Random Ballot

Search organisations working on Random Ballot in the Democracy Landscape →

A random ballot (also called random dictatorship) decides a vote by drawing one cast ballot at random and adopting whatever it chose. In Wikipedia's words, it "is a randomized electoral system where the election is decided on the basis of a single randomly selected ballot."¹ An option backed by 30% of voters therefore wins 30% of the time.

Not the same as sortition. Both use a lottery, but for different jobs: sortition draws people (a jury, a citizens' assembly), while a random ballot draws an outcome from the votes people have already cast.

The case for it is fairness. Allan Gibbard showed in 1977 that it is "the unique social choice rule that treats all voters equally while still being strategyproof in all situations"²: nobody gains by voting for anything other than their true favourite. The case against is just as plain: in Gibbard's own words, such a method "leaves too much to chance",³ and it is "rarely, if ever, proposed as a genuine electoral system".⁴ It is mostly discussed as a thought experiment, notably Akhil Reed Amar's 1984 Yale Law Journal article "Choosing representatives by lottery voting".⁶

Majority rule versus a fair chance for the minority

Majority rule gives the losing side nothing, however large it is and however often it loses. A random ballot gives every side a chance of winning in proportion to its support: 30% of the vote means winning 30% of the time, not never.

Philosophers have argued about the same trade-off under the name weighted lottery, as a fix for the objection that "the greatest good for the greatest number" ignores the minority entirely. Philosophy Vibe's short explainer puts it in voting terms with a group of ten choosing dinner: seven want burgers and three want sushi, so instead of a show of hands, draw from a bag of seven red balls and three blue.⁷

The same video gives the case against, and it applies directly to elections: the lottery leaves a real chance of an outcome nearly everyone agrees is worse. Its example is a rescue where the draw sends the helicopter to save one person instead of five.⁸ That is why a random ballot is easier to defend for small, repeated group choices, where an unlucky draw is cheap and the minority's turn comes round, than for one-off, high-stakes decisions. Mathieu Doucet's "Playing Dice with Morality" (below) is a fuller academic critique of the weighted lottery.

Extensions

Ways a random ballot can be combined with other mechanisms:

After another voting method, as an acceptability filter. Any method that can tell which options most voters reject can screen the field first, with the winner then drawn by lot from what's left, weighted by support. With Majority Judgment, strike off any option whose median grade is "Reject"; with approval voting, any option fewer than half the voters approve; with ranked ballots such as preferential voting, any option a majority ranks last. The minority keeps a fair chance, but only among options the group as a whole can live with, which answers much of the "clearly worse outcome" objection above.
As a tiebreaker. A common real-world role: another method decides, and a random ballot settles only a tie.⁵

Random Ballot

Majority rule versus a fair chance for the minority

Extensions

Further reading

See also