Technology and the Voting Process
VI. Canadians' Attitudes to New Voting Technologies
A. 1997 Canadian Election Study
Elections Canada commissioned a number of questions in the 1997 Canadian Election Study aimed at determining the extent to which Canadians were open to the idea of electronic voting and various voting alternatives. Respondents were asked a series of questions about their willingness to use, and preferences among, five possible methods of voting. The methods of voting were mail, telephone, computer, using a touch screen computer, and using a counting machine.Footnote 19
The main findings are summarized below.
1. Vote by mailFootnote 20
Overall 31 percent of survey respondents indicated a willingness to vote by mail, with respondents aged 18-24 most willing to use this method (40 percent). Fewer unemployed (30 percent) and retired (20 percent) persons were willing to vote by mail, in comparison to students or homemakers (36-39 percent), and only 22 percent of persons with a disability were willing to vote by mail. Voting by mail was the most likely method for 20 percent of respondents and the least likely method for 45 percent of respondents. The most common reasons for choosing mail as the least likely method were: confidentiality issues (31 percent), slow/delays (19 percent), and a concern that the vote would lose its importance (10 percent).
2. Vote by telephone
36 percent of survey respondents indicated a willingness to vote by telephone, with 44 percent of respondents aged 18-24 willing to use this method. Fewer retired persons (18 percent) than others were willing to vote by telephone, and 39 percent of respondents with a disability that affects their mobility or agility were willing to vote by phone, compared to 16 percent with a disability that affects their sight or hearing. Voting by telephone was the most likely method for 26 percent, and the least likely method for 40 percent of respondents. Common reasons for choosing vote by telephone as least likely were: confidentiality (43 percent), risk of fraud (21 percent), and accuracy (9 percent).
3. Vote by personal computer
Overall 29 percent of survey respondents were willing to vote by personal computer. That proportion increases to 43 percent among those aged 18-24. More students are willing to vote by computer (50 percent), compared to working (33 percent), unemployed (31 percent), or retired (10 percent) people and homemakers (22 percent). Only 16 percent of respondents with a disability reported a willingness to vote by computer (all types of disability). Voting by computer was the most likely choice of method for 18 percent of respondents, and the least likely method for 43 percent of respondents. The most common reasons for choosing computer as the least likely method were: problems with access (37 percent), and confidentiality (24 percent).
4. Vote by touch screen computer
54 percent of survey respondents were willing to vote using a touch screen computer. The touch screen method was popular among working people (56 percent willing to use), students (64 percent), unemployed persons (50 percent), and homemakers (51 percent). 30 percent of respondents with a disability that affects their sight or hearing were willing to vote this way, compared to 52 percent with a disability affecting their mobility or agility. Voting by touch screen was the most likely choice of 37 percent of respondents, and the least likely method for 16 percent of respondents. The problems cited most often by those who chose touch screen as least likely method were: a risk of fraud or tampering (15 percent) and confidentiality (22 percent).
5. Vote using a ballot counting machine
Overall, 67 percent of survey respondents report being willing to vote using a counting machine. The highest proportion of willing responses came from respondents aged 65 and over (77 percent). 73 percent of retired persons were willing to vote by this method. 64 percent of respondents with a disability report a willingness to vote using a counting machine. Counting machines were the most likely choice of 52 percent of respondents and the least likely choice for 16 percent. Inconvenience (32 percent) was the most commonly cited reason for selecting counting machine as the least likely choice.
Based on these results we set out to model the voting process using the telephone, Internet (vote by computer), and electronic kiosk (vote by touch screen). We did not model ballot counting machines as they are more geared toward tabulation and less the actual process of casting a decision, nor did we model voting by mail since it is not a technologically innovative option.
6. Summary of voting preference survey results
|Method||phone||computer||touch screen||counting machine|
|% willing to use||31%||36%||29%||54%||67%|
|Method||phone||computer||touch screen||counting machine|
|most likely (%)||20||26||18||37||52|
|least likely (%)||45||40||43||16||16|
|total||phone||computer||touch screen||counting machine|
|risk of fraud||14||14||21||12||15||6|
B. Public Policy Forum Roundtables
Elections Canada engaged the Public Policy Forum to convene a series of roundtables with key electoral stakeholder groups – Members of Parliament and Senators, persons representative of voting groups that have a major stake in the use of new technologies for electoral purposes, and selected opinion leaders. These roundtables took place between March and May 1998 and are summarized below.
1. Members of Parliament and Senators
The roundtable discussion with Members of Parliament focused on ways of complementing or enhancing the current voting system with technology. Three areas of concern associated with technological change emerged from the discussions.
- Technological issues – participants considered Canada's technological requirements for elections, focusing on the issues of security and suitability.
- Social issues – some discussion was devoted to the social aspects of technological change particularly the fact that not all segments of the population would be able to experience or benefit from changes to the same degree.
- Political issues – the political aspects of technological change were debated including the issue of whether or not technological change would alter Canada's system of representative democracy.
The discussion opened with the question of whether or not technological change would actually benefit or enhance the voting process. While it was recognized that technologies like the telephone and the Internet could help to overcome some of the accessibility problems faced by the disabled, the elderly, and individuals with limited sight or hearing – and possibly help to increase voter participation – it was also recognized that electronic voting attempts in the recent past have met with varying degrees of success, and that public opinions on new technology are varied.
Concerns were also raised that the new technologies might alter the fundamental nature of Canadian politics. Some participants felt that lessening the effort required to cast ballots (deformalizing the process), might prompt voters to take an equally casual view of the issues at stake in the election. Thus, the advantages of using technology to increase voter participation could potentially be offset by unintentional changes in the vote itself.
In addition to the immediate questions of the necessity for change, two main issues emerged during the discussion concerning the implications of adapting new technologies to the election process – security and applicability.
In terms of security, concern was expressed that the integrity of an electronic voting option could be compromised in the event of an elector's password or other personal identifier getting into the wrong person's hands. The recent North York telephone plebiscite was stated as an example where personal identification numbers (PINs) were distributed through the mail and then used by electors to access the telephone voting system. It was argued that individuals might be tempted to intercept these PINs and to corrupt the vote. On a smaller scale, it was suggested that a single voter might coerce members of the household to vote for a particular candidate. Similar security issues were discussed for other technologies like the Internet, and kiosk-voting.
In terms of applicability, there was much discussion as to which technology should be used (the telephone, the Internet, or kiosk) given their differing penetration levels in Canada. The respective advantages and disadvantages of the technologies were discussed.
It was recognized that the telephone has the highest penetration rate among the three technologies, and as such, seems the obvious choice. However, the security issues raised in respect to PINs concerned Parliamentarians, as did the high number of rural telephone customers connected by multiple party lines who would not be able to securely take advantage of this option.
While kiosks, in the form of automated teller machines (ATMs) can be found on almost every street corner, Parliamentarians noted that ATM fraud is still common and that such a system, which like the telephone relies on the activation of a personal identifier, has security issues. Concerns were also raised that an ATM-style system would be difficult to implement in low-density, rural areas.
The idea of personal computer or Internet-based voting was also discussed given the proliferation of these devices. Participants praised the Internet's potential to inform voters of the issues and candidates' platforms; however, they questioned its use as a voting tool given that it is the least secure of the three technologies.
Parliamentarians and Senators were, however, enthusiastic about the possible utilization of various voting technologies to improve accessibility for specific groups within the electorate.
Discussion of the technical implications of various technologies led to the identification of a number of social issues which might accompany any technological change. Parliamentarians and Senators noted that technologies rarely affect different segments or age groups of the population in the same ways. The introduction of new technologies inevitably requires some learning on behalf of the users which could negatively impact their ability to take part in the voting process. As a result, participants agreed that care must be taken to prevent new tools from excluding any sector or strata of society.
Parliamentarians and Senators engaged in a discussion of the nature of Canadian democracy and the role that technological change could play therein. The principal issue raised was that the Canadian political system is based on representative, not direct democracy, and that the potential for decentralization in an electronic voting system (i.e., one in which voting takes place in remote locations as opposed to the polling station) could potentially blur the lines between the two.
Participants agreed that new technologies like the Internet are excellent means of informing the electorate and incorporating greater numbers of people in policy discussions. However, in terms of increasing accessibility to the voting process through the use of such technology, some participants argued that decreasing voter turnout is more likely due to a lack of engagement in political issues, attitudes toward politicians, and apathy, due to the accessibility to the process.
Voter participation received considerable attention as participants discussed the value of raising voter turnout through enhanced technology. Some members of the group questioned whether or not increased participation implied greater representation. In short, while the use of new technologies could be perceived to be part of the "inevitable march of democracy," there was concern that introducing new means of voting without understanding one's objectives for doing so could endanger the system in place.
The participants in the roundtable discussion agreed that there exists a need to establish objectives for enhancing the voting process with new technologies.
A consensus emerged that Elections Canada should arm itself with as much knowledge as possible on the advantages and disadvantages of the new technologies, as well as the different perspectives of Canadians toward the introduction of new means of voting. Parliamentarians and Senators suggested that Elections Canada should make efforts to: track attitudes toward technological enhancement using time series data; keep abreast of technological developments which could improve the voting process; create a database to track innovations and experiments in electronic voting; and remain on the leading edge in its use of new voting technologies. They were also supportive of the idea of pilot testing the new technologies for groups for whom accessibility to voting is a major concern.
2. Representatives of voting groups
Roundtable participants were identified and specifically invited to the session because of their ability to represent the views of various groups within Canadian society. This consultation process provided a preliminary sample of the challenges faced by individuals with low literacy skills, with physical disabilities, with visual and hearing impairments, the elderly, new Canadians and aboriginals. Each representative was asked to assess three potential technologies as they relate to the voting process – the telephone, touch screen kiosks, and the Internet.
Literacy groups expressed their need for a voting process which incorporates "plain language" and which is consistent from year to year. In evaluating the three proposed technologies, literacy groups emphasized that it is not so much an issue of accepting new technology but understanding how it works. Voting by telephone was generally supported as long as the process would be easy to understand and participate in. Kiosks were also popular as they incorporate pictures and use voice activated technology. The use of the Internet was seen as a potentially useful tool as long as individuals were properly educated about, and had access to, the system.
Representatives of persons with disabilities supported the telephone as the most viable option for their stakeholder group. This option allows electors to stay in the comfort of their own homes. Kiosks were seen as useful, but not easily accessible to people with certain disabilities. The Internet was recognized as potentially beneficial, although few stakeholders within this group have access to the Internet.
The most important characteristic of any voting alternative for individuals with visual impairments is the ability to increase their independence and confidence in participating in the electoral process. It was recognized that voting by telephone would be a possibility so long as the process was easy to use and provided enough time to comfortably negotiate the various options and steps throughout the process. Kiosk and Internet technologies were seen as less useful if they relied on reading a screen.
The hearing impaired community expressed substantial concern with the current TTY system (telecommunications devices for the hearing impaired). The most accommodating method of communicating with this group is through one-on-one sign language. Kiosks were largely supported by this group because the interface between the technology and the user does not rely on the spoken word.
While the seniors community in Canada is increasingly embracing Internet technology, in general, many seniors are fearful that technology will take away their access to human services. Numerous individuals within this group are fearful that if they support technology alternatives, the more traditional, and in many cases, preferred service options will be taken away. Voting by telephone was largely supported by representatives of this group as it would facilitate voting for those confined to their homes or nursing home facilities. The only concern expressed about telephone voting was that it not be designed in such a way as to require use of the touch pad. Many seniors have trouble with fine motor skills. This could be ameliorated through the use of integrated voice technology.
Participants representing new Canadians focused on education and awareness issues. Many new Canadians come from different cultures and are not fluent in French or English. Technologies which offered multilingual capabilities would assist this group.
Representatives of the aboriginal community were largely supportive of the possible voting alternatives. It was pointed out that on many reservations, telephone penetration rate is not high, and the potential of the Internet has not yet been realized. Another concern tabled by aboriginal representatives was that of language and communication. High quality translation services are seen as essential.
The participants in the roundtable discussion agreed that the possibility of improving accessibility to the voting process should be further explored. Participants cautioned Elections Canada to proceed in a consultative fashion however and to not limit the voting options available to groups within the electorate. This comment reflected some dissatisfaction among representatives with the introduction and proliferation of technology with a perceived lack of choice in the banking industry.
3. Selected opinion leaders
The roundtable discussion with selected opinion leaders covered all segments of the voting process from registration and vote casting, to validation, collection and tallying of results. In addition, the group discussed issues such as privacy, transparency, accessibility, integrity, universality, accuracy and anonymity. The most salient points from the discussion of verification of eligibility to vote and the use of technology in the act of casting a decision are summarized below.
Voter identification technologies
This step in the voting process ensures that the individual who is registered to vote is the person who actually casts the vote. Four technologies were put forward by the participants to deal with this challenge: cryptography, voice recognition, biometrics and smart cards.
Cryptography (the recognition of hand writing and signatures) is seen to be a promising technology that is developing very rapidly. Voice recognition was cited as a growing and increasingly popular medium of verification, where current systems can now transact thousands of calls daily within a secure and verifiable environment. Biometric identification is also becoming more practical, and it was noted that Canadians seem to be less concerned with using the face as a biometric identifier than fingerprints because fingerprinting continues to be associated with the criminal system.
While these options were recognized as valid and likely able to meet the needs of the voting system, it was generally agreed that a fourth option – smart cards – may be even better suited to the requirements of Elections Canada. A smart card is a transaction card, such as a health card, that has an embedded electronic chip with the capacity to carry a large amount of data. It would be possible to issue all Canadians with such cards such that the chip would be activated as soon as the individual reached eligible voting age. Smart cards could be integrated with kiosks or more advanced telephones that have card swipe technology. The general preference of the participants was towards use of smart cards with telephones since they are readily accepted and accessible to society and would be less costly to implement than individual kiosks across the country.
The policy issues associated with the introduction of many of the alternatives were seen as considerably more challenging than technological feasibility.
While linking federal and provincial databases to produce a more accurate voter registration list may be beneficial to Elections Canada, the broader societal reaction to this process could be very negative. To what degree do Canadians want their movements tracked and recorded by governments? Smart cards are also a technologically superior product, but they may be viewed by individuals as intrusive and a threat to privacy. Citizens may view technology as being too smart – potentially threatening the anonymity of the voting process.
Participants recognized that there will be trade-offs between certain attributes of the voting system. Focusing on the accuracy and integrity of the registration lists may in fact create more barriers as the system becomes more complicated and potentially more intrusive to the users. Introducing new technologies that create greater verifiability within the system may threaten anonymity. Participants recommended that Elections Canada be clear about the specific aspect of the voting system that is intended to be addressed by introducing new technologies. If accessibility is the key issue, then there must be a recognition of the impact that voting alternatives will have on the other attributes of the election process.
Drawing from the experiences of the participants in introducing new, technologically advanced alternatives to various clients and stakeholders, it was agreed that a hybrid approach would be beneficial. Different services meet the needs of different people and Canadians want to have a range of choices in accessing services. The objective should be to maintain the current level of service to the public, and use new technologies to build multiple channels to access the voting system.
Return to source of Footnote 19 The Canadian Election Study is a three wave survey of the Canadian electorate. The first survey was conducted during the election campaign period and 3,925 interviews were completed by telephone. The second survey was conducted in the period following the election and 3,163 interviews were completed, also by telephone. The third study was a self-administered mail-back survey; 2,500 questionnaires were sent and approximately 1,850 were returned. The methods of voting questions were asked in the mail-back questionnaire.
Return to source of Footnote 20 It should be noted that voting by mail is already permitted under Schedule II of the Canada Elections Act (the Special Voting Rules).