04 Mar Primavera De Filippi

Primavera De Filippi is a permanent researcher at the National Center for Scientific Research (CNRS) in Paris and a faculty associate at the Berkman Klein Center for Internet & Society at Harvard University. She is a former member of the Global Future Council on Blockchain Technologies at the World Economic Forum and co-founder of the Internet Governance Forum’s dynamic coalitions on Blockchain Technology (COALA). Her fields of interest include legal challenges raised by decentralized technologies (with a focus on blockchain), their potential to create new governance models and participatory decision-making, and the concept of governance-by-design.

In this interview, we discuss the biggest challenges facing blockchain governance, the seminal “law is code” concept, and some of the regulatory hurdles slowing mass adoption of blockchain tech.

Conducted December 16, 2019 by Daniel Kees

How did you get interested in blockchain governance?

As a legal scholar, I was looking mostly at the legal questions and regulatory challenges raised by blockchain technology. Because of its decentralized and transnational nature, blockchain technology somehow introduces similar challenges as the Internet, whose regulation cannot be achieved without taking into account Internet governance. 

With blockchain technology it’s a similar problem. The regulation of a blockchain cannot be done independently of its current governance structure. First, we need to identify who are the actors with the most influence and power, as they are those whom regulation should focus on. At the same time, however, we need to realize that regulating specific actors involved in the maintenance and operation of a blockchain is not tantamount to ensuring that the whole blockchain network will operate as prescribed by the law.

The field of Internet governance studies emerged as an attempt to address these regulatory and governance challenges of a global and decentralized communication network. As such, it may provide today relevant insights on how to address the regulatory and governance challenges of blockchain technologies.

How would you explain blockchain governance to someone who isn’t familiar with that area?

First, it is important to distinguish between “on-chain” and “off-chain” governance. On-chain governance means governance “by” the infrastructure. That is the codification of specific governance rules directly within the technological infrastructure of a blockchain: how we design the protocol, the consensus algorithm, and so forth. At the same time, on-chain governance necessarily relies on off-chain governance, which is governance “of” the infrastructure. That is the more social and institutional layer where decisions are made about what rules are going to be codified into the blockchain protocol. On-chain governance is more about enforcement and execution of rules, whereas off-chain is more about deliberation, decision-making; the latter is essential to establish when particular on-chain rules need to be changed, either because things have evolved or because things are not working as expected.

Today, most of the focus is on-chain governance, which mostly relates to game theory and mechanism design. But there also needs to be more focus on the social aspect of governance, which consists of all the rules and procedures to decide on the design of the blockchain infrastructure, and as such directly affect the operation of that infrastructure.

By that “social” element, are you referring to the biases and assumptions that developers bring to the table when they’re writing code? Or, is it broader than that?

Not quite. When I talk about the social layer of governance I refer to the various actors involved in the development, maintenance, and more generally in the operation and use of blockchain networks.

The developers are developing blockchain solutions; but we also need to account for the miners and validators, who are adopting these solutions, with the task of choosing one version over another. Then, there are specific actors like blockchain explorers, cryptocurrency exchanges, and all those commercial players that do not actually have any specific power on the network (in that they are just one node among many), but because of their commercial power, if they make a particular decision, that decision is likely to be followed by other people as well. And then, finally, there are the charismatic leaders, who have a particular influence over the community, along with all these people who are just being very loud on social media. 

It’s actually a very complex, polycentric mechanism of governance, and it’s really hard to know who actually has power and influence over which decisions. 

Related to that, what are some of the biggest challenges to a well-governed blockchain ecosystem? What types of challenges do widely adopted ecosystems like Bitcoin still have to figure out?

At the moment, the main problem is that governance is not transparent. There are a lot of works describing the governance of specific blockchain systems—at least from a theoretical level—but there is very little information about how governance actually works in practice. Oftentimes, the theory refers to on-chain governance, but to the extent that off-chain governance operates according to different principles, the theory of off-chain governance does not have real practical implications.

The big problem today is understanding what is the practical governance of blockchain communities and who are the people who actually control and influence decisions—whether those are token holders, miners and mining pools, or just a few people with strong influence in these communities.

This is related to the “tyranny of structurelessness”—that if you don’t have a formalized governance structure, it’s really hard to hold people accountable for their decisions because you don’t know who is actually making these decisions. It’s difficult to improve a governance structure if you don’t have proper information about how it actually operates.

So it goes beyond how the code is run, to the very transparency of the system?

Yes.

In thinking about those big challenges, how do you see those governance challenges being resolved, or is that an issue to be addressed once blockchain becomes more widely adopted?

I don’t have the magic solution, if that’s what you’re asking. [laughter]

In that case, given the wide variety of these governance issues, what is a good way to judge the effectiveness of a governance mechanism?

Again, it depends on what you mean by “effective.” Do you mean whether a decision gets made as opposed to not getting made? Do you mean that there is proper representation of the various stakeholders? Do you mean that there is legitimacy in the process? 

It does seem very broad. 

You cannot adopt an engineering approach in trying to design governance because it’s not a “Does it work/not work?” approach. You can have governance systems that are completely legitimate but that don’t work, or working governance systems that are considered as illegitimate. So, it really depends on what is the objective you want to achieve, and different objectives will require different set of compromises. It’s ultimately a policy question.

I wanted to talk about your scholarship, specifically a paper that you wrote in which you discussed a shift from the traditional notion that “code is law” to “law is code” in the blockchain space. Can you elaborate on what that means?

That’s basically the idea of blockchain as a regulation technology. With the Internet, people understood that it is much easier to dictate and organize behaviors by means of technological constraints, as opposed to traditional legal constraints. And so, “code is law” means that code is slowly assuming the same function as the law. However, because of its characteristics, code has very different properties and therefore operates in a very different manner than the law. Most importantly, the process by which technical rules are designed and implemented is very different from the more democratic process by which laws are elaborated and enforced. 

The shift from “code is law” to “law is code” occurs when the idea is no longer just to use technology as a means to dictate behavior but rather to use technology as a tool to facilitate or support regulatory compliance. Today, there is a lot of complexity and costs involved in regulatory compliance—in terms of auditing and formalities—mostly to make sure that those actors that we do not trust actually act in a constructive manner. With blockchain technology, insofar as we can create technological guarantees to ensure that specific operations are implemented as prescribed by the law, we might reduce the level of formalities, audits, etc. necessary to guarantee that no one is abusing their power.

On the other hand, some issues might emerge to the extent that codifying the law into a technical system means we are also changing the nature of these legal provisions because we are moving away from the ambiguity and flexibility of natural language, which can only be enforced ex post, to a formal and strict code-based language which is enforced ex ante. So enforcement of the law is more effective, but it is not exactly the same law that is being applied.

Relatedly, you’ve written extensively on the potential regulatory implications of blockchain in various contexts. Broadly speaking, are you hopeful for a global regulatory framework of some sort, or will blockchain continue to advance as this patchwork of national jurisdictions? 

I think as we start understanding the various applications of blockchain technology, some kind of regulatory harmonization will naturally emerge. The problem right now is that different jurisdictions are coming up with different solutions to address emerging applications of blockchain technology. 

Right now regulation is done in a prospective manner. There needs to be more practical use cases to understand what could go wrong and what exactly do we need to regulate. Right now, there is little empirical data that we could rely upon to assess the actual risks and benefits of the technology. I would assume that, once we get more of that empirical data, regulation will naturally evolve in a more convergent manner.

So guiding principles may emerge as people gain more experience with the technology.

That’s my hope.