Ambition made to counteract ambition - DAO governance and bicameralism
On the parallels between Lido's dual governance proposal and bicameralism and what DAOs can learn from constitutional engineering.
My friend at Friday Catalyst recently picked up on an interesting DAO proposal on Lido, the largest liquid staking protocol in DeFi. The crux of the proposal argues for introducing a Dual Governance scheme, in which stETH holders will also be granted governing power in addition to LDO holders.
In unpacking the details of this proposal, we will see how DAOs inevitably face the same problems that has baffled organized communities since the dawn of time. We will see how despite that DAOs are equipped with novel bells and whistles such as “self-executing smart contracts” and “governance tokens on an immutable blockchain”, governance eventually remains an issue of institutional design that technology alone is unable to solve.
We will then explore how the Dual Governance scheme resembles the structure of a bicameral legislature (with specific reference to the U.S. Congress) and inherits the spirit of check and balance to address moral hazards.
What are Lido and liquid staking?
Normal staking on the Ethereum blockchain works like this: you lock your ETH into a staking pool and gradually receive more ETH as a reward for helping to secure the blockchain network. But once you lock your tokens for staking, you can no longer use them for other purposes.
A liquid staking protocol like Lido tells people: stake your ETH with us, and we will give you stETH (staked ETH) in return. The new stETH is a tokenized share of the amount of ETH you have staked. Now, on top of accruing staking rewards, you can take your stETH and deploy it elsewhere across the DeFi ecosystem freely (e.g. to transfer, trade, post as collateral for borrowing, etc.).
A liquid staking protocol is helpful for a couple of reasons. First, to stake on Ethereum you need to meet the 32 ETH minimum threshold. Lido makes staking more accessible by pooling the money together so that not just the whales enjoy staking rewards. Second, staking and unstaking on the network takes time. Depending on the blockchain, it could take any duration, from a few days to a few months. For Ethereum, the early stakers may not be able to unstake until the Merge's1 later phases, which could even take years.
As of the time of writing, more than 4.2 million ETH (~$5.8 bn) are staked with Lido. This means Lido roughly controls about 32% of the 13 million ETH staked. Thus, the holders of LDO, Lido's governance token, hold significant governance power not just on the protocol itself but also on the Ethereum network.
Lido’s Dual Governance Proposal
Given the sizable amount of stake the protocol controls, core developers at Lido believe something has to change with how the Lido is governed to prevent moral hazard. The proposal seeks to address the principal-agent problem that arises from the current state of governance in which LDO holders, who are the agent, are motivated to make decisions in their own self-interests, contrary to those of the stETH holders, who are the principals.
The stakers' interests, in this case, are better aligned with the Ethereum network; those of LDO holders, however, are not so much. Fear of staker coercion, censorship, and cartelization (concentration of node operators) naturally arises. In a worst-case scenario, LDO holders could even stage a heist and steal the ETH staked in the smart contract, abusing their controlling power over liquid staking codes. (For more details, you can refer to this blog post from Friday Catalyst).
The Dual Governance scheme aims to better align the two parties' incentives by ensuring that this worst-case scenario would not happen. Under the scheme, protocol changes are still proposed by LDO holders, but stakers would be granted veto rights and can shut down proposals passed by LDO governance. The veto power would be essential to protect stakers' interests by preventing protocol misalignment or instances of governance capture.
The below flow chart describes the new proposed process. After a Lido proposal is passed, a time lock will be applied. During this period, the stETH holders could reach a quorum and convert the entire governance module to a Veto State, where no passed proposals can be executed. To ensure governance does not completely remain in impasse, stETH holders can create an anti-veto vote for specific proposals that are passed but not yet executed.
The proposal includes more details on unstuck mechanisms and protection of veto abuse, which you can read more about in the original forum discussion.
Bicameralism - Dual Chamber Legislature
The design of Lido’s dual governance scheme mimics principles of bicameralism. Bicameralism refers to a type of government legislature in which the law-making body consists of two chambers or houses. For instance, the U.S. Congress comprises the House of Representatives and the Senate. During the conception of the U.S. Constitution, the Framers deliberately designed different membership and scopes of governance for the two Houses and incorporated the principle of check and balance to prevent abuse of power and to protect civil liberties.
For instance, the number of representatives in the House is directly related to each State's population. The representatives are elected every two years. On the other hand, Members of the Senate are appointed by the state legislatures for six-year terms, which are staggered so that one-third of the senators will be re-elected every two years. The number of senators is also fixed as two senators per State, without regard to each State's population.
The Constitution assigns distinctive scopes to the House and the Senate regarding functions and power. The Senate holds power to ratify treaties and approve presidential appointments, while The House has the sole authority to originate revenue bills (tax bills). And at the end, each legislative body's approval is required to enact a law.
In devising the structure of Congress, the Framers also faced their own version of the principle-agent problem, which is one between the elected officials (agent) and citizens (principal). The bicameral design seeks to strike a balance between preventing the government from being captured by populist “mob rule” through the Senate and, at the same time, from being too distant from the people’s popular opinion and becoming detached from its electorate.
Of course, such a constitutional arrangement is as much a result of deliberate design as a historical inevitability since it is deeply rooted in realpolitik and the tug-of-war between populous and less populous states. Nonetheless, the Great Compromise of 1787 eventually allowed for lower chamber representation based on population and upper chamber representation based on equality among the states.
Dual Governance and Bicameralism
Many parallels between dual governance and bicameralism can be observed. On a high level, both seek to mitigate the principal-agent problem through better alignment of interests and both seek to limit the power of the governing body by introducing elements of checks and balances. One level deeper, we can observe four main rationales for a bicameral legislature and dual governance scheme: representational diversity, virtues of delay, specialization and predictability.
1）Representational Diversity. For the U.S. Congress, the Senate serves as a check on popular passions and thus on the possibility of majority tyranny from the house of representatives. In the case of Lido, dual governance increases representational diversity by incorporating stETH holders' interests in addition to LDO Holders. stETH holders here act as a check to prevent governance capture by LDO whales.
2) Virtue of Delay. The bicameral legislature and dual governance scheme increase the complexity of the governance process. The need to conciliate a bill in both chambers in the case of Congress and the introduction of a time lock for Lido reduce the prospect of arbitrary changes and therefore thwart the impulse for governing parties to resort to quick fixes when complex problems would benefit from more deliberative and considered treatment. Of course, on the flip side, such a design can also lead to more gridlock situations in which nothing gets passed.
3) Specialization. Hamilton and Madison, in The Federalist Papers #62, argued:
“It is not possible that an assembly of men called for the most part from pursuits of a private nature, continued in appointment for a short time, and led by no permanent motive to devote the intervals of public occupation to a study of the laws, the affairs, and the comprehensive interests of their country, should, if left wholly to themselves, escape a variety of important errors in the exercise of their legislative trust.” (The Federalist Papers #62)
Senators with longer terms are better positioned to accumulate expertise and human capital relevant to governing. And in fact, a significant responsibility of the Senate is to review and revise matters originating from the House. Members of the House are closer to their electorates and may better reflect popular opinion. For Lido’s dual governance scheme, it is not unreasonable to posit that LDO holders are better at decisions relating to protocol parameters and maintenance while stETH holders are better suited to evaluate proposals from the perspective of Ethereum’s network security.
4) Stability and Predictability. In Federalist #62, Madison also articulated, "No government, any more than an individual, will long be respected without being truly respectable; nor be truly respectable, without possessing a certain portion of order and stability". Bicameralism reduces the fickleness of policy-making, while Lido's dual governance adds to the sense of security for its stakers, which is important as the protocol matures.
Constitutional Engineering & DAO Design
Bicameralism is, of course, not a uniquely American phenomenon as its historical roots can be traced back to classical societies in Greece and Rome. Modern bicameralism originated in the U.K. and is also adopted in many other countries, with variations from country to country on the exact design.
The comparison above is drawn at a rather micro-level, looking specifically at the U.S. Congress and Lido’s dual governance proposal. On the broader level, designing a DAO is not dissimilar from conceiving a constitution. In essence, they are institutional arrangements consisting of systems, processes, and policies for the purpose of coordinating activities efficiently for a common objective. As such, principles in constitutional engineering, a topic that had been much more well-studied, can serve as great reference material for inchoate DAO designs.
One area of comparative constitutional structures is the evaluation of veto gates and veto players. Veto Gate refers to a formal institution that serves as a point in the legislative process where a proposal can be blocked, and Veto Player refers to any person or group who has the capacity to block a proposal.
For instance, the presidential system with a bicameral legislature in the U.S. has three veto gates: presidential veto, the two houses and the Supreme Court. However, the number of veto players is determined bythe dynamics of political parties. Depending on the relative dominance of a party, one veto player can control all three veto gates.
Lido’s dual governance scheme is perhaps the first DAO to attempt to install a veto gate in its institutional design. How well the scheme will serve its designed purpose, nonetheless, is less certain and will hinge on the interactions of veto players. For instance, whether stETH holders will act as a monolith with homogenous interest remains to be seen. As Lido also provides liquid staking on multiple other chains such as Polygon, Solana and Avalanche, the potential inclusion of non-ETH stakers in Lido DAO’s governance will further complicate the dynamic of LDO holders and stakers.
State of DAO - Reality vs. Expectation
While at this relatively early stage we are unable to make claims about what a perfect DAO would look like, we can certainly discern the bad ones. An example of poor DAO governance is the recent Solend drama regarding the attempt of a forceful liquidation2. Solend is a decentralized lending and borrowing protocol on the Solana blockchain. The Solend governance team, in anticipation of the cataclysmic consequence of a potential $100Mn position liquidation, proposed to use the power of smart contract upgrade and force a liquidation through over the counter sales.
While the proposal (known as SLND#1) was passed by an overwhelming majority of SLND token holders, it quickly raised serious concern among the crypto community as it was seen a blatant violation of the decentralization ethos. Faced with harsh criticisms, the Solend team quickly drafted SLND#2 and put it thorugh another vote. Eventually, SLND#2 invalidated the decision of the first proposal within a day. The erratic nature of immature DAO governance harms the credibility of the protocol. In Solend’s case, the non-binding nature of passed proposals in retrospect make the whole purpose of the DAO seemed fatuous: whats the point of DAO voting if its decisions lack any enforceability and it is simply up to the admin team to do make the final call?
Of course, it might be too harsh to pick on one poor example of DeFi protocol governance since DAO is still nascent and everyone is largely experimenting and learning through failures. In Post-Capitalist Society (1993), Peter Drucker, who is considered the father of modern management theory, hypothesized the emergence of an autonomous community organization that fits between the private sector and public sector.
“Every developed country needs an autonomous, self-governing social sector of community organizations. It needs it to provide the needed community services. It needs it above all to provide the bonds of community and to restore active citizenship. Historically community was fate. In the post-capitalist society and polity community has to become commitment.”
Blockchain's core innovation comes down to governance — the new mode for the distribution of trust. DAOs, powered by blockchain, have served as the foundation for many organic formations of communities, and they certainly have the potential to live up to Drucker's vision of spawning a "new center of meaningful citizenship". The path to get there, however, is circuitous and surely comes with perilous detours.
Conclusion - Ambition made to counteract ambition
“Ambition must be made to counteract ambition.” — Federalist Paper No. 51.
The spirit of checks and balances propounded by James Madison is not just a timeless canon for polities, but also a tenet for any populous community organization seeking to balance the interests of diverse stakeholders. It is therefore encouraging to see that Lido DAO is actively incorporating more complex institutional arrangements in its governance process.
At the time of writing, Lido’s Dual Governance scheme has been pushed to next stage for implementation based on the generally positive feedback solicited. For DAO to live up to its vision of a new form of social organization without capture by centralized authority, innovation must materialize on the institutional design level in addition to the technological level. To live up to its full potential, DAO should not shy away from looking into the rich history of constitutional engineering and fish out valuable insights from the wax and wane of political systems.
The Merge refers to the process of transitioning the Ethereum’s consensus mechanism from Proof-of-work to Proof-of-Stake.
For more details on the Solend drama, you can check out this other blog post from my prolific friend at Friday Catalyst.