Reputation instead of coercion: the justice system that worked for centuries

A Mediterranean merchant of the eleventh century entrusts a cargo to an agent who sails for a port weeks away. There the agent will sell the goods, collect payment and -supposedly- remit the money. No court within the merchant's reach has jurisdiction over that distant port; no public force will go and recover whatever the agent decides to keep. Between the two there is no bond but a promise. What stopped the agent, then, from simply stealing?

The answer is not law, nor force. It is information. And the merchants who turned it into a system of enforcement built, without setting out to, a justice that pronounced and enforced its decisions autonomously. It worked for centuries. It is worth understanding how, because it is the best map of what technology makes possible today.

The coalition that disciplined itself

Avner Greif (1993) reconstructed one of the clearest cases from the documents of the Cairo Geniza: the coalition of the eleventh-century Maghribi traders. Their problem was that of any long-distance agency relationship -the principal cannot monitor the agent, and the agent has every incentive to defraud-, aggravated by the slowness and opacity of maritime commerce. Going to a court was, in practice, useless: by the time the merchant could even file his claim, months after the transaction, no judge was in any position to verify the condition in which the goods had arrived, the price they fetched or what bribe had to be paid at the port. And, in fact, they scarcely did.

In place of courts operated a reputation mechanism sustained by a rule as simple as it was devastating: a merchant would never again employ an agent who had defrauded any member of the coalition. Not only the one who had cheated him: anyone. Dishonesty toward one brought exclusion by all.

What turned the gears was not the threat but the fact, and it travelled by letter. Around 1055 it became known in Fustat that Abun ben Zedaka, an agent settled in Jerusalem, had embezzled the money of a Maghribi trader. The response was unanimous: the traders ceased all dealings with him, and his own correspondence reveals that the ostracism reached merchants as far away as Sicily. Only when he compensated the injured party was he readmitted. That was the whole system: the agent who cheated did not lose a client, he lost the market, because everyone came to know. The sanction was not handed down by a judge; it was imposed by the community, and enforced in the only way that truly hurts in commerce: by ceasing to trade with him.

When trade grew, reputation needed an institution

That system had a limit, and it was one of scale. It worked because the coalition was a closed group in which everyone knew everyone and information travelled among acquaintances. But the European commerce of the twelfth and thirteenth centuries overflowed that frame: at the great fairs -those of Champagne, crossroads of continental traffic- gathered merchants from different kingdoms, who did not know one another at all and who perhaps dealt only once. When anyone can be your counterparty, how do you know whether the stranger before you left an unpaid debt at the other end of the continent?

The answer was an institution of remarkable elegance. The judges of the lex mercatoria were not, above all, a force that punished, but an archive that could be consulted. The mechanism was almost a protocol. Before closing a deal, the merchant asked the judge whether his counterparty carried any unpaid judgment; if there was one, he did not deal. If, once the deal was closed, he was cheated, he went to the judge, who handed down his ruling and entered it in the register, where the next person to ask would find it. The judge seized no one's goods: he kept the record and made it available. And that was enough, because it returned reputation to its function with a fraction of the information the old coalition had required: it was no longer necessary for everyone to know everything about everyone, only that it be on record who had not complied. The collective boycott did the rest, and whoever defaulted was left, quite simply, with no one to trade with.

A pattern that became institutionalised

What the fairs improvised eventually settled into place. Over the late Middle Ages, the lex mercatoria crystallised into permanent merchant courts -the Consulados del Mar in the Mediterranean, the Courts of the Staple in the Anglo-Saxon world- which not only settled commercial disputes but enforced their own decisions with the same instruments as ever: reputation and exclusion (Trakman, 1983). It was, then, no isolated or ephemeral experiment, but the ordinary way in which commerce administered justice to itself for centuries, until the consolidation of the modern State claimed for itself the monopoly of enforcement and subordinated that order to the public courts. But that is another story; what matters here is that, long before that nationalisation, the system existed and worked.

What actually bound contracts

Behind all these cases lies one and the same mechanism, and it is worth naming it precisely, because that is what makes it transferable. It was not coercion that held those contracts together: it was a signal. Reputation worked as an invisible bond -what Klein and Leffler (1981) would formalise much later-: a party complies when the future business it would lose by being excluded exceeds what it would gain by defrauding today. In that light, breach ceases to be profitable and compliance becomes the rational strategy, even for the most self-interested.

But that signal depended on one condition, and one alone: that information circulate. Reputation is worth nothing if the cheating is not known. That is why each of those systems was, at bottom, an information technology: the letter network of the Maghribi coalition, the consultable register of the fair judge, the memory of the guild. And that is why its limit was always informational. They worked within a community or a fair, where what was known reached everyone; they weakened as soon as the cheating could hide in distance or anonymity. The reach of the system was, exactly, the reach of its information.

The precedent, and the one thing that has changed

Here history ceases to be antiquarian. If enforcement without a State worked for centuries, and if the only thing that limited it was how far information reached, then the pertinent question is not whether it can work again, but what has changed in that limit.

It has changed completely. What it cost a fair judge to maintain in a parchment register -who complied, who did not, and that anyone could consult it- is today trivial on a planetary scale. A shared, permanent and immutable record -a blockchain- of contractual conduct is nothing other than the archive of the Champagne fair judge rebuilt without its limits: without borders, without requiring the parties to know one another, without depending on a guild to keep memory. Reputation can, for the first time, travel as far as commerce itself.

And there is no need to imagine it, because the mechanism is not only still alive: it is the same. When a dealer defaults on an award of his tribunal in the international diamond trade, his photograph is displayed and that information circulates through every bourse in the world, and he is readmitted only if he makes full restitution to the injured party, without a single state judge intervening (Bernstein, 1992). It is, feature for feature, what happened to Abun ben Zedaka in 1055.

Those merchants were not naïve: they were, without knowing it, excellent economists. They solved with reputation and information a problem we are posing ourselves again, and they did so without asking permission of any power. The only difference is that what limited them -how far what was known could reach- has ceased to be a limit.

At IMPERA we work on these ideas. And I leave open the question they could not have posed: if reputation sustained commerce wherever information reached, and information no longer meets a border to stop it, what does the state monopoly of enforcement have left to contribute? I look forward to reading you in the comments.