Recounts, missing ballots and secret software enliven the Western Australia Senate election

The Australian preferential voting system is not the easiest thing in the world to understand. The elector must show a preference for all candidates listed on the ballot paper.You have “formal votes” and “informal votes”. An informal vote is a ballot paper which has been incorrectly completed or not filled in at all. Informal votes are not counted towards any candidate but are set aside. But it seems that on recounts formerly informal votes can formally be declared to be formal after all. Between one count and another votes can apparently get lost which suggests that there are physical votes which are counted. But it also seems that “software” –  which is held secret – is used to count the votes.That itself makes me wonder as to what discretion and what margin of error such “software” can have? And who does the programming? Especially when margins of victory are swinging between 14 and 12 for opposing candidates and where 1375 ballot papers have suddenly gone missing.

Australia went to the polls on September 7th. In Western Australia, however, senate seats were decided by a mere 14 votes, a result that saw a recount. During the recount, two important events took place: first, a number of votes originally treated as “informal” (that is, incorrectly cast) were re-classified as formal and included in the count; secondly, and most importantly, however, the AEC stated that more than 1,300 ballots had been mislaid between the original count and the recount.

Lots of folks therefore intend to appeal the result and it looks likely Western Australian will need to stage another poll for the state’s six senate seats.

Following the shambolic Western Australian Senate election, the Australian Electoral Commission (AEC) has turned down a citizen’s FOI request to look at the software it uses to count Senate votes.

The decision, published yesterday at, was in response to a request made by Michael Cordova. …… The software, called EasyCode, is not used for voting, but for ballot counting only. The AEC believes EasyCode would be easy to decompile if it fell into the wrong hands (that is, members of the public).

This reminds me of a science fiction short story I read many, many years ago where software has developed to such an extent and where social behaviour is so well understood that voting is reduced to the use of one single elector and his/her preferences. If my memory serves, this single voter casts her vote and the software does all the rest.

The Western Australia election where the number of votes “lost” is about 100 times greater than the winning margin immediately gives rise to conspiracy theories which are being played down:

The Federal Government says it is unlikely “skulduggery” is behind the disappearanceof more than 1,300 ballot papers from the Western Australian Senate recount. Special Minister of State Michael Ronaldson says he has personally expressed his view to the Australian Electoral Commission (AEC) that the loss of the 1,375 votes is completely unsatisfactory.

“People have got to have trust in the AEC, that’s one of our pivotal democratic institutions. These sort of incidents I think effectively can, if you like, negate that trust that people have in the commission,” he said. …..

One of the potential options being canvassed to resolve the impasse is for Western Australian voters to return to the polls.

But Senator Ronaldson has cautioned against jumping to conclusions before the poll is declared. “We are in completely uncharted waters,” he said. “I think people have got to be cautious about jumping to a decision which is way at the end of the process.” Senator Ronaldson says he is worried about the impact of the incident on the reputation of the AEC. “These sort of incidents effectively can, if you like, negate the trust that people have in the commission.”

Dr Stewart Jackson, a politics lecturer at the University of Sydney, says he has never seen anything like it. “This may have happened before but we’ve just never noticed it because we’ve never had a recount in the same way,” he said.

I am still a little perplexed as to why software is necessary for counting physical votes. It’s not as if the numbers involved are all that large. Is a “software” count more accurate than or does it rank higher than a physical count? Or is a “software” count less susceptible to cheating?

Once upon a time humans had very limited number skills and were limited to “One, two, three, many”. Software would clearly have been superior to a human count in those times (except that it would have had to be programmed by the same humans)!

Tags: , , , ,

2 Responses to “Recounts, missing ballots and secret software enliven the Western Australia Senate election”

  1. Bill Says:

    From AEC statements it appears the software number crunches the preference flows of above the line ballots. But maybe it does more than that.

    Until 1984 preference flow in senate elections was calculated by “random sampling” of second preferences after counting all the first preferences. This system is still used for the NSW upper house. I was under the impression a full count of all votes is now used but maybe that’s wrong,

    There are no specific rules for a recount in the Electoral Act. It’s a repeat of the first process. Dr. Stewart Jackson is almost right in saying there has never been anything like this before. The Full Bench of the High Court judgement in Lack McManus in 1965 gives excellent guidance on the nature of a recount, the relevance of the first count and even the purpose of counting votes.
    To paraphrase the judgement, the purpose of counting the votes is not to arrive at an exact tally but to declare a result. Because the tally is going to be different every time the ballots are counted any comparison between different counts is irrelevant. They ruled that when a recount is called the first count is “disregarded”, The judgement also states that a recount is not a count of the same votes as this is not the definition in the Act.

    • ktwop Says:

      Thank you.
      I am always a little wary about “number crunching” for an election. It sounds much too complex for the simple addition that it ought to be.

Comments are closed.

%d bloggers like this: