Listing candidates alphabetically on ballot papers has been proven to be unfair

Having failed to get elected in two earlier campaigns, the candidate previously known as Cheryl Talamaivao has wisely rethought her tactics. But deciding to elevate her position on the ballot paper by tacking her great-grandfather's surname in front of her own may turn out to be something of an own goal in her bid for election to the Henderson-Massey Local Board.

As Brown-Talamaivao, she is now number six on the ballot paper, instead of the second-to-last at number 27 she would have been as plain Talamaivao. But the experts, in what in Australia is referred to as the "donkey vote" advantage, reckon that being second to bottom is a good place to be, and argue that candidates at the tail end of a long list of contenders share a similar advantage to those at the beginning, over the poor suckers stranded in the middle.

Of course, in Australia, a mix of both compulsory and preferential voting forces the reluctant and the ignorant to front up, with some just going through the motions, ticking randomly from the top or bottom of the list.

Still, Ms Brown-Talamaivao and a rival candidate in the ward, Ann Paia - who has tagged her maiden name Delga to the front of her surname - are correct in their belief that the traditional alphabetically ordered voting papers are not fair to all contestants.

Advertisement

Existing Auckland councillors were warned about this in April, but chose instead to save the $100,000 extra - from a total electoral bill of around $5 million - it would have cost to print ballot papers in truly random order, and voted to stick with the tried, but definitely untrue, alphabetical ordering.

The unfair advantage, in particular to those at the top of the ballot paper, is well studied and proven. The Local Government Commission examined the outcome of the 2007 New Zealand local elections and found that those listed alphabetically at the top of the ballot papers and candidate profile booklets "were up to 4 per cent more likely to be elected than those whose names were later in the alphabet". Names like Anae, Brown, Brewer, Casey and Coney.

The commission also noted "a significant bias in favour of candidates in the left column of voting documents".

A similar effect was found in an analysis of the 2010 Greater London local elections by City University researchers. After examining the fate of 5000 candidates, they found that "ballot position did indeed strongly influence the number of votes received by candidates". There was evidence that "the strength of this effect is sufficient to overcome voter preference for party ..."

Dunedin councillors voted this year to go random. Before a unanimous vote in favour, Mayor Dave Cull said, "If anyone were to vote for the continuation of alphabetical listing they would have to be voting for a lack of evidence, putting perceived convenience above fairness, or possibly acting in self-interest."

The southern politicians were following the example of the local Otago Health Board, which adopted random ballot papers in 2007 after chairman Richard Thompson analysed the 2004 board election results and found that if your name began with A you had a 53 per cent chance of being elected, compared with the overall chance of 29 per cent. "If your name is Atholl the Aardvark," he wrote, "you had more than twice as much chance of being elected in an alphabetical listing than poor Rocky the Raccoon."

Dunedin and Wellington city councils will be among the 25 or so local administrations using this fairer system in elections. Some, like Dunedin and Wellington, will use the fully random system, where the list on each ballot paper is individually randomised. Others have replaced the unfair alphabetical listing with the equally unfair pseudo-random system. This involves drawing the candidates' names out of a hat, and listing them on all ballot papers according to the luck of that one draw.

To Auckland Council, this is all too hard. If they want hard, they should look to Afghanistan, where well over half the voters in this week's election are illiterate. Instead of a debate over their place on the ballot paper, candidates had to participate in a draw for a symbol to represent them on the voting sheet. Candidates literally had to draw three symbols out of a hat and choose one to best represent them - an apple, cow, icecream cone, duck, bicycle, cellphone among them. Now that is hard.