open Secondary menu

The Burden of Voting in the 2019 Canadian Federal Election

Chapter 1: Introduction

The objectives of this study are to determine how easy or difficult the act of voting is and to ascertain the impact of this "burden" in the decision to vote or not. Our study deals specifically with the 2019 Canadian federal election. We wish to find out whether, in that election, some/many Canadians did not vote because it was too complicated/difficult.

Voting must be as easy as possible for every citizen. Electoral Management Bodies such as Elections Canada have to work toward this, while also considering other goals, such as minimizing the risk of fraud, that may come into conflict with that of minimizing the burden of voting.

Voting is a fundamental democratic right. Exercising that right should be as easy as possible because turnout should be as high as possible. A high turnout contributes to the quality of democracy and is often considered to be a crucial criterion of democratic performance (Powell 1982). The higher the turnout the greater the attention that policy-makers pay to citizens' preferences (Martin and Claibourn 2013; Verba and Nie 1972; Hansen 1975; Hill and Matsubayashi 2005; Blais, Dassonneville and Kostelka Forthcoming). As a consequence, all qualified electors who wish to vote should be able to cast their ballot. This means that all the barriers that prevent people from casting a vote should be removed to the extent possible.

Our objectives are thus simple and straightforward. The first is descriptive. We need to know how much of a burden casting a ballot is for Canadians, how many find it (somewhat or very) difficult. We also need to find out how many do not vote because of that burden.

We distinguish two basic reasons why a person who has the right to vote decides not to vote in an election. The first is that they do not want to; they are not interested, and feel that it is OK to abstain, that is, do not feel that they have a moral obligation to vote. A person might also be disappointed in politics in general and might not be willing to endorse any of the political parties running in the election. In short, they are not motivated to vote (Blais and Daoust 2020). The second is that they are not able to; they would like to, but voting is too difficult or complicated. This is the burden of voting, which corresponds to the famous C ("cost") term in the rational choice model (Downs 1957; Riker and Ordeshook 1968). Our focus is on the second factor, the burden, but we need to consider motivations when we ascertain the relative weight of the burden in the decision to vote or abstain.

There are "objective" and "subjective" burdens. There is the actual distance between my residence and the polling station. And there is my perception of the time that it would take to go to the polling station, to possibly wait in a queue, to cast my ballot, and to come back home. This perception is informed by things like past experience with similar situations, age, physical abilities and limitations, and the environment. What matters in the end is the elector's subjective perception, however accurate or inaccurate it may be, since the decision to vote or not is necessarily based on that subjective assessment. We therefore focus on citizens' subjective perceptions, though we wish to determine to the extent possible whether they are misguided or not.

There is a substantial amount of research on "objective" burdens. Many studies have shown that distance from the polling station affects turnout (Gimpel, Lay and Schuknecht 2003; Dyck and Gimpel 2005; Gimpel, Dyck and Daron 2006). Brady and McNulty (2011) also demonstrate that changing the location of a polling station reduces turnout. There is also vast literature about the impact of registration laws or early voting (Rosenstone and Wolfinger 1978; Highton 1997, 2004; Nagler 1991).

Using survey data, Rosenstone and Wolfinger (1978) examine the effect of state registration laws on voter turnout across 50 states in the United States, from 1960 to 1973. States in the United States vary greatly in their registration hours and locations. They find that if all states had the registration laws that prevail in the most permissive states, turnout would have been about nine percentage points higher in the 1972 presidential election. They suggest that registration requirements are the most substantial barriers to voting and that turnout would be much higher in a European-type system where registration is the state's responsibility.

Highton (1997) offers a similar observation. Using data from the 1980 and 1992 Voter Supplements of the Census Bureau's Current Population Survey with states that offer election-day registration and registration prior to election day, he finds that turnout is 10 percentage points higher in states with election-day registration. Highton also compares the impact of education on turnout in states that have election-day registration with states that do not, and he finds that education has a stronger impact on turnout in states that do not have election-day registration. On the other hand, Nagler (1991) uses the 1972 and 1984 Current Population Survey to examine if poorly educated individuals are more deterred from voting by registration laws than highly educated individuals, and he finds no evidence that more permissive registration laws differentially increase registration by lower educated individuals.

In yet another article Highton (2004) reviews the recent literature on registration in the United States and argues that a number of reforms, including poll taxes, literacy requirements, registration closing dates and motor voter programs, influence voter turnout rates. He suggests that states could extend polling place hours and allow mail-in ballots to further reduce voting costs.

Much less is known about the "subjective" burden of voting, that is, about whether people feel it is difficult to vote. The Making Electoral Democracy Work (MEDW; see Stephenson et al. 2017) project, which includes 27 single-election surveys conducted in Canada, France, Germany, Spain and Switzerland, includes a question asking respondents to indicate how (very or somewhat) easy/difficult it is for them to vote. All in all, 40% responded that it is very easy and 38%, somewhat easy; only 4% answered that it is very difficult (Blais and Daoust 2020, 67). It thus seems that few people see the act of voting as a burden. This is an encouraging sign.

We cannot rule out, however, the possibility that the burden is perceived to be substantial in some subgroups of the population. Furthermore, the MEDW question provides only a general assessment. Voting entails a number of steps. In order to vote, one has to be registered on the voters' list, to research information about the candidates and parties, to find out where, when, and how one can cast a vote, and, finally, go to the polling station and cast one's vote. Each step involves physical and cognitive efforts; some require information and documents that are more or less accessible to different electors. As a result, many of these steps may be very easy but others may be challenging, at least for some voters.

In short, we need to distinguish different kinds of burden. In another study, Blais et al. (2019) distinguish what they call direct and indirect burdens (or costs), the former being associated with the act of voting as such and the latter with the effort to make an informed choice. The authors find that the great majority of people find it very easy to go to the polling station and to make up their mind about the parties and their leaders. As expected, both types of burden have an impact, though small, on the turnout decision. The distinction between direct and indirect burden is interesting though not fully satisfactory, since it does not take into account the various steps, starting with registration, that a voter needs to undertake in order to cast their ballot. We fill some of that gap in this study.

Using data from the National Electors Study (NES) 2019 (see methodological details in Chapter 2), which was conducted by Elections Canada, our analysis will proceed in three steps. The first step is purely descriptive (see Chapter 3). We provide an accurate assessment of the subjective burden of voting among Canadian electors. Is it correct to say that, all in all, voting is viewed to be quite easy by the great majority? We then distinguish the various steps to determine which ones are perceived to be the easiest and the most difficult. We then compare pre-electoral perceptions of the burden of voting with post-electoral assessments of the actual act of voting to determine whether Canadian electors tend to overestimate the burden of voting.

In a second step, we examine whether the burden of voting is higher for some subgroups of the population (see Chapter 4). We focus on four groups: Indigenous electors, persons with disabilities, those who have never voted before, and people with less education, particularly those not in education, employment or training (NEET) who also are within the 18–34 age group. From the existing literature and the previous Surveys of Electors conducted by Elections Canada 2 , we know that turnout is lower among youth and those with less formal education (Wolfinger and Rosenstone 1980; Blais 2000; Leighley and Nagler 2014; Blais and Daoust 2020) as well as in Indigenous communities (Fournier and Loewen 2011). We look at whether members of these groups perceive voting to be more difficult and which specific steps leading to voting are deemed to be particularly burdensome.

Finally, we address the "Does it Matter" question (see Chapter 5). Some people may find voting easy but abstain because they are not really interested or don't care. Others may find voting difficult but vote anyway, because they are strongly motivated to vote. We thus need to sort out how much of the abstention, among the whole electorate but also among different subgroups, is due to lack of motivation (people do not vote because they are not interested and/or they do not feel they have a civic duty) and how much is due to the perceived difficulty (people would like to vote but they are not able to because of the obstacles). The bottom line and toughest question is thus about the impact of burdens on turnout. How many more people would vote if the burden was nil (or minimal)? Which specific burdens are the most consequential for the decision to vote or abstain? Finally, for which groups is the burden most important, that is, in which groups would turnout increase the most in the absence of any obstacle?

The final chapter (Chapter 6) summarizes the main findings and makes suggestions about what could be done to make the act of voting as easy as possible for as many Canadians as possible.