Cathy Barrera, a CoinDesk columnist, is a founding economist at Prysm Group, an economic advisory group, and was chief economist at ZipRecruiter. She has a PhD in business economics from Harvard.
In January, when Facebook announced the board's by-laws, critics asserted they did not go far enough in giving power to the board. More bluntly, some critics declared the board toothless.
With this latest announcement, it is clear the by-laws are not what will be enforcing the decisions of the board. Instead, the reputation of the board's members will do the heavy lifting. Members of the board risk damage to their reputations if they appear to be bending to Facebook's will. At the same time, Mark Zuckerberg will face a hell-storm of angry users if he goes back on his commitment to implementing the board's decisions.
It seems a missed opportunity that such an esteemed group has to rely on these types of informal enforcement mechanisms to get their will enacted. Could a better-constructed set of by-laws have fixed this problem? In short, no. No other set of rules could have provided the board with more formal power than it has in this situation.
The reason for this is an economic concept called contractual incompleteness. All blockchain and DLT projects should understand this concept and the limitations it puts on formal governance.
Economics tells us that who owns an asset – in the above case, Facebook's platform and network – is just as important as contracts and agreements when it comes to making governance decisions regarding that asset. The Board and Facebook might agree to delegate some decisions to the Board under specified circumstances. However, it is impossible to predict every situation that could arise in the future and write a contract that covers all of them. This is the essence of incompleteness. When a situation arises that was not anticipated in the agreement, Facebook – the owner – gets to decide what to do. So, no matter the contents of the contract, rules, or by-laws, the owner will always have more control than other parties to the agreement.
Siva Vaidhyanathan, a professor of media studies at the University of Virginia, summed it up well in the Guardian: "The power of Facebook is its ability to choose what everyone sees." Whoever owns the platform and network has the power. And in this case, that control remains with Facebook.
This concept has broad implications for how organizations and markets are structured. Oliver Hart won the Nobel Prize in 2016 for his work in this area. Those implications are equally relevant to blockchain projects as they are to Facebook.
Blockchain projects – whether permissionless or permissioned – attempt to redistribute power and control in order to overcome a single point of failure, circumvent intermediaries, or reduce hold up. Multiple nodes (hardware assets) owned by different entities each independently maintain the shared ledger with the assistance of a consensus mechanism. Each node owner has control over its hardware, but not over the shared ledger itself.
Since the node operators own the hardware that maintains the ledger, they naturally play an outsized role in governance. No matter the process used to make a governance decision or who is involved in that process, if the node owners do not want to implement that decision, it will not be implemented. Projects or consortia that include other non-node entities in the governance process will face the same challenges as Facebook's oversight board. Informal mechanisms will be required to assist formal governance rules to ensure decisions are implemented in an orderly manner.
For blockchain projects, intellectual property (IP) ownership is even more important than hardware ownership. One of the critical decisions any enterprise blockchain consortium will make early on is how to address the IP rights over its software. We typically recommend that the most crucial elements of the network's software be made open source specifically to avoid the type of governance problem Facebook is now facing. Giving ownership control of essential IP assets to a single member is at odds with the decentralization benefits of using DLT and poses governance risks for the consortium.
It is possible to design governance rules and ownership allocation of nodes and IP rights in a way that distributes power effectively. Doing so requires understanding the participants in the consortium and the degree to which they currently wield power in the industries and markets in which they operate. Participants may have ongoing relationships outside the network that make them want to cooperate with each other as consortium partners. Like Facebook, a company’s commitment to certain actions may be reinforced by its reputation with its customers. These issues will impact the final design of the consortium’s governance.
Facebook's inability to create a genuinely independent body with real control over content decisions is an essential lesson for blockchain projects. Blockchain projects have an advantage that Facebook lacks; IP and hardware ownership rights can be allocated purposefully in order to optimize the distribution of power. However, even with these additional tools, distributing power is challenging, and the results will not be perfect.
The key to achieving the best outcomes possible is to utilize agreements and rules to target gaps left by ownership rights, accept the limitations of these formal tools, and embrace the role that informal mechanisms, such as reputation, can play. The ultimate governance design of a project will depend on the existing relationships and reputations of the parties involved in the project.