If the dismal spectacle of the government shutdown has created any openness to changing how we elect leaders—and how they make decisions—a young economist at the University of Chicago is ready with a proposal.
The rather extreme notion, floated by E. Glen Weyl and amplified by his Chicago law-school colleague Eric Posner, would scrap one man, one vote in favor of a complex system of vote-purchasing. Yes, vote-purchasing. Under this system, called quadratic voting—which very much bears the market-oriented stamp of University of Chicago-style social science—citizens could cast one or more votes, but each vote would cost them an amount equal to the square of how many they were buying. So, one vote for a candidate would cost $1, two would cost $4, four would cost $16, and so on.
After the votes were tallied, the majority would win and the money would be evenly divided and returned to voters, serving as partial compensation for the people on the losing side.
Paying to vote is a taboo notion, of course: The Harvard political scientist Michael J. Sandel is just one scholar among many who have argued that the ballot booth should be a market-free zone. But Mr. Weyl argues that, once you get past the knee-jerk opposition, his proposal has advantages that might appeal across the political spectrum.
He and Mr. Posner have promoted it in a series of working papers, articles in academic journals, and in The New York Times and (in a piece by Mr. Posner alone) in Slate. They’ve applied it, with modifications, to votes for candidates, votes within legislatures, and votes by shareholders. Mr. Posner has suggested it could resolve debates over whether New York City should run a bike-share program, and Mr. Weyl has proposed applying it to departmental decisions about faculty hiring.
They’ve even formed a venture called Collective Decision Engines to build software that would help institutions manage nonpolitical versions of quadratic voting. In the case of departmental hiring, fake money might be distributed that professors could spend on the candidates of their choice, according to the quadratic rule. Such a system would cut through lots of politicking and favor-trading, Mr. Weyl and Mr. Posner say. Likewise, leaders of small companies might distribute money to their employees, to be used to vote on initiatives (espresso machine or gym equipment?) or saved.
Among the problems quadratic voting is meant to resolve is that one man, one vote fails to take account of how strongly people feel about an issue—a problem long recognized by economists. For example, a few voters feel extremely strongly about gay marriage (notably gays and lesbians and a small number of activist traditionalists) while the vast majority leans mildly one way or the other. Quadratic voting would allow the passionate minority to make that passion known at the voting booth.
Back-of-the-envelope math, Mr. Weyl and Mr. Posner argue in a working paper called “Voting Squared,” finds that California’s anti-gay-marriage Proposition 8, which passed in 2008 by 52 percent to 48 percent, would have failed if quadratic voting had been used. (They assume that the average member of a gay couple would value the right to marry at $50,000, and the average single gay person would value it at $20,000, but the numbers are far less important than that they far surpass plausible estimates for most heterosexuals, whether supporters or opponents.)
Then there’s the tyranny-of-the-majority problem. The American political system does have elements designed to prevent abuse of power by the 51 percent, such as judicial review, separation of powers, and supermajority rules. But these can be extremely narrowly tailored—courts will review some political decisions involving African-Americans but not all, for example—and they can cause gridlock, as in the case of the rule that requires 60 senators to assent to end a debate before holding a vote.
Quadratic voting, Mr. Weyl argues, “precisely achieves efficient bargains when there are a large number of conflicting parties involved,” something “our political system seems chronically incapable of accomplishing recently.”
Inspired in Rio
Mr. Weyl, 28, is something of a prodigy: He effectively completed the requirements for a Ph.D. while an undergraduate at Princeton University. He earned it in 2008, a year after he graduated, largely because of a university rule requiring that graduate students enroll for a year first. He conceived of quadratic voting while on trip to Rio de Janeiro in 2007. He was considering the favelas, or slums, that sprawl across highly desirably real estate within the coastal city. How, he wondered, would you decide what proper compensation would be for encouraging impoverished people to move from those areas, improving their material well-being while opening the land to productive development?
“Majority rule wouldn’t work at all” as a way of determining a fair deal for such enormous displacement, Mr. Weyl says. You could give 51 percent of the people an attractive incentive “and screw over the rest,” who would nonetheless have to leave. As he experimented with adding market elements to the vote, linking intensity of preference to vote pricing in ways that led to efficient outcomes, a regular quadratic pattern emerged in scenario after scenario.
Only later did he realize that Richard Zeckhauser, a professor of political economy at Harvard University, had co-written an unpublished paper that also hit upon a version of quadratic voting, albeit one that didn’t use money or apply to elections. It used “influence points” through which citizens could “buy” votes for a set of public-works projects under consideration, and the cost of additional votes grew in quadratic fashion.
Mr. Posner says that when Mr. Weyl first described the idea to him, at lunch, “it struck me that if he’s right about it, then it’s relevant to—the solution to—countless legal and public-policy problems I work on.” The idea, he thinks, is most useful in cases where minority and majority interests are in tension but there is no clearly just answer, such as affirmative action, or stricter voter-ID laws, or security rules that disproportionately inconvenience one group. Quadratic voting would provide information about how different groups weight different policies that is far better than any survey could, in addition to producing the result with the maximum societal benefit.
At first blush, a world with quadratic voting may sound like a libertarian fantasy, but neither Mr. Weyl nor Mr. Posner is ideologically predictable where markets are concerned. In a 2013 article in the Northwestern University Law Review, they proposed creating an agency modeled after the Food and Drug Administration that would approve financial instruments before banks could sell them. The Bloomberg View columnist (and former investment banker) William D. Cohan called it “one of the sillier ideas to come down the pike since the onset of the financial crisis five years ago.”
Mr. Weyl has consistently argued that much of what Wall Street does represents a squandering of intellect and energy, and he’s explored ways of tweaking the tax system to discourage young people from entering finance. “On and off, I’ve been accused of being a fascist and a communist, depending on the setting,” he says.
Other economists have devised “efficient” alternatives to voting, but many of these tend to be hugely complex, involving bids, calculations of which voters are pivotal, and payments by the winners to the losers.
Impact of Wealth
To the obvious objection that the wealthy get too much say under a votes-for-cash system, Mr. Weyl and Mr. Posner have several responses.
First, the cost of votes rises so fast that the effect of wealth dissipates quickly. (One thousand votes cost $1-million.) Second, on questions like taxation—you may have to trust them on the math here—the bottom half will tend to protect its social insurance net more avidly than the wealthy will push for a few extra dollars (the marginal value of a dollar is less for a rich person). The impact of wealth would reveal itself most strongly on social issues, but it’s far from clear the rich would form a coherent block. Moreover, the postelection refund would blunt the inequality.
Most importantly, the two scholars say, the proposal must be compared not with an ideal of equality, but with our current messy system —"super PACs” and all. There is “no reason to think” that majority rule produces economic equality, Mr. Weyl says; indeed, it’s “odd” that we assume that’s the case. He even thinks his system would raise the level of political discourse, because political ads today are often aimed at uncommitted, low-information voters; if candidates instead had to persuade people who felt passionately, their ads would be more substantive.
Within legislatures, the system would work differently than in general elections. It wouldn’t make sense for lawmakers to spend their own money. Each lawmaker might get a certain amount of money each term, to be spent either on votes or sent home to the district. Lawmakers would be trading off influence and money for their constituents, but through a market mechanism rather than favor trading: The rationalization of pork!
Alessandra Casella, a professor of economics at Columbia, describes herself as “sympathetic” to the goals of quadratic voting but thinks money introduces incentives that could cause it to unravel: A well-off voter might be tempted to pay uncommitted voters for their (cheap) votes rather than buy (very expensive) additional ones for himself. “When you put money into the equation, I get nervous,” she says.
Mr. Weyl and Mr. Posner say they see no reason why wealthy voters would risk the legal penalties for vote buying or why potential cheaters wouldn’t be scared off by the fear that the poor would just take the money and not vote.
Ms. Casella’s own preference is for “stored votes": In the U.S. Senate, for example, each Senator might get three votes to cast for or against three judicial nominees—but a Senator could save them up and use them against the least liked nominee. The minority would win some fights but couldn’t obstruct the system completely.
Although Mr. Weyl and Mr. Posner think quadratic voting would usher in an era of electoral efficiency, they know they have a lot of persuading to do—even at the level of departmental hiring. “At this point we have no plans to introduce the rule,” says John List, chairman of Chicago’s economics department, in an e-mail, although he says he admires Mr. Weyl’s work in this area. But Mr. Weyl remains hopeful. After all, he points out, “300 years ago, nobody thought democracy was a reasonable form of government.”