Delegation of Authority - PreMIP discussion

Maker governance is still in a beautiful early phase - call it springtime. One defining feature of such a early phase is that there is little separation between roles in the organization nor much in the way of hierarchy. Maker is however outgrowing this phase at a rapid pace and there could be time for a move towards a slightly more conventional structure. The amount of information that needs to be processed is steadily increasing and putting Maker governance under stress.


How did we end up in this situation? Largely because of two things: first of all because decentralized organizations have never really done much delegation and second because on a strategic level the Maker Foundation provided the long term guidance needed. As a result the Maker community was stuck with what remained, which has largely been day-to-day operational issues. The net result is the governance we have today.

When stressed organizations have two options: either automate or delegate. The first steps towards automation are already underway in the form of MIP27, I will in this article argue we start doing the same with regards to delegation.

My motivation behind this is simple - the need to scale. With the present state of affairs Maker is not able to onboard more than roughly two types of collateral per month, which is far below what we need in order to dethrone Tether. Simply taking the present process and multiplying it will not be sufficient as Maker governance voting is already an expensive clickfest and multiplying this x10 or x20 is simply not workable. Therefore we need to both delegate and automate.

A spoonful of delegation will remove the issue of voting for proposals that sail through with 95%-100% of the vote anyway, an easy picking with regards reducing the overhead cost of Maker governance. It will also empower Domain Facilitators and so facilitate future automation in their respective domains.


I hereby propose MIP_delegation_of_authority for a pre-MIP discussion. The MIP empowers Domain Facilitators to send specific issues straight to Executive Vote, bypassing the Governance Poll. Which issues? The typical issues that get close to 100% of the vote anyway. I plan to detail this with Domain Facilitator input if this pre-MIP passes. One example are Oracles where I have yet to remember an oracle proposal that did not go through. Another is Collateral Onboarding where the bad apples can be separated out by anyone that did not invest in OneCoin. I want to delegate my voting power to the Risk Team on this issue, I fully trust them to do the right thing. At the moment all the November 30 polls fall into this category where the onboarding of renBTC, 6s Capital and Uniswap as well as whitelisting of Yearn are all looking to be approved with respectively 99.85%, 100%, 98.7% and 100% of the vote. Let’s streamline this!

The MIP_delegation_of_authority is not imperative, meaning Domain Facilitators are free to subject decisions to a Governance Poll if they feel the decision is important or potentially divisive.

Feedback appreciated.

Do you think this is a good idea?
  • Yes
  • No
  • Abstain
  • Other - please explain

0 voters

6 Likes

I would say that non contentious proposal (like whitelisting for oracles) don’t take too much governance intensity. We should streamline them just because adding administrative work doesn’t make sense.

The best example for “useless” governance work in my mind is setting SF for stablecoins with two failed polls. I think (may be wrong) that no result would have changed the world. The monetary and risk impact was low and temporary anyway. Thinking on how to improve governance like your post can change stuff significantly and impact Maker for the years to come. Just like the decision to onboard RWA or the decision (not taken) to decrease the collateral value per DAI. We probably should have spent more time on those strategic items.

For RWA (just an example, because I’m struggling with this one), I try to put minor decision to the community like here. Participation is low as I expect people have better to do than go deep in the implementation. Nevertheless, sometimes you get quite interesting feedbacks.

The plan is to have a gov poll of all these micro-decisions (to avoid having x governance polls). But participation is low and having something set in stone is painful as well. We also might need to change stuff the next day because real life is not a clean theory.

Transparency on what the domain teams is working is important so Governance can check and yell if needed. But having everything to be discussed with all governance is not efficient.

It’s somewhat moving from a Command and Control model to a Mission Command one.

2 Likes

I am fully behind the idea of getting rid of unnecessary bureaucracy. DeFi moves fast and Maker DAO needs to be nimble to keep up which at the moment its hands are tied a bit.

2 Likes

I think this is on the right path. You’re correct - we’re not nimble enough, and this type of delegation could help us fix that. It shouldn’t sacrifice too much decentralization either with the chief 2.0 contract since we won’t need to bundle all of the executives.

One additional thought, not sure if this is the right place for it, but I was wondering if term limits for domain facilitators be included in this MIP? This would give them a commitment timeline, and would also be easier for governance to express a different direction by electing a new domain facilitator when their term is over.

I’ve been one of the main complainers about the peg. I have to admit, it is pretty annoying. I think this meme explains the situation fairly accurately (me on the left):


Like it or not, I’m never going to change! I will continue to be the annoying peg pest until it improves. I firmly believe it is the most important problem we’re up against today. If you felt the pain I felt in early 2019 when we were at 0.96, you might think differently, @SebVentures. Giving the facilitators more power to send votes straight to executives should help with future peg problems - I’m in.

2 Likes

I fully support that. Maybe I didn’t express myself well, I was speaking about small details (SF on USDC-A at 4% or 0% when most vault are near 100% CR and any fees is probably just an accounting trick). The PSM might solve the issue but much discussion is needed to know exactly on how we will handle the DAI surplus/demand imbalance in the future. Using stablecoins as today? Expecting RWA solving the situation? How to avoid going back to DAI below $1.0 and having RWA where SF can’t be changed quickly?

1 Like

The problem with term limits long term is that it incentivizes short term thinking of mandated actors in order to implement their authority. If you know you’re not going to be in a position long term you have more incentive to think about what can be done sooner rather than seeing the forest through the trees.

“A society grows great when old men plant trees in whose shade they shall never sit.” - Greek Proverb

2 Likes

Good point. I guess I was thinking that term limits might make them work even harder to compete with others to get re-elected. MKR holders have most of the burden of being long term focused, and would they should vote for facilitators with that in mind.

1 Like

IMO, as a Yankee, I believe once MAKER is officially label a “commodity” we can do what you are suggesting–for know we need to include the entire community to build the protocol and decentralize it via the governance component of the DAO, as much s possible. Otherwise, you’re only pin-pointing the Risk to a certain numbers of actors as oppose to All. Hence, why I opted for “no”.

We can say that Delegating voting is a more legitimate approach–but we can also say MKR voters allowing DoA is the same. I just don’t know how the SEC would interpret it, IMO. Need more clarity, from these Gov Agencies.

All for one, one for All. BTW, was this post inspired by the Fred Erhsam op-ed?
Or, just something you personally have been thinking about?

2 Likes

@ElProgreso thank you for the link - that is a good article, I hope other people read it as well. The problem with government minimalization is not the theory, it is how to implement the practicalities. I hope we get there one day. Moving towards a protocol is the way to scale in my opinion.

1 Like

FYI: In my experience you can’t delegate Authority without also ‘assigning responsibility’. In too many organizations authority to get something done DOES NOT lay with a single authority, it often requires coordination of multiple authorities to complete a single responsibility.

Just keep this in mind with this MIP.

I completely agree with this observation. Far too often there is this idea that someone with power (i.e. authority) should never be allowed to hold this for any length of time out of concern for abuse of power, forms of nepotism. While these are issues, given that governance is the one to carry out the final stamps of approval on many key actions, I don’t see this as a problem.

Today far too many organizations don’t even have long term planning groups with people who themselves come from, not just one but ideally, more than one authority/power roles as these people have the best view of what the organization does because they have done most of it. These groups tend to act in advisory capacities to corporate boards or high level management.

Now if this is going to be a who can win at politics contest then perhaps there could be reason to have ‘term limits’ but honestly if there are clear ways an ‘elected’ authority can be removed there shouldn’t be too much problem. The real issue here will be conflict of interest issues and how to be transparent about those as they crop up.

I have argued Domain Facilitators should just do the micro work defining a proposal (say changes to SF, DC, collateral onboarding) and put their report up for review for a reasonable length of time for community to review and provide feedback. Then just on-chain poll yes-no the resultant DF package. After this process completes governance should just yeah/nay the executive. This whole governance micromanaging everything Maker in the end will become wholly impractical and this is why some form of heirarchy is desirable.

The debate won’t be around whether to delegate authority and responsibility, but how to structure it.

Lastly, I want people to carefully consider robustness against decentralization in this authority/responsbility model. In my own analysis on this I conclude that decentralized organizations will be less efficient in resource allocation (people) than centralized organizations because it will take at least 2x amount of people and effort to get that 1 level personel redundancy. Quality assurance and fall back costs more in decentralized vs. centralized organizations to make sure that authorities are acting consistently and have checks/balances. 24/7/365 organizations will need to be at least 3 people deep in key positions (often this is really 5) with trainees and retirees simply due to nature of 24/7365 operations and staffing rotation (vacation, sick, changing jobs, training considerations/time etc.).

3 Likes

Completely agree. Glad we’re discussing these challenges now. Another topic that I think will require its own thread (many) is expanding past the dunbar number without eventually collapsing. Building out a decentralized organizational framework is definitely one of the largest issues we will face and I look forward to seeing how that evolves.

1 Like

Thanks for the thoughts Planet X! Slightly off topic from the purpose of the post, but…

Something to consider is how we define the value of a decentralized org. Centralized organization are measured through efficiency, and I argue that’s the wrong value to measure to decentralized organizations. C-orgs will always win at efficiency. While we should try to be as efficient as possible, IMHO, we should optimize the opportunity for some of the values of distributed networks: more creative solutions (from the variety of people in the network), resilience through redundancy, and personal/team autonomy (with the right structure).

5 Likes

Great points Amy. I’m glad you put a spotlight on this. The whole reason crypto exists today is because we saw that the current centralized system is broken and we need more sustainability, resiliency, and accountability. If we lose some efficiency in the process, it will be worth it.

3 Likes

Spoken like a true community leader @amyjung ! In fact one of the most exciting VCs to launch recently IMO, is The Community Fund led by Lolita Taub. Their goal is to invest in early stage “community centric companies”. Assuming they will be community driven–I got to believe some of their portfolio community projects will follow an orthodox style of decentralization–by the community, for the community. Hence, I doubt this VC team believes decentralized orgs will be less efficient, IMO.

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.