Musings on electoral systems: Bucklin voting

The United States has a reputation for being a barren wasteland in terms of electoral system design: every office at every level of government seems to be chosen by first-past-the-post or its multi-member variant (the block vote), with the occasional use of the party block vote (eg. for presidential electors) and the two-round system (most associated with Louisiana, but recently adopted in California). Even a simple reform such as the alternative vote is limited to such über-progressive hubs as San Francisco and Cambridge, Massachusetts. The nation has, however, experienced periods of frenzied electoral reform, with the 1910s (the height of the Progressive Era, and of the socialist and single-tax movements) being perhaps the most experimentative decade.

Bucklin voting takes its name from James W. Bucklin, a prominent civic leader in Grand Junction, Colorado. From biographical information available online, he appears to have been the very model of a turn-of-the-century progressive, hostile to political machines and saloons, and supportive of prohibition, non-partisan elections, and commission government. Bucklin voting, or at least one of its variants, was invented by him for use in his home town after it adopted a reformist city charter in 1909. It spread rapidly across the country (often adopted simultaneously with female suffrage or new city charters), and was used in large cities such as San Francisco, Portland, Seattle, Denver, and Cleveland before its advance was abruptly stopped and rolled back around 1920 (more on that below).

In the version used in Grand Junction, voters were able to cast a first-, second-, and third-preference for candidates for an office (write-in votes appear to have been allowed). If no candidate had a majority of first preferences, the second preferences of all candidates would be added, and then the third preferences, at which point the plurality winner would have to be elected if no majority was found. Other variants involve eliminating low-placed candidates, allowing more preferences, or allowing multiple preferences at the third stage. (Oklahoma briefly implemented a variant, the only Bucklin method ever tried at state level, which counted lower-preference votes at fractional values: one-half for the second preference and one-third for the third preference.) Bucklin voting was most commonly compared to the alternative vote, which also involves preferential voting for a single office. Unlike its competitor, Bucklin voting doesn’t necessarily eliminate low-placed candidates, allows the voter to cast multiple votes at later stages of counting, and allows for supporters of the two main candidates to express preferences for those in third place or lower. The latter feature was decisive in the 1909 mayoral election in Grand Junction, in which the second and third preferences of the candidates who finished first and second on first preferences helped to elect the candidate who place third on first preferences.

Groups which lobby for electoral reform in the United States typically call for the alternative vote for single-winner offices. The largest such group, FairVote, claims that Bucklin voting was dumped because voters failed to express preferences below the first, making elections function the same as they would under first-past-the-post. Actual data from major elections, however, demonstrates otherwise. The reason for Bucklin voting’s demise was probably the reason that so many other electoral reforms were repealed in the United States: racism and anti-communism. Heterodox electoral systems such as the single transferable vote were often blamed for letting African-Americans and socialists into municipal legislatures; a key feature of the campaign for a ballot measure to repeal STV in Cincinnati in 1957, for example, was the threat of a ‘Negro mayor’. In that regard, it is interesting to consider that the first experiment with Bucklin voting, in Grand Junction in 1909, produced a Socialist mayor, that the 1915 Cleveland mayoral poll almost elected an anti-war, pro-labour Democrat, and that Bucklin voting’s rollback coincides exactly with the First Red Scare. (It’s no coincidence that the other wave of repeal of alternative voting systems occurred in the 1950s – the era of McCarthyism, or that the widespread banning of electoral fusion took place during the 1890s, after the rise of the Populist Party and the entrenchment of Jim Crow.)

Bucklin voting has the following advantages: it can provide for multiple candidates of the same party or political tendency to contest a single office, thus negating the need for primaries; it avoids the lesser-of-two-evils tactical voting made necessary in first-past-the-post and runoff systems; it shares with the alternative vote the incentive for candidates not to alienate supporters of other candidates, whose preferences they might need; and it requires, in most cases, the winning candidate to obtain votes from a majority of voters. The experience from the United States in the 1910s suggests that it might be a way of breaking Duverger’s Law – instead of being mere sources of preferences for the major parties, as they are under the alternative vote, third parties can, given a double-digit percentage of first preferences and good transfers from those finishing above them, win single-winner elections. In the French presidential election of 2007, for example, it is conceivable that second and third preferences from supporters of Nicholas Sarkozy and Ségolène Royal might have catapulted François Bayrou (the third-place-getter, who would have beaten either Sarkozy or Royal approximately 60-40 in the runoff) to victory.