2020 Poll Worker Recruitment Survey: Results and Preliminary Analysis

PACE Research, January 2021

Introduction

From November 2020 to January 2021, Elections Canada (EC) conducted an online survey, collecting data from 72,198 poll workers in the 43rd federal general election (GE) about whether or not they would be interested in working again during the 44th GE.^{Footnote 1} The primary objective of the survey was to provide returning officers with a refreshed inventory of potentially interested poll workers in their electoral district whom they could contact for recruitment purposes in the event of an election. The survey also provided information about the safety measures EC has adopted at the polls to ensure the health and safety of poll workers during the COVID-19 pandemic, so respondents could give an informed response. For research purposes, the survey included a small number of questions to measure poll workers' level of concern about the COVID-19 pandemic.

This research focus presents the selected survey findings and a preliminary analysis of the impact of poll workers' concerns about COVID-19and their age on their level of interest in working the next GE. For completeness, a set of tables in the appendix provide a fuller set of survey results, broken down by region of Canada.

Subsequent analysis will examine the relative impact of multiple factors in combination, as well as the impact that a change in concerns about COVID-19 might be expected to have on poll workers' interest in working the next election.

Considerations and Limitations

Considering that the primary purpose of the survey was recruitment, caution is advised when interpreting the results, since they are provided unweighted and survey respondents may not be entirely representative of the population of all poll workers from the 43rd GE. In particular, the reader should note that:

The proportion of respondents who are interested in working the next GE is expected to be higher than it would be if all poll workers from the 43rd GE had responded to the survey. Those who were interested in working had a personal incentive to complete the survey in order to be contacted about potential work in the next election, while those who were not interested did not have a similar incentive to participate.

In addition, the changing context of the COVID-19 pandemic requires that the results be considered in relation to the time period of the data collection:

The vast majority of the data were collected in December 2020, before the holiday season. Respondents' concerns about COVID-19 and interest in working the next election may have evolved since the time they were surveyed, depending on various developments in different regions: for example, changing incidences of COVID-19 cases, new government measures, availability of vaccines, etc.

Findings

Interest in working the next GE was very high among the previous poll workers who responded; issues with the hours or working conditions were ahead of COVID-19 as the main reason for not being interested.

With the caveat that those interested in working the next election had more incentive to respond to the survey, interest in working in the 44th GE was very high among poll workers from the 43rd GE: Overall, 92% (66,038) of respondents said they are interested in working the next GE, including 77% (55,935) who said they are very interested. Only 4% of respondents said they were not at all interested.

Among the 8% (5,851) of respondents who were not interested in or not sure about working the next GE, the main reason for their reluctance was the hours or working conditions, mentioned by one in four (24%), followed by one in five (21%) who cited concerns about COVID-19.

Figure 1: GE43 poll worker interest in returning to work for EC in the next GE — Q. How interested would you be in returning to work for Elections Canada during the next general election? Base: all respondents; n=72,198. Source: 2020 Poll Worker Recruitment Survey.

Figure 2: Main reason for being not interested in/not sure about working the next GE — Q. What is the main reason you are (not interested in/not sure about) working in the next general election? (One response allowed.) Base: those who are not interested in or not sure about working in the next GE; n=5,851. Source: 2020 Poll Worker Recruitment Survey.

More personal concerns about COVID-19 have a small but meaningful impact on interest in working.

When asked various questions about their level of concern regarding COVID-19, at least two-thirds of respondents said they were at least somewhat concerned about the virus in each case. However, people were more often very concerned about others, such as friends or family (40%) or people in their community (34%), becoming sick from COVID-19, rather than about themselves (24%).

Figure 3: GE 43 poll worker general concerns about <span class= — Q. In general, how concerned are you about each of the following? Base: all respondents; n=70,203. Source: 2020 Poll Worker Recruitment Survey.

GE 43 poll worker general concerns about COVID-19
Level of concern	Friends or family becoming sick	People in your community becoming sick	Personally becoming sick from COVID-19
Very concerned	40%	34%	24%
Somewhat concerned	39%	46%	43%
Not that concerned	13%	12%	22%
Not at all concerned	5%	3%	8%

Higher levels of concern about COVID-19 tended to be linked to a small but still meaningful decrease in interest in working the next GE—but the decrease in interest tended to be larger when concerns were more personal:

Those who were very concerned about becoming sick from COVID-19 themselves were less likely to be very interested in working the next GE (74%), compared with those who were somewhat concerned or less (79%). ^{Footnote 2}
Those who were very concerned about friends or family becoming sick were also less likely to be very interested in working the next GE (75%), compared with those who were somewhat concerned or less (80%).
In contrast, there was no significant difference in being very interested in working the next GE between those who were very concerned about people in their community becoming sick (77%) and those who were somewhat concerned or less (78%).

Being at high risk for COVID-19, or having an immediate contact who is, had the greatest impact on interest in working the next GE.

About a quarter (24%) of respondents said they consider themselves or someone in their immediate social circle to be at high risk for COVID-19. This factor had the single greatest impact on respondents' interest in working the next GE:

Those who consider themselves or an immediate contact to be at high risk for COVID-19 were much less likely to be very interested in working the next GE (65%), compared with those who do not (83%).
Furthermore, those who consider themselves or an immediate contact to be at high risk for COVID-19 were more likely to be not all interested in working (8%) compared to those who do not (2%).

Relatively small decreases in interest in working the 44th GE among older respondents are meaningful, given the high median age of poll workers from the 43rd GE.

Similar to the profile of election officers from the Survey of Election Officers for the 43rd GE, respondents to the survey on interest in working in the 44th GE were much more likely to be women and/or older in age disproportionate to the Canadian population:

More than two-thirds (69%) of respondents were women.
The median age of respondents was 60 years, with nearly a third (32%) aged 65 or older.
Only about a quarter (24%) of respondents were between the ages of 16 and 44, inclusive.

Figure 4: Respondents by age and gender — Base: all respondents; n=69,933. Source: 2020 Poll Worker Recruitment Survey.

There were no significant differences between men and women in their levels of concern about COVID-19 or their interest or disinterest in working the next GE.

However, respondents were more likely to say they are very concerned about personally becoming sick from COVID-19 when they were aged 65 to 74 years (28%) or 75 or older (33%), compared with all younger age groups (a range of 16% to 22%). They were also more likely to say that they consider themselves or someone in their immediate social circle to be at high risk for COVID-19 (27% of those aged 65 to 74 and 30% of those aged 75 or older, compared with a range of 21% to 25% among younger age groups).

Recalling that these two factors were associated with somewhat lower interest in working the next GE, it is notable that interest in working the next GE decreased after age 64:

Respondents aged 55 to 64 were more likely to be interested (93%) or very interested (81%) in working the next election than those aged 65 to 74 (90% interested, 77% very interested) and 75 or older (88% interested, 73% very interested).
Respondents aged 16 to 24 and 25 to 34 were equally likely to say they were very interested (76%) as those aged 65 to 74 (77%); however, they were more likely to say they were at least somewhat interested (95% and 93% compared with 90%, respectively).

Figure 5: Interest of GE 43 poll workers in working the next GE, by age — Base: all respondents; n=69,933. Source: 2020 Poll Worker Recruitment Survey.

Interest of GE 43 poll workers in working the next GE, by age
	16 to 24 years	25 to 34 years	35 to 44 years	45 to 54 years	55 to 64 years	65 to 74 years	75 years or more
Very interested	76%	76%	80%	83%	81%	77%	73%
Somewhat interested	19%	17%	14%	12%	12%	13%	15%
Total interested	95%	93%	94%	93%	93%	90%	88%

These relatively small differences in interest by age could be expected to have meaningful impacts on recruitment levels, given the number of poll workers required and the past tendency of poll workers to be disproportionately older. In ordinary times, increasing age would itself already be a factor in a person deciding not to work another election—and it is worth mentioning that among respondents aged 75 or older who said they did not want to work again, they most often said they were simply not interested in working another election as the reason (25%, compared with 21% who cited concerns about COVID-19). Still, the impact of concerns about the COVID-19 pandemic on interest in working an election seem likely to exacerbate existing challenges associated with relying disproportionately on recruiting older persons to work the polls.

Methodological Notes

The 2020 Poll Worker Recruitment Survey was conducted by the Research Division of the Public Affairs and Civic Education branch, in collaboration with and on behalf of the Field Governance and Operational Readiness branch of the Electoral Events and Innovation sector.

The primary objective of the survey was to develop a refreshed inventory of potential poll workers that could be shared with returning officers to directly support their recruitment activities. Given this objective, an attempt was made to invite all poll workers from the 43rd GE to participate, provided they had consented to be contacted by EC via email and had a working email address on record. Fielding of the survey was conducted on a rolling basis across all provinces and territories from November 26, 2020, to January 13, 2021. All regions of Canada had received the initial invitations by December 9; all final reminders were sent by December 15.

Out of a total list of 208,997 poll workers for the 43rd GE, 118,125 (56.5%) had what were assumed to be working email addresses and were therefore invited via email to complete a five-minute online questionnaire hosted on the Voxco survey platform. Emails were successfully sent without bouncing to 105,360 (89.2%) of these addresses.

Ultimately, 72,198 of invited poll workers (61.1%) responded to at least the question about being interested in working the GE, while 70,364 (59.6%) completed the entire survey.

No random procedures were used to select survey participants, so the results have no margins of sampling error. However, there are two known and expected non-random sources of error that may cause bias and need to be considered when interpreting the results from respondents as representative of the population of poll workers from the 43rd GE:

Coverage bias: The quantity and quality of available email addresses was not uniform across all electoral districts, with districts having varying proportions of poll workers who provided their email address and varying data quality for the emails that were collected. One known source of coverage error is that no email addresses were available for one of the 338 electoral districts. The extent of coverage error for the remaining 337 electoral districts that had at least some email addresses has not yet been evaluated. It is possible that these gaps in coverage could bias the aggregate results, since regions that have better coverage would have more relative weight than regions with poor coverage. That said, these errors may not significantly bias results if cases are effectively missing at random at the aggregate level—for example, results for the missing electoral district could be made up for by responses from nearby electoral districts that have good-quality coverage, assuming the views and characteristics of poll workers are similar across that neighbourhood. It would be possible to correct for coverage error by weighting the data that adjusts for uneven coverage by electoral district; however, for the sake of timeliness, the results have been provided unweighted.
Non-response bias: A degree of bias is expected in favour of those who are more interested in working the next GE, since respondents were told their information would be shared with their local Elections Canada office so that they could be contacted about potential work in a future election. They therefore had an incentive to participate. As a corollary, those with less interest in working the GE had no similar incentive and are therefore expected to have participated less often. As a consequence, the proportions of results for interest in working the GE are in particular considered to represent survey respondents only—they are not generalizable to the broader population of poll workers for the 43rd GE.

Appendix: Survey Results by Region of Canada

Figure 1: Interest in working in the next GE
Figure 1: Interest in working in the next GE	Total (n ≤ 72,198)	Region of Canada
Figure 1: Interest in working in the next GE	Total (n ≤ 72,198)	Atlantic provinces (n ≤ 4,974)	Quebec (n ≤ 14,309)	Ontario (n ≤ 25,896)	Prairie provinces (n ≤ 5,341)	Alberta (n ≤ 10,710)	British Columbia (n ≤ 10,740)	Territories (n ≤ 228)
How interested would you be in returning to work for Elections Canada during the next general election? (n = 72,198)
Very interested	77%	77%	79%	79%	71%	80%	72%	68%
Somewhat interested	14%	14%	12%	14%	18%	13%	18%	18%
Not that interested	2%	2%	2%	2%	2%	2%	3%	3%
Not at all interested	4%	4%	5%	3%	6%	3%	5%	6%
Not sure	2%	3%	3%	2%	3%	2%	2%	4%
What is the main reason you are not interested in (or not sure about) working in the next general election? (n = 5,851)
The hours or working conditions	24%	19%	20%	25%	22%	22%	33%	28%
Concerns about COVID-19	21%	22%	27%	23%	17%	18%	16%	17%
Not interested in working another election	15%	19%	15%	13%	22%	16%	13%	10%
Not available to work another election	14%	13%	15%	14%	10%	16%	14%	17%
Pay reasons	8%	8%	5%	9%	7%	11%	8%	3%
Not sure when the election will be	5%	5%	6%	4%	6%	6%	4%	14%
Other	5%	7%	3%	5%	8%	5%	7%	3%
No particular reason	4%	5%	6%	3%	5%	3%	3%	3%
Which days would you be interested in working as a poll worker? (n = 65,851)
Polling day	13%	14%	13%	13%	16%	15%	12%	13%
Advance polling days	5%	4%	5%	4%	5%	4%	7%	3%
Both polling day and advance polling days	71%	71%	71%	73%	67%	71%	68%	67%
No preference	9%	10%	10%	8%	10%	8%	10%	14%
Don't know	2%	1%	1%	2%	2%	1%	3%	3%
Which positions would you be interested in working? Select all that apply. (n = 65,606)
Deputy returning officer	50%	46%	62%	49%	46%	50%	44%	40%
Central poll supervisor	30%	32%	29%	32%	30%	27%	29%	28%
Registration officer	35%	34%	30%	37%	34%	34%	40%	34%
Information officer	27%	27%	22%	28%	28%	26%	30%	22%
Other (e.g. poll clerk, office worker, etc.)	8%	3%	3%	10%	13%	11%	9%	9%
No preference	18%	24%	14%	17%	19%	21%	20%	29%

Figure 2: Concerns about COVID-19
Figure 2: Concerns about COVID-19	Total (n ≤ 72,198)	Region of Canada
Figure 2: Concerns about COVID-19	Total (n ≤ 72,198)	Atlantic provinces (n ≤ 4,974)	Quebec (n ≤ 14,309)	Ontario (n ≤ 25,896)	Prairie provinces (n ≤ 5,341)	Alberta (n ≤ 10,710)	British Columbia (n ≤ 10,740)	Territories (n ≤ 228)
How concerned are you about… personally becoming sick from COVID-19? (n = 70,203)
Very concerned	24%	30%	11%	28%	24%	24%	28%	22%
Somewhat concerned	43%	46%	38%	43%	46%	44%	45%	41%
Not that concerned	22%	18%	33%	19%	20%	19%	19%	23%
Not at all concerned	8%	4%	11%	7%	6%	9%	6%	9%
Don't know	1%	1%	3%	1%	1%	1%	1%	2%
How concerned are you about… friends or family becoming sick? (n = 70,203)
Very concerned	40%	48%	23%	45%	43%	41%	46%	41%
Somewhat concerned	39%	40%	40%	38%	41%	40%	38%	36%
Not that concerned	13%	9%	23%	10%	10%	11%	10%	13%
Not at all concerned	5%	2%	8%	4%	4%	5%	3%	5%
Don't know	1%	1%	3%	1%	1%	1%	1%	2%
How concerned are you about… people in your community becoming sick? (n = 70,203)
Very concerned	34%	43%	17%	39%	37%	34%	41%	41%
Somewhat concerned	46%	45%	46%	45%	48%	47%	46%	41%
Not that concerned	12%	8%	21%	9%	9%	11%	8%	9%
Not at all concerned	3%	1%	7%	3%	2%	3%	2%	6%
Don't know	2%	1%	6%	2%	1%	2%	1%	2%
Do you consider yourself or someone in your immediate social circle to be at high risk for COVID-19? (n = 70,102)
Yes	24%	24%	21%	23%	28%	25%	25%	27%
No	61%	63%	61%	63%	56%	61%	61%	60%
Don't know	11%	10%	15%	11%	12%	10%	11%	11%

Figure 3: Respondent characteristics
Figure 3: Respondent characteristics	Total (n ≤ 72,198)	Region of Canada
Figure 3: Respondent characteristics	Total (n ≤ 72,198)	Atlantic provinces (n ≤ 4,974)	Quebec (n ≤ 14,309)	Ontario (n ≤ 25,896)	Prairie provinces (n ≤ 5,341)	Alberta (n ≤ 10,710)	British Columbia (n ≤ 10,740)	Territories (n ≤ 228)
Gender (n = 70,033)
Female	69%	75%	66%	69%	73%	70%	69%	72%
Male	29%	23%	33%	29%	25%	28%	30%	26%
Non-binary/transgender	0.2%	-	-	-	-	-	-	-
Age (n = 69,933)
16 to 24 years	9%	5%	7%	11%	6%	8%	13%	4%
25 to 34 years	7%	5%	7%	9%	4%	8%	6%	11%
35 to 44 years	8%	7%	6%	9%	7%	10%	6%	9%
45 to 54 years	10%	10%	8%	11%	9%	13%	9%	10%
55 to 64 years	24%	26%	28%	23%	26%	23%	22%	23%
65 to 74 years	26%	34%	31%	23%	32%	22%	26%	22%
75 years or more	5%	6%	6%	5%	7%	5%	6%	6%

Footnotes

Return to source of Footnote ¹ See the Methodological Notes at the end of this report for details on the survey methodology.

Return to source of Footnote ² Throughout this report, differences in the results between groups are statistically significant at the 95% confidence level or higher.

Media