Impact of Bill C-19 on Securing Polling Places – CEO Appearance on Bill C-19, An Act to amend the Canada Elections Act (COVID-19 response)

Key Messages

• Signals from returning officers (ROs) are that securing polling places is more difficult during a pandemic.
• Educational facilities and seniors' homes are less likely to be available to serve as polling stations during a pandemic.
• However, as of May 20^th, ROs have completed a sites availability verification exercise for a Spring election held on a Monday and are confident they could lease at least 13,340 facilities for a Monday (against 15,487 in October 2019).
• Data gathered so far suggest fewer facilities would available over three days than over two days or on a Monday
• Canadians should expect more polling places that do not meet the standards of proximity, familiarity or accessibility.

Facts

• In 2019, there were 3,802 polling places for advance polls and 15,482 for ordinary polls.
• Educational facilities represented 19% of voting locations on Election Day, serving 46% of the voters, owing to the generally large number of polling stations that can be installed in a school gymnasium.
• In a pandemic context, traditionally used polling places such as schools and community centers may no longer be available.
• Election days over three days further restrict the use of schools and churches, and require ROs to identify alternate locations.
• The impact of a three-day polling period on the availability of polling places used in 2019 can only be extrapolated from observations made by returning officers in the Fall of 2020 and through some interpretation of the pre-event assignment completed in the winter of 2021 for a potential spring election.

Inventory of available polling places

• In fall 2020, ROs made preliminary contact with potential landlords. In total, 28,455 polling places were assessed. Of those, 1,512 are new locations which have never been considered for polling places in previous general elections (GE).
• In winter 2021, ROs contacted potential landlords to reassess the availability of sites for a potential spring election. Many educational landlords were not contacted as schools were deemed as unavailable during a pandemic. The ROs primary focus was to identify new sites and only revalidate the availability of sites that would likely be used for a spring election. Sites that were deemed as not available in the fall were not pursued. In total, 20,524 polling places were assessed. Of those, at least 3,144 are new locations which have never been considered for polling places in previous GE.
• There is a high level of uncertainty in securing sites which can be attributed to the limits imposed by provincial health and education guidelines. These limitations change from time to time due to status of the pandemic in each region. In addition, the unknown polling date in the minority government context contributes to the difficulties in securing spaces.

If Voting is held on Saturday and Sunday (CEO Recommendation)

• Of the locations contacted last fall, 11,057 polling places (39%) would be available to host a polling station on the weekend (no change from Monday's polling day), while 8% would not be available.
  • The sites contacted during the fall represent previously used polling places and new locations. Of the 39% listed as available, these locations hosted 52% of the total ordinary polls during the 43^rd GE.
  • For 53% of the sites contacted, availability are yet to be determined (TBD) as the landlords could not respond at the time of the initiative.
• It can be estimated that up to 80% (39% available +53% TBD for up to 92% maximum) of the polling places contacted in the fall 2020 would be available for a vote held on a weekend (Saturday and Sunday), in the context of a pandemic.
• This option was not assessed in the winter – as C-19 would not have been part of a spring election.

If Voting is held on a Monday

• During winter 2021, ROs contacted 20,524 polling places to revalidate availability for a potential spring election, under current legislation. Of the sites contacted, 13,082 (64%) would be available to host a polling station on a Monday and 15% would not be, while the rest are still unknown.
  • The sites contacted during the winter represent polling places used during the 43^rd GE (14,085) as well as potential new locations (6,739).
  • Of the polling places listed as available, 55% (7,354) of those sites were used during the 43^rd GE. These locations hosted only 42% ordinary polls during the 43^rd GE while 4,466 of the available locations are new and have not been used in the past. Many of these new buildings have not been fully assessed for accessibility requirements or to confirm exactly how many voting desks they will safely be able to hold, as many parts of the country were in lock-down, and site visits were not permitted by the landlords. Some of these locations may be less familiar to electors and in more commercial areas located outside of residential communities.
  • For 21% of the locations contacted, availability is yet to be determined (TBD) as landlords and owners were not able to confirm availability at the time of the initiative.
• It can be estimated that at least 64% and up to 85% of the polling places contacted during the winter 2021 would be available for a vote held on a Monday, in the context of a pandemic, but ROs did not contact any educational facilities, or landlords that said No during the fall. The total polling place count of 13,082 available is still well below the 15,482 used in the last GE, and the number of physical desks may be lower in some polling places to accommodate for physical distancing, so the hope is to have more polling places, not less. In addition, with the unknown election date there are always some sites that are unavailable when the writ is issued. The hope is to have a contingency of sites, which can be leveraged as needed. Regions that are most impacted by the loss of education facilities, are those that do not have available commercial spaces and polling places may be limited.

If Voting is held on three days – Extrapolation from existing data

• It is a different distribution of sites which would be available on a Saturday/Sunday versus Saturday/Sunday/Monday. The ability to direct electors to different sites for one of the days, and even secure alternate availability of a site on Monday that wasn't available on the weekend would be difficult. As such, this assessment looks at polls remaining open for all three days at the same site.
• Based on the locations contacted last fall, if a three-day polling period on Saturday, Sunday and Monday is held, only 9,463 polling places (33%) would be available to host a polling station.
  • The sites contacted during the fall represent previously used polling places and new locations. Of the 33% listed as available, those that were previously used hosted 44% ordinary polls during the 43^rd GE.
  • For 43% of the locations contacted, availability was yet to be determined as owners/managers were not able to respond at the time of the initiative. However, if we look at the more recent data from the winter evaluation to see if polls could be held on a Monday, commercial landlords have committed to allowing polls at a higher rate.
• Preliminary analysis indicates that fewer polling places will be available for 3 days (estimated at 70%) compared to a two day weekend vote (estimated around 80%), but many of the polling places available for 3 days are new polling places, not fully assessed for accessibility, number of polls that can be assigned, or fully evaluated for distance that electors will have to travel to the polls.

Impact to electors of a 3-day election period – Projections and simulation

• Elections Canada completed simulations using information on polling sites availability collected last fall and assumptions on voting behaviour. For polling sites where the availability could not be determined, a large number of random scenarios were generated. The possible variability of the overall results was also estimated.
• Projections based on statistical models of site availability show that the addition of a third polling day could cause about 3.3M voters (of the expected 13M in person voters) to be relocated at a different polling sites with a further distance from their homes, with approximately 200,000 electors not being able to vote in person on polling day due to a lower number of available sites or none within a reasonable distance to their home.

Other information recently gathered during both the winter and fall pre-event assignments.

• Education facilities (typically our most common site-type) has one of the lowest positive response rates (1474 available of the 7265 total in our database), however only facilities where positive conversations with school boards were already underway, were contacted in the winter.
  • Manitoba has communicated that no schools will be available if an election were to be called during a pandemic.
  • The Toronto District School Board (TDSB), who during the 43^rd GE leased 380 schools (representing 81 advance polls and 1,680 ordinary polls) across 24 EDs in Ontario, had confirmed their schools will not be available for use unless polling day falls on a holiday, a professional development day or on a weekend. This represents 400,000 electors for advance polls and 759,600 electors for ordinary polling day that will be relocated to a poll further away and will likely wait longer to cast their ballot in person.
  • The trend of unavailable schools is documented across the nation and impacts urban EDs the most.

By-elections in the fall of 2020

• We expect a wide variation around the averages provided above: For instance, during the Toronto by-elections last October, ROs could only secure 55% of the polling places used during the 43^rd GE. While ROs were able to add a few alternate and larger sites in commercial facilities such as hotels and a banquet hall, the overall number of sites was decreased, and electors had to travel further than within their immediate neighbourhoods and in some cases, outside their electoral district.
• While using polling places outside of the ED was possible during the by-election this will not be possible in a GE as the neighbouring EDs will be using all available locations.

Other factors and dimensions associated with the expected reduction:

• Implementing protective measures and physical distancing at the polls require more space, which further excludes traditional locations or reduces the number of polling stations and desks which can be set up.
• Reduced availability may result in having larger polling places with more polling stations.
• ROs recently conducted a revalidation exercise to assess polling place availabilities and identify new locations. To date, 4,866 new locations have been identified to which some unconventional locations such as arenas, theatres/cinemas, retail locations, convention centres, museums, etc. have been identified as potential polling places.
  • The purpose of this exercise was to identify additional locations should polling places used during past elections not be available for a potential spring election under the current legislation.
• Larger polling places could include alternate commercial sites such as hotels, conference centers, golf/country clubs, exhibition/convention centres, arenas/athletic facilities, vacant retail spaces, restaurants, etc. that can also provide entrances large enough to manage line-ups and to triage the flow of electors to and from their designated polling stations.
  • Cineplex Cinemas have agreed to host polling stations should an election be called this spring, before they return to normal operations.
  • Starwood Hotels and Resorts Worldwide, Inc have also agreed to host polling stations in available hotels across Canada.
  • Elections Canada has contacted Service Canada and DND, and PSPC to get a list of all federal government buildings that could potentially be used.
    • S. 122 of the CEA allows for ROs to use gov't facilities, but it is not always easy to determine who should be contacted and who has the authority or requirement (landlord if leased, government if owned, building manager, etc.) to authorize use of the premises as a polling place.
• Larger polling places ("super poll" locations) present advantages but also challenges in terms of proximity, familiarity or accessibility:
  • Advantages include:
    • having multiple entrances which will control the flow of electors
    • larger hallways and foyers which allow for adequate physical distancing
    • ample parking
    • generally good accessibility measures onsite
  • Disadvantages may involve:
    • having electors travelling further to their polling place and to areas in their community that are less familiar
    • locations not close to residential areas or with easy access to public transit, which may cause issues for electors who would normally walk or take public transit
    • increased walking distances from the parking area or from the nearest public transit stops to the voting room inside a large building
• If the election coincides with vaccination campaigns, a number of large facilities – sport arenas, community centers and recreation centres are not available for polling locations, and without educational facilities, these represent the highest number of alternate locations.
• The rental cost of polling places is expected to be higher in a pandemic because of:
  • the increased number of polling days (C-19);
  • the extra costs for cleaning and disinfection protocols;
  • pandemic induced disruption of supply in the rental market, pushing prices upward; and
  • the use of larger commercial or alternative sites that cost more than traditional polling places (like schools, community centers or churches). The recent planning reports show that these sites will cost up to $3.45M more than our standard rental rates for a one-day election, and we are not yet at the targeted number of polling places to serve Canadians.
• The costs for multiple days of voting were estimated based on directive rates and historical fees. Based on recent information, a three-day election period could drive these costs up more than estimated.
• During the October 26, 2020 by-elections in Toronto Centre and York Centre, ROs secured polling places at a much higher cost than during the 43^rd GE:
  • With the loss of traditional polling places such as schools, polling place rental fees increased by $19,494 for advance polling days and $24,188 for ordinary polling day.
  • For advance polls, there were fewer polling places (from 22 to 15) as compared to 2019, but the average rental fees per unit went from $959 to $2,706, a 182% increase.
  • For ordinary polls, there were fewer polling places (from 157 to 84) as compared to 2019 but the average rental fees per unit went from $333 to $911, a 173% increase.
  • 4 polling places located outside of the ED were used, compared to 9 locations during the 43^rd GE (over 338 EDs). Neither of the by-election EDs used sites outside of their ED during the 43^rd GE.