Alphabet Soup: Randomized Ballot Order and the Representation of Marginalized Candidates

Ballot order has the power to shape electoral outcomes: even when randomly assigned, candidates listed first receive more votes. Meanwhile, candidates from historically marginalized groups—such as women and ethnic minorities—tend to fare worse in low-salience or low-information contexts. We examine whether being listed first—an increase in salience—benefits these candidates more or less than their white and male counterparts. Leveraging a natural experiment from over 29,000 California local elections between 1995 and 2021, in which ballot order was randomly assigned, we estimate the causal effect of being listed first on vote share by candidate gender and ethnicity. We find that while all candidates receive a premium from appearing first, non-white candidates benefit significantly more than white candidates. This advantage varies with the partisan and racial composition of the electorate and with election timing. We close by discussing the implications of election design for equality of representation.