Robson Rotated

As Sinclair Davidson kindly noted in comments the other day, Gary Nairn, the Special Minister of State, wrote a letter to the Australian Financial Review last Friday to criticise my paper with Amy King on ballot order:

Your “Ballot change urged to counteract donkey vote” (September 18) quotes research claiming to have found that male candidates who were listed first on the ballot received an additional 1.4 per cent of the vote, while the ballot order had no effect for female candidates.

I have some reservations about this research from the Australian National University and University of South Australia.

Donkey voting is generally understood to mean a vote where the elector numbers preferences in sequence down the ballot paper, starting with a first preference for the candidate whose name is at the top. Thus, if there are four candidates on the ballot paper, a donkey vote is cast by numbering the ballot paper 1, 2, 3 and 4, from the candidate whose name appears first on the ballot paper to the candidate whose name appears last.

The article does not discuss subsequent preferences down the ballot paper, it discusses only the possible advantage to a candidate whose name appears first on a ballot paper. For these researchers to suggest that our electoral process is “manifestly unfair” because of the impact of donkey voting is quite frankly, manifestly misleading.

Since 1984, the Commonwealth Electoral Act has required the order in which candidates’ names appear on a ballot paper to be determined by a double randomisation draw conducted after the close of nominations. This means there is no advantage for a candidate from the alphabetical order of their last name, the name of their party or the time their nomination was lodged with the Australian Electoral Commission.

Introducing a Robson rotation system, as proposed by the researchers – printing multiple versions of each ballot paper, rotating positions so each candidate tops the ballot as often as every other candidate – would serve only to cause voter confusion and increase the rate of informal voting.

Gary Nairn,
Special Minister of State,
Canberra, ACT.

Nairn’s suggestion that the current system is fair depends on how you think about fairness. As we point out, randomisation is fair before the ballot draw takes place. After the draw, the first placed candidate has a better chance of winning for reasons entirely unconnected with competence. And my understanding is that the introduction of Robson rotation in the ACT and Tasmania wasn’t associated with a rise in informal voting in those states.

Then again, perhaps if Amy and I really wanted to get the law changed, we probably should not have opined that Nairn would have lost his seat in 1998 but for the fact that he drew the #1 spot on the ballot.

This entry was posted in Australian Politics. Bookmark the permalink.

6 Responses to Robson Rotated

  1. Sacha says:

    In his letter, Gary doesn’t really discuss the point of your research.

  2. Fairness to candidates is one issue. A more important one in my opinion is a robust process that gives electors the representative they truly chose. That means minimising the risk that the outcome will be influenced by random / anomolous factors such as ballot order. And that would suggest the Robson rotation. Having a candidate benefitted by ‘a double randomisation draw’ rather than by his surname or some other arbitrary criterion may be considered fair by Gary Nairn. But it’s not fair on the voters.

    The only time Nairn engages with the real argument at all is in his last paragraph. If he’s right and multiple versions of ballots confuse people (which is certainly feasible as it makes it harder to follow ‘how to vote’ cards) then that’s certainly a downside of Robson rotation, but the risk of that needs to be weighed against the undeniable problem of having one candidate get an arbitrary donkey vote boost.

  3. Western Suburbs Magpies says:

    I am sure there needs to be consideration of the fact that an increase in the number of candidates increases the rate of informality. (This is understood in anecdotal terms anyway).

    An increase in accidental informal voting surely has a major negative impact on the democratic process (as does a high rate of preferences exhausting), and if you make it harder to distribute and understand how-to-votes, then you are likely to get that.

  4. David Walsh says:

    Can I use this thread to advocate optional preferential voting?

    I think Robson rotation has plenty of merit. As everyone seems to agree that the donkey vote is a factor. But as others have pointed out, it makes it difficult to follow how-to-vote cards. And HTV cards serve an important role in making sure voters lodge a formal vote, under the difficult formality rules of the CPV system.

    Which leads me to conclude that OPV is the answer. It makes voting simple for voters; without the need obtain HTV cards. Which in turn makes Robson rotation viable.

    It also addresses WSM’s concern of a high informal rate. (A high exhaust rate isn’t really an issue. As these would be voters who, under CPV, were only stating an artificial choice anyway.)

  5. Andrew Leigh says:

    David, sounds sensible to me. I’ve never understood the rationale for CPV (why should the validity of my first choices be cointingent upon me making subsequent choices?).

    Incidentally, my understanding was that informal voting didn’t increase with the introduction of Robson rotation in the ACT and Tasmania. Surely if this is true (and I haven’t checked the data), then it’s the decisive rebuttal to Nairn.

  6. backroom girl says:

    It’s even more decisive if you consider the fact that Robson rotation is currently only used in multi-member electorates, which means that you have much larger ballot papers with lots more names on them. While fixed-order ballot papers and how-to-vote cards obviously make it much easier for illiterate people to vote (as long as they know which party’s HTV they hold in their hand), most people who can read should be able to cope with a rotated ballot paper with or without HTV cards. You just need some other strategy for capturing the preferences of illiterate voters, as I think already exists anyway.

Comments are closed.