25 Mar Guido Molinari
Guido Molinari is a Managing Partner at Prysm Group with a focus on blockchain governance. He has led Prysm Group’s consulting, education and research work with corporates, academic institutions, investment funds, and government entities including Microsoft, IBM, MIT Sloan, UC Berkeley, Polychain Capital, Hashed, DARPA, and the World Economic Forum. He is a former investment banker at Merrill Lynch and has been a visiting lecturer in entrepreneurship at Columbia Business School and the London School of Economics, and an Imperial College Master in Innovation business coach.
In this interview, we discuss hot topics in governance including Proof-of-Stake vs. Proof-of-Work, evaluating the effectiveness of different governance models and “the Governance Wheel”.
Conducted February 12th, 2020, by Shayaan Khanna
Which projects does your team typically spend most of their time working on?
We generally spend most of our time working on enterprise consortium projects. These are a group of large enterprises coming together to deploy blockchain technology. The underlying motivation is because the technology makes them willing to share data in the way that before they could not before (from a technology perspective) but they were being unwilling to do so. So we work with this group of companies thinking about questions around incentives around tokens and around governance, which is the topic that we’re here today to discuss.
What got you interested in blockchain governance in the first place?
We got interested because it was a natural continuation of our work that focuses a lot on the economics and incentives of these networks: When we think about enterprise consortia we think deeply about its economics. The rules of the system inform what you get if you join the system, such as the resource allocation required to run a node and the transaction fees. The rules are the economics. The governance is the rules to change the rules. It’s essential to think about governance as a key component of the overall design of the network and it’s so interconnected with the economics.
That makes a lot of sense. Just to give us an example, how would you go about explaining this to someone who’s not clearly an expert in blockchain?
I think the easiest way to think about it is to imagine that somebody comes to you and says, “Would you like to join the market/network I created?” When you look at what they are proposing today it seems like an attractive proposition. The value you receive from joining is higher than the cost of adopting the technology. But before you join the network, you ask yourself what benefits you receive today.
But few people ask: “What am I going to get tomorrow and the day after tomorrow.” Such questioning demonstrates you are thinking about how the network will evolve over time when you don’t have the sole decision to make changes. And then the next logical question is “What is my voice in those decisions?”. And that’s really the key governance piece! As a stakeholder in the network, I want to make sure that my views have reflected in the way the network evolves over time because I don’t want to end up in a situation where I like what I see when I joined and made a financial commitment, and then the rules change against me. Now I am being stuck in something where I actually don’t get the same level of benefit that I would have thought about when I first joined.
This is why we think about governance in terms of how decisions around change are reflected as changes happen in a natural progression of time. We know that in any technology changes will happen – no system remains the same over time. And so the governance system will help to align the interests of different stakeholders to drive the long-term development of the network itself.
That rings true, especially right now with the whole debate on proof-of-work proof-of-stake in the limelight with Ethereum. I think this is entering the public consciousness now for public blockchains – would you mind telling us what you see on the enterprise side?
On the enterprise side, generally speaking, the consensus mechanism is more defined: Each player or each enterprise will run a node. But some key questions still remain. A lot of enterprise networks are being launched and they ask themselves “Do we give the same weight to every member whether they join as founding members versus if they joined later?” They may want to reward people for joining early. If you can join as a founding member you will have a certain weight in the decision-making process and a certain stake. On the other hand, if you made the decision to join later that will be lower since you did not incur the cost when the project had a higher risk.
This is a big question that, depending on where you sit at the table, you might want to answer in different ways. If you want to encourage people to be among the founding members, then you want to give more weight to their voice in the governance of the system.
But instead, if you want an effort where you are an early adopter but need other adopters, leading to much broader adoption within the industry then you might want to actually say “Well, even if you join later we’re just going to come to the same because we really do need you. Adding you will increase the value to even the founding members through network effects”. We see a lot of these questions being tossed around in our Boardrooms right now as a lot of these networks are going towards “launch” and ultimately these decisions are going to impact whether these projects succeed or not in the long term.
So you touched upon the challenges around when you join and who joins as well. Are there any other challenges that you see for a well-governed blockchain system?
To the point I was previously making – a good governance system will ultimately face the challenge that as it progresses towards the goal the stakeholders involved will change over time. And the more players enter the preferences among these players are going to be different from where you were when we first started maybe with a few entities. And so, what we talked about the governance system, is it a living and breathing animal right it’s not something static – the “rules to change the rules” will also need to change over time. We don’t think there are many systems that are going to be with the same exact rules in, say, five years as they are now. We have an internal framework called the Governance Wheel. It’s a Wheel because once you get to the end, you have to go through it again – continuously thinking about whether the state of your current governance system reflects the preferences of the current set of stakeholders. As your stakeholder base changes you think about what changes need to be made to the system to align the preferences of the current stakeholders. It’s been a continuous process and it is pretty challenging given that these are networks that eventually will have a lot of very high-value transactions on them. These are not trivial decisions!
Going back to the example of the wheel. So, what would be the best way to address these challenges that stem from the notion that governance is a moving target. What examples of that could you share with us?
We not only look at the statistics using our framework but also various different networks for the last few years. The first step is always thinking about how governance is the “rules to change rules”, what the rules are, and how they will change over time. For example, let’s say live in networks without transaction fees. Ultimately, the node validators will need to be paid and will most likely determine how the networks potentially change over time. We then ask what the decision process is around making such changes. Let’s say we want to put together a potential change. First, a party needs to conduct formal research around the proposed alternative and take a stance around the potential change. Then you need to go through a vetting process for those potential changes.
In permissionless networks, you tend to have a situation where anybody in the community can propose a change to a system. Permissions to change enterprise networks is through more formalized ways to propose changes to the system. You might see an investing committee or a technical committee.
Under the first method it’s determined whether this change is viable, and then eventually gets put to a vote by the stakeholders, and there is where, to my earlier point, you decide the weighting each stakeholder has in the voting process. And then after that, to close the wheel, you have to think about the implementation of the change. We often have seen these alignments where something gets voted in by a certain group of stakeholders, but then ultimately the people who have to make the change say “Well actually we don’t want this to be implemented”.
And if let’s say, as an example, this is an event that involves node operators and those operators, even if they’re a minority, they vote against it. You cannot really force their hand. It’s going to be very difficult so you have to think about whether the governance system is aligning all the stakeholders and even minority interests in a way that again changes can happen over time.
And again, if we think about permissionless systems we’ve seen varying degrees of success of governance. In the case of Bitcoin probably not so successful. A lot of proposals for change that can happen: A lot of the hard forks which happened were because proposals people wanted to change died off. Other networks like Ethereum have big changes happening. There’s been a question of how centralized the governance is. I think permitted enterprise is a lot clearer because often all such items are codified in the documentation. This is all pretty clear because enterprises are used to writing down the rules right this is not an informal type of governance that could break down easily.
If we think about how the markets evolving so quickly, both from a market perspective and also from a regulatory perspective. Is there any kind of wish list that you have 2020, looking at the year ahead? Is there anything you would like clarified or confirmed from a regulatory or market perspective?
I think there is a need for clarifying the language that we use around this question of governance – often you find that people use the same terms but mean different things, or they use different terms to mean the same thing. The result is that this is a very distributed industry with a lot of the different issues without, a really centralized place where people can agree to a common terminology that will be used to define certain things. And so even the word ‘governance’, itself, is often used with various different meanings.
And we’re often, in a situation where we’re talking about governance, and we’re actually addressing the rules to change the rules. And some people think about governance, as the operational rules so what we will call the ‘economics’ of the system. Hence we have sort of this disconnect just because we’re talking about two different things using the same term. My wish would be that there is greater clarity among industry players, about what we mean by the governance of the blockchain and autonomy. I believe that there are long term benefits to having greater clarity and more consensus among industry players around our terminologies.
I think it’s really fascinating as well that could because you have so many different fields converging – it’s philosophy, it’s economics, it’s legal when you think about governance, it’s also the tech. All these things are being enabled by so many different discourses nowadays, as you said, that the tech is ready to drive some of that innovation right now. There is that scope for the industry to hopefully harmonize what the words themselves mean when even the words ‘decentralized’ and ‘distributed’ mean different things to different people.
Last point. How can we judge the effectiveness of governance methods?
So the way we currently evaluate governance goes back to the very first point. And the way you can ensure the governance system that you have in the network reflects the preferences of the current stakeholders is whether the stakeholders feeling that they are represented in the decisions that are being made to evolve the system over time. And you can easily measure that by seeing whether they’re continuing to contribute actively or whether they are leaving the network entirely.
From our standpoint, whenever there’s a hard fork in a permissionless network and we believe that in the failure of the governance system: Clearly the governance system was not able to reflect the preferences of those people and those people had to leave. Now in a permitted system that we’ve seen in many enterprise networks that corporations that have initially joined, they might later leave because they don’t see their preferences are no longer represented. And so the question is whether you keep providing the same level of incentive to the stakeholders compared to when they initially joined. Is the value of the network still similar?
And often, of course, when you have certain changes there’s contention. You need to ensure that you have a governance system that is able to reflect the preferences and still keep the players that might vote against the production to get implemented inside the system, and this is not easy. And again, this is the challenge not only for blockchain initiatives: If you think about the state of the US politics we see similar systemic challenges because a certain party wins over the other and you want to make sure they are somehow represented and the opposition doesn’t want to leave. Leaving the system could possibly look like a secession or a hard fork. Such an event is counted as a loss for the existing governance system.
You get decay in the network effect and overall development goes down significantly. So that’s what we focus on when we do work with networks is really thinking about the types of questions. And we are seeing among corporates, a lot of interest in these topics as they go more towards the implementation of their system.