This article was originally posted as Orca Protocol on orca.mirror.xyz.
Written by Dan Wu (@itsdanwu)
Improving member participation in governance remains one of the hottest and most nuanced problems facing DAOs today. However, these are not new problems we are facing. Roy Lerner wrote ‘Blockchain Voter Apathy’ describing the barriers for effective governance and potential solutions for increasing voter participation over two years ago. Despite his plea for change, governance participation has made little to no meaningful progress.
The purpose of this piece is to provide a high level primer on the current state of governance and participation. Some topics explored include areas where governance is needed, analysis of on-chain voter participation, common issues with governance today, and recommendations on how to resolve some of these issues identified.
It is important to take a step back and understand the “why” behind governance before we reexamine issues and potential solutions.
As outlined in ‘Pods: The DAOnfall of Token Voting’, governance is simply a tool for organizations to decide where power resides and who gets to click what button. By placing rails around operations, organizations can quickly get from point A to point B while minimizing risk.
An ideal governance model is one that maximizes a DAO’s meaningful participation and the realization of its mission.
Tarun Chitra provided an excellent description of when governance is needed within DAOs in his recent piece for a16z,
DAOs work best when the governance burden related to curation, security, and risk can be reduced faster than the natural increase in coordination costs that accompanies the need to have members involved in voting on every decision made.
The need for governance and diligent decision making goes up as:
So now that we know when governance is needed, the next logical question is where is governance usually applied within protocols/DAOs?
Governance is usually required for decisions in which the result is unknown and cannot be derived from on-chain data, thus requiring some human input. These decisions should be made with the mission, vision and goals of the DAO in mind.
A few key areas in which governance is likely needed for the foreseeable future:
It’s crucial to leverage the vast amounts of on-chain data available to help set some ground truths about governance participation’s current state.
Some of the questions this data can help answer include:
On-chain voter participation by DAO size seemed like the most basic place to start—how does voter participation change as DAOs scale?
The results are not too surprising—as DAOs scale (# of members), their voter participation (# of voters as a % of members) falls. On-chain voter participation starts out around 50% for the smallest DAOs and falls to around 5% for the largest DAOs. While a simplistic view, this gives a sense of the trend for DAO voters that cast at least one vote within a DAO as it scales.
Examining on-chain voter participation of a large protocol also helps draw more concrete conclusions—Compound turned out to be a good candidate based on its size and maturity. Compound is a decentralized blockchain-based protocol that lets users lend and borrow assets—while also providing these same users a say in governance via their native $COMP token.
Compound’s on-chain voter participation (unique wallet addresses voted as a % of total delegates, including self delegates) has averaged ~3% across the ~70 on-chain proposals that have been executed.
The majority of Compound’s proposals have been related to DeFi-based technical parameter setting and average relatively lower voter participation (e.g., Proposal 65 was a follow-up fix to the COMP bug and had ~3.5% voter participation).
However, there have been a few proposals related to other topics that resulted in different voter participation behavior.
Based on the behavior of Proposals 40 and 43, it’s clear that not all proposals are created the same. The subject matter of the proposal, materiality and breadth of impact all play a major role in whether voters feel qualified and empowered to vote.
Voter participation as a percentage of total circulating supply is one perspective to understanding governance participation. Delegation is an arguably more important perspective—how are those who have been deemed decision makers for the DAO/protocol participating in governance? While quite popular, delegation is sometimes seen as a solution for low voter turnout (e.g., ENS including delegation at the end of the airdrop claiming flow). The truth is there is more nuance to delegation than meets the eye.
There are two aspects of delegation to analyze given most of the major protocols have token weighted voting:
Average token voting power of delegates who have voted: ~3900 COMP
Average token voting power of delegates who haven’t voted: ~500 COMP
Shockingly, 3/4 of all $COMP delegates have never cast a single on-chain vote. Despite this, 80% of all delegated $COMP have participated in governance directly identifying a few "whale" delegates with massive amounts of voting power (such as a16z, Gauntlet, and Getty Hill). This is a major gap: the majority of delegate voters—who assume the responsibility of participation—do not participate in governance.
Regardless of governance participation levels, Compound has had success from a financial metric perspective—growing from ~$1.5B in TVL to ~$12B in just under one year. However, it seems that there is still ample room to improve participation where needed, harden governance structures where relevant and minimize governance involvement where possible.
Now that we have some context and benchmarks on governance participation, we can further explore some primary issues and paths towards a better tomorrow.
Three primary issues with governance were identified on Week 3 of a recent ‘DAO Vibe Ops’ Twitter space with several DAO / governance thought leaders and operators.
Information accessibility problems may best be portrayed by some recent direct user feedback provided in the Discord server of Boardroom, a platform used by DAOs to engage, signal and vote on decisions:
In due time, the UX for governance management platforms will improve to solve some of the pain points identified here. Out of the three governance issues identified, information accessibility is the easiest one to solve.
We are already seeing builders within the ecosystem tinker and attempt to solve some of these pain points:
HAL has built a no-code tool to help teams remain informed on new Snapshot proposals by automating some of the workflow required today. Between the proliferation of workflow automation platforms and tools in the web2 world (e.g., Zapier) and projection that the workflow automation market is expected to balloon to $55B by 2028, there is major promise for tools like this.
Engagement and participation is the flip side of information accessibility. Standards for what proposals look like and processes/tools for how they are shared are important to governance accessibility, but governance thinkers also need to determine the relevant people required to vote for a specific proposal in the first place.
Optimizing for the right people to participate in governance processes is an ever-evolving discussion in governance circles today. It may not be pragmatic to assume that all voters should participate in every governance decision. Participating in governance is a lot of work—reading and digesting a proposal, providing feedback for discussion and going to vote can be mentally fatiguing. In addition, many do not have the subject matter expertise or desire to participate in proposals outside of their domain. I.e., An operations lead may not be able to form a sound opinion on a technical protocol development proposal and vice versa.
Friends With Benefits (FWB) had a recent example of poor engagement related to voting on applicants for the FWB Fellowship Program. The FWB Fellowship Program is intended to create on-ramps for creative individuals who lack the financial resources necessary to buy tokens on the open market. As long as the FWB Fellows meet participation requirements, they receive a token grant for continued membership.
FWB member sirsu recently made a public plea encouraging FWB members to vote on Season 4 Fellow applicants:
While sirsu rightfully points out that 46 votes out of 400+ is a little low, asking for more members to vote is a request that can be challenged:
FWB operates in a specific context and may warrant higher numbers for participation—no two DAOs/protocols are the same. However, as seen in some of the data above, voter participation rate is harder to maintain as the organization scales. With fewer than 500 members and low participation rates today, it’s clear that the “everyone votes on everything” model causes problems in both big DAOs and small-ish social clubs.
Instead of expecting to drive higher engagement across all of FWB, it may be more worthwhile to understand what a trustworthy 10% could look like in the context of the FWB Fellows program. Finding ways to delegate and distribute decision making to the right groups in the future will be essential for FWB to scale.
Accountability within a DAO is all about setting and holding contributors and governance participants to a common expectation aligned to the DAO’s mission, values and goals.
One of the problems with DAOs and accountability today is that many are operating too fast, have fuzzy priorities and vague expectations in alignment to the DAO’s mission. This leads to performance minimizing organizational habits in the form of poorly defined roles and responsibilities.
Tactically, good accountability for governance participants within a DAO starts with:
And when accountability is upheld, reputation is forged. Reputation is one of the most important pieces to driving accountability. And as you might imagine, reputation is one of the more important intrinsic motivators for individual and collective performance.
Lucky for us, we are starting to see innovation in the reputation management space. @KashDanda recently outlined how SuperteamDAO is leveraging a federated organizational model to build accountability, reward and reputation mechanisms.
SuperteamDAO’s accountability and reward structure revolves around projects. All projects are assigned 1000 “XP Points” that a Scope Squad and Project Lead allocate across all tasks needed to be completed for the project. Tasks are grouped by skill required to perform the task. Project Leads are ultimately accountable for the completion and success of the project. Once complete, project contributors are rewarded the XP points assigned to the tasks they have completed. The plan is that these points will flow to a transparent “Reputation Leaderboard”.
The transparency and intention of SuperteamDAO’s process should promote performance, ownership, and trust across all of their project teams - and can serve as a benchmark framework for others to follow.
Information accessibility, engagement and accountability are critical issues for DAOs right now. DAOs are increasingly recognizing the need for governance primitives to address these issues. The elegance of governance primitives, such as pods, is precisely in their simplicity.
The pod primitive is designed with small working groups in mind. With these smaller groups, internal trust naturally increases and the need for formal communication decreases. Information becomes more accessible for relevant individuals as the workflows, tactics and communication behaviors required to close pending decisions simplify - they simply take less time and energy.
Pods greatly help with avoiding the governance engagement traps we have seen in the past, such as the FWB Fellows example outlined above. FWB Fellows selection and overall program progress was delayed due to lack of voter participation. A podified FWB Fellows committee can be used to limit the scope of engagement required for FWB Fellows program decisions. This can be a good thing for FWB: to delegate operational decision making to smaller working groups and simultaneously decentralize ownership of a DAO. This type of structure should empower contributors and allow for increased clarity and focus across the community.
The last thing we will highlight here is that pods drive accountability within and between other pods, which promotes higher quality governance. There are two primary ways in which pods drive accountability:
It’s clear that there are issues with governance participation today. Oftentimes governance participation is low and dominated by the few delegates with the majority of the voting power. Other times, governance participation may be too high and stifle the agility of a protocol/DAO.
Many DAOs and protocols have ample opportunity to improve how they:
The ecosystem is already seeing innovative solutions being built to address some of these problems, such as HAL’s no-code solution to automate notifications for Snapshot proposal creation and SuperteamDAO’s reputation management system.
The pod primitive will be foundational in improving governance by pushing decision making out to aligned stakeholders, hardening roles and responsibilities, identifying existing power systems and elevating contributors through autonomy.
Metropolis is the on-chain permissions layer for DAOs.