Governance And Decentralization

A detailed primer on various decentralized governance frameworks and their benefits and pitfalls.

Governance And Decentralization

What is Governance?

Governance is defined as “the act or process of governing or overseeing the control and direction of something (such as a country or an organization).”

Before we discuss any sort of governance, let's first define two models of governance: decision function and coordination.

The decision function model treats governance as a function for which the inputs are the wishes of various stakeholders and/or delegates, and the output is the decision.

Decision function is often used in approximation, meaning that it often leaves room for interpretation, breaking or bending of the law, and outright revolutions. This is perhaps by design. Systems like this allow the possibility of destruction, renewal, and iterations, all of which can be beneficial.

The coordination model sees governance as a layered structure. The bottom layer is always the ultimate deciding layer and cannot be stopped. However, the action of those above this layer will influence it. The second layer is an institutional coordination layer with the purpose of creating behavioral coordination among individuals – i.e., how individuals should act.

To illustrate how the coordination model works, let's use an example of a general and troops. A general rallies the troops not to inspire each individual to be brave, but rather to reassure each of them that everyone else feels brave and will excitedly charge forward.

Coordination institutions can operate formally or informally. However, signals are often ambiguous and conflicting. The pivotal questions to ask when considering governance should at least include:

  • What and/or where should layer 1 be?
  • What are the blockchain, protocol, and/or platform features, and how does this influence the ability to change the levels and kinds of power agents may hold?
  • What and/or where should layer 2 be?
  • What should be the focus and intention of governance, and how can beneficial participation be encouraged?

What is Decentralized Governance (DeGov)?

DeGov is often used to refer to a type of governance in which no one person owns or controls the organization, so decisions must be implemented by a diverse set of public and private actors. The concept of decentralized network governance relies upon social network theory.

There is a conflict to note here. One one hand, we have the values of coercion minimization and efficiency maximization. On the other hand, the logic of private ownership and markets optimizes for production and consumption.

Is DeGov Necessary?

The needs for DeGov come in two primary forms:

  • Funding public goods - how do valuable projects which do not have a business model get funded?
  • Protocol maintenance and upgrades - how are upgrades and regular maintenance attended to and agreed upon?

Projects should strive to accomplish a combination of sufficient developer funding and credible neutrality of funding.

A “true” DAO, for example, should in theory be able to achieve this balance. However, it would require a fair launch, and a perfectly fair launch – completely devoid of unfairness, information asymmetry, etc. – would be near-impossible. Nonetheless, significant steps have been made in some projects to realize funding of public goods in the form of sufficient and more credibly neutral developer funding.

The second matter is protocol maintenance and upgrades (although many advocate for minimizing all non-automated parameter adjustments).

Governance is unavoidable in a protocol’s early stages. There must be a system in place for events that require restructuring, and governance must remain in place while maturation takes place.

It is important to note that the ability to utilize off-chain governance may be restricted in specific cases. While a base-layer blockchain may facilitate the majority of its governance off-chain, some application-layer projects may not, since their smart contracts often inherently control external assets that cannot simply be forked.

A base-layer example would be Ethereum forking from Ethereum classic. As an application-layer example: if MakerDAO’s on-chain governance were captured, it's true that the community could fork the code; however, the assets under management (AUM) would be held in the original CDPs (collateralized debt positions). From this we can deduce that application-layer projects require a form of on-chain governance.

The Need for DeGov and Its Dangers

Many of the current iterations of decentralized governance are risky, especially coin voting. There are two primary issues with coin voting:

  • Misalignment of incentives and inequalities
  • Vulnerabilities through vote buying

Before we dive into these two primary issues, let's first explore types of coin voting.

Coin voting can be loosely coupled or tightly coupled. If it's used as a sole governance mechanism, these are equivalent to one another. Let's distinguish between the two. If a loosely coupled vote proposes a change, it then serves as a “green light” to encourage participants to vote in favor of a given proposal, so it is more of a coordination tool. A tightly coupled vote would seek to implement a change first, and if a group of participants disagreed, they could implement a hard fork.

However, there is a cost associated with a hard fork: in tightly coupled voting, the minority must sacrifice effort – and potentially capital – to preserve their decision. Thus a default bias is created for adoption in favor of the majority.

Now let’s return to the two primary issues mentioned above. Problems concerning the misalignment of incentives and inequalities in coin voting can be divided into three main areas:

  • A minority of wealthy participants holding large amounts of capital are more successful in executing decisions than a large majority of participants holding small amounts. Due to the tragedy of the commons, individual smallholders each have an insignificant influence on any outcome. So even with reward incentives, they may have little incentive to research, remain informed, and vote.
  • Governance coin holders and their interests are empowered at the cost of other parts of the community. Each community consists of a diverse constituency of different values and goals. Coin voting is biased toward only one constituency: the wealthy. This often leads to the prioritizing of price inflation and subsequent rent extraction.
  • Conflicting interests. The wealthy constituency holding the majority of power brings the inevitable risk of excessive exposure to the interests of that one group.

The vulnerability of vote buying is another problem which is becoming increasingly apparent, especially within decentralized finance (DeFi). Many protocols are bundling the right to participate in governance with some form of economic interest, such as a share of the protocol’s revenue. The objective of this approach is to attempt to align power and responsibility. Problems occur when those two rights are unbundled from each other.

As an example of this unbundling, consider a borrower using Compound. A user deposits ETH in a CDP which allows them to borrow an amount of XYZ for up to a percentage of value deposited. The user then can do whatever they want with this XYZ. If they wish to recover their ETH they must pay the XYZ back, usually with interest. The user can then take XYZ with no financial exposure to it and vote on a governance decision affecting XYZ’s price.

Note that the user may no longer care if the value goes up or down, because they have the nominal amount of XYZ to pay back into the CDP regardless of the outcome. Thus unbundling is achieved: the borrower has governance power with no economic interest, and the lender has economic interest with no governance power.

Some projects use timelocks to limit the potential vulnerabilities introduced by this unbundling. Timelocks require users to lock their coins in a contract for a period of time in order to vote. Some projects even use time-weighted incentives; Frax, for example, multiplies vote weight by time of lock, for up to four years. Although this strategy may limit the buy-vote-sell attacks, they can be bypassed via users holding and voting with their tokens through a contract which issues a wrapped version of that token.

Vulnerabilities and Issues of Coin Voting

There are many criticisms concerning coin voting mechanisms. Although this is not an exhaustive list, let’s discuss the main criticisms.

  • Low voter participation. Coin voting often exhibits low voter participation, which can interfere with perception of legitimacy since the vote only reflects a minority of holders’ views. A voter with only a small percentage of all coins can sway or manipulate the vote. This problem remains regardless of whether the vote is loosely- or tightly-coupled.
  • Game-theoretic attacks. Since the probability of any given voter individually having an impact on the results is low, their incentive to vote may be reduced. This means bribes may be readily applied. For example, exchanges may incentivize users to deposit coins in exchange for interest rates, and then use this deposit as they see fit. Exchanges profit from chaos; it is fair to assume their motivations are misaligned with users and holders.

It is important to note that loosely coupled voting, direct or indirect, still allows for bribing. However, if a community decides to, they can socially agree to ignore it or blocklist certain addresses. In a tightly coupled vote there is no ability to create a blocklist, because that would be a governance decision in itself. This may suggest that tightly coupled voting is more inherently vulnerable to bribe attacks.

  • Non-representativeness. A coin holder’s vote represents only one user. In Bitcoin’s case, its store-of-value (security) use case conflicts with its medium-of-exchange (usability) use case. The same argument can be made for Ethereum and other such blockchains; perhaps the conflict is even greater. This issue applies to both loosely- and tightly-coupled voting.
  • Centralization. Centralization risks also apply to both loosely- and tightly-coupled voting. As an example of a potential centralization vector, consider Ethereum miners initially following the direction favored by the community (core developers, Ethereum foundation, etc.) If voters are not technically competent, they often defer to a small and dominant group of “experts,” thus exposing the system to increased risk of centralization. (Over time, however, Ethereum miners began to adjust the gas limit without consulting core developers of the Ethereum Foundation).
  • Digital constitutions. “Digital constitutions” use verifiable proof to satisfy specific properties which act as coordination flags to help reinforce the desired values of a protocol. It’s a good idea, and it’s better suited to operate as a loosely coupled (layer 2) vote rather than a tightly coupled (layer 1) vote.

Potential Solutions and Strategies for Governance and Coin Voting

Within governance, a multi-factorial consensus may be recommended for the sake of balance. A system solely controlled by core developer consensus can result in control disseminated from “ivory tower intellectuals” with a focus on technically impressive solutions and abstract philosophies rather than day-to-day concerns. Ideally, to help inform decisions, different groups, mechanisms, and coordination flags should be used and pooled into a collective result. These could include:

  • Coin holder votes
  • User votes via a sybil-resistant system
  • Core developer team consensus
  • Project roadmap
  • Established norms pertaining to that specific project

Whether or not coin voting can remain sustainably resistant against attacks, potential solutions and strategies can be implemented.

  • Delegation. Delegation seeks to address the low participation rate of smallholders by mitigating conflicts of interest introduced by wealthy minorities. With delegation, smallholders don’t personally judge and participate in each decision. Instead they elect to delegate their votes (or “proxies”) to a member of the community whom they trust. However, this does still expose the system to coin-holder-centrism – a problem seemingly baked into coin voting. Although not the intention, it nonetheless poses a challenge.
  • Limited governance. Limited governance reduces vulnerabilities by placing a limit on what coin governance can do. Viable options include:

    - Using on-chain governance for applications, but not base layers
    - Limiting governance to fixed parameters, which could be reduced over time
    - Adding time delays
    - Being fork-friendly

    Limited governance may not be an acceptable solution when used on its own, since public goods funding is still vulnerable to malicious and manipulative attacks.
  • Non-coin-driven governance. There are alternatives to coin voting that can be used separately or adjunctively:
  1. Proof of personhood systems verify that there is a unique individual human that corresponds to an address, which helps enforce one vote per human.
  2. Proof of participation systems track and attest that an address corresponds to a human who has taken a specific action: voting, participating in an event, completing training, submitting work, paying off debt, etc. This could be synergized with soulbound tokens (SBTs), for better or worse. One challenge with these systems is determining what and how much constitutes participation.
  3. Hybrid systems such as quadratic voting (a combination of coin voting and proof of personhood). Quadratic voting weights a single voter using the square root of the economic resources the voter has committed. The idea is that even if an individual splits their resources across multiple identities, the financial component still exists and helps to protect against manipulation.
  • Skin-in-the-game. (Insert inspirational Warren Buffet quote here). While coin voters are often held collectively accountable for their decisions, each voter is not individually accountable. Fork-friendliness is a built-in line of defense against extremely undesirable decisions: if such a decision succeeds, users may produce a fork that can destroy all coins voting for the bad decision. This creates individual responsibility when voting, and simultaneously increases the risk of lending governance tokens. However, this does not mitigate low-scale manipulation, nor low voter participation for that matter. Selection pressure must be encouraged somehow to create preference for higher-quality opinions and decisions.

    Futarchy, which creates a betting system, is a popular solution to this dilemma. To vote for a proposal, users wager their coins on whether their option of choice (for or against) will lead to a good outcome or a bad outcome. Good outcomes are rewarded; in bad outcomes users forfeit their coins. Pure futarchy systems are difficult to implement, but hybrid futarchy systems which include a form of non-futarchy governance can become more usable. Some examples of hybrid futarchy:
  1. Votes as buy orders require an enforceable buy order for additional tokens at a lower price, which helps ensure that those who support a bad decision are forced to buy their opponents out. Furthermore, it gives “normal” decisions a greater degree of leeway.
  2. Retroactive public goods funding enables users to buy the project tokens and retroactively fund their project after objectives and results are realized. If the project is successful, users are rewarded.
  3. Escalation games use the potential appeal of a higher-effort, higher-accuracy, higher-level process to help align values on low-level decisions. Users whose votes agree with the ultimate decision gain the reward.

Hybrid Solutions

  • Time delay + elected-specialists governance. Coin holders may elect providers or specialists (this could be oracles in some projects, advisors, team members, etc.) If the elected party is later deemed unworthy for whatever reason, they have a specific amount of time to appraise and rectify the situation or complaint before the project switches different provider or specialist.
  • Futarchy + anti-collusion = reputation. Users may vote with “reputation,” which may come in the form of a non-transferable soulbound token. In this case users improve their reputation through beneficial decision-making, while undesirable outcomes harm their reputation.
  • Advisory coins are tokens that do not directly implement change. Instead, they improve off-chain governance by indicating decision outcomes. Advisory coins enable the benefits of coin voting, but mitigate the risks.

Closing Thoughts

With consideration to all of the above, coin voting may function as one of many coordination methods that help voters decide on implementing changes in a project. However, coin voting should not be the only means of coordination. If coin voting becomes part of governance rather than the dominant form of governance, we can build more legitimate forms of decentralized governance.

Coin voting may appear neutral since anyone can buy a governance token in an open market, but many large positions are in the hands of elites.

In this primer we have not touched on different forms of consensus, plutocracy, and collusion as seen in governance, DeFi, and the wider crypto ecosystem. We will explore these in later articles.

Join us in the new #dao-governance channel to take part in the decentralized revolution! If you are interested to hear a bit more about the challenges of governance, check out the latest episode of the Starbased Podcast.

twitter starbased