Criticism of quadratic voting with creating lots of small issues

How do quadratic voting proponents address the gamesmanship of setting the agenda of what to vote on? Specifically protecting against creating a lot of similar proposals on the same topic, so those issue voters have to spend a lot of credits protecting their interests, relative to other issues.

For example, consider the following two ballots:

Issue 1. Vote on topic about the environment
Issue 2. Vote on topic about civil liberties

Issue 1. Vote on subtopic about the environment
Issue 2. Vote on another subtopic about the environment
Issue 3. Vote on another subtopic about the environment
Issue 4. Vote on another subtopic about the environment
Issue 5. Vote on another subtopic about the environment
Issue 6. Vote on topic about civil liberties

Ballot 2 is artificially rigged such that people who care about the environment have to spend a lot more on their issues, since one issue has been split into many issues, leaving less left over for other topics.

1 Like

Your question is a great one…

All issues need to be created in a multi-level structure, that is nested which allows flexibility within different subgroups. My talk, ‘Physical Framework for Healthy Governance’ references this in terms of Village, neighborhood, district and city levels of governance.

Like Vitalik sees humans and their social activities as playing a significant role in Ethereum, any governance tools must have some relationship to our human physical world. Audrey Tang’s wonderful work, exists within a very strong level-based system with centralized control: Digital Minister of Taiwan appointed by the presentient.

These are essential issues.


The criticism from @jrp (vote splitting) is not exclusive to quadratic voting, but applies to most voting methods except scoring (although scoring methods would also suffer from rigged ballots for other reasons).

This highlights a fundamental difference between decision-making (eliminative) and proposal-forming (creative) processes, and the directional dependency from the latter to the former.

As such, the original criticism stands: I would go so far to call it a feature of QV. Should we start a new topic to discuss how collectives could construct ballots in the spirit of QV?

1 Like

Hi @doug, it might apply to many other voting methods, but I wouldn’t agree with “most”…

For example, simple one-person-one-vote + majority rules wouldn’t have this problem. It doesn’t matter how many different issues there are to vote on, since every person gets one vote per issue, so creating new issues wouldn’t matter. If, however, you mean creating multiple alternatives in a single vote, then yes vote splitting is an issue there, but many other systems like approval voting don’t have that problem.

Here is a chart I found, where a criteria we’re discussing is “cloneproof”:

Ah yes, I think you’re right :sweat_smile:

If I return to your original post, you state that quadratic voting could suffer vote-splitting when facing a ‘poisoned ballot’ of 3+ options, to which I agree.

I meant to reply that this is would also be true for any cumulative method facing a ‘poisoned ballot’ of 3+ options: plurality/‘vote-one’, cumulative/‘allocate-n-voting-points’ or QV.

Quadratic voting is the only interesting cumulative voting method for me, because it encourages economizing on extreme opinions, while expressing them in a scalar way: this is AFAIK not possible with scoring/ranking methods.

In any case, a voter is limited to the options on the ballot, whatever method they’re given! Ballot-forming seems to be worth investigation in any case then.

1 Like