In May 1925 the five most successful dealers in the French scrap-metal business each received an invitation to an "official" but "highly confidential" meeting with the deputy director general of the Ministry of Post and Telegraphs.
The venue was the Hotel Crillon, then the most luxurious hotel in Paris. When they arrived it was the director general himself who met them, in a suite on the top floor. His name was Lustig.
After drinks he explained. *"Gentlemen, this is an urgent matter that requires complete secrecy. The government is going to have to tear down the Eiffel Tower."*1
And then the reasoning, which is the part that works: the tower had recently been in the news needing repairs; it had only ever been meant as a temporary structure for the Exposition of 1889; maintenance costs had soared; the government was in a fiscal crisis and would have to spend millions. Many Parisians thought it an eyesore. Over time even the tourists would forget it — "it would live on in photographs and postcards."1
"Gentlemen, you are all invited to make the government an offer for the Eiffel Tower."
The apparatus is minimal and every piece of it is doing one job.
Sheets of government stationery filled with figures, including the tonnage of the tower's metal.2 Not a claim — a work surface. The dealers' eyes "popped as they calculated how much they could make from the scrap," and a man doing arithmetic is not a man evaluating a premise.
A waiting limousine, which converts a proposition into a logistics.
A flashed official badge at the tower, and a guided tour "spic[ed] with amusing anecdotes." The badge is shown, not examined. The anecdotes are the tell: a man conducting a fraud has no time for charm, so charm reads as a man who is not conducting one.
A four-day deadline for offers, delivered to his suite.
Nothing here is forged at scale. The stationery is the only manufactured object, and its content is a set of figures the marks are invited to check against a tower they can see.
Greene's interpretation is the sharpest sentence in Law 28.
*"Had Count Victor Lustig tried to sell the Arc de Triomphe, a bridge over the Seine, a statue of Balzac, no one would have believed him. But the Eiffel Tower was just too large, too improbable to be part of a con job."*3
Read the inversion. Improbability is normally a cost — the less likely a story, the more evidence it needs. Here it is a credential, because the mark's reasoning is not is this true but would anyone attempt this.
And the answer to the second question is genuinely no. Nobody would attempt it. Which means the audacity is not covering the con's weakness; the audacity is the evidence.
Greene's general form: "Largeness of scale deceives the human eye. It distracts and awes us, and is so self-evident that we cannot imagine there is any illusion or deception afoot."3
There is a further piece the chapter passes over. The dealers were not asked to believe something absurd about the world. They were asked to believe something mundane — that a fiscally strained government would scrap an expensive temporary structure — and every element of that was independently plausible in 1925. The absurdity is entirely in the object, not in the argument, and the argument is what they checked.
A few days after handing over a certified cheque for more than 250,000 francs, Monsieur P. began making telephone calls. There was no deputy director general Lustig and no plan to destroy the tower.4
"Monsieur P. never went to the police. He knew what kind of reputation he would get if word got out that he had fallen for one of the most absurdly audacious cons in history. Besides the public humiliation, it would have been business suicide."5
That is the most important sentence in the case and Greene does not treat it as a design feature.
The scale that made the story credible is the same property that makes reporting it unbearable. A man defrauded of 250,000 francs over a plausible contract has been unlucky. A man who bought the Eiffel Tower is a joke for the rest of his working life, and the joke is proportionate to exactly the audacity that convinced him.
So the con contains its own silencing mechanism, and the mechanism scales with the con. Bigger lie, better credential, quieter victim.
The proof that the silence was structural rather than lucky is what happened next.
"In fact it was so improbable that Lustig was able to return to Paris six months later and 'resell' the Eiffel Tower to a different scrap-iron dealer, and for a higher price—a sum in francs equivalent today to over $1,500,000!"3
Six months. Same city. Same landmark. Same trade — the Paris scrap-metal business, which cannot have been large.
For that to be possible, the first mark must not have told anyone: not the police, not his competitors, not the four other dealers who had been in the room at the Crillon and knew there had been a sale.
Which means the operation's real product is not the sale. It is a market in which the loss cannot be reported, and Lustig priced the second one higher because he had learned what that market would bear.
There is a detail worth holding against the chapter's framing. Greene presents this as boldness. It is at least as much patience — a six-month wait for the story to fail to circulate, which is a bet about other people's shame rather than about his own nerve.
You are about to present something and you have quietly reduced it.
Not the price — the ambition. The pilot instead of the programme. The three-month trial instead of the two-year build. And the reason you reduced it, if you are honest, is that the larger version felt like it would invite a question you could not answer.
Now look at what the reduction actually costs. A small ask gets evaluated as a purchase: line by line, against alternatives, by someone whose job is to find the flaw. A large one gets evaluated as a proposition — would anyone bring this if it weren't real — and that is a different and much easier test to pass.
This is Greene's mechanism and it is genuinely available to honest people, which is precisely why it needs a check.
Here it is, and it is one question. If this fails, can the other party say so out loud?
Lustig's mark could not. That is what made the con repeatable, and it is the property that separates ambitious honest work from the thing in this chapter. If your proposal is structured so that the person who backs it would be humiliated to admit it went wrong, you have built the silencing mechanism whether or not you meant to — and you will not hear about the failure, which means you will do it again.
Watch for it in the specific form it takes: a proposal whose champion's reputation is fused to it before any result exists. That fusion is what Lustig manufactured with the Crillon suite and the government stationery. It is also what a certain kind of internal project deck does by accident, and the tell is the same — nobody ever reports the outcome.
Set the case against Rommel's distinction between a risk and a gamble — the same action, the same odds, differing only in whether you survive losing.
Lustig's operation was, on that test, a risk rather than a gamble, and deliberately so.
If Monsieur P. had gone to the police, Lustig would have been hunted. But the whole architecture made that outcome improbable: an unreportable humiliation, a victim with a business reputation to protect, four other dealers who did not know which of them had won, and a story so absurd that telling it convincingly would itself be difficult.
The audacity that made the con work also capped its downside, and that conjunction is what let him run it twice rather than once.
Which is the thing Greene's chapter cannot say, because his six mechanisms grade boldness by its effect on observers and contain no term for recoverability. On his account Lustig is bold. On Rommel's, Lustig is prudent, and the boldness is the visible surface of a carefully bounded exposure.
This is the third page in this build carrying Count Victor Lustig, and the recurrence is worth recording because the PRD's table does not list him.
The vault already holds Count Victor Lustig and the Blank Telegrams — the Japanese chauffeur, the deliberately misplaced medal or armband, "seen not as tasteless but as odd and intriguing" — and Lustig and Al Capone, where he takes $50,000 for sixty days, does nothing with it, and returns it with an apology.
Three appearances, three laws, three different mechanisms: manufactured intrigue, manufactured trustworthiness, and manufactured improbability. He joins Bismarck, Talleyrand, Duveen and Erickson as a figure Greene uses repeatedly to prove separate things, and — like all of them — the book never assembles him into a single practitioner with a method.
Which is a loss, because assembled, the three cases show one consistent principle: Lustig's product is always an inference the mark makes unprompted, and he never states the claim the mark ends up holding.
🚩 SINGLE SOURCE · 🚩 SECONDARY WITHOUT PRIMARY. No citation anywhere. Monsieur P. is unnamed; the four other dealers are unnamed; the sums are given in modern dollar equivalents without a conversion basis. The Lustig legend is substantially a product of later popular retelling, and Greene supplies no way to distinguish the documented parts from the polished ones.
🚩 The second sale is the least verifiable and most load-bearing claim. "Return to Paris six months later and resell the Eiffel Tower" is the fact that makes this an argument rather than an anecdote, and it is asserted flatly.
Tension with the chapter's own Reversal. Greene's Reversal says boldness "should never be the strategy behind all of your actions. It is a tactical instrument." Lustig's career, across the vault's three pages, is a standing posture — the misplaced medal, the Continental accent, the count's title — maintained continuously, not deployed at a moment.
Open question. The con's silencing mechanism scales with its audacity: the bigger the story, the more unbearable it is to report. Does that produce a general rule about which frauds get detected — and would it predict that the most-reported cons are the modest ones?
Greene's Law 12 Reversal tells the bribe half of this same episode, and the build-once list assigns it here. It is built separately at The Bribe That Restored Confidence because the mechanism is distinct — a request for money functioning as proof of authenticity — and cross-referenced from both laws rather than duplicated.
There is a sharper convergence with Greene's own Law 6, Court Attention at All Cost. Lustig's blank-telegram page runs on being conspicuously odd; this one runs on being conspicuously plausible. Same operator, opposite visibility strategies, and Greene never notices that his exemplar of one law is his exemplar of its neighbour.
And a tension with Law 3, Conceal Your Intentions. Lustig conceals nothing about his intention — he says outright that he intends to sell the Eiffel Tower. What is concealed is his standing, not his purpose, and the chapter's vocabulary has no way to mark the difference.
Behavioral mechanics — Risk versus Gamble
Rommel's distinction, on that page, is that a risk and a gamble are indistinguishable from inside the decision — both are actions with only a chance of success, both improved by acting boldly, both undertaken on incomplete information. What separates them is a single external fact: whether you can survive losing.
Greene's six mechanisms have no such term, and the Eiffel Tower case is where the omission shows most clearly.
Read as a gamble, Lustig's operation is insane: he impersonates a senior official, in the most conspicuous hotel in Paris, to five men who all know each other, about the most famous structure in Europe. Discovery means prison.
Read as a risk, it is carefully bounded — and the boundary is the victim's shame. The same audacity that made the story credible made its reporting unbearable, so the failure mode was closed off by the very property that produced the success.
What the pair yields is a correction to how this case is usually taught. It is not evidence that boldness works. It is evidence that boldness which simultaneously closes your escape route and closes theirs is a different instrument entirely — and that the operator's real skill was identifying a mark whose loss could not be spoken.
Which supplies a diagnostic Greene's chapter cannot generate. Before any bold move, the question is not do I have the nerve but who has to be able to talk about this if it fails, and can they? Lustig ran that calculation. His mark did not.
Business — We Don't Negotiate With Terrorists — Never Lower Price
The rule there is close to absolute. A prospect asks for five hundred off and the answer is we could do it for more. The mechanism is anchoring — the prospect has tried to anchor downward, the closer counter-anchors upward, and the original number becomes the reasonable middle. And the deeper claim: a discount does not cost margin, it reprices you, because the moment the number moves, every number you have ever given is understood as an opening position.
Lustig runs that move at the exact moment of maximum danger, and Greene records it precisely.
Monsieur P. arrives with the cheque and starts to doubt — why a hotel instead of a government building, why no other officials, is this a hoax? And Lustig's response is the one nobody expects: "Instead of backing down, or lowering his price, he simply raised his price higher, by asking for and getting a bribe."6
The pairing shows why that works, and it is not bravado. A person about to walk away is running a hypothesis: this is not real. Every accommodation confirms it, because accommodation is what an anxious fraud offers. An increase is the one response the hypothesis does not predict — a fake official does not create new obstacles between himself and a certified cheque.
So the counter-anchor is not primarily about price. It is about which story your behaviour is evidence for, and lowering is evidence for exactly the story the doubter is entertaining.
And it gives the honest version of the move, which the sales page already holds and this case makes vivid: when someone doubts your legitimacy, the useful response is to behave in a way a fraud could not afford — add a condition, slow the process, invite a check. Lustig's version was a bribe. The structure is available without the crime.
Sharpest implication. The property that made the con credible and the property that made it unreportable are the same property. A story too improbable to be a con is also too humiliating to admit having believed — so audacity does not merely improve the odds of success, it suppresses the evidence of failure, and it does so in proportion to itself. Which predicts something uncomfortable about the detection rate of frauds generally: the ones we hear about are not the incompetent ones, they are the modest ones, the ones a victim can describe without becoming a joke.
Generative questions