The results so far of the Guardian’s bird of the year poll have not been without controversy. The Australian white ibis, a bird that is disliked by many who encounter it, took an early lead and has maintained that lead for more than a week. While this seems like a strange result, it makes sense when you think about the options provided to the voters. With so many birds to choose from, the voting system used has a tendency to produce a winner who has a committed support base, even if that option also has a lot of opponents.
The vote was conducted using the first-past-the-post (FPTP) voting method – everyone gets one vote, and the candidate with the highest vote wins. This system works fine when there are just two candidates running in an election, but when you get more than two, it inevitably results in candidates winning with less than a majority of the vote. The ibis is sitting on 13.6% of the total vote, with the magpie coming second on 11.1%. The top two birds combined have received less than a quarter of the total vote.
FPTP causes problems even when there are a small number of candidates and we know who has a chance of winning. Voters in the UK are used to voting tactically to ensure their vote ends up with someone with a chance of winning. But this method requires voters to accurately predict who has a chance of winning, and will hinder the growth of smaller parties who may not immediately have a chance of winning. Even with tactical voting, candidates regularly win with less than 50% of the vote.
The ibis was helped by the large number of birds on offer. If you like the ibis, you had a clear choice, but anti-ibis voters were scattered among many options.
The vote for some birds was split between similar species. The fairywren vote was split between the east coast superb fairywren, who is coming fifth, and the splendid fairywren, who is coming 10th. They would come close to overtaking the kookaburra for third place if their votes were combined.
At least four kinds of cockatoo were included, with none of them making the top 10. Combine their vote and you get 7.5% of the total, enough to tie for third with the kookaburra. Indeed, eight types of parrot received a total of 14.8% between them – more than the ibis.
FPTP elections with many candidates tend to favour divisive candidates who may draw a lot of opposition but have a concentrated support group. The ibis has played such a role – many dislike it, with a minority passionately voting for it.
The University of Melbourne student union was so outraged by lack of preferential voting in the bird of the year poll that it passed a motion condemning Guardian Australia and calling the decision “un-Australian”.
“UMSU affirms the right of all Australians to rank people, places and things, but not necessarily in that order, at any time, at any place, anywhere in this great southern land,” the motion reads.
There are parallels the early stages of the 2016 Republican primary in the US. Donald Trump was one of 17 candidates. While Trump was the clear leader in most polls, he rarely passed 40% of the vote, and sometimes polled in the 20s. His supporters rallied to his position but his opponents were scattered among numerous other candidates. Trump later picked up more support, but lacked majority support for most of the primaries.
It’s fun to compare the bird poll to an election, but a more relevant comparison might be to how award shows decide their winners.
The Academy Awards are decided by an electorate of 6,000 industry professionals. As with the bird poll, there are many potential nominees to choose from. There are two stages. Firstly, members of the Academy vote on potential nominees, to choose five nominees for each category (and 10 for best picture). The Academy uses the single transferable vote to decide the winners – similar to the system used for the Australian Senate.
Academy members then choose from among the nominees using FPTP (although preferential voting is used for the best picture winner).
A fairer voting system for a bird poll wouldn’t need to be too complicated to be reasonably fair. You could conduct it in two rounds. The first, say, five options (in this case, the ibis, magpie, kookaburra, tawny frogmouth and superb fairywren) would go through to a second round of voting, with voters then choosing a single winner from that group.
Anyone for a runoff round?
