people back to that.
The twenty fourteen UN Commission of Inquiry found that forcibly returned North Koreans face crimes against humanity. A UN Human Rights Office report last September reiterated it and called on states to respect non-refoulement. China's response has been to expand the machinery.
Let's stay with the mechanics for a minute, because I think there's something specific to this border that Daniel's getting at. The information leak problem. It's not just that a defector is a lost worker or a lost soldier. It's that they've seen the inside. They know the food situation, the surveillance, the executions.
And the regime treats that as a capital offense. There was a twenty ten Ministry of People's Security decree that made defection treachery against the nation, punishable by death. The criminal code distinguishes between economic crossing and political defection — two years forced labor for the first, minimum five years for the second. But the practical reality is much harsher. The BBC interviewed a defector who described a public execution in April twenty twenty-three of a twenty-two-year-old man. His crime was listening to about seventy South Korean songs and watching three films, then sharing them. That was under a December twenty twenty law allowing execution for sharing South Korean content.
Seventy songs. That's a playlist. That's someone's afternoon.
The same defector described the state's food policy as starving your baby, then giving them a small amount so it would thank you. That's the information environment. People inside know it's bad. The regime knows they know. And the border is the place where that knowledge becomes dangerous, because once you're across, you can say it out loud.
So the border isn't just a physical line. It's an epistemic boundary. Inside the line, the state controls what's true. Outside, anyone can contradict it. And the people who cross become evidence.
Which is why the crossing stories matter so much. The literature Daniel mentioned — the memoirs, the testimonies — those aren't just adventure stories. They're the counter-propaganda. Every defector who publishes a book is a breach in the information wall.
Let's talk about what crossing actually looks like now. The DMZ crossings, when they happen, are bizarrely low-tech. A soldier walks south across the Military Demarcation Line at ten at night, unarmed, picked up by thermal cameras. That happened in June. Another soldier crossed in October twenty twenty-five. A civilian crossed the MDL on the central-western front in July twenty twenty-five. Another was rescued from the Han River neutral zone.
The Han River one is interesting because it's a water route through the DMZ itself. The river runs through the neutral zone, and the median line matters. He was pulled from the water south of the median line. So even within the DMZ, the river is a potential corridor.
But these are one-offs. The DMZ is mined, fenced, guarded on both sides, and the North has shoot-to-kill orders along much of it. The numbers reflect that. Direct crossings are rare enough that each one makes news.
The land border with China is the real historical route, and it's also a river border. The Tumen River separates North Korea from China's Jilin Province. It's narrow enough to wade across in places, and in winter it freezes. People crossed on the ice for decades. The Yalu River, further west, is the other major crossing point.
And the Russian border?
That's the gap in the research, honestly. The North Korea-Russia border is only about seventeen kilometers long — it's the shortest of North Korea's land borders, running along the Tumen River to the sea. There are far fewer documented defections that way. It's less populated on the Russian side, the terrain is different, and the Russian security services have their own relationship with Pyongyang. I couldn't find dedicated sourcing on Russian border policy toward North Korean escapees, and I'd rather say that plainly than invent a number.
Seventeen kilometers. That's a rounding error compared to the fourteen hundred or so with China. But it's also the border that's gotten more attention lately because of the Russian relationship — the workers sent to Russia, the military cooperation.
And the workers are a whole other category. North Korea sends laborers abroad, including to Russia, and they're effectively hostages to their own state. Their families stay behind. The state takes most of their wages. They're not defecting because the cost is borne by the people they left.
So the land routes, in order of use, are China first, Russia a distant second, and the DMZ a symbolic third. But the sea route has become the main event since twenty twenty.
The sea escapes are the ones that read like novels because they basically are. The Kim family of nine in May twenty twenty-three — that story's been told in the BBC and CNN. They spent a decade planning. They crossed a minefield in the dark. They hid children in grain sacks. They carried swords and poison. The mother carried eggshells filled with chili powder and black sand to throw at coastguards' eyes.
Eggshells filled with chili powder. That's the detail that tells you how desperate this is. You're not bringing a weapon, you're bringing a kitchen spice and hoping it buys you a second.
They crossed the maritime border in under two hours and were rescued near Yeonpyeong Island. That's the Northern Limit Line, the de facto sea border in the Yellow Sea. And they made it. Four other suspected defectors were found in a small wooden boat near the Northern Limit Line in October twenty twenty-three. Another family, Kang Gyu-rin and three relatives, escaped in a rickety wooden boat in October twenty twenty-three and got to South Korea.
This isn't a smuggling operation with a fast boat and a professional crew. This is people building something in a shed and hoping it floats long enough.
Sokeel Park from Liberty in North Korea said these sea escapes take meticulous planning, incredible bravery, and for everything to go miraculously well. And then he added: there must be many more who have tried but not made it.
That's the uncounted number. We know the sixty-three who arrived in South Korea in the first half of this year. We don't know how many boats went out and didn't arrive anywhere.
The sea route is the most lethal and the least documented. A boat that sinks in the Yellow Sea doesn't leave a body for anyone to count. The families left behind can't ask questions. And the North Korean coast guard isn't publishing its interception numbers.
Let's talk about the South Korean side of this, because Daniel asked about what policies the neighbors apply. South Korea's policy is, in principle, unconditional acceptance. Any North Korean who reaches South Korean soil or waters and expresses a desire to defect is taken in, debriefed, and resettled.
The resettlement program is called Hanawon, and it's a two-to-three-month re-education and integration program. Defectors get citizenship, housing support, a settlement grant. The South's constitution claims the entire peninsula, so a North Korean is technically already a citizen. That's the legal framework.
But the practical experience is harder. Defectors face discrimination in the South. They're marked by their accent, their mannerisms, their lack of familiarity with South Korean life. And the incarceration numbers tell a story — one hundred seventy-two North Korean defectors in South Korean prisons last year, up eight point two percent.
Some of that's economic desperation, some of it's the trauma of the escape and the adjustment. The South's policy is generous on paper, but the integration is uneven.
And then there's the third-country problem. The sixty-three who arrived this year all came through third countries — mostly China, then Southeast Asia, then South Korea. The route is: cross into China, hide, make contact with a broker, travel overland through China to Laos or Thailand or Vietnam, then fly or take a boat to Seoul. The entire journey can take years.
The CNN story about the Kim brothers is exactly this. They spent ten years planning. The younger brother who made the escape possible died in an accident months after reaching freedom. Ten years of planning, and he got a few months of the life he'd been working toward.
The China leg is the most dangerous part. That's where the quotas come in, the surveillance cameras, the informants. Reuters called it a fishing net. You're in a foreign country where you don't speak the language, you have no documents, and the local police have a numerical target for finding you.
The women are especially vulnerable. Human Rights Watch documented two hundred twelve trafficked women returned from Kunming, Nanning, and Pingxiang during twenty twenty-four. These are women who were trafficked into marriages or sex work in China, then rounded up and sent back to North Korea, where they face punishment for having left.
Two hundred twelve trafficked women sent back to be punished for being trafficked. That's the border policy in one sentence.
The NKDB report's framing is useful here. Donghwi Shin said the forced repatriation is a structural processing system, not a simple act of administrative enforcement. It's not a border guard making a bad call. It's a pipeline: detention, identification, transfer, repatriation, punishment. And it runs on quotas.
Let's circle back to something Daniel raised — the comparison to other volatile borders. He's right that most border security is about keeping threats out. Drug smuggling, human trafficking, terrorism. The border is a filter for what enters. North Korea's border is a filter for what leaves.
And the thing being filtered out is information. That's the unique part. A drug smuggler crossing into the United States isn't carrying a narrative that undermines the Mexican government. A North Korean defector is carrying exactly that — a first-person account of life under the regime. The regime's propaganda machine depends on controlling the narrative, and every defector is a hole in that narrative.
Which is why the punishment for defection is so severe, and why the punishment for sharing South Korean media is execution. It's not about the individual. It's about maintaining the monopoly on reality.
The twenty fourteen UN Commission of Inquiry called the situation in North Korea without parallel in the contemporary world. The border is the enforcement mechanism for that. The camps, the executions, the propaganda — all of it is held in place by the fact that people can't leave.
And yet people do leave. Sixty-three in six months. A family of nine in a fishing boat. A soldier walking south at night. The border isn't perfect. It's just brutal enough that the success stories are extraordinary.
The sea route is the wild card. The land borders are militarized and surveilled, but the coastline is long and the sea is dark. North Korea's coast guard has boats, but it can't cover every inlet. The escapees use the same logic as smugglers everywhere: go where the patrols aren't.
The Kim family crossed the maritime border in under two hours. That's how close it is. Two hours in a small boat, and you're in South Korean waters, and a South Korean patrol picks you up, and you're free. But the two hours are the entire thing. The planning, the minefield, the grain sacks, the chili powder — all of it is compressed into those two hours.
And the margin for error is zero. If the boat's too slow, if the weather turns, if the coast guard spots you, if the engine dies — any of those, and you're either dead in the water or back in North Korea facing the consequences.
I want to touch on the Russia border again, because it's the least understood part of this. Seventeen kilometers along the Tumen River to the Sea of Japan. There's a rail bridge at Khasan. It's the only rail connection between the two countries. And it's become more important because of the labor exports and the military cooperation.
The workers in Russia are a specific category. They're not free agents. They're sent by the state, their wages are confiscated, their families are effectively collateral. Defection from a Russian worksite means your family back home pays the price. So the Russia route, even though it's physically accessible, is constrained by a different kind of pressure.
And the Russian authorities have their own reasons not to encourage defection. They've got a working relationship with Pyongyang, they're buying artillery shells, they're hosting North Korean workers. A high-profile defection through Russia would be diplomatically inconvenient.
The China route is different because of scale. Fourteen hundred kilometers of border, millions of ethnic Koreans on the Chinese side, a long history of cross-border movement. China's policy is the single biggest factor in how many North Koreans get out and how many get sent back.
And China's policy is, bluntly, to send them back. The nineteen eighty-six protocol is the legal instrument, but the quotas are the operational reality. A border police officer in Jilin has a target for how many undocumented North Koreans to find and expel. That's not border security, that's a production line.
The October twenty twenty-three deportation is the clearest example. Up to six hundred North Koreans deported by China in a single month, and then they vanished. Rights groups lost track of them. That's the "vanish" in the Reuters headline. Six hundred people handed over to North Korean authorities, and no one outside knows what happened to them.
Six hundred people. That's ten times the number who made it to South Korea in all of twenty twenty-one.
The scale of the repatriation machine dwarfs the scale of the escape pipeline. For every person who reaches Seoul, there are dozens who get caught in China and sent back.
Which brings us to the question Daniel asked about attempts versus successes. How many try the DMZ versus the land borders? The honest answer is that we don't have a clean count. The Unification Ministry's statement that there were zero direct crossings in the first half of this year tells us the DMZ is nearly unused. The individual cases — the soldiers, the civilians — are reported one by one because they're rare enough to be news.
The land border attempts are also uncountable in the other direction. We don't know how many people try to cross the Tumen and get caught on the North Korean side before they even reach the river. The North doesn't publish interception numbers. The only numbers we have are arrivals in South Korea and documented repatriations from China.
And the sea attempts are the least countable of all. A boat that sinks leaves no record. A coast guard interception on the North Korean side leaves no public record. The only sea escapes we know about are the ones that succeeded.
Sokeel Park's line keeps coming back to me. There must be many more who have tried but not made it. That's not a statistic, it's an inference. And it's almost certainly right.
Let's talk about the literature, because Daniel raised it and it's the human dimension of all this. The memoirs — Yeonmi Park's In Order to Live, Blaine Harden's Escape from Camp 14 — these are the texts that made the world pay attention. I should say I couldn't verify specific claims from those books in the time I had, so I'll speak to the genre rather than the individual details.
The genre is testimony. It's a specific literary form that emerged from the defector experience. The books are written with co-authors, often Western journalists, because the defector's English isn't usually strong enough for a full manuscript. The co-author brings the structure, the defector brings the experience. It's a collaboration, and it's been criticized for exactly that reason — the authenticity gets filtered through an outside voice.
But the core of it is the same thing the BBC interview captured with Mr. Kim. The details of daily life — the starvation, the surveillance, the executions — told in the first person. That's the information leak made permanent. Once it's in a book, it can't be rounded up and sent back.
The NKDB testimonies are rawer. They're collected for legal purposes, for documentation, for the eventual possibility of accountability. The ninety-six survivor testimonies in the March report — those aren't polished memoirs, they're statements. The man who ate rats and maggots in the labor camp, the one who just wanted a night of peaceful sleep. That's the border from the other side.
And there's a class dimension to who gets to tell these stories now. The pastor's line — only the rich and well-connected can defect — means the testimonies we're getting are increasingly from people who had resources. The poorest North Koreans, the ones in the worst conditions, can't afford the broker, can't pay the bribe, can't get across the river. They're the ones who get caught and sent back.
That's a real shift from the two thousands, when the defectors were a broader cross-section. The economic collapse of the late nineties pushed people across the border in large numbers, and the defector population in the South reflected that. Now the flow is smaller and more selective.
Which means our window into North Korea is narrowing at exactly the moment the regime is most opaque. Fewer defectors, more filtered by wealth, and the ones who do get out are increasingly coming by sea, which selects for people near the coast.
The coastal population has its own characteristics. They're more exposed to smuggling networks, more likely to have access to boats, more familiar with the sea. The Kim family were from a coastal area. The Kang family too. The sea route isn't available to someone in the interior — they'd have to cross half the country just to reach the coast.
So the map of escape routes is also a map of who can escape. The Tumen River for the northeasterners, the sea for the coastal population, the DMZ for the soldiers and the desperate. And the interior — the majority of the population — has no route at all.
The interior is the prison. The borders are the walls. And the people in the interior know it. The information environment inside North Korea is changing — more smuggled USB drives, more South Korean content getting through — but the physical reality is unchanged. You can know the outside world exists and still have no way to reach it.
I want to push on one thing. The South Korean policy of unconditional acceptance — does that create a moral hazard? Does it incentivize the North to make the border even more lethal, knowing that anyone who reaches the South will be welcomed and their testimony broadcast?
That's a real question. The North already treats defection as treason, already executes people for sharing South Korean media. The South's policy doesn't change the North's calculus much, because the North's position is already maximal. But the South's acceptance does mean that every escapee becomes a propaganda asset for the South, which the North knows, which means the North has every incentive to prevent escapes at all costs.
So the South's humanitarian policy and the North's information control are locked in a feedback loop. The more the South welcomes defectors and amplifies their stories, the harder the North works to keep people in. The more the North works to keep people in, the more desperate the escapes become.
And the more desperate the escapes, the more remarkable the stories. The Kim family's eggshells full of chili powder — that's not a detail from a well-resourced operation. That's improvisation under extreme constraint.
The chili powder is the whole thing in miniature. It's a weapon made from what you have, designed to buy you a moment against an armed coast guard. It's not going to work. But you bring it anyway, because you're bringing everything you have.
Hilbert: The number was four hundred six.
What?
Hilbert: Human Rights Watch. Forced returns since twenty twenty-four. At least four hundred six. You said it earlier and then you said over a thousand since twenty twenty. Both are right, but the four hundred six is the recent number. I was listening.
No, that's right. Four hundred six since twenty twenty-four, over a thousand since twenty twenty. I should have kept them separate.
Hilbert: I knew a man who crossed the Tumen in eighty-seven. He was a truck driver on the Chinese side, ethnic Korean, spoke the dialect. He'd take loads up to the border and come back with stories. One winter he drove a North Korean woman across in the cab of his truck. Hid her under a pile of blankets. She made it to Seoul eventually. He never told me her name.
Eighty-seven is a different world. The border was looser then, the Chinese side less policed.
Hilbert: He said the river was the easy part. The hard part was the hundred kilometers after the river. Checkpoints, informants, the police looking for anyone who didn't belong. He drove her past three checkpoints with her under the blankets. Said he aged ten years in one night.
That's the China leg. The river is a moment. The journey after the river is the thing.
Hilbert: He got a letter from her in ninety-one. She was in Seoul, working in a restaurant. She thanked him for the blankets. He kept the letter in his truck for years.
The blankets. That's what she remembered.
Hilbert: He said the blankets smelled like diesel and she said it was the best smell she'd ever known, because it meant she was moving.
The details people hold onto. A smell, a blanket, chili powder in an eggshell.
The border strips everything down to what you can carry and what you can remember.
Hilbert: He died in ninety-eight. I went to the funeral. There were a dozen people there, all Korean, none of them related to him. I didn't ask who they were.
The people he drove across.
Hilbert: Probably. Nobody said anything. They just stood there.
The border keeps a ledger that nobody publishes.
Hilbert: One of them left flowers. That's all I know.
The uncounted number. Sokeel Park's point exactly. We know the sixty-three who arrived in the South this year. We don't know how many truck drivers carried how many people across how many winters.
And the truck drivers aren't in any database. They're not defectors, they're not repatriated, they're not counted. They're the infrastructure of escape, and they're invisible.
Hilbert: He never charged anyone. Said he did it because his mother was from the North. She crossed in fifty-one. He never knew how.
The border runs through families. That's the thing the statistics don't capture. Every defector is someone's child, someone's sibling, someone's parent. The border doesn't just separate countries, it separates bloodlines.
And the people who help — the truck drivers, the brokers, the boat builders — they're taking the same risk as the escapees. They're just doing it for someone else.
Hilbert: He said the worst part was the silence. He couldn't tell anyone what he did. Not his wife, not his friends. He carried it alone for thirty years.
Until he told you.
Hilbert: He told me on his deathbed. Said he wanted one person to know.
The information leak, inverted. The defector carries the story out. The helper carries the story in silence. And the border keeps both of them.
Hilbert: I think about the blankets sometimes. What it must have been like under there. Hearing the checkpoint voices, not knowing if this is the one where they find you.
The Kim family hid children in grain sacks. Same logic. You make yourself small, you make yourself cargo, you wait.
And you bring the chili powder, knowing it probably won't work, because you have to bring something.
Hilbert: He said she never made a sound. The whole drive. Not one sound.
That's the discipline of the desperate. The border teaches it.
Hilbert: Anyway. That's what I know about the Tumen.
The question Daniel asked about the sea — whether there have been stowaways, people making it into international waters — the answer is yes, but it's not the main sea route. The main sea route is small boats from the North Korean coast to South Korean waters or to the Northern Limit Line. Stowaways on cargo ships happen, but they're rarer and harder to document.
The cargo ship route is interesting because it's more like the classic stowaway story. A North Korean port worker hides on a ship bound for China or Russia or Southeast Asia, then tries to slip off at the other end. The problem is that North Korean ports are heavily surveilled, and the ships themselves are often crewed by North Koreans who have every incentive to report a stowaway.
So the small boat is the more reliable option, if you can call it reliable. It's the difference between hiding in someone else's infrastructure and building your own.
And building your own is what the coastal people do. They know the water, they know the patrol patterns, they know the weather. The Kim family crossed in under two hours because they'd spent a decade learning exactly when and where to go.
Ten years of planning for a two-hour crossing. That ratio tells you everything about how hard this is.
The brothers in the CNN story, same thing. A decade. And the younger one died months after arriving. The border takes even the people who make it across.
That's the thing I keep coming back to. The border doesn't end at the border. It follows you. The trauma, the family left behind, the adjustment to a society that doesn't quite accept you. The one hundred seventy-two defectors in South Korean prisons — that's the border still working on them, years after they crossed.
The resettlement program is good on paper, but it can't undo what the border did. The man who wanted one night of peaceful sleep — that's what he asked for after nineteen months in a labor camp. Not money, not revenge. Just sleep.
And the border is still there, still running in the wrong direction, still keeping people in. Sixty-three got out this year. Sixty-three stories that made it. And the uncounted number that didn't.
The sea is dark, the river is cold, and the quotas are waiting on the other side. That's the border.
So where does this leave us? The DMZ is nearly closed as an escape route — zero direct crossings in six months. The China land border is a repatriation machine, with quotas and patrol boats and a treaty from nineteen eighty-six that China uses to send people back to labor camps. The Russia border is short, under-documented, and increasingly tied to the labor export system. And the sea has become the last real route, the most dangerous and the most unknown.
The policy picture is just as stark. South Korea accepts anyone who arrives, but the arrivals are dwindling and increasingly filtered by wealth. China sends people back in industrial quantities. Russia is a black box. And North Korea itself treats every escape as a capital crime against the information regime.
Daniel asked how many try the DMZ versus the land borders. The honest answer is that the DMZ attempts are now rare enough to make individual news, while the land border attempts are uncountable but clearly larger, and the sea attempts are the great unknown. We know the successes. We don't know the failures.
The failures are the real number. The boats that sank, the people caught at the river, the repatriated who vanished. That's the border's true body count, and nobody's publishing it.
But the stories that do get out — the Kim family, the BBC's Mr. Kim, the ninety-six testimonies — they're the counterweight. Every one of them is a hole in the information wall.
And the wall is still standing. That's the thing. Thirty-four thousand six hundred defectors in the South, and the regime is still there, still running the propaganda, still executing people for seventy songs. The border is doing its job.
The border is doing its job. That's the sentence that should keep you up at night.
Thanks to Hilbert Flumingtop for producing.
This has been My Weird Prompts. Email us at show at my weird prompts dot com.
We'll be back soon.
