(This is really quite an engaging conversation!)
It seems to me we’re talking somewhat at cross-purposes here. The purpose of QV is modest and limited: gauge the intensity of the preferences of the voters. You could try to achieve that simply by giving everyone a number of votes > 1 and allowing them to allocate these accordingly. But just like a log/exponential scale amplifies the effect of small changes, quadratic voting allows you to produce this insight without requiring to fix the number of votes (both in total and per voter): demanding that voters pay for an extra vote, and pay ever more for subsequent votes really means they’re serious when they do so.
The fact that the quadratically increasing cost also makes domination of the vote by wealthy people unlikely is a side-effect, and I would argue one that does not materially affect the democratic process.
There’s absolutely no need for me to limit your power.
In a market, where transfers of tokens are a sign of demand certainly not. But as far as democracy is concerned, I think there are other considerations. Most modern democratic societies (and I’d say all of them but I am simply being cautious here :-)) start from the principle that every eligible citizen has an equal stake in it, irrespective of their wealth, status, intelligence etc. A system that allows people to buy an unlimited number of votes would violate this. A poor guy may feel as strongly about abortion, prohibition, environmental protection or low taxation as a wealthy guy, but one of them will be able to buy many votes to push his ideas, and the other won’t.
I think that is quite a different situation than one in which they might both desire a Bugatti, but one can afford it and the other can’t.
So in a democratic process, there is a need to limit power — and to ensure that every citizen has (more or less) equal power to influence policy.
You support quadratic voting but oppose full vote buying/selling.
I don’t think I said I oppose vote buying/selling (I don’t — or at least not under certain conditions.
A good few years ago I had a series of discussions about road congestion and the use of road pricing to try and allocate the scarce resource better. The typical left-wing objection to road pricing is that “it’s the rich who will be able to drive when they want”. I think this is correct, but it’s not a good argument.
My counter-argument was: why not give every person over a certain age a nominal share in the road network. They can use this share to purchase access to the road — or they can choose to sell it to someone else. So you start with equal shares in a common good, and then you operate a market in which each owner of such a share can decide to use it themselves, or sell it to someone to whom it is more valuable.
Something similar could be used to set up vote trading. Give everyone the same number of votes, and allow them to either use them or sell them.
I think your model really breaks down because your dedicated power transfer to Frank, signalling that you want him to grow more artichokes, becomes generic. That is not a design feature of pragmatarianism, I think. The argument that there must be other artichoke growers doesn’t cut much ice. There is no guarantee that there will be, or that they will produce the same quality of artichokes as Frank. And the argument that that is just tough, and you need to make a trade-off between what is more important to you: the best artichokes but abortion made illegal, or legal abortion but poorer (or no) artichokes, is equally weak.
There is no need to tie commercial markets with policy making in this way, which achieves lower overall welfare than if they are kept separate.