The Arbitration Paradox: Why the Best Mechanism for Justice Remains a Niche

Picture a product that solves a problem better than its dominant alternative. It is faster. It is designed by specialists in the field, not by overstretched generalists. It bends to the will of those who use it instead of imposing a rigid procedure. And its outcome is recognised in more than 170 countries thanks to an international treaty that state justice, bound by its borders, does not have.

That product exists. It is called arbitration. And yet, despite all these advantages, it remains a niche: reserved for large commercial contracts, for million-dollar disputes, for a club of companies and institutions that can afford it. For the vast majority of economic relationships —a supplier and its client, a small business and its distributor, two companies signing a supply agreement— the default route is still the courthouse, with its years of waiting.

Why? If arbitration is so superior, why has it not displaced state justice on the very ground where it should win comfortably?

I call this the Arbitration Paradox, and I believe the usual answer —“it's a cultural matter,” “there's a lack of awareness,” “it will come in time”— is comfortable and mistaken. Arbitration is not stalled by inertia. It is stalled because it operates within an architecture of incentives that, furtively, rewards precisely the opposite of what it should. And as long as that is not understood, no procedural reform is going to change it.

Let me take the paradox apart into its four pieces.

First piece: litigating is artificially cheap

When a company decides whether to take a conflict to court or to arbitration, it does the maths. And that calculation is rigged from the outset, because the price it pays to go to court does not reflect what that service truly costs.

The administration of justice is funded through taxes. The fees paid by the litigant cover a minimal fraction of the real cost of the process: the salaries of judges and clerks, the buildings, the infrastructure. The rest is paid by the taxpayer, whether they litigate or not. This means the litigant receives a kind of subsidy on litigation: a substantial part of the bill is paid by someone else.

Arbitration, by contrast, is paid in full. Arbitrators' fees, institutional charges, the seat. Everything.

So the comparison the company makes is not “slow, free justice versus fast, expensive justice.” It is a comparison between a service whose price is artificially reduced and another that shows its full cost. On equal terms, many would choose to pay for speed and specialisation. But the terms are not equal: the State's thumb is on the scale.

Added to this is a human factor that every lawyer knows. The professional who advises the company tends to know the courthouse —they have set foot in it a thousand times— and not so much arbitration. Recommending the familiar is safer for the one doing the recommending. The result is a bias toward the usual that reinforces itself: because it is used little, it is known little; because it is known little, it is used little.

Second piece: information does not circulate

Here there is something that is, curiously, at once arbitration's greatest virtue and its greatest structural limit: confidentiality.

That proceedings are private is an enormous advantage for the parties. No one wants their commercial disputes on the front page of the newspaper. But that same privacy has a collective cost: accumulated experience does not circulate.

Each arbitral institution is an island. Its awards, its criteria, the conduct of the parties that have passed through it —all of it remains locked within its walls. There is no way to know whether a company I am about to contract with has a history of complying with awards or of fighting them to the end. There is no way for the criteria developed by one institution to nourish the system as a whole.

Compare it with what happens in a market that works: prices aggregate and transmit information. A high price warns of scarcity; a low one, of abundance. No one needs to know the details: the signal travels on its own. In arbitration, that signal does not exist. Information is atomised into silos, and without information that circulates there is no reputation to build nor scale to reach.

Third piece: there are no prices, and without prices there is no market

The third piece is the most abstract, but it is the decisive one.

Justice is what economics calls a private good: rival and excludable. That a judge devotes the morning to my case means they do not devote it to yours (it is rival), and one can perfectly well charge for the service and exclude whoever does not pay (it is excludable). Goods with these characteristics are allocated efficiently through a mechanism we have spent centuries perfecting: the market and its prices.

But state justice has suppressed that mechanism. It is a monopoly without real prices. And without prices, something more profound than a simple inefficiency occurs: it becomes impossible to calculate. There is no way to know how many judges specialised in intellectual property are needed, nor where, nor how much it should cost to resolve one type of dispute versus another. A central planner, however well-intentioned and competent they may be, cannot gather that information, because the information does not exist until a market generates it. It is not a problem of better managers or more budget. It is an impossibility at the root —what economists call a problem of calculation.

Arbitration could have brought prices back to dispute resolution. It could have created a true market for justice, with specialisation guided by demand and competition among providers. But it has not, trapped by the paradox's other pieces: without information that circulates and without scale, prices never come to form.

Fourth piece: in the end, you always have to call the State

And here is the knot that tightens it all. Suppose the three previous barriers are overcome. Suppose a company chooses arbitration, obtains a fast and expert award, and wins. What happens if the other party simply does not pay?

Then the award, on its own, is not worth much. To collect, one has to go to court to request its enforcement. One has to ask the State —the same slow jurisdiction to which arbitration meant to offer an alternative— to seize assets, to enforce, to use its force.

Arbitration resolves the conflict, but it cannot make its own decision be complied with. It depends on the state monopoly on coercion for the final step, the one that truly matters. And that final step reintroduces all the slowness, all the costs and all the uncertainty that had been avoided.

There is a historical irony here worth pointing out. The great treaty that universalised arbitration —the 1958 New York Convention, which requires more than 170 countries to recognise and enforce foreign awards— was a monumental success. But by building the recognition of arbitration upon state judicial enforcement, it enshrined its dependence. It turned that dependence into the very architecture of the system. Arbitration became global hand in hand with the State, and with that it was left chained to it.

The paradox, whole

Put it all together and the picture explains itself.

An epistemologically superior mechanism —faster, more expert, more flexible— remains confined to a niche because: litigating is deceptively cheap, which discourages change; information does not circulate, which prevents reputation and scale; there are no prices, which makes it impossible for a true market to form; and enforcement remains in the hands of the State, which nullifies at the final step the advantages gained in the earlier ones.

Each piece reinforces the others. It is a stable equilibrium. And that is why —this is the uncomfortable conclusion— there is no procedural reform that breaks it. One can speed up a procedure, lower a fee, train more arbitrators. None of that touches the architecture of incentives that sustains the paradox. It is like pruning a tree's branches expecting to change its root.

So what now?

If the problem is not procedural but architectural, the solution cannot be procedural either. It has to be a different architecture: one that makes information circulate without breaking confidentiality, that builds reputation out of conduct, that forms prices where today there are none and —above all— that resolves enforcement without having to call the State in the final step.

Is that possible? For centuries, the answer was no. But there were communities of merchants —from the merchants of the medieval Mediterranean to the contemporary diamond industry— that built private systems of justice that worked without leaning on state force, sustained only by reputation and information. What they lacked was the technology to do it at scale.

That technology exists today. And it is on that idea —how an architecture of incentives, and not one more reform, can finally dissolve the Arbitration Paradox— that we work at IMPERA.

In the coming entries I will unpack how.