Electronic voting machine used in Brazilian elections. Agência Brasil
As more countries are discussing the adoption of voting technology, it is important to fully explore the challenges and drawbacks of e-voting. To this end, election experts gathered at the Fourth International Conference on Electronic Voting, which explores some of the challenges presented by e-voting. The conference took place in Bregenz, Austria from July 21-24, 2010.
Every two years the Conference brings together experts on electronic voting from the fields of political science, computer science, election administration, voting rights and technology suppliers to discuss current issues and recent experiences related to electronic voting and counting. Unfortunately international observation organisations, which are still struggling to find ways to properly observe e-enabled elections, were not as well represented.
More than 50 participants attended the event, including the keynote speaker, Donetta Davidson, Chair of the US Election Assistance Commission. Nearly twenty papers were presented and discussed at the Conference, covering some critical challenges facing electronic voting at the moment, as well as recent experiences of electronic voting and counting in the US and Austria. The Council of Europe, a co-organiser of the Conference, also used the opportunity to discuss its own work developing papers to guide states in the implementation e-enabled elections. These include papers on ‘Implementing E-enabled Elections: The Key Steps’, ‘Initial Draft Guidelines on the Transparency of E-enabled Elections’ and ‘Clarification of Certification Procedures for E-voting Systems’, and will be completed by the end of 2010.
Key themes which permeated the Conference concerned the trust of voters in e-enabled elections, the certification of electronic voting and counting systems, and the verifiability of results under e-enabled elections. The idea of verifiability as potential answer to the lack of observability seems to be shared among the presenters.
Election administrators are faced with two tasks when it comes to elections, not only must they implement elections according to international standards, but they must also be perceived to have done so. In some ways the perception of a good election could be argued to be more important, as the credibility of the election and the elected institutions rely on the trust of the voters in the results and that defeated candidates and parties will accept the results.
While paper based elections are far from perfect and have a long history of fraud, the system of paper based voting is inherently more transparent to the voter, and this helps to build trust in the process – however misplaced. While e-enabled elections can be conducted in a way that maintains and builds voter trust, this trust is a fragile commodity and is far easier to maintain than regain. Therefore election administrators need to extremely careful in implementing e-enabled elections so that they address the inherent lack of transparency and trust involved with such elections. As Thad Hall, a leading US scholar on electronic voting, perceptively stated, “The technology you use is less important than how well you implement it”.
This point can be aptly demonstrated by the recent trial of internet voting in the Austrian Federation of Students election. Austrian students were allowed to choose between using traditional paper voting and Internet voting in the election. Although the election was a technical success it was perceived as a failure, largely because the administrators failed to have internet voting endorsed by a key stakeholder in the process, the student body. This led to organised spoiling tactics by student and political parties, and an incorrect perception that the internet voting platform had failed. This has set back the use of e-enabled elections in Austria many years.
It is partly to address the issue of trust that such importance is placed in certification procedures for electronic voting and counting systems, and that the debate about the verifiability of electronic voting systems has arisen. The Conference was briefed by the Chair of the US Election Assistance Commission on the comprehensive testing and certification procedures implemented for US voting systems through its Independent Testing Authorities. This certification process is meant to ensure that electronic voting and counting systems meet the specifications defined for them, do not contain program or system errors, and are secure against malicious attack. The independent certification of these standards is an important mechanism for building trust among voters.
However, it was accepted by all that no certification and testing process can ever fully identify all possible bugs and problems in counting and voting systems, especially as these systems are becoming increasingly more complex in order to deal with language, ballot design and accessibility issues. Therefore, computer scientists have devised another mechanism to attempt to ensure trust in the process where electronic voting or counting systems are used – verifiability.
Verifiability seeks to give the voter using electronic voting and counting systems the tools to confirm that his or her vote is “cast as intended” and “counted as cast”. Here the attempts of computer scientists run into contradictory electoral standards. On the one hand the aim is to provide transparency for, and therefore trust in, the e-enabled process, a process which is by nature less transparent. However, the ultimate aim of allowing the voter to check that his or her vote has been included as intended in the count seems to be incredibly difficult to do without violating the secrecy of the vote and introducing the possibility of voter coercion and vote buying.
This challenge has been testing the minds of computer scientists in recent years and while some novel approaches to this were presented to the Conference, for example the Scantegrity and Sigma Ballot solution, these systems seem very complex and may face challenges of scalability if attempted for larger elections.
It is clear that these issues still need to be developed further and considered in more detail before an effective and implementable solution can be found. The poor attendance of the observation community at the Conference, a key stakeholders in trust building, is also indicative of this community’s inertia in building a capacity to understand the technology behind e-voting. The Conference provided an excellent forum for an exchange of views between experts of varying backgrounds and it is to be hoped that through such open exchanges of knowledge solutions can be found to raise the level of voter trust in e-enabled elections. Only in this way can the potential benefits of e-enabled elections be truly realised.
Special thanks to Election Guide for sharing this article with us.