Open source software has gained enthusiastic advocates because it can be maintained by any person without fear of vendors going bust. There are many other extremely positive traits however. One of these is that open source software can be understood and trusted.
The underlying program code is available for any interested person to scrutinise, establishing confidence and trust.
In a similar fashion, despite being an electric world, many stable democracies are reluctant to turn to electronic voting systems because a traditional paper and pencil voting system has transparency. Voters see their name marked off the electoral role, they mark their ballot, they deposit it in a locked box. Later, scrutineers from differing political parties may view the ballot box being unlocked and ballots counted.
This process offers many checks and balances and its transparency provides confidence and trust.
Yet, it may (or may not) be a surprise to learn the Australian Electoral Commission (AEC) employs in-house custom software for counting Senate votes. The Senate ballot paper has become an enormous tablecloth sized ballot with many different parties and candidates. This is partially to do with the low first preference vote required to gain a Senate seat provided a sufficient number of preference swapping deals can be made with the other hopeful candidates.
Can the ordinary person on the street understand how the Senate vote is calculated? The AEC is sending out mixed messages.
Yet, on the other hand, the AEC has told one Tasmanian lawyer the Senate vote counting algorithm is a “trade secret”.
Specifically, Michael Cordover – aka mjec – who holds a Bachelor of Science, Bachelor of Laws and a Graduate Diploma in Legal Practice from the University of Tasmania – submitted a Freedom of Information request to the AEC on the 4th October 2013. Cordover’s request was for the source code of the software used to conduct Senate counts. This software is known as EasyCount and was developed internally by the AEC.
On the 5th November 2013 the AEC refused this request. Cordover requested an internal review of this decision, which the AEC again rejected in December 2013. Cordover has continued to seek a review of this matter.
The AEC’s basis for denying the request is that it would disclose “trade secrets” and risk destroying or diminishing the commercial value of the information released.
It is important to note EasyCount is not a voting system, but purely a ballot counting system. There ought not to be any “trade secrets” in how Senate votes are counted; that is one of the fundamental components of a stable democracy.
Rather, the AEC asserts EasyCount was not initially designed for Senate vote counting but for use in industrial elections where trade unions have elections conducted by the AEC; it then became adapted for use with the Senate as well as for fee-for-service elections in which bodies conducting elections contract the AEC to perform the service.
The AEC thus maintains if it disclosed the source code for EasyCount it would potentially lose a revenue stream as other, competing, organisations could produce software to perform election counting and management of complex preference distributions.
While on the surface this argument may hold merit, it is important to note again that voting and preference distribution should not and cannot be a “secret” or else the entire democratic process – be it for a company, a union, a volunteer organisation or a government – has no integrity. Additionally, it is potentially dubious that a company would hire the AEC for its software as opposed to its brandname and the credibility that name gives.
Nevertheless, Cordover’s FoI request was denied for these reasons, despite Cordover only seeking the component of source code which is used for the Australian Senate.
Cordover has documented the story so far and has commenced a crowd funding campaign to reach a target of $948 to make a further appeal via the administrative appeals tournament.
At the time of writing Cordover has reached $1,055 pledged with 30 days to go. The target has been reached, clearly, but the matter still deserves support to provide additional funding for any further action. Those who espouse the values of freedom – be it software or an accountable, transparent democratic process – would do well to review the matter and campaign.