Random Ballot

Random ballot (also called random dictator or lottery voting) is a non-deterministic electoral procedure: every voter ranks or names a preferred candidate, one ballot is then selected uniformly at random, and that ballot determines the outcome. Although rarely used for high-stakes public elections, it is a benchmark method in social choice theory because it satisfies several properties that deterministic systems cannot jointly achieve.

The method is strategyproof in the sense formalized by Allan Gibbard's 1977 theorem on randomized voting rules: Gibbard showed that the only strategyproof and ex post efficient probabilistic voting rules are random dictatorships (with random ballot as the canonical example over voters). It also satisfies proportionality in expectation — a faction comprising x% of the electorate wins with probability x% — which makes it appealing to theorists concerned with minority representation.

Random ballot is used as a tie-breaker in several real jurisdictions. Under New Mexico law, for example, tied state-level races may be resolved by a game of chance; similar provisions exist in other U.S. states for local offices. The Republic of Venice historically used a complex multi-stage process combining lot and election to choose the Doge from 1268 until 1797, partly to frustrate factional manipulation.

Critics raise several objections:

• Legitimacy: outcomes determined by chance can appear arbitrary and undermine the perceived mandate of the winner.
• Volatility: a single drawn ballot can elect a fringe candidate supported by a small minority.
• Accountability: because any voter's ballot might be decisive, incentives to cast an informed vote may shift in unpredictable ways.

In contemporary practice, random ballot appears mainly as a tie-breaker, as a thought experiment in mechanism design, and occasionally in sortition-adjacent proposals for citizen assemblies. It contrasts with sortition, which selects officeholders randomly from the population rather than from cast ballots.