Great article; very interesting and well written.<p>A bit of a tangent, but something that bothers me is the assumption in economics that the amount of money one is willing to pay for a good is equal to the utility provided by that good. And then it follows that if I am willing to pay €2 for a coffee and another person is only willing to pay €1 for that same coffee, then I value the coffee twice as much as the other person.<p>But to me that doesn't make much sense. What if I am a billionaire and the other person is very poor? It could very well be that the other person values the coffee a lot more than I do, but because I have so much more disposable income than they do I am willing to pay more for it. To me, the assumption should be something like the amount of money <i>as a percentage of my total wealth</i> that I'm willing to pay is equal to the utility of the good.<p>And because this assumption is so fundamental to economics, it determines the conclusions that the field reaches. For example, the definition of an "efficient market" or the optimal level of production of some good both embed this assumption that willingness to pay equals utility. And then one can imagine that policy decisions and such are influenced by this assumption. (And in fact the article itself mentions this problem: "Another challenge is that quadratic payments, being a payment-based mechanism, continues to favor people with more money.").<p>So I'm wondering if anyone has looked into an alternative formulation of economics with a different fundamental assumption (perhaps something closer to the one I proposed above). If so, do certain things that are determined to be "optimal" in classical economics turn out not to be optimal in this alternative formulation (and vice-versa)? Apologies if this is a bit of a naive question; Econ 101 is the most I've ever studied the subject.
This post describes a system of voting/payments where you can make multiple votes, depending on how strongly you value your preferred outcome, but each vote costs more than the vote before.<p>Identity management and collusion are big issues with these systems, as the essay points out.<p>For programmers, one way to think about this is that each voter has a state, n, that tracks how many of times they've voted. And each vote is more expensive than the last. Eventually you either run out of money or you lack the desire to spend more money on another vote.<p>But if you can create a new voter account as easily as you can create a new Gmail account, then once n gets sufficiently high, you just switch to a new account to lower the price of your vote.<p>Or if a really rich voter with a high n-value pays someone with a low-n value to make a vote on their behalf, the system collapses as well.<p>Enforcing strict identity management (e.g. requiring valid state-issued ID cards) and implementing secret voting can help address these problems, but my guess is that if there is a strong enough incentive, people will try their hardest and come up with novel ways to thwart these safeguards.
This is a fascinating collective decision making scheme. One potentially serious nitpick. In the "one dollar one vote" scheme, it makes a simplification that seems really bad. It's the same flavor of mistake as when you have the option of making a bet. If the expected return is positive, it says you should bet all of your money. In reality, people don't do this. Rational actors shouldn't even do this.<p>I'd love to see what an attempt at this same sort of number-of-votes-is-proportional-to-value comes to with a less simplified model of behavior. If I value outcome A at price $x and outcome B at price $y, I might not be able to afford $x^2+y^2, and this model doesn't say what I would or should do. A corresponding model that talked about how people <i>allocate</i> their finite money rather than a unit by unit <i>spending</i> description seems like it would better apply to reality.
I don't think this idea is actually implementable in a large-scale, real-world setting. One person, one vote democracy is pretty understandable to an average Joe who doesn't even remember what "squared" means. I think that a normal person would have no idea how much to spend on a particular decision, as the formulas would seem like black magic to them. Normal voting is pretty intuitive, quadratic is not. This problem is not that noticable for HN people, as our average intelligence is probably way above the societys', but before deploying this on a large scale, it has to be considered. I don't think that internal workings of democracy should be something that an average school child can't easily grasp.
We tried quadratic voting for our team choosing a new name, and we did not find it to work well. I think the analysis that the scheme even converges makes simplifying assumptions that aren't well justified, particularly that everyone knows the marginal probability of each candidate winning (when that itself is behind a fairly inscrutable voting method.) Without that assumption I think it falls apart.<p>The optimal strategy is to vote maximally for your preferred most likely contender, and maximally against the nearest competitor, but if you don't know who those are, the system provides no means of discovery. So strategic voting is not just possible, it is critical to have any influence, but the evidence needed for optimal strategic voting is neither obvious, nor revealed by the system, (and from our experience I don't think it has any stable Nash equilibria.)<p>After a few days with the system, we finally abandoned it and switched to a condorcet election.
While reading I wasn't convinced this was going to work due to the collusion problem, until I read the comments on collusion which add an extra condition: "we need votes to be so private that even the person who made the vote can't prove to anyone else what they voted for."<p>However, this condition is not enough, it doesn't cover the case where someone is being watched while they are voting. The condition should be something like:<p>"we need votes to be so private that only the voter can know what he voted on"<p>I don't see many solutions to this, except the '19th century' way of voting at a voting station, or a voting machine that can read minds. Obviously it would be pretty strange inputting your vote through thoughts, especially since this system might never allow you to confirm what you voted on.<p>However, I guess even ballot voting could suffer from collusion since it's not fundamentally private. You could take a hidden camera into the voting box and record your vote, and afterwards get a payout.
I've always enjoyed the concept of exponential costs for incremental gains. I've had the benefit of only really utilizing them in strong-identity systems (games) but I've considered some of the drawbacks of using them in weak-identity systems and come to a few ideas:<p>1) A participation barrier can be used to prevent identities from participating until they've overcome something. A simple example would be an age barrier (identity must exist for some period of time before participating) which prevents spinning up multiple identities on-demand to try to increase voting power. Ideally, in practice there would be multiple barriers that would be naturally-occurring for a real participating identity but too expensive to create/maintain several identities.<p>2) Similar to (1), have an ongoing cost to maintaining identities. Something such as a "subscription fee" may serve as a deterrent as the value of one identity needs to be weighed against the cost of maintaining it. This can be made additionally effective if the issue being voted on recurs every period rather than one-time (i.e. revisiting regulation votes every cycle instead of just voting once and having the regulation remain until stricken). For the normal participant, the value of this subscription could be offset by access to a non-scaling benefit - i.e. access to private content/events.<p>3) While the article focuses on the economics in terms of dollars there's the very real question of allowing votes to be purchased with other forms of currency that either complement or replace traditional currency. This is common in MMOs where you can have multiple characters (weak identities) each with their own in-game currency that can be acquired from in-game activities and may or may not be exchanged with real-world currency depending on the stance the owners of the game take.<p>Suffice it to say there aren't really silver bullets to "general purpose quadratic payments with weak identities" but you could create some limited-purpose constraints that are particular to the problem/community and make some strides from there.
It seems like solving the Sybil problem (a prerequisite to this scheme having any use whatsoever) might be something better on which to focus attention first, as it has other useful applications, even in one-person, one-vote democracies.
Really great to see Vitalik working with Glen Weyl's ideas. Weyl's collection of "radical ideas" may actually become more than just thought experiments.
The main topics of interest here are how non “quadratic” voting converges to quadratic via influence peddling or subquadratic in the form of rigged/kabuki elections.