Minimal Viable Governance

Just posted in the CIP-1694 PR:

My vision of a “minimal viable governance” system would be much more minimal: No constitution, no committee, no dReps, no SPO votes, just pure voting by ADA holders.

Then the only governance actions left that are needed are hard forks, parameter updates, treasury withdrawals, and polls.

Do these rather mundane actions need a constitution, a committee interpreting and guarding this constitution, dReps spending significant time evaluating them (and maybe even getting compensated for that)? I’d say: Definitely not.

Up to now, hard forks and parameter updates are pretty rare events.

Hard forks necessarily will stay rare, because they need implementation in the nodes. (If it is a realistic/desirable scenario that a hard fork is implemented in the node, installed by enough SPOs, and then shut down by on-chain governance, is an interesting question there. How to get a change implemented if the cardano-node team doesn’t want to implement it, another one. But both are outside the scope of this CIP, probably.)

Hard forks are also the only action left for SPOs to vote on. But why? At the moment, where a governance action is proposed, they should already have voted “with their feet” and installed the implementation for that hard fork. Why should they then have a special role in determining if it is also enacted?

The long-standing issues with minPoolCost and nOpt/k might lead to a lot of competing parameter changes shortly after governance is enabled, but once that is resolved, they will also be rare and pretty boring again.

Treasury withdrawals might be controversial, but they will be much rarer than they are now: A system like Catalyst will withdraw the whole budget for a funding round at the beginning of that round to be able to guarantee that it can fulfill its promises according to its own rules. “You will get voting rewards according to this formula (if the governance system approves).” and “You will get funding for your project (if the governance system approves).” seems much too uncertain.

These are a very manageable number of Yes/No/Abstain decisions per epoch which all interested ADA holders should be able to do in their wallet apps without intermediary dReps. If we implement it right, it’s one transaction per ADA holder active in governance per epoch. The chain will survive that.

There already are provisions proposed above to make sure that dReps vote in time that delegators can withdraw their delegation before the tally and that “everyone can be their own dRep” is also promised to all critics. Why then even introduce them in the first place?

If the proposers of governance actions and people interested in the development of Cardano cannot rally enough holders to vote on important actions, they probably also should not be enacted.

Yes, the thresholds should be low enough then, but also safe enough. I have proposed a dynamic margin requirement for that here:

The idea is to require No-Percentage < 2×(Yes-Percentage)² If the Yes-Percentage (share of total registered voting stake) is low, very few No-Votes can keep the action in an undecided state. If the Yes-Percentage is higher, this requirement on the margin gets lower and lower.

For unproblematic actions, we would not need much participation then, but if an action turns out to be controversial, both sides have to rally.

How much registered stake do we want before we arm that system? Probably not that much. We have a lot of stake owned by people who are not even interested enough to move it off the exchanges. And additionally also quite some stake on wallets that have no activity at all for a long time, so probably long-term holders who only look at this “investment” every few months or years. They could join governance anytime they want, but do we need them to? Probably not. So, I’d guess registration for governance voting at something between 10% and 30% of total stake should be enough.

Path to active would simply be: One hard fork enabling the vote registration and once enough have registered another hard fork disabling the old governance keys and activating this.


thanks @HeptaSean - just read it on GitHub. There does seem to be community support for a flat governance model according to what I’ve read in the CIP-1694 PR and social commentary. As much as I hate to add to the review burden it does seem like your suggestion would justify its own CIP.

On that note I wonder if the acceptance or boot strapping process (apologies if I can’t keep the nomenclature straight, if there’s a preferred term for the transition from off-chain to on-chain governance) could begin with a choice of which CIP to adopt for its governance rules… since it looks like we will already have 2 or 3 posted CIP alternatives. I haven’t read every corner of CIP-1694 or every comment yet, so I don’t know if that’s included or been suggested before.

(will re-post this in the GitHub thread if this question hasn’t already been addressed before)

1 Like

Will have the same problem as all the RSS CIPs and as also highlighted for hard forks in my post/proto CIP above:

Without a commitment to implement whatever is successful, such a choice would not make much sense.

To minimise community uproar, it’s probably even advisable to not frame it as “Meh, we don’t like it! We are just doing our thing.”, but rather as: “No, that alternative is not viable! It’s irresponsible to do it that way!” And it will be a matter of belief to either buy “Drama and FUD” framings or “They are oppressing viable alternatives!” framings.

And only starting the implementation after a successful vote, will also mean a huge delay. So, if n possibilities are there and proven to be ready for deployment, n−1 of them were developed for the bin. … or not, but then all not implemented alternatives will have an uphill battle against: “Not even implemented!”

This Gordian knot does not seem to have an easy solution.

Last thought for the moment: I have to be honest here: I do not know how relevant that support is in terms of people as well as in terms of stake. Might well be a noisy minority.

I am offended by framing it as “Drama and FUD”, of course. Might be that a vast majority is content with everything in the current course of CIP-1694 development. But I have seen a lot of thoughtful reasonings from the group that is not so content with it. Just dismissing that does not do them (us) justice.

Interestin read @HeptaSean and thanks for posting it.

Would you be able to give some examples of how this would work in practice, and am I correct to assume that in your model, the total registered voting stake would be all ada at all Cardano addresses, or would you say that only ada that is delegated should be considered?

1 Like

Neither. My preferred way would be that people explicitly register for governance voting (maybe even with an automated expiry). And I would take that (minus explicit abstains for this particular vote) as the 100% to not make it too hard.

If you meant delegating to a dRep that is basically the same, but I’m arguing for not having dReps, here, so I’m formulating it as “registering for voting”.

But in my view, participation of a stake in governance should be completely independent of participation in staking, delegation to a stake pool: There might be tax or other reasons for not doing staking (although the big discussion in Germany seems to be resolved now), but still participating in governance. And long-term holders might want to stake, but not participate in governance, because they are not that actively monitoring the Cardano community and its discussions.

Hmm, let’s see. Suppose 1 billion ADA are registered for governance. (I still have on my ever-growing to list to run some numbers if that should be a satisfying and/or realistic number.)

If there are no “No” votes, any number of “Yes” votes could pass a proposal. Which is okay, even one of the goals to allow uncontroversial, boring topics to pass easily.

For example 1 million ADA voting “Yes” would be 0.1% of the registered voting stake. 0.1%=0.001, 2×0.001²=0.000002=0.0002%, 0.0002%×1 billion=2000 ADA voting “No” are enough to prevent the proposal from passing for now.

Let’s say 8 000 ADA vote “No”. Those are 0.0008% of the registered voting stake. We would need more than sqrt(0.0008%/2)=0.2% (the inverse of the 2×Yes² formula), i.e., 2 million ADA voting “Yes” for the proposal to pass.

If we have low participation, the majority has to be quite overwhelming. If it turns out that a proposal is controversial (even if only a quite small minority opposes at first), both sides need to rally for participation of their side in the vote.

If 100 million ADA vote yes, those are 10% of the registered voting stake. Then, 2×10%²=2% “No” votes, which are 20 million ADA, are needed to keep the proposal undecided.

If now 200 million abstain, the total voting power relevant for this proposal goes down to 800 million ADA. 100 million ADA of 800 million are 12.5%, 2×12.5%²=3.125%, 3.125%×800 million=25 million ADA voting “No” are needed to block. Abstaining makes decisions easier.

When we approach 50% of the total voting power (registered minus abstain), we approach the point, where 50% + 1 vote is enough to pass.

Still with the 200 million ADA abstaining, 380 million ADA voting “Yes” are 47.5%. 2×47.5%²=45.125% “No”, which are 361 million ADA, are needed to still keep the proposal undecided.

Does that make it a bit clearer?


Before anyone asks: I am not aware that this has ever been proposed or used somewhere, unfortunately. Quora seem to be set rather ad-hoc almost everywhere, even in quite important “official” rules. I’d be glad if someone finds in-depth sources.

For comparison, these are the alternatives often found:

Even the – objectively bad – participation quorum is not only used in Catalyst, but also in the rules for state- or nation-level referenda.

CIP-1694 up to now includes what I have called “Fixed Margin Quorum”, which is quite okay, but requires a certain level of participation at the lower end and does not allow tight decisions even at 100% participation (might be considered good or bad). And the margin is an arbitrarily set number – like in the other examples.

The reason for the minority < 2×majority² formula is that it is – arguably – the simplest mathematical function allowing decisions even at low participations and approaching no margin requirement at 100% participation.

I’m not into the details of this topic at all, so I may be wrong and please correct me, but I see a point having the ability to give my vote to a representative that I believe I trust and with whose current opinions I agree with, to someone who I believe had the time and strength to put the effort into research of the topics with enough knowledge to make good decisions.

The reason for the majority is the reality, life and lack of time and resources to get deeply involved with every topic being voted for. Yes it does look like the governments we have now, but with one major advantage and that is the ability to withdraw my voting power instantly from the representative and to get involved myself into voting if I feel like this. No need to wait four years to change the government that promised one thing but eventually did something else once in power.

That’s what Charles mentioned in his video few days ago: Drama and FUD with Governance - YouTube

Something that I’m not sure how it will be solved is the accumulated power of the representatives and possible bribery ie lobbying… also how do we ensure those people get a reward for the time invested into proper research so they can make a living and not fall for corrupt deals and bribery, other than having trust they are sacrificing their time for the greater good.

Yes, it absolutely does. Thanks a lot for the clarification.

Okay, I understand. What you are describing is the DRep system minus the delegation part. Why would you want to design a system where we cannot have liquid democracy? I would love to understand your reasoning. It’s a fascinating topic for me.

The way I see it is that having the possibility to delegate voting power is just the next step in the evolution of voting systems. But my knowledge is limited, and I’m happy to be shown reasons why my thinking could be flawed.

1 Like

Without a constitution, if the ADA holders are misinformed or under-informed about particular matters, that would have potentially large effects upon the Cardano ecosystem. While ADA holders may have the best interest for the community, there may be external actors that intend on jeopardising Cardano. There has been countless fud and noise that misrepresents Cardano in the media that thankfully Charles has addressed such issues. But what happens if Charles is for a myriad of reasons (including that where he intends on retiring and be a spectator of Cardano) whereby he cannot address the misrepresentations of Cardano.

For the medium and longer term, I believe a constitution would help in addressing potential external threats that hinder the growth of the Cardano ecosystem.

I agree that proper research, and from my perspective based off the academic peer-reviewed findings amalgamated from the fields of political science and anthropology would be required to mitigate potential bribery and corruption.

Such academic fields have research carried out for decades that have concrete findings of which the community and the founding entities of Cardano could openly discuss. I believe that we should have CIP-1694 workshops that include political scientists and anthropologists to determine the most viable governance structure.

I don’t see how a constitution prevents voting being influenced by misinformation and fud?

Does the US government constitution prevent misinformation from influencing a vote…

Where did I say that? What I said was without a constitution, there would be potentially large effects upon the ADA ecosystem. There has been countless fud and noise in the media that misrepresents Cardano that jeopardises the blockchain’s interest. A constitution would assist in addressing potential external threats…

Otherwise what other effective measures should we implement to safeguard against external institutions that serve their own interest?

With the caveat that your withdrawn voting power will be quite irrelevant compared to the dReps. Also see answer to @Lovecoach below.

It doesn’t have to be bribery or lobbying – and lobbying isn’t always bad – trying to convince legislators that your interests are valid is lobbying – and it is typically only seen as something bad from the ones with other interests.

The accumulated power can even be a problem if one of the more powerful dReps (even temporarily or just for one specific proposal) doesn’t employ the due dilligence expected from them.

We already have vastly different voting powers due to whales (but in my opinion have no viable alternative to prevent that right now, perhaps never). dReps make that worse either by accumulating voting power of multiple whales or by becoming another class of voters just as powerful as the whales or both. The promise of “You can always undelegate and vote yourself.” gets a little hollow if your voting power than still stands against the remaining voting power of the dReps.

The claim is that people have no time, energy, or expertise to vote themselves, so they should be able to delegate to a person they “trust”. But assessing who to delegate to is just as cumbersome. And monitoring/controlling if that trust is still justified also requires engagement. I don’t even know if I would do it, but, worse, I can always have the suspicion that the others don’t do it, that all the delegated votes I am trying to stand up to are just fire and forget delegations based on popularity.

(This is all much worse for the planned Catalyst dReps. It is even harder to assess if someone who seems competent on first sight will really employ due diligence in assessing thousands of Catalyst proposals and will vote on all proposals – I don’t want my meager voting power wasted by only voting on a handful of proposals and downvoting the competition is imperative given the current voting system.)

In both cases – governance as well as Catalyst – my alternative – which does not have to be implemented in the core system – are just recommendations how to vote. The voting interfaces – wallet apps or dApps or whatever – should just allow to import recommendations of several people I trust. It can then allow to set my vote automatically if they agree, highlight the points where they disagree for special attention, I can mark the proposals I have a strong opinion on myself, and combine all that to result in my very personalballot.

People could still blindly follow just one person if they want, but at least they sign it off as their vote themselves. And there is a good chance that they look at what they are signing there, that they notice if something is controversially discussed in the community and check in what camp their favourite recommender is, that they not only choose one recommender, but several. It opens the possibility of a continuum between choosing just one person to trust completely under incomplete information and forming a detailed own opinion on all the proposals yourself.

Moreover, Liquid Democracy typically includes that delegates can delegate the voting power further for specific proposals or whole subject matters. Such transitive delegations are not planned up to now, so true LD fans could say that we are doing some fraudulent labelling here. But transitive delegations are also even more problematic, I think. That I trust someone making good decisions does not mean at all that I trust them to find someone else who makes good decisions. And it accelerates the trend to accumulate voting power in very few hands.

I don’t believe in Charles at all nor in supposed “truths” he purports.

The “FUD” – or just other opinions than yours or mine – is quite irrelevant for the decisions handed over to governance in this step – hard forks, parameters, and treasury spendings, which is also why I don’t see a “constitution” necessary for these simple decisions handed over to governance at this point. (The more complex questions – how to manage development, who gets to decide what is developed and what not – are out of scope of CIP-1694 and, hence, also of this proto-alternative.)

A constitution in the form of an informal text document doesn’t add much safety. The noble motives laid down there have to be weighed against each other. That’s what institutions like supreme courts usually do – and probably what the “constitutional committee” is supposed to achieve.

Depending on country and personal stance, those courts work more or less – in some countries heavily depending on the current majority in the court, in others being a quite consistent check on the other branches.

But I don’t think electing such a body by popular vote with the power to no-confidence it at any time is a wise choice. I don’t see something like a court coming there, more an unpredictable something that can stop anything at will with the claim that it is for some obscure reason “unconstitutional” – to be then no-confidenced if the electorate is angry enough about it.