Professor Barry Weingast is one of the world leading scholars in institutional economics, a field that studies how human institutions affect economic behavior. Trained as a political scientist and an economist, he is a professor at Stanford University and a Senior Fellow at the Hoover Institution, a public policy think tank promoting the ideas of individual, economic and political freedom.
Throughout his career, Weingast has conducted extensive research in the application of rational choice theory to legal, legislative and constitutional institutions. His work was very influential in my thinking about decentralized justice, in particular, the seminal article titled “The Role of Institutions in the Revival of Trade: the Law Merchant, Private Judges and the Champagne Fairs” he coauthored with Nobel Memorial Prize laureate, Douglass C. North, and Paul Milgrom, an economist and game theorist at Stanford University.
This paper analyzes the work of medieval merchant guilds and the ways they provided security by developing trade relations and securing merchants’ property rights. This article presents one of the deep roots of Kleros and the future of decentralized marketplaces, so we sat down one day to discuss how these two concepts intertwine.
Tell us a little bit about your background and how you became interested in the type of topics you are working on?
I started out studying the politics and economics of regulation in the United States. I moved on to broader topics at the intersection of economics, law, politics, and history as I became interested in the age-old question about why a few countries are rich, while so many others are poor. What makes developed countries so different from developing countries and on so many dimensions?
We in the developed world take many things for granted, that we don’t even think about. Whereas a look at medieval societies reveals the absence of various features central to the developed world, such as property rights, the ability to alienate land, and especially an absence of the everyday threat of violence and predation. Studying political, economic, and legal history allows us to see the pieces that are missing from the current way that people think about development.
You wrote this great paper on how the Champagne Fairs enabled a revival of trade in the Middle Ages by providing a system of private law…
We were interested in the question about the rise of trade. In particular, why was trade so slow getting restarted after the fall of the Roman Empire? At the time we wrote, historians argued that an important role was played by the Law Merchant.1 I think we wanted to have a sense of how trade could work in the absence of territorial-nation-states with extensive territories.
There were governments in the Middle Ages, but they were ineffective, what we today call "low state capacity". The king was up against the great lords, sometimes called "the barons" in English history, and the barons were often almost as powerful as the king; and sometimes even more powerful than the king. So, the king could not impose his law on the land, thus property rights were subject to constraints and mischief by the great barons.
The set of institutions which arose back then had elements of private governance where a private group of merchants in different countries came to agreements on how to set the rules.
How did that system of private law work? Who created the rules? Who were the judges? How was enforcement done?
All long-lived constitutions have to be self-enforcing in the sense that those that live under them have to have incentives to abide by the rules. Part of what emerged in the early Middle Ages were mechanisms by which these traders devised to enforce long-distance trade where the opportunities for cheating, fraud, and malfeasance were legion.
In medieval northern Europe, trading cities were necessarily mindful of foreign merchants who sought to defraud local merchants and depart, never to be seen again. The home merchant might get a judgment in the home courts, but what good would that do if the defrauding merchant never returned?
Medieval trading cities developed a complex system that mitigated the problem of fraud/absconding in long-distance trade (as Grief 2005, Ch. 9, explains). These cities carefully controlled which foreigners were allowed to trade. Cities did not grant individuals rights to trade. An individual merchant from Cologne, Germany, could not show up in, say, London, England, and trade his goods.
Instead, England gave the rights to trade, not to an individual, but to an organization: a merchant guild comprised of the merchants from a given city. The guild was jointly liable for the behavior of their members – if one of them were to cheat and disappear, the other guild members were responsible for making good on his contracts, debts, and any judgments against the merchant who absconded after defrauding a local merchant.
This system of joint liability, in turn, gave the guilds the incentives to police who among them got to trade in their name. To trade, a merchant had to be part of a merchant guild and not everyone could do that. The guild closely controlled its membership. To become a member, a merchant had to have local property that could be confiscated; and typically the merchant’s family would remain in the city, so there were consequences for that, as well.
This system is an example of self-organization, sometimes with the help of governments. In England, for example, the king was involved with the trading community. But the central idea is that merchants did not rely on the king and the state for enforcement. They developed a complex set of institutions that created incentives to trade honestly.
In principle, this system could be designed by and operated by merchants. Or it could be built in cooperation with local governments. The first approach is a purely private order, while the second involves one or two governments. What’s the difference? As long as the system operated the same way, the result is the same public good.
What we would call these days a reputation system...
Yes, it was a reputation system, but one in which institutions were necessary to make it work. The nature of interaction among merchants had to be rigged in a way that mitigated the problem of enforcement in a complex environment where cheating is not observed.
Let me explain. In the 1980s, many political-economists began studying repeated games, especially the repeated prisoners’ dilemma. This model attracted scholars because it suggested a way in which cooperation could be sustained even if the players had short-term incentives to cheat one-another.
In the single play prisoners' dilemma, both players had incentives to cheat rather than cooperate, leading to the failure of cooperation. But scholars discovered that when repeated infinitely, the prisoners' dilemma afforded cooperation. The reason is that, if player 1 cheated player 2 in this round, player 2 could retaliate against player 1 in the next round and forever thereafter.
Repetition combined with retaliation to force the would-be cheater to choose between: (i) cheating now and foregoing the benefits of cooperation forever, and (ii) refraining from cheating now and capturing the benefits of cooperation into the indefinite future.
The “folk theorem” holds that when people do not discount the future too heavily, the benefits of on-going cooperation exceed the short-term gains of cheating now but not cooperating thereafter. Defection today might be tempting, but if it leads to costly retaliation, the potential cheater will be deterred from cheating.
This illustration explains how the basic reputation mechanism works. But it is not the last word.
Consider a context of a merchant community whose members trade with one-another. If a given member of the community cheats, he gets a reputation as a cheater, and members of his community will avoid trading with him.
A fundamental but hidden assumption limits the applicability of the basic reputation model. The model assumes that behavior is common knowledge. Everyone knows who cheated, so they effortlessly focus their retaliation on the cheater. But what happens if members of the community do not costlessly observe who cheats?
Consider a merchant community whose members trade bilaterally. No one knows what goes on in an exchange between two merchants. Instead, members of the community typically observe a dispute in which both trading partners blame the other. Herein lies the problem: if members of a community observe disputes, not cheating, against whom are they to retaliate?
The problem is even worse because opportunist cheaters may engineer disputes that masquerade their subterfuge. The lack of knowledge about cheating disables the reputation mechanism.
This is where the Law Merchant model comes in, as a mechanism for preventing that kind of cheating by creating a centralized investigation, adjudicating the dispute, and then making a common knowledge announcement to the community about whether cheating had occurred.
In our paper, we show that, if the members of the community use this common knowledge signal for their retaliation decisions, they can sustain cooperation. But absent a mechanism like the Law Merchant, the community cannot sustain cooperation.
The Middle Ages taught us that both governments and private actors can play a role in providing a secure environment for trade and other social interactions. But, what’s the difference between them? How do we know when we have a government? Take Facebook or eBay, for example. They are huge digital platforms with users from across the globe. They write the law of the land (terms and conditions). If a user violates this law, he can be banned, similar to how ostracism worked in the Middle Ages. And this decision is made by a moderator, a kind of private judge appointed by the “platform king” to adjudicate on disputes which happen under this private system of law. Would you say that Facebook, eBay and other Internet platforms are states? Or is that going too far?
This is a deep question and I don’t think anyone has a definitive answer. Part of the difficulty arises because the answer depends on the way you define a state.
Generally scholars think of the state as many components, such as sovereignty, the type of state (authoritarian, republican), etc.
A defining characteristic of a state is sovereignty: No legal authority exists above a sovereign state that can order the state to do something or prevent the state from doing something. If we accept this premise that a state is a sovereign entity, then we get a syllogism that answers your question:
-Every state is a sovereign entity.
-Facebook is not a sovereign entity.
-Therefore, Facebook is not a state.
Facebook is subject to the host country’s laws. It cannot set aside American commercial law or contract law or criminal law. Nonetheless, Facebook is a governance structure, to use Oliver Williamson’s terminology.
What can we learn from these institutions that can function without a government?
A central lesson is that we have to look at the incentives of the actors, the ways they participate, and what are they willing to do. Another piece is the perspective Gillian Hadfield and I have developed which we call the “what-is-law?” framework (Hadfield and Weingast 2012, 2014). When does a body of rules constitute law? A related question is: what characteristics must a set of rules have in order to generate third party enforcement of the rules in a given community?
A range of 20th century legal philosophers have thought deeply about this question. One of the things that emerges from literature dating from the ancient Greeks are characteristics of the rule of law. These include characteristics such as stability, feasibility, prospectivity, clarity, consistency, and impersonality. These characteristics are called “legal attributes”.
Using a simple game, Hadfield and I show that legal attributes emerge as part of an equilibrium of third party enforcement of the rules. The reason is that individual A has no incentive to help punish a violation of the rule against person B if the rule in question applies to B, but not A. In a system of privilege, where each member of the nobility has a different set of privileges and in which commoners have few, a community finds great difficulties in sustaining law.
The "what-is-law" framework has important consequences for projects like Kleros. Do the mechanisms underlying the products and services you're providing satisfy the conditions that people associate with the rule of law? I assume this is desirable. If so, how is Kleros attempting to implement and sustain the rule of law?
An example has to do with a characteristic we call "impersonality", which means that the law treats like people alike. All citizens have the same rights. Your social position does not determine your rights as a citizen. In the rule of law, it corresponds to the idea that like cases should be treated alike. Extraneous details about your personality, race, or circumstances should not enter into a decision judging you when you have been accused of wrongdoing.
This argument leads me to a question for you and Kleros – what kinds of thoughts you have about issues having to do with the rule of law?
Kleros seeks to provide rule of law for an environment which doesn’t have it yet. For example, in today’s global and digital economy, lots of disputes happen between agents across jurisdictional boundaries. Mostly small claims between freelancers and clients, etc. There is no court system able to solve a dispute for a couple hundred dollars between, say, a client in the United States and a contractor in India. It’s simply not effective.
Kleros is a system of private law which provides small claim arbitration. When parties agree on the contract, they also agree on the jurisdiction for resolution in case a dispute happens.
In the Middle Ages, merchants agreed to have their disputes solved in fairs that had a reliable court system, which had some rule of law institutions that they could trust. The Champagne fairs were very successful because they managed to develop a legal system that provided rule of law. Merchants trading there knew that it was likely that contracts would be enforced.
In some way, Kleros is a digital age equivalent of this private court system. Parties can agree on using Kleros (and particular courts within Kleros) knowing that the dispute will be solved in a fair manner, by a panel of jurors selected in a transparent way. After the panel makes a decision, this is automatically enforced. It provides rule of law for a global society where there is no rule of law.
I would agree with everything you say, except for one piece. It is not clear to me you provide a rule of law in the sense of what legal scholars think of as law. In part, because it is not clear what mechanisms lead your system to treat like cases alike.
That is not to say you provide no value, you clearly have a system that is fair and that is not obviously biased towards either party. It is also a system of third party enforcement, meaning nobody making the decision is directly involved with the people in the case.
In that case, you have some important pieces of the rule of law, but it is not clear to me you have all of them, especially issues like precedence. Precedence is one of the central ideas underlying Anglo-American legal systems. What, if any, is the role of precedence in your system?
Jurist Oliver Wendell Holmes famously said: “The prophecies of what the courts will do in fact, and nothing more pretentious, are what I mean by law”. Kleros also has this feature. After many cases are ruled in a specific court in the system (e.g., the website disputes court), we will start to observe a regularity on how cases that look alike are decided in a similar way. Then agents would know, even before going to arbitration, how a specific case is likely to be ruled by that court.
That is a very important piece to have, especially if you think about Kleros growing into a larger platform, one that moves beyond your initial focus. People are going to want to know how Kleros decides any given case, especially as you move into more complex interactions between businesses. This poses a challenge for any company trying to do these kinds of things.
I have been following TaoBao, the world’s largest e-commerce platform, owned by Alibaba. They are creating the legal system for their platform. A big difference between TaoBao and eBay or Facebook is that the latter have all of American law behind it.
Imagine you are an entrepreneur cyclist and you invent a new digital brake system, which you want to sell it on the Internet. In the United States, three out of four people will create their own independent website. In China, only 10% of people would do that. Most of the people use platforms like TaoBao.
The reason for this is that America has a rule of law system which underlies commerce and exchange, so there exists mechanisms to deal with problems like intellectual property rights. In China, this backdrop of American law does not exist.
Part of what TaoBao is doing as a platform is creating a legal system in which disputes are resolved wholly within the system. eBay has that too, but it is truncated and if you exhaust the remedies, you can go to court and challenge any ruling made by eBay’s system of adjudication. Whereas in China, no legal system and courts exist that can play this role.
To pick up on our previous discussion: is TaoBao a state, then? It would seem like a type of monarchy which makes the laws and enforce them, and where there is no other place to go if you don’t like a decision.
In the 17th century, Louis XIV could overrule any ruling from his courts he didn’t like. In modern democracies with rule of law, the executive power cannot overrule judicial decisions, because of the principle of the division of power.
But TaoBao’s management would seem in a position to overrule any of their courts decisions. And it all happens within a walled garden. In the case of Kleros, there is no "owner" who could overrule the rulings of a court. And also everyone knows how the system works: how jurors are selected, what are their incentives, how evidence is handled, etc. This is a big difference between a private law system like TaoBao’s and Kleros, which is also a private law system, but operating with a procedurally fair framework.
In the end, users have to trust that TaoBao will be honest. Most of the times, it will probably be honest. But they still hold the monopoly of a key that can change rulings. That is a substantial difference. In Kleros, nobody holds that key and anyone can verify how the system works.
That is good, because one other characteristic of the rule of law is that it is transparent, that people understand how it works. Nonetheless, your system, like that of Taobao, is not free from potential biases. Part of your reputation growth will be trust that Kleros has not rigged the system, as you just pointed out with respect to Taobao.
One of the biggest problems in the evolution of law has to do with the arrival of unforeseen circumstances, when existing laws are inadequate to deal with it. This is a major reason why Hadfield and I argue a system of law must have some sort a legal steward, an individual or group that helps resolve disputes over how to extend the law into new areas and novel circumstances.
Often, there are multiple ways to do it, and different parties have different beliefs and preferences on how the rules should be adjusted to cover the new circumstances. Does Kleros have a way of managing such circumstances?
Rules for each court are decided by users through a voting mechanism. If there is a change in how a specific industry decides to appraise evidence or make decisions, then this will be reflected into the rules of the specific Kleros court where these disputes are settled. If the court doesn’t adapt to the new ways of doing things in the industry, it is likely that it will be left behind and users will switch a to another court that is better adapted to the new law…
One final question - could you recommend some reading for our community?
These are works mentioned in this interview.
Avner Greif, Institutions and the Path to the Modern Economy. Cambridge: Cambridge University Press (2005).
Gillian Hadfield and Barry R. Weingast, "What is Law? A Coordination Model of the Characteristics of Legal Order" Journal of Legal Analysis. (2012); and idem.
“Microfoundations of the Rule of Law” Annual Review of Political Science (2014)
Emily Kadens, “The Myth of the Customary Law Merchant” Texas Law Review (2012).
Klerman, Daniel. (2007). “Jurisdictional Competition and the Evolution of the Common Law”. University of Chicago Law Review.
Paul R. Milgrom, Douglass C. North, and Barry R. Weingast. "The Role of Institutions in the Revival of Trade: The Medieval Law Merchant, Private Judges, and the Champagne Fairs". Economics and Politics (1990).
Lizhi Liu and Barry R. Weingast. “Law, Chinese Style: Taobao and the Private Building of Legal Market Infrastructure”, Working Paper, Hoover Institution, Stanford University, 2018.
Douglass C. North. Structure and Change in Economic History. Cambridge: Cambridge University Press. (1981)
All of my papers, including those mentioned here, are listed on my website.
Here's some further reading on this perspective of self-organizing institutions:
Bernstein, Lisa. 1992. “Opting out of the Legal System: Extralegal Contractual Relations in the Diamond Industry”. Journal of Legal Studies.
Alfred Chandler. Visible Hand: The Managerial Revolution in American Business. Cambridge, Mass: Harvard University Press. 1977.
Robert Ellickson. 1991. Order without Law: How Neighbors Settle Disputes. Cambridge: Harvard University Press.
Leeson, Peter T. 2009. The Invisible Hook: The Hidden Economics of Pirates. Princeton: Princeton University Press.
Munger, Michael C. 2018. Tomorrow 3.0: Transaction Costs and the Sharing Economy. Cambridge: Cambridge University Press.
Skarbek, David. 2014. The Social Order of the Underworld: How Prison Gangs Govern the American Penal System. Oxford: Oxford University Press.
Stringham, Edward Peter. 2015. Private Governance: Creating Order in Economic and Social Life. Oxford: Oxford University Press.
Weingast, Barry R. 1983. “The Industrial Organization of Congress: or Why Legislatures, Like Firms, Are Not Organized as Markets”. Journal of Political Economy.
1 In recent years, modern legal historians have challenged the older view of the Law Merchant, questioning whether the phenomenon existed (see especially Kadens 2012). Regardless of that debate, merchants needed various types of institutions to underpin long-distance trade. Moreover the Law Merchant model – apart from whether the law merchant existed in practice – applies to such trading institutions as the Champagne Fairs, where merchants came to trade via contracts to deliver wares over the course of the coming years.
Before you go... would you like to become a Kleros Juror?
You can always sign up to become a juror adding your email to a mailing list. Use of our juror platform is fully decentralized of course but this lets us get in touch to notify of any pilots or real Dapp's ready for you to arbitrate on.
Join the community chat on Telegram.
Visit our website.
Follow us on Twitter.
Join our Slack for developer conversations.
Contribute on Github.