(This essay was published in South China Morning Post on 7 January 2015.)
“One person, one vote” is a great invention, but can it solve today’s practical political problems?
Hong Kong is inflicted with political gridlock that has stalled numerous public policies and projects. The pan-democrats blame the lack of fully democratic political arrangements, the establishment blames the pan-democrats, and some blame recent progress towards more democracy.
Even mature democracies struggle with gridlock and how to weigh the views of supporters and foes on issues. Many now let voters decide on big questions, such as same-sex marriage, through ballot referendums. But this allows an indifferent majority to outvote a passionate minority. The outcome can also be influenced by recent events.
Consider the outcome of the Occupy Central civil referendum in June, which returned close to 800,000 valid votes supporting the public nomination of candidates for the 2017 Chief Executive election. In opposition, “The Alliance for Peace and Democracy” campaign in July and August collected more than 1.5 million signatures.
What could one conclude from the results of these unofficial exercises? Many people probably voted (or signed) without a strong stake in the outcome, although it was of considerable signiﬁcance to minorities on both sides. If the minorities care a lot, and the majority a little, it is not obvious that “one person, one vote” can tell us the true balance of thinking.
Common belief also suggests voters in the Occupy referendum were probably heavily influenced by the release of the white paper on Hong Kong’s “one country, two systems” policy. As a consequence, the vote count might not be robust.
Economists have given much thought to the problem of aggregating people’s preferences, taking into account their intensity. A University of Chicago economist, Glen Weyl, recently developed an ingenious new mechanism to do this, called Quadratic Vote Buying (QVB).
It works this way: Three persons, Ann, Bo, and Chou, can each vote on whether the government should develop a new town by purchasing as many votes as he wants at a price equal to the square of the number of votes that he buys. So, one vote would cost $1, two votes $4, three votes $9, 10 votes $100, and so forth.
The outcome would be determined by majority rule, so if Ann bought eight votes (for $64) in favor of development, Bo bought four votes ($16) against, and Chou bought two ($4) votes against, the program would be approved by a vote of eight to six. By contrast, with ordinary voting the project would be rejected two to one.
The payments would then be distributed — to each other, in equal shares. The total amount paid would be $84 so each would receive $28. Ann’s net loss would be $36, but he beneﬁts from the new development, while Bo and Chou would net, respectively, $12 and $24, which would compensate for the impact of a loss, in contrast to the winner-take-all under “one person, one vote.”
Bo would be a little worse off, receiving $12 versus the $16 he paid out, but over time, as votes were held on other policies and projects, he most likely would gain more than he lost. This again contrasts with “one person, one vote” where a minority can be outvoted repeatedly.
QVB works because it forces voters to pay for the impact of their votes on other people, in the same way a waste charging scheme forces people to pay for the effects of waste disposal on others.
An objection to QVB is that wealthy people would have greater inﬂuence on outcomes than poor people. But this concern can be mitigated.
First, because the price of voting is the square of the number of votes cast, the price goes up extremely quickly. Second, the collected funds are returned to all voters so if Mr. Li Ka-shing pays $1 million to support a new development, he will get little of his money back, while everyone else will collect whether he wins or not.
And third, a cap on the number of votes people could cast would weaken the inﬂuence of wealth.
The growing prevalence of diverse minorities with intense interest in narrow issues and a largely disinterested public require an updated method for determining citizens’ preferences.
QVB is forward thinking, the first to target the core voting method rather than tampering with regulatory constraints to mitigate the negative consequences of “one person, one vote”. It is able to reflect variations in preferences and intensities among voters.
This method is especially useful for resolving highly controversial decisions: developing the Northeast New Territories, redeveloping part of the country park, finding landfills, preserving heritage structures, siting columbaria, banning life chicken imports, and numerous other issues.