Summary of the ISF report

Evaluation of the e-voting trial in 2011. Accessibility for voters, trust, secrecy and election turnout

The report presents the findings from an evaluation of the internet voting trial (i-voting) that took place in ten Norwegian municipalities in the 2011 local elections. The evaluation is based on different types of qualitative as well as quantitative data such as representative voter surveys, data on people with disabilities, the youngest voters and the local medias’ coverage of the trial. The analysis is structured in five main parts which shed light on the following aspects: voter turnout, accessibility for the voters, faith in the electoral process and i-voting, secret voting, and younger voters’ relationship with i-voting and voting in general.

The aim of the report is to assess the  i-voting trial as a democratic project and provide a basic understanding about the democratic effects of the trial.

Chapter 1: Introduction

No summary

Chapter 2: Election turnout and internet voting

The aim of this chapter is to test whether internet voting has a positive effect on electoral participation in the trial municipalities. We investigate this, first, by looking at aggregate turnout levels in the trial municipalities over time, and compare these to the country as a whole. We find that in previous elections, the trial municipalities have had somewhat lower turnout levels than the county as a whole. That remains the case in the 2011 elections. The three percentage-point increase in turnout in the country as a whole from the 2007 to the 2011 elections is replicated in the trial municipalities. There are no indications that the trend in turnout in the trial municipalities deviates from the country as a whole in 2011. Hence, the aggregate-level numbers suggest that the trial had no effect on turnout.

The second approach to analyzing turnout in the trial municipalities is to look at individual level data. One may envisage that some groups of voters have been mobilized as a result of internet voting, even if that does not have an effect on aggregate turnout-levels. We look at voters’ social background, attitudes, and participation in previous elections to find out of internet-voters have other characteristics than paper-voters. When we identify a group which is overrepresented among internet-voters, we test whether that group has higher turnout in the trial municipalities than in the country as a whole.

The results of the individual-level analyses are, first, that internet-voters are quite similar to paper-voters on most of our variables. Second, when there are differences between internet- and paper-voters, we find that those who are especially likely to vote online do not have higher turnout in the trial municipalities than in the country as a whole. Thus, we find no evidence that groups of voters have been mobilized to take part in the election as a result of internet voting.

Finally, we find that 89 per cent of internet-voters respond that they would have voted even in the absence of the online voting-option. The remaining 11 per cent claim that they would not have cast a vote if they could not do so on the internet. We argue that this result overestimates the share of internet-voters who would not have voted if there was no trial.

The analyses, in sum, indicate that the trial did not have an effect on voter turnout. Since this is an analysis of a single trial only, we should caution that there could be long-term effects of internet voting that we are unable to uncover.

Chapter 3: Availability and accessibility for voters in general

A partial goal of the Norwegian internet voting trial was to «increase access for voters in general» to the voting act. This chapter asks if access has increased in a way that impact on voters’ decision to take part in the election or not. We find that internet voting is widely used, by about a quarter of all voters. However, internet voting has no effect on overall turnout. People in the trial municipalities who do not vote are as likely to say that they abstained from voting because of time-constraints, as those in the rest of the country. Thus we conclude that internet voting did not impact on voters’ decision to take part in the election.

Nevertheless, people who voted online are very happy with internet voting, and report that it was easy to cast a ballot in this manner. We therefore argue that internet voting did increase access to the election for people who actually voted.

Chapter 4: Availability and accessibility for voters with disabilities

One of the primary objectives of the Norwegian e-voting trail was to ensure that the actual act of casting a vote is accessible. The main questions in the report concerns the extent to which people with disabilities experienced the e-voting system as accessible, easy and free from significant hurdles, i.e., universally designed. Further, the study sheds light on these voters’ experiences, thoughts and opinions, both in relation to e-voting and elections in polling stations. Another question is whether e-voting can make it easier for people with disabilities to exercise their democratic rights. To answer these questions, a qualitative study among 30 people with disabilities was conducted. The study was conducted during the e-voting period of the 2011 municipal and county elections in Norway. Participants in three municipalities, Sandnes, Ålesund and Re, were observed and interviewed while they used a copy of the e-voting system to cast a vote. The study shows that the e-voting system was inaccessible for users of certain types of assistive technology, many participants had problems related to logging in to the system, and the solution had several deficiencies in relation to universal design. On the other hand, many participants expressed that they wanted to have e-voting as a supplement to voting in polling stations. For the voters to exercise their democratic rights, many aspects are important. The participants in this study emphasised the positive aspects of e-voting, such as being able to vote on their own, independent of others, and free from stress and time pressure.  The report indicates that e-voting may contribute to this goal provided that the solution is universally designed.

Chapter 5: Trust in the Norwegian election system

Voter confidence in the electoral process in Norway is extremely high. This chapter asks if the introduction of internet-voting can affect these high trust-levels. We start by looking at attitudes toward internet-voting, and find overwhelmingly positive views. Norwegian voters in general, and especially those in the municipalities where internet-voting has been tried, are very favorable toward online voting. Given this favorable attitude, one would expect confidence in the electoral process to be unaffected by the introduction of online voting. However, trust in the electoral process is somewhat lower in the municipalities where the internet voting trial was held, than in the rest of the country. The lower trust levels are not the result of negative attitudes toward internet voting. We argue that the unusual attention given to the electoral process in connection with the internet voting trial promotes some critical thinking on the issue, which in turn can explain a small reduction in overall trust levels.

Chapter 6: Local media coverage of the e-voting trial. Local media as opinion makers

In this project we have studied how the local media has covered the e-vote trial in the ten trial municipalities and whether there is any correlation between the media coverage and people’s attitude towards the arrangement. We have conducted a content analysis of all media postings in the most frequently read newspapers in the trial municipalities, for a period of two months. The results from the content analysis have been compared to people’s attitudes towards and confidence of the e-vote trial, measured in a representative population survey in the trial municipalities, the e-vote survey. Moreover, we have interviewed journalists and editors from these same newspapers. The content analysis shows that the coverage of the e-vote trial has not been very extensive, although it has somewhat varied between municipalities. The majority of posting have had a positive or neutral angle, but a substantial part of postings have been balanced, in the sense that they have contained both positive and negative arguments about the e-vote trial. Few postings, however, have had a negative angle, and the bulk of these appeared after the election and discussed why the e-vote trial had not contributed to increased voter turnout, as was expected. In many articles, the e-vote trial has been related to voter turnout, especially among young people, and it has been described as a solution to the problem with low turnout. There is no clear connection between local media’s coverage of the e-vote trial and people’s attitudes towards the arrangement. A reason for the lacking correlation may be that the coverage has been moderate. Another reason may be that people do not feel strongly enough about the e-vote trial for the media coverage to have an effect upon their attitudes.

Chapter 7: Arguments for and against the e-vote experiment in the local online newspapers

This chapter summarizes the most frequent arguments mentioned in the local online newspapers in the municipalities that participated in the e-vote experiment. In total, this study examined 85 articles, roughly from August 1-September 20, 2011. There were more arguments in favour of the e-vote than there were against. The most frequent positive arguments were that the e-vote would 1) increase participation (which did not correspond with the KRD’s stated aims), 2) modernize Norway’s democracy, and 3) have safeguards to uphold the secret ballot. The negative arguments were that this alternative would 1) jeopardize the secret ballot, 2) be vulnerable to hacking and viruses, and 3) detract from the solemnity of voting. Especially notable was the lack of intensity in the local online newspapers compared to the debate in the Storting. This omission in the online newspapers is suggestive of a high level of political trust and therewith trust in the electoral system in Norway.

Chapter 8: Secrecy of the vote

In the first part of the chapter, the principle of the secret ballot is discussed as a background for the following empirical analyses: to what extent are internet voting and the secret ballot compatible, and to what extent should the secret ballot be considered a duty – and not only a right? The following sections deal with the voters’ behaviour and attitudes to the secret ballot. It is difficult to estimate the extent of buying/selling votes and undue influence. Both problems do apparently exist, but to a very limited degree. The norm prescribing that internet voting shall take place in a private room does not seem to have taken strong roots among Norwegian voters, neither in attitudes nor in behaviour. There is widespread support for internet voting among the citizens of the trial municipalities, also among those who did not vote via the Internet. This support declines somewhat when the principle of the secret ballot is presented as a counter-argument, but there is still a large majority in favour of internet voting. There is, however, also a large majority that agrees that the state, not the individual voter, should secure the secrecy of the vote. There are some differences between social and demographic groups regarding internet voting, but the issue is not politicized along party lines at the voter level.

Chapter 9: Traditionalists with trust in technology. A case study of young voters' attitudes to Internet voting

In this article, the author address young voters attitudes and acting towards e-voting. The analysis is based on a field work in the municipality Mandal two weeks after the local election in 2011. The municipality did also participate in the trial of reduced voting age to 16. Nine informants in the age of 16 –18 were interviewed in two focus groups. The article explores the following research questions: (1) Do the young informants experience that online voting increase their accessibility to vote? (2) Do they care about the potential security flaws with e-voting and do they trust the technology? (3) Are they aware of the potential threats of core values of democracy, like secret voting and no protection against intimidation or vote selling? The findings imply that the young informants are traditionalists with optimistic attitudes towards online voting. Among them, online voting is well known, noncontroversial and not a subject they discuss with their peers. However, these young voters prefer to walk to the polling station on Election Day. They defined traditional voting as a symbolic and ceremonial act that indicates adultness. Over all, these informants are more concerned about the question why young people should vote than how they will vote.