MakerDao needs a Treasury to provide Compensation to its Workers post foundation (focus on workers for now)

This first section is a bit of a preamble

MakerDAO (Decentralized Autonomous Organization) as it stands today relies heavily on a centralized organization called the Maker Foundation. In order for MakerDAO to realize itself the organization must create transparent and decentralized functions to support its existence, necessary components include worker and treasury management. Intuitively it does not make sense to expect the foundation to create these frameworks. To create decentralization we must design tools and frameworks that render the foundation irrelevant (but maybe hope they build them ;)).

Right now we have three or four major teams and a couple proposed. Risk, Governance facilitator, technical and oracle team. A validation team has been proposed to offer reports, support and dissent for other teams work and the system as a whole. And a collateral coordinator may also be a necessary position (more on that will come soonish). The ultimate goal is to create a framework to review, accept, compensate and validate these teams and future ones, outside of any support from the maker foundation.

Each teams mandate was ratified by a governance poll. Rich Brown as interim governance facilitator, Nik k as interim oracle team and Cryrus ( and Vishesh + others?) as interim risk team. In order to bootstrap these roles the Maker foundation compensates and supports the individuals work. Considering MCD as the real birth of the financial aspect of the protocol, the state of decentralized management accompanying it is disproportionately underdeveloped.

The (majority) members of these teams work for the foundation, so while formal mandates are a positive step towards transparency and decentralization, little functional difference emerges from a foundation control standpoint (not that they are arguing otherwise). The future of the DAO depends on our ability to evolve beyond these concerns. While more pressing issues might exist, picking up the slack regarding worker compensation-relations confers more legitimacy as a DAO and resilience to regulation.

What would happen if the foundation and its operations experienced extreme disruption from some authority. What would happen to the protocol? (I know its on ethereum and whatnot, but what happens if all these business, marketing and technical development teams paid by the foundation all suddenly ceased operation?) The recent announcement of the Dai Foundation was a little hard to understand at first, but probably a good move covering against malicious actors attacking the protocol through symbol usage.

Here are some of the questions I see as necessary to formulate a resilient decentralized worker-operations framework.



  • Who does the recruiting?
    • Could existing workers get bonuses for successfully recruiting somehow that stays employed for some time
  • An advantage is that people who work in their industry (weather, risk, legal, technical, logistics) likely know others that have adequate skills.
    • Some social bias I guess
  • Reduces voter fatigue if current team heads facilitate the recruitment
    • Reduces control mth have on the individual worker

To avoid social bias in recruitment , an application process could allow interested and qualified parties to apply. Fairly basic/normal list of qualifications would be applied. Would we want people to apply as teams or individuals applying to pre existing teams? I’m thinking it would make sense if only team heads applied

Thinking about how an application system would work with our current voting system (considering the foundation has over 100 employees I believe) manifests a headache. Personally I see recruitment by domain teams as the most sense. Their expertise and social networks likely have access to the individuals we would want. A finders fee could be paid to a risk team or individual that successfully recruits a new member (once that new member starts making contributions).

In a recruitment scenario, the heads of each team however should still be voted in by MTH. Just like we have now. If a head decides to leave they can propose a replacement (likely from within the team) or people of maker can reach into their circles to find someone suitable.


  • What is fair compensation for each team?

    • 10 months ago I wrote about this issue after a governance meeting . In that post I considered more competitive incentives. Rune commented proposing a framework more similar to academia. Comparing working for makerdao as a public service he also mentioned that looking at how researchers at universities are compensated as a potential model. Interesting (almost philosophical) question about how we want to compensate in general
  • Would individual teams be better compensated for “better” performance? What is better performance?

  • How to negotiate terms and pay?

    • For the domain team member compensation would negotiated with the team head. If a workers work is short term maybe be better off as a lump sum.Potentially different in different situations
    • For the Domain team head I think we would need a bit more sophisticated incentive structure.
      • Maybe a monthly salary with yearly raises
      • “Stability” bonus for keeping things “stable”
      • “Growth” bonus for demonstrated enhancement of the protocol
  • Part of compensation paid in mkr, giving workers more voting rights over time and another monetary incentive?

    • Domain heads will need to specify an address for compensation, not sure how changing the address could occur (in the case of compromised keys)
  • How do maker holders hold workers accountable? Who conducts the ‘performance’ reviews and how?

    • Domain team head can cut out work of a team member and revoke whatever access they previously had
    • I’m thinking we need a procedure for MTH to call for a review of a domain team head. Maybe this would occur after some botched operations or something else.
      • If we have an operational treasury I think MTH would just vote to cancel the contract, thus ending payment from the treasury


Here are the questions I don’t have solutions. All I think is that we really need to start specifying how Maker will control its finances post foundation. This is a prerequisite for compensation of domain teams.

  • When are we going to seriously consider a treasury under the control of maker holders?
    • This is obviously imperative for compensating various teams and viability of the DAO in general. How does the DAO have budget awareness?
  • Who monitors expenses, assets and liabilities?
    • Will we try to keep them private or public?
  • How do token holders ensure an effective strategy for employing revenue?
    • Do we need to get some accountants?
  • Why shouldn’t the DAO take out safe CDPs for financing?
    • Should we invest in tokens and maintain cdps of various collateral types?

What sort of smart contract capability is necessary for these ideas. Development has to start at some point. What best ensures the fininacle future of the DAO?

In general I suspect MTH will allocate budgets to domain heads who determine how to best use those funds. A report on how the funds were used should happen every so often.


Unforeseen circumstances can reduce or eliminate an individual’s ability to do work for Maker. How should governance structure agreements and prepare for these eventualities? This is a broader question that’s definitely being discussed in other contexts, but especially pertains to managing people since they can be kinda wild.


How do we settle disputes? What will be the first big governance split?

Aragon is working on their decentralized court system based on various economic incentives. Anyone know of other possibilities? Heres an except from the github:

Aragon Court handles subjective disputes that cannot be solved by smart contracts. For this, it employs jurors that need to stake a token to the Court which allows them to get drafted to adjudicate disputes, that can earn them fees. The more tokens a juror has activated, the higher the chance to get drafted and earn more fees.

"Aragon Court attempts to find what the subjective truth is with a Schelling game. Jurors are asked to vote on the ruling that they think their fellow jurors are more likely to vote on. To incentivize consensus, jurors that don’t vote on the consensus ruling have some tokens slashed. Jurors that vote with the consensus ruling are rewarded with ruling fees and juror tokens from the jurors that voted for a minority ruling.

A design goal of the mechanism is to require very few jurors to adjudicate a dispute and produce a ruling. A small number of jurors is adjudicated by default to a dispute, and their ruling can be appealed in multiple rounds of appeals."

It unclear whether Maker needs this sort of mechanism. I haven’t heard much talk of it in the community, but I know the team is well aware of Aragon. It doesn’t hurt to bring this up again. Disputes occur, and some will likely challenge the legitimacy of an executive outcome at some point. How will we deal with those disputes externally from the voting protocol.

I’ll leave it at that for now. Sorrry if I was roundabout. Basically all I am doing is pointing out that to be a DAO we need to control the finances and obligations to workers that power the protocol. Implicit in this is a recognition that translating between various national communities is essential in creating a more diverse and skilled pool of talent.


This is a pretty deep topic.

I think their are two first steps that can be made.

  1. Actually creating a treasury (even if it is used by the Maker Foundation)
  2. Consider creating DAO treasurer position.

When it comes to positions a few DAOs I am aware of are at least 2 - 3 people deep in them and most real decentralized companies have a primary position, then an assistant, then kind of an assistant in training. Often times these DAOs are required to run 24/7 and have staff available at all times so they can tend to get a bit deeper in these positions. Even so if a person gets sick or can’t continue as the primary head of a company position then quite often there assistant stands in as an interm person to handle the primary responsibilities and authorities associated with the position.

In the loosest sense the DAO still has well defined positions - and they do elect or have by appointment (depends on how the DAO operates in terms of governance) people who assume those responsibilities and powers even if it is temporary. The idea of a community running all the details of a CEO, CIO, CTO, etc. doesn’t really work that well in practice, but having a governance structure where people can line up in position (as in training, assistant(s), etc.) provides for a model where there is a well defined organizational structure with authorities and responsibities but that anyone within the DAO can move into any position.

I’ll have to look for examples because I know I was researching this a long time ago and found a few DAO type companies even then. I remember reading that these types of companies were well positioned to compete because of the benefits of the DAO model vs. the typical centralized corporate model. BUT there were significant legal as well as operational issues that needed to be hurdled before such organizations were to become ubiquitious. The DAO model was good for a product that did not have to have manufacturing facilities - hence perfect for software, or for charitable activies, etc. Think tanks, whatever. Any business type that required real facilities vs. virtual would likely face more challenges than the others discussed above.

All in all great post. We probably have to break this down into bite sized reasonable steps but the idea of defining those steps as well as trying to imagine an end goal target to achieve would be useful. I will have to mull on this a bit more and see if I can dig up my old research on the topic.


Great post. I absolutely agree on the goal of creating a framework that is self managing and has budget awareness.

Regarding budget spend for building improvements…I could see such requests being vetted either through this forum or a more formal manner e.g. EIPs where timelines, resources and costs are transparently agreed upon just like technical aspects of formal verification, and development could be based upon integration with a service like Sablier to ensure iterative build and delivery.

Further hypothesising about a future state, I would expect some revenue generating groups to form - for example trading desks that run bots, and similarly, cost centres devoted to server upkeep or website management. Would a treasury function sit across these groups or could there be a smart contract relationship between the now independent internal/external groups due to mutual interests…perhaps the treasurer is voted in by Governance for a year - definitely keen to see what your research reveals, MakerMan.


I fully support the drive towards decentralization. If we are going to have a tonload of workers employed by Maker this is never going to scale.


Yeah those seem like manageable steps we can start to take. A shared multisig with the foundation and the DAO could act as starting point? Both need to sign off for any fund use, and both the DAO and the foundation contribute to the fund (stability fee and dev fund). We’ll need to talk to the technical people to see how much work this would require. I’m guessing it wont be to much on the technical side.

Creating the treasurer position could happen after the treasury is first established.

I guess the next step would be to make a concise signaling post to gauge people opinions on an initial joint treasury.

Intriguing, the idea of having an assistant or two for the main roles in the DAO makes a ton of sense. We can’t expect that Cyrus or Rich operate 24 hours a day, things will come up in their lives that inhibit their ability to work. Right now there are key people (visible and invisible) whose absence would likely be very detrimental.

In general I’m realizing that a pre requisite to scientifically creating roles and structure in the DAO requires a basis in organizational theory. I’m starting to work through some of that stuff. Trying to understand how we might we want to model tasks and their coordination in relation to the DAOs goals.

By Maker do you mean the foundation or the DAO?

I would start with proposals about creating Treasury. I assume there is strong consensus we need it, so it’s not really important at this moment how exactly we will spend it. I think payment should be in DAI, receivers can always buy mkr on the market.

Could it be implemented something like this?

  • create additional buffer (let’s call it Treasury).
  • create poll vote on variable called ‘treasury rate’. Let’s say mkr holders vote on 0.2 (20 percent) and confirm it with executive vote
  • after system surplus buffer reaches 500k, divert 20 percent of surplus to buffer, other goes to auction
  • executive vote is needed to transfer arbitrary amount of dai from the Treasury

I am just throwing idea naively, but probably developers could provide some proposals.

I think if we want to start with the whole scope of “hire the workforce that will power the DAO” then we probably need to start at a really high level and there are a lot of details that will need to be debated and agreed on before we can even start.

It’s probably a better idea for us to start at “What is the first task that we can pay a person to complete” and figure out how to make that happen. (That will still probably be very difficult)


Yes @stable ,

Or beside this, what do you really need to get done between all of this.

That would be a good logic instead of giving headache and setting standard too high for a point of start. By that i mostly mean giving the chance to everyone to compete and contribute.

Have a great morning… Coffee time.

Totally, I liked @MakerMan idea on starting with the treasury/treasurer. Thats a basic part of this process that has to happen before the DAO can employ people.

1 Like

Made a signal request Signaling Request: Should Maker make a Treasury to manage revenue?

A decentralized protocol will probably not stay decentralized if there are people who are able to control parts of it. This proposal creates an administrative class that can potentially negatively affect future efforts towards even further decentralization of the protocol.

The people most financially invested in the long term success of makerdao are those who hold the largest amounts of MKR tokens, so they will be the ones most incentivized to fulfill the necessary roles by employing the appropriate people to assist the project or fulfill these roles themselves with their own resources.

Employing people through the makerdao itself introduces social biases that probably cannot be fully ameliorated because it’s known that people will engage humans who agree with themselves. People with opposing views tend to be eliminated over time.

This proposal has the potential to significantly interfere with the direct financial interests of large stakeholders and the protocol by introducing an administrative class who could be incentivized by other types of biases, such as social biases, reputational biases and even financial bribes.

There are no real solutions to the removal of these biases therefore the intended future effect of this proposal is unknown but most likely it will weaken the future further decentralization of the makerdao protocol. It has the potential to centralize the power of the administrative group to the disadvantage of other stakeholders.


This post isn’t a proposal. Where do I say it is? Its clearly just my thoughts on how the DAO could pay workers. The other signaling request post purely asks about formation a treasury and dosent ask about access or any other detail yet.

Yeah hows that working out? In that logic large holders would pay oracle fees too, but they dont. What your suggesting is a more centralized process compared to large holders actually going through a governance process to incentivise work.

You also imply that makerdao would have more social bias than large holders. Why is that justified?

People have to do work at some point for maker to make, but even then nothing in my thoughts implies more control over the protocol for employees. The process would be controlled by governance (ak large mkr holders). An accountant/treasury position would only have control of the treasury if mkr holders wanted them too (which I severely doubt they would). Their job would be to recommend budgets and other financial tools to maximize the effectiveness of revenue.

We simply must figure out a way to incentive the work we want done. Right now the foundation control that completely. How could employment from the DAO be considered more centralized or adversarial to large holders?

I dont see a clear justification for this point. The foundation controlling workers is a non speculative centralized aspect of the system; open to attack from regulators. I am working on trying to push employment to the DAO which would promote transparency in roles, pay and performance. Why would transparency weaken decentralization?

1 Like

By interfering with the incentives for large stakeholders to be involved via the introduction of a communal pool, it would essentially create an internal “governing authority” that would be more vulnerable to social and corruption attacks.

For example, the governing authority is initially paid a small amount that increases to say $1 million a year. However, an external party like ripple or bsv offers a hidden payment of $10 million to the governing authority to green light the XRP or BSV coins. The potential for such attacks should be carefully analyzed, assessed and credibly neutralized before proceeding further.

I don’t think that the suggestion is to create a governing authority above MKR Holders. Any employees would be funded and employed at the decision of MKR Holders themselves. Additionally, risk models will almost certainly be public, so it should be easy to see if a risk team is favouring a certain asset (perhaps evidence of a bribe.)

I’m also not sure that your objections aren’t also true of a group funded by an individual large MKR holder. Wouldn’t they also be vulnerable to bribery in the same fashion?

And I find myself agreeing with Mitote here. If any large MKR Holders wanted to fund risk teams, development teams, etc, they are currently able to do this. As was pointed out, we haven’t see this happen yet, and frankly due to the diffusion of responsibility across token holders, I’m not sure that we ever will.

However, if we do see this happen, great the more people working on the system the better! The employment of individuals using DAO funds doesn’t prevent this from happening.


There is a greater level of accountability if the risk team answers to a large stakeholder who directly employs them. One could assume greater levels of alignment if the employer is also a large stakeholder and there is direct accountability. Disrupting direct accountability could lead to unforseen results, corrupt government authorities are prime examples. No system is perfect, which is why checks and balances will still be needed as you pointed out.

The determination of the risk team, even if public can have unexpected outcomes, simply because they are the incumbent governing authority. We can’t assume non risk specialists understand everything they’re doing. Even now, US pension funds are investing in unaudited Chinese companies.

Lastly, if the existing resources within makerdao continue to do their functions, there is no incentive for large stakeholders to step in. A reasonable suggestion would be to ask known large stakeholders to employ human resources to protect their own investments. This should ensure better alignment of stakeholder interests with administrative functions, minimizes the risk of social and bribery attacks by maximizing accountability to the stakeholder via direct accountability.

Why? What makes being accountable to a single large stakeholder better than being accountable to all the large stakeholders, and all the small ones, and everyone in-between?

I agree, I made a post suggesting that we employ community validation teams for this very reason. There are multiple solutions to this problem.

However, I am not convinced that having the risk teams directly employed by a large MKR Holder is one of them, because:

  • They are still subject to bribery.
  • They are reporting to a single Entity there is less chance that their output will be validated by multiple entities.

You are arguing against your own point here. The key point of your argument is that it is better for large MKR Holders if they fund their own risk teams (safer for the project, less risk of bribery, accountability etc.) Either:

  • It is better, and large MKR Holders have an incentive to do this.
  • It is not better, and they don’t have an incentive to do this.
  • It is better, but they all have the incentive to let a different large holder do this (This is the diffusion of responsibility problem I mentioned earlier).

Imo, it’s option 3, which is why we need to setup a system in which these positions are funded by all MKR Holders (ie funded by the stability fee income)

1 Like

The disintermediation of accountability by getting all stakeholders to fund the administrative function implies less accountability because they will effectively be answering to no one. This is the equivalent of a municipality electing municipal representatives who spend municipal money on useless things. The point is that the direct accountability to a single stakeholder is disrupted.

There doesn’t need to be a single stakeholder for there to be accountability, they won’t be answering to no one they will be answering to everyone. If funds were only distributed via executive vote then it would require a majority of MKR to sign off. If employees were not able to justify their salary then it would not be provided.

The problem with elected representatives in municipalities is that their actions aren’t always publicly visible, and that voters are often not able to remove them from office outside of specific times or truly outrageous actions. Neither of these things have to be true for MakerDAO.