1. Park Sang-ho's Back Room
September 2025, Hai Phong. Seven in the evening.
In the back room of a Korean restaurant on Nguyen Van Linh Street, the fluorescent lights come on. Three tables have been pushed together into one long surface, and fourteen plastic chairs are arranged around it. On the wall hangs a soju poster; beyond the window, motorbike horns blare without pause. The air conditioning is running, but it cannot fully lift the September humidity of Hai Phong.
Park Sang-ho — 51 years old — arrived first. He has set a whiteboard against the wall and written three lines in marker: three new assignments. Press-die maintenance for a supplier within the LG cluster's radius. QC system construction for a second-tier Samsung vendor. Equipment inspection at a Korean textile factory on the outskirts of Jakarta, Indonesia.
The smell of marker ink mingles with the smell of soju glasses. Below the three lines, the results of last month's assignments are still written out: four completed, one in progress, one cancelled. The cancelled one involved a local Vietnamese company that hired Chinese technicians instead of Korean ones. Price competition. It has already started.
People arrive one by one. Ages 48 to 63. Former occupations — textiles, shipbuilding, automotive, steel. Their last titles in Korea are all different; their present position here is the same. They are people standing on the far side of where formal institutions end.
Park Sang-ho — a former welder at Hyundai Heavy Industries — started this gathering in early 2024.
At first there were three: Park Sang-ho himself, a 58-year-old former POSCO equipment engineer, and a 54-year-old former QC specialist from a Samsung supplier. Over soju at a Korean restaurant in Hai Phong, someone said: "Couldn't we share work among ourselves?"
That was the beginning.
Lee Jung-hoon opens the back-room door and steps inside. Park Sang-ho gestures toward him. Two seats remain. Lee Jung-hoon takes a plastic chair by the window. The 63-year-old beside him — a former Daewoo Shipbuilding pipefitter — holds out a soju glass.
Lee Jung-hoon does not take it. He is still here to observe. The man beside him has welding-burn scars on the backs of his hands. Lee Jung-hoon's hands carry the same kind of scars. The same era, the same industry, the same body's language.
"Back in Korea they say people like us are useless. But here, they can't run the line without us."
Park Sang-ho says this, pointing at the whiteboard. Lee Jung-hoon reads the three lines. The first assignment — press-die maintenance. Something he can do. Something he did for 28 years. In Korea it is work that AI quality-prediction systems have replaced, but here it is still work that needs a human ear.
The third assignment catches his eye. The outskirts of Jakarta, Indonesia. Four hours by plane from Hai Phong. This network's radius is expanding beyond Vietnam to cover all of Southeast Asia.
Twelve people fill the seats. Soju goes around, pork belly grills, and from the front of the whiteboard the assignment distribution begins.
Is this business? Mutual aid? For some it is a source of income; for others a cure for isolation; for others still, a role gained here in place of the title lost in Korea. Lee Jung-hoon cannot yet define the character of this gathering.
He knows only one thing. It is different from when he was alone.
Chapter 8 followed Lee Jung-hoon's individual strategy — positional arbitrage, domain crossroads, the personal edition of the Indispensable Node. But we also wrote that individual strategy alone cannot sustain structural instability. When a contract dispute arises, he must handle it alone. When a visa problem arises, he must solve it alone. When work runs out, he must find it alone.
The twelve people in this back room are an attempt to change that "alone" into "together."
2. The Seed
The process by which an individual's adaptation becomes another person's path is not planned. Park Sang-ho's decision to come to Hai Phong was a personal one. His call to three colleagues for soju was a personal need. But when three became five, and five became twelve, it was no longer an individual act.
A structure had formed.
Informal transition-support institutions share the same characteristics wherever they appear.
Voluntarism — they arise not because someone ordered them into existence, but because a need existed. Reciprocity — not one-directional benefit but two-directional exchange. The person who receives a work introduction this month makes an introduction next month. Weak ties — the core of the structure is new relationships that extend beyond existing strong ties: family, classmates, former colleagues. In Park Sang-ho's network, all twelve came from different former workplaces, different home regions, different specialties. What binds them is not a shared past but a shared present condition.
And gradual formalization — rules are not written down, but repetition has made them practice. Monthly gatherings, an assignment-distribution order, the principle of no brokerage fees. Not decided by anyone — solidified by six months of repetition.
Korea has formal institutions for mid-career transition support. Since opening in 2016, Seoul's 50+ Campus (서울 50플러스 캠퍼스) has expanded to five locations and is used by 150,000 people a year. It provides life-redesign counseling, career-transition education, and community connection to Seoul citizens between 50 and 64. The Korea Chamber of Commerce and Industry's Senior Expert Utilization Program runs a matching service that connects retired specialists from large companies to small and medium-sized enterprises.
These institutions exist. Yet structural gaps remain.
Transition support for mid-career workers aged 40 to 55, with skilled-technical backgrounds, at the stage immediately before or just after formal retirement, is the thinnest of all. The 50+ Campus is centered on group education for those over 50; senior clubs are centered on simple service work for those over 60. There is no institution that manufacturing engineering workers like Lee Jung-hoon or Park Sang-ho — in their late forties and early fifties, needing a technical transition — can actually use.
Park Sang-ho's network grew up from that absence. KOTRA's K-Move program is specialized for youth employment in their twenties and thirties. The age ceiling at the Overseas Employment Support Center is mostly 39.
Where the state could not design coverage, individuals fill the space. The seeds of informal institutions always germinate in precisely this kind of gap.
3. Olson's Paradox and Park Sang-ho's Threshold
Yet the existence of a gap does not guarantee that a seed will germinate. There is a more fundamental question. Everyone knows the problem — so why does no one move first?
Mancur Olson answered this question in 1965 in The Logic of Collective Action. Rational individuals do not bear costs for collective benefit. The cost falls to oneself; the benefit is shared by all. Free-riding becomes the rational choice.
The larger the group, the worse this problem becomes. Estimates suggest more than 200 Korean technicians are in Hai Phong. Most are members of a Facebook group. But no one posts job-lead information. The same holds in Seoul: the silence of a KakaoTalk open chatroom with 300 members, where no one is first to share work. Olson's paradox.
Olson offered two solutions. Small groups — when membership is small, each person's contribution is visible and free-riding is difficult. Selective incentives — when there are concrete benefits that flow only to participants, motivation for action arises.
Park Sang-ho's twelve people fit Olson's first solution precisely. With twelve members, it is immediately apparent who introduced work and who did not. Selective incentives also exist: those who do not join the gatherings are excluded from assignment distribution.
Elinor Ostrom, in Governing the Commons (1990), overturned Olson's conclusion. The tragedy of the commons is not inevitable. There are conditions under which self-governance is possible. Ostrom analyzed cases from around the world and derived eight design principles: clearly defined boundaries, rules adapted to local conditions, collective-choice arrangements, monitoring, graduated sanctions, conflict-resolution mechanisms, recognition of the right to self-organize, and nested governance.
Park Sang-ho's network satisfies four of these.
Clearly defined boundaries — the group is limited to the twelve who attend the monthly gathering. Rules adapted to local conditions — assignments are distributed by specialty, and there are no brokerage fees. Monitoring — in a small group of twelve, contributions and free-riding are immediately visible. Collective-choice arrangements — assignment distribution is decided by consensus at the gathering.
The remaining four — graduated sanctions, conflict-resolution mechanisms, autonomy from external interference, and nested governance — have not yet taken shape. For an informal network of twelve to develop into a sustainable institution, all four are needed.
Mark Granovetter proposed the threshold model in 1978. Each individual has a unique number: "at least this many people must act before I will act."
Imagine 100 people in a public square. One person has a threshold of zero — that person acts even if no one else does. When that person acts, the one person with a threshold of one joins. Then the person with a threshold of two joins. A chain reaction begins.
But if there is no person with a threshold of zero — if everyone thinks "I'll move if someone else goes first" — nothing happens.
Park Sang-ho is a person with a threshold of zero. He had already been pushed out of Korea, had already pioneered a new path in Hai Phong, and had little to lose — so the cost of acting was low. He was the first to call three people. Those three each brought one or two others. When the group reached five, rules for distributing assignments appeared; when it reached twelve, the monthly gathering became regular.
Granovetter's chain had worked.
There is a paradox here: when the person most faithful to the previous system gets pushed aside, that faithfulness becomes the energy for building a new system. Park Sang-ho's "irrational" choice — investing his own time and connections in the network without pay — was what made the structure of twelve possible.
4. The Institutional Mirror: Sweden, Denmark, Singapore
There are places where the state has built into institutions what Park Sang-ho's twelve must solve for themselves.
TRR (Trygghetsrådet, the Swedish Job Security Council) is a transition-support body established in 1974 through a labor-management agreement — not by the government, but by employer associations and white-collar unions through collective bargaining. It serves approximately 40,000 workers facing redundancy each year.
When an employer notifies a worker of impending redundancy six months in advance, a TRR consultant conducts a one-on-one capability assessment, draws up a personalized transition plan, and stays alongside the worker until re-employment. "When did you feel you were doing your best work?" TRR consultations begin with this question. The starting point is not a job description but the individual's experience.
Re-employment rate within 12 months: 83 percent. Of those re-employed, 69 percent maintain wages equal to or above their previous level.
These figures represent something more than mere efficiency. They are the result of a system designed not as "a safety net for the displaced" but as "infrastructure for transition." TRR begins by recognizing the displaced worker's past — the years spent being faithful to the system.
Denmark's flexicurity is called the golden triangle. Flexible dismissal, generous unemployment benefits, and active labor market policy — the three vertices support one another.
In exchange for easy dismissal, unemployment benefits cover up to 90 percent of the previous wage for up to two years. During that period, job-seeking activity or vocational training is mandatory. It is both an obligation and a right.
Active labor market policy spending as a share of GDP is approximately 1.9 percent — near the top of the OECD. Korea's is approximately 0.4 percent. A gap of nearly fivefold.
Singapore's SkillsFuture is a system through which the state has defined lifelong learning as a survival strategy. Every citizen aged 25 and above receives a learning credit of S$500. Those 40 and above receive additional credits, and subsidies cover up to 90 percent of education costs.
The Skills Framework — which maps "the skills needed now and the skills that will be needed in five years" across more than 30 industries — determines the direction of training. Supply is driven by demand. The direction is the reverse of Korea's, where training institutions design courses and recruit participants.
The three systems have one thing in common: the state or labor and management jointly defined "transition" not as an individual problem but as a matter of institutional infrastructure.
Korea has none of this.
5. The Limits of the Naeil Baeum Card
The Naeil Baeum Card (Tomorrow Learning Card) — Korea's public reskilling subsidy — is issued to 1.52 million people per year, with 980,000 participating in training. The employment rate six months after completing training is 54.6 percent. The median wage of those re-employed is 91.3 percent of the pre-training level — meaning it actually declines.
There are four structural problems.
First, it is supply-driven. Training courses are operated through a model in which training institutions design them and the government endorses them — the opposite of TRR, which works backward from "what does the market need?" to design training. Of those in their forties and fifties participating in reskilling education, 41.2 percent chose their course because they "heard that employment demand was strong." Selection by rumor, not by structural need.
Second, there is no user sovereignty. TRR provides individualized capability assessment and transition planning. The Naeil Baeum Card provides only the card — there is no assessment. In a structure where a single employment center counselor handles hundreds of job-seekers, personalization is impossible.
Third, it does not reflect the dual structure of the labor market. Wages at small and medium-sized enterprises run at 63 percent of large-company wages. When a restructured fifty-year-old from a large corporation completes training and re-enters a small company, they must accept that gap. The rational calculation — "why go through training if wages will fall anyway" — leads to low participation.
That TRR achieves wage recovery for 69 percent of re-employed workers is a consequence of Sweden's wage-compression structure — where the wage gap between large and small companies is smaller than in Korea. The effect of an institution is not determined by the institution alone. The structure of the labor market in which it operates determines it as well.
Fourth, there is social stigma. In Denmark and Sweden, job changes and career transitions are accepted as a normal career path. In Korea, job loss in one's forties or fifties still carries the stigma of failure. As long as training is perceived in the language of comeback rather than growth, the psychological barrier to participation remains high.
The trajectory of the Rider Union (라이더유니온) illustrates these limits from a different angle.
In the summer of 2018, Mapo-gu, Seoul. Park Jeong-hoon — a McDelivery rider of three years — held a sign reading "This year, give us a heat-wave allowance, not cup ramen" and staged a one-person protest for a month. He was a person with a threshold of zero.
In 2019, 41 people gathered to launch the Rider Union. It grew to 8,000 members. In 2023 it was absorbed into the KCTU Public Transport and Service Workers' Union, and its chairperson became a member of the Minimum Wage Council.
The leader entered the formal institutions. Yet the legal status of 400,000 delivery riders remains "self-employed workers." No guaranteed minimum wage, no paid sick leave, no protection against unfair dismissal.
The leader crossed into the institutions, but the 400,000 who log into a delivery app every day, collect per-delivery fees, and ride the roads without industrial accident insurance remain outside them. A case where the transition from informal to formal stopped at the leader's transition.
6. The Law Draws a Line; Capital Dances on the Line
There are cases where informal institutions have made the transition to formal ones, and cases where the transition has failed. What determines the outcome?
Spain. In 2017, delivery riders in Barcelona and Madrid created a Facebook group, "RidersXDerechos" — Riders for Rights. It was not a union. It was not a legal entity. It was an SNS group accessed by smartphone while waiting for orders.
In 2019, a Glovo rider died in a bicycle accident in Barcelona. There was no industrial accident insurance. He was a self-employed worker.
In 2020, the Spanish Supreme Court ruled: Glovo riders are workers, not self-employed contractors. The algorithm controls work assignment, pricing, and performance evaluation, so an actual relationship of subordination exists. In 2021, the Rider Law (Ley Rider) was enacted — the first platform labor legislation in Europe.
Yet even after the law was passed, the transition was not complete. Glovo kept 80 percent of its riders as self-employed workers. The state did not retreat. Labor inspectors applied repeated sanctions; the Social Security Administration levied cumulative fines of 770 million euros; and a 2022 amendment to the criminal code made bogus self-employment a criminal offense. In 2024, facing the threat of criminal prosecution, Glovo announced it would convert 15,000 riders to regular employment.
From an informal Facebook group to a Supreme Court ruling: three years. From the ruling to legislation: one year. From legislation to substantive conversion: another three years. Seven years in total. Drawing the line through law was not enough. The state had to repeatedly sanction the companies dancing on the line.
At the other end stands the Amazon Mechanical Turk (AMT) worker. In 2009, researchers Lily Irani and Six Silberman created Turkopticon — a browser plug-in through which workers could share assessments of requesters. In 2014, the collective-action platform Dynamo was built. But campaigns stopped at two, and the platform effectively ceased to exist.
Why did it fail? 500,000 workers scattered across 140 countries were doing the same tasks but did not share the same space. Amazon did not permit communication channels between workers inside the platform. With an effective hourly wage of $2, there was no time to devote to organizing.
Spatial blockade, structural blockade, economic blockade. Three walls operating simultaneously. It was not a physical wall — the architecture itself blocked solidarity.
Park Sang-ho's back room is possible because all twelve people are in the same city. Because they can drink soju in the same restaurant. Physical proximity is a precondition for solidarity.
In Britain, the IWGB (Independent Workers' Union of Great Britain) contested the legal status of Uber drivers and won unanimously before the Supreme Court in 2021. Uber drivers are "workers." But at the same time, Deliveroo drivers were classified as "independent contractors" because of a single line in their contracts — a substitution clause.
The court drew a line. The company danced on the line.
All three cases teach one lesson. The transition from informal to formal does not happen automatically. It happens only when judicial breakthrough, political will, and social pressure converge simultaneously. In Spain, all three came together. In Korea, not one has yet fully arrived.
7. Formula Checkpoint — The Trajectory of the Formula
This series has followed a single formula. Technological innovation causes an explosion in productivity. Capital concentrates. Society grows unstable. Institutions are redesigned. A formula established in Book 1 and confirmed in every volume since.
In Book 6, The Last Profession, this formula has been modified. From technological innovation to concentration of capital: the same. After social instability, two additional terms are inserted. Individual adaptation — and the formation of informal institutions. Formal institutional redesign comes after these — if it comes at all.
Tracing the arc through Chapter 9: Chapters 1 through 4 followed technological innovation — AI dismantling cognitive labor, restructuring the meaning of occupations, the leading signals appearing in the undergrowth. Chapters 5 and 6 revealed the contours of social instability. Chapters 7 and 8 analyzed individual adaptation strategies.
And Chapter 9 — the formation of informal institutions. Park Sang-ho's twelve people. The Rider Union's 41 people growing to 8,000. Spain's RidersXDerechos Facebook group.
Two chairs stand here.
First century BCE, Rome. At a time when Crassus's private fire brigade was capitalizing on fires, citizens formed collegia — mutual-aid associations organized by trade and neighborhood. The collegium of bakers, the collegium of carpenters, the collegium of weavers. Each had regular assemblies, communal meals, and a funeral fund.
In Rome, to die without a proper funeral was to suffer social annihilation in the afterlife. The collegia were the minimum apparatus against that annihilation.
From the disappearance of Crassus's private fire brigade to the creation of the vigiles: 59 years. During those 59 years, the collegia filled the space left vacant by the absence of a formal fire brigade. Informal institutions were the soil from which formal institutions grew. When Augustus established the vigiles in AD 6, he was absorbing into the state what the collegia had been doing on an improvised basis for 59 years.
2025, Hai Phong. Park Sang-ho's twelve are the modern edition of the collegia. While Korea has no formal institution to support the overseas technology transfer of retired technicians, these twelve are filling that empty seat. Like the 59 years before the formal fire brigade arrived.
In Book 4, Slow Justice, Fast Order, we analyzed the process by which adaptive regulation fills the gaps in existing institutions. Just as a regulatory sandbox experiments with rules within a bounded space, informal institutions conduct social experiments before formal institutions arrive.
Park Sang-ho's network is a social sandbox for the question "how should Korea organize the overseas technology transfer of retired technicians?" When the results of this experiment accumulate, they become a model the state can reference. Just as the regulatory sandbox in Book 4 was a space for experimenting with innovation, informal institutions are a space for experimenting with transition. British friendly societies worked this way. Korea's Saemaul Geumgo credit unions worked this way.
But we have not yet reached the formula's last term — the redesign of formal institutions. The seeds of informal institutions rise from below. The conditions for formalization are made from above.
8. Four Conditions of Transition
Some informal institutions are formalized; others disappear. Four conditions determine the difference.
First, Critical Mass. According to the mathematical model of Pamela Oliver and Gerald Marwell, a small, highly motivated critical mass triggers collective action.
Britain's friendly societies grew from 700,000 members in 1801 to 4 million in 1872. By the early twentieth century, more than half of adult British men were enrolled. This scale was the precondition for the National Insurance Act of 1911. Lloyd George did not abolish the friendly societies — he incorporated the already-existing scale as "approved societies."
Park Sang-ho's twelve have crossed the threshold within the small group, but fall short of the social threshold.
Second, Institutional Entrepreneur. The concept defined by DiMaggio. An actor with sufficient resources who initiates a departure from existing institutions and participates directly in implementing that change.
Lloyd George used the power of the Chancellor of the Exchequer to negotiate directly among medical associations, insurance companies, friendly societies, and trade unions to produce the National Insurance Act. In Korea, there is as yet no actor who has taken it upon themselves to formalize the overseas technology transfer of retired technicians into a formal institution.
Third, Crisis Visibility. In John Kingdon's multiple-streams framework, a policy window opens when the problem stream, the politics stream, and the policy stream converge. In Spain, the death of a Glovo rider opened the policy window.
AI-driven job displacement is a "gradual crisis." Small-scale layoffs repeat somewhere every week, but each event is too small to appear in macroeconomic employment indicators. Without a triggering event, the policy window does not open.
Fourth, the visibility of existing institutional failure. The Naeil Baeum Card's employment rate of 54.6 percent and wage recovery rate of 91.3 percent represent "middling performance." Neither complete success nor complete failure. A result in which half are employed and half are not is too ambiguous to demand fundamental institutional redesign.
When failure is not visible, the social pressure for institutional change does not form.
According to historical patterns, formalization of informal institutions proceeds when at least three of the four conditions are met. The transition from British friendly societies to the National Insurance Act was a case where all four were satisfied. The Spanish Rider Law satisfied three.
For Korea's retired-technician transition support, all four conditions remain unmet.
Park Sang-ho's twelve have not reached the social threshold. There is no political actor who would formalize this network into a formal institution. AI displacement is gradual, so the triggering event to open the policy window is absent. The Naeil Baeum Card operates at a middling level, keeping the visibility of failure low.
All four conditions unmet. This is Korea's current coordinate.
In Book 5, The Strategy of the In-Between, when we discussed national indispensability, we wrote that transition is not a matter of individual awakening but of institutional infrastructure. Individual awakening has already happened — to Park Sang-ho, to Lee Jung-hoon. What is absent is the infrastructure.
9. The Threshold Question
On his way back from Park Sang-ho's gathering, Lee Jung-hoon video-calls his daughter. Sunday evening. On the streets of Hai Phong, motorbikes stream past; from a roadside pho shop, steam rises. Lee Jung-hoon walks with the screen held up.
His daughter is seventeen now. The poster on the wall beyond the screen has changed again. Lee Jung-hoon notices, but does not ask. His daughter's world is changing at a speed he cannot know.
His daughter asks.
"Dad, what do the people there do?"
Lee Jung-hoon thinks for a moment. He could say: people who were pushed out of Korea. People who have skills Korea no longer needs. But that would not be accurate. These people were pushed aside, but they did not stop. Korea may have said it no longer needs them, but here, without these people the factories cannot run. The people most faithful to the previous system were pushed aside, and in the place they were pushed to, they are building a new system.
"People I work with together."
His daughter's question is simple, but for Lee Jung-hoon it is not. The word "together" becoming possible — that happened in Park Sang-ho's back room. Until two months ago, Hai Phong was a place Lee Jung-hoon endured alone. From today it is different.
Just as citizens formed collegia in the era when Crassus's fire brigade was capitalizing on Rome, people seek each other out where formal institutions do not exist.
Lee Jung-hoon's 28 years were "obsolete data" in Korea. In Hai Phong they are "knowledge that still cannot be done without a person." But for individual knowledge to become a sustainable structure, "together" is required. Just as the collegia's funeral fund made the death of an individual the responsibility of a community — just as Park Sang-ho's whiteboard made the survival of an individual into the structure of a collective.
His daughter smiles on the other side of the screen. Lee Jung-hoon smiles too. The first time since coming to Hai Phong that he has smiled in front of his daughter.
The next chapter's story begins from Lee Jung-hoon's daughter. "Dad, what should I study?" What does the transition Lee Jung-hoon is living through in Hai Phong mean for his daughter's generation? Informal institutions help the individual's transition, but the next generation's transition begins in education.
After the call ends, Lee Jung-hoon looks out the window. Hai Phong's night is humid. What is the informal institution you belong to? When did it start, and from whose threshold of zero? And for it to become a formal institution — what is still missing?