Electronic Voting

Despite electronic voting having been implemented in political elections more that 30 years ago – in the Netherlands – and the fact that around 25 countries have carried out experiments on electronic voting in various forms, with almost half of these having been started more than 10 years ago, its regular usage is currently very restricted. Only four countries (Brazil, India, Estonia, Venezuela) nowadays (2008) use direct electronic voting everywhere (via the Internet in Estonia), and only two other countries use it to a reasonable extent (around 50% in Belgium in 2004 and 2007; 38% in the United States of America, in 2006).

Of these countries, voting via the Internet only takes place in Estonia, whereas in other countries voting is carried out using machines set up in polling stations and “mobile voting” is not possible.

Therefore only in Estonia has electronic voting brought about the possibility of “mobile voting” for all the electorate, although it is little used, and up to now the maximum use of voting via the Internet has been by 3.4% of voters.

Switzerland has been enlarging the possibility of voting via the Internet in 3 of its 26 cantons since 2003. In June 2007 17% of the votes cast in a federal referendum were via the Internet, and it has decided to carry out a gradual move over to allowing voting via the Internet throughout the whole country.

Even making voting via the Internet available to citizens resident abroad in countries where they may vote by correspondence has been rare: besides Estonia, and in part of the three Swiss cantons where voting via the Internet has been introduced, this was made possible in the Netherlands in 2004 and in France in 2006.

The security problems which may arise have led a number of countries to delay or interrupt the introduction of electronic voting and, in some cases, to abandon it, the most noteworthy case being the Netherlands which, after steady growth throughout more than 30 years and having made electronic voting almost fully available through machines in polling stations in 2002, decided to completely ban this in May 2008 and return to paper balloting.

The first experimental piloting of electronic voting in Portugal was carried out during the 1997 Local Elections, in the ward of São Sebastião da Pedreira (Lisbon), by the Technical Secretariat for Electoral Procedures - STAPE  without the results counting towards the official results. As such, Portugal was the fifth European country to experiment with electronic voting, following the Netherlands, Belgium, France and Spain, and the eighth in the world after these and the United States of America, Brazil and the Philippines.

The second experimental piloting of electronic voting was also carried out by STAPE during the 2001 Local Elections, in the wards of Sobral de Monte Agraço (Sobral de Monte Agraço) and Campelo (Baião), without the results counting towards the official results.

The third and fourth experimental piloting of electronic voting took place during the 2004 European Elections, involving the testing of three different technologies using around 150,000 voters in 9 wards, where 61,083 people voted, 9% of whom were participating in the experiment, and in the 2005 Parliamentary Elections, which increased the previous voting level and included technology to provide support for voting for citizens with special needs, and was carried out in 5 wards with around 38,000 voters, where 26,515 persons voted, 33% of whom were participating in the experiment, as well as experimenting with voting via the Internet for the around 147,000 registered voters abroad, 3% of whom participated in the experiment. The votes cast in these projects also did not count towards the official results.

The remote voting system from any Internet access point could be seen as an alternative to postal votes, but, as there is no system to allow voters to state that they were not been coerced at the moment of voting, it is difficult to accept the possibility of this being generalised.

The other alternatives tested indicate a possible generalised future use of electronic voting machines in all polling stations, which would be extremely difficult both logistically, in terms of the need to ensure the installation and maintenance of voting machines on a large scale throughout the country, without any problems of power cuts of connectivity, and also in terms of the high costs of associated goods and services.

It should be added that both voting using electronic voting machines and via the Internet have shown tendencies for security vulnerabilities, the resolution of which is complex, as has been seen in electronic voting experiments in other countries, and in particular in Germany, the United States of America, the Netherlands and Ireland, where electronic voting has been suspended or substantially altered.

On the other hand, Estonia started remote electronic voting using any access point to the Internet in October 2005 and has continued to do so, as have some cantons in Switzerland, and Latvia is intending to introduce this form of voting shortly, despite the fact that providing voting via the Internet to all the electorate has been excluded in many European countries, since it has been judged not to satisfy security requirements and guarantee independence and voting secrecy for generalised use and as such, should at the most only be considered in situations such as voters who are resident abroad (see the electronic voting experiments in other countries).

Experiments which at the time were considered exemplary, such as making remote electronic voting via the Internet available to 80% of the electorate in Estonia in 2005, have now been shown to present serious difficulties, which in the case of Estonia were amply illustrated in the local elections of 16 October 2005 when only 0.9% of the electorate voted electronically. These difficulties were confirmed in the parliamentary elections of 4 March 2007 when only 3.4% of the electorate voted electronically although electronic voting had been made available to 100% of the electorate.

In political elections it appears that it can be stated at the present time that there is the need to keep general voting in places where it can be witnessed that voting is carried out by voters themselves, without any coercion. In addition to this, certain countries which tried out electronic voting systems, such as Ireland or the Netherlands, have concluded that it was necessary to provide systems which offered a paper trail to verify and recount votes.

Situations such as those in Brazil or in India are widely known, where voting machines are made available in all the polling stations, but which do not allow for mobile voting, with the aim of shortening the period in which the votes are counted from many days to less than a day. However, this is a problem which we do not have in Portugal, where traditional counting of paper ballots normally takes less than 6 hours, and the solutions mentioned do not resolve the problem of "mobile voting". Our present aim should clearly be to find a reduced costs solution to "mobile voting”.

It should be noted that one of the reports auditing the experimental piloting carried out in the 2004 Local Elections (Consultancy Report on the experimental piloting of Electronic Voting carried out on 13 June 2004 (in Portuguese)), André Ventura Zúquete, IEETA / UA and Paulo Jorge Pires Ferreira, INESC ID / IST) mentions that:

"The most significant advantage of the systems being considered, when compared with the current system (i.e. the traditional, paper-based one), is greatly reduced, since it is on the whole limited to presenting an interface which is (perhaps) more agreeable to the voter and potentially reduces the time to determine the results. Neither of these aspects appear to us to justify the investment, however reduced in might be, in the relevant technologies. In effect, as far as we are concerned, an IT solution would seem justified if it would provide mobility of voting, i.e. if the right to vote could be carried out in a place which was not the polling station where one was registered."

"Mobile voting” does not necessarily require electronic voting. In truth, as we know from the votes of those eligible to vote who live overseas, it is viable for the country to introduce a simple and economic paper-based system of "mobile voting", with the need to bring the “mobile vote” forward by a few days between it and Election Day, a process which would be facilitated by the digitalisation of the electoral register.

The possible contribution of information and communication technologies to “mobile voting” in the country is essentially a result of simultaneously allowing “mobile voting” as well as traditional voting on Election Day itself. In fact, this possibility requires an electronic electoral register for all polling stations, and instantaneous and secure means of communication between polling stations which could be made possible through encrypted electronic communication with the aim of checking whether a certain voter who comes to a polling station to cast a “mobile vote” has already voted elsewhere and, if not, recording in the electoral register that he/she cast a "mobile vote".

Generalised electronic ballots for voting situated in polling stations has the inconvenience of requiring complex large-scale technological logistics, which raise technical problems of security and secrecy of voting and obscure the transparency of the electoral act for the general citizen, whilst having the advantages of facilitating and shortening the process of vote counting with maximum rigour. The other possible advantage of electronic voting is it facilitates “mobile voting” in the cases of local elections through allowing the secrecy of a small number of “mobile votes” to be effectively ensured by the electronic merging of a large number of votes.

Within this context, and with the aim of obtaining data to consider the adoption of a system of presencial electronic voting in polling stations The Knowledge Society Agency (UMIC) commissioned a studied from Deloitte in 2006. The resulting report was published in March 2007 with the title “Final Report - Analysis of the Financial Impact of Electronic Voting in Portugal (in Portuguese) ”.

Responsibility for the development of electronic voting services was moved from The Knowledge Society Agency (UMIC) to the Agency for Administrative Modernisation - AMA (site in Portuguese), on 1 May 2007, following the start of activities by this new body, to which was transferred the competencies which UMIC had held in the field of electronic public administration, following the publication of the organic law and the statutes of the AMA (site in Portuguese) in the Official Gazette Diário da República, on the  27 and 30 April 2007, respectively, with continuity ensured by the transfer of the member of the UMIC Management Board responsible for electronic public administration projects to become the Director of AMA (site in Portuguese).