AGP discussion: Support Locked Voting

Discussion thread for https://github.com/aragon/AGPs/pull/63/files

I have previously suggested this idea and received some early support, so I figured I’d get ANT holders’ official opinion in an AGP.

3 Likes

Thanks John.
One question around practicality: why would this just be a proclamation? What if this one was successful, we all spent resources and efforts in developing this, and then the “ratifying” AGP would be unsuccessful?

I like the direction of this proposal. Though I can’t imagine the cryptoeconomics network effects of this decision and how will affect ANT.

In the first order of consequence seems to benefits the users of the network that been involved for at least 1 year. Is this a behaviour we would like to incentivice? :thinking:

In my mind it seems that a progressive model would make more sense in general, for example:

  • 1st ANV at least 3 mont lock
  • 2nd ANV at least 6 month lock
  • 3rd ANV at least 9 month lock
  • 4th ANV onward at least 1 year lock
1 Like

I think you might have misunderstood how it works (which I did as well at a first read).

The thing is that people would commit to locking at the time of the vote, not have had their tokens locked for the prior year.

At least this is how I understand it now, and think it would solve your issue.

@light maybe can clarify if that’s not the case

1 Like

I love the idea of this, but perhaps we can use the Survey app outside the AGP process to get the signal instead of a proclamation? Then that gives more than 2 days to gauge sentiment. I feel like we should do more surveys in general, not sure what the roadmap is like for getting that app running again — @sohkai?

2 Likes

Totally agree on this! I like the proposal but would like to see that progressive approach to it as well

Yes, ANT holders would lock their ANT for a period of one year or more to be eligible to vote. They will not be able to vote if their ANT is not locked for at least one year at the start of the vote.

I think there’s a risk that we invest the resources to develop this feature and ANT holders reject adopting it whether we get a signal from them using this AGP or not. The benefit of this AGP is that we at least get a signal that they either support it or not, the idea being some data is better than no data.

Maybe, we would have to formalize and build some legitimacy / expectation around using the survey app for signalling which currently does not exist. Since ANT holders are currently most engaged around the AGP vote, I think this is the best tool we have right now to gauge sentiment on this issue.

1 Like

I actually feel like locked voting will be a feature requested and useful for other DAOs. I know that it’s definitely important for Space Decentral! And if we have it as an idea for Aragon Network… I would say that it will likely be requested by many other orgs. We will be expanding the governance capabilities of the platform!

Yes I also agree a more progressive approach to begin with would be good. Why I mentioned the survey app, is because I thought that’d be a good way to gauge the length of the lock, vs. a binary yes/no vote.

Can you clarify how you see there is legitimacy to using a proclamation for signaling – or why for the purpose of signals, the duration in which the signal is taken needs to match the AGP process? Which part of AGP-1 specifies that?

If the Survey app is ready, I don’t see the harm in adding it to the governance DAO and having a survey that’s open for a longer period of time. [Correction: I get it now! Can’t add the survey app without an AGP] They can participate in the survey during the regular AGP vote, it would just be open longer so we can get really good quorum on the signal.

And yeah, agreed, due to this pondering this does make me think a meta AGP to clarify the signaling process in general would probably be useful. I guess not sure if there’s enough time to hash that out for now though.

There is legitimacy because this is the “official” governance process, and the Proclamation track is specifically to give ANT holders a voice on arbitrary subjects - in this case, to see if they support the idea of adopting Locked Voting. If they approve this AGP then that would send a signal to anyone who would fund / build a Locked Voting app that ANT holders are willing to adopt it for the AGP process.

Does this clarify?

1 Like

The thing is that people would commit to locking at the time of the vote, not have had their tokens locked for the prior year.

You are right, now that I keep thinking about, it make sense to be that way.


Yes, ANT holders would lock their ANT for a period of one year or more to be eligible to vote. They will not be able to vote if their ANT is not locked for at least one year at the start of the vote.

Thanks for the clarification.


With 1Hive team we are starting to work on a first implementation of a lock voting kind of app:

1 Like

Yes I see the use of this now! Thanks for the clarification.

I think the AGP you propose here makes sense.

But based on the feedback, I wonder if it would be better for the lock length to be specified in the later AGP that brings it to life?

@luis idea here about the lock length increasing your voting weight can be another dynamic to exlore.

1 Like

I support this! It sidesteps the ‘buying voting power’ problem, but 1 year seems like a sledgehammer.

I am cognizant of the low voter turnout which would most probably decrease. Another approach could be some kind of time-weighted voting, although this would increase the complexity

+1

1 Like

I’m very glad to see the plutocracy problem in Aragon being acknowledged and to realize there are already initiatives/ideas that are willing to tackle it, but I have to say I’m not personally sure about this being a good solution, as this could end in having votes only from the usual suspects as it is basically a counter-incentive for new people to come to the Aragon ecosystem, understanding and adding diversity by their participation.

This proposal has nothing to do with addressing plutocracy. It’s a measure to stop someone buying voting power and using it without having exposure to the results. But I do agree with the need to look closely at how various lock times change the incentives of all participants

3 Likes

Plutocracy is a tricky theme last days, and I have realized lately that people tend to handle different concepts (just like it happens with terms such as capitalism).

Anyway, in my own perception, as long as people can acquire voting power with economic wealth we’ll be in the presence of plutocracy. Like it or not, governance by ANT holding is precisely that, for the time being I think it’s ok until we discover and validate better mechanisms.

Anyhow, @light I’d be curious about your view in implementing something like conviction voting instead of locking tokens to tackle this very same issue.

1 Like

The 1Hive Lock app as its currently intended would not help address this issue (Since the current voting app does not have a role for cast vote, and I believe the intention is for voting to be limited to Locked ANT). It more intended as a way to minimize spam by requiring some cost associated with taking an action (in this case its an opportunity cost of locking the asset).

Perhaps a seperate thread on conviction voting would make sense? Its one of the most interesting voting mechanisms I’ve seen proposed but feels a bit off-topic for this thread…

This is my main concern with this proposal as well, I talked about this in the other thread that this came up in: Whale watching on ANV-2 🐳

A question that comes up for me is how this would interact with delegation? Presumably it means that in order to delegate your vote you would also be subject to the 1 year lock requirement.

Another challenge would be for organizations that use Fundraising using the default template which includes an ANT reserve requirement. One of the benefits of this is that the organization can choose to participate in the AGP process, but it would not be possible for them to lock ANT held in the bonding curve reserve, as it would break the bonding curve sell function.

I think time-locking mechanics are an interesting and would be supportive of similar proposals, but would not support an implementation of locked voting as currently described for the AGP process.

One approach that I think could be interesting would be to add a bonus weight for locking ANT (perhaps also have this bonus be quadratic and tied to some sort of identified user registry). That way people can still participate without locking, but those that do lock have a greater say.

2 Likes

I think the only effect this would have is to stop people from voting with their ANT until they are comfortable exposing themselves to the consequences of a vote for at least one year, which is precisely the goal of the proposal. If you aren’t comfortable exposing yourself to the consequences of a vote for at least a year after the vote, then you probably shouldn’t be participating in Aragon governance. We want long-term participants.

I’m not sure what is possible from a technical standpoint but the way I would do it is that everyone who wants their tokens counted in a vote has to have them locked for at least one year, whether the tokens are delegated or not. So the UX would be: lock for one year -> delegate. Then the weight of the delegate’s vote is measured by the number of correctly locked tokens that have been delegated to them (and the number of their own correctly locked tokens).

Would it be possible to have something like once someone sells their org tokens back to the curve, if the ANT they are trying to buy from the curve are locked then they will receive the ANT once it unlocks?

I had considered adding a “weighted” mechanic to the proposal i.e. the longer the lock, the more weight the vote gets. But as people are already sensitive to the economic weighting of ANT votes I figured it would be best to not increase the voting-weight disparity between larger and smaller holders. I’d prefer to keep voters on an “even footing” so to speak.

1 Like

Its not just that you are exposing yourself to the results of the votes you choose to participate in but also the results of all votes within the next year. (presumably it also means that in order to eventually divest you have to have not participated in any votes for a year).

Its very reasonable (especially for a smaller holder) to prefer liquidity over participation because they have relatively little say in the outcome anyways.

The importance of exit is why in Moloch for example, votes are done on a regular basis but in sequence, if you vote yes you are not able to exit (the shares are non-transferrable and the ragequit option is not available) until after that decision has been executed. Because votes are in sequence, you will always have the option to exit if you choose to vote no on a subsequent proposal.

Perhaps if we do go down the locking route it would make sense for us to lock only for 2.5 months, that way if you are voting you forfeit your ability to sell before the next ANV, but you are not exposing yourself to future proposals that you have never reviewed, and possibly only if you vote to approve proposals are you subject to the restriction.

Agreed it seems like it would make sense for those delegating to also be subject to the lock requirement…

Though it makes me think about the tradeoff space for delegation implementation (and one of the reasons we decided to implement delegation as an upgrade to the voting app rather than as a seperate proxy voting contract) is that delegation should be possible without having to transfer tokens out of cold storage. Presumably the locking requirement would mean that users would need to send their tokens to a contract, which can deter certain groups (like funds which have policies on asset storage) from participating.

I’m not sure, it seems like it would be quite complex though, especially if locking requires moving a token into a separate contract.

I updated this AGP draft to remove the one-year requirement so we can focus on the general concept of Locked Voting:

This may be the better option, if we decide to go down this path. Though I think 2.5 months is not really a long time for the negative consequences of a vote to be realized, the same could be said for a one-year lock time too. I had originally considered four years, inspired by the typical 48 months vesting in startup equities :wink:

That’s fine, they just wouldn’t be eligible to vote then. This is the explicit trade-off this AGP is asking people to make. Trade liquidity for commitment and voting rights.

If this Locked Voting contract becomes a standard then I don’t see why they couldn’t add it to their storage policy, then they can still vote in orgs that required tokens to be locked to vote.

That does sound complex!

Based on all of this feedback it sounds like the options are:

  • Implement Locked Voting and exclude a bunch of ANT holders for various technical (hard) or policy (soft) reasons, and maybe mitigate the Bad Voter attack vector (what I’m calling the attack where someone buys a bunch of ANT, makes a bad decision in a vote, then dumps the ANT before the market has time to react), or
  • Don’t implement Locked Voting, remain fully exposed to the Bad Voter attack vector

ANT holders then have to weigh whether excluding a bunch of ANT holders from voting is worth the risk of being exposed to the Bad Voter attack vector.

Agreed. I’m just pointing out that this tradeoff is worse for you the less of a whale you are. Which is I think important to identify since the other main concern I’ve seen vocalized is that small holders feel disenfranchised by straight token voting.

And while I’m very much of the opinion that stake weighted voting is the right choice for the Aragon Network right now, I also recognize the importance of individual contributors to the ecosystem and I would hate for them to feel even more marginalized by the project’s governance.

This is a false dichotomy, there are other ways to potentially approach the “Bad Voter” attack vector, that may have completely different tradeoffs.

1 Like