1. Introduction

In March 2020, COVID-19 hit Canada, and over the past two years the pandemic has affected all aspects of people’s lives, from school closures to the implementation of mask mandates and quarantines, and the use of vaccine passports. The long-term consequences for democratic politics are unknown, yet Canada provides a fascinating case study to examine the impact of the pandemic on one core aspect of democracy: the administration of elections.

Since the start of the pandemic, there have been five provincial elections and one federal election. Election management bodies (EMBs) were forced to conduct these elections in a highly uncertain environment: it was unclear how citizens would feel about voting in person, leading many EMBs to offer new modes of voting or to simplify existing alternatives. COVID-19 rates across the country had the potential to change rapidly, and poll workers and electors were required to abide by public health measures, which evolved as the waves of the pandemic ebbed and flowed. Despite a relatively high level of satisfaction with the ways in which governments handled the pandemic early on (Harell et al. 2020), there have been increasing signs of dissatisfaction over time and some protests against restrictions, such as the so-called Freedom Convoy protests that occupied Ottawa and other cities across Canada in early 2022.

In this report, we seek to analyze how these elections were run and the impact this has had on electors’ perceptions of the electoral process. What can we learn from these pandemic elections about how electors respond to participating in democratic processes during a crisis? What impact did the choices made by EMBs have on the health of democratic politics in Canada? We first provide information about the context of each of the elections: what was the pandemic context in which they took place, and what measures did each EMB put in place to accommodate electors? We then turn to a detailed analysis of electors’ use of these accommodations, and how this impacted their experience and overall satisfaction with the process.

Our analyses draw upon a series of two-wave election surveys that were conducted by the Consortium on Electoral Democracy (C-Dem) using online samples of citizens. Public opinion surveys were run before and after each of the provincial elections (n=854-1,505 by province) as well as the 2021 Canadian Election Study (CES) (Stephenson et al. 2021a) The provincial studies were designed to mimic the content of the CES.^[1] By comparing experiences across these elections, we can draw lessons about the impact of election administration decisions on voter behaviour and satisfaction in Canada. Our focus is on how voters’ electoral experience changed during the pandemic and whether previous experience with a provincial pandemic election had consequences for voter behaviour in the 2021 federal election. This report concludes with a series of lessons learned and recommendations for the future of election administration in Canada.

2. Elections in Context

From 2020-2021, six elections took place in Canada: five provincial elections and a federal election. Each electoral management body (Elections New Brunswick, Elections Saskatchewan, Elections BC, Elections Newfoundland and Labrador, Elections Nova Scotia, Elections Canada) took a tailored approach to the election depending on the risk profile in the province, their resources and their level of preparedness. In some cases, such as Saskatchewan, the election was expected. In others, like New Brunswick, British Columbia and federally, it was not. Figure 1 shows the timing and campaign length for each election.

The pandemic context has varied widely across the country since March 2020, with some places experiencing huge spikes while others had relative calm.^[2] Figure 2 provides some insight into the prevalence of the virus during the time of each election, measured by the number of active cases per 100,000 people. In New Brunswick, Premier Blaine Higgs called a snap election in August 2020, the first COVID election in Canada. The prevalence of the disease was low in terms of cases throughout the course of the election period, and even declined (see Figure 2). However, according to the Stringency Index calculated by Breton et al. (2021),^[3] the restrictions put in place to mitigate the risk actually increased somewhat over the same period.

Elections were held in British Columbia and Saskatchewan shortly after, with October election dates in both locations. BC’s snap election was called by the NDP minority government shortly after the New Brunswick election, and it was followed closely by Saskatchewan’s scheduled election on October 26. In both cases, the elections took place when there were more documented cases of the virus circulating in the communities than had been the case in New Brunswick (Figure 2). Saskatchewan, in particular, saw a marked increase in cases between the time the writ was issued and election day (Figure 2). However, in both provinces the restrictions remained stable according to Breton et al.’s Stringency Index.

Newfoundland and Labrador, of course, had a very different experience. The snap election was called on January 15 to be held on February 13, 2021 in a context of very low case counts in the province. This all changed the week prior to the election (see the dramatic spike in Figure 2), which led to the chief electoral officer cancelling in-person voting on February 11 and a rise in pandemic restrictions. This change necessitated a shift to mail-only voting and several adjustments to when mail-in ballots had to be received, finally ending on March 25, 2021. Nova Scotia’s election, in contrast, took place in a period of relative calm, both in terms of the level of infection circulating and the stability of the restrictions. An election had been largely expected in the province, the only one without fixed election dates.

These five provincial elections preceded a fall federal election. On August 15, Prime Minister Justin Trudeau called a snap election to be held on September 21, 2021. While many observers expected an election call in the fall, given the minority government situation, the short five-week election campaign called at the end of summer required Elections Canada to mobilize quickly and provide electors with information about the options available. There was a clear increase in COVID-19 cases over the course of the campaign across the country, although the local circumstances that were likely most relevant for voters (including the restrictions in place) varied greatly.

Each EMB developed a series of health and safety provisions for their workers, their offices and polling stations. Common practices included using fewer poll workers, erecting plexiglass barriers, establishing personal protective equipment protocols, providing hand sanitizer for voters and workers, minimizing physical contact with ballots, requiring masks when appropriate, facilitating physical distancing as well as providing single-use pencils for marking ballots. In each situation, existing public health regulations were followed. For example, BC could not require masks but encouraged them. In the federal election, mask use varied by jurisdiction, but poll workers were required to wear them. Other public restrictions were important considerations for planning too. Elections BC notes in its report on the 2020 election that clarifications by the public health officer that size-gathering restrictions beyond distancing requirements would not apply to polling places made a difference for the sizes and types of locations that were sought for voting (Elections BC 2021).

Generally, each election took steps to ensure health and safety measures were followed while upholding the mandate of administering fair and accessible elections. Various ways of casting a ballot were used in each province (Figure 3). For example, Elections New Brunswick prioritized reducing lineups and congestion with “flatten the election curve” advertising, which encouraged people to vote early at advance polls or returning offices, or by mail (Elections New Brunswick 2021). In Saskatchewan, “Elections Saskatchewan was committed to providing a safe in-person voting experience but also needed to offer an option to vote remotely for those who did not feel comfortable voting in person” (Elections Saskatchewan 2021). Elections BC identified three specific challenges: ensuring no one was disenfranchised for lack of safe and accessible opportunities to vote; maintaining safety at all Elections BC locations, including polling stations; and ensuring capacity for mail voting was met (Elections BC, 2021). Elections Nova Scotia benefited from the experiences of other jurisdictions:

Due to the COVID-19 pandemic, Elections Nova Scotia included public health measures in all aspects of planning for the 41st provincial general election to ensure the safety of voters, election staff, candidates, and their campaign teams. Several voting options were provided and promoted to build voters’ confidence and to show they could vote safely either in person or by mail using a write-in ballot. Elections Nova Scotia learned from the experiences of other Canadian election management bodies that conducted elections during the pandemic and incorporated lessons learned from these jurisdictions. What remained paramount was to ensure the delivery of a safe, fair, inclusive and balanced election. (Elections Nova Scotia, 2022)

Communicating  plans to the public was also important. For example, Elections Saskatchewan had a strong media campaign that utilized newspapers, billboards and ­outdoor advertising, social media, radio and television. The message was about “voting safely,” ensuring that voters were aware of how to cast their ballot by mail. Each of the EMBs had its own materials and plans for publicizing the available voting options and safety measures taken in the polling stations.

3. The Voting experience across elections

Each election varied in terms of timing, the local pandemic context in which the election occurred and the previous experience electors had with various voting options. A key component — and challenge — for election management bodies was to ensure that electors could vote in a way that not only followed health guidelines put in place by the provincial and local authorities, but also allowed electors to feel safe participating in the process. As discussed, EMBs varied in how they dealt with this challenge, with some provinces encouraging mail-in ballots, some utilizing assisted telephone voting and all limiting exposure at the polling place through hand sanitation, social distancing and other public health measures.

Despite efforts to plan accordingly, the ability of electoral management bodies to predict how voters would respond to the pandemic context was largely unknown. While there was evidence prior to the pandemic of increasing use of advance voting options by electors (for example, in the 2019 federal election), election planning is based largely upon prior experience: looking at previous elections as well as long-term trends to estimate organizational needs.

Yet the pandemic had the potential to destabilize electors’ regular behaviour. Elections New Brunswick served as a first test case of pandemic voting; other EMBs were also able to closely observe each pandemic election as it unrolled. Which options would be popular with electors and in what proportions was largely unknown. Elections BC tried to get a sense by conducting surveys of how many mail-in ballots might be requested to help with their planning. Elections Canada also did public opinion polling in spring 2020 which led it to estimate demand for approximately 5 million mail ballots; it revised its estimates after the Saskatchewan and British Columbia elections to 4 million (Elections Canada 2020). In Newfoundland and Labrador, preparations were made for 60,000 special ballots, which seemed generous given that less than 10,000 were used in the 2019 election, but it ended up being less than half of the total number of ballots required when in-person voting was cancelled. The EMBs planned in accordance with the best information they had available to provide efficient voting experiences for electors as well as capacity for the postelection tabulation of votes with the technical and staff needs to conduct them. But the amount of uncertainty due to the pandemic was ever-present, and unpredictable.

In this section, we look at how electors navigated the options presented to them by EMBs. We examine to what extent specific measures put in place allowed vulnerable voters to participate in the election, despite the pandemic, and whether electors’ experiences varied with the different forms of voting.

3.1 Participating in pandemic elections

We begin by focusing on how electors engaged with the pandemic elections. First, we consider turnout. Figure 4 shows turnout rates for each election, compared to the previous election. Turnout in COVID elections declined everywhere but Nova Scotia and declined most in Newfoundland and Labrador, by over nine percentage points. That turnout was lower there is not surprising in light of the specific situation of the Newfoundland and Labrador election. However, the fact that turnout did not decline more precipitously in a decidedly riskier environment (higher levels of infection) demonstrates that voters were willing to make the effort to have their voices heard. The efforts of the EMBs to provide safe voting options paid off, though it is not clear what the participation rate would have been with no safety measures in place.

Second, there was a clear shift in voting mode across these elections, with a decline in election day voting and an increase in advance voting. Figure 5 shows how people chose to cast ballots compared to the previous contest. The data were taken from the various public reports of the EMBs.

In some provinces, more people voted prior to election day than cast their ballot the traditional way. This trend would be striking were it not for a recent trend towards advance voting across the country. This was particularly notable in Canada’s 2019 ­election when more than a quarter of electors cast their ballot before election day (Elections Canada 2019).

Finally, the third panel in Figure 5 demonstrates an increase in the use of special ballot provisions, which include mail voting, in all elections. There was approximately 6300 percent higher demand in BC (Elections BC). Mail voting stands out because it was not commonly used prior to the pandemic in general elections, and when breakdowns are provided, such as in Saskatchewan and British Columbia, the volume outpaces other special ballot provisions. In Saskatchewan, only about 5 percent of special ballots were not mailed; in British Columbia the percentage of mail ballots was 31.4 percent of all ballots cast. However, it is important to recognize that BC had experience with mail ballots from the 2018 referendum and the media coverage of mail voting in the American election may have made voters more familiar with the concept of mail ballots. Also striking is that the amount of special ballot voting was lowest for the federal election, at 6 percent. While this is almost a 100-percent increase from 2019, it pales in comparison to the provincial elections. In total, Elections Canada issued almost 1.3 million special ballots in 2021, of which 84 percent were eventually counted toward the outcome (Elections Canada). This is far fewer than its initial estimate for public demand.

These numbers provide an overview of how voters voted, yet statistics cannot provide information about how the decision to vote, and in which way, was made. For this, we can mobilize the unique pre-election/post-election panel studies that were conducted during each of these elections by the Consortium on Electoral Democracy. As noted, one of the challenges of planning during COVID elections was the number of unknowns in expected behaviour. In Figure 6, we explore how electors planned to vote compared to how they ended up voting. What is clear is that most people ended up voting the way they had intended, except in Newfoundland and Labrador where in-person voting was cancelled. In-person and advance polling were the most common choices, except in BC where there was a large substantial portion of the population who intended to — and did — vote by mail.

Shifts in voting mode are important to understand because in uncertain elections, how people intend to vote is a key piece of information for planning purposes. Figure 6 makes clear that despite the changing context during the pandemic, most people voted the way they planned to vote. And, importantly, the largest shifts appear to be between in-person voting options (on election day versus advance polling).

When we analyze the profiles of different types of voters, clear patterns emerge of who chose to use alternative vote options. Combining the surveys across the five provincial elections into a single dataset, we ran a single model to look for patterns in mode choice. Comparing advance polling to election-day voters in the provincial elections, advance poll voters tended to be older, were more likely to be women, and were more likely to report that they or someone in their family faced heightened health risks because of COVID.^[4] We find no evidence that university-educated voters were more likely to choose advance polling over in-person voting on election day. In contrast, university-educated electors were more likely to vote by mail or at a returning office, compared to in-person voting. Older voters more likely to choose these options as well. Those over 65 were about 20 percentage points less likely to vote in person on election day. Those with heightened health risks (either to themselves or family), perhaps not surprisingly, were less likely to pick in-person voting on election day, and more likely to use mail voting.

Federally, we see many of these patterns re-emerge. Older voters, especially those over 65, were less likely to vote on election day and preferred advance polling stations. We also see a clear educational difference, with university-educated electors taking advantage of alternative voting methods. The perceived COVID-19 risk question was not asked in the CES, but respondents were asked to assess their overall physical health. Those who reported worse overall physical health were more likely to vote by mail, consistent with the idea that risk assessment led voters to select lower-risk modes of voting. Finally, there are also interesting trends across provinces. Table 1 provides the predicted probabilities of using each vote mode by province (Nunavut, Northwest Territories and Yukon are excluded due to small sample sizes). What is interesting to note is that in the three provinces that had substantial numbers of mail voters in the provincial pandemic elections (BC, Saskatchewan, and Newfoundland and Labrador), we see a higher tendency to vote by mail in the federal election. Although the use of special ballots was also high in Nova Scotia in Figure 5, it was largely due to voting at the returning office rather than using write-in ballots (Elections Nova Scotia, 2022). From our analyses, then, it seems that experience with voting by mail at the provincial level spilled over to influence the way that people voted in the federal election.

3.2 Satisfaction with democracy in a pandemic

More important than how people voted is whether voters were satisfied with how the election process was run. This is important to gauge confidence in democracy and the democratic process and it can also have an impact on participation. The pressure on the EMBs during highly unusual times to fulfill their mandates and administer fair elections was heightened especially given the stakes if they were unable — or perceived to be unable — to fulfill that responsibility.

In Figure 7, we show responses to a question about overall satisfaction with how the relevant EMB ran the election, 0 — “not at all” to 1 — “very satisfied”. In general, people reported being satisfied with how the election was run, and there is very little variation across four of the five provincial elections. Newfoundland and Labrador, unsurprisingly, is a clear outlier, with the average satisfaction response being below the midpoint, suggesting an overall dissatisfaction with how Elections Newfoundland and Labrador ran the election. The provincial election was the only election that was directly disrupted due to changing pandemic circumstances, with in-person voting on election day being cancelled on short notice, forcing electors to vote by mail and requiring multiple logistical changes. Perhaps what is more interesting is that there was still a middling level of satisfaction in Newfoundland and Labrador after that election. Of all the elections held during the pandemic, Elections Canada received the highest mark for overall satisfaction for how the election was run.

In Figure 8, we show respondents’ overall satisfaction with democracy and how it was affected by the election in their province. Here, we take advantage of the two-wave design of the election studies to assess satisfaction before and after the election. Once again, overall satisfaction is consistently above the midpoint. Interestingly, satisfaction with democracy in each province tends, if anything, to increase after the election (looking only at respondents who provided responses to both the campaign period and post-election surveys). The exception among provinces is again Newfoundland and Labrador. During the election campaign, respondents were relatively satisfied with democracy in their province, and their satisfaction rates look similar to other provincial contexts. However, the particularities of Newfoundland and Labrador’s election led to an erosion of satisfaction with democracy in that province. Post-election, it had the lowest level of satisfaction compared to the other four provinces that also experienced pandemic elections. However, a similar dynamic is also observed at the federal level both in 2021 and, when looking back, in 2019 as well (2019 results not shown). Clearly, satisfaction with democracy draws upon more than just how the election was run. It likely captures people’s satisfaction with the results based on their expectations of how their preferred party would do as well as more general sentiments about democracy (Blais and Gélineau 2007).

Who was most satisfied with how the election was run and with democracy in their province? Again, we can pool the provincial election studies together to examine who is most likely to be satisfied and consider whether the way that someone cast a ballot had an impact on their attitudes. We find very little evidence that the way people voted influenced their satisfaction with the management of the election, nor did it impact how they felt about the state of democracy. The one exception is the (relatively few) voters who used the returning office, which shows weak, positive effects on satisfaction (both the running of the election and democracy). Potentially this relates to the influence of engaging more directly with the EMB, but we cannot tease this out with the data.

We find some demographic differences with how the election was run; older, more educated respondents tend to be more satisfied. Interestingly, we also find that those who reported a higher COVID risk were somewhat more likely to be satisfied with how their provincial EMB ran the election. However, turning to the model for overall satisfaction with democracy after the campaign, we find no effect of education or risk, though we do again see that older voters are more satisfied in general. Finally, we can directly test the election outcome on satisfaction by looking at electors who voted for the winning party. Not surprisingly, electoral winners were on average more satisfied with both how the election was run by the EMB and with democracy in their province.^[5]

When we look at overall satisfaction with how Elections Canada ran the election, we find similar effects as for the provinces. Older, university-educated voters and those who voted for the winner tended to be more satisfied with how the election was run. We find no differences by voting mode. Interestingly, we also see no differences based on the province of residence (except for PEI, where there was slightly higher satisfaction). Thus, there is no evidence that provinces that had experienced a pandemic election were more (or less) satisfied with the performance of Elections Canada during the federal contest. Elections Canada is held in high esteem by Canadians, and the pandemic has not changed that estimation.

In general, the efforts of the EMBs resulted in a high level of confidence among voters that voting was safe. Over 95 percent of those who voted in person in each province reported that they somewhat or strongly agreed that they felt safe voting in person, and strong majorities across four of the five provinces agreed that their EMB took the necessary precautions (75 percent to 95 percent of respondents). Here, the exception was again Newfoundland and Labrador, with only a slim majority (50.5 percent) agreeing that Elections Newfoundland and Labrador took the appropriate safety precautions.

4. Findings and recommendations

In the face of a world-wide pandemic, election management bodies were tasked with running free and fair elections in an environment marred by unknowns. How would citizens feel about waiting in long lines surrounded by others? Would a largely older workforce staffing the polls be available and willing to work? Would electors want alternative ways to vote to minimize their risk of infection?

Our analysis shows that despite variation across election administrations and the context in which elections took place, the responses were somewhat similar across jurisdictions. Basic health and safety measures were commonplace, such as the use of masks by staff (and, when required, by electors), hand sanitizers and physical distancing. These measures allowed electors to feel safe voting in person, while information campaigns rolled out on how to vote at advance polls or by mail to service those who felt safer staying away from the polls on election day.

Our analysis also demonstrates that voters were generally quite comfortable using in-person voting options (either in advance or on election day) across jurisdictions when it was an option. Mail voting ­— a highly politicized alternative in the recent US election ­— was used substantially more during the pandemic than in the past, particularly in jurisdictions like BC that had previous experience with vote-by-mail. Regardless of the method used, overall levels of satisfaction with the process were high.

An important caveat to our results is that it is difficult to know the opinions of non-voters. Public opinion surveys like those conducted by C-Dem are plagued by overreports of voting and non-response by uninterested citizens, many of whom are also non-voters. For that reason, it is difficult to use the data on hand to provide strong statements about how non-voters experienced the election campaign and whether their attitudes about the pandemic and/or the EMBs’ preparations related to the pandemic contributed to their lack of participation. Specific research into the reasons for non-voting, in extreme as well as normal elections, is of high importance and, unfortunately, in short supply. Nonetheless, we believe our analysis still has much to add to our understanding of Canadian elections during the COVID-19 pandemic.

4.1 Key findings

A few key points stand out from our study.:

• Turnout levels for the pandemic elections in Canada declined everywhere but in Nova Scotia, and it declined most in Newfoundland and Labrador, by over nine percentage points.
• The pandemic elections brought a decline in election day voting and an increase in advance voting. In some provinces, more people voted prior to election day than on the actual day.
• Most people ended up voting the way they intended to, except in Newfoundland and Labrador where in-person voting was cancelled.
• Most people reported being satisfied with how the election was run after the election. We found very little variation across four of the five provincial elections.
• Satisfaction with the performance of Elections Canada during the federal contest was not different in provinces that had experienced a pandemic election.

4.2 Recommendations

Our analysis also reveals an important lesson. The provincial election in Newfoundland and Labrador shows the importance of process for maintaining high levels of elector confidence. While we saw that, across elections, those voting for the winning party tended to be happiest, the drop in satisfaction with both how the election was run and with democracy in Newfoundland suggests that the shift to mail-only voting during the campaign led to a drop in public confidence in both the process and in the election management body. Indeed, when asked how much confidence people had in the EMB after the election, only 6.6 percent of respondents reported a great deal of confidence. The amount was at least four-fold higher in every other jurisdiction, with about three-quarters having either quite a lot or a great deal of confidence, compared to just 42 percent in Newfoundland and Labrador.

What lessons can we learn from this analysis? We see two recommendations emerging from our analysis. First, it is clear that EMBs cannot plan for every eventuality, especially in a crisis. Yet one of the key takeaways from this analysis is that electors can handle new ways of voting and new procedures at the ballot box. Providing options in an uncertain situation allows voters who may be more at risk to choose different ways of participating. This, of course, comes with organizational challenges because planning requires estimates of the uptake of various options: guessing the rate of in-person voting incorrectly may mean long lines that turn some electors away, as were observed during the federal election. A huge, unexpected uptake in mail voting (as became a necessity in Newfoundland and Labrador) meant that capacity to get ballots into electors’ hands in a timely fashion was a challenge. And unexpected increases in voting by mail means counting ballots may take longer than electors are accustomed to, or even that not enough ballots are available to send out. Nonetheless, thinking creatively about how voters can cast their ballots in extreme situations is an important step toward ensuring that voters feel their voices are important and desired in the electoral process.

Second, clear communication with voters about what to expect appears to be particularly important. In the case of Newfoundland and Labrador, one of the causes of the hit in public opinion was likely that changes were made during the campaign, and in fact right before election day. Such changes, when unexpected, can lead electors to blame key actors, regardless of whether the situation was in their control. It also risks the impression that the management of elections in this country has become politicized, if one party is seen to be advantaged above others when rules change. This, in particular, is something that all of the non-partisan election management bodies would be well advised to avoid.

Elections Ontario, in preparation for the anticipated June 2022 election in that province, seems to have taken these lessons to heart. A CBC news story on April 10 reported that the EMB will increase the number of advance voting days to 10 and has created an online portal to simplify requests for mail-in ballots. The story quotes the Chief Electoral Officer of Ontario saying, “I think one of the things that we’re really trying to do this election is really look at flattening the curve of the vote” (Jones 2022). Those who vote in person will find similar health and safety measures as those used in the federal election.

We also see some larger trends in the data that may affect future elections. We have seen increases over time in the use of advance polls by voters, and this trend continued during the pandemic elections. As voting becomes something that happens over several days or even weeks rather than on a single day, how does this affect the ebb and flow of campaigns and the activities of candidates? Are voters making up their minds earlier? Are they paying attention to campaign events? Are they satisfied and able to make decisions with minimal information? Campaigns are designed to both mobilize and inform voters, but voters who decide to vote early may be making their decision based on information that is different from what they would have had at the end of the campaign. If a scandal breaks mid-campaign, voters cannot take their ballot back and change their mind. What is more, earlier voting may shift campaign activity earlier and create longer, more permanent campaign cycles for parties.

Furthermore, many more Canadians than ever before have now voted by mail. As we saw in the federal election, such experience at one level may well spill over into other jurisdictions so future elections in Canada may well see more and more electors comfortable with this option. Mail-in voting is not without its challenges. When this option requires a special ballot that does not include a list of candidates within a voter’s riding, this can create barriers to filling in the correct information. It can also slow the announcement of results to the public after the election. How much voting by mail is encouraged in the future, and steps taken to make the process as easy as possible for electors, will have important implications for how Canadians engage in the electoral process in the future. Ensuring that ballots are as simple and accessible to fill out as possible avoids voter frustration and ensures that electors can cast their ballots as they intend.

There are also security considerations with mail voting that may need to be considered if used more broadly outside of a crisis. When asked during the 2021 federal election if they thought voting by mail was equally as trustworthy as voting in person, Canadians were divided. Only a slim majority (54 percent) somewhat or strongly agreed with this statement, though not surprisingly, agreement was highest in British Columbia, which had the most previous mail-voting experience (in both the pandemic and the 2018 referendum). Mail voting tended to be used more by certain partisans than others. The risk that this method becomes politicized, as has been witnessed in the United States, should be closely monitored to ensure that the legitimacy of the voting process is not undermined by short-term political discourse. From our analysis, it seems that no matter which voting options are provided, when the process runs smoothly, is clearly communicated, and electors are given a reasonable set of choices, confidence in electoral democracy in Canada is likely to remain high.

^[1] Provincial surveys were fielded two weeks prior to election day in each province, and then all respondents were re-contacted in the two weeks following the election for a follow-up survey. The 2021 CES was run throughout the campaign period and then immediately following the election. Each survey was approximately 20 minutes and was self-administered online via the Qualtrics platform. Respondents were recruited by online panel providers (Léger for SK, BC and Canada, and Narrative for NB, NL and NS). Each study was quota-sampled based on the distribution of key demographic variables. Full technical reports for each survey are available from the authors. For a description of the CES approach, see Stephenson et al. 2021b.

^[2] For a detailed case study of each of the first four provincial elections, see Garnett et al. (2021).

^[3] Breton et al.’s (2021) Stringency Index takes a variety of different public health measures into consideration, including gathering restrictions; mask mandates; whether schools are open and whether masks are required in schools; restrictions in care homes, dining and restaurants, personal care services, cultural attractions, and intra- and inter-provincial travel; curfews; and vaccine mandates (https://centre.irpp.org/2021/07/covid-19-stringency-index-new-codebook/).

^[4] This analysis is based on a multinomial logit regression with controls for provinces. Note that “other” voting methods were removed from the model due to very few observations.

^[5] Note, the weighted pooled provincial models included controls for province as well.

Berdahl, Loleen, Laura B. Stephenson and Allison Harell. 2020. 2020 Saskatchewan Election Study. Dataset.

Bittner, Amanda, J. Scott Matthews, Laura B. Stephenson and Allison Harell. 2021. 2021 Newfoundland and Labrador Election Study. Dataset.

Blais, André, and François Gélineau. 2007. “Winning, Losing and Satisfaction with Democracy.” Political Studies 55 (2): 425-441. doi:10.1111/j.1467-9248.2007.00659.

Breton, Charles, Ji Yoon Han, Mohy-Dean Tabbara and Paisley Sim. 2021. COVID-19 Canadian Provinces Measures Dataset. Center of Excellence on the Canadian Federation. https://centre.irpp.org/data/covid-19-provincial-policies/

Elections British Columbia. 2021. Provincial General Election: Report of the Chief Electoral Officer. (Victoria, BC: Elections British Columbia).

Elections Canada. 2022. Report of the 44^th General Election of September 20, 2021. (Ottawa, ON: Elections Canada).

———. 2019. The 2019 Federal election by the numbers. https://www.elections.ca/content.aspx?section=med&dir=bkg&document=num&lang=e

Elections New Brunswick. 2021. Fortieth General Provincial Election, September 14, 2020: Report of the Chief Electoral Officer. (Fredericton, NB: Elections New Brunswick).

Elections Newfoundland and Labrador. 2021. 2021 Provincial Election Report. (St. John’s, NL: Elections Newfoundland and Labrador).

Elections Nova Scotia. 2022. Statement of Votes and Statistics, Volume 1: 41^st Provincial General Election, August 17, 2021. (Halifax, NS: Elections Nova Scotia).

Elections Saskatchewan. 2021. A Report on the Twenty-Ninth General Election, Volume II: Administrative Review. (Regina, SK: Elections Saskatchewan).

Everitt, Joanna, Laura B. Stephenson and Allison Harell. 2020. 2020 New Brunswick Election Study. Dataset.

Garnett, Holly Ann, Jean-Nicolas Bordeleau, Allison Harell and Laura Stephenson. 2021. “Canadian Provincial Elections During the COVID-19 Pandemic.” Electoral Integrity Project Report, Case Study 17 (December). https://www.idea.int/sites/default/files/multimediareports/canadian-provincial-elections-during-the-covid-19-pandemic-en.pdf

Harell, Allison, Laura Stephenson, Daniel Rubenson and Peter Loewen. 2020. “How Canada’s Pandemic Response is Shifting Political Views.“ Policy Options. Montreal: IRPP. https://policyoptions.irpp.org/magazines/april-2020/how-canadas-pandemic-response-is-shifting-political-views/

Jones, Allison. 2022. “Elections Ontario offering more options for voting other than in person on June 2.” CBC News, April 10. https://www.cbc.ca/news/canada/toronto/elections-ontario-voting-june-1.6415277.

Pickup, Mark, Laura B. Stephenson and Allison Harell. 2020. 2020 British Columbia Election Study. Dataset.

Roy, Jason, Laura B. Stephenson and Allison Harell. 2021. 2021 Nova Scotia Election Study. Dataset.

Stephenson, Laura B., Allison Harell, Daniel Rubenson and Peter Loewen. 2021a. 2021 Canadian Election Study. Dataset.

———. 2021b. Measuring Preferences and Behaviours in the 2019 Canadian Election Study. Canadian Journal of Political Science. 54 (1): 118-124. https://doi.org/10.1017/S0008423920001006

ABOUT THIS REPORT

This report was published as part of the research of the Centre of Excellence on the Canadian Federation, under the direction of Charles Breton and assisted by Ji Yoon Han. The manuscript was copy-edited by Shannon Sampert and proofread by Rosanna Tamburri, editorial coordination was by Étienne Tremblay, production was by Chantal Létourneau and art direction was by Anne Tremblay.

Allison Harell is a professor of political science and a fellow at the Canadian Institute for Advanced Research (CIFAR). Her research focuses on how social diversity affects the political world, especially the ways in which prejudice influences public opinion formation. She is co-director of the Consortium on Electoral Democracy.

Laura B. Stephenson is a professor in the Department of Political Science at Western University. Her research focuses on political behaviour and the effects of institutions, including the issues of partisanship, voting, and the relationships among voters, parties and electoral systems, from a Canadian and comparative perspective. She is co-director of the Consortium on Electoral Democracy.

This report was co-written equally by the authors. Their names appear in alphabetical order.

To cite this document:
Harell, Allison, and Laura B. Stephenson. 2022. Elections During a Health Crisis: Voter Involvement and Satisfaction across Pandemic Elections in Canada, IRPP Report (May). Montreal: Institute for Research on Public Policy.

ACKNOWLEDGEMENTS

This report draws on research conducted by the Consortium on Electoral Democracy (C-Dem). The data collection was supported by a SSHRC-funded partnership grant (Grant # 895-2019-1022) and a SSHRC Insight grant (Grant # 435-2018-1208), and we are thankful for this support. The views and analysis presented in this paper are solely those of the authors.

---

The opinions expressed in this essay are those of the authors and do not necessarily reflect the views of the IRPP or its Board of Directors.

If you have questions about our publications, please contact irpp@irpp.org. If you would like to subscribe to our newsletter, IRPP News, please go to our website, at irpp.org.

Illustrator: Luc Melanson