Quadratic Voting (QV) is a new way of allowing people to express their preferences. Unlike traditional voting systems, where voters simply vote yes/no or at best rank their choices, QV lets people speak louder on the issues most important to them and agree to compromise elsewhere. This improves information quality and protects minority interests while giving people more democratic freedom and discouraging polarization.
The real breakthrough here is that people can influence issues in proportion to how much they know and care about them. And by gathering more nuanced perspectives on what really matters to people, even large and diverse groups can quickly find what everyone agrees on, consensual solutions with broad and strong support.
How does QV work?
Every person can vote on a set of collective decisions using tokens earned on the platform. Instead of a rigid yes/no vote on each issue, people can move their voting power (tokens) across the ballot to place multiple votes on some issues and fewer or no votes on others. But the twist is that it becomes increasingly expensive to put more votes on any particular issue.
If someone wants to vote just once on each issue, it works like traditional voting: they would spend one of their tokens on each vote. But to place a second vote, that costs three tokens. A third vote, an additional five tokens. And so on.
The underlying principle is that the total cost in tokens is calculated as the square of the number of votes cast. Note that placing three votes costs a total of nine tokens because three squared equals nine. The graphic below illustrates this dynamic.
What are the benefits?
Quadratic Voting is the best way for groups to make decisions because of its unique property that the marginal cost of each additional vote increases linearly with the number of votes. Remember that placing a third vote costs five more tokens, and a fourth vote costs another seven tokens (for a grand total of 16 tokens, which is four squared).
To understand why this is a good property, let us first consider how it works in traditional voting, where every person gets only one vote per issue. Here the marginal cost of another vote is effectively infinity; someone who values a vote more than others cannot buy more votes. Thus the system cannot distinguish between strong and weak preferences and ends up favoring weak and diffuse majority interests over strong and concentrated minority interests.
The alternative is usually a market system where the marginal cost of a vote is constant, meaning people can buy as many votes as they wish at the same price per vote. This time the advantage goes to the strong, concentrated interests; those with the strongest preference can simply purchase the most votes and tip the scale in their favor.
QV perfectly balances these two systems. Because every vote is more expensive, people purchase votes in proportion to how much they care about the group choosing one option over others. This way there is no advantage to those who care a lot or those who care only a little. The total number of votes for a given issue is simply the sum of the strength of the preferences of the people who voted, creating the possibility that a small but passionate minority can prevail democratically against an apathetic majority. QV leverages the fact that signals that are more costly are more credible to improve information quality for participatory governance.
QV extends democracy by letting our voices be equal overall, but also flexible, so that we may speak to the different aspects of the world where we actually have interest or expertise. Asking voters to know about every issue is unfair, inefficient, and fails to tap into natural democratic intelligence. The genius of QV is that it incentivizes everyone in a bottom-up process to self-select and coherently weigh in on the issues they know and care about.
One benefit of self-selection is that porous communities without exact boundaries can have legitimate votes on important group decisions. It also gives a balanced advantage to core members of a community who may have strong preferences across many issues. This is because spreading their votes across many issues is a more efficient use of their votes, and so the more votes they will end up making. For a small sacrifice to voting power on their highest priority issue, they can have considerable influence elsewhere.
Traditional voting lets people express extreme views at no cost and cannot rule out those who most people have strong objections to; voters are either on the winning side or the losing side. QV makes zealotry increasingly expensive, for placing more votes on an option quickly exhausts a voter’s budget. Everyone is on the winning side of their most important issues, and still supportive of other choices that win, giving them a rich sense of agency and ownership over the process.
We’ve already seen this outcome in practice, in the Democratic Caucus in the Colorado House of Representatives, who used QV in 2019 to prioritize a long list of spending bills. Before using QV, and so without imposing constraints on representatives’ priorities, every bill seemed to be the most important: Representative Chris Hansen could not see clear preferences between the “big blob” of bills with roughly the same number of votes.
For the prioritization exercise with QV, each member was given 100 artificial tokens to place, for example, 9 votes on one bill (81 tokens) and 3 votes on another bill (9 tokens), or 5 votes each (25 tokens) on 4 different bills. The process captured nuanced preferences on every bill, and rather than a few bills at the top followed by a mess, the result was a smooth prioritization curve. The Representatives discovered the bills they could broadly agree on should be funded.