Welcome to a new occasional series covering what academics have to say about politics, elections and public opinion. As with most things in life, academic research comes in various flavours, including the good, the bad and the stating the bleeding obvious (though investigating ‘what everyone knows’ does have a role, as just sometimes it isn’t true after all).
Today’s selection is about the order in which names appear on ballot papers can affect election results.
First, we have some venerable research from the last 1970s which, after highlighting that way back in 1910 there was concern in the US about people changing their names in order to get high up the ballot paper, looked at the issue and concluded that the order of names has an impact. The impact it found in American elections was lower the more high profile the election is, but even so, “In general elections the bias is the smallest and may be confined to the first place on the ballot; the advantage is thought to be about 5 percent above the expected results under the assumption of totally unbiased position voting”. (Source: Avichai, Yakov, “Equity in Politics: Name Placement on Ballots.” American Bar Foundation Research Journal 4.1 (1979): 141-178.)
The idea that ballot paper ordering matters more for candidates who the public know little of is reinforced by a study of the Californian Governor recall election from a few years ago (the one that saw Arnold Schwarzenegger elected). Minor candidates were helped or hindered by their ballot paper location, but the front runners were not affected. (Source: Ho, Daniel E. and Kosuke Imai. “Randomization Inference With Natural Experiments: An Analysis of Ballot Effects in the 2003 California Recall Election.” Journal of the American Statistical Association (2006).)
There is more on the US research at Larry Sabato’s Crystal Ball website, but turning to British research the effect of ordering has been repeatedly found in different research.
The twist in the British context is that with multi-member ward elections under first past the post in many parts of the country, the impact isn’t just on choice between parties (Party A candidate at top of the ballot paper does better than Party B candidate half-way down) but also within parties: “Within each party’s slate of candidates those placed higher on the ballot paper have a clear advantage over those lower in the alphabetic order and hence lower in ballot paper order … A candidate positioned highest on the ballot order was thus more than twice as likely to finish at the top rather than in second place on the party slate. The advantage is even greater when first is compared with third alphabetically: the former is four times more likely to have finished ahead in votes received.” (Source: Rallings, Collin, Thrasher, Michael and Borisyuk, Galina, “Unused Votes in English Local Government Elections: Effects and Explanations.” Journal of Elections, Public Opinion and Parties 19.1 (2009): 1-23, which also includes a long list of references to other research.)
Less researched are the possible drawbacks of moving from alphabetic ballot papers to ones where the candidates are in a random order. The main one is that if you know who you want to vote for, it is harder to find their name on a random ballot paper. That matters very little in, say, a single vacancy, three candidate election, but as the number of seats and candidates rises, putting extra hurdles in the way of completing ballot papers may cause the number of votes recorded on them to drop.
Conclusion for our current system? The lower profile the election, the more extra effort is required to promote candidates lower down the ballot paper. Oh, and get that deed poll form ready. Aaaron-Aardvark has a certain ring to it, don’t you think?