Secondary menu

Consultation and Evaluation Practices in the Implementation of Internet Voting in Canada and Europe



In recent years, the Internet has quickly penetrated our economic, social, and political way of life. While it took radio 38 years to reach 50 million users, the Internet amassed the same user base in a mere four years. By 2012, more than 2.3 billion people were using the Internet. In developed countries, about 70 percent of households are online (ITU 2012). This uptake has far-reaching consequences for the activities that are carried out online and the goods and services citizens expect to be able to access digitally. As an increasing number of services are made available online, governments and election agencies around the world have been exploring the possibility of Internet voting systems in hopes of enhancing the accessibility of the electoral process, improving rates of electoral participation, and moving toward a more citizen-centred service model. In the past decade a number of countries have embraced Internet voting in elections at various levels of government. The focus of these jurisdictions, and of those studying Internet voting development, however, has primarily centred on the outcomes of these attempts, operational details of the models, impact on voter turnout, and security components of the systems (for example, Alvarez and Hall 2004; Alvarez, Hall and Trechsel 2009; Bochsler 2010; Goodman, Pammett and DeBardeleben 2010; Jones and Simons 2012; Madise and Martens 2006; Trechsel and Vassil 2010). There has been little focus on consultations undertaken prior to the deployment of Internet voting systems and the evaluation protocols undertaken afterwards.


The lack of attention to these topics in the literature is in part a reflection of the unsystematic way in which consultations have been undertaken in the various jurisdictions that have implemented Internet voting trials. Post-trial evaluations have been sporadic as well, particularly in the Canadian jurisdictions. This report focuses on consultation and evaluation procedures and processes used in some European and Canadian jurisdictions where Internet voting has been introduced or studied. Specifically, it examines the nature of public and group consultation that has been carried out in a variety of contexts in which Internet voting was established. Most jurisdictions have continued to use Internet voting methods after the initial trials. However, this situation should not be regarded as an endorsement on the part of the authors of this report of the consultation or evaluation methods used, or of Internet voting itself. The Internet voting trials and studies examined in this report took place in the European countries of Estonia, Norway, and Switzerland, and the Canadian provinces of Nova Scotia, Ontario, Alberta, and British Columbia.

Elections Canada's Strategic Plan 2008–2013 focuses on three key objectives: Trust, Accessibility and Engagement. All three of these aims are intimately involved with any evaluation of Internet voting. Trust in the fairness, security and transparency of electoral administration needs to be maintained, if not enhanced, while voting is being undertaken over the Internet. Only if the voting method meets these tests can it be successful. Accessibility is an important goal of all remote voting operations, including those using the Internet. Finally, the degree to which Internet voting promotes engagement of the citizenry, particularly youth, is central to debates over its utility. Based on these considerations, this report pays particular attention to the three key dimensions of trust, accessibility and engagement as reasons for public consultation about the introduction of Internet voting trials, and as bases for the evaluation of those trials.


The focus of this report is on Internet voting, also referred to as I-voting, online voting, or online ballots for purposes of stylistic relief.Footnote 1 This denotes the casting of a secure ballot using a device that is connected to the Internet, which typically takes place in an uncontrolled environment (Carter and Bélanger 2012; 2013). The term "electronic voting," by contrast, may refer to casting a ballot via an electronic channel, which may or may not be using an Internet connection. In this report, "electronic voting" is used to reference a combination of Internet and telephone voting (in the Canadian cases) and in other instances to signify methods of casting electronic ballots that are not necessarily supported by an Internet connection. "E-voting" is also used as a short form for electronic voting.


Internet voting has been used in a number of European countries. For example, a number of the constituency trials in the United Kingdom employed this voting method between 2003 and 2007, often in combination with other experiments, such as widespread use of postal voting, voting by telephone, and by text message (Boon, Curtice and Martin 2007). France, too, employed Internet voting for the parliamentary elections of June 2012 for its citizens abroad (OSCE-ODIHR 2012c). The European illustrations in this report are drawn, however, primarily from three European countries – Estonia, Switzerland and Norway – which have seen the most extensive uses of Internet voting to date and where, in the first two cases, this procedure has been employed over a longer period of time. Additional interview material was obtained from these three countries in the fall of 2012, which allowed for specific questions to be asked about consultations. Evaluation frequently results in published reports, where the criteria employed may be readily ascertained, but interviews also included this topic.

Canada also has made extensive use of Internet voting in regular local elections. This first occurred a decade ago in twelve municipalities in the Province of Ontario. Since then there has been notable uptake by about 60 municipalities in the provinces of Ontario and Nova Scotia. This report closely examines the two largest municipalities to have used Internet voting, Halifax and Markham, Ontario. Halifax has offered Internet voting in two regular elections and one by-election and Markham has incorporated online ballots in three consecutive elections. Cape Breton Regional Municipality, the second-largest municipality in Nova Scotia, is also studied here, given the research and development that municipal officials carried out prior to implementation. As well, a smaller Nova Scotia community, the Town of Truro, is included because of some of the unique features of its model and application. Finally, the City of Edmonton and Province of British Columbia are examined based on their consideration of Internet voting and the different types of consultation processes they have employed to assess the appropriateness of online ballots for elections.

Some information for this report was gathered from secondary sources, including printed materials and reports. Primary research comes from personal interviews conducted by the authors during field trips to the locations in the fall of 2012.Footnote 2 These interviews were formally unstructured, but followed an overall script designed to address relevant aspects of consultation and evaluation procedures (see Appendix 1 for specific questions). Questions addressing consultations were designed to inquire about the nature of the consultations, which groups and individuals were consulted, the stage of the proceedings at which consultations occurred and whether official reports were made. Questions regarding evaluation examined the criteria and indicators that were used, the degree of formality of the process and whether written reports were produced. All interviewees were made aware of the nature of the research prior to speaking. Those interviewed were selected based on their role in a particular organization, agency, or jurisdiction and their knowledge and experience with consultation and evaluation procedures related to Internet voting.

Footnote 1 This differentiation in terminology is reflected in the literature addressing Internet voting (see for example, Bélanger and Carter 2010; Carter and Bélanger 2012; Carter and Campbell 2012; Goodman, Pammett and DeBardeleben 2010; Stenerud and Bull 2012).

Footnote 2 A list of these interviews is included at the end of the bibliography.