Joseph Duveen, the greatest art dealer of his era, had a habit that looks like bad business. He bought whole collections and put them in his basement.1
Not to sell later at a markup. To keep them off the market — so that the paintings he did sell would be rare.
His own line on it: "You can get all the pictures you want at fifty thousand dollars apiece—that's easy. But to get pictures at a quarter of a million apiece—that wants doing!"2
Read that as a job description and the whole thing reorganises. Duveen did not think his job was finding buyers, or finding paintings. His job was producing the price, and the price is a function of how many exist to be had.
Greene's summary of the effect: "The paintings that he sold became more than just paintings—they were fetish objects, their value increased by their rarity."3
Law 16 is a chapter about your own presence — withdraw and your value rises. Greene's move here is to note that the same mechanism runs on things, and that economics already has a name for it.
"Another, more everyday side of this law, but one that demonstrates its truth even further, is the law of scarcity in the science of economics. By withdrawing something from the market, you create instant value."4
Then he generalises it back to the reader: "Extend the law of scarcity to your own skills. Make what you are offering the world rare and hard to find, and you instantly increase its value."5
That last instruction is where the chapter is doing more work than it can support, and it is worth separating the three claims it fuses.
Natural scarcity. There are only so many Vermeers. Nobody made this true.
Withheld scarcity. Duveen's basement. The supply exists and is being held back by someone with an interest in the price.
Manufactured desire. The tulip case, below, where the scarcity is real but the wanting was engineered first.
These have completely different stability properties, and Greene treats them as one law.
Natural scarcity is durable — nobody can add Vermeers. Withheld scarcity is durable only as long as the withholder holds, and collapses the moment they need cash or die. Manufactured desire is the least durable of the three, because the thing being priced is a belief, and beliefs are not stored in a basement.
Greene's other case is the more interesting one, and his framing of it is a claim most people have never heard.
"In seventeenth-century Holland, the upper classes wanted to make the tulip more than just a beautiful flower—they wanted it to be a kind of status symbol. Making the flower scarce, indeed almost impossible to obtain, they sparked what was later called tulipomania. A single flower was now worth more than its weight in gold."6
Note what that account does. It presents the most famous speculative bubble in history as a deliberate product-design decision by an identifiable group — the upper classes wanted a status symbol, engineered the scarcity, and got a mania.
That is a strong causal claim, and it is given in one sentence with no source. 🚩 It also runs against the general shape of the historiography, in which tulipomania is usually described as an emergent speculative episode in a functioning futures market, with the extremity of the crash itself substantially disputed. Flagged rather than repeated: the case as Greene tells it is doing a specific job in his argument — showing that scarcity can be authored — and the authored version is the contested part.
What survives the flag is still useful, and it is the sequencing. In the tulip account the desire came first. The upper classes wanted a status symbol and then made a flower scarce to serve that want. Scarcity did not create the demand; it converted an existing appetite into a price.
That is the difference between the two cases. Duveen sold to people who already wanted Old Masters. The tulip account claims the want itself was installed. The first is a supply operation and the second is a demand operation, and only the first is a thing a dealer can actually do.
Greene's instruction — make what you are offering rare and hard to find — inherits a condition from Duveen's case that it does not carry with it.
Duveen could withhold because he owned the supply. He bought entire collections. The scarcity was real, enforceable, and expensive: capital sitting in a basement earning nothing, for years, to hold a price.
An individual applying this to their own skills owns their supply too, in a sense. But there is a difference that decides everything: a painting in a basement does not stop existing, and an unexercised skill does.
The Old Masters were finished. Withholding them cost storage. Withholding your own work costs the work — the reputation that would have been built, the practice that would have compounded, the evidence that would have accumulated.
So the same manoeuvre has opposite balance sheets depending on whether the asset is complete. Duveen's basement is a warehouse. A person's is a fridge.
There is a second condition in the Duveen case that the general advice quietly drops.
Withholding only raises the price if buyers cannot go elsewhere. Duveen's position rested on having cornered a specific market — the American millionaire collector — and on the Old Masters being a genuinely closed set.
Take away either and the basement becomes a loss. If the buyer can get a comparable painting from another dealer, withholding does not raise your price; it moves the sale to someone else and teaches the buyer they have options.
Scarcity is not a property you can apply to an offering. It is a property of the market the offering sits in, and the chapter's instruction addresses only the half that the reader controls.
You are being asked to do something, again, for free or nearly.
The advice in the air is to be less available — raise your price by being harder to get. Before you do, run Duveen's two conditions, because they are the whole of it.
Do you own the supply? Not are you good — is the thing you provide actually only obtainable from you? Write down who else could do it acceptably. Not as well; acceptably. If that list has names on it, withholding does not create scarcity. It creates a referral.
Is the asset finished or is it a practice? Duveen's paintings sat in a basement and were exactly as valuable when they came out. If what you would be withholding is a completed thing — a body of work, a product, a manuscript — the basement is available to you. If it is a capacity, holding it back does not preserve it. It degrades it, and the market forgets what you were good at faster than it wonders where you went.
Then the move that actually follows from the case, which is not withdrawal. Duveen did not become scarce; he made the alternatives scarce. He bought the substitutes. The analogous move is not making yourself hard to reach — it is doing the specific thing that has no acceptable substitute, which is slower, and which makes the availability question stop mattering.
The strongest evidence is Duveen's own quoted line, because it states the strategy in the operator's voice rather than in the analyst's: getting pictures at a quarter of a million "wants doing" — the price is the product.
Tension: the chapter fuses three kinds of scarcity. Natural, withheld and manufactured-desire have different durabilities and different requirements, and the law treats them as one mechanism.
Tension: the instruction drops the case's own preconditions. Duveen owned the supply and faced a closed substitute set. Neither transfers automatically to a person's skills, and the chapter's extend the law to your own skills carries neither condition.
Tension: a completed asset and a practised capacity behave oppositely under withholding. Greene's cases are all of the first kind and the advice is aimed at the second.
🚩 SINGLE SOURCE · 🚩 SECONDARY WITHOUT PRIMARY · 🚩 CONTESTED — the tulipomania account is given in one uncited sentence and attributes the episode to deliberate engineering by the upper classes. That is a strong causal claim standing against the general shape of the historical literature, in which the episode is treated as emergent speculation and the severity of the crash is itself disputed. [CONTESTED] — do not use this framing in output without checking it against a scholarly source. The Duveen material is [POPULAR SOURCE] and uncited but internally consistent with the same dealer's documented practice elsewhere in the book.
Open questions. How long can withheld scarcity hold before it inverts — is there a point at which a basement full of unsold Old Masters becomes a known overhang that suppresses the price rather than supporting it? And: does the tulip claim have a source anywhere, or has a contested economic episode been recruited as a clean illustration because it makes the point neatly?
Within Law 16, this section is where the chapter's mechanism is most testable — objects have prices, and prices are observable in a way that a courtier's standing is not. Which makes it the useful check on the rest of the law.
Against Deioces the parallel is exact and the correction is sharp. Deioces' withdrawal worked because the Medes' alternative judges were corrupt — no acceptable substitute. Duveen's basement worked because he had bought the substitutes. Both cases are substitutability stories that Greene reads as scarcity stories, and the difference matters because scarcity is something you can perform and substitutability is something you have to actually arrange.
Against Death Before Death, the pairing exposes the law's biggest unstated split. Duveen's withholding is repeatable — he did it as standing practice, across many paintings, for decades. Charles V's withdrawal was a one-shot conversion of a person into a finished thing. The chapter recommends both under one law, and only one of them can be run twice.
Behavioral mechanics — False Scarcity
That page's territory is scarcity claimed rather than held — the countdown timer, the only three left, the artificial deadline that exists to compress a decision.
Duveen is the honest version of the same instrument, and the comparison isolates what actually does the work.
Duveen's scarcity was real. The paintings genuinely were in his basement and genuinely were not available. Nobody was being told a false fact. And the price effect was identical to what a false claim would have produced.
The insight the pair produces: the buyer cannot distinguish them, and that is not a detail — it is the reason the false version exists. A price responds to believed availability, not to actual availability, so the two techniques are indistinguishable from the demand side and differ only in cost to the operator. Duveen paid capital and years for his scarcity; a countdown timer costs nothing.
Which gives a cleaner account of why the false version is corrosive than the usual one. It is not primarily that it lies. It is that it debases the signal for everyone, including operators like Duveen who are paying full price for a real one. Once buyers learn that scarcity claims are cheap to fabricate, they stop reading scarcity as information — and the person who genuinely owns the supply loses the ability to demonstrate it. The forgery does not merely deceive; it destroys the market for the authentic instrument.
Psychology — Loss Aversion and Asymmetric Valuation
Losses loom larger than gains, consistently. The same quantity registers more strongly on the way out.
That supplies what Greene asserts — by withdrawing something from the market, you create instant value — with a mechanism, and it also predicts the shape of the effect.
Scarcity does not make people want a thing more in the abstract. It reframes not-getting-it as a loss rather than as a non-gain. A painting nobody can buy is not an opportunity foregone; it is something being taken away, and the curve is steeper on that side.
What the pair produces is a limit the chapter does not have. The reframe requires the buyer to have already placed themselves, imaginatively, in possession — you cannot lose what you never held, even hypothetically. Which is exactly the sequencing the tulip account describes and Duveen relied on: the upper classes wanted the status symbol first; Duveen's clients already wanted Old Masters. Scarcity is an amplifier with no output of its own, and applying it to something nobody has yet imagined owning produces nothing at all. That is the failure mode behind most of the instruction's real-world misuse — people withdraw an offering the market had not yet learned to want, and interpret the silence as needing more scarcity.
Sharpest implication. Duveen's move was not becoming scarce. It was buying the substitutes — his own availability was never the variable. Scarcity is a property of a market, not a posture you can adopt, and the version of this advice aimed at individuals quietly swaps one for the other.
Generative questions.