Master students in Public and Nonprofit Management at the Luzern University of Applied Sciences and Arts organized a podium discussion on the topic “E-voting in Switzerland”. Initially due to take place on 21 April, the event was cancelled because of COVID19 restrictions. My foreseen input talk took the form of a written interview. I received thought provoking questions by the students, ranging from constitutional compliance of e-voting to its relevance in corona crisis times. Here is the full written interview.
Would e-voting in Switzerland comply with the principles of free and fair elections and would you see it consistent with our democratic principles?
A. Driza Maurer – What I find interesting about the use of digital technologies in elections is that such use questions our understanding of the constitutional principles. It’s not just e-voting, but also e-counting, the use of biometrics or blockchain that raise questions about the exact meaning of the basic principles of free and fair elections and their application to the technology.
The constitutional principles of universal, equal, secret, free and direct elections grew organically during the nineteenth century – when democracy based on universal citizen participation as we know it today started – and developed in the twentieth century, where paper voting in polling stations and voting by raising hand, mainly at the local level, were the main techniques to express the vote. The exact content of the principles was defined mainly by the judge, through interpretation. The case-law of the Federal Supreme Court has played a pivotal role in identifying the content of the principles and in evaluating their application. Interpretation relies on a good understanding of contextual elements, among which the technique employed to vote. Paper or raising hands, or even mechanical voting – where it existed – can be understood and evaluated by the layperson. However, this is not true for digital technology.
The first question when it comes to constitutional compliance of e-voting is: What are the e-voting specific requirements that result from the general principles for democratic elections? In 2014 the Federal Supreme Court decided not to conduct an abstract review of cantonal legal provisions on e-voting leaving it to politicians, informed by academia, to decide on the merits of e-voting and define the conditions that apply to it (decision of 22 July 2014 – 1C_136/2014). Other courts, namely the German Constitutional Court or the Austrian one, have taken another approach.
The German Court held in its judgement of 3 March 2009 on the use of e-voting machines in polling stations (BVerfGE 123, 39) that the principle of election transparency required that all the essential steps in an election be subject to public scrutiny. It interpreted this to mean that it must be possible for citizens to check reliably and without specialist knowledge the essential steps in the act of voting and the ascertainment of the results. The detailed regulation of e-voting should provide for this. As it did not, it was considered unconstitutional.
The Austrian Court, in its judgement of 13 December 2011 on the use of internet voting to elect the representative bodies of the Austrian Students’ Union in 2009 (VfSlg. 19.592/2011) followed a different reasoning path, to reach a similar conclusion: the electoral principles require that a voting system could be verified and controlled by the voter and the public – represented by the electoral commission. The Court quashed the specific regulation as being too indeterminate, i.e. not detailing sufficiently the technical system as well as the procedures to be applied so that a reliable control, without specialist technical knowledge, was possible.
The Swiss Federal Supreme Court decision of 22 July 2014 was certainly influenced by the fact that a new federal regulation came into force in January 2014, which is rather detailed and introduces minimum requirements on verifiability and transparency of e-voting. That regulation takes into account several concerns expressed by the German or the Austrian Courts. Its provisions were considered to reflect state of the art. As for the extent of public scrutiny, the Federal Supreme Court indirectly endorses the regulator’s choice to rely on independent technical aids and experts, provided experts are appointed in a democratic process.
A second question relates to controls and to ensuring that an e-voting system as well as its implementation during a specific vote or election comply with the principles and the requirements. The e-voting 2014 regulation foresees several controls, from formal certification by an independent and competent body, to tests that should be performed by the canton and/or its e-voting service provider, to individual and universal verifiability checks to be conducted by the voter or by experts. The transparency exercise that took place in 2019 on the Swiss Post/Scytl e-voting system showed that some checks did not detect some important flaws affecting the verifiability solution. As a result, in June 2019 the federal Chancellery received the task of redesigning the conditions for e-voting trials (including for controls).
When it comes to detailed regulation of digital technologies, a third important question relates to ensuring that detailed requirements, especially those on security, reflect good practice and state-of-the-art over time. Detailed requirements that are in line with state-of-the-art are expected to best ensure compliance with democratic principles. However state-of-the-art in the digital field evolves rather quickly. Regulatory provisions must be updated to integrate changes. Even when the update is done quickly, there is always a time lag between regulation and technique. This problem needs to be discussed and solved at the regulatory level.
To summarize, compliance of e-voting with democratic principles for elections requires legal analysis of the exact meaning of the principles when applied to the specific e-voting context. Furthermore, principles should be “translated” into detailed requirements that govern e-voting, including requirements on controlling compliance of e-voting with democratic principles. Combined legal and technical expertise is necessary when designing an e-voting policy and regulation. Major political decisions such as those on the scope of secrecy, the exact meaning of public scrutiny or the financial implications of security, must receive broad political backing. Technical choices must reflect good practices and other recommendations of the community (peers). When “translating” principles into detailed requirements, decisions may vary between countries, to reflect legal specificities. Such decisions must be taken by the competent authority, usually the legislator/regulator. Given the rapid changes in digital technology, an important aspect is ensuring compliance over time. These considerations and others, if handled correctly, contribute to ensuring compliance of e-voting with democratic principles.
Your work focuses on the regulatory aspects of e-voting / use of digital solutions in voting and elections. Which findings do you consider to be particularly relevant?
A. Driza Maurer – Documents and processes related to political rights are increasingly being digitized. This is so also for less discussed elements like pure e-counting, registers, results transmission systems, etc. Digitizing and interconnecting the different elements of the political rights system may make it more efficient but makes it also more vulnerable, especially to external threats. Now, e-voting covers just two – although very important – procedures: vote casting and counting. It can be considered as the tip of a bigger iceberg. In my view, it is important to consider the bigger picture (the entire iceberg) and envisage e-voting and its regulation in the broader context. Votes and elections involve sensitive and critical data and processes to which stricter conditions of data protection, information security, transparency, verifiability, etc. apply. Regulation should address, in a systematic and harmonized way, all sensitive data and processes involved in political rights.
Which positive effects could e-voting have in a direct democracy like Switzerland?
A. Driza Maurer – As an additional voting channel, e-voting helps ensure universal suffrage for Swiss abroad and secret suffrage for the visually impaired. Depending on the evolution of postal mail and of social habits, it may help guarantee that voting from home remains a possible option, also in the future. Voting from home and during a relatively long period prior to voting day are important to enable voters to participate in a direct democracy like Switzerland.
The e-voting experiment has furthermore provided valuable insights on what the use of digital technologies in the political rights field entails and how to proceed with it.
Why are there so many opponents to e-voting?
A. Driza Maurer – As mentioned, a number of decisions of principle are necessary in order to develop e-voting. They relate to a sensitive field – political rights – and to a complex technology – the digital one. Ideally, they require broad political consensus. Understanding the issues involved, identifying appropriate solutions and building broad consensus takes time.
An explanation may be the fact that e-voting addresses only some issues, namely it offers an effective voting channel to the Swiss abroad and a voting channel that respects secrecy to the visually impaired. Some may feel that other important elements, for instance opinion formation or signature collection, are being neglected and may wish to prioritize digitization of the last ones.
The important financial means that are needed to develop and maintain an e-voting system that is state-of-the-art may offer yet another explanation.
What are the main obstacles to implement e-voting? How could they be resolved?
A. Driza Maurer – The main challenge in my view is to embed e-voting and other uses of digital technology in votes and elections into a coherent project and to consider the development of the whole political rights system in the digitized era. The e-voting experience offers important insights into that.
Part of the solution is strengthening multidisciplinary work. True dialogue between legal, technical and social experts is more than a one-off conference (although these are also important). It requires mutual understanding and iterative exchanges between the different disciplines and this should accompany the project from its design, through legislation and regulation, to operation and control. An adequate framework for this important work to take place is needed.
Another important challenge is the effective use of state-of-the-art solutions by voters. It is not enough to elaborate technically secure solutions. They should also be effectively used by the voter. Certain attacks, for instance, can only be detected if enough end-users conduct the verification, understand the problem revealed by the verification and then complain. The solution here is to inform and instruct voters. Achieving this is difficult but important. It may be more realistically achieved if such (verifiability or other) solutions are used more broadly throughout the political rights system and, why not, in other sensitive areas too.
In comparison to other countries how would you describe the attitudes and the current state of affairs concerning e-voting in Switzerland?
A. Driza Maurer – Swiss authorities deserve praise for sticking to the principles. The organization of the source code publication and of the public intrusion test of the Swiss Post/Scytl internet voting system in 2019 is a good practice and so far, to our knowledge, the most complete transparency exercise worldwide on an internet voting system used for political elections.
The step-by-step, cautious approach adopted by Switzerland in introducing e-voting is considered a good practice by the Council of Europe and is part of its recommendations. Cooperation with technical experts, e.g. the BFH and UZH is another positive example.
Generally speaking, Switzerland is the country that has done the most work and gathered the most experience in the field of e-voting regulation. That experience is considered by international organisations, like the Council of Europe.
We are currently in corona crisis mode. Federal votes must be postponed. To what extent could e-voting guarantee the continuation of democratic processes in exceptional situations such as the current corona pandemic?
A. Driza Maurer – Democratic processes include a wide range of activities. E-voting would guarantee that the vote takes place, however it cannot ensure that the campaign takes place. Recent examples illustrate this. The impossibility of organizing a correct campaign was the main argument for cancelling the 17 May 2020 federal vote. In Geneva, on the other hand, the fact that the campaign for the first round of local elections already took place normally before the corona crisis started, had a great impact on the decision to hold the second round as scheduled, on 5 April, despite the corona-related restrictions. By the way, only postal voting was allowed at that occasion.
Again, it seems important to consider the whole political rights system, its resilience and functioning under exceptional circumstances like the present ones.
Can you briefly say something about yourself and your activities in relation to e-voting?
A. Driza Maurer – I am a jurist, writing a PhD (UZH/Zentrum für Demokratie Aarau) on the genuine expression of a citizen’s will (art. 34.2 Cst.) and e-voting. I was a member of the e-voting project at the federal Chancellery from 2006 to 2012 and I am an independent consultant since 2013. I led the group of technical experts of the Council of Europe who drafted the new Recommendation on e-voting: Recommendation CM/Rec(2017)5of the Committee of Ministers to member States on standards for e-voting. I co-edited “E-voting case law: A comparative analysis”, 2015/2017, Ashgate/Routledge, which discusses legal aspects of e-voting in some thirteen countries. I am currently expert consultant mandated by the European Committee on Democracy and Governance (CDDG) of the Council of Europe to work on the issue of possible guidance by the Council of Europe on the use of new technologies in the electoral cycle.
Ardita Driza Maurer, 28 April 2020
Questions were prepared and submitted by Jana Kobler, Tabea Diener, Alexandra Dändliker and Sonja Leguizamon
