Electronic Voting in Victoria: Background, Comparisons, and Implications
05 July, 2006| By Peter Chen |
Sign Up for free e-mail updates!
In Mid 2006, the State Government of Victoria (ALP) introduced a piece of legislation into the Legislative Assembly to empower the Victorian Electoral Commission (VEC) to undertake a limited experiment in electronic voting in the forthcoming State election.
This legislation, Electoral and Parliamentary Committees Legislation (Amendment) Bill ("the Bill"), sets out a range of requirements and specifications for the introduction of electronic voting (including new offences related to the implementation of this technology). In short, the Bill, when assented, will:
- Empower the VEC to implement electronic voting machines for use by electors with a visual impairment. This provision will not be regarded as a right (e.g. an elector can not demand to have a facility provided for them in their electorate, nor demand to use a system).
- Specify a range of performance requirements for the software of the system implemented, essentially to ensure that the system provides the same level of utility as the existing system (preferential voting, capacity to lodge an informal vote, ability to review the vote prior to casting the ballot, etc.).
While the legislation is ambiguous in that it provides for electronic voting to be available through a limited number of designated electronic voting centres (similar to the voting "super centres" employed by the VEC in previous elections to offer additional services to disabled voters), it is clear that the intention of the legislation is to provide the legal basis for stand alone (or local area networked) electronic voting machines to be implemented rather than the more controversial use of internet-based voting systems.
The advantage of these systems are that they allow complete physical control over the electronic voting system (if implemented correctly) by electoral officials and therefore significantly reduce the risks of vote tampering that have been hypothesised around internet-based systems.
Background to Electronic Voting in Victoria
On the one hand, the introduction of electronic voting can be seen to be a result of the Parliamentary Inquiry into Electronic Democracy (2004-5) in that limited electronic voting was recommended in this report for the purposes of providing a private vote for people with a visual impairment as well as those with limited or no English reading skills, such as migrants or the illiterate.
However, it should be noted that a number of electoral commissions throughout Australia have been interested in the possibility of electronic voting in Australia based on observations of international activities. Also the introduction of the EVACS system in the ACT provided a significant precedent for the spread of the idea beyond the progressive city state of Canberra.
Thus, in the wider history of electronic democracy in Australia, the introduction of Victorian electronic voting could probably represent yet another example of incremental policy learning and transfer common in many federal nations.
EVACS – A Summary
The EVACS electronic voting system used in the Australian Capital Territory is worthy of significant note given its status as the first binding electronic voting system in Australia and its innovative approach. The system has been used successfully in two Territory elections to date.
The key elements of the EVACS system that make it interesting are:
- EVACS was not based on purpose-built electronic voting machines (such as the ATM-style systems popular in the United States), but around a set of low cost, second-hand standard personal computers. The computers served as relatively "dumb" terminals recording votes on a single site-based hardened server, allowing a large number of low cost terminals to be deployed in the field, with the expectation of replacing defective machines quickly and easily.
- The core of the EVACS system was its open-source voting software based on a Linux OS, allowing the code to be inspected by members of the public (including programmers from the Australian National University, who identified bugs in the software prior to implementation).
- EVACS was open to all electors using a pre-registration system. It should be noted that the longer voting period in the ACT changes the nature of the voting process, allowing the EVACS machines to have higher levels of use than would be expected in jurisdictions with only a single day of voting.
Wider Background
The number of countries that have introduced some form of electronic voting system has been increasing over recent years and now includes (non-exhaustively) the United States, Canada, the United Kingdom, Ireland, and India. Given the wide variations in electoral systems (both structurally/institutionally, as well as in terms of the voting system itself), it is unsurprising that the range of models introduced includes relatively simple systems like the use of a basic hand-held single button model (India) to an attempt to build a globally-accessible Internet-based system (the US government SERVE approach, latter abandoned).
Overall, the variation in technical approaches also reflects considerable variation in motivation, for example:
- The primary motivation for a wide-scale multi-technology set of pilots (internet, kiosk, SMS) in the United Kingdom related to declining voter turnout;
- The primary motivation of electronic voting in parts of the US (including the SERVE system) was an attempt to improve public perceptions of the integrity of the voting system following the handing outcome of the Gore-Bush race; and
- The key motivation of the Indian government is to provide for voting in areas of limited security and where there are high levels of illiteracy.
With the possible expectation of India, few of the attempts have been free from controversy, which have mainly revolved around three key issues:
- Technical: can the system be made secure from tampering?
- Institutional: can the system be securely and effectively administered by electoral officials (particularly problematic in countries that devolve responsibility for electoral management to low levels of government and volunteers, like in the US)?
- Perceptual: if the system is secure, will electors accept this to be the case?
Overall, the security issue sits at the core of the debate, making the more robust pilot process of the UK (which showed ambiguous outcomes regarding the primary motivation for implementation, namely increased voter turnout) a standout in that time was taken to study the value of the system in detail beyond its technical robustness.
Differences to the ACT Model
While the exact nature of the electronic voting system is yet to be determined (the VEC has only just begun - at the time of writing - to have preliminary discussions with potential vendors or equipment suppliers), there are two notable observations that should be made regarding the implementation of electronic voting in Victoria, when compared with the ACT. Both differences revolve around the security and integrity of the system.
First, the Bill explicitly requires that the system implemented has the capacity to produce a paper record of the vote. This issue has been highlighted in the implementation of electronic voting systems in the United States over a number of years, with critics of electronic voting machines in that country calling for the introduction of "voter verifiable" paper trails for each machine. These systems do not provide the voter with a "receipt" (which would allow the possibility of vote buying), but allow the voter to physically observe a record of the vote, which can be manually recounted if disputes arise regarding the election outcome.
Second, while the Bill is not specific in its requirement for the security of the system by the VEC, the Bill does not include a requirement for an open-source approach to be used for the electronic voting system. An open-source system was recommended by the electronic democracy inquiry, given the value of this approach observed in the ACT. The lack of this requirement may not be problematic, should an open-source model be used (such as re-use of the ACT codebase, a logical approach).
Rationale for a Comparatively Conservative Approach
It is easy to criticise the Victorian approach, namely:
- The approach is limited in only offering the system to the vision impaired and not including those with limited written English skills; and
- The lack of security specificity in the legislation.
However, a number of factors should be considered that explain Victoria's (and Australia's more widely) conservative approach to electronic voting:
- With compulsory voting there is a limited need to experiment with technologies that are claimed to increase participation rates;
- There is no evidence that the general public (leaving aside recent changes by the Commonwealth to reduce the franchise based on dubious "integrity" rationales) lacks confidence in the system; and
- In general (though changes to Vicotiria's upper house make this slightly less clear), the system of voting in Australia is not so complex as to need to add "intelligence" to prevent spoiled ballots.
Future Implications
Overall, the implications are mild:
- Victoria has taken a small step towards wider use of electronic voting, should the pilot in 2006 be successful. With a population density much lower than the ACT, it is unlikely electronic voting will be employed as a general voting system in the near term;
- The rights of the vision impaired to cast a private ballot will be improved, something that should be extended as soon as possible to those with limited written English language skills; and
- To a small extent, the organisational capacity of the VEC to run technologically-facilitated elections will be improved.
Overall, with compulsory voting, the issue of electronic voting in Australia has limited scope to become an issue of interest to most Australians. Proposals to use electronic vote scanners possibly have more significant implications for electoral outcomes but these issues tend to be more at the administrative/institutional capacity level. Where interest in wider use of electronic voting (such as internet-based systems) may come from is recognition that, as an increasingly number of younger Australians become "gold-collar workers", a growing expatriate electorate is emerging.
Disclaimer - This article represents the personal views and opinions of the author and does not necessarily reflect the views or beliefs of my employer. Please note that, at the time of writing, the system for electronic voting in Victoria has not been determined, thus details in this paper may be subject to change.
Related Links