ELECTION LAW JOURNAL

Volume 19, Number 2, 2020

# Mary Ann Liebert, Inc.
DOI: 10.1089/elj.2019.0568

Protecting Electoral Integrity in the Digital Age:

Developing E-Voting Regulations in Canada

Aleksander Essex and Nicole Goodman

ABSTRACT

As elections around the world become digital, governments have begun adopting regulations to govern the
Downloaded by 223.187.185.185 from www.liebertpub.com at 08/04/21. For personal use only.

use of voting technologies and protect electoral integrity. Canada, however, is an exception. Despite the
prevalence of voting technologies in Canada’s local elections, notably online voting, no regulation frame-
work has been initiated. In particular, there are no guidelines or standards surrounding the use of online
voting. While research documents online voting has positive effects for participation, implications for
the integrity, accountability, and transparency of elections are stark. Canada’s multilevel governance struc-
ture has meant municipalities mostly deliver elections on their own terms, resulting in a patchwork of
online voting models and cybersecurity requirements. Many municipalities also lack the resources to vet
vendor solutions adequately, and an increasing number of cities are eliminating paper voting. These con-
ditions highlight an urgent need to regulate the design and procurement of election technology in Canada.
To proactively respond to these developments, this article draws upon interviews with select officials and
experts and regulation models in other jurisdictions to argue for a new model of electronic voting regulation
that would be a good fit for Canada.

Keywords: election law, voting standards, electronic voting, online voting, Canada, municipalities

INTRODUCTION for example, have become commonplace in the

United States whereas electronic voting machines

I n recent years, governments and election

management bodies around the world have
begun to digitize voting in elections. In some cases
are the norm in India. Use of online voting, by com-
parison, has grown in places like Estonia, Switzer-
land, and Canada while trials have been cancelled
this has involved the adoption of electronic tabula- in Norway, France, and the UK. Notwithstanding
tors or ballot scanners, while in others it has meant differences in approaches across jurisdictions, adop-
voting remotely online. Digital voting solutions tion of electronic voting has been on the rise.
vary by country and context. Electronic tabulators, These technological changes to the way people
vote prompt questions about regulation and
Aleksander Essex is an associate professor in the Department of whether different governance and policy arrange-
Electrical and Computer Engineering at Western University in ments are required to manage electronic voting,
London, Ontario, Canada. Nicole Goodman is an assistant pro- which we define as including ‘‘Direct Recording
fessor in the Department of Political Science at Brock Univer-
sity in St. Catharines, Ontario, Canada. Authorship is equal and
Electronic (DRE) voting machines, ballot scan-
authors are listed alphabetically. The authors sincerely thank ners, digital pens, and Internet voting systems’’
interview participants for their time and insights, and also (Council of Europe 2017). In many cases countries
extend their thanks to Evan Young for research assistance and have responded by putting standards or regulatory
Matthew Bernhard, Michael Boda, Ardita Driza Maurer,
Beata Martin-Rozumi1owicz, Poorvi Vora, and the three anon- arrangements in place, either proactively or reac-
ymous reviewers for helpful feedback. tively, but typically before electronic voting usership

162
 DEVELOPING E-VOTING REGULATIONS IN CANADA 163

has reached a critical mass. Canada, however, is an were fully electronic, some municipalities were
exception. Despite the fact that local elections forced to declare a state of emergency and extend
across the country are highly digital, there are no voting a full day to ensure electors were not disen-
safeguards in place such as standards, guidelines, franchised. This called electoral legitimacy into
or bodies that provide certifications to regulate question in many of the affected communities. It
electronic voting. also raises questions about whether it could have
Electronic tabulators, for example, have become been prevented with the support of a regulatory
the norm for vote counting in municipal elections, framework. Had technical standards existed, for ex-
and in the provinces of Ontario and Nova Scotia, ample, would the bandwidth issue have occurred?
online voting is used by more than 240 communities. The multilevel governance structure in Canada
In many cases, voting by telephone is also an option. gives municipalities autonomy to use electronic vot-
These election technologies are seen as enhancing ing, providing overarching legislation written by
administrative efficiencies and improving voter ac- the provinces including a provision that allows for
cess. While online voting in particular has been ‘‘electronic’’ or ‘‘alternative’’ voting. The result of
found to increase voter turnout (Goodman and this governance structure is that election technolo-
Stokes 2018), especially for marginalized popula- gies are contracted from private vendors by munic-
Downloaded by 223.187.185.185 from www.liebertpub.com at 08/04/21. For personal use only.

tions (Germann and Serdült 2014; Goodman et al. ipalities, oftentimes without rigorous evaluations
2010), and deliver counting efficiencies (Goodman because many communities do not have the resources
and Pyman 2016), concerns have been raised about or capacity to adequately vet vendors or put in place
the lack of guidelines to shepherd its use. The ab- measures to protect who has access to, and owns,
sence of regulation is additionally problematic in election data. As the scenario above illustrates, the
the Canadian case since many local governments circumstances surrounding use of electronic voting
eliminate paper ballots and run fully electronic in local elections in Canada present the perfect
elections. Voting is either carried out exclusively storm to potentially compromise electoral integrity.
online or via a combination of online and telephone As the only country in the world running fully un-
ballots. While some countries run in-person paper- supervised electronic elections, Canada is a unique
less voting, such as India’s electronic voting ma- case other jurisdictions can learn from. This article
chines (EVMs), or the numerous U.S. states using draws upon interviews with experts and officials,
paperless DRE touch-screen voting machines and other international regulation models to argue
(Verified Voting 2018), this voting takes place in for a way forward to regulate electronic voting in
supervised locations and does not rely on an Inter- Canada. Specifically, we propose a new regulation
net connection. To our knowledge, Canada is the model structured around actionable principles and
only country to use remote online voting in binding guidelines. While the article discusses electronic
elections without a paper ballot option. voting broadly, as defined above, it primarily focuses
The absence of electronic voting regulation leaves on online voting since we argue it presents the
local governments vulnerable and poses an increas- greatest risk to electoral integrity. At times, we
ing threat to the health of democracy in Canada. Use have employed different terms for stylistic relief.
of online voting in the 2018 Ontario municipal elec- The article proceeds in six sections. First, we re-
tions provides a case study highlighting this risk. In view relevant literature addressing electronic voting
these elections, 49 of 177 municipalities were af- regulation. Next, we discuss why Canada is a com-
fected by a technical issue that severely limited pelling case and what scholars, election administra-
server bandwidth and prevented online votes from tors, and practitioners can learn from Canadian
being cast for several hours during the evening of developments, or lack thereof. Third, the article de-
Election Day. The vendor, Dominion Voting Sys- scribes the approach taken to carry out the inter-
tems, issued a statement explaining their network views that inform the discussion. We then explore
infrastructure subcontractor (e.g., co-location pro- what other select jurisdictions and intergovernmen-
vider) placed ‘‘an unauthorized limit on incoming tal organizations are doing to regulate electronic
voting traffic’’ which resulted in 1/10th the pre- voting in the United States and Europe. Fifth, we dis-
agreed network bandwidth being available (Domin- cuss what regulation could look like in Canada along
ion 2018). This caused election websites to slow four dimensions: election principles, and technical,
down and even appear offline. Since many elections operational, and legal considerations. We conclude
 164 ESSEX AND GOODMAN

by discussing implications for electoral integrity, approaches to regulation have tended to organize
takeaways for scholars and practitioners, areas of ac- proposals for improvement around one of these di-
tion for government, and topics for future research. mensions (see, for example: Maurer, 2017), albeit
some recent writings are more balanced and assess
the progression of regulation in a particular jurisdic-
EXISTING LITERATURE tion over time (Maurer 2017) or offer a cross-
ON ELECTRONIC VOTING REGULATION disciplinary perspective by bringing together experts
from multiple fields (Grimm et al. 2007).
To date, there is a modest body of literature on These contributions highlight that the regulatory
the regulation of electronic voting, and these contri- environment of electronic voting is evolving but still
butions are mostly international. Aside from a 2013 very much in its infancy. Many models of regulation
report commissioned by Canada’s federal election remain siloed to the extent that they differ by juris-
agency, which outlines a proposed legal framework diction and context, and also since some often focus
for e-voting in national elections (Schwartz and on certain aspects (e.g., security) more than others
Grice 2013), we are not aware of any studies that ex- (e.g., operations). The approach presented by the
plore the topic of electronic voting regulation in a Council of Europe, which centers on the principles
Downloaded by 223.187.185.185 from www.liebertpub.com at 08/04/21. For personal use only.

Canadian context, despite mention of the need for of democratic elections, is broader in scope but its
standards in the reporting of some provincial elec- voluntary nature means that each application differs
tion agencies (e.g., Elections Ontario 2018) and and focuses on the aspects desired by a particular
other reports (Gabel and Goodman 2020). Many government. Much like electronic voting deploy-
of the international contributions that address elec- ment may work in one country or region but not
tronic voting regulation are focused on the Council in others, regulation presents similar challenges.
of Europe’s work (Stein and Wenda 2014; Driza While other countries have taken action on regula-
Maurer 2014; Maurer 2017) and European stan- tion, Canadian authorities have been slow to do so
dards more broadly (Remmert 2004), or specific likely given the limited use of electronic voting at
country cases such as Germany (Schmidt et al. higher levels of government, albeit provincial au-
2009). Some also compare European developments thorities have recently begun using tabulators for
with certification standards in the United States (Pran- sub-national elections, and territorial governments
dini and Ramilli 2012). In addition, one study exam- are experimenting with online ballots for absentee
ines potential security requirements for online voting voters (Elections BC 2018; Elections Ontario 2018).
in non-political contexts (Grimm et al. 2007), while Lack of action on the regulatory front is not un-
others outline the necessary security measures for like other areas where the application of technology
technical implementation (Langer et al. 2008). for governance has raised concerns and legislators
There are different ways to approach discussions have tended to be reactive rather than proactive.
of regulation—some studies focus solely on secu- For example, there have been increasing calls for
rity (e.g., Chang-Fong and Essex 2016) and/or oper- Canadian authorities to regulate social media
ational aspects (Langer et al. 2008), whereas others firms like Facebook and Google, particularly in an
address legal and constitutional aspects (Mitrou election context (Thompson 2019), and take action
et al. 2002; Schwartz and Grice 2013). In fact, the on data governance in high-tech development pro-
bulk of contributions on regulation, standards, or jects (CBC News 2019). Discussions regarding the
processes of e-voting are organized around either regulation of new technologies affecting govern-
legal, technical, and/or operational elements, and ment operations are evolving and seem to differ
political or policy considerations. This likely has by policy area. Since elections are a sacred demo-
to do with the separate nature of scholarly disci- cratic institution and the ‘‘lifeblood’’ of democracy
plines where computer scientists have tended to (Franklin 2004), regulating the use of technology in
focus on the more technical aspects (Culane et al. voting contexts should continue to develop and is a
2017), legal scholars have addressed constitutional necessity to maintain public confidence in free and
elements and voting principles (Schwartz and fair elections.
Grice 2013), while political scientists have been As noted above, contributions addressing regula-
more likley to explore policy (Gabel and Goodman tion of electronic voting in Canada are limited. Two
2020). Even contributions that critique existing exceptions are a study by Schwartz and Grice
 DEVELOPING E-VOTING REGULATIONS IN CANADA 165

(2013), which focuses on establishing a legal frame- A second feature of the Canadian case is its mul-
work for e-voting at the national level, and a 2020 tilevel governance structure wherein, as noted in the
report by Gabel and Goodman that discusses the introduction, the legislation governing municipal
importance of electronic voting standards for Indig- elections is written by the provinces. Beyond author-
enous communities. Beyond these studies, this ar- ing this legislation, however, provincial and federal
ticle is the first to propose a model for tackling authorities have no involvement in local elections.
electronic voting regulation in Canada while paying Providing the Acts include the necessary provisions,
greater attention to security and operations. municipalities have the latitude to customize their
elections to fit the unique needs of their community.
While legislation in each province differs, electronic
ELECTION TECHNOLOGY IN THE voting is typically made possible via a clause that
CANADIAN CONTEXT allows for ‘‘alternative voting methods’’ or other
language to that effect. The openness of this lan-
Electronic voting developments in Canada make it guage and lack of accompanying regulations has
a unique case that can shed light on the development resulted in a patchwork of electronic voting models
of election technology regulation in an era of digital and local policy led by municipalities (Goodman
Downloaded by 223.187.185.185 from www.liebertpub.com at 08/04/21. For personal use only.

governance. Specifically, the Canadian case is distin- and Pammett 2014).

guished by several features: (1) the types of voting In most other cases where voting technology is
technology and approaches to implementation, (2) used, governance and/or regulation is led or shared
the multilevel governance structure, (3) the unique by national or sub-national governments. In Swit-
local characteristics and municipal resources, and zerland, for example, where online voting has
(4) the absence of electronic voting regulation. been readily used at all three levels of government
First, as noted at the outset of this article, Canada primarily for referendums, there is substantial in-
presents a unique case in terms of the voting tech- volvement from the federal government, and munic-
nologies used in local elections. For one, Canada ipal votes are run in cooperation with the cantons
is the only country where voting by telephone is (Germann and Serdült 2017). In Norway, by com-
an option. While other jurisdictions have experi- parison, online voting pilots in select municipalities
mented with SMS text message voting (e.g., the in 2011 and 2013 were orchestrated by the Ministry
United Kingdom) (Norris 2004), the use of a tele- of Local Government and Regional Development,
phone line to cast a ballot in a binding election is the body responsible for national and local elections
presently not actively offered in government elec- (Segaard et al. 2013; Saglie and Segaard 2016).
tions in any other country of which we are aware. Involvement from higher levels of government is
In Canada, specifically, some technology vendors typical in jurisdictions where election technologies
also offer multi-channel voting, which allows a are used in communities, specifically online voting.
voter to commence their ballot with one voting chan- This upper-level involvement is usually responsible
nel (e.g., telephone) and then complete it using an- for initiating regulation, which helps explain the ab-
other voting method (e.g., online voting). A second sence of such frameworks in Canada. While in some
aspect is the widespread elimination of paper voting. cases a provincial elections body (e.g., in Ontario)
In a majority of municipalities that offer electronic has made suggestions to the provincial legislature
voting (either online ballots or a combination of to address the lack of regulation, no action has
online and telephone), paper voting is being halted, been taken. Likewise, calls from academics for the
and elections are run with solely electronic options. federal government to take a role on this issue
While research has begun to examine the effects of have gone unheeded (Essex and Goodman 2017).
these changes on voters (Goodman et al. 2018a), it A second consideration regarding multilevel gov-
presents an issue for electoral integrity if the only ernance is the growing use of electronic voting by
method available encounters a technical problem Indigenous communities, specifically First Nations.
and there is no alternative. As mentioned, this spe- There are approximately 634 First Nations in
cific fear was realized in the October 2018 municipal Canada, and more than 80 have used voting technol-
elections in Ontario. The absence raises the stakes for ogies to support their elections, referendums, and
potential problems and puts the integrity of local other types of votes (Gabel et al. 2016; Goodman
elections in Canada at greater risk. et al. 2018b). First Nations elections and referendums
 166 ESSEX AND GOODMAN

can be governed by legislation written by the federal have few to none. This highlights another layer of
government or a self-government agreement or cus- the piecemeal nature of election technologies in mu-
tom community election code passed by communities nicipalities in Canada.
that allows them to resume control over their election
and referendum processes. This aspect is important
to point out, since many First Nations using elec- APPROACH AND METHOD
tronic voting face similar implementation challenges.
While beyond the scope of this article, there is a pos- This article draws on semi-structured interviews
sibility that the regulations developed for municipal- with 23 officials from varied sectors to better under-
ities would be valuable for First Nations, providing stand the challenges governance faced when using
they were not imposed (Gabel and Goodman 2020). electronic voting in local elections. Interviewees
Third, the unique characteristics and resources of included a mixture of municipal officials, security
municipalities have exacerbated ad hoc develop- experts, scholars, representatives from other rele-
ment. Issues of technical capacity, budgets, and vant municipal and provincial organizations, and
the vetting of vendors differ significantly across mu- technology vendors working in Canada to provide
nicipalities. While one community may have an IT contextual insight as well as broad perspective re-
Downloaded by 223.187.185.185 from www.liebertpub.com at 08/04/21. For personal use only.

team and the financial means to carry out a risk- garding what type of regulatory solution might
limiting audit or specialized technical assessment, work best in Canada’s unique political environ-
other places have no IT staff and rely on the good- ment. The goal of engaging a varied group of par-
will of vendors to deliver a safe election with a ticipants was to gain perspective from many
sound product. Said another way, some communi- stakeholders to ensure a proposed solution was a
ties run elections for more than 300,000 electors reflection of a diversity of voices in the electronic
and have election budgets that exceed $500,000 voting space in Canada. Participants were asked to
(CAD), whereas others have fewer than 10,000 elec- provide insight into the type of guidance that
tors and can have a budget of approximately would be most appropriate in Canada (e.g., guide-
$25,000. This variance in the local government re- lines, standards, principles), what these should
sources has contributed to uneven development. look like, and whether these should be voluntary
Finally, the legislation governing municipal elec- or mandatory. Questions also focused on issues
tions is dated and provides no guidance to local with the language of the Municipal Elections
bureaucrats regarding the administration of technol- Acts and specific recommendations regarding
ogy in elections. Ontario, which has the largest use technical, operational, or legal standards, guide-
of electronic voting, presents a useful example. The lines, or principles that should be considered in
Municipal Elections Act in Ontario is highly pre- Canada.
scriptive regarding the use of paper ballots in elec- Interviewees were contacted by e-mail with the
tions. The statute clearly outlines how to conduct questionnaire and asked if they would be willing
a polling location and provides guidance on a mat- to take part in a 45-minute interview. To encourage
ter such as what voters and candidates can and can- interviewees to be as open as possible they were
not do in a voting place (e.g., respect for privacy, assured their comments would not be cited directly.
the level of assistance that can be provided, etc). While we provide a list of interviewees who con-
Beyond permitting electronic equipment for the sented to having their names and titles appear in
counting of ballots or alternative voting methods, an interview list in the appendix, there is no direct
there is no guidance about how municipalities attribution in the article. Interviews were conducted
should govern the electronic aspects of elections. by phone between January 7 and 28, 2019. While
One consequence is that some candidates go door- efforts were made to contact municipal officials out-
to-door with iPads encouraging people to vote for side of Ontario, these individuals did not respond
them. In such situations, having guidance on rights within the interview period. All interviewees were
to privacy and undue influence that govern candi- made aware of the nature of the project and recieved
date behavior would be an asset for promoting elec- a draft copy of the article prior to publication to fa-
toral integrity. This lack of overarching direction cilitate any comments and fact checking. There was
has meant that while some municipalities write no conflict of interest between the authors or any
their own bylaws governing these aspects, others interviewees.
 DEVELOPING E-VOTING REGULATIONS IN CANADA 167

In the sections that follow we present our own ar- began following controversies during the 2000
gument regarding a proposed regulatory solution for U.S. presidential election with the Help America
Canada that may also work in other political environ- Vote Act (HAVA, 2002), which in turn established
ments, particularly ones characterized by multilevel the Election Assistance Commission (EAC).
governance arrangements. Material from interview- The EAC has multiple roles. It operates as a clear-
ees was used to provide necessary insight to formu- inghouse for information about election administra-
late a proposed regulatory framework, but did not tion and administers the distribution of federal funds
dictate the parameters of the framework itself. to states for the implementation of HAVA objectives.
With the help of the National Institute of Standards
and Technology (NIST), the EAC serves as the cre-
APPROACHES TO REGULATION
ator and maintainer of the Voluntary Voting System
Guidelines (VVSG), which provide guidelines both
Before thinking about which approach is best to
for voting system performance and national certifi-
regulate electronic voting in Canada, it is essential
cation and testing. The EAC also oversees a national
to examine strategies used in other countries and or-
program for testing and certifying voting systems.
ganizations. Three models are notable. First, the
A key feature of this regulatory environment is its
United States takes what we refer to as a prescriptive
Downloaded by 223.187.185.185 from www.liebertpub.com at 08/04/21. For personal use only.

voluntary nature; states may decide to adopt ele-

approach, which places a strong focus on system cer-
ments of the VVSG in whole or in part. This has
tification and testing. Although adoption is voluntary,
resulted in states adding or subtracting from the fed-
most states legislate some aspect of federal require-
eral standard, ultimately using it as a template to ex-
ments. Second, the broad-based approach taken by
press their own needs. Several categories of state
the Council of Europe provides recommendations
legislation exist including requirements to test fed-
for intergovernmental standards for a range of voting
eral standards, requirements for testing to be done
technologies that can be voluntarily applied by 47
by federally accredited laboratory, and requiring
member nations and non-member states. This strat-
federal certification (U.S. EAC 2009). Currently 42
egy offers sufficient latitude to be widely applied in
states legislate some aspect of the federal testing
a variety of contexts and legal frameworks. Finally,
and certification requirements, and 12 require full
Switzerland embodies a top-down, shared approach
federal certification (NCSL 2018). Even states that
where the direction is given by the Federal Chancel-
do not require federal certification typically still
lery, but the cantons are responsible for compliance.
rely on the federal program to some extent, and
This model is focused on certifying voting technol-
many still choose to use voting systems that have
ogy and providing the tools for auditing elections,
been federally certified.
limiting the number of electors that can be offered
This approach has led to mixed success and has
online ballots until security thresholds are met.
taken years to have an impact. The FEC updated
Table 1 outlines these regulation models.
its 1990 standards in 2002, creating the Voting Sys-
tem Standards (VSS). Most voting systems currently
The United States: A prescriptive approach
in use in the U.S. are certified to this standard. The
U.S. federal voting system standards date back to initial VVSG was adopted by the EAC in 2005
1990 under the Federal Election Commission (VVSG 1.0). Few systems (about a dozen) are pres-
(FEC). The history of modern standards, however, ently certified to the 2005 standard. An effort to
Table 1. Regulation Approaches to E-Voting

Prescriptive approach Top-down, shared approach Broad-based approach

Country/ organization United States Switzerland Council of Europe

Mandatory or voluntary Voluntary standards Mandatory, based on different Voluntary standards
levels of use
Primary focus Certification, voting Certification, voting standards Recommendations of standards
standards (hardware) (software)
Characterized by Very specific and Specific technical standards Taking a broad approach to be
prescriptive to ensure the reliability and applicable and relevant to
safety of online voting member and non-member nations
 168 ESSEX AND GOODMAN

completely rewrite the VVSG to address future international organizations have also developed reg-
voting systems began in 2007. The draft was recom- ulation frameworks for e-voting use.1 The most
mended by a Technical Guidelines Development widely followed international document is produced
Committee (TGDC) and incorporated feedback from by the Council of Europe, an international human
election officials, academics, and vendors, but was rights organization with 47 member countries,
not adopted over several points of contention. Ele- which establishes standards for e-voting along
ments from the 2007 draft were eventually inte- legal, operational, and technical dimensions. Since
grated with the VVSG 1.0, producing the VVSG 2002, the Council of Europe has compiled many re-
1.1 in 2009. However, the EAC lost quorum of its sources addressing e-voting, including a Handbook
commissioners from 2010 until 2014, which delayed on the steps to follow during adoption, guidelines
adoption until 2015. Whereas the VVSG 1.1 is on certification and transparency (Driza Maurer
two volumes totaling approximately 300 pages, 2014), and e-voting standards Rec(2004)11 and
the EAC has chosen to redefine its role, and the Rec(2017)5. These standards are the only intergov-
current draft, VVSG 2.0, is a compact five-page ernmental documents that focus on regulation and
set of principles and guidelines (U.S. EAC 2017). standardization of voting technologies. All other
Ultimately, the U.S. approach can be character- international documents can be characterized as
Downloaded by 223.187.185.185 from www.liebertpub.com at 08/04/21. For personal use only.

ized by its prescriptive nature, focusing primarily guidelines, efforts to formalize procedures, or pro-
on testing and certification of election technology vide advice regarding good practices. Even com-
such as vote capture and tabulation technology, pared to domestic documents, the recommendations
how ballots are designed and presented, and how are unique because of the broad framework they
ballot data is represented and exported. This high de- present. Although voluntary, the Council’s e-voting
gree of specificity is evident in detailed testing re- recommendations have been widely followed by
quirements such as which types of pens and inks member and non-member countries. Norway, for
are permitted, or how precisely to present election example, relied closely on the Rec(2004) for its
results to news agencies (U.S. NIST 2016). This 2011 and 2013 deployments of online voting (Barrat
specificity can be very expensive for vendors to im- et al. 2012). Finland’s use of voting machines is an-
plement and has resulted in a consolidation of the other case that was monitored and evaluated using
market to a small number of companies. It can the Rec(2004) (Whitmore 2008).
also be argued this approach has an overly narrow The creation of the Recommendations was the re-
scope that does not account for all aspects of the sult of a multidisciplinary ad hoc group of specialists
broader threat environment. The voluntary nature formed by the Council of Europe in 2002 in response
of standards has also created tension between state to a growing international buzz about electronic vot-
and local election officials over national security ing. The goal of the group was to discuss standards for
considerations (U.S. DHS 2018). ‘‘e-enabled voting’’ as part of the Council’s project,
Finally, it is worth noting that despite the pre- ‘‘Making Democratic Institutions Work’’ (Stein and
scriptiveness and attention to detail in the U.S. Wenda 2014). The group collaboratively developed
model, there are types of ‘‘electronic’’ voting not standards along legal, operational, and technical cate-
regulated by, or covered under, current standards. gories, which were adopted in September 2004.
Many states, for example, permit variations of vot- Nearly ten years later, in response to rapidly changing
ing by e-mail whereby a voter receives his/her ballot technology and the fact that many lessons learned had
via e-mail and then, depending on the state, can emerged from e-voting experimentation, the Council
return the marked ballot by mail, fax or e-mail. Sev- revisited the Rec(2004) (Driza Maurer 2014).
eral states also offer an online portal to distribute and
receive votes (Thompson 2018). It is important to
point out this loophole to illustrate limitations of cur- 1
The Institute for Democracy and Electoral Assistance (IDEA),
rent approaches. the Organization for Security and Co-operation in Europe, the
Organization of American States, and others, for example,
have outlined guidelines regarding election observation of tech-
Council of Europe: A broad-based approach nologies. Other organizations, such as the International Founda-
tion of Electoral Systems, IDEA, and the European Union, have
While some countries have standards or certifi- developed principles to inform the use of e-voting and election
cations that are crafted and enforced domestically, technologies (Driza Maurer 2014).
 DEVELOPING E-VOTING REGULATIONS IN CANADA 169

The updated version broadens the definition of for executing elections and referendums with the
e-voting—the same definition embraced by this cooperation of municipalities. The cantons also
article and outlined above. It consists of three docu- run federal referendums and elections (Serdült
ments: the Recommendation, which outlines central et al. 2015; Serdült 2016; Goodman 2017). The Fed-
aspects of e-voting; an Explanatory Memorandum; eral Chancellery is a federal agency that oversees
and guidelines to inform the implementation of pro- and coordinates matters related to elections and en-
visions in the Recommendation. By parsing out sures cantons are complying with national norms
the Recommendation into these three documents, (Serdült et al. 2015). This differs from the Canadian
the Council aimed to improve upon weaknesses of case where local elections are more decentralized
the 2004 version, which was criticized for overlap, and municipalities implement election technologies
redundancy, and having specific language in some without oversight. Furthermore, Canada’s federal
areas while being overly vague in others (Barrat election agency, Elections Canada, is only respon-
et al. 2012). sible for running national votes. There is no federal
The 2017 version, for example, outlines 143 pro- authority monitoring elections at sub-national and
visions, 49 standards in the Rec(2017)5 and 94 in local levels. In Switzerland, the Federal Chancel-
the accompanying Guidelines, compared with 142 lery has been involved in electronic voting pilots
Downloaded by 223.187.185.185 from www.liebertpub.com at 08/04/21. For personal use only.

in the 2004 version (112 in the Rec(2004)11 and from the very beginning, serving as ‘‘central coor-
30 additional items in the guidelines). While the dinator and mentor of the whole project’’ and
number of standards did not change much, the con- funding 80 percent of initial trials (Serdült et al.
tent has been substantially altered, including the 2015, 128).
addition of several new standards. Key changes in- Although e-voting has been used in Switzerland
clude the removal of language that elections sup- since 2003 when online voting was first piloted,
ported by e-voting should be ‘‘as reliable and 2014 marked a significant change in regulation of
secure as’’ paper-based options, the need to specif- the voting technology (Puiggalı́ and Rodrı́guez-
ically assess risks related to e-voting and compile Pérez 2018). The regulation, initiated by the Federal
strategies to mitigate or counter these risks, and Chancellery, focuses on specific technical require-
the need to update e-voting guidelines on a regular ments which, if not met, limit the proportion of
basis via biannual review meetings (Maurer 2017). the electorate that can vote online. This approach
Despite these changes and the international prom- is not new to Switzerland. Initially, the Federal
inence of the document, the standards remain broad Chancellery limited the extension of online voting
in scope. This approach is understandable since the to 10 percent of electors, and this was gradually in-
document needs to apply to multiple types of tech- creased to 30 percent. This limitation was estab-
nology and be relevant for legal frameworks in lished to mitigate risk under the assumption that
many countries. While its breadth is well suited for the more electors casting online ballots, the greater
international regulation, there is too much latitude the incentive for interference (Goodman 2017).
for applicability in Canada. We discuss this further The new regulation requires cantons to apply to
below. use online voting before every vote. They must
meet specific technical requirements to proceed,
which mostly focus on achieving verifiability targets
Switzerland: A top-down, shared approach
(for a more detailed account of this process see Puig-
Finally, Switzerland has advanced its approach to galı́ and Rodrı́guez-Pérez 2018). At the lowest level
e-voting regulation. The Swiss model focuses on (30 percent of the cantonal electorate and 10 percent
providing the mechanisms for auditing elections, of the federal electorate), online voting can be of-
recognizing that technology changes over time. It fered without certification, although authorities
differs from the U.S. approach, which is more cen- point out that while this is legally possible, it
tered on hardware. The Swiss model is a useful would no longer be accepted politically (Spycher
comparison for Canada since they also have a fed- 2018). If the online voting system is certified as hav-
eral system of government with different levels ing individual verifiability, meaning an auditor can
holding distinct responsibilities for running elec- verify a ballot has been cast-as-intended, the voting
tions and because of the growth in online voting method can be offered to 50 percent of the cantonal
use. In the Swiss case the cantons are responsible electorate and 30 percent of the federal electorate. If
 170 ESSEX AND GOODMAN

complete verifiability is certified; that is, ensuring electronic voting methods to prevent further disjunc-
a ballot has been recorded-as-cast and counted-as- tion in voting approaches and policies.
recorded, online voting may be offered to 100 per- The definition of electronic voting outlined by the
cent of the electorate. An Ordinance on Electronic Council of Europe’s Rec(2017)5, which includes
Voting issued by the Federal Chancellery outlines vote-counting equipment and online voting, provides
the specific requirements that must be met for a more comprehensive strategy to regulate the array
each of these limits, which capture functionality, of voting technologies used in Canada, albeit the reg-
security, and usability requirements (Puiggalı́ and ulations are overly broad. While the breadth of this ap-
Rodrı́guez-Pérez 2018). proach makes it widely applicable, that same leeway
Whereas previously system security was left up to makes it more challenging to hold technology vendors
the cantons, it is now federally controlled and these accountable. It also lacks the specific guidance mu-
new standards are adapted from several sources, in- nicipal officials articulated is needed for vetting ven-
cluding the Council of Europe’s Recommendation. dors or managing issues associated with unsupervised
This shift focuses on certifying the technology and voting. Also, this approach would likely require inter-
ensuring the means for auditing elections through vention from the federal government as an overarch-
verifiability. The Swiss model differs from other in- ing authority. While national standards would be the
Downloaded by 223.187.185.185 from www.liebertpub.com at 08/04/21. For personal use only.

ternational approaches given that its system guide- best way forward, they are not as practical given the
lines are mandatory. Some interviewees made system of divided jurisdiction. They would also take
reference to the U.S. approach being ‘‘too hot’’ and more time and require dedicated funding.
the Council of Europe model being ‘‘too cold,’’ Finally, while the U.S. approach is too rigid and
while the Swiss regulations are ‘‘just right.’’ This ap- the Council of Europe model has too much latitude
proach builds in latitude allowing cantons time to for application in Canada, standards in both scenarios
test and develop their voting technology but still are voluntary. Interviews carried out for this article
use online voting even if it does not yet meet the indicate support among municipal officials for man-
gold standard of complete verifiability. For Canada, datory technical standards, while other aspects such
this approach is most relevant given the similarities as operational considerations could be more discre-
of divided jurisdiction and the mandatory technical tional. Strengthening the legal framework for elec-
standards. tronic voting in Canada is also best achieved by
amending the statutes in each province, complicating
the issue of a federal body taking over all aspects.
The Swiss approach, which outlines mandatory
DISCUSSION: BRIDGING MULTILEVEL
technical requirements and is applied in a federal
GOVERNANCE AND ENHANCING
system where responsibility for elections is divided,
ELECTORAL INTEGRITY
also presents some key takeaways for Canada. The
greater focus on certifying system software (as op-
Applicability of approaches to election
posed to the emphasis on hardware in the U.S.) is
voting regulation
more suited for the range of technology used in
Based on the above discussion, which is the right Canada. Furthermore, the standards outlined in
approach for Canada? The U.S. model with its spec- Switzerland provide mechanisms for auditing elec-
ificity regarding the types of pens and inks that can tions, which would strengthen technical capacity
be used is overly prescriptive. There is also no polit- in communities. The three levels of certification
ical will federally in Canada to create something akin also provide goals to work toward while allowing
to the EAC, nor would that necessarily be the best local governments to continue using the technology.
way forward as it would require the development Despite these benefits, the focus is mainly on tech-
of a new governance structure for local elections. nical aspects and not operational considerations,
Additionally, there are electronic voting methods which is an area of guidance sorely needed in
(e.g., e-mail voting) not covered under existing U.S. Canada. Additionally, the active role the federal gov-
frameworks which would pose problems for election ernment has played in Switzerland is not guaranteed
governance in Canada given the freedom of munici- in Canada, and the provinces need to be given a more
palities to develop their own models. A solution for significant role. Finally, the inclusion of a more pre-
Canada needs to target all identified loopholes and scriptive legal framework is needed in Canada.
 DEVELOPING E-VOTING REGULATIONS IN CANADA 171

A way forward for Canada do not even bring technology vendors in for an inter-
view before awarding the contract. To promote elec-
We argue the best way to move forward with reg- toral integrity and enhance technical knowledge in
ulation of electronic voting in Canada is to begin communities we propose a two-pronged technical ap-
with principles. While election principles are not proach that includes the creation of (1) mandatory
enshrined in legislation, they have become a con- minimum standards for cybersecurity, and (2) vol-
vention based on court cases. Commencing with untary procurement guidelines. The former would
principles as a base is a useful starting point. ensure minimum technical regulation while the lat-
There are then three dimensions that require consid- ter would promote technical capacity across munic-
eration: legal, technical, and operational. ipalities, but especially in places with no IT team or
With respect to a legal framework, these statutes technical resources. The question of whether these
are written by the provinces with municipalities technical elements are led by federal or provincial
that have the ability to pass by-laws through local authorities is one we cannot answer in this article
councils to further customize the governance of their given the complexities of multilevel governance.
elections, so having federal direction over this would Municipal support regarding who should take lead-
not make sense. We assert the best way forward on ership is divided and requires further discussion.
Downloaded by 223.187.185.185 from www.liebertpub.com at 08/04/21. For personal use only.

the legal front is for the provinces to change the Finally, we argue for the creation of voluntary
Municipal Elections Acts to provide additional proto- operational guidelines directed by the provinces in
cols around the use of electronic voting. This process consultation with municipalities and municipal as-
should be undertaken with the robust consultation of sociations. These guidelines would be grounded in
municipalities. Some preliminary suggestions regard- the many good practices that have been tried and
ing which aspects to tackle are outlined below. Updat- tested during nearly two decades of experimenta-
ing of legal frameworks should also be shared across tion. Given the regional character of the Canadian
provinces to benefit those that are slower to update state, what may work well operationally in one
their statutes, or in cases where current Acts do not area (e.g., Ontario) may not be as directly applicable
provide for electronic voting. in another (e.g., Nova Scotia). Furthermore, taking
Technically, we argue there is a greater need to de- the unique contextual circumstances of communi-
velop and enforce mandatory standards to hold ven- ties into consideration is an important rationale for
dors accountable and boost technical capacity in why operational items should be voluntary as it
municipalities, which in some places is strong while would allow communities to make operational
in others it is minimal at best. Some communities, choices that work best for their context.
for example, conduct risk assessments, contract out- The approach we suggest to regulate electronic
side experts, and review security reports, whereas voting in Canada is directed from higher levels of
other local governments, according to interviewees, government but administered locally, see Table 2.

Table 2. An E-Voting Regulation Approach for Canada

Prescriptive Top-down, Top-down, locally

approach shared approach administered Broad-based approach

Country/ United States Switzerland Canada Council of Europe

organization

Mandatory or Voluntary standards Mandatory based on Mandatory technical Voluntary standards

voluntary different levels of use standards, voluntary
procurement and operational
guidelines, a renewed legal
framework
Primary focus Certification, voting Certification, voting Standards and a legal Recommendations
standards (hardware) standards (software) framework of standards
Characterized by Very specific and Specific technical Broad focus on electronic A broad approach
prescriptive standards to ensure the voting with specific technical for wide applicability
reliability and safety standards and latitude on to member and non-
of online voting operational items member nations
 172 ESSEX AND GOODMAN

Unlike other country models, this approach includes Voting Systems Guidelines offer a more specific
a mixture of voluntary and mandatory elements to wording: ‘‘The voting process preserves the privacy
maximize the promotion of electoral integrity of the voter’s interaction with the ballot, modes of
while being sensitive to the unique contextual con- voting, and vote selections’’ (U.S. NIST 2017). Per-
siderations in the approximately 3,500 municipali- haps the principle of ballot secrecy is most suc-
ties across Canada and providing some discretion. cinctly expressed in the Canada Elections Act:
It also includes a legal element, which we argue ‘‘The vote is secret’’ (Canada Elections Act, 2001,
should be explicitly incorporated into existing stat- sec. 163). It is unclear, however, whether applying
ues by the provinces upon consultation with local secrecy to the vote itself could preclude the possibil-
governments. We explore the development of prin- ity of a post-election audit, and, nevertheless, the se-
ciples, legal, technical, and operational components crecy of the vote itself is not specific property we
in further detail below. require. Rather we are interested in the secrecy of
the association between a voter’s identity and vote.
Second, to the principle of fairness and for rea-
GUIDING PRINCIPLES FOR CANADIAN sons of clarity the ruling in Cusimano v. Toronto
ELECTIONS (2011) suggested replacing the term ‘‘non-biased’’
Downloaded by 223.187.185.185 from www.liebertpub.com at 08/04/21. For personal use only.

with the words ‘‘and must not favor one candidate

There was wide agreement among interviewees over another.’’ This wording, however, becomes re-
that any effort toward a standards-based framework dundant with the other fairness principle that ‘‘vot-
should begin with an articulation of key principles. ers and candidates shall be treated fairly and
Such principles should be broad enough to be inde- consistently.’’
pendent of any particular election technology or ju- Third, to the principle of certainty, the ruling of
risdiction, and preferably have some pre-existing Cusimano v. Toronto (2011) took issue with ‘‘cer-
foundation in case law. tainty of results’’ arguing:
Ontario’s municipal use of both conventional and
advanced election technology provides us with a If (it) means that the numerical results of an elec-
starting point for our discussion of principles. We tion must agree with the actual votes cast, it sets
a test of perfection that can never be met. But if
observe, however, the Ontario Municipal Elections
it means that there must be certainty that the winner
Act does not outline a list of principles. A set of
of an election actually received the most votes, then
principles that has been informally applied by it is acceptable.
many Ontario municipalities was originally formu-
lated by the clerk of Vaughn in Di Biase v. Vaughan Instead, the ruling favored the following phrasing
(2007): (Montgomery v. Balkissoon 1998): ‘‘the proper ma-
jority vote decides the election. That principle is
The secrecy and confidentiality of the voting pro- achieved by ensuring, so far as is reasonably possi-
cess is paramount; The election shall be fair and
ble, that valid votes be counted and invalid votes be
non-biased; The election shall be accessible to
the voters; The integrity of the process shall be
rejected.’’
maintained throughout the election; There is to While establishing the priority of an election out-
be certainty that the results of the election reflect come ahead of a numeric result, these particular
the votes cast; and Voters and candidates shall be phrasings are not broadly applicable since the no-
treated fairly and consistently. tion of a ‘‘proper majority’’ or ‘‘the most votes
cast’’ implies a plurality (e.g., first-past-the-post)
These particular principles have evolved in case system. The 2018 municipal election of London,
law over time, and we make additional suggestions Ontario, for example, held the first ranked-choice
for improvement here. ballot in which the winner was determined by a sin-
First, the principle expressing the ‘‘paramount’’ gle transferable vote, not a simple majority. This al-
nature of ballot secrecy is essential, but given the in- ternate electoral system remains an option for
creasing multi-national character of election infra- municipalities in the province.
structure (see e.g., Culnane et al. 2017), a more The principle of certainty as worded neither
unqualified phrasing is needed. The U.S. Voluntary specifies whom the certainty inhabits (e.g., election
 DEVELOPING E-VOTING REGULATIONS IN CANADA 173

official, candidate, electorate, etc.), nor how or 6. Accuracy. A cast vote must accurately reflect
whether the sense of certainty is related to the elec- the voter’s intended preference;
tion’s outcome. While the notion of election offi- 7. Correctness. The election outcome must re-
cials ‘‘ensuring’’ the correct outcome should be flect the correct combination of valid cast
considered necessary, it is not sufficient. Ultimately votes;
the high-level notion of certainty should be arrived 8. Confidence. Voters and candidates must be
at through a combination of principles, such as ac- confident in the correctness of the election
curacy, correctness, transparency, and confidence. outcome; and,
Fourth, as pointed out in Harper v. Canada 9. Transparency. The election is to be conducted
(2004), ‘‘Maintaining confidence in the electoral in a manner that is transparent and account-
process is essential to preserve the integrity of the able to voters and candidates.
electoral system which is the cornerstone of Cana-
dian democracy.’’ It goes on to cite the following Observe that neither this list of principles nor
(R. v. Oakes 1986): those of the original (Di Biase v. Vaughan 2007) in-
clude factors such as cost or convenience. Principles
If Canadians lack confidence in the electoral system, aside, the cost to meet security requirements will be
Downloaded by 223.187.185.185 from www.liebertpub.com at 08/04/21. For personal use only.

they will be discouraged from participating in a a significant issue, especially if it divides provinces
meaningful way in the electoral process. More im- and territories along a line of who can and cannot
portantly, they will lack faith in their elected repre- afford to implement a minimum cybersecurity stan-
sentatives. Confidence in the electoral process is, dard for online voting. Switzerland’s approach of
therefore, a pressing and substantial objective.
graduated levels of verifiability may offer Canada
a starting point to explore this issue.
Furthermore, any proposed changes to voting sys-
tems must be implemented in a manner ‘‘safeguard-
ing public trust in the election process, by ensuring
LEGAL CONSIDERATIONS
reliable and verifiable results obtained through an
effective and objective process that is secure and
Presently, there is not much of a legal framework
preserves vote secrecy for individual Canadians’’
guiding electronic voting use at the local level in
(House of Commons 2016). Applying the principles
Canada. As noted, legislation includes a provision
of confidence, verifiability, and public trust, we can
to allow for use, but beyond that it is not prescriptive
conclude that a burden of proof exists, one that rests
regarding procedures. Meanwhile, procedures are
with the election administrators. The Organization
quite detailed when it comes to paper ballots and
for Security and Co-operation in Europe (OSCE),
the structure and conduct of polling locations.
of which Canada is a participant, offers a set of elec-
In Ontario, the Municipal Elections Act, for ex-
tion principles that explicitly include requirements
ample, clearly envisions what voters and candidates
for accountability and public confidence (OSCE
can and cannot do in a polling place such as respect
2013).
for privacy or the level of assistance someone can
Drawing on these various sources we propose the
provide to a voter, but the same is not true for elec-
following principles:
tion technologies or unsupervised voting. The Act
1. Secrecy. The association between a voter’s requires that clerks establish procedures for ‘‘any
identity and vote is secret; voting and vote-counting equipment’’ and ‘‘any al-
2. Equality. A voter shall not be able to cast more ternative voting method’’ (Ontario Municipal Elec-
votes than another, nor be prevented from tions Act, Section 42(3)). The result is that clerks
casting a ballot; have established procedures in their communities
3. Accessibility. The election shall be accessible where needed, creating a further patchwork of pol-
to voters. Each voter must have the means and icy across the province.
opportunity to participate; Drawing upon the example of regulating voter
4. Fairness. Voters and candidates shall be trea- and candidate behavior, some interviewees men-
ted fairly and consistently; tioned one of the biggest complaints they encounter
5. Dependability. The election shall deliver an is candidates going door-to-door with iPads encour-
outcome in a reliable and timely manner; aging electors to vote for them. Another example
 174 ESSEX AND GOODMAN

mentioned by a clerk was a candidate in their com- TECHNICAL CONSIDERATIONS

munity who set up a makeshift office and put an ad
in the local newspaper encouraging electors to visit As noted, in Canada no minimum cybersecurity
and vote there. Clerks have responded by reminding standards or procurement guidance exist for elec-
electors of their right to privacy, educating about the tion technology, including electronic voting. This
legalities of undue influence, and closing down any has resulted in many jurisdictions operating largely
makeshift voting centers, but there is nothing in the in isolation, which is problematic given their signif-
Act they can draw upon to regulate this behavior. icant variance in resources. Scholarly literature is
These examples illustrate a clear need for statutes replete with of examples of new voting technologies
to be more prescriptive regarding protocols for elec- leading to unforeseen and unintended consequences
tronic voting. impacting election principles, such as ballot secrecy
Another element of elections that has been af- (Springall et al. 2014, Culnane et al. 2017, Cardillo
fected because of the upsurge in electronic voting and Essex 2018); accuracy (Mercuri 2001); correct-
is scrutineers, whose historical role at polling places ness (Halderman and Teague 2015, Chang-Fong
has been essential for promoting electoral integrity. and Essex 2016); and dependability (Porter 2018).
While supervised voting with paper ballots provides Online voting in particular presents a unique and
Downloaded by 223.187.185.185 from www.liebertpub.com at 08/04/21. For personal use only.

many opportunities for scrutineers, use of electronic potentially extreme challenge to these principles.
voting has limited or, in some cases, done away with To that end, the OSCE suggests an ‘‘incremental ap-
this role. Some municipalities that have opted for proach to introduction, together with thorough test-
fully electronic elections (remote online and tele- ing, verifiability and full transparency’’ as a means
phone voting) have resolved that scrutineers no lon- of developing public confidence in new voting tech-
ger have a role to serve in modern elections, whereas nologies (OSCE 2013).
others, albeit a minority, make a point of continuing While interviewees disagreed substantially about
to include scrutineers as a convention of municipal the degree to which voting technology should be al-
elections. They do so by including scrutineers in lowable, there was widespread agreement that any
voting demonstrations, showing them the vote election technology (especially online voting sys-
count before the election, allowing them to test tems) should be subject to minimal technical re-
the technology to illustrate a vote cannot be cast quirements pertaining to cybersecurity, and that
twice, and including them as part of the tabulation they should be mandatory.
process. While the inclusion of scrutineers could In terms of the cybersecurity standards them-
be optional for municipalities as they innovate elec- selves, one interviewee with standards development
tions, if this aspect of the voting process is viewed experience warned that online voting was ‘‘a whole
as important for Canadian democracy the Act other ballgame,’’ and that they would not even begin
should address it. to know how to approach the question. Another,
While the current statute serves municipalities with similar experience, warned of the importance
well for supervised, paper ballot voting, it does of ‘‘getting it right,’’ since machines and technology
not provide sufficient guidance regarding electronic will be built to those standards and will take years to
voting, especially in an unsupervised environment. go through even a single iteration of refinement. In
Without being overly prescriptive and limiting inno- the U.S., for example, systems are built with an an-
vation, which was a concern articulated by many we ticipated lifetime of 10–15 years and take 3–4 years
spoke with, legislation could be updated to more to be certified. They recommended, as we argue,
formally include some basic protocols for managing that a Canadian approach be much more broad in
issues such as the decline of scrutineers and candi- scope than the current U.S. approach.
date and voter behavior. This would provide munic- A number of interviewees, however, pointed to
ipalities with enforcement tools and empower them potential starting points, including the Security
to more easily promote electoral integrity. It would Considerations for Remote Electronic UOCAVA
also help to level the playing field between munic- Voting (Hastings et al. 2011); software and security
ipalities that have established more advanced poli- requirements of the VVSG 1.1 volume 1 (U.S. EAC
cies compared to those that have none. This is 2015); the U.S. Vote Foundation’s report on cryp-
best handled on a provincial basis since provinces tographic end-to-end verifiable internet voting
are responsible for writing and updating the statute. (Dzieduszycka-Suinat et al. 2015); and Toronto’s
 DEVELOPING E-VOTING REGULATIONS IN CANADA 175

2013 Internet voting request for proposal (Toronto OPERATIONAL CONSIDERATIONS

2013). It was also suggested Canada consider estab-
lishing a body similar to the Technical Guidelines There was widespread agreement among inter-
Development Committee (TGDC), which brings to- viewees that while operational guidance is an
gether election administrators, standards develop- asset that enhances governance capacity, the diverse
ment experts, and technical experts, and include and unique characteristics of communities suggest
provisions for public-comment and testing. this component should be voluntary. By providing
Some practical suggestions for mandatory guide- optional operational guidelines, communities have
lines included elements such as encryption of the a resource they can apply and draw upon where
voting browser, encrypted back-ups and data recov- applicable.
ery, specification about the location of servers, third There are a number of areas where communities
party tests (e.g., penetration), independent verifica- could benefit from operational guidance. In light of
tion of a vendor’s source code, verifiability (there is the 2018 Ontario municipal election issues, one
disagreement regarding the level of verifiability), suggested aspect is protocols for how to manage
and the need for risk limiting audits, among other an election in situations where a technical issue oc-
items. Striking a multidisciplinary committee with curs that is not related to the security of the voting
Downloaded by 223.187.185.185 from www.liebertpub.com at 08/04/21. For personal use only.

local government administrators, scholars, technical technology. Had clerks had access to such sugges-
experts, and election officials is one way to draft an tions, the extension of voting may have gone more
initial version of this document. smoothly and not raised as many questions about
A second necessary reform regarding technical electoral integrity. Other operational items include
aspects is the need for voluntary procurement guide- suggestions regarding the length of time electronic
lines. Such a document would outline good practices voting is available. In 2018 one Ontario municipal-
for vetting technology vendors with the goal of en- ity offered online voting for a single day while oth-
hancing the technical capacity and knowledge of ers had the voting method available for a 22-day
diverse local governments. Whereas the City of Tor- period. The average length of time across all
onto, for example, the most populous municipality municipalities was 10 days. Finally, some commu-
in Ontario, has a dedicated department and full- nities adopt what is commonly referred to as a
time staff to handle election administration and pro- one-step approach, which allows an elector to
curement, one of the smallest municipalities, the cast a ballot after receiving their voter card with
Township of Brethour, is approximately 25,000 a unique PIN, whereas other places prefer to im-
times smaller, has no website, technical support, or plement a two-step approach which requires regis-
team. Although some cities have used a version of tration to vote online before gaining the ability to
Toronto’s requirements (Toronto 2013), each munic- access a ballot (Goodman and Pyman 2016).
ipality ultimately makes its own procurement deci- Guidance regarding which approach is useful in
sions about election technology, which has led to particular contexts was articulated as being help-
hundreds of disparate requests for proposal. ful. Other items could include advice regarding
Before standards can be developed, however, we audits, testing, reporting requirements, and infor-
must establish who has the authority to develop and mation sharing.
enforce them. While a federal standard would be The intent of operational guidelines would be to
ideal in theory, any form of mandatory standard provide good practices and boost local knowledge
would realistically only be achievable at the provin- regarding the implementation of electronic voting.
cial level, though there would be the possibility of This document could be built from a wealth of
coordinating these efforts. Two organizations in- good practices that have developed from nearly
volving federal, provincial, and territorial chief two decades of local innovation. Crafting opera-
electoral officers that could be leveraged include tional guidelines could be led by the province or
the Association of Canadian Electoral Partners municipalities, with funding and leadership from a
(ACEP) and the Canadian Council of Election Offi- higher level of government. In addition, operational
cials (CCEO), which have been developing a Secre- guidelines should be regularly updated given the
tariat for Electoral Collaboration (SEC). This group rapid pace with which technology is changing. It
could be a candidate for coordinating an eventual is suggested to have a working group that revisits
standards development effort. these items every two years.
 176 ESSEX AND GOODMAN

CONCLUSION AND RECOMMENDATIONS - Any regulation must be developed and in close

cooperation with municipalities and updated
The issue of regulation of election technologies regularly with their input to keep pace with
is a new area of governance that will likely grow in technological change and ensure it is providing
importance as elections around the world digitize. the best instruction possible;
The lack of attention paid to regulation in Canada - The federal government should commit to
specifically has been raised in public forums investing in academic research collaborations
(Essex and Goodman 2017). While some munici- with governmental organizations at all levels
pal and provincial officials have also spoken out toward the development of new secure election
about the importance of putting a framework in technologies.
place, action has yet to be taken by provincial or
federal authorities. As outlined, part of the chal- Electronic voting has been readily used in munic-
lenge in taking action on regulation is the complex- ipal elections in Canada for over 15 years with no
ities of multilevel governance in Canada and oversight. Continuing to take no action on regula-
ensuing questions about whose responsibility it is tion puts local bureaucrats in a difficult position
to develop and enact standards. This article has when situations do arise, even if they are unrelated
Downloaded by 223.187.185.185 from www.liebertpub.com at 08/04/21. For personal use only.

presented a way forward and, in doing so, outlined to the security of the voting technology. It also pres-
a new approach to the regulation of election tech- ents a tangible threat to electoral integrity, which
nologies that includes both mandatory and volun- could compromise electoral modernization, reduce
tary elements and builds upon principles, while public trust in elections, and stifle government legit-
addressing legislative, technical, and operational imacy. While a lack of regulation has allowed for in-
aspects of elections. By building off of existing novation and the testing of a variety of models,
models of regulation and incorporating feedback uptake of electronic voting has exceeded a critical
from interviewees, we make the following recom- point, and not taking action on regulation poses
mendations regarding the regulation of electronic more risks than benefits. By developing regulations
voting in Canada: for electronic voting, Canada is not only taking
steps to protect its own elections, but also creating
- Create mandatory minimum cybersecurity a new model other jurisdictions can learn from as
requirements at the sub-national level, and elections around the world modernize and govern-
coordinate efforts where possible between ments grapple with how to govern these new voting
federal, provincial, and territorial chief elec- technologies while maintaining the integrity of po-
toral officers; litical institutions.
- As a second component to technical standards, Future research could focus on specific technical
develop a guidance document and standardized standards and what these should look like to ensure
requests for proposal for local election officials they both enhance electoral integrity and provide
(where applicable) on the procurement of elec- flexibility. Scholars could also work on developing
tronic voting (especially electronic tabulators secure voting technologies. An important compo-
and online voting); nent of this effort is government willingness to
- Update legislation, the Municipal Elections fund research into new, secure voting technologies.
Acts, to address the use of electronic voting In Canada, in particular, this would require the na-
and unsupervised voting; tional scientific funding body, the National Science
- All of the above items should be crafted by and Engineering Research Council, to reexamine its
drawing on the election principles presented focus on funding industry-based research and con-
in this article, and take an approach that strikes sider other collaborations that involve government
a balance between the flexibility of the Council partners working in the public interest. In addition,
of Europe’s guidelines, and the specificity of future research could further tease out aspects of
the U.S. Voluntary Voting System Guidelines. electronic voting regulation and consider whether
The model should be prescriptive enough to there is an ‘‘ideal’’ model that could be applied
give municipal officials the tools and guidance across countries. Finally, scholars could evaluate
they require, while not being overly specific to whether different approaches to regulation have
stifle local innovation; an impact on public and administrative attitudes
 DEVELOPING E-VOTING REGULATIONS IN CANADA 177

toward elections and trust in government. We would Elections Ontario. (2018). Modernizing Ontario’s Electoral
also encourage researchers to bridge disciplinary Process Report on Ontario’s 42nd General Election June
7, 2018. Retrieved from <https://www.elections.on.ca/content/
silos (e.g., computer science and the social sciences)
dam/NGW/sitecontent/2019/Reports/2018%20General%20
and come together to tackle the challenge of regu- Election%20-%20Post-Event%20Report.pdf>.
lating election technologies in the digital age. Essex A., and N. Goodman. (2017). ‘‘Internet Voting Will
Happen. Let’s Make Sure the Research into How It Hap-
pens Is Sound.’’ Ottawa Citizen. January 30. Retrieved
REFERENCES from <https://ottawacitizen.com/opinion/columnists/essex-
and-goodman-internet-voting-will-happen-lets-make-sure-
Barrat, J., B. Goldsmith., and J. Turner. (2012). International the-research-into-how-it-happens-is-sound>.
Experience with E-Voting. Washington, D.C.: International Franklin, M.N. (2004). Voter Turnout and the Dynamics of
Foundation for Electoral Systems. Electoral Competition in Established Democracies Since
Canada Elections Act. (2009). S.C. 2000, c. 9. Retrieved from 1945. New York, NY: Cambridge University Press.
<https://laws-lois.justice.gc.ca/eng/acts/e-2.01/FullText.html>. Gabel, C., N. Goodman, K. Bird, and B. Budd. (2016). ‘‘What
Cardillo, A., and A. Essex. (2018). ‘‘The Threat of SSL/TLS Does Internet Voting Mean for First Nations? A Case Study
Stripping to Online Voting.’’ In In Electronic Voting: Inter- of Whitefish River First Nation.’’ International Indigenous
national Joint Conference on Electronic Voting, E-Vote-ID Policy Journal, 7(3): Article 3.
2018. 35–50. Cham: Springer. Gabel, C., and N. Goodman. (2020). Indigenous Experiences
Downloaded by 223.187.185.185 from www.liebertpub.com at 08/04/21. For personal use only.

CBC News. (2019). ‘‘‘Who Owns the Data?’: Questions with Online Voting. Ottawa: Crown-Indigenous Relations
Emerge at Sidewalk Labs Consultations About 1,500- and Northern Affairs Canada.
Page Proposal.’’ CBC, July 18. Retrieved from <https://www Germann, M., and U. Serdült. (2014). ‘‘Internet Voting for
.cbc.ca/news/canada/toronto/who-owns-the-data-questions- Expatriates: The Swiss Case.’’ JeDEM—eJournal of eDe-
emerge-at-sidewalk-labs-consultations-about-1-500-page- mocracy and Open Government 6(2): 197–215.
proposal-1.5215990>. Germann, M., and U. Serdült. (2017). ‘‘Internet Voting and
Chang-Fong, N., and A. Essex. (2016). ‘‘The Cloudier Side of Turnout: Evidence from Switzerland.’’ Electoral Studies
Cryptographic End-to-End Verifiable Voting: A Security 47: 1–12.
Analysis of Helios.’’ In Proceedings of the 32nd Annual Goodman, N., J.H. Pammett, and J. DeBardeleben. (2010). A
Conference on Computer Security Applications. 324–335. Comparative Assessment of Electronic Voting. Gatineau,
New York, NY: ACM. QC: Elections Canada.
Council of Europe. (2017). ‘‘Council of Europe Adopts New Goodman, N., and J.H. Pammett. (2014). ‘‘The Patchwork of
Recommendation on Standards for E-Voting.’’ Electoral Internet Voting in Canada.’’ In 2014 6th International Con-
Assistance Newsroom. June 14. Retrieved from <https:// ference on Electronic Voting: Verifying the Vote (EVOTE).
www.coe.int/en/web/electoral-assistance/-/council-of-europe- 1–6. IEEE.
adopts-new-recommendation-on-standards-for-e-voting>. Goodman, N. and H. Pyman. (2016). Internet Voting Project
Culnane, C., M. Eldridge, A. Essex, and V. Teague. (2017). ‘‘Trust Report: Results from the 2014 Ontario Municipal Elec-
Implications of DDoS Protection in Online Elections.’’ In tions. Toronto, ON: Centre for e-Democracy.
Electronic Voting: International Joint Conference on Elec- Goodman, N. (2017). Online Voting: A Path Forward for Fed-
tronic Voting, E-Vote-ID 2017. 127–145. Cham: Springer. eral Elections in Canada. Ottawa: Privy Council Office.
Cusimano v. Toronto (City). (2011). ONSC 7271. (CanLII). Goodman, N., and L. C. Stokes. (2018). ‘‘Reducing the Cost of
Di Biase v. Vaughan (City). (2007). 43 M.P.L.R. (4th) 287. Voting: An Evaluation of Internet Voting’s Effect on Turn-
(S.C.J.). out.’’ British Journal of Political Science: 1–13. DOI:
Dominion Voting Systems Inc. (2018). ‘‘Statement Regarding https://doi.org/10.1017/S0007123417000849.
Internet Voting Service Slowdown Affecting Ontario Goodman, N., M. McGregor, J. Couture, and S. Breux. (2018a).
Municipalities.’’ Dominion Voting. Retrieved from <https:// ‘‘Another Digital Divide? Evidence That Elimination of
mail.meaford.ca/forms/purchasing/6455-statement-dominion- Paper Voting Could Lead to Digital Disenfranchisement.’’
voting-ontario-iv-issue-221018-002/file.html>. Policy & Internet 10(2): 164–184.
Driza Maurer, A. (2014). ‘‘Ten Years Council of Europe Goodman, N., C. Gabel, and B. Budd. (2018b). ‘‘Online Voting
Rec(2004)11: Lessons Learned and Outlook.’’ In 6th Inter- in Indigenous Communities: Lessons from Canada.’’ In
national Conference on Electronic Voting, 111–117. Electronic Voting: International Joint Conference on Elec-
Dzieduszycka-Suinat, S., J. Murray, J. Kiniry, D. Zimmerman, tronic Voting, E-Vote-ID 2018. 67–83. Cham: Springer.
D. Wagner, P. Robinson, and S. Morina. (2015). ‘‘The Grimm, R., R. Krimmer, N. Meißner, K. Reinhard, M. Volkamer,
Future of Voting: End-to-End Verifiable Internet Voting- M. Weinand, and J. Helbach. (2007). Security Requirements
Specification and Feasibility Study.’’ U.S. Vote Foundation. for Non-Political Internet Voting. No. 06/2007, Work Report
Retrieved from <https://www.usvotefoundation.org/E2E- from the Department of Computer Science, University of
VIV>. Koblenz-Landau. Retrieved from <https://hbz.opus.hbz-
Elections, BC. (2018). Report to the Chief Electoral Officer on nrw.de/opus45-kola/frontdoor/deliver/index/docId/115/file/
Recommendations for Legislative Change. British Colum- arbeitsberichte_6_2007.pdf>.
bia. May. Retrieved from <https://elections.bc.ca/docs/rpt/ Halderman, J.A., and V. Teague. (2015). ‘‘The New South
2018-CEO-Recommendations.pdf>. Wales iVote System: Security Failures and Verification
 178 ESSEX AND GOODMAN

Flaws in a Live Online Election.’’ In International Confer- Saglie, J., and S.B. Segaard. (2016). ‘‘Internet Voting and the
ence on E-Voting and Identity. 35–53. Cham: Springer. Secret Ballot in Norway: Principles and Popular Under-
Harper v. Canada (Attorney General). (2004). 1 S.C.R. 827, standings.’’ Journal of Elections, Public Opinion and Par-
2004 SCC 33. ties, 26(2): 155–169.
Hastings, N., R. Peralta, S. Popoveniuc, and A. Regenschied. Schmidt, A., D. Heinson, L. Langer, Z. Opitz-Talidou, P.
(2011). ‘‘Security Considerations for Remote Electronic Richter, M. Volkamer, and J. Buchmann. (2009). ‘‘Develop-
UOCAVA Voting.’’ NISTIR 7770. U.S. National Institute of ing a Legal Framework for Remote Electronic Voting.’’ In
Standards and Technology. 1–64. Retrieved from <https:// International Conference on E-Voting and Identity. 92–
www.nist.gov/sites/default/files/documents/itl/vote/NISTIR- 105. September. Berlin, Heidelberg: Springer.
7700-feb2011.pdf>. Schwartz, B., and D. Grice. (2013). Establishing a Legal
Help America Vote Act (HAVA). (2002). Pub. L. No. 107–252. Framework for E-Voting in Canada. Elections Canada.
House of Commons. (2016). Strengthening Democracy in Can- Segaard, S.B., H. Baldersheim, and J. Saglie. (2013). ‘‘The Nor-
ada: Principles, Process and Public Engagement for Elec- wegian Trial with Internet Voting: Results and Challenges.’’
toral Reform. Report of the Special Committee on Electoral Revista General de Derecho Público Comparado, 13.
Reform. Retrieved from <https://www.ourcommons.ca/ Serdült, U., M. Germann, M. Harris, F. Mendez, and A. Porte-
Content/Committee/421/ERRE/Reports/RP8655791/errerp03/ nier. (2015). ‘‘Who Are the Internet Voters?’’ In Electronic
errerp03-e.pdf>. Government and Electronic Participation, eds. E. Tambou-
Langer, L., A. Schmidt, and J. Buchmann. (2008). ‘‘Secure and ris, et al., 27. Washington, D.C.: IOS Press.
Practical Online Elections via Voting Service Provider.’’ Serdült, U. (2016). The Swiss Experience with Online Voting.
Downloaded by 223.187.185.185 from www.liebertpub.com at 08/04/21. For personal use only.

Proceedings of ICEG 2008, 255–262. Online Voting Roundtable, Centre for e-Democracy.
Maurer, A.D. (2017). ‘‘Updated European Standards for Springall, D., T. Finkenauer, Z. Durumeric, J. Kitcat, H. Hursti,
E-Voting.’’ In International Joint Conference on Electronic M. MacAlpine, and J.A. Halderman. (2014). ‘‘Security Anal-
Voting. 146–162. Cham: Springer. ysis of the Estonian Internet Voting System.’’ In Proceedings
Mercuri, R. (2001). ‘‘Electronic Vote Tabulation Checks & Bal- of the 2014 ACM SIGSAC Conference on Computer and
ances.’’ PhD diss. University of Pennsylvania. Communications Security. 703–715. New York, NY: ACM.
Mitrou, L., D. Gritzalis, and S. Katsikas. (2002). ‘‘Revisiting Spycher, Oliver. (2018). ‘‘Verifiable E-Voting in Switzerland.’’
Legal and Regulatory Requirements for Secure E-Voting.’’ Presentation delivered to the 2018 Evote-ID Conference,
In Security in the Information Society. 469–480. Boston, October 4.
MA: Springer. Stein, R., and G. Wenda. (2014). ‘‘The Council of Europe and
Montgomery v. Balkissoon. (1998). ONCA, 38 O.R. (3d) 321. E-Voting: History and Impact of Rec(2004)11.’’ In 2014
(CanLII). 6th International Conference on Electronic Voting: Verify-
National Confrerence of State Legislatures (NCSL). (2018). ing the Vote (EVOTE). 1–6. IEEE.
‘‘Voting System Standards, Testing and Certification.’’ Thompson, J.R. (2018). ‘‘Email Voting in Indiana Elections.’’
Retrieved from <http://www.ncsl.org/research/elections-and- E-Vote-ID 2018, 98.
campaigns/voting-system-standards-testing-and-certification Thompson, E. (2019). ‘‘Canada Should Regulate Social Media
.aspx>. Firms to Protect Election, Says International Group.’’ CBC
Norris, P. (2004). ‘‘E-voting as the Magic Ballot for European Par- News, June 11. Retrieved from <https://www.cbc.ca/news/
liamentary Elections? Evaluating E-Voting in the Light of politics/social-media-election-disruption-disinformation-
Experiments in UK Local Elections.’’ In The European 1.5171462>.
Union and E-Voting (Electronic Voting). 72–102. Routledge. Toronto. (2013) ‘‘Internet Voting for Persons with Disabilities
Organization for Security and Co-operation in Europe (OSCE). in the 2014 Regular Municipal Election. Request for Pro-
(2013). Handbook for the Observation of New Voting Tech- posal No. 3405-13-3197.’’ Purchasing and Materials and
nologies. Warsaw: OSCE/ODIHR. Management Division. Retrieved from <https://www
Porter, K. (2018). ‘‘Towns Seek Answers After Voting Snafu.’’ .verifiedvoting.org/wp-content/uploads/2015/01/RFP_
CBC News. Retrieved from <https://www.cbc.ca/news/ Internet_Voting_Toronto_2013.pdf>.
canada/ottawa/ottawa-valley-towns-seek-answers-voting- Unt, T., M. Solvak, and K. Vassil. (2017). ‘‘Does Internet Voting
snafu-1.4875096>. Make Elections Less Social? Group Voting Patterns in Es-
Prandini, M., and M. Ramilli. (2011). ‘‘Taking the Best of Both tonian E-Voting Log Files (2013–2015).’’ PloS One 12(5):
Worlds: A Comparison and Integration of the US and EU e0177864.
Approaches to E-voting Systems Evaluation.’’ In 2011 U.S. Department of Homeland Security (DHS). (2018). ‘‘Critical
44th Hawaii International Conference on System Sciences. Infrastructures Sector—Government Facilities Sector.’’ Cyber-
1–10. IEEE. security and Infrastructure Security Agency. Retrieved from
Puiggalı́, J., and A. Rodrı́guez-Pérez. (2018). ‘‘Defining a <https://www.dhs.gov/cisa/government-facilities-sector>.
National Framework for Online Voting and Meeting Its U.S. Election Assistance Commission (EAC). (2015). Voluntary
Requirements: The Swiss Experience.’’ Electronic Voting: Voting System Guidelines 1.1. Retrieved from <https://
International Joint Conference on Electronic Voting, www.eac.gov/assets/1/28/VVSG.1.1.VOL.1.FINAL1.pdf>.
E-Vote-ID 2018, 82. U.S. Election Assistance Commission (EAC). (2017). Volun-
R. v. Oakes. (1986). 1 S.C.R. 103. tary Voting System Guidelines 2.0 Principles and Guide-
Remmert, M. (2004). ‘‘Towards European Standards on Elec- lines. 1–5. Retrieved from <https://www.eac.gov/assets/1/
tronic Voting.’’ In Electronic Voting in Europe, 13–16. 6/TGDC_Recommended_VVSG2.0_P_Gs.pdf>.
 DEVELOPING E-VOTING REGULATIONS IN CANADA 179

U.S. Election Assistance Commission (EAC). (2009). State Whitmore, K. (2008). Information Report on the Electronic
Requirements and the Federal Voting System Testing and Voting in the Finnish Municipal Elections. Congress of
Certification Program. Retrieved from <https://www.eac Local and Regional Authorities. Retrieved from <https://
.gov/assets/1/1/State%20Requirements%20and%20the% wcd.coe.int/ViewDoc.jsp?id=1380337&Site=Congress>.
20Federal%20Voting%20System%20Testing%20and%20
Certification%20Program.pdf>. Address correspondence to:
U.S. National Institute of Standards and Technology (NIST). Nicole Goodman
(2016). NIST SP 1500-100 Election Results Reporting Department of Political Science
Common Data Format Specification Version 1.0. February. Brock University
Retrieved from <https://pages.nist.gov/ElectionResults
Reporting/>. 1812 Sir Isaac Brock Way
U.S. National Institute of Standards and Technology (NIST). St. Catherines, ON
(2017). VVSG Principles and Guidelines. Retrieved from L2S 3A1 Canada
<https://collaborate.nist.gov/voting/bin/view/Voting/VVSG
PrinciplesAndGuidelines>. E-mail: nicole.goodman@brocku.ca
Verified Voting. (2018). ‘‘Polling Place Equipment.’’ Verified
Voting. Retrieved from <https://www.verifiedvoting.org/ Received for publication March 24, 2019; accepted
verifier/#/year/2018/>. November 26, 2019; published online May 4, 2020.
Downloaded by 223.187.185.185 from www.liebertpub.com at 08/04/21. For personal use only.

Appendix A1. List of Interview Participants

Please note one interview participant is not listed 12. Rick Johal, Director of Member and Sector
below as they wished to remain anonymous. Relations, Association of Municipal Manag-
ers, Clerks and Treasurers of Ontario
1. Chris Bantock, Legislative Coordinator, Town 13. Mark Keating, Senior Technical Director,
of Whitby Neuvote
2. Susan Beckel, Clerk, Town of Greater 14. Kimberley Kitteringham, Clerk, City of
Napanee Markham
3. Richard Catahan, General Manager, Scytl 15. Brian Lack, President, Simply Voting
Canada 16. John Meraglia, Manager, Election Services,
4. Nicole Cooper, Clerk, Town of Ajax City of Toronto
5. Greg Essensa, CEO, Elections Ontario 17. Fiona Murray, Deputy City Clerk, Election
6. Joshua Franklin, Founder, OutStack Technol- Services, City of Toronto
ogies 18. Michael Pal, Associate Professor, Faculty of
7. Diane Gould-Brown, City Clerk, City of Sarnia Law, University of Ottawa
8. Joseph Lorenzo Hall, Chief Technologist, 19. Martha Petit, Deputy Clerk, City of Markham
Center for Democracy and Technology 20. Jordi Puiggali, CSO and SVP Research and
9. Chris Harris, Clerk, Town of Whitby Security, Scytl Secure Electronic Voting
10. Matthew Heuman, CEO, Neuvote 21. Gwendoline Savoy, Senior Marketing Direc-
11. James Jenkins, Deputy City Clerk, City of tor, Scytl Secure Electronic Voting
Sarnia 22. Ulli Watkiss, Clerk, City of Toronto
 This article has been cited by:

1. Holly Ann Garnett, Sean Grogan. 2021. I Came, I Saw, I Voted: Distance to Polling Locations and Voter Turnout in Ontario,
Canada. Canadian Journal of Political Science 54:2, 316-334. [Crossref]
Downloaded by 223.187.185.185 from www.liebertpub.com at 08/04/21. For personal use only.