Dear Seeders,
Welcome to the Weekly Seed
🌿Community Highlights
There’s no denying it; the bear market is upon us. But if we perch on a high ledge and look out over a longer horizon to see where crypto is going, things are looking just plain rosy. Projects are launching every day, and protocols that have been around for a few years are showing us that they are built for the long haul. And then there’s DAOs.
Unlike the bear market of 2017-18, when DAOs were still a novel concept, DAOs today are expanding their contributor lists and use cases. That’s right - DAOs are buidling. But with that building comes certain growing pains, even during a bear market. One of the issues that have beset DAOs, just fiat-space organizations, is less-than-ideal governance participation reflected in low voter engagement - Snapshot is less snappy and a bit through-shot.
🔬Seed’s Research
At an old wooden desk in 1880s England, the writer of Alice in Wonderland and The Jabberwocky was hunched over a manuscript. No, he wasn’t writing about Alice and the Cheshire Cat—that had long been published. He was penning a pamphlet on Parliamentary representation that would form the baseline of thought on liquid democracy.
Liquid democracy, the idea that people can assign their voting power to different representative delegates at any time, is a type of delegated voting. In a broad sense, delegated voting just means that constituents entrust their vote to a representative who can vote or otherwise act on their behalf. Most democratic political systems use some form of delegated voting.
It’s been well-documented that DAOs have a voter participation problem. To moderate an inherent tension of direct democracy, that not everyone reasonably can be expected to vote on everything, there's been a lot of talk about Lewis Caroll’s delegated voting model.
The idea is simple — assign your voting rights to a trusted community member who can vote with your tokens on your behalf. You enjoy the proxy voter participation and the delegate earns social and political capital for participating in governance. All good, right?
But, I’m wary of using delegated voting without proper strategies and safeguards.
Why am I wary of this seemingly best-case-scenario solution to voter apathy? Because we use a form of delegated voting in the U.S., and while there are many advantages, voters are historically terrible at re-allocating their votes when a delegate isn’t serving them well.
For example, the American Congress approval rate is typically around 30%, meaning less than one-third of Americans approve of how Congress members are doing their jobs.
But with that low approval rating, reelection rates are startlingly high.
Reelection rates in the U.S. House of Representatives (two-year terms, representation based on state population) are generally over 90%.
U.S. House reelection rates haven’t dipped below 80% since 1964. Chart from OpenSecrets.
The purple bars in the chart below show the percentage of incumbents re-elected each year. Reelection rates haven’t dipped below 80% in the last 56 years.
And the Senate (six-year terms, two senators per state) is startlingly high, too. It’s typically above or around 80%, with a few years around 60%. But never 50% or below.
Picture: U.S. Senate Reelection Rates between 1964 and 2020. Chart from OpenSecrets.
And for those who manage to unseat the incumbents? They need to put in a lot of effort and money to do so. Just look at how much more money challengers had to spend in the 2019-2020 Senate election cycle to unseat an incumbent: $73 million dollars. But for an incumbent to win, they only need to spend $19 million.
Picture: To defeat an incumbent, challengers had to spend an average of $73 million dollars, compared to incumbents spending $19 million. Chart from OpenSecrets.
The takeaway: once someone is a delegate, it’s extremely difficult to get them out of that position, even if they have low approval ratings.
Even though the trends of the United States are not representative of how the entire DAO space will operate, we should be looking at how existing representative democracies function if we are going to build blockchain-based representative democracies in DAOs.
Given the inability of U.S. voters to unseat delegates, even when they don’t like what those delegates are doing for them, I foresee DAOs having the same issue.
But I believe we can make delegated voting better. There are a few areas we can explore to prevent some unwanted results:
Preventing career delegates: term limits
Preventing a single delegate from getting too many tokens: percentage maximums and delegate minimums
Preventing delegate bribing or vote-skipping: slashing and “socialware”
Preventing parties and factions: forking and quitting
Delegate Term Limits
Term limits place a cap on the amount of time, or number of terms, a delegate can serve. This prevents career politicians—people who turn politics into a full-time job and never venture to work in the “real” world—from taking over every office in the capital.
U.S. voters largely believe term limits should be implemented in Congress. One survey found that 82% of voters would be in favor of term limits.
Here, you get in an odd situation. The people who have the power to propose term limits are precisely the ones who would not want them. How, then, can you possibly remove ineffective incumbents?
In DAOs with delegated voting, term limits should be put in place before delegates are chosen.
You may be wondering: But why require term limits when voters can simply re-delegate their tokens whenever they see fit?
That requires active DAO members who are keeping up with how their delegates are voting.
If DAO members aren’t voting widely now, how will they be keeping up with people voting for them?
Imagine this situation.
A charismatic-leader type of delegate emerges. The delegate often votes against the grain, and they become popular because of this. People delegate their tokens to them because the person resonates with them in some way. All is well: it’s important to have delegates who represent the will of the people. But then that delegate’s term limit ends. The charismatic-leader-delegate has to step down. But, since so many people supported them, there must be another delegate who had those same views coming down the pipeline, right? Surely someone will take up the mantle.
But this is when the effects of the charismatic leader begin to wear off. No one quite fills the old delegate’s shoes. People aren’t interested in what the new delegate is doing. So, they slowly re-delegate their tokens.
If there wasn’t a term limit, the charismatic leader would continue exercising power even if their “power” was simply their charisma, not the ideas that current voters want. And if the delegate was allowed to stay, they may begin to dissociate with the idea and simply focus on maintaining the following they’ve built. Think of any charismatic leader, and how “pure” their ideas remain over time.
Term limits on delegated voting protect against the charismatic leader issue.
Ideally, delegates would operate entirely in the open. Meetings and conversations would be made public. Voting choices would be shared—maybe delegates would write a couple sentences explaining their choice for each vote. This would help transfer knowledge to new delegates who must step up in their place.
Delegates may not want term limits. And people may worry over loss of knowledge. But for delegated voting to work, I believe term limits are necessary.
Preventing Delegate Whales
With purely liquid democracy, it’s possible that delegates could get so much power that they control a massive percentage of tokens in the network.
What if a single delegate controls 20, 30, 50, 70% of the vote? What could they use their power to do, especially if the token holders are checked out?
The U.S. breaks delegates up geographically, so they would never be able to gain too much of the vote. But in liquid democracy and DAOs, there’s no stopping someone from getting an outsize portion of total votes under their control. And when that delegate controls a huge piece of the voting system, well, we’re screwed.
Some solutions for this would be to cap the percentage of tokens that can be delegated to a representative. Otherwise, delegates may obtain a proportion that becomes so high they can 51% attack the protocol or organization.
Alternatively, a DAO could set up a delegate “minimum.” Maybe there must be at least 50 actively voting delegates. There would still be a scenario of some delegates having much more voting power than others, but because there must be 50 delegates voting to reach quorum (the minimum number of votes or participants required to enact a vote or conduct business), there would be some degree of safeguard.
Percentage maximums and delegate minimums may feel counter to the ethos of decentralization and self sovereignty, because it is placing constraints on the pure, unchecked “will” of the people. But if a delegate received so many token delegations that they could use their super-mega-whale status to abuse their power, then safeguards are needed.
Preventing Delegate Bribing or Vote-Skipping
Another unfortunate scenario coming out of delegated voting is bribing.
Delegates could potentially “win” voters by bribing them, such as with financial promises. You could imagine a situation where a delegate says, “Delegate your votes to me, and I’ll make sure a proposal goes through that gives you a good role in the DAO next season.” Or, “I’ll pay you X amount if you delegate your tokens to me.”
With enough money, the system could be abused.
The value of bribery could be mitigated by delegate percentage caps, as mentioned above. But even then, multiple delegates could collude and attack the DAO with their voting power.
How do we prevent this?
Firstly, there could be token slashing —which is when tokens are confiscated because of an unwanted action —for delegates who use bribes to gain power. The slashing would prevent people from choosing to delegate to someone who is bribing them, because they could lose their own tokens. But delegates would also shy away from bribing, knowing their voting power could evaporate in an instant.
How would slashing be implemented, technically? I predict it will be difficult. Because there’s no trustless way to determine if someone is bribing (as far as I know), then there would need to be human intervention. Someone would need to manually block the delegated wallet from voting, which is against the ethos of a trustless system.
How, then, could vote bribing be prevented?
Social pressures could be used. In systems trying to operate in a trustless manner on the outside, trust-based social pressures on the inside tend to be the way to deal with squishier human issues that can’t be codified. (Coined “socialware” in this Orca piece.)
In other words, there may need to be some “socialware” involved in order for delegate token slashing to work.
For example, if a delegate was performing suspicious behavior, members of an operations guild or governance committee could ask people to remove their delegated votes from that particular delegate. This would be a form of social pressure—no delegate would fully be removed this way, because they’d still have their own tokens to vote with and maybe those of the people who did not want to apply pressure. But, given that social pressure does run many DAOs today, it could make an impact when needed.
Delegates also may miss votes. For example, many of the top delegates for ENS DAO have voted in half or fewer of the proposals since the DAO started. In a blockchain, if validators get behind, they may be slashed. But, would a DAO really want to slash all of those tokens? Is it as critical as a validator in a blockchain?
Again, we’d likely need some kind of socialware here. How many votes is too many to miss? Does that number change as the DAO matures? Again, slashing could be used, but it would likely need to be in the realm of humans, not code.
I’d like to point out that there’s no concrete way for U.S. Senators or Representatives to get punished if they’re using sketchy vote-gaining tactics. Sure, there was a series of expulsions of Senators around the Civil War, but other than that, Senators and Representatives don’t get “slashed” or “banned” because they gave out a few favors to supporters along the way. They would get punished or removed from office if they violated the law or accepted a bribe (depending on what state you’re in). But politics in America is all about making promises and logrolling all sorts of stuff into bills that the public might never even see.
I’m excited about the better forms of governance we’re building in DAOs. But, we’re treading into uncertain territory here.
So, would preventative measures against bribing be groundbreaking? Maybe. Would it be possible to implement the necessary socialware to make it possible? Maybe. Are there more questions than answers right now? Definitely.
Preventing Parties and Factions
Once a DAO gets large enough, it’s easy to imagine factions forming. Just like a modern state, when large groups of people work together, factions naturally form.
The oft-quoted Federalist Paper #10 by James Madison warns against the formation of factions, but says that “the latent causes of faction are thus sown in the nature of man.” In other words, we are naturally going to form tribes and groups based on our opinions, ideologies, and identities.
Groups of like-minded individuals will naturally form — there’s no way around it. But what about when those groups start forming parties that lobby, fight, conceal, and hurt? That’s not the world we’re trying to build in DAOs.
One option for factions is forking, a term typically used to describe blockchains, but that can be used in the DAO space as well. Unlike modern states, a group of like-minded individuals can relatively easily split into their own DAO if they’re not happy with the way their own DAO is run.
Sometimes, forking can even be considered a growth strategy. As it gets more and more difficult to coordinate large groups of people within a DAO, the best strategy might simply be to start a new DAO out of the old one. Maybe if a small group of delegates were, naturally, going to start a faction, they would just fork the DAO and start their own.
But forking does not come with the social capital, contributors, resources, treasury, token, and voting processes. Forking means you basically lose it all for the sake of building a DAO in the way you want it.
I see forking as a philosophical alternative, but not necessarily a concrete, actionable one that will be used often. Blockchains can fork, but can DAOs, which are just made up of humans trying to coordinate, really fork?
Another alternative to forking is rage quitting. This is when a contributor forfeits their stake in the DAO, essentially giving up their membership tokens in exchange for that token’s value. Rage quitting could look like selling your tokens on an exchange, or it could look like forfeiting your tokens or NFTs back to the DAO in exchange for your monetary stake in the DAO via a Rage Quit function, like DAOhaus.
The problem with rage quitting is that it doesn’t take into account the human aspects of DAO work. What if you have a role and people are depending on you? What if that’s your main source of income? What if you don’t have a job lined up next? There are so many factors that go into the simple idea of “rage quitting,” that I don’t know if this is a good solution for preventing factions and parties.
But I do think forking and rage quitting should both still be options, even if they might not be employed very often. Like DAOhaus said, rage quitting is a right. My reservation is this: I wonder if people will truly give up everything they have in their DAO to exercise that right.
DAOs as Grand Experiment
DAOs are the grandest voting experiment we’ve had in centuries and we can use them to test hypotheses that we can’t in fiat-space politics. There’s a big problem with states: you can’t experiment. The stakes are far too high. So, even though there are frequent pleas for eliminating the electoral college, or using ranked-choice voting to prevent extremism in parties, or using a multi-party system in the United States, the reality is that most of those ideas will never be tested because the cost of experimentation is too high.
In DAOs, we can actually experiment. We can actually try new governance systems with a small group of people without having to overhaul a state’s entire system.
And we have that opportunity with delegated voting. And, we have the time to do so. The sky won’t fall on DAOs if we take another few months to iron out details of delegated voting that will pay off years and centuries in the future. Bear markets are for building, and it looks like we may have time on our hands that we can put to a great purpose.
We don’t need to use a basic system without putting thought and planning into it ahead of time. We can create a delegated voting system that truly works for everyone: whether it’s Lewis Carroll’s concept of liquid democracy, or something that hasn’t been thought of yet. As the tech advances and voting experiments are put to the test, we might come up with something just as otherworldly as Carroll’s Alice in Wonderland and as advanced as his voting schemes.
🧮Platform status
Everything is now IP protected. The demo is done by a person from the team using the software. While betatesting is autonomous, currently at IP blocked so those who want to participate need to contact us.
From July 4, 12/13 startups who give their IP will be able to start testing. Ditto for the investing side. Currently there are three demo campaigns running in testnet (no difference with mainnet).
🎬Action
Don't stand still join the community, stay up to date with the latest news and discover opportunities
Subscribe the newsletter
Get in touch with the team