dimecres, 14 de març de 2007

Feedback on Estonian e-voting (and more...)

Our colleague Letizia Caporusso has sent us some feedback on her electoral observation in Estonia during the last electoral process. We have her permission to "copy & paste" her mail, because it's really interesting. She went also to the Netherlands... so don't miss the last part of the "report". (Thank's once more, Letizia!)

I am finally back in Trento after ten days of e-immersion and I am currently trying to put together all the pieces of the puzzle. Here’s a very brief résumé of my journey and of my impressions about elections in Estonia and in the Netherlands, as promised.

Parliamentary elections in Estonia took place, as you know, on March 4th by means of paper ballots, whereas during the advance voting period both electronic and paper ballot were available. It was the first time that internet-voting was implemented on a national scale, though it had already been tested in October 2005 during municipal elections [1]. As pointed out during the presentation held for the international observers on the day preceding the elections [2], i-voting is just one of the 14 ways in which Estonians can cast their vote. The rationale behind it is that it might expand citizens participation by allowing voting when on the move or from home. But alongside this effort, you should also bear in mind the communicational and commercial consequences of crafting E-stonia own made voting solution. Estonian economics and IT-related activities boomed in the last decade [3] and my feeling is that the technophilia spread to all aspects of everyday life and to the interaction between citizens and institutions. I witnessed how common it is for commercial activities and individuals alike to advertise their own websites and how proudly people, regardless of their education and background, talk about the communication infrastructure that has been set up (I had very nice conversations with the staff of my hotel and with the cab driver on my way back to the airport, besides chatting with our “official” guides). The widespread trust that citizens have in technology, the pride for the goals achieved in such a limited amount of time, the unusual possibility to set up new voting procedures without transitioning to e-voting from older, traditional and well established forms of balloting, as well as the small size of the country and the solidarity within the community, have resulted in a general acceptance of automated systems. There is no news of any pressure group against electronic voting and I haven’t heard of debates about the outcome of elections or the management of the entire election process.
What I really appreciated from the election committee is the total openness to hosting observers from abroad and from diverse backgrounds (I was there as an independent researcher, not as a member of any national or international auditing institution) and I wish to underline how helpful, precise and… picky were the local guides that volunteered to join us in the polling stations, answer to our queries and translate our questions to voters and scrutinizers. My regrets is not to have the skills to appreciate and evaluate the technical solution and, above all, to have observed only “traditional” (paper and pencil) elections, as when the observers’ programme started, advance voting was already over. The counting of electronic ballots that took place late in the evening was a highly dramatic moment, a climax of mixed emotions that shook the audience until results were shown and an applause soared. On my account, well, though excited by the novelty I must admit that not being in control of the whole process (having missed some of the phases and being unable to decipher others) made me feel slightly unease. My arguments are not that original: not knowing where servers and ballots are stored, how votes were encrypted, and so on, elicited some insights on the sense of “trust” that citizens must have in the entire process – or, put in another way, on the relative distance that they might take from it.
The bottom line is that e-voting (and i-voting even more!) should be a completely auditable process not only to the international observers and professional auditors but above all to citizens, who should be able to follow every single step of what is happening, if interested.

The observation mission continued in the Netherlands, where provincial elections were scheduled for March 7th [4]. I was lucky enough to be invited by Nedap [5], which is one of the largest European producers of voting machines and the only provider for Dutch elections. As Italy is a potential market, they showed us the computers they use around the globe and explained in plain words (intelligible to non technicians) the solutions they designed for voting in a supervised environment as well as for voting on the web. We watched elections in Enschede, the town closest to their main place of business, and also observed the arrival of data transmitted via GPRS as part of a pilot. The overall impression was that people are just “used to” automated voting procedures, so scrutinizers and voters alike seem to be content with the system as it is. Nonetheless, I am aware that pressure groups against electronic computers are playing an increasingly strategic role in the interplay between governments and commercial providers of voting solutions. There’s a very active one in the Netherlands, and I was lucky or cheeky enough to schedule a meeting with people of the “We don’t trust electronic computers” campaign [6]. Their report [7] on security issues related to the machines used for voting in the Netherlands elicited much attention from the media, the government and the public. At the time being, all the people I had the chance to talk to in Amsterdam, though not a representative sample, knew about the hack and about potential threats of casting a vote electronically. Nedap’s “scientific” reply to the hackers, a report compiled by the University of Twente (based in Enschede, too) dealing with reliability and usability of their voting machine [8], apparently failed to intercept citizens. I am actually very curious about the position that the electoral commission will take into the debate. What the hackers want to point out is the potential risks involved in the present election system (I don’t think they are truly aiming at destabilizing the state, as some accuse them of) and the need of VVPAT, though I had the impression that they somehow fail to propose a better solution in a consistent way (there must be all sorts of geeks involved in the campaign ranging from extremist to moderate positions). Anyway, Rop Gonggrijp arguments were clear, straightforward and non ideological and Anne-Marie Oostveen [9], whose articles on electronic voting and electronic democracy represent an important benchmark for my own research, confirmed that the nature of the campaign is non economically-driven but rests only on volunteers’ engagement.
There are obvious problems with paper ballots, too, and I personally wish that activism concerned with voting procedures could consider all options available, electronic and paper alike, without focusing solely on a scapegoat and failing to address more overarching problems. I guess we should consider risks and opportunities on a relative basis, i.e., comparing different solutions, rather than just opposing or supporting electronic or paper voting tout court.

The ideas expressed above are obviously very subjective and susceptible to further developments. The issue is so intriguing and stakeholders so diverse that it might take more time until I get it to grips competently; in the meanwhile, I hope you enjoyed this brief account of my own experience!

Best greetings,

[1] http://www.vvk.ee/english/report2006.pdf
[2] http://www.vvk.ee/engindex.html
[3] http://www.stat.ee/
[4] http://www.verkiezingsuitslagen.nl/
[5] http://www.election-systems.eu/website/Read.php
[6] http://www.wijvertrouwenstemcomputersniet.nl/English
[7] http://www.wijvertrouwenstemcomputersniet.nl/images/9/91/Es3b-en.pdf
[8] http://www.election-systems.eu/website/Read.php?PageID=1212
[9] http://www.social-informatics.net/evoting.htm