The Districts: What Every Country Decided to Do About This

Begin with a fact that reverses what most readers of this chamber expect, including the Japanese ones.

Sweden is widely admired for a law passed in 1999 that criminalised paying for sex while leaving the person selling it free of penalty. It is called the Nordic model, it has been exported to Norway, Iceland, France and elsewhere, and it is usually presented as the moment a country finally stopped punishing women for their own exploitation.

Japan did that in 1956.

The Anti-Prostitution Law declares prostitution prohibited and attaches no criminal penalty to the act — not for the woman, not for the man. Everything punishable sits around it: procuring, soliciting in public, coercing, taking the money, providing the premises, running it as a business, financing it. The woman who sells and the man who buys are, in the narrow matter of the act itself, outside the reach of the criminal law, and have been for nearly seventy years.

Japan is almost never credited for this, and the reason is worth understanding, because it explains the whole strange architecture that followed. Japan did not arrive there by feminist design. It arrived there through a post-war abolition movement that understood prostituted women as victims of poverty, and therefore found it politically impossible to prosecute them. The result looks, from a distance, like the Nordic model minus the part that punishes the buyer. Nobody planned it as a model. It was a compromise, and compromises do not get credited.

Here is the finding that makes this an Around the World essay rather than a complaint about Japan. There is no country that solved this. There are five distinct things states do, each one a different answer to the question of what to protect when you have accepted you cannot make the trade disappear.

Prohibition. Criminalise all of it — buying, selling, organising. Most of the United States, China, most of the Middle East, much of Africa. The thing protected is the public moral position. The cost is that the person with the least power in the transaction is also criminally liable, which makes her unable to report violence against her without confessing to a crime.

Abolitionism, the Nordic model. Criminalise the buyer, free the seller. Sweden 1999, then Norway, Iceland, France 2016. The thing protected is the principle that buying a person’s body is an act of harm regardless of consent. The contested cost is whether it drives the trade into places where the seller is less safe, which the evidence genuinely does not settle in either direction.

Regulationism. Licence it as a business. The Netherlands lifting its brothel ban in 2000, Germany in 2002 and again with tighter rules in 2017, parts of Australia, one American state by county. The thing protected is the worker’s formal status — contracts, health access, tax standing. The cost is that licensing creates an unlicensed remainder, and the remainder gets darker as the licensed part gets cleaner.

Decriminalisation. Remove the criminal law entirely and apply ordinary labour, health and safety law instead. New Zealand, 2003, which is the clearest single example anywhere. The thing protected is the worker’s ability to use the same institutions as everybody else — to call the police, to sue, to refuse a client and have the refusal mean something. The cost is political: it requires a society willing to say out loud that this is work.

And then there is the fifth, which is Japan’s, and which almost nobody else runs.

Japan prohibited prostitution without punishing it, and then defined it so narrowly that an entire industry could grow in the space the definition left.

The offence is built around one specific act. Everything that is not that act is not prostitution in Japanese law — not as a loophole, not as a tolerated evasion, but because the statute simply does not reach it. A great deal of what the districts sell was never illegal and never needed to be excused.

Then, in the mid-1980s, a second statute did the part everyone assumes is missing. It gave the industry a name, required every business in it to register with the prefectural authorities, zoned it by local ordinance at fixed distances from schools and hospitals and childcare facilities, and froze new premises-based openings. Existing shops continued under acquired rights. In most of Japan, a new soapland cannot legally open. The ones that exist are the ones that already existed.

This is not an absence of policy. It is a policy with a clear and arguable logic: a trade the state can see, count, locate and inspect is preferable to one it has driven somewhere it cannot follow. It is the containment answer, and it was legislated openly, debated in the Diet, and written down.

What distinguishes it from the other four is what it optimises for. Prohibition optimises for the moral position. The Nordic model optimises for the principle. Regulationism optimises for formal status. Decriminalisation optimises for the worker’s access to ordinary institutions. Japan optimises for the contradiction being manageable — visible enough to police, quiet enough not to require anybody to resolve it.

The question worth asking is what keeps these districts from simply being abolished, and there are four answers operating at once. None of them is protection in the sense the question implies.

The first is a legal doctrine, and it is the literal mechanism: acquired rights. A business lawfully established under a previous rule is not retroactively destroyed when the rule changes. This is not a favour done for this industry; it is a general principle of administrative law that protects pharmacies and petrol stations by the same logic, and it is why the 1980s freeze stopped new openings rather than closing existing ones. Closing them would have required compensation and a political appetite nobody had.

The second is the containment argument, which is the honest policy case and deserves to be stated at its strongest rather than dismissed. Abolition does not end demand. It relocates it. A trade conducted in a registered premises in a zoned district, with an operator who has a name and an address and something to lose, is a trade where a violent client can be identified and a minor can be discovered. Drive it into apartments and private arrangements and none of that remains. Whether this argument is correct is contested. That it is an argument, rather than an excuse, is not.

The third is definitional, and it is the deepest of the four: you cannot abolish what you have not defined, and every definition leaves a remainder. Narrow your definition and the trade reorganises just outside it. Widen it and you begin criminalising things no society is willing to criminalise. Every country discovers this. Japan simply discovered it early and then built on the remainder instead of chasing it.

The fourth is electoral arithmetic, and it is the least dignified. No politician’s career improves by reopening this. There is no constituency demanding it, there is a constituency that would be damaged by it, and the current settlement has the enormous political advantage of already existing. Things persist because nobody gains from ending them, and that is true of a great deal more than this.

This is the detail that makes the arrangement feel like conspiracy, and the explanation is duller and more interesting than conspiracy.

The districts predate the zoning by centuries. Yoshiwara was a licensed quarter from 1617. The red-line system ran until 1958. When the modern regulatory map was drawn in the 1980s, it was drawn around places that had been there for three hundred years. These are not where the trade moved to. They are where it already was when the law arrived to find it.

Police boxes are not on the exclusion list. The ordinances keep these businesses a set distance from schools, hospitals and childcare facilities, because the concern was children and patients. Nobody wrote a setback from a police box, because proximity to police was never the thing anybody was worried about.

And the causation runs backwards from how it looks. A police box in a nightlife district exists *because* it is a nightlife district — dense crowds, alcohol, cash, assault, touting, fraud. Heavy police presence in such a place is the staffing response to a high-incident area, not a blessing conferred on it. Reading the police box as approval gets the arrow the wrong way round.

There is a real double standard here and it is almost always described in the wrong place.

It is usually framed as men versus women: men may buy, women are shamed, and the state indulges the first and polices the second. There is something to this, and it is not the main structural asymmetry.

The main one runs between the operator and the worker. Look at what the Japanese arrangement does. The business is registered, zoned, legally recognised, able to advertise, able to bank, able to sue and be sued, protected by acquired rights, and shielded from the prostitution offence by a structure of independent contracting and separated payments that makes the operator’s involvement hard to prove. The worker, inside that same arrangement, is an independent contractor in a trade the state has officially declared prohibited — which means there is no occupational health regime for her, no labour-rights framework, no collective bargaining, and no mechanism that addresses coercion before it has become a criminal case.

This is not a Japanese failing. It is the structural signature of every model except decriminalisation. Prohibition makes her a criminal. The Nordic model declares her a victim, which removes her liability and also removes her standing to describe her own situation as anything else. Regulationism gives formal status to whoever can obtain a licence and leaves everyone who cannot in a darker place than before. In four of the five models, the person with the least power ends up with the least protection, and the reason is the same in each: you cannot write labour law for work you have officially abolished.

That is the double standard worth naming. It is not primarily about who is allowed to want. It is about who the arrangement is built to protect.

Now the question this chamber exists to ask. Every society has some form of this. Why does each one look so specifically like itself?

The lazy answer is that men want sex everywhere, so a trade appears everywhere, and the local colour is decoration. That answer cannot be right, because it does not predict the variation. If the demand were simply for a sexual act, the cheapest and most available form would dominate everywhere, and it does not. What commands a premium differs by country, systematically, and the pattern is legible.

Japan built an enormous industry around conversation. The hostess club sells a woman’s attention across a table, with no sexual contact at all, at prices that make no sense unless attention is the product. The host club does the same for women. An entire tier of this economy exists in which nothing happens physically and people spend a great deal of money.

The Anglophone escort market names its premium tier the girlfriend experience — which is an astonishing thing to say out loud. It is an admission, printed on a price list, that what is being purchased is the relationship rather than the act.

Korea’s 방 culture, the Japanese snack bar, the various forms across Europe that organise themselves around a long dinner: in each case a society has built a market at the precise point where it refuses to supply something freely.

Which gives the thesis this essay is actually for. The shape of a country’s version is a photograph, in negative, of what that society has decided not to give away. Where adult touch outside a romantic contract is culturally unavailable, a market in touch appears. Where a man cannot be uncertain or tired or frightened in front of his wife without losing standing, a market appears in permission to put the role down. Where being listened to without being advised is not something friends do, a market appears in listening. The districts are not evidence of appetite. They are an inventory of scarcity.

Follow that inventory one step further, because it leads somewhere that matters more than the regulation does.

If you look at what costs the most in these markets, across every country, it is not the most explicit thing. It is duration, exclusivity, unhurriedness, being remembered between visits, being treated as a particular person rather than a customer. The premium attaches to the parts that resemble being known.

That is a finding about human beings and not about any industry. People do not pay a premium for release; release is cheap and getting cheaper and has been thoroughly solved by technology that requires no one else to be present. What stays expensive is another consciousness paying attention on purpose. The market prices it high because it is genuinely scarce, and it is scarce because it cannot be manufactured.

Which means the districts are telling us something unflattering and fairly precise about modern life: that attention has become a commodity, that loneliness is a by-product every society manufactures while optimising for something else, and that wherever a society produces it in volume, a market will eventually arrive to price it.

And here the essay has to stop and refuse its own momentum, because everything above can be read as an argument it is not making.

The argument it is not making is: this appears in every society, therefore it is natural, therefore it is legitimate, therefore the objections are prudishness. That inference is invalid and it is worth being blunt about why.

Near-universality across human societies demonstrates pressure, not legitimacy. Slavery was near-universal. Child marriage was near-universal. Judicial torture was near-universal, for millennia, in societies of every level of sophistication. The fact that a practice appears everywhere tells you that something about how human societies organise themselves keeps producing it. It tells you precisely nothing about whether it is good, and a pattern that recurs under conditions of inequality is not thereby excused by its own frequency.

So the honest position is two-handed and uncomfortable. The demand underneath these markets — for touch, for attention, for a few hours of not performing — is ordinary, human and not shameful, and treating the people who feel it as degenerates is both cruel and analytically useless. And the industries that have grown around that demand contain real and serious harm: coercion, debt bondage, trafficking, the targeting of minors, and a structural arrangement in which the person taking the physical and legal risk is the one the law protects least. Both of those things are true at once, and any account that holds only one of them is propaganda for somebody.

This chamber’s rule applies here as everywhere: the job is to understand the society, not to exonerate it.

Step back from the statutes and there are three things the global pattern says about human beings, and none of them are about sex.

The first is that every society knows more than it can admit. Each of the five models is a device for holding a contradiction without resolving it. Japan narrows a definition. Sweden relocates the guilt onto one party and declines to look closely at what happened to the other. The Netherlands licenses the visible part and accepts a darker invisible part. Prohibition declares the thing absent and funds a vice squad to manage its absence. Nobody has solved it, and the variety of the evasions is a better guide to each national character than any tourist board.

The second is that states are far more willing to regulate location than conduct. Japan knows exactly where every one of these businesses is, to the metre, by ordinance — and has almost nothing to say about how anyone inside them is treated. That ordering of priorities is not unique to this industry or to this country. It is what states do when they want a problem contained rather than addressed.

The third, and it is the one worth carrying away, is that the trade is a measuring instrument. If you want to know what a society withholds, look at what it sells at a premium behind a discreet door. A country with an enormous market in paid conversation is telling you something about how hard ordinary conversation is to come by there. A country whose escort market advertises the girlfriend experience is telling you what has become scarce inside actual relationships. Read that way, the districts stop being a moral curiosity about other people and become a report on the place you live.

There is a coda to all of this that almost nobody notices, and it is the most hopeful thing in the essay.

Every premises-based institution described above is old. The districts are centuries old; the 1980s freeze means the shop-based businesses in Japan are, by law, the ones that were already there. That industry inherited its geography, its assumptions, its vocabulary and its habits from a world that made them.

The services aimed at women arrived afterwards. They appeared in recognisable form only in the last fifteen or twenty years, and they arrived after the freeze — which means they could never be premises-based, could never have a district, and will not acquire one. There is no women’s Yoshiwara and there never will be, and the reason is not modesty or discretion or anything about women at all. It is an accident of timing in administrative law.

The consequence is unexpectedly large. This is the first category of intimate service in Japanese history being built with no inherited geography — no historic quarter, no red-line legacy, no four hundred years of accumulated assumption about what the encounter is and who holds the power in it. It does not have to reproduce any of it, because it did not inherit any of it.

Which is a rare situation. Most institutions are trapped by their own history. This one, by pure regulatory accident, gets to decide what it is.

The statutory descriptions here are outlines of the structure, not legal advice, and the architecture is summarised rather than cited. Japanese statutory provisions, and especially the prefectural ordinances that do the zoning, vary and are amended; anyone with a practical reason to rely on any of this should read the current text rather than this essay.

The comparative models are characterised at a level of generality that necessarily flattens them. Germany’s 2002 framework and its considerably stricter 2017 revision are not the same law. “Most of the United States” conceals real state-by-state variation including one genuine exception. New Zealand’s 2003 decriminalisation is the cleanest example of its category and not a claim that it eliminated harm.

On outcomes, this essay deliberately declines to adjudicate. Whether the Nordic model reduces total harm or displaces it, whether licensing improves worker safety on balance or mainly sorts workers into a protected tier and an exposed one — these are live empirical disputes with serious researchers and serious advocacy organisations on several sides, and the underlying data is poor for the obvious reason that the activity is stigmatised and partly criminal and therefore badly measured everywhere. An essay that announced a winner would be overstating what anybody knows.

The claim that Japan’s 1956 decriminalisation of the seller prefigured the Nordic model by decades is a reading of the statute’s effect, not a claim that the two laws share a purpose. They do not: one declines to punish the buyer and the other exists specifically to punish him.

The negative-image thesis — that a country’s version is a photograph of what it refuses to supply freely — is this essay’s own framing. It is offered because it has explanatory power and it is not a finding from any literature. The observation about premium pricing attaching to duration and recognition is drawn from how these markets advertise themselves, which is a cultural signal and not a measurement.

And the essay speaks only about societies and arrangements. It says nothing about the moral situation of any individual who has bought or sold, which is not a judgement this chamber is equipped to make or has any business making.

The question to leave a reader with is not whether any of this should exist.

It is the one the measuring instrument asks. If the districts are an inventory of what a society declines to supply freely, then a reader in Japan can run the instrument on her own life without going anywhere near one. What is on your list? Which ordinary human thing — being touched without it meaning a contract, being listened to without being advised, being permitted to be uncertain out loud, having a few hours where nothing is required of you — has become something you would have to arrange, and pay for, and probably not mention?

That list is not a symptom and it is not a confession. It is just a reasonably accurate report on the society you happen to live in, which built itself around other priorities and left those things unsupplied.

This Library can help with one part of it, and is honest about which. What we can offer is language — a way of naming something that used to arrive with its name attached, so that it becomes a thing you can say rather than a vague weather inside you.

And the other thing is smaller and more specific than anything in this essay. An evening in which somebody is paying attention, the terms set in advance and by you, where saying what you want is the activity rather than the risk, and where nothing is inferred from your silence. Not a district, not a solution, and not a position on any of the five models. A few hours of being accompanied, which is the one item on the list that can be arranged for a Tuesday.