The Joint Standing Committee on Electoral Matters has just released its third interim report on the inquiry into the conduct of the 2016 federal election. The report covers the topic of AEC modernisation.

Much of the report refers to the Australian Electoral Commission's submission to the inquiry, which included a range of recommendations for integrating electronic systems into the voting process. This line from the report seems to acknowledge the tension between efficiency and integrity:

"the AEC's submission noted that it operates in a 'high risk environment which has zero tolerance for error'."

This noble sentiment does not mesh with one new part of the electoral process: the Senate vote counting algorithm.

Distribution of Senate votes: a black box

The redistribution of 2016 Senate preferences was done by a secret algorithm. The AEC refused a freedom of information request and a Senate direction to reveal this algorithm. Their reason? It is a "trade secret".

This is the code that decides which candidates are elected to parliament. The public should have the opportunity to scrutinise the program that decides who will make our laws. 

The Science Party condemns the refusal to publicise this source code. Keeping the algorithm secret cannot be justified and is counter to our principle of open government.

It is a shame that this contentious issue was not mentioned in the committee's report.

Electronic data input

The 2016 Senate votes were digitised using optical character recognition (OCR) software and verified by a human. This process should improve integrity by reducing human error.

The AEC recommended expanding this system to the House of Representatives ballot papers. However, the report noted that this move would need public support, particularly in marginal seats. The public sees that scrutineers do an important job and public confidence would need to be high in an automated replacement system.

Casting votes electronically

Our submission noted that electronic voting provides privacy for vision-impaired voters and allows easier vote counting, but adds potential for wide-scale vote corruption. This is why we recommend a hybrid system where the voter prepares their vote on-screen, prints the ballot paper and casts it into a box.

It may be appropriate for people to opt-in to fully electronic voting, especially if the alternative is to be assisted by another person who would then know the voter's choices. The report quotes Rikki Chaplin of Blind Citizens Australia:

"The introduction of an electronic voting system, even if limited to certain demographic groups, is the preference widely held by people who are blind or vision impaired. The New South Wales system of electronic voting continues to be greatly appreciated by people who are blind or vision impaired and, to our knowledge, has not experienced any breaches of security."

Preventing wrongful voting

Fraudulent and multiple voting is a frequent concern, with voter ID often cited as the way to end it. However, multiple voting in Australia is rare and most cases appear to be accidental. It is, and should be, taken very seriously when it occurs.

Requiring voters to produce ID may further disenfranchise people from the most marginalised groups. The Queensland experiment suggests that rural voters are worst affected.

The AEC's submission recommended wider use of an electronic voter roll to mark off voters' names, to prevent both multiple voting and voting in the wrong electorate. This technology was trialled with success in 2016. Unfortunately the AEC's current budget does not permit Australia-wide use of the electronic roll.

Other technological wins

The AEC noted the success of online enrolment and online postal votes. Online enrolment may account for some of the recent increase in youth participation, and online postal votes eliminate postal delays.

We can only hope that the Australian government sees the value in expanding electronic systems where safe to do so, such as by expanding the use of the electronic voter roll.