Consultation and Evaluation Practices in the Implementation of Internet Voting in Canada and Europe
Consultation and Evaluation in Context
Consultations come in many forms. Approaches to public consultation and tools used to engage citizens vary depending on the specific context, policy process or program development (Sheedy 2008). Widespread public consultations are a relatively open process, in which submissions are invited from any group or individual wishing to express a view. This could be organized in any number of ways, but the common denominator would be a process that does not select a particular kind of person, with a particular kind of view, as a participant in advance. A completely open process is the least restrictive method for selecting participants, whereby participants are a self-selected subset of the population (Fung 2006). Ideally such a process would attract people with set views, occasionally extreme views, as well as a number of people who have not yet made up their minds about the issue. Public consultation processes can produce sharply conflicting results, or even deadlock, but at other times the prevailing opinion can point in one direction. Deliberations can often point to recommendations that might be generally subscribed to, or at the limiting extreme, form a consensus. If public consultation processes can suggest a generally arrived at opinion or result, support will be generated for a policy choice that goes far beyond that of a select group of representatives (in a parliamentary committee, for example).
A different kind of support for a policy decision is that provided by a committee of experts, whether these experts are technically proficient in the field concerned, or more diverse. Consultations can therefore be deliberately structured with groups who have expert knowledge of a subject, or who are most likely to be interested in the subject, potentially because they might benefit from it.
Evaluations, by comparison, can be formal or informal. In more formal, written, evaluations, criteria are specified to consider whether, in this case, experiments with Internet voting have been "successful" and are worthy of being continued. Formal evaluations such as those conducted by the Office for Democratic Institutions and Human Rights of the Organization for Security and Co-operation in Europe, which are outlined later on with respect to the three European countries referenced in this report, have a number of established criteria, such as the conformity of Internet voting with legal or constitutional provisions in the country concerned, and the ways they have helped the country meet provisions of human rights statutes like those requiring equal access to the ballot for all citizens. There are also other evaluation criteria that make a formal or informal appearance on the scene and that are heavily dependent on the context.
The types of evaluations and consultations employed, or indeed whether they are employed at all, are closely related to the context in which Internet voting is being considered. From our investigations of the cases specified above, we have identified different kinds of contextual factors that are related to the consultation and evaluations methods used and help account for why jurisdictions may or may not have engaged in consultation or evaluation and why they made the particular choices they did. A summary of the primary rationales for the consideration of the adoption of Internet voting can be found in Table 1. These factors help put the following discussion into perspective, demonstrate commonalities, and highlight differences.
|Jurisdiction name||Voting turnout||Leadership role in e-government||Accessibility||Convenience||Citizen-centred service||Greater youth involvement||Counting efficiency|
|City of Edmonton||X||X||X||X|
|City of Markham||X||X||X||X||X|
|Halifax Regional Municipality||X||X|
|Cape Breton Regional Municipality||X||X||X|
|Town of Truro||X||X||X||X||X|
First, there are particular problems or situations for which Internet voting is suggested to provide some solutions. These include a drop in voting turnout, a desire to increase accessibility of the electoral process, and a wish to increase opportunities for external voting, especially for those not present in the country. A second group of factors is more broadly contextual, and includes factors that are raised in support of or opposition to Internet voting operations. This often involves a general extension of e-government operations, desires to stimulate technological development in a country or jurisdiction, concerns about the security of Internet voting, and concerns about voting secrecy where voting does not take place in the polling place. Finally, there are overall elements of context that are more general still, such as a general expectation among the population for more personal control over voting, similar to that which has occurred in other areas of life.
When some of these contextual considerations are more prominent, authorities may be more or less motivated to undertake public consultations. Concern with voting turnout, for example, might be widespread and lead authorities to mount public consultations. Desires to increase accessibility for persons with disabilities, or for groups of citizens residing outside the country, could motivate consultation with specific groups representing those affected. Concerns for security, because of the technical nature of online voting, may lead only to consultation with specific experts who understand the technical aspects of the situation. This section of the report considers several of the most important contextual factors and examines how they relate to the degree to which consultation and evaluation mechanisms have been used. Recommendations will later be made about the desirability of including them in a framework for evaluation.
Concern Over Voting Turnout
Over the last two decades, many countries have experienced a decline in their levels of voting turnout. It is not the task of this current report to document this trend, but reference may be made to previous reports for Elections Canada (Pammett et al. 2001; Pammett and LeDuc 2003). The situation of low and/or declining turnout has been an important background factor to motivate examination of ways to improve access to the vote, on the premise that such structural changes may bring about changes in the participation situation. Thus, examinations of the electoral system have occurred in several Canadian provinces, and different systems have been put to public referendums, on the premise that public interest in voting might be revived by electoral systems that are better at representing their choices than the current one.
The interest in Internet voting fits into the general context of concern over voting participation because it is perceived by many as a means of increasing voting turnout by making the electoral process more accessible for potential voters. In particular, there is an assumption that remote Internet voting will be especially appealing to young citizens given their heightened rates of use of digital and mobile technologies compared with other cohorts, and their frequent participation in social networking forums. Without reviewing the literature on the turnout decline in detail (reviews can be found in Blais and Loewen 2011; Milner 2010), we can note that the much lower voting rates among young people and their tendency to congregate on and use the Internet makes online voting especially tempting for governments and policy-makers. For many, it has seemed a natural connection to examine the possible contribution Internet voting could make to boost the voting rate of the young. The straightforward hypothesis suggests that if young people could vote using their personal computers or other mobile devices, more might do so. This possibility is further supported by the reasons young non-voters give for failing to vote, which includes reasons of being "too busy" or the reduced accessibility associated with being away at college or university and faced with registration procedures that are perceived to be arduous. Furthermore, given that many young persons typically live outside their home constituency while completing their post-secondary education, offering them a simple opportunity to cast a ballot as a voter in their home riding presumably addresses registration concerns.
Hopes to raise, or at least stabilize, voting turnout are always in the background of discussions of Internet voting. Sometimes they are overt goals, as in the Canadian municipalities of Cape Breton, Markham, Peterborough (Ontario), and Truro. This was also the case in Switzerland (Chevallier 2009), where the frequent scheduling of referendums has meant relatively low levels of turnout at any given voting occasion, and has further meant that elections are seen as less important and therefore attract fewer voters than they might otherwise. Switzerland had implemented a system that fostered widespread access to postal voting beginning in the 1990s, and officials concluded that this helped to increase turnout. This history of improved access to the ballot being perceived as a success in Switzerland has led authorities to consider Internet voting as a further extension.
We will see when we examine the European cases in more detail in the next section of this report that any hoped-for "turnout effect" has been elusive; in some cases it was minor and in others non-existent. Caution because of prior findings led Norwegian authorities to explicitly state that increased turnout was not an overt reason for undertaking the 2011 experiment that used Internet voting trials in municipal elections. Nevertheless, it seems fair to say that the hope that it might do so was still there, particularly among advocates for Internet voting. Canadian cases have experienced mixed results where turnout is concerned, though the option of Internet voting is often well used by electors. In instances where telephone voting was also offered, Internet ballots have been the preferred choice by far (Goodman in press).
Contrary to expectations, however, those jurisdictions that cited increasing voter turnout as a goal for the implementation of Internet voting did not undertake public consultations about this proposed plan. This may have been because it was simply assumed that Internet voting opportunities would produce more participation, or it may have been that turnout increases were a secondary goal. It is also possible that authorities did not foresee how consultations would provide useful input on how the policy could be implemented.
Desires to Improve Accessibility
Persons with disabilities can have difficulty accessing polling stations because of mobility issues, vision problems, or other difficulties. Solutions involving postal ballots can be time consuming to arrange if special applications need to be made and materials delivered. Election officials have been conscious of this problem for a long time and have made efforts to improve accessibility of polling stations with ramps, and other modifications designed to enhance the equality of the electoral process. A number of devices to assist persons with disabilities have been made available in poll locations, and there are provisions for people requiring assistance to have someone of their choice help them cast a vote. It is in this context of widespread public and official consciousness of the needs of persons with disabilities that Internet voting is discussed as a further means of improving accessibility of the vote.
Consultation with this goal in mind has primarily been with groups representing persons with disabilities. These groups have been active in demanding easier access to the vote, among other services, in most countries. Mobility and vision issues have been the most prominent disability areas where groups have been active. In Ontario, for example, the Accessibility for Ontarians with Disabilities Act Alliance (AODA Alliance) issued a letter to Elections Ontario in December 2012 expressing "serious concern" for the fact that the agency has not made better progress on trialling Internet and telephone voting given that "the Ontario Legislature directed Elections Ontario to study these alternative voting options over two and a half years ago" (AODA Alliance 2012).
In 2010, the Canadian Human Rights Tribunal heard a case on behalf of elector James Peter Hughes regarding his inability to vote because of accessibility problems in a 2008 federal by-election and subsequent regular general election. Part of the Tribunal's ruling stipulated that Elections Canada should make efforts to improve facility accessibility for potential voters. Improvement of accessibility is a common theme in electoral discussions at all levels of government because it is necessary to ensure equality of the vote and maintain the integrity of the electoral process (Hughes v. Election [sic] Canada). As a result of increasing numbers of requests from citizens with disabilities and advocacy groups to move to electronic voting, many Canadian municipalities have made the decision to do so primarily to bolster efforts to accommodate persons who may encounter difficulties exercising their democratic right to vote. The communities of Cape Breton, Edmonton, Halifax, Markham, Truro, and the other cases featured in this report have expressed these sentiments.
The context of external voting is also relevant to the topic of accessibility. The impetus to develop I-voting systems in several countries has been stimulated by a need to provide improved electoral accessibility for citizens of that country living abroad than the traditional voting methods allow for. These traditional methods have generally required citizens abroad to request a paper ballot and return it through the mail at a point in advance of the election. This is a cumbersome procedure, even when election dates are known well in advance. In voting systems where choice of specific candidates is a requirement, the names of these candidates may not be known long in advance of the election. Where choice of candidates within a party is an option, external voters may not be able to exercise that option, and may be limited to choice of party. In either case, the full range of privileges of democratic electoral choice is not available to them, simply because they are not in the country at the time of the election.
Voting from Abroad: The International IDEA Handbook, published in 2007 by the International Institute for Democracy and Electoral Assistance, lists five methods of external voting (International IDEA 2007). These are voting in person (generally at embassies or consulates in the foreign countries), voting by proxy, voting by post, voting by fax transmission, and I-voting. Sometimes these methods are available in combination. At the time the handbook was published, a few countries were experimenting with I-voting systems and several others (i.e. Switzerland, France, Spain, Austria) had expressed an interest in it. Since then, interest in offering this method of external voting has increased.
In Estonia, the use of I-voting systems by foreign situated citizens has been limited by the necessity of having the Estonian national identity card and the required card reader attached to their personal computer to read it. Since Estonians residing abroad on a long-term or permanent basis had little need for this card, most had not acquired it. Furthermore, foreign countries where large numbers of Estonians lived lacked importers that would bring in the card readers for sale (Martens, October 15, 2012). For Estonians living in Canada, these problems have since been resolved.
In Switzerland, a group for overseas voters called the Organisation of the Swiss Abroad, headquartered in Bern, the federal capital, has been very active in demanding voting rights for their membership. Switzerland has since implemented I-voting for the Swiss living abroad, and its use has not been encumbered with the restrictions implemented for Internet voting in the territory of Switzerland itself.Footnote 3 In Canada, many municipalities have noted that citizens living abroad or travelling have made use of Internet voting to cast ballots. In the 2012 Halifax election, for example, questions from those residing afar even came in via Twitter (McKinnon, December 11, 2012).
Another group of people with a special concern about being away from their polling stations during election time is students. Elections, whether regularly scheduled (a growing trend in Canada) or held irregularly after a dissolution of the legislature, are frequently held at precisely the time when students are most likely to be away from home and least able to vote conveniently in them. This is most evident in the fall, which is the most popular season for Canadian federal elections (LeDuc and Pammett 2006). For university students attending school in locations other than their hometown, this presents difficulties in either voting at home or changing their registration to permit voting in their school location, if they are able.
In Nova Scotia, for example, recent legislation allows a student from another municipality who has lived in the community for at least one year and is enrolled in the second year or greater of a post-secondary school program to be added to the voters' list and cast a ballot in that community (White, October 15, 2012). Voting in school locations can be made arduous, however, by requirements for identification connecting students with their new address. For these reasons, attention to student voting demands has involved examination of the possibility of Internet voting. An Internet voting option can allow students to maintain their home registration address and cast a remote vote for the contest in that location (Goodman 2011).
Extension of E-Government
E-government is a development that has occurred in most countries, but to varying degrees. It signifies the use of Internet technology to connect government with citizens and citizens with government. The government-to-citizen connections relate to such things as information about citizen entitlements to government services like pensions, employment insurance, and health care. As well, citizens increasingly have the option to file income tax returns electronically and get answers to questions about government policies. While the government-to-citizen component of the relationship is relatively well established, the citizen-to-government aspect of e-government has unrealized potential. The Internet technology of communication can theoretically allow citizen input into government as well as the reverse.
European governments that have piloted Internet voting all have extensive e-government operations. Estonia has coordinated these into the application of a national identity card, which can be inserted into a card reader attached to a personal computer and used to access all government departments where e-government services are offered. It gains access to items such as personalized tax services, and pension and health information. It can also be used to access banks and other personal information outside the government arena. Interaction between the public and government takes place on a regular basis over the Internet, and in this case the extension to voting seems both expected, and normal (Tallo, October 16, 2012). In Canada, this trend is also established, and some cities and townships have tried to use social media as a tool to inform and engage electors at election time (Goodman and Copeland 2011).
If there is a system of e-government in place, electronic voting is often presented as a natural, almost organic development. In this way, it is seen as an extension of the ways in which citizens and government interact over the Internet. The establishment of one service after another is treated in such states as a normal and expected development. Furthermore, if citizens are accustomed to interacting with government electronically and are satisfied with doing so, they will expect further electronic interactions. In such circumstances, formal public consultations may be perceived as unnecessary. Furthermore, the results of such consultations might be demands that the process be implemented fully, or right away, whereas governments may want to proceed cautiously.
Furthermore, another set of attitudes accompanied the implementation of Internet voting trials and systems in the European countries under examination, notably a faith in technology and consequent high Internet penetration. Internet World Stats, a marketing organization, ranks Norway second in the world in Internet penetration (97.2 percent of the population), Switzerland 13th (84.2 percent) and Estonia 28th (77.5 percent) (Internet World Stats 2012). Cellphone use is virtually universal. (Norway uses cellphones to receive codes that verify voting choice, and Estonia is developing such a system.) Government policy in these countries supports Internet expansion and use. In such an atmosphere, arguments for extending voting methods to include the Internet are regarded by many as natural extensions of current practice.
Rates of Internet penetration in Canadian municipalities are fairly robust. Recent studies report that Canadians are the most frequent Internet users worldwide, spending an average of 45 hours per week using the Internet (Ladurantaye 2012). Statistics Canada data from 2009 confirm extensive use among Canadians, reporting an average usage rate of 80 percent among citizens, 96 percent of whom access the Internet from their homes. Worries about unequal access to fast connection speeds are also dwindling, as 94 percent of those from communities with more than 10,000 residents say they have a high-speed Internet connection. Rates of Internet use among Canadians are reportedly highest among British Columbians and Albertans (at 85 percent), followed by Ontario residents (at 81 percent) (Statistics Canada 2011). In Nova Scotia, where there has been significant uptake of Internet voting, about 76 percent of residents have Internet access, 74 percent from home (Statistics Canada 2010). Despite this slightly lower rate of penetration, the provincial government in Nova Scotia has made delivering high-speed Internet access a priority, declaring it "as essential as electricity and telephone service" on the government website (Government of Nova Scotia 2012).
Stimulating Technological Development
For some European governments, such as those in Estonia and Switzerland, the development of Internet voting systems has played a part in stimulating the local science and technology industries. For Estonia, its position in Northern Europe in proximity to Finland, home to Nokia cellphone development, gave an impetus to the development of its own technology industry. Electronic voting in Estonia was a way to provide a challenge to homegrown experts (Martens, October 15, 2012); experts were assembled to consider the feasibility of e-voting, and presented with the challenge to develop it. Switzerland, as well, with its long tradition of developing precision instruments, was home to a pool of experts, who could be tapped to develop an Internet voting system as an outgrowth of the establishment of other e-government services. Norway, very advanced in terms of Internet penetration, was stimulated by the remoteness of some of its hinterland communities.
For the three European countries under examination, Internet voting has become something of a symbol, an internationally recognized manifestation of the advanced developed nature of the country. We may refer to this as a "branding strategy" whereby the country becomes known throughout the world for its advancements in technology generally. It creates an image of a country that is in touch with its own people, and with its experts. The eye-catching, and democratically important, Internet voting area becomes a leading edge of a country's technological prominence generally. This is important, not only for development of indigenous entrepreneurs and business, but also for attracting foreign capital to invest in technology industries. Estonia in particular is known worldwide for the fact that it has the most extensive system of Internet voting in all elections. People visit Estonia to study it, and national pride becomes involved. In the case of a perceived need to stimulate technological development, an assembly of experts would be a way of ascertaining the particular means of doing so. In such a case, the policy direction has been determined in advance, and the consultations are only about the means to accomplish it.
For Canadian municipalities, the advent of Internet voting has not necessarily been seen as a means of stimulating development (because they all contract out their I-voting operations), but for many it has been regarded as a means of becoming a leader in e-government at the municipal level. Being the first major Canadian municipality to introduce Internet voting was a motivator for the City of Markham. Recognition as a leader in e-government is continuing to influence the way Markham uses the Internet to reach out to electors, particularly through social media (Goodman and Copeland 2011). On the east coast, Halifax, Truro, and Cape Breton are examples of communities that hoped adopting Internet voting first in their respective areas would help demarcate their municipality as progressive, trendsetting, and focused on keeping pace with other services in society. Finally, after studying the landscape of Internet voting in Canada, the City of Edmonton identified Markham as a municipal leader it would like to emulate (Kamenova, September 9, 2012).
The discussion of Internet voting inevitably takes place in the context of extensive discussion regarding the security of the system. Our purpose here is not to debate the merits of these concerns, but rather to consider the place they play in the deliberations that precede decisions to undertake the enterprise or the evaluations that are made of it. The very fact that security issues – whether they relate to the ability of hostile forces to disrupt an election by "denial of service" or "hacking," or a more insidious possibility, that changes may be inserted into the choices made by voters so that a false result may be obtained – are always in the background and have affected the ways in which the issue has been presented to the public.
Opponents of Internet voting (Jefferson et al. 2004; Jones and Simons 2012; Simons and Jones 2012) often adopt a stance that asserts that Internet voting cannot (and should not) be safely implemented because security can never be assured. Discussion of this position is not possible for those lacking technical knowledge, so the argument becomes a kind of "take it or leave it" opinion, in which lay audiences are forced to make a yes-no decision with no possibility of compromise. Trust in the operations of the system is attacked, and if the attack is accepted, the project is stopped or delayed indefinitely. Such attacks and commentary halted the US military from proceeding with its plans to use Internet voting in the 2004 election, and though efforts have been made to keep the project alive, a decision was made once again in the 2012 election to not use the system.
The way in which security concerns have been presented has had an impact on the willingness of jurisdictions considering Internet voting trials to involve the public in a deliberative process leading up to the decision. Processes of widespread public consultation will inevitably bring about submissions maintaining that security concerns trump all other factors and that the efforts should be stopped immediately. In this context, some jurisdictions are reluctant to use open public consultation processes. It might be thought that security matters represent threats to public trust that must be overcome if Internet voting is going to be implemented successfully over the long term. Therefore, one might reason, they should be thoroughly discussed and public agreement to proceed secured if such trust is going to be sustained. Our experience, however, is that this does not seem to prevail in most of the cases we have observed. When the topic of security is discussed in public forums, everyone expresses concerns relating to it. "Laypersons" (which includes most citizens) are presented with arguments about either the ability or inability of authorities to maintain the secure transmission of a ballot by Internet, but have no basis for judgment between such arguments. The larger the consultative group, the less likely it is to come to a consensus to accept or ignore security arguments, and the more likely it is to experience deadlock and frustration. Fear of such an outcome can prevent authorities utilizing such consultative forums.
Internet voting is a system that operates on a shared understanding between voters and administrators that the intended voter is the person who is casting the vote. When voting occurs in person, having the voter mark the ballot alone behind a screen and then place the completed ballot in the box imposes privacy. When votes are cast remotely, voters may be instructed to operate under the same regime of privacy, but there is no observed assurance that this is occurring. When postal ballots are filled out and mailed, and when Internet votes are cast, this process may not have occurred in secret.
Voting was not always done in secret, but this practice changed near the end of the 19th century and it has now become customary to do so. Lack of secrecy is considered to be a problem because there is the possibility coercion may take place during the casting of the vote, by mail or Internet. This is seen as more likely to occur when dominant family members exercise undue influence on other family voters, or when employers attempt to direct the votes of their employees. Some places are quite sensitive to this issue. For example, the two largest cities in Norway, Oslo and Drammen, originally applied to join the Norwegian pilot project in 2011, and those conducting the trials were keen to have them in order to see how the Internet voting system worked in large cities. Both cities later pulled out of participation, a development interpreted as being connected with the fact that these areas are the largest points of settlement of immigrant families in Norway.
Estonia has been sensitive to potential criticism that there may be some influence on Internet voters, and has established a procedure that goes at least some distance toward alleviating it. There is the possibility for an Internet voter to cast subsequent votes, which annul the previous ones. The reasoning for this approach is that if on one occasion a demand is made that a voter cast a ballot a certain way and be observed to do it, another ballot may be cast for the voter's true choice on a later occasion where the influence is not present. There is always the possibility, however, that those keen to influence votes will focus their coercion tactics near the end of the election, where because of time the chance of changing votes is reduced. To prevent this, a voter who has previously voted on the Internet may go to the polls on election day and cast a secret ballot there, which cancels all the Internet votes. Norway implemented a similar system in its 2011 Internet voting trials.
In Canada, voter secrecy has been imposed by reminding electors of the penalties for coercion or not voting in secret or by adopting regulations that provide for strict penalties. In the Town of Truro, for example, pamphlets were distributed to every household informing and educating electors about Internet voting and also reminding them of the importance of keeping their vote private and the consequences of not doing so. In Markham, penalties for breaches of voter secrecy were increased to fines of up to $10,000 or a potential two years in jail, but to date no charges have been laid.
Secrecy concerns can suffer difficulties when placed in the arena of public discussion. It is easily understood that any form of remote voting cannot ensure the secrecy of the voting process, so discussion tends to be channelled into a couple of corollary issues. One is whether secrecy in a remote voting situation is really important, essentially a values question that is not resolvable. The second issue, and potentially more serious, is the (often implicit) connection of a lack of secrecy with immigrant groups, where patriarchal families supposedly exist, or where patron-client relationships mean that votes are determined by others. Both discussions are difficult to conduct openly, and are not easily resolvable, since anti-immigrant attitudes can be rationalized and expressed as points of principle, tainting the discussion for both those who are expressing them and those who are genuinely arguing a point of principle.
The computer has not only made it possible for people to do things online, but it has also changed the culture of individual interaction with institutions. This process has been implemented with public acceptance but has been spurred by the actions of those institutions themselves, through positive or negative inducements (lower prices or "Internet bonuses" on the one hand, and reduced service hours or staffing on the other). One good example is Internet banking, not long ago an occasional activity for many, and now the norm. This situation began to change as a result of the bank machine, but evolved into use of the Internet for paying bills, checking account balances, transferring money, and accessing other services and information. Internet banking allows individuals to do without direct personal contact with the bank altogether, to do things themselves, at the time and in the way they want to. People have become personally empowered with regard to financial transactions, as these actions have become largely under their own control. Other major extensions of the empowering effect of the Internet involve shopping and travel planning.
These increases in personal empowerment have an electoral dimension, even outside the issue of the Internet. There has been a discernible change in the way in which voters interact with electoral institutions. This is most noticeable in the increased use of advance voting options, Internet or otherwise. According to Elections Canada, the number of voters at the advance polls rose in the 2011 federal election to over 2 million, from approximately 1.5 million in the previous two federal elections. Where postal ballots are easily available in order to vote in advance, they are very popular. In Switzerland, up to 95 percent of ballots are cast by mail, a figure that drops to around 80 percent if Internet voting is an option. Recent Canadian provincial elections support the conclusion that advance turnout is on the rise. Increases in advance turnout, however, do not always translate to a rise in overall voter participation. In Canadian municipalities that have offered Internet voting in advance polls, there has been a dramatic change in turnout. Before its introduction in the City of Markham, for example, advance turnout was limited to a couple of thousand votes. For the past two elections, around 10,000 voters have cast their ballots online in the advance portion of the election (Goodman in press). Advance turnout in Halifax has also noted a dramatic change. Whereas 27 percent of those who voted, voted online in the 2008 advance polls, in 2012, 59 percent of voters made use of the Internet option (McKinnon, December 11, 2012). Where available, Internet voting seems to promote voting in advance polls.
Return to source of Footnote 3 France has also implemented an Internet voting system for the French living abroad. A feature of the French election system calls for French citizens abroad to elect representatives to 11 special seats in the National Assembly.