Campaigning in Quebec Municipal Elections: When the Party Defines Door to Door Canvassing as “The Right Way” to Campaign

Electoral campaigns are one of the key moments in political life. Yet, there is little Canadian work on this subject when it comes to municipal elections. However, the international literature on campaigning provides an opportunity for useful questions on the transformations of this aspect of local political life by bringing together political sociology and electoral sociology. That being said in this context, we present the results of an exploratory case study of campaigning in a municipal political party in a Canadian city, more specifically situated in the Province of Quebec. Municipal political parties are usually considered as electoral machines. It is therefore important to study in detail the way these organizations conduct election campaigns. More specifically, we are looking to explore how municipal political parties influence the campaign by providing electoral techniques. To achieve this, we closely examine the door to door canvassing strategy, which hints at what the party considers a “good campaigning” standard and helps us observe the behaviours of candidates and the different ways they fulfill the party’s requirements. It is therefore not a question of measuring the effectiveness of partisan electoral devices but of understanding how the party produces campaigning norms and puts them to work. The results presented here offer an original insight into the internal workings of a municipal political party—something that has never before been documented in the Canadian context. First, they help to open the black box that is municipal political parties and to better understand their internal modus operandi. Second, the results illustrate that election campaigning is still fundamentally based on one-on-one encounters.


INTRODUCTION
How to campaign in the municipal election in Quebec? Surprisingly, this question has not yet been raised in academic literature. It should be noted that municipal elections are generally understudied in Canada and Quebec. While municipal voting behaviour has been analyzed recently (Couture et al., 2014;Breux et al., 2017a;Breux et al., 2017b;Lucas and McGregor, 2021), the electoral supply angle remains largely unexplored. There are a few, mostly quantitative studies on candidates' profiles (Couture Gagnon, et al., 2019;Tremblay, 2014). Except for their media coverage (Théberge-Guyon and Bourassa-Dansereau, 2019) and the oversight of election expenses (Taylor and Vanhooren, 2021), municipal election campaigns are non-existent in the field of Canadian political science. While observers regret the low turnout in Quebec municipal elections (around 40% on average), the issue of voter mobilization is not clearly addressed. And this is precisely what we are investigating here: door to door canvassing as the central strategy in a municipal election campaign.
Municipal elections in Quebec offer a rather original field of study, combining a first-past-the-post (FPTP) electoral system, which entails a strong individualization of the process, with the presence of municipal political parties, which imply a collective rise to power and governance. In Quebec municipalities, the FPTP system implies one election per district 1 . Consequently, campaigns tend to be extremely localized. Only mayoral races are municipality-wide. In the traditional Canadian municipal model (Chiasson and Mévellec, 2014) all candidates run as independents. Provincial and federal political parties are not explicitly involved in these local races. The province of Quebec, however, stands out as it allows and even encourages the presence of municipal political parties in municipalities that have more than 5,000 residents. These parties, without brushing off individualism entirely, do add a "big city" territorial and political vibe with teams of candidates in line to win the mayoral race, blurring the ward/at-large distinction generally in place in other Canadian municipalities (Koop and Kraemer, 2016). In the field, municipal political party campaigns use both party strategies (presenting the party as a city-wide project, e.g., in press conferences with the party leader/mayoral candidate) and individualized strategies (having councillors elected with programs for their own constituencies). Municipal political parties are common in Quebec's major cities, but are not the standard everywhere. Elections are still largely between « party labelled" and independent candidates, whether for mayor or district races. This is the situation that prevails in the case under study here.
In larger cities, the growing presence of municipal political parties goes with what (Sawicki, 1994) described as a sort of modernized way of campaigning. Several elements reflect this trend. First, the financial resources they collect outside election periods are invested in professional election consultancy services (including communications and marketing strategies, sociospatial analysis of the voting population). Second, municipal political parties benefit from the electoral expertise of their various candidates and campaign managers who have experience at the federal, provincial or municipal level. With such tools, they are more likely than independent candidates to deploy a variety of strategies to "get out the vote", i.e., mobilizing the voting population on election day and making sure their supporters cast their vote. Mévellec and Tremblay (2013) have highlighted that municipal political parties in Quebec are mostly used as vehicles for campaigning, providing candidates with collective tools such as financial, material, organizational and social resources. Our objective is to look at the 2017 election and analyze the inner workings of a municipal political party campaign in a larger city in order to better understand the partisan effect in a crucial period of political life. In concrete terms, we conceive of campaigning as a social process in which the party seeks to impose its political vision but also its conception on the campaign process itself.
More specifically, we are looking to explore how municipal political parties influence the campaign by providing electoral techniques. To achieve this, we closely examine the door to door canvassing strategy, which hints at what the party considers a "good campaigning" standard and helps us observe the behaviours of candidates and the different ways they fulfill the party's requirements. It is therefore not a question of measuring the effectiveness of partisan electoral devices (for this, see Green and Gerber, 2019), but of understanding how the party produces campaigning norms and puts them to work. In other words, we transpose into Quebec's municipal field the question raised by Rémi Lefebvre when trying to understand what were "the good ways" and "the bad ways" to campaign (2016: 115). The results presented here offer an original insight into the internal workings of a municipal political party-something that has never before been documented in the Canadian context. First, they help to open the black box that is municipal political parties and to better understand their internal modus operandi. Second, the results illustrate that election campaigning, even when "modernized" and "collectivized" by the presence of parties, is still fundamentally based on one-on-one encounters. In our opinion, this finding sheds light on the difficulty of the Quebec municipal level to be recognized as a full-fledged, politicized level of government.
First, we look at the way the scientific literature frames the canvassing door to door strategy, i.e., either as an electoral engagement strategy or as the candidate's political work itself. Based on this literature, we observe a municipal political party's campaign. The second section explains the investigation protocol. We then show how the municipal political party uses door to door canvassing as their main campaign tool for district candidates. In fact, the strategy is normalized along with a method and the party's expectations of candidates and their campaign teams. In a final section, we focus not on party prescriptions and rationales, but on how candidates in the districts behave. Thus, we note a variety of behaviours in response to how this standard is implemented, ranging from compliance to transgression.
to convince targets of the merits of their proposals, be they political (get out the vote, vote for a particular party) or thematic (environmental protection, same-sex marriage). Canvassers follow a systematic methodology based on a list of names and addresses to be visited and a relatively short script that allows them to strategically and quickly guide the interaction with the target (Green et al., 2003). Further to the work of (Gerber and Green, 2000), the academic work seeks to demonstrate the conditions under which this door to door canvassing, led by volunteer canvassers or by the candidate themselves, helps get out the vote in their favour. This work in the United States focuses on campaigns with a first-past-the-post voting system, which emphasizes each candidate. This research confirms the effectiveness of this campaign strategy on people's electoral behaviour since they will vote more than those who have not been canvassed. According to the Nickerson survey (2003:1094), one additional vote is acquired for every 12 successful face-to-face interactions. Furthermore, according to (Green and Gerber, 2019), face-to-face meetings are also more effective than telephone contacts. Several other research studies have refined early findings or tested the model in other countries (for Canada, see Rudny et al., 2011;for Europe, Bhatti et al., 2019;Pons, 2018, and more specifically for the municipal level, Cantoni and Pons, 2020). In these latter cases, door to door canvassing by volunteers appears less effective in Europe than in the United States; however, the authors did not reach consensus on the reasons for this distinction. Thus, this literature globally seeks to evaluate the performance of campaign techniques on electoral turnout or on the actual vote. With this North American literature, (Lefebvre, 2016;Talpin, 2016) examined the transfer of American know-how, particularly implemented during Obama's first campaign, through the work of the French Socialist Party. Their perspective aimed to show how these new door to door buzzwords interact, and sometimes conflict, with the party's relatively rigid militant practices. In a comparative study, (Talpin, 2016) makes it clear that while French socialists have been inspired by the "O'Bama method," their field practices differ greatly from those in the United States. In the U.S., canvassers are required to make a scripted speech about their own experience, guided by the question, "Can we count on your vote?". The discussion should be kept brief and not get into the details of the election platform. In France, socialist party activists are mobilized. They approach voters by presenting their candidate's platform and criticizing that of their opponents, following old door-to-door routines. This French work allows us to reflect on the social and political context in which door-todoor canvassing is implemented.
The second corpus explores the candidate's point of view during door to door canvassing, generally from an ethnographic perspective. This research proposes to examine door to door canvassing as one of the rituals of political activity in the broader sense (for small French municipalities, see Catlla, 2014; large French cities: Lefebvre, 2005;rural England: Massard-Vincent, 2009). In this literature, door to door canvassing is perceived as a knowledge/behaviour specifically related to political activity. For (Hastings, 1987), door to door canvassing is a ritual in itself (a repetition of actions until the election) and an initiation rite in the sense that it is an opportunity for novice canvassers to learn from experienced canvassers, to establish themselves, or to speak out in interactions with citizens. This idea is taken up by Catlla (2014) who observes how the group door to door approach enables novices to learn the trade by imitation. These new canvassers learn through these social opportunities to assert themselves as candidates/incumbents. More generally, viewing door to door canvassing as a social activity allows for discussion of the framework and the issues of proximity in municipal politics (Le Bart, 2017).
This quick look at the literature on door to door canvassing inclines us to explore it not only as a campaign tool, aimed at getting the candidate elected, but also as an opportunity for the candidate to socialize in politics. As we will discuss below, the municipal level allows for the integration of these two bodies of work. Door-to-door canvassing is practiced there as a systematic technique for getting out the vote, carried out jointly by candidates and their volunteers.

METHODOLOGY
The methodology used in this exploratory study is based on nonparticipating observations and semi-structured interviews conducted within a municipal political party in a large Quebec municipality, i.e., with a population of over 100,000. The study protocol was subject to an ethics certificate issued by the University of Ottawa, guaranteeing the anonymity of the participants. For this reason, the name of the party under review cannot be disclosed. It is possible, however, to mention that the party was in its second general election campaign and submitted a complete list of candidates (i.e., with as many names as positions in the election).
In a first phase, we had access to the party's three main election meetings, before, during and after the campaign, throughout 2017. The first meeting brought together invested candidates and previously recruited local campaign organizers. During the day, plenary sessions and more specialized workshops were held. The first related to the general elements of the campaign (communication strategy, party election platform, presentation of campaign tools, socio-electoral portrait of different neighbourhoods in the city, etc.). The workshops focused on the respective tasks of the candidates' campaign organizers in the districts (funding, communication plan, press conference calendar, volunteer recruitment, etc.) and the candidates themselves (election platform). A second election meeting during the campaign brought together the same people. The purpose was to provide an overview of the campaign at the party level (funding, polling, election ads, communications plan for the past few weeks) while promoting what were considered good practices by some candidates. (i.e., whether or not objectives were met for door to door canvassing, funding, media visibility). Finally, a third meeting, this time post-election, was observed. The organizers presented the analysis of the election results, the gains achieved and the reasons for certain expected and unexpected losses, along with the lessons to be learned and the actions to be taken to better position the political party in the next election period. These observations were used to identify the stakeholders and their roles, to understand the party's electoral organization, its resources, its strategy, the instructions given to candidates and their local district organizations.
In a second phase, we conducted 13 interviews with 8 candidates (victorious or not), 1 local candidate campaign manager, and 4 party election committee members. The purpose of these interviews was to gather factual information about these individuals' experiences, their role during the 2017 campaign, and information on the Universe of meaning surrounding the campaign (Pinson and Sala Pala, 2007). The interviews were processed using NVivo12, whose door to door "nodes" support the analysis presented here. Verbatims were coded inductively, highlighting several topics related to the conduct of the campaign e.g., entry into politics, previous electoral experiences, media relations, campaign financing, campaign strategies. Door-to-door canvassing emerged as one of the central topics, and several nodes were created as shown in the following list (including the number of references for each node): preparation (5), topics discussed during door-to-door canvassing (61), timing (20), voter identification (38), importance (44), doing it alone or with others (36), physical elements (24), advice (11), alternative venues (8), liking or disliking door-to-door canvassing (14), opponents (8), reception (10).

IMPOSE THE STANDARD: "GET OUT THE VOTE, ONE DOOR AT A TIME"
Door to door canvassing has long been part of Canadian election campaigns. Although there is little research into the practice, it is an expected element of campaigns, regardless of the political level involved, as are debates, phone calls or postal mail. In this case, in 2017, the door to door approach is prioritized by the party as the electoral strategy to be adopted in the districts. This change in electoral strategy is the result of input from new party stakeholders bringing their knowledge acquired elsewhere. Combined with "modern" instruments for voter monitoring, it is legitimized by an argument based on its effectiveness and efficiency.

Door to Door Canvassing as a Collective Practice
The political party being studied includes activists and organizers from all walks of life, although its core stems from the Quebec sovereignty movement (Parti Québécois at the provincial level and Bloc Québécois at the federal level). The practices of campaign organizers are influenced by the practices of previous provincial and federal election campaigns and referendum campaigns. For example, there are no surprises in the electoral techniques favoured during the 2013 election, such as telephone calls and town hall meetings, spaghetti dinners or other BBQ and bowling nights organized by various associations, or by senior clubs. Because of its sustainability, the municipal political party has built up a solid financial reserve available to organize the 2017 election campaign. Also, a by-election that took place halfway through the mandate was an opportunity to mobilize supporters and test a new approach to the election campaign, focusing foremost on the door to door campaign. This strategic shift towards door to door canvassing is the result of the arrival of new players among party organizers. Socialized in federal politics (Liberal Party of Canada, New Democratic Party), they succeeded in convincing municipal party organizers of the importance of the door to door campaign as a central tool. In particular, one of these newcomers has dual experience in federal election campaigning and environmental activism. He has received training in canvassing from 350.org, a U.S.-based activist nongovernmental organization that fights climate change and trains activists in different countries. With the support of other members of the party, who also came from federal politics, he managed to negotiate a new, more collective and systematic door to door approach during the by-election campaign. This organizer describes in his words the ideal modus operandi: My method is, you have five or more people, but you try to have as many people as possible canvassing door to door, if there are new people they can work in pairs initially, let's say the first hour and then everybody takes their own sheet, everyone on the same street, but one person per door. Then you knock and introduce the candidate, the candidate is not necessarily at a door, so they don't have a list that they carry with them. They walk around, and then as soon as a door opens they come . . . , and then we start by introducing them while we're waiting for them to arrive, but as soon as they arrive the volunteer places themselves in the background, they only listen, they take notes. So, the candidate never has to deal with the logistics of taking notes and then remembering who we talked to, what number we got to, they just show up, they talk, and then the volunteer is sort of their secretary. The others continue, so the volunteers move all the time, the candidate is usually in the back, and then it optimizes their time because they're not there knocking on empty doors. (Interview 11, electoral committee).
While door to door canvassing was very much a part of candidates' practices in 2013, in 2017, the party is making it an injunction on the proper way to campaign. The method requires the availability of a significant number of volunteers, who can be mobilized throughout the campaign. However, not all candidates have that resource. Recruited members of the municipal political party proved to be unlikely to participate in such activities, thus confirming that the party is more hybrid in nature than a mass party (Duverger, 1976). Candidates had to use their personal networks. That is why the term volunteer is more fitting than activist for those who get involved in this campaign. Some of them are friends, colleagues, who did not live in the neighbourhood or even in the city, as this candidate can attest: . . .a former boss when I was working in (name of another city nearby) who knows what an election campaign is, because he was involved for years in the Liberal Party of Canada, who calls me out of the blue and says, I know what you have to do, how tough it is, so I'm giving you two afternoons. (Interview 2, candidate).
Candidates with little social capital found themselves in a more difficult position to apply the method. Although at the end of the campaign, available volunteers were "redistributed" to the most strategic areas by the political party.
The role of the volunteer is threefold: canvassing (knocking on doors, introducing the candidate), taking notes and identifying voters (i.e. noting whether they are supporters or opponents), and eventually closing the conversation: In principle, the volunteer was . . . right behind me with our documents . . . they would give me the name of the person . . . I would do the "Hello" and so on, and then the person would be identifying voters behind me, they would take notes on what the person was telling me and then they would be identifying voters as much as possible. And sometimes, if it was necessary for them to intervene because it was too long, it was (them) who usually cut off the conversation, because I find it easier to get [the volunteer] to do it than the candidate who says: I apologize that I really have no more time for you, I have to go to the next door. (Interview 7, candidate).
While volunteers are used strategically during door to door canvassing, the focus remains on the candidates. In the above excerpt, the candidate was at the centre, supported by their volunteer. In the following excerpt, it is clear that the candidate's presence is non-negotiable: I've had at least, at least 30 people who have walked with me, I've never walked alone, and no one has ever walked for me. (Interview 13, candidate).
The collective nature of door to door canvassing must not mask the strong individualization of election campaigns. In that sense, we are quite far removed from the American practices of canvassing for electoral participation. In the Quebec context, the candidates' presence is considered essential because the goal is to obtain the voter's vote. No candidate sends their volunteers to represent them.
One of the reasons for the expected effectiveness of canvassing door to door is that it quantifies the electoral behaviour of the people canvassed. While often imperfect, voter identification is, according to the party, the best tool it has at its disposal. Face-toface meetings allow for realistic identification. According to some organizers, the telephone does not allow for this clear verdict, which is why it was abandoned in 2017.
In 2017, we had decided to identify voters in person, not by telephone. In 2013, a number of volunteers identified by phone. Now we said it has to be the candidate, because you're still in the personality cult, so it has to be the candidate who (sees) people, it's as silly as that. And I think that, at the technical level, it's a winning situation. (Interview 12, electoral committee).
Surveying carried out during door to door canvassing is not just a one-time campaign activity but also part of a longer-term strategy. The results of the voter identification are entered into software (Democratik 2 ) that enables this data to be tracked and cross-referenced with data on voter turnout and sociodemographic data from the census. As summarized by one of the organizers, voter identification becomes a much more attractive tool, not only for the campaign, but also for a strategic vision on the part of the candidates and the party as a whole: You are much more able to identify your typical voter profile. What does it look like in this neighbourhood . . . in that neighbourhood . . . , which then lets you adjust the message a little bit to talk (to them) about something that really matters to them. (Interview 9, electoral committee).
The above organizer has a door to door vision that very closely resembles the activities promoted during Barak Obama's first election campaign, including access to and the cross-referencing of personal and political data, at the individual level.

Justifying Door-Door Canvassing
Since this more collective and systematic door to door approach represents a change from past practices, the members of the electoral committee developed a justification narrative based on two arguments: effectiveness and efficiency-even if these terms are never explicitly mentioned.

The Most Effective Way to Get out the Vote
For the party's electoral committee, the door to door approach is the best strategy to get potential voters interested in voting, because it's based on personal contact.
. . . There are really no other factors that are more important than that because the people who are going to see you canvassing door to door are going to remember that . . .. At the local level, it's a lot like that, you know. As I said, some of them are going to vote, they're going to line up their votes, but the vast majority of people are going to be really, do they seem good for the neighbourhood? He came up to my house, I saw him or her, I didn't see the other one. It's kind of depressing sometimes, but at the same time . . . I can understand that the voter who's not super-informed . . . they're going to say, the one who deserves my vote is the one who worked for it, to come and get it. (Interview 8, electoral committee).
You win an election canvassing door to door, people will still have the reflex to say, I don't know the candidates, I'm going to vote for the person who came (to our house), who went out of their way, who put the time and energy into it, and knocked on our door. (Interview 7, candidate).
These two excerpts illustrate the individualization of electoral races in the districts. The physical presence of candidates is considered essential, since, as elsewhere, "the personal and physical commitment of the candidate to see his or her availability and pugnacity plays a central role in this legitimization work" (Lefebvre, 2005:199). Indeed, being present is already interpreted as a "job" that deserves pay, in this case, as a vote in favour of it. Direct door to door contact between the candidate and the voter reinforces the individualization of the election campaign by creating a "visual reference" (Interview 12, electoral committee). This focus on face-to-face contact in 2017 is effectively paired with a decommissioning of a more traditional tool, that of telephone calls (Interview 11, electoral committee).
The individualization strategy is also explained by the municipal electoral context in Quebec. It compensates for the low media coverage of candidates in the districts 3 and the low interest in municipal politics among voters. The local party label does not provide instant visibility as in the case of the major provincial or federal parties. Thus, the municipal election remains primarily candidate-centric and the party label is only one of the candidate's resources (e.g. social capital). This context explains the importance of door-to-door canvassing for all candidates, regardless of whether they are running for a municipal political party. So, for the two candidates who are speaking below, it is the in-person contact that is important to make themselves known and get out the vote: So many people tell me, oh, you didn't come to our house, or oh yes, you came to our house. Really, it's so important to meet us in person. . . . I had zero media exposure, no televised debate. There is no alternative to going out and meeting people. (Interview 2, candidate) On voting day, the people I met (while canvassing door to door) voted . . . the people I didn't meet didn't vote for me, but the people I met voted for me. (Interview 6, candidate).
If individualization of the campaign comes through canvassing door to door for candidates in the districts, it relies on other avenues in the race for mayor. The candidate for mayor is already well known since they are the incumbent candidate. The name of the political party includes the leader's name. And finally, the party finances and organizes a communication campaign featuring them on regional media. So, when the candidate for mayor goes door to door, they do so less for their own campaign and more to support their candidates in the districts: The leader fires at the media level, but the campaign on the ground is done by the districts. (The leader) can, yes . . . He was made to canvass door to door in every district, he went everywhere. (Interview 12, electoral committee).
The Most Efficient Way to Get out the Vote: Targeting the Voters and Only the Voters Canvassing door to door is seen not only as effective, but also as the most efficient way to get out the (right) vote, for at least two reasons. The rational logic and management of the return on investment is reflected in the speeches of the political organizers. Unlike other strategies, the door to door approach makes it possible to target potential voters for each candidate in the districts, speak to them directly, and then compile the information accurately.
Canvassing door to door is considered the only way to ensure contact with potential voters. District voting requires a precise election strategy. As such, the party will always advise candidates to favour the door to door approach in their neighbourhood, rather than other forms of meetings (partys, markets, etc.) that can bring in people from outside the neighbourhood. Party organizers therefore reiterate their warning at internal meetings, which could be summed up by the motto "don't waste energy doing other things in other neighbourhoods." This clarifies their logic: If I meet you at an event, I don't know where you live, I can't go back and do a follow-up whereas when you go door to door and you know exactly what the address is, where the person is, so on D-Day, election day, you can go back. (Interview 11, electoral committee).
The candidates' campaign is basically going door to door in the neighbourhoods, I would say, that's more what we were trying to say: Did you go door to door? Are you doing what you need to do to win your district? Because I mean, there are candidates who think that just by talking to resident associations, by doing a couple of activities that it's in the bag for them, but the crux of the matter at the municipal level is, you go door to door if you want to win, you have to be seen by people out in the field. (Interview 8, electoral committee).
For municipal council candidates, the first-past-the-post system disqualifies a number of campaign devices observed elsewhere, such as the distribution of leaflets or other representational activities at city-wide events (Vignon, 2016). For local candidates, the distribution of leaflets, door hangers or participation in social events must therefore be done in such a way as to ensure that it targets voters living in their constituency. Therefore, these activities are only conducted within the boundaries of the electoral district. Only candidates for mayor are encouraged to go to larger public gatherings (festivals, markets, championships, etc.). Being a candidate for a party has little impact. If media activities are organized around the leader and their candidates, the campaign can be divided into as many districts as the city contains. The party's communication strategy has a city component, which means showing the group around the leader in the fight for mayor, and a district component that focuses on individual candidates. This dichotomy is also sometimes discussed by some candidates who would have liked more opportunities to showcase themselves as belonging to a team of potential councillors (for example around issues in territorial areas that go beyond the boundaries of their district).
To conclude on discourse by party organizers, it is clear that candidates in the districts must prioritize investing in the door to door approach. It is highly standardized in its procedure; it must be collective, the script must be oriented to the question allowing for voter identification and the pace must be sustained. However, the instructions do not include content (talking about the political party or the campaign themes to be addressed as a priority). Interaction itself seems to be the party's priority, while the content of the exchange is left to the discretion of each candidate. In the party being studied, this shift towards door to door canvassing as the main campaign device stems from the arrival of new party members who manage to convince organizers on the merits of their know-how acquired in other campaigns (political or activist), thus testifying to a certain emulation (Dolowtiz and Marsh, 1996) of US American campaigns toward the municipal political party. This emulation is all the more likely to exist because the welcoming environment (municipal politics, voting system in place) favours the door to door approach as an equally traditional tool of election campaigns in Quebec cities. Moreover, using the potential of a tool like the Democratik software, the approach is surprisingly modern. The party is therefore less focused on using an innovative tool than it is on reinvesting an existing tool in a more professional electoral strategy.
The effectiveness and efficiency hammered home by campaign organizers is used to justify the workload and time cost of door to door canvassing, of which candidates and campaign organizers are aware. 4 Contrary to what was observed in Lille (Lefebvre, 2005), party organizers have full confidence in the productivity of door to door canvassing. It is also clear that the door to door approach shows that the party organizes several types of campaigns concurrently, the mayor's campaign and those in the districts. Since constraints and opportunities are not of the same order, electoral strategies are also distinct, even though the party must ensure a certain consistency for all its candidates.

DOOR-TO-DOOR CANVASSING: A USEFUL TOOL FOR CANDIDATES BUT DIVERSE PRACTICES
This section looks at what is happening with the candidates. To capture what is at play in the door to door canvassing from the candidates' perspective, two phases of reflection are suggested. The first relates to the candidate's (future incumbent's) ability to politically represent their district, because they know it. This legitimacy of the land is acquired or demonstrated by a thorough knowledge of one's area, its inhabitants, and the issues unfolding there. The second phase of reflection looks at how the door to door approach allows candidates to practise social interaction, impose their personality, even if it means deviating from the party line.

Walking the Neighbourhood: Survey Your Area to Represent it
Door to door canvassing is somewhat akin to survey a territory, in a mapping and planning sense. It consists of walking around the neighbourhood ("walking the block"), knowing its characteristics ("measuring it") and being able to interpret them, and thus embody them. According to the party, this systematic surveing is ideally conducted during two rounds of door to door canvassing. The first is exploratory. The second must be conclusive.
The ideal door to door approach starts very early, at least 6 months before the official election campaign (which begins 40 days before the November 5, 2017, election). A first round of the neighbourhood lets candidates take ownership of the district's ultra-local issues, of which knowledge is key: there are no other ways to get to know the micro issues in your district. Macro (in the city sense) unless you live under a rock, it's hard not to know them, but micro like the corner, the park, speeding, you can't know that if you don't meet people, it's impossible. . . . And it's an exercise, I can't understand that someone is campaigning without going door to door, it's, I just don't understand it, I don't know what they're building a platform on, I don't know what they're basing their district statements on if there isn't, it's not true you can know everything if you don't meet people, I don't believe it. (Interview 13, candidate).
This knowledge of the field is recognized by the party, which is generally more focused on "wide city" issues. For example, one of the electoral committee members is designing a first round of door to door canvassing to test ideas or proposals for the neighbourhood: Hello, I am running for this issue . . . what do you think about it? . . . Am I mistaken, is that true? A good door to door experience is an opportunity to test the waters. (Interview 9, electoral committee).
For example, the first door to door tour is used to write the candidate's local platform in their district. It will be added to the party's website, as well as to the candidate's pamphlets that they can distribute during a second round of canvassing. This is what this candidate describes: We did a tour, I already had ideas, but we did a complete tour of the neighbourhood before writing the platform, so we had time to validate. And then after that I started my platform just before the official start of the campaign, it seems in that ballpark that the campaign officially starts. But that's it; I had consulted with everyone, I had knocked on all my doors. (Interview 3, candidate). I wanted to do my area program quickly, and then I realized fairly early on that it was impossible to do that without going door to door without knowing a little bit about the opinions of citizens in the district. So, I went door to door for a good month I'd say before I sat down seriously, but the platform was written super quickly too because of that. (Interview 7, candidate).
Here again, the institutional and partisan context explains this non-traditional use of door-to-door. First, Quebec's municipal political parties generally develop platforms that are not very detailed (Mévellec and Tremblay, 2013). Second, the districtbased election of municipal councillors implies a hyperlocalization of the issues and electoral promises of candidates. Third, the general disinterest of the population leads candidates to focus on very concrete issues that they believe they can address during their term. Thus, in a grassroots approach, the first doorto-door canvass is used by candidates to take the pulse of their constituency and adapt their electoral discourse to the issues identified.
During this first round of door to door canvassing, certain characteristics of the territory clearly appear to the candidates and influence how the exercise is carried out. Two main characteristics are noted: the density of the neighbourhood and its socio-demographic composition. In rural districts, commuting time between homes is not considered costeffective by candidates (interview 4), or by organizers, "You tell them to start with dense neighbourhoods, you never start with (rural neighbourhoods), it's not worth it." (Interview 11, electoral committee). Conversely, in high-density neighbourhoods, buildings are a barrier to candidates. They pose the problem of building access (see also Catlla, 2014:541;Lefebvre, 2005). The party's instructions are fairly murky on this. The legal nature of door to door canvassing in a collective building is left to interpretation by candidates and volunteers. If the pragmatic solution is put forward (once you enter the building, you might as well try to knock on doors), there are hesitations about not announcing yourself on the intercom.
While the building is considered a physical barrier to the smooth operation of door to door canvassing, it also reveals socio-demographic data. Dense and popular neighbourhoods have the lowest participation rates and the highest tenant turnover rates. This abstention and mobility of inhabitants are two essential vehicles for candidates and their campaign teams. Since the party has in hand the lists of voters who voted in 2013, the door to door approach is directed, on a priority basis, to higher voter-to-vote neighbourhoods. Thus, like French strategists (Liegey et al., 2011), abstentionists are not seen as a priority target of the party, but rather as inevitable. One candidate described how their colleague does not go door to door in buildings as "every apartment block that doesn't vote, where the rate is below . . . he wasn't going." (Interview 3, candidate) The municipal party's strategy clearly differs from that of the Democratic Party in the United States, which sought to convince non-traditional voters to vote. Yet the rate of abstention from municipal elections often exceeds 60%. In addition, the party's December 2017 campaign record showed that without new voters, it could never win a majority on the city council.
In concluding this subsection, it is important to remember that not all candidates were able to comply with the party's instructions to start door to door canvassing very early. Indeed, the party rules impose a nomination contest in each district to select the local candidate who will carry the party colors. While many of them were nominated early enough to start door to door canvassing in early 2017, or before, a second group of candidates was not nominated until spring 2017. Finally, some late candidates were recruited only during the summer. So those three groups of candidates had to look at their door to door canvassing separately. The first group was usually "model students" and were able to walk around their district twice. The second group generally completed one round. The third group had to strategically select "the right doors" because there was not enough time to do all of them. The only common thread is that everyone went door to door until the last day of the election campaign.
Hello . . . Can We Count on Your Vote?
In this last section, we revisit what is at play for candidates canvassing door to door. To do so, we take another look at the idea that local elected officials are continuously interacting (Faure, 2016:80). Door to door canvassing is a baptism by fire, particularly for those entering politics, since they must acquire sufficient dexterity to come into contact with strangers, introduce and impose themselves, make an electoral proposal and validate their target's electoral behaviour, in under 3 minutes. In doing so, the candidate is not alone. Exploring these learnings also shows that the "ideal" door to door effort is both collective and highly individualized, but the result is not guaranteed. This individualization means that the candidate quickly acquires certain campaign-related skills, such as showing up before the voters, interacting with a variety of stakeholders on a variety of issues, in a word, demonstrating their behavioural plasticity, to use Lefebvre's expression (2005). For candidates of a municipal political party, it may also mean quickly explaining and justifying the relevance of political parties in cities, so that they can move on to other parts of the conversation. These learnings take place at the beginning of the campaign under the microscope Frontiers in Political Science | www.frontiersin.org November 2021 | Volume 3 | Article 743252 of experienced people, as evidenced by two candidates and one campaign organizer: At first I . . . was (canvassing door to door with) a (former) federal Member of Parliament also to learn some things . . . The first time, I had never done that, he said to me, you knock on the door and you talk to the person, and then he would listen to me talk and then he would often make comments to me, that, that was great, you know when the person stubbornly insists, you see it's an opponent, you can stop and leave. I feel like sticking to my guns so that it's more things like that, or sometimes we say yeah, that's fine, that's why it takes one or two evenings, and then after that you're having success. (Interview 4, candidate).
You take the seasoned (people who have already canvassed door to door) who come with you on two or three outings, and then you get the hang of it pretty quickly. Anyway, I've gotten the hang of it, it wasn't, it's not rocket science. (Interview 5, candidate).
I went door to door with all my candidates to validate, "Are you doing well with door to door canvassing, should you say more of this, a little less of that," those (are the) kinds of things I was doing.
The interviews showed two things: First, the candidates all said they had picked up the pace quickly, within a few days. The interaction flow is standardized to maximize the few minutes allotted to each door. However, beyond this standardization, a number of candidates occasionally transgressed this way of doing things when they took the time to talk to people at greater length. I liked it all by myself because it's a one-man show. I had no one to restrain me, I had no one to hear me, I just had the person so I quite enjoyed it, and then I (canvassed) on streets where I also met people who saw me and they said: Are you Mr. X? And then I sat on the balcony for an hour and a half (with) influential people in the neighbourhood . . . because I was free, I was by myself allowed to do things like that, and then that's what I liked, that's the freedom I loved . . . I don't know if that wins, you know if I had met 1,000 more people maybe I would have won, but listen . . . (Interview 6, candidate).
This last excerpt shows that, when accompanied, candidates see door to door canvassing more as a standardized and compulsory exercise, which does not quite correspond to their personality. Socializing and meeting with others is experienced as artificial when it is done in a standardized way. Another candidate, who most often insisted on going door to door alone, also insists on discussing with neighbourhood residents. I don't know how I did it or I don't know how others do it, but I spent 6 months canvassing door to door in an extremely intensive way, I was among those who worked the most hours, and then I wasn't even able to finish the whole thing. I would never have been able to do a second round, of course, I talk a lot at the doors with people, but at any rate, that's my vision, I can't come and talk for 10 s and then leave because it's an opponent, for example, but otherwise I often spoke with people, that's what you want to improve, and then sometimes even when I had opponents, I tended to struggle with it and then when I heard people going to fifty doors an hour. I was happy when I got to twenty, twenty-two doors in an hour . . . The time to say hello, (ask) what your issues are about, I tried for a while to speed up, then I didn't feel good about it, and then I thought it felt like, here's my card, I'm off, it wasn't me. So, I stayed like this. (Interview 4, candidate).
These excerpts confirm that there are two competing visions about how to go door to door. On the one hand there is an electoral strategy that door to door canvassing must be efficient, involving talking to as many people as possible in their homes and being able to identify them as sympathizers or not. Volunteers are there to maximize the time and enhance the candidate's exposure. That is the message the party organizers have been sending. On the other hand is a door to door approach more designed as a rich social interaction, a time of exchange. This element is expressed by several candidates. The cost effectiveness of each discussion is then considered not in terms of the vote, but in terms of its social relevance and the ability of the candidate to connect with the inhabitants. This can no doubt be linked to the fact that these candidates did not win the election. In fact, valuing the quality of social exchange gives meaning to a failed door to door campaign.
Surveying is the grand finale of door to door canvassing. It is both a crucial and delicate exercise since the lists of "identified voters" are used to plan election day efforts. On election day, polling station slips will allow campaign organizers to see if supporters have gone out to vote, and if not, to solicit them to do so. Many recommendations were made at election committee meetings to ensure that voter identification was done conservatively. Democratik helps classify the people you meet into three groups: Sympathizer-Undecided-Opponent. 5 However, many candidates used these three groups as a creative jumping off point. Some had used another system in a previous election (sympathizer or opponent) and preferred to operate as in 2013 rather than mould to the software. Others inserted subcategories: "potential" or wrote "undecided+," "undecided -" as in the following excerpt.
So, I created a tag for myself . . . I called it the U + (undecided +), those who tell me, ah! But they don't tell me they're going to vote for me. For me to tag sympathizer, I'd have to hear you say, "I'm going to vote for you." If you didn't tell me that, I didn't tag you sympathizer. But if I felt you were one based on the exchange we had, I have created the tag U plus, and I have created the tag U minus, . . . because they didn't tell me, I'll never vote for you. You know my opponents, we didn't really have that many, it was the ones who told me, "Never," I didn't have that many. (Interview 13, candidate).
As election day approached, the undecided category needed to include fewer and fewer people according to the organizers' instructions. It's a question of the party knowing as precisely as possible who the voters are on election day, so that they don't run the risk of soliciting opponents.
It was really a hard-and-fast rule to headquarters to say that we don't want any undecideds, we want a yes or a no. (Interview 10, local organizer).
However, several candidates point to the fact that not everyone asks the question explicitly, or that they are unable to categorize the canvassed voters: And then the problem, let's say in campaigns like (another candidate's name), is that he never asked the question, he was uncomfortable with that question. (Interview 11, electoral committee) Within a few weeks of the campaign, we ended up with most people having 30 percent sympathizers, 10 percent opponents, and then the rest undecided . . . So, what we had been told to do a few weeks ago (was) you put your opponents on and then it's normal, don't be afraid to increase your ratio of opponents because it doesn't matter that you have 30 percent opponents, that doesn't make you lose. But what we really want is to identify the people we should call and then those we should not call on voting day. (Interview 3, candidate).
Surveying needs to explicitly ask the question that will enable us to clearly classify the voter but also get them to commit, or as Catlla proposes to "enlist the inhabitant" (2014:543). However, the exercise remains unsatisfactory in many districts as one candidate suggests: But I found that there were limits to that, because having got out the vote in three districts on election day, there were lots of places . . . that had been identified as sympathizers, (. . . but finally) you bring out sympathizers who are actually 50% opponents. (Interview 3, candidate).
The link between the door to door process and the outcome of the final vote differentiates between the Quebec context and the French context. If, in the French case, Catlla concluded that "the tour evaluation is less focused on measuring or anticipating results, that it expresses an understanding of and a relationship to political work" (2014:546), the same is not true in Quebec cities. In this case, the door to door approach corresponds as much to a certain standardization of the campaign (as a "professional custom" of the political world that should not be disregarded, for lack of not being elected), as an instrument for predicting and monitoring electoral behaviour. These two notions overlap in the divide between organizer and candidate.

CONCLUSION
This exploratory study has revealed a few dynamics at play in Quebec's municipal election campaigns. Even though it could not possibly have gone to the bottom of this complex aspect of democratic life, we were still able to identify a few important elements and establish three series of conclusions with regard to the partisan, individual and municipal levels.
The role of the municipal political party as a campaign organizer had already been identified (Mévellec and Tremblay, 2013) and is further outlined in this study through the devices and technology used by the parties as well as the practices that candidates are required to adopt. This paper helps to open the black box of the municipal political party in campaigns. Door to door canvassing might not be new; however, the streamlining of its modus operandi (collective, systematic, scripted) is. The party transforms into a source of modernizing standards for local candidates and their teams in order to win as many seats as possible. The party that was studied as part of this analysis also showed that campaign after campaign, the devices used evolve based on past experiences. A professionalization of the campaigning process can be observed based on, for instance, the backgrounds of several electoral committee members. However, some of the party's strategies are actually duplicated as two distinct types of campaigns are unfolding simultaneously. While the mayoral race is subject to relatively intense media coverage depending on local media outlets, district races are completely different. Door to door canvassing is considered the only option to create a more personalized bond between candidates and voters. Also, unlike big parties with strong support at other political levels, Quebec's municipal political parties cannot rely on a highly mobilizable base during campaigns. As such, it turns out that "the right way to campaign" depends not only on the means and expertise of parties, but also on candidates' behaviour and individual ability to recruit volunteers. However, it is clear that in the party studied here, if an important work of standardization of the campaign is engaged, the organization also remains very dependent on the behavior of its candidates. The electoral campaign period thus shows that the party has a relative power over its members. Some previous studies have shown the reassuring role of the party for many elected municipal officials, including in running election campaigns (Mévellec and Tremblay, 2016). That being said, these results must be nuanced.
At the individual level, candidates are encouraged to campaign in a certain way and favour certain devices (e.g., canvassing door to door, voter identification with the Democratik software), but behave in differing ways. Some embrace the party's strategies and logistics while others "cherry pick" standards and make them their own. Although we could not establish any connection between a candidate's attitude and their political capital from the interviews, we can propose the following hypothesis: the fewer social and political resources a candidate has, the more likely they are to follow the proposed standards. As such, "novice" candidates at the municipal level who have a weaker foothold in their districts seem more willing to follow the partisan guidelines. Incumbents and candidates who are known in their districts or have experience with unions are less likely to do the same. However, these findings are limited to the small size of our sample. Moreover, many studies have shown how local elected representatives consider their first campaign a key milestone in their career, a building block even (Faure, 2016;Mévellec and Tremblay, 2016). It proves more difficult to define door to door canvassing as a way to earn credibility as a politician when interviewing candidates as opposed to elected representatives. Walking around the neighbourhood, for instance, is generally known as a strategy used by a befitting representative-they know their district because they knocked on every door. On the contrary, for many interviewees, the ability to listen and cultivate a personalized relationship did not necessarily translate into an electoral win (Faure, 2016). Finally, this paper on the electoral moment shows that in Quebec, municipal politics remains fundamentally marked by interpersonal relationships between candidates and voters. The district voting system, coupled with the ideological weakness of most municipal political parties in Quebec, means that these oneon-one encounters often focus on mundane issues. As a result, the presence of a political party is not always sufficient to politicize municipal politics, at least during campaigns.
The contemporary literature on door-to-door canvassing focuses primarily on its effectiveness in mobilizing voters. However, the data presented here also invites to investigate what election campaigns do to the party and to the candidates. As an illustration, we have shown that door-todoor canvassing allows the party to propose a standard to its candidates. Even if the instructions are followed in an uneven way, it contributes to the institutionalization of the party and its expertise. It has also been shown that door-to-door canvassing allows candidates to develop their own electoral platforms, which they then defend to voters. This "programmatic" function of door-to-door canvassing should be found in the practices of independent candidates, in Quebec as elsewhere in Canada where ward-based voting is the norm.
Those who follow Quebec's municipal politics have surely observed the overlap between the municipal, provincial and federal levels during election campaigns. The case presented here demonstrates this overlap through the diverse backgrounds of campaign managers who apply to the municipal level strategies they learned at other levels. Whether at the provincial or federal level, these key actors are largely unknown to the academic field (Mc Grane and McGrath, 2019). We are unaware of any study done in Quebec or in Canada on municipal campaign managers. Our goal is to bring attention to these key actors, whether they work for municipal political parties or independent candidates, who represent the majority of municipal candidates in Quebec.

DATA AVAILABILITY STATEMENT
The datasets presented in this article are not readily available because dataset will be destroyed in 2023, according to University of Ottawa ethics policy. Requests to access the datasets should be directed to mevellec@uottawa.ca.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by the Office of Research Ethics and Integrity. The patients/participants provided their written informed consent to participate in this study.

AUTHOR CONTRIBUTIONS
The author confirms being the sole contributor of this work and has approved it for publication.