A New Way of Voting That Makes Zealotry Expensive

(Bloomberg Businessweek) -- An intriguing new tool of democracy just had its first test in the real world of politics, and it passed with flying colors.

The tool is called quadratic voting, and it’s just as nerdy as it sounds. The concept is that each voter is given a certain number of tokens—say, 100—to spend as he or she sees fit on votes for a variety of candidates or issues. Casting one vote for one candidate or issue costs one token, but two votes cost four tokens, three votes cost nine tokens, and so on up to 10 votes costing all 100 of your tokens. In other words, if you really care about one candidate or issue, you can cast up to 10 votes for him, her, or it, but it’s going to cost you all your tokens.

Quadratic voting was invented not by political scientists but by economists and others, including Glen Weyl, an economist and principal researcher at Microsoft Corp. The purpose of quadratic voting is to determine “whether the intense preferences of the minority outweigh the weak preferences of the majority,” Weyl and Eric Posner, a University of Chicago Law School professor, wrote last year in an important book called Radical Markets: Uprooting Capitalism and Democracy for a Just Society. (I wrote about the book last year in this article in Bloomberg Businessweek.)

This spring, quadratic voting was used in a successful experiment by the Democratic caucus of the Colorado House of Representatives. The lawmakers used it to decide on their legislative priorities for the coming two years among 107 possible bills. (Wired magazine wrote about it here.)

“It’s been a really fun collaboration,” says Representative Chris Hansen of Denver, who is chairman of the state House Appropriations Committee and led the experiment. In the past, Hansen says, it was hard to get a read on what lawmakers really wanted. Each would pick 10 or 15 bills, and the bills that were chosen by the most lawmakers would be adopted as the platform, but it was always hard to get a sense of what really mattered to them. Then, Hansen says, he was introduced to Weyl and quadratic voting by a mutual friend. “He said, ‘Oh, gosh, have I got the thing for you.’ ”

In this year’s experiment, the 41 lawmakers in the Democratic caucus were given 100 tokens each to allocate among the 107 bills. No one chose to spend all 100 tokens on a single bill. Many of them spread their votes around widely but thinly because it was inexpensive to do so—one vote is just one token. The top vote-getter by a wide margin turned out to be a bill guaranteeing equal pay to women for equal work. “There was clear separation” of the favorites from the also-rans, Hansen says.

The computer interface and other logistics were provided by Democracy Earth, which describes itself as a borderless community and “a global commons of self-sovereign citizens.” The lawmakers had more immediate concerns—hammering out a party agenda. “Some members were more tech-savvy,” Hansen says. “Some started skeptical but came around. I was pleasantly surprised. There was this feeling of ownership—your voice being heard.”

I recently wrote about the democratic benefits of ranked-choice voting, in which voters rank all the candidates in a race and votes are reassigned from the lowest vote-getters to the higher finishers until someone winds up with a majority. But although ranked-choice voting is gaining in popularity, it traces its roots back to the 19th century. Quadratic voting is much more of a break from the past. “This is a new idea, which is rare in economic theory, so it should be saluted as such, especially since it is accompanied by outstanding execution,” George Mason University economist Tyler Cowen wrote in 2015. (He did express some cautions about it as well.)

The math behind why quadratic voting works is a bit tricky, but the intuition is clear. Imagine one extreme, in which expressing a strong preference was not costly at all—you could use your 100 tokens to buy 100 votes instead of just 10. That would make it too easy for highly opinionated voters to dominate. At the other extreme, if you made extra votes very costly—say, by a cubing rule that made two votes cost eight tokens and three votes cost 27—then people would be overly discouraged from expressing their preferences.

Quadratic voting “is the one vote pricing rule under which voters who intend only their own gain are led, as if by an invisible hand, to advance the interests of society,” Weyl and University of Chicago statistics professor Steven Lalley wrote in a paper last revised in 2017.

To contact the editor responsible for this story: Eric Gelman at egelman3@bloomberg.net

©2019 Bloomberg L.P.