Computational Legal Futures is a tri-monthly series exploring the promise of computational law: digital transformation and extended intelligence in the law. The series is authored by Sandy Pentland, MIT Toshiba Professor and Director of MIT Connection Science, and Robert Mahari, JD-Ph.D. Student at Harvard Law School and MIT. This piece was co-authored by Tobin South, Ph.D. Student at the MIT Media Lab.
From the Palaeolithic era until the mid-1800s, tally sticks represented an international system for recording debts. Debtors and creditors would break a piece of wood or bone in half, and the unique fracture memorialized the debt. Of course, the record of these legal obligations was only as good as the social contract that guaranteed their enforcement. The era of tally sticks ended in a blaze when the British Parliament decided to abandon the ancient form of record keeping, burning the remaining tally sticks under the House of Lords and accidentally setting light to the Palace of Westminster…
Legal obligations are worthless without a judicial system to enforce them. Without a way to arbitrate disputes, creditors were left with nothing but a jagged piece of wood. The world of cryptocurrencies faces a similar issue today. Although cryptography can be used to create immutable records of legal obligations, it cannot always enforce them. Some transactions, especially those involving offline world assets, require an additional mechanism to enforce obligations and adjudicate disputes.
Proponents of crypto paint an exciting vision of a trustless and decentralized world in which transaction risks can be minimized through immutable agreements. In this piece, we will begin to explore what it would take to make these visions a reality: how can legal obligations be enforced independently of contemporary legal systems to provide justice in a vacuum?
1. The limitations of decentralized agreements
Cryptocurrency ecosystems provide cryptographic guarantees of exchanges in decentralized finance, token swaps, and other decentralized smart-contract mediated behavior. However, any interaction with the physical world, like buying real property, represents a failure point for the ‘trustless’ system the crypto-world relies on. Smart contracts are touted as trustless solutions to this problem, but they, too, cannot create a reliable link between the physical and virtual worlds. This fundamental limitation becomes painfully apparent when exchanges mishandle user funds (FTX), or stable-coin providers fail to properly manage asset reserves (Luna-Terra), leaving investors with little more than a splintered piece of wood.
The key problem is exemplified well with the example of real estate transactions. If we enter into an agreement to purchase a physical asset – like a plot of land – in exchange for cryptocurrency, there is no native and universal legal mechanism to ensure that the transaction is consummated. We can rely on existing jurisdictions to enforce the transaction, but this limits us to jurisdictions with a (strong) rule of law. Moreover, jurisdictions with a reliable rule of law may not be well-versed in how crypto works.
Escrows represent one solution to this issue: upon entering into a contract, the seller of the land sends money to a third party which is released once the buyer takes possession. However, the escrow solution leaves a lot to be desired: it does not scale well, ties up capital (potentially for long periods), and still requires a trusted third party (or trusted blockchain oracle). Although smart contracts can mediate escrow agreements, they themselves cannot reliably introduce real-world information to the crypto space.
The promise of crypto is that it provides a medium of exchange that anyone can use anywhere to buy anything. If we are to see reliable transactions independent of existing jurisdictions, for example in places with high corruption, international waters, or even deep space, then rule-of-law must be maintainable wholly within the crypto world.
2. The state monopoly on violence and justice
Laws work because states have a monopoly on the use of violence. If a party to an agreement is deemed to have violated the agreement, she may be ordered to pay the other party. If she refuses, she may be held in contempt, and the state may force her to give up her property or liberty. Justice in a vacuum is hard to achieve simply because tribunals without the backing of a state have no way to enforce their judgments.
Arbitration represents a relatively new legal innovation that side-steps the traditional need for state enforcement. Most countries have signed onto the Convention on the Recognition and Enforcement of Foreign Arbitral Awards (the New York Convention), promising to enforce each other’s arbitration decisions. However, arbitration merely outsources enforcement to existing state power; it does not create a new enforcement mechanism. Therefore, arbitration awards are virtually impossible to enforce against a party with insufficient contact with a member state. A truly decentralized form of justice would enable enforcement against parties no matter where they are.
3. Dispute resolution through networks
To illustrate the monumental challenge of decentralized legal enforcement, let us take a closer look at one way such a system could be operationalized.
Alice lives in an unincorporated territory where she operates a mine. She enters into a 10-year contract with Bob to supply metal every month at a fixed price. Several years into the partnership, the price of metal skyrockets, and Alice insists on increasing the contract price. There are no courts or police forces in the unincorporated territory, and Alice does not stop doing all her business virtually. How can Bob enforce the original agreement, or recoup his losses?
Imagine a decentralized tribunal system that can be used to adjudicate these types of disputes. As part of each agreement, transaction parties select a tribunal that will rule on any future disputes between them. A plurality of tribunals exists, with self-defined rules and procedures. When transaction parties enter into a new agreement, they select a tribunal, agree to resolve disputes arising under the agreement through the tribunal, and undertake to enforce judgments issued by the tribunal. If a dispute arises under their agreement, the parties approach the relevant tribunal, and the tribunal issues a monetary judgment. The losing party can pay the judgment amount, ending the dispute.
If the losing party refuses to pay, the tribunal system can enforce the judgment by leveraging the community of members who have agreed to enforce the tribunal’s rulings. Each member in a given tribunal network will impose a penalty on future transactions with the losing party. These penalties are used to compensate the winning party until the judgment has been satisfied. If a rogue member of the tribunal network refuses to levy the penalty, the remaining members will levy a penalty against this rogue member.
FIGURE 1 – A figure illustrating how members participate in different tribunals to create enforcement power that leverages relationship networks (black lines).
A network-based approach to justice in a vacuum hinges on at least three prerequisites: identity, reputation, and mass adoption.
If individuals can easily change their identity in the network, it becomes effectively impossible for network-based penalties to be applied. A central authority can provide identity, but this would fly in the face of the core principles – decentralization, trustlessness, pluralism – that define the crypto space.
One approach to maintaining purely network-based identity is to make identity changes costly. This could be accomplished by locking assets (making it hard for Alice to transfer her mine) or attaching attestations to identities that maintain reputation (such as those suggested by soulbound tokens in a Decentralized Society). For large institutions with relationships with many business partners, changing identity is costly. However, smaller and less influential individuals can more easily assume a new identity. We expect that in a decentralized tribunal system, individuals’ perceived cost of changing their identities (proxied by connection to assets and reputation) would be an important signal to potential transaction partners.
Currently, there are numerous initiatives working on universal digital identity. In the cryptocurrency space, Worldcoin is exploring the use of biometrics as the basis for a universal identity. Of course, digital identity is not limited to crypto. The ID2020 Alliance has defined technical requirements for user-managed, privacy-protecting, and interoperable digital identity. Meanwhile, the Open Wallet Foundation is developing the open-source infrastructure needed to store and verify different identity solutions.
State violence is inherently universal: no one wishes to have their liberties curtailed. By contrast, the reputation mechanisms that underpin the system we outline are inherently weaker. Reputation builds on top of identity. Without a permanent form of identity, reputation mechanisms are pointless because low-reputation individuals can always opt to ‘start over’. Reputation only matters insofar as the network ecosystem is relevant in everyday life. As long as the crypto space is relatively niche, it will be easy for bad actors to shirk virtual legal obligations and drop out of the decentralized world entirely.
Identity is required for a reputation to be permanent. Decentralized tribunals use reputation to operationalize penalties. But only mass-adoption of decentralized networks gives these penalties any weight. As long as dropping out of the crypto space is relatively costless, bad actors can only be held accountable by traditional state courts – this is playing out today in the FTX case. Achieving pluralistic decentralized justice necessarily implies a world in which cryptocurrency is used broadly, giving rise to a significant ‘cold-start’ problem.
Formalizing distributed dispute resolution
Even if the described prerequisites were available, a distributed dispute resolution system would still face several challenges. To explore these further, we will take a slightly closer look at the incentive mechanisms underpinning decentralized justice.
The ability of a tribunal to enforce its judgments is a function of the number of members in the tribunal and the strength of the relationships (naively, the number of transactions) between them. This gives rise to a recursive incentive mechanism described by the equation below:
If a member of a tribunal, X, is penalized by the tribunal and refuses to pay, then X’s future transaction partners who are part of the tribunal (X’s neighborhood) enforce a penalty on X during future transactions. The more transactions X has with her neighbors, the larger the total penalty that can be enforced. If any member, Y, in X’s neighborhood, refuses to levy the penalty, Y will similarly be penalized by her neighborhood, and so on. The recursive mechanism ensures that any member that is linked to X has a financial incentive to enforce the tribunal’s verdict.
A key element of this formulation is the choice of decay function. In a case where no decay is applied, the penalty on anyone who transacts with rogue actors is the same as the original penalty. This approach creates strong incentives to enforce the tribunal’s verdict, but it also leads to a rapid spread of punishment through the network, chilling economic activity. While decay functions are mathematically convenient (leveraging network results such as eigenvector centrality), they may not meet the preferences of a community. Ultimately, the choice of decay function would be made by tribunals, and different tribunals would compete based on how exactly they operationalize the recursive incentives.
5. Three challenges
Creating an alternative to the state monopoly on violence through network-based incentives is clearly a bold ambition. We will outline three significant challenges.
Just as a single dominant state power can have detrimental effects on world order, a single dominant tribunal can undermine the order of the decentralized ecosystem. A single tribunal could enforce arbitrary or self-serving judgments. Meanwhile, a diversity of tribunals can serve as a laboratory for justice and cater to differences among market participants. For example, some entities might value rapid and unintrusive dispute resolution, while others prefer more thorough adjudication.
Similarly to nation-states, there are increasing returns to power for tribunals. A large tribunal has stronger enforcement powers and will thus tend to attract more members, in turn increasing its enforcement power. This growth benefits all members of the large tribunal, while weakening all other tribunals. Importantly, the increasing returns to power create a lock-in effect that makes it challenging for members to leave the tribunal. Even if it is democratically governed, a single tribunal cannot represent market participants’ conflicting preferences — this requires a plurality. Preventing this failure mode is analogous to solving antitrust, and structural design choices may help avoid the emergence of disproportionately dominant tribunals.
Dominant actors (too big for justice)
While the dominance of tribunals leads to disproportionate enforcement power that squashes diversity, large actors within a network present an opposing threat. In a network where a single actor has outsized wealth and influence, it can effectively threaten to impose a cost on anyone who levies a penalty against it. Due to the recursive incentives, an individual will choose not to impose a penalty on a large actor if the consequence of this action is greater than the cost of incurring the recursive penalty. In an ecosystem of tribunals, anyone is free to create a new tribunal. Although this creates a diverse pluralistic landscape, it also means that a single dominant actor can effectively create a thiefdom.
While a dominant agent in a system may be too big for justice, there is a large reputational cost associated with threatening an entire community and openly flaunting communal norms. As a result, it is likely that a whale would only consider this course of action in exceptional circumstances or if it is possible to drop out of the tribunal system entirely.
Cold start problem
The challenges considered above assume a steady-state tribunal system, but the genesis of justice in a vacuum represents a challenge in its own right. As with other types of networks, the utility of a tribunal system is tied to the number of users, and there is little incentive for individuals to join initially. It is conceivable that a tribunal system would grow out of an existing dispute resolution organization or arbitration body or that it is the result of a digital community embracing network justice systems for self-governance.
Even once a tribunal system exists, new tribunals face an analogous cold start problem. Without a large network of members, a new tribunal has low enforcement power and will fail to attract new members. Partnerships between tribunals could address the cold start issue by allowing tribunals to ‘lend’ membership networks to one another for enforcement while maintaining separate dispute resolution mechanisms; this type of collaboration is the decentralized equivalent of the New York Convention.
Popular pieces like The Network State and Decentralized Society paint a vision of a distributed pluralistic technological future enabled by a truly decentralized and universal means of payment. If cryptocurrencies are to become a universal medium of exchange, then their reach must extend beyond the online realm. This requires a method of enforcing agreements and resolving disputes that are as universal as the currencies themselves. Leveraging existing courts and arbitration tribunals provides a limited solution, as these existing bodies cannot enforce verdicts against individuals outside of their jurisdictions.
We explore the feasibility of a truly decentralized form of dispute resolution that relies entirely on network effects to enforce judgments. In doing so, we surface a number of onerous requirements – like the need for identity and reputation mechanisms – and consider several challenges to decentralized dispute resolution. Though a fully decentralized dispute resolution system does not currently appear feasible, elements of this network-based justice system may be applied today, especially in contexts where the rule of law is weak.
|Citation: Sandy Pentland, Robert Mahari and Tobin South, Justice in a Vacuum?, Network Law Review, Winter 2023.|