DAO Governance Lessons Thus Far

We have already written a primer on DAOs. If you’re unfamiliar with DAOs, we strongly encourage that you read that article first before continuing.  

Since Vitalik Buterin first proposed the concept of a DAO in the Ethereum whitepaper (and a quiet period after the DAO hack), DAOs have re-emerged to play a prominent role in how projects are governed and run on the blockchain. Crypto-native organizations and communities love to identify themselves as a DAO, but sometimes this appears to be a marketing strategy more than anything. 

The reality is that DAOs are just like almost everything in the crypto space, nascent and still evolving. This makes discussions surrounding DAOs a messy topic, lacking coherency and full of nuance. We look at some of the recent significant events surrounding DAOs, compare how this would have played out within traditional organizations, and see if there are any lessons to be gleaned from such circumstances. 

The DAOs of today

Before analyzing the DAO model, we must first consider the ‘ideal DAO.’ If we could disregard all the impracticalities of reality, what would the perfect DAO look like? While there is no hard consensus, a safe assumption would embody the core components of the DAO abbreviation itself: purely decentralized and autonomous. 

Decentralization is meant to minimize the concentration of power within a central entity or group while being autonomous retains the right to govern oneself. Underpinning these two terms are the principles of fairness and equitable decision-making. Thus, when read together, we can extrapolate that an ideal DAO is decentralized and has the means to govern itself or control its affairs fairly and equitably. 

Since DAOs were first incepted as a new form of governance for crypto communities, there is also an expectation that DAOs should embrace blockchain technology and its underlying principles (though there is no reason why the DAO model cannot expand beyond the blockchain realm). 

Since we have considered the ideal DAO, the question then becomes; what do DAOs look like today? So far, despite the many different structures and setups, DAOs seem to be a never-ending cycle of tradeoffs between ideals and practicality, mirroring the blockchain trilemma. We look at the following events through this lens of tradeoffs.  

DAO Tradeoffs

Accessibility v Accountability

Blockchain technology grants us the ability to be pseudonymous, reaffirming the mantra ‘trustless yet secure.’ We need not know anyone’s real identity because they can be monitored and regulated through the blockchain. DAOs are typically always open - anyone can start or participate in a DAO, paving the way for global access to different projects and communities. Participants are judged based on their merits and contributions, not their background. However, that easy accessibility is a dual-edged sword.

In an ideal DAO, we would have the tools to regulate every action and remove centralized layers of power. But in practice, the DAOs of today are still fundamentally dependent on individuals/entities, some of which may be pseudonymous. The community can claim as much control as they want, but if the multisig holders of the treasury wallet decide collectively to rug the project, the community will get rugged. Or, if the developers abandon the project without handing over the private keys to the community, there’s not much they can do about it. Parties who have known identities may be subject to legal enforcement, but that detracts from the nature of DAOs in the first place on having to rely on external parties (e.g., regulators) to help preserve the DAO’s sanctity.

A great example is the ongoing saga in Frog Nation, a loose conglomerate of DeFi projects led by prolific DeFi figure Daniele Sestagalli. On January 27, 2022, @zachxbt revealed a slew of damning revelations about 0xSifu, the ‘de facto’ CFO and multi-sig holder of the Wonderland treasury at the time. 0xSifu’s real identity was Michael Patryn, a figure formerly convicted for identity theft and co-founded infamous Canadian exchange QuadrigaCX, whose co-founder disappeared with $169M. Sifu was accused of removing funds from the Wonderland treasury and laundering it for his gain. The event and subsequent revelations sent the price of $TIME tumbling. 

Sestagalli later admitted that he was aware of 0xSifu’s origins but believed an individual’s past doesn’t determine their future. This belief, however, was not shared by other members of the community; after the news broke, the Wonderland DAO voted overwhelmingly to remove 0xSifu as the Treasury Manager effectively immediately. Sestagalli subsequently proposed to “wind down Wonderland, and give the treasury back to its holders,” which was narrowly defeated after several whales lost out to the majority of smaller wallet holders, who had suffered massive losses on their $TIME holdings. 

Immediately after the vote, Sestagalli seemed to indicate that he would unilaterally dissolve the project against the wishes of the community. However, he subsequently backtracked and decided that he would manage the DAO directly, effectively becoming the sole dictator decision-maker for Wonderland. 

The Frog Nation Saga is far from over and will undoubtedly be an important case study for governance. There are glaring shortcomings regarding the nature of anonymity and governance. People should not be discriminated against based on their past, but there are many counter-arguments to that point. There is a reason why the judicial system recognizes the importance of knowing an individual’s prior convictions; for example, in the UK, prior convictions may be adduced as admissible evidence under certain circumstances to help prove the increased likelihood of the individual to commit an offense. On the other hand, sentences are meant to be rehabilitative as a person ‘should not be judged strenuously by reference to the awesome specter of his past life. The juxtaposition here is challenging to balance.

Although crypto should be open to anyone, there is another issue regarding disclosure. As a  member of management in a traditional company, 0xSifu would have been required to disclose his history. He may even be barred from becoming a director, depending on the nature of the crime. While crypto has no such inhibitions, does the community have the right to know? Should Sestagalli have disclosed this crucial piece of information once he found out instead of hiding it from the community? Would this conflict with crypto’s principles of respecting someone’s pseudonymous identity and judging someone based only on his track record in crypto? The conflict between the right to privacy versus the right to know never ends.

The Frog Nation is but one example. The reality is that all members of a DAO rely almost exclusively on faith in select individuals who have the option of foregoing any accountability for their actions. 

Decentralization v Efficiency 

DAOs usually employ a hybrid democratic approach, possessing a flat hierarchy that incorporates elements from representative democracy where specific individuals are elected to act on behalf of the community. These individuals will have control over operational matters such as managing social media accounts or special “powers” exercised in emergencies.

There is little debate that this tradeoff makes sense for operational efficiency. Some decisions may be too trivial or highly time-sensitive, and cannot wait for a two-week governance vote. However, how the individuals choose to exercise their power to act in the best interest of the DAO / project is often up for debate. 

In a galaxy far, far away, we have the Curve wars. Curve is a stablecoin exchange that rewards its liquidity providers (LP) with rewards in its governance token, $CRV. Holding a minimum of $CRV tokens and locking it up allows users to influence the protocol through governance votes, including adjusting LP/yield farming rewards for different trading pairs. Many DeFi protocols have sought to take advantage of this feature by accumulating $CRV tokens and attracting liquidity for their native stablecoins. However, an upstart named Mochi Finance (now Mochi Inu) decided to take this to the extreme. 

Mochi is an overcollateralized lending protocol that accumulates and lists collateral assets. Borrowers can mint $USDM, Mochi’s native stablecoin, in return. On November 11, 2021, Mochi announced the acquisition of 1M $CVX, Convex’s native governance token. Being the primary “yield aggregator” for Curve, Convex had accumulated the largest share of $CRV tokens and hence wielded the most influence over LP rewards for trading pairs on Curve. Mochi used the CVX tokens to boost its $USDM LP rewards on Curve. It then allowed users to mint free $MOCHI (their native governance token), redeem $MOCHI for ‘free’ $USDM, and then use it to buy more $CVX to boost the project’s $CRV gauge rewards. This created an infinitely repeatable positive feedback loop, “gaming” Curve’s incentive system. Through this, Mochi managed to accumulate over $100M in TVL in $USDM pools very quickly. Now, there are obvious risks here. Mochi is essentially leveraging its token, putting their LP users at risk if the price of any of the assets involved ($CVX, $CRV, or $MOCHI) drops. Furthermore, there were concerns about undercollateralization for $USDM since some regarded $MOCHI as ‘worthless’ and was used for gaming the Curve system, as opposed to being an actual governance token. 

You could imagine this sparked outrage from other pools in the Curve ecosystem. If Mochi were allowed to continue, rewards for other trading pairs would eventually dry up as Mochi accumulated more $CRV. To prevent this, Curve’s Emergency DAO executed their first-ever governance proposal to kill all $CRV rewards to the $USDM pool. 

There are two sides to every coin. The existence of Curve’s Emergency DAO is questionable in the first place. While the members were appointed by the main Curve DAO, they have the power to take action without consulting its members; it can be seen as a subDAO or a committee in many ways. Nevertheless, they purportedly acted with good faith on behalf of their community. 

That’s fine and all but let’s consider another scenario. What happens if they do the opposite and do nothing? Maybe there are malicious actors in the Emergency DAO looking to take advantage of the entire debacle? Wouldn’t the Curve community be increasingly at risk? Alternatively, the decision to kill off the $USDM rewards could have been left to a broader community vote, but that would take a long time and could have led to more severe consequences for all stakeholders, including Mochi users. 

There is no clear answer here. The utilitarian mindset (greatest benefit to the greatest number of users) justifies the existence and actions of Curve’s Emergency DAO. Still, some would argue that this goes against “total” decentralization principles. 

Openness v Control

The general expectation for DAOs is that they eventually adopt a governance token model. Usually, the weight of one token equals one vote, similar to how shares in a company or voting in a democracy traditionally worked. There are variations, but for the most part, the more tokens you have, the more voting power you have. In other words, whales have the largest influence. We have seen this occur in the Frog Nation Saga and the Curve Wars. However, Curve takes it one step further as the Emergency DAO effectively limited the ability of CRV tokens to ‘govern’ the USDM pool. Although it can be reinstated, the point is that should such action be accepted first? After all, Mochi was only taking advantage of a ‘feature’ available for anyone to use.  

Lately, there have been alleged instances of ‘governance attacks’ by Justin Sun, Tron’s infamous billionaire founder. On-chain activity suggests that Sun has been making governance proposals on several DeFi lending protocols and borrowing vast amounts of the protocol’s governance token during the voting period to help improve the adoption of Tron’s native stablecoin, TUSD. The most recent one was to add $TUSD as a collateral asset on Compound using $COMP. Should this behavior be encouraged or prohibited? Where does the line get drawn? What if it was a more risky proposal that definitively endangered Compound’s users? Should the team refuse to implement the proposal if it passes? 

We have also seen this in the recent Building DAO fiasco where a malicious actor managed to slip in a governance proposal which allowed him to mint an unlimited number of the protocol’s governance token, $BUILD. Since Building DAO utilized a self-executing on-chain contract, the team had no control once changes to the smart contract were accepted. Disregarding the poor governance controls (e.g., low quorum threshold and short voting periods), if the team had control before the smart contract changes took effect; would it have been acceptable for them to intervene?

Governance attacks are not new. Publicly listed companies are always at risk of hostile takeovers, which is a natural consequence of the share/token model. However, there is something else entirely that only DAOs experience, namely, forking.

Openness in blockchain takes on another meaning because code is open-sourced. Smart contracts can be forked. Movements can be forked. Even communities can be forked. Nothing can stop a decision to fork if there is enough momentum and mindshare behind a decision to fork. Intellectual property rights do not exist in crypto. Everyone has the freedom to steal borrow ideas from each other, and users can vote with their money on the winners. Ethereum had done it when ideologies diverged. Even established protocols like Sushiswap emerged from a fork (through a vampire attack on Uniswap). There are also countless projects in the NFT space that have created derivatives of an established collection. However, none have done it quite like Wrapped Penguins, a fork of the Pudgy Penguins NFT project.

From the start, the Pudgy Penguins team was a doxxed centralized team that made many unfulfilled promises to the community. Over time, dissatisfaction with how the Pudgy Penguins were run led to calls for further decentralization and conversion into a DAO. While the team agreed to step down, they were only willing to do so if they sold the project and received the funds from the sale. Offers came flying in from different parties. Nevertheless, a portion of the community was unhappy with the way things were progressing, especially since the proceeds from the sale would still accrue to the founders. Not to mention that the team was still collecting royalties for every Penguin sale. Led by a group of rebellious individuals such as VincentVanDough, a new smart contract was made whereby existing holders could wrap their Penguins. Any NFT sales made on marketplaces like OpenSea would be redirected to a wallet controlled by the newly formed Wrapped Penguins DAO. It remains to be seen as to whether there will be any reconciliation between the two communities. The original Pudgy Penguins project has not yet been sold at the time of writing, and it will be interesting to see how the situation develops. In any event, this showcases how DAOs can be used as an ‘oversight body,’ keeping project teams on their toes as they face the threat of being forked.

The lesson here is that while the DAO concept of “one token one vote” may seem the fairest on the surface, it opens the DAO up to parties “gaming” or taking unintended advantage of these features. While there may be fail-safes in place to prevent such “attacks,” these run contrary to the principles of a DAO. And behind this constant vie for control is the ever-present risk that you might get forked.

Current DAOs are still full of flaws and contradictions, and we can do better

We have to accept the oxymoronic truth; there will always be some form of centralization in DAOs. DAOs are great at kickstarting short-term movements but eventually require more centralized mechanisms to manage them efficiently. Protecting the overall interests of a community often means disregarding the underlying principles of a DAO. Even for DAOs that truly embrace decentralization, there will always be indirect concentrations of influence. 

Evolutionary leadership theory teaches us that humans are inherently bad at cooperating without a leader. Our success as homo sapiens can largely be attributed to group coordination facilitated through leaders; in other words, centralized powers. It is hard to imagine projects getting to the height of their success without their respective ‘spiritual’ leaders: we had 0xMaki and Sushi, Andre Cronje/banteg and Yearn Finance, and even Charles Hoskinson and Cardano.

This is, of course, not an advocation piece for benevolent dictators. Rather, it is an exercise to highlight the shortcomings of decentralization for social coordination, especially for DAOs which exists outside the protections offered by centralized regulators. Just take a look at the decline of ancient Greece. Constant internal disputes and the lack of a unified central government (Greece adopted a highly decentralized state-city system) eventually led to their downfall when foreign powers invaded. And yet, despite all this, decentralization is still the future.

It is striking how in 1999, a World Bank article entitled “From Centralized to Decentralized Governance” predicted the rise of decentralized governments in developing countries. The conclusion, however, perfectly encapsulates the current state of DAOs:

“Strategies aimed at stopping decentralization are unlikely to succeed. The pressures to decentralize are beyond government control. … Rather than resisting these pressures, countries in these regions should learn from countries that have gone before.”

The DAO model is still developing and still has a long way to go; however, there are many things that we can learn from centralized models of governance. On top of that, there are unique advantages that DAOs can leverage to improve accountability and governance efficiency. Here are some suggestions:

1. Adopt more traditional corporate governance practices such as governance audits 

2. Introduce performance-based incentives for the team instead of/in addition to timelock incentives

3. Isolate pools of treasury capital for individuals to manage

4. Implement succession planning, private key management, and password management (e.g., social media accounts) strategies

5. Consider alternatives to token-weighted votes, e.g., address-weighted or time-weighted votes

6. A decentralized reputation system to vouch for anonymous individuals (e.g., Meritverse)

7. Develop and implement DAO tools that promote decentralization and remove concentrated points of failure

These suggestions are merely food for thought and should not be construed as advice. If there was one key takeaway, however, it is that we should set the bar higher. We need to be more proactive about governing ourselves, DAO or no DAO.

Some might see DAOs as a cop-out on the path towards true decentralization, and in many ways, it is. However, until we reach the point where 'immutability and the lack of human interaction' becomes a feature, DAOs are perhaps the best we got for on-chain governance. We might as well make the best of it and learn from experience.

Watch our Video Guide