At TQuorum NY someone asked @adrian if people could be rewarded with XTZ for voting. He responded saying that that would lead to undiscerning people voting just to get paid, and not taking their votes seriously, which would corrupt the outcomes.
I agreed with that, it made sense. In my mind, I imagine a bunch of bakers who otherwise didn’t care about voting, didn’t pay attention, seeing voting was about to close and thinking “Oh crap! Effectively, I’ll lose money if I don’t vote” and in the last minute just voting, in any direction, blindly saying ‘yes’ or ‘no’ without even knowing what the vote was.
Likewise, wouldn’t this method have the same potential problem — with bakers thinking: “I’ll be losing money if I don’t vote, so I’ll just pick something, anything!”?
I am curious though how exchanges would respond to this because in my conversations with the bigger ones they vehemently said that they definitely want to be governance agnostic; to abstain from voting.
TF also sees itself as a neutral force in governance but always participates in votes by voting pass.
Exchanges can learn to vote ‘pass’ and still remain ‘governance agnostic’. That’s also true for bakers who don’t have an opinion or interest in voting.
Right, but that assumes all unmotivated participants would vote ‘Pass’, which is not the case even in a secret ballot as studies have shown, let alone when one’s vote is public on the blockchain.
When people are compelled to the ballot box with 3 choices in front of them — Yes, No, Pass, there are still cognitive biases in the form of groupthink that would not be suffered otherwise. — bandwagon voting, for example. Basically, many people see a proposal is ‘winning’, and whereas they come into it uninformed and agnostic, they shrug and vote ‘Yes’ assuming it must be good and they want to go the way the wind is blowing (wanting to vote for the winner).
This could be exacerbated in a voting environment like ours in which votes are public. No one wants blame for a certain outcome, so going with the flow / hiding in the crowd becomes intuitively attractive.
Or maybe their personality dictates that when agnostic and seeing a winning option, they’ll vote the opposite because they see an underdog — the underdog effect
BUT when voting of any kind is non-compulsory; if those going to the ballot box in the first place are from a self-selecting pool, motivation is much more organic.
Not voting can be a sign of inactivity, but sometimes it can be used to express dissatisfaction. Maybe a voting period can be followed by two cycles where non-bakers can send a special proof-of-activity operation, in order for them to avoid being marked as inactive. After that the nudge can go into effect and it would be clearer that it hit inactive bakers, rather than non-voters.
As a refinement to the idea above, I would suggest “automatic pass”, an option that bakers can select to implicitly and automatically vote
What if bakers want to express dissatisfaction by not even acknowledging the need to send a proof-of-activity? It’s turtles all the way down.
If that’s really a use-case create an operation, call it “displeased” and make it semantically equivalent to pass.
There was a baker that abstained from voting recently and expressed it as a passive statement. But what it really boils down to is that I don’t think punishment (lose baking & endorsement rights) is a good way to encourage participation. It would most certainly result in a higher participation, but how would the previous non-voters act? I see a risk that they would just throw a random vote in order to not lose any baking and endorsement rights. That wouldn’t add any value, only create noise. I think genuine participants will make wiser decisions.
I don’t think they’d toss a random vote, I think they’d toss pass. It does seem marginally better that people explicitly pass and the quorum stays high than they don’t vote at all.
In my opinion, nobody can be forced to vote even if it were to vote pass because this vote is still counted for reaching the quorum.
I would propose an alternative, we can create a new type of vote, “abstain” !
Voting “abstain” demonstrates activity but does not affect the quorum or the vote.
At this point, if you do not even vote abstain, you deserve the penalty.
If automated voting is a thing, “abstain” would be a danger to protocol progress.
With the current options just a “pass” vote would be possible to be automated (which does not “jeopardize” the progress of the protocol).
I think this is ethically justifiable in so far as much that the incentive does not create “undue influence” in participation. This concept parallels some of the issues faced in medical research. Scientific studies need participates so that the study is not “underpowered” (i.e. study does NOT have enough data to detect an adequate signal to noise ratio, called type II error in statistics) so it is ethically justifiable to allow incentives to increase participation; however, if the incentive is too great then it creates data bias towards low social economic status patients jeopardizing the ability to generalize the data to the general public. So one really has to look at the whole picture (pros - increase participation versus the cons - jeopardizes data/voting integrity). I think on balance, just as practiced in medical research today, some level of compensation for voting participation is reasonable such that the rewards (or penalty for not participating in governance) does not create undue influence.
See New England Journal Paper for more info on this topic from medical science perspective: https://catalyst.harvard.edu/pdf/regulatory/Gelinas%20nejmsb1710591.pdf
This is a wonderful idea, bakers need to vote or at least officially pass to be considered active.
I’d like to see this being followed up by somebody.
A 5 cycles penalty seems too high. I would only do 1-3 cycle max, preferably 1 to start. While this is in effect we become more centralized therefore our security potentially drops.
Kevin brings up some good points about people just randomly choosing the leading vote rather than voting pass. And if we default to pass then we don’t need to implement the penalty. We can tell if a pass was implicit or not by whether it was cast at the last block of the cycle (default), or somewhere in between the cycle.
I agree with the automatic pass too, with 1-3 cycles without rewards though.
Good point about the risk to the consensus. But if there’s autopass what is the point of pass or even the quorum?
I’m for this idea. We need the voting system to be more used!
I support this idea, however it would be helpful to fix the voting feature on Tezos kiln before implementing.
Autopass would only be useful if it is not the default and therefore requires a one time manual setting. If implemented it would be cool if the autopass option had a one time burn fee to deter it’s use.
Let’s do it.