Introduction – National Electors Study on the 43rd Canadian Federal General Election: Report on Voter Experience
Elections Canada commissioned Phoenix Strategic Perspectives Inc. (Phoenix SPI) and Advanis to conduct research to help evaluate the 43rd federal general election.
Background and Objectives
Elections Canada (EC) is an independent, non-partisan agency that reports directly to Parliament. The agency is mandated to conduct federal general elections, by-elections, and referendums; administer the political financing provisions of the Canada Elections Act; monitor compliance; and enforce electoral legislation.
In the context of a federal general election (GE), EC conducts studies of electors that are used as part of the evaluation and development of EC's programs and services and to inform the Chief Electoral Officer's reports to Parliament.
The 2019 National Electors Study (NES) is EC's primary public opinion research study conducted for the 43rd GE held on October 21, 2019. The NES measures electors' values, opinions, and attitudes toward various election-related policy issues; their knowledge of, expectations toward, and experience with the electoral process; and their satisfaction with the agency's communications, services, and programs.
Many measures in the 2019 NES provide continuity with previous EC surveys conducted in 2015 for the 42nd GE, including the Survey of Electors, Evaluation of the Electoral Reminder Program, and the National Youth Survey.
The NES consisted of two components: 1) a national longitudinal survey of electors, and 2) a series of post-election focus groups and interviews. For the first time, the survey of electors included: a large-scale longitudinal sample in addition to a discrete random sample; integrated use of mixed online and telephone surveying; and three waves of data collection conducted before, during, and after the election period.
This report presents the survey results on electors' perspectives on the administration of the 43rd GE, in terms of:
- perceptions and experiences with registration and voter identification
- expectations and experiences with voting
- experiences and satisfaction with Elections Canada services
- overall confidence in Elections Canada and trust in the fair administration of the election
Two other reports present the findings of the NES on other topics, including a report on the voter information campaign and elector awareness during the 43rd GE and a report on electors' views on election-related policy issues.
The results from all reports will be used to assist in evaluating and refining Elections Canada's programs and services to the electorate.
A brief overview of the 2019 NES quantitative methodology is provided in this section. A detailed description of the research methodology, including the research instruments, can be found under separate cover.
The National Electors Study was conducted by telephone (by live interviewers) and online (via Advanis's online survey platform) between June and December 2019 in three survey waves. All respondents were eligible electors—Canadian citizens who were at least 18 years of age on polling day (October 21, 2019). The questionnaires varied in length from 15 to 20 minutes.
The longitudinal sample was recruited for the pre-election survey (W1) in June 2019 using probability sampling (random-digit dial phone recruitment using an overlapping dual frame, including landlines and cellphones) and non-probability sampling (web panel). Two-thirds of respondents were obtained using probability sampling. Electors were recruited in proportion to the population by province, age, and gender. To ensure sufficient final sample sizes, the recruitment targets took into consideration expected attrition across each sample source. Respondents in the longitudinal sample were invited back to participate in subsequent survey waves. A discrete random-digit dial sample was recruited solely for the post-election survey wave to offset attrition in the longitudinal sample.
|Wave||Sample||Method of data collection||Field period||Sample size|
|W1||Longitudinal||Online, telephone||Pre-election: June 12 to July 14, 2019||49,993|
|W2||Longitudinal||Online||Election period: September 3 to October 20, 2019||23,880|
|W3a||Longitudinal||Online, telephone||Post-election: October 23 to December 9, 2019||19,435|
|W3b||Discrete||Telephone||Post-election: October 22 to November 12, 2019||2,000|
The W2 election period survey was fielded as a rolling cross-section and divided into five phases. Questions changed based on the survey date to correspond with milestones in the election period, as follows:
- September 3 to 17: Early election phase (W2a)
- September 18 to October 1: Registration phase (W2b)
- October 2 to 8: Voter information card phase (W2c)
- October 9 to 15: Early voting phase (W2d)
- October 16 to 20: Election day phase (W2e)
The survey data have been weighted to correspond to the demographic composition of the full population of electors. Weighting was done in two stages: adjustments for factors related to the study design, including differences in probability of selection between sample frames, the in-scope rate, non-response, and household size; followed by post-stratification/calibration to align the results with known population characteristics of age, gender, and province/territory. Different weights were calculated at each wave to account for attrition in the longitudinal sample over the course of the study.
The inclusion of the non-random web panel means no estimate of sampling error can be calculated for the entire sample, and results are not statistically projectable to the entire elector population. A margin of sampling error and statistical estimations can be obtained if the panel is excluded and only the random samples are considered, in which case all samples are of a size such that overall results across all waves would have a margin of sampling error less than ±1%, 19 times out of 20. The margins of error for subsamples would be larger.
Notes to Readers
- This research relies on self-reported voter turnout, which historically is over-reported in public opinion surveys: In this survey, self-reported turnout was 90%, while the official turnout rate for the 43rd GE among registered electors was 67%. A limitation of this current research, therefore, is that it over-represents voters in the survey sample. Two factors may be responsible for the over-representation of voters: 1) people who vote may be more likely than non-voters to participate in a study about voting (response bias); and 2) people who did not vote may report doing so because they think to present themselves in a more positive light (social desirability bias).
- The term elector denotes survey participants who were eligible to vote in the 43rd GE (i.e., all respondents). The term voter denotes survey participants who reported that they voted in the 43rd GE.
- All results in the report are expressed as percentages, unless otherwise noted. Percentages may not always add up to 100% due to rounding or multiple mentions.
- The number of respondents varies where questions were asked of sub-samples of the survey population and during different survey waves.
- Statistically significant subgroup differences are identified in the report. Subgroup reporting includes a variety of demographic, behavioural, and attitudinal variables. Particular subgroups were defined as follows:
- Indigenous: Electors who self-identified as First Nations, Inuit, or Métis
- new Canadians: Immigrants to Canada who became citizens after the 42nd GE, and therefore were newly eligible to vote in the 43rd GE
- NEET youth: Electors 18 to 34 years old who were “Not Employed, in Education or in Training” during the fieldwork, compared to youth attending school (i.e. 18 to 34, full-time or part-time secondary or post-secondary students) and youth working full-time or part-time (and not attending school)
- electors with disabilities: These respondents were identified using a functional disability approach based on reported difficulty with various activities, combined to a single measure of overall level of disability, on a scale ranging from no disability to mild to very severe disability
- The results of significance tests establish the extent of relationships among variables, but cannot be generalized to the population, given the inclusion of respondents collected from a non-probability sample. When reporting subgroup variations, only differences that are significant at the 95% confidence level and that pertain to a subgroup sample size of more than n=30 are discussed. If one or more categories in a subgroup are not mentioned in a discussion of subgroup differences, it can be assumed that significant differences were found only among the categories reported.
- Comparisons to results from the 2015 Survey of Electors Following the 42nd General Election are included when possible.