Saturn DAO: Democracy that owns a decentralized exchange. Voting rules discussion


#1

This post is a placeholder for the current voting rules and latest proposals of the Saturn DAO. Later on this topic shall be forked into chain-specific DAO discussions. However, at least this year all DAOs will launch with identical set of rules.

Open questions:

  • Quadratic voting was proposed by Glen Weyl as the economically optimal method of corporate governance, and could thus be applicable to DAO governance as well. The problem with application of quadratic voting to governing a DAO is the lack of one human - one identity concept. An individual could potentially own multiple private keys that have control of voting tokens. As such, we will initially roll the DAO our with flat voting: one token = one vote. I would be curious to hear a discussion of other voting mechanisms.
  • What is the appropriate threshold? Obviously, we could say if 50% + 1 vote is in favor, then the change is approved. However, we do need to take laziness into account - will people actually show up and vote? Maybe a test vote could provide us with some estimates.

If you have more questions - ask them and let’s keep the discussion going.


What is SATURN token?
#2

I feel like decentralization is a goal, and a journey. It is not only binary, yes or no, black or white. Two nodes running a common system is decentralized, 100 nodes running a system is more so.

The idea can be applied to token holders as well.

The problem is this idea of economic disadvantage. That it’s not “fair” that some people have more than others. The idea that nature is somehow seeking to make your life “fair” is preposterous and dangerous.

The only way to make Quadratic voting feasible, as far as I see, is to introduce KYC, which would nullify an advantage Saturn has over it’s competition.


#3

I agree with your statement. Even with KYC it would still really not work as good as a good old centralized system - so why even bother building it?

We must focus on decentralization, explore the possibilities to the fullest and make no unnecessary compromises.

But here’s a twist on quadratic voting that might work (edit: it doesn’t :laughing:):

Quadratic Voting as presented by Weil

Let’s say there is a vote going on to make a decision X. If you want to put 1 vote towards it, you “pay” one token. If you want to put two votes in, you gotta “pay” 4 tokens. If you want to put three votes in you gotta “pay” 9 tokens, etc.

This only works under the KYC assumption, because if a party splits their nine tokens into nine wallets they can make 9 votes instead of three.

Inverse QV

What if we had KYC not on users, but on votes?

Let’s say there is a vote to make a decision X with two outcomes - yes and no.

The voting power can be a complex function of coins accumulated for the vote, instead of a simple sum. Consider this application of quadratic voting:

starting state

Yes: 0
No: 0

vote 4 tokens for yes

Yes: 2
No: 0

vote 9 tokens for no

Yes: 2
No: 3

vote 5 tokens for yes

Now, here are the two options. Either we do {Yes: 3, No: 3}, and then it is effectively the same thing as doing a simple 1 token = 1 vote. Or we do {Yes: (2 + sqrt(5)), No: 3}, in which case the system can again be gamed by splitting the tokens among many wallets.

In short, it seems that one token = one vote is the optimal way to govern a decentralized system.

My homework is to read this paper by Vitalik Buterin and Glen Weyl.


#4

One token one vote sounds fine for now. If the majority of tokens vote against the majority of the community on a super serious issue then we can always fork which would not be good for saturn. That should be incentive enough to keep big holders from abusing their vote. Also if the community wants to protect itself from “malicious votes” they can always buy more coins. A large community of ppl holding only a few coins each can out vote even the biggest whale because unlike many other coins the supply is capped, with no new tokens created they will all eventually end up in the hands of ppl who will use them for governance. Also if the price skyrockets quadratic voting could be added later to give smaller holders more of a voice but even still ppl could form collectives to push votes one way or the other. Who decides what the votes will be on? Can anybody initiate a vote at any time? There is no perfect solution but bringing in ppl who really want to see saturn do well and believe in decentralization early on can prevent most issues later


#5

I agree with your sentiments. Indeed, at this point we should focus on launching the simplest upgradable DAO system. Then we can debate whether improvements are needed. If they are, we can adopt a new system.

Running a DAO is an iterative process. It’s a thrilling journey! I am very excited for the moment when independent communities on different chains choose different paths for their governance.

Who decides what the votes will be on?

There will be a fixed amount of things that can be voted on. At a minimum,

  1. Pay out money to an address. This is how the dividends will work. A snapshot block will be communicated in advance. Somebody will create a smart contract with a certain address => reward allocation according to the snapshot. Token holders will validate the correctness of the smart contract. If the smart contract is correctly created, then the DAO will vote to transfer funds to it, and people will withdraw tokens from it via a dApp (similar to our HODL dApp).
  2. Change fees. This is pretty obvious. Initially there will be flat fees for all tokens, but at a later date the DAO smart contract can be upgraded to include more granular controls.
  3. Upgrade the DAO contract. DAO upgrades need to be voted in.
  4. Upgrade order book contract. The order books (that host the liquidity and perform the actual exchange) can be upgraded as well.

Did I miss anything crucially important?

Can anybody initiate a vote at any time?

Yes. However, initiating a vote, and voting itself, must include token lockup. This acts as a rate limiting mechanism against vote spam. There are only so many tokens - which means there can only be so many votes at the same time.


#6

Initially, upon researching voting mechanisms - the two most recognized systems are:

  1. Plutocracy: One man = One vote. *
    correction 7/11/18: One coin = One vote establishes plutocracy.
    One man = One vote is ideal of democracies, sometimes now can be expressed as One wallet = One vote or One KYC…

  2. Reputation-based: Higher weights are given to higher scoring metrics as quantified by the society values, such as Bitnation’s proof-of-nomic system (of automated and human scores, ratings, collaborative filtering, and digital representations of sentiment, opinion and thought) which attaches a reputation value to every smart contract executed, and evolves the most utilized set of laws to the surface.

Both entail differing sets of challenges, and by no means should we be content to accept their limitations by opting to enforce behaviour away from their pitfalls. Plutocracies are effective in equalization of voices, but must enforce KYC policies to ensure no repeated voting occurs. Reputation, even despite the difficulty and time intensity required to amass, could still theoretically be “sold” to the highest bidders and we wind up no farther than the grasp of economic power in its discretion. Nonetheless, we should be aware of both the progress that has been made, but also the effect that these thought paradigms will have on any creativization which takes these ideas further either in part or as a part of.

An idea I have been working on builds on the liquid democracy framework with voting weights and assignments given towards a holocratic representation structure. It puts every mechanism into play, and leaves room for new mechanisms to be introduced as well. I call it, a

“Liquid Holocracy”:
We start with the plutocratic one man one vote, and enable transferability of voting right into a representative holon that is most closely associated with the voter’s interest profile. Interest profiles are based on an impact assessment or by closest association ie. let voting be assigned to the holon that is most relatable to the person by stake or value assessment (this needs much more explanation).

Then, we must also go into the rules forming the Holocracy in order for this optional delegation to make sense. In this case, a person can opt to also be a contributing member of a Holon, and also needs to be qualified to do so (reputation score). Holons are elevated positions capable of vote representation and defined as specialist groups serving a specific function for the entire whole. Responsibility is given by reputation metrics, however it does not filter through directly to voting power as you are either in or out of a Holon position and amassing reputation only ensures your ability to participate as a contributing specialist to the collective needs. Examples of Holons for Saturn Network include:

  • Governance Holon (carries out administration, evaluation, oversees voting process and maintains DAO constitutions)
  • Development Holon (dev teams, product life cycle etc.)
  • Marketing Holon
  • Knowledge Management Holon
  • Finance & Treasury Holon

The concept is a work in progress, as should be all matters of proposition and exploration into the field of DAO governance. This is what is most exciting about it, encouraging all critiques, comments and more; should it eventually lead to one day solving the problems we face with governments around the world as a result of archaic parlimentary systems and the true cost of centralization to social advancement as a whole…

KR