Why Collect Institutional Data on Local Governments?

The study of local politics in Canada has a long tradition of focusing on institutions (see Eidelman and Taylor, Reference Eidelman and Taylor2010). The merits of institutional features such as official nonpartisanship and weak political executives have been the subject of extensive inquiry and debate (Moore, Reference Moore2017; Urbaniak, Reference Urbaniak2014). The relatively frequent restructuring of local institutions in some, but not all, provinces has also sparked scholarly interest (Cashaback, Reference Cashaback2001; Sancton, Reference Sancton, Young and Leuprecht2006). At the same time, the fact that local institutions are unstable and functionally limited has led many scholars of urban politics in Canada to see them as just one piece of a bigger governance puzzle (Sancton, Reference Sancton2008; Horak and Young, Reference Horak and Young2012; Taylor and Bradford, Reference Taylor, Bradford, Moos, Vinodrai and Walker2020). In the early 2000s, many urban politics researchers adopted a governance approach in their work, focusing on how state-society interaction, as well as interaction across levels of government, shaped urban policy processes and outputs. Studies of individual policy fields, such as immigrant settlement (Good, Reference Good2009) and homelessness (Doberstein, Reference Doberstein2016), as well as work on multilevel governance (Horak, Reference Horak2013), all moved away from a focus on local institutions as objects of analysis.

Recently, though, the interest in local institutions has re-emerged. There is a boom in local election studies, which explore issues such as the dynamics of nonpartisan elections (McGregor et al., Reference McGregor, Moore, Jackson, Bird and Stephenson2017; Breux and Couture, Reference Breux and Couture2018; Lucas, Reference Lucas2020) and the structure of electoral finance regimes (Taylor and McEleney, Reference Taylor and McEleney2019). The historical development and policy effects of special-purpose governments have also received attention (Lucas, Reference Lucas2017; Lyons, Reference Lyons, Anderson and Stephenson2021), as have municipal accountability offices (Sancton, Reference Sancton, Breux and Couture2018). There is a re-emergence of interest in questions of local autonomy (Good, Reference Good2019; Alpert et al., Reference Alpert, Rosiers and Flynnforthcoming), and there is new historical scholarship that focuses on longer-term patterns of institutional development in Canadian cities (Lucas, Reference Lucas2017; Taylor, Reference Taylor2019).

Despite this institutional turn, we have a limited empirical base of comparative information on local institutions. In many other countries, national and/or subnational governments collect systematic data on local institutions (see, for instance, United States Census Bureau, 2023), but this is not the case in Canada. As a result, for example, we do not know, even approximately, how many urban municipalities in Canada conduct their elections by ward, as opposed to at-large. While Lucas (Reference Lucas2017) has compiled a historical dataset of governing structures across five policy fields, it is limited to six Canadian cities. In a recent overview, Taylor and Dobson (Reference Taylor and Dobson2020) map variation in municipal legislation across Canadian provinces, which sets out parameters for some local institutional arrangements but by no means determines them completely.

In this context, a database of local institutions in urban municipalities in Canada can help us advance our understanding of urban politics and local governance in several ways. First, it can give us a basic descriptive portrait of formal institutions in Canadian cities, so that we know the extent to which institutions that we take as characteristically Canadian actually are characteristic, and so that we can map the variation that exists. Second, it can allow us to explore the correlates of institutional variation—for example, the extent to which the powers of mayors differ by province, as opposed to by municipal population size. Third, it can allow us to examine how different aspects or dimensions of local institutions vary (or do not vary) together—for example, whether there is any correlation between a concentration of executive authority on council and the way in which service delivery is governed in a municipality. Finally, it can identify interesting cases for detailed work—be it unusual institutional arrangements in one municipality or similar or contrasting cases that researchers may not have been aware of before.

The Canadian Municipal Attributes Portal (CMAP)

CMAP is a database that contains information on over 50 institutional features of Canadian municipal governments. It has been built since 2019 by a research team based at the University of Western Ontario. Information is gathered from local government websites and other online sources using a standardized question template, which produces an institutional profile document for each municipality. Each profile is then verified either by a local expert on that municipality (if possible) or by a second researcher. Key datapoints from all profiles are then inputted into an Excel datasheet. As of September 2023, all of these elements are publicly available through an open-access web app: https://westernurban.shinyapps.io/CMAP/. In addition to viewing and downloading the template, profiles and datasheet, users can employ the app's interactive features to directly compare the institutional characteristics of selected municipalities and to produce charts that visualize the distribution of institutional arrangements across municipalities. This makes the app useful for both research and teaching applications.

CMAP currently contains information on 91 municipalities. At its heart is a “core” national sample of 65 municipalities. To identify this core sample, we balanced two distinct criteria—municipal population size and representation from all provinces. Selecting only the most populous municipalities would produce a sample largely composed of Ontario and Quebec municipalities, whereas selecting an equal number of municipalities from each province would produce a sample heavy on small municipalities. Instead, we selected the 10 largest single-tier or lower-tier municipalities in each province with a population of 25,000 or more, plus every provincial capital (if not already one of the 10 largest municipalities).Footnote ¹ If there were fewer than four municipalities with over 25,000 people in a province, the top four were included regardless of population. The municipalities in our resulting core sample include over half of the Canadian population and anywhere from 30 per cent to 69 per cent of the population in each province.Footnote ²

In addition to the core sample, CMAP currently contains information on the three territorial capitals, as well as all municipalities in the Greater Toronto Area. A living database, CMAP will continue to be expanded and updated with the assistance of the scholarly community and the broader public. As a first step, we are currently expanding CMAP to include all municipalities with a population greater than 25,000, which will add an additional 104 municipalities. Since some of the institutional characteristics documented in CMAP (such as service delivery arrangements) can change quite frequently, keeping the database up-to-date will be an ongoing challenge. In the first instance, we invite users of the database to submit any errors or updates that they notice directly to a dedicated CMAP correspondence address. In the longer term, our aim is to leverage scholarly and practitioner interest to secure funding support for a periodic comprehensive review of database information once every five years.

An Index of Authority Concentration for Canadian Cities

Making governing decisions is at the heart of what local politicians do. Local institutions influence how decisions are made by distributing formal decision-making authority in various ways across offices (for example, mayor versus council) and venues (for example, council versus special-purpose authorities) and by shaping the processes through which it is exercised. Yet researchers lack an overarching vocabulary that could be used to characterize this variation. We propose that a focus on the concentration of political authority can serve this purpose. Just as some metropolitan areas are more fragmented into multiple municipalities than others, so is decision-making power within municipalities diffused to varying degrees across public offices and decision venues.

Among scholars of Canadian local and urban politics, it is common to portray formal decision-making authority as highly diffuse (Sancton, Reference Sancton2015). In municipalities that feature nonpartisan councils elected by ward and a mayor with no significant executive powers, decision-making authority is distributed across the council as a whole, and councillors are heavily incentivized to focus on local constituent interests as opposed to citywide policy concerns (Moore, Reference Moore2017; Couture et al., Reference Couture, Breux, Bherer, Breux and Couture2018). Yet there are many urban municipalities in Canada—especially in Quebec and British Columbia, but also since 2023 in Ontario—that do not fit this pattern. Moreover, the overall distribution of political authority within a municipality is shaped by institutional variables beyond partisanship, electoral system and executive structure.

To produce a nuanced and integrated portrait of variation in the concentration of local political authority, we propose here a multidimensional index of authority concentration that can be used to systematically locate municipalities relative to each other on a continuum of concentrated to dispersed authority. In what follows, we draw on existing scholarship on Canadian local government institutions to identify four dimensions on which the concentration of political authority varies: geographical, executive, legislative, and service delivery. For each dimension, we identify one or more empirical indicators drawn from the CMAP database.

Applying the Index: Sample, Scoring and Limitations

To apply our index of authority concentration, we used data on the CMAP core national sample of 65 municipalities. Our scoring system is outlined in Table 1. For each indicator, 0 indicates dispersed authority, and 1 indicates concentrated authority. For the final composite index of authority concentration, we weighted each dimension equally as indicated in the right-most column. The raw service delivery scores (out of 10) were converted to a 0 to 2 scale by dividing the 65 sample scores into three roughly equal groups (those that score 0 to 5, those that score 5.5 to 7, and those that score 7.5 to 10). The index thus ranges from 0 to 8.

Table 1. Dimensions, Indicators and Scoring for Authority Concentration Index

Our measures focus on elements of the municipal institutional landscape that have been identified as substantively significant in prior research. Our scoring is designed to allow us to capture variation in institutional authority while also being relatively simple to apply. The four dimensions are weighted equally in our composite index because they are qualitatively different from each other, so there is no compelling reason to suggest that one dimension is more important than another. All that said, we do not claim that our measures, scoring and weighting are somehow definitive. On the contrary, we consider our index to be a first cut, one that illustrates the analytical value of such a tool and that we hope will spark debate and discussion about how variation in the structure of local decision-making authority in Canadian municipalities might best be conceptualized and measured.

We also do not mean to suggest that the institutional features that are the focus of our index fully determine the dynamics of local decision making. This is clearly not the case. It is well documented, for instance, that in Canadian “weak mayor” systems, a mayor's personal leadership and communication skills are of critical importance (Sancton, Reference Sancton2015; Graham, Reference Graham2018). Indeed, a wide variety of factors—ranging from the policy preferences of political leaders to the structure and mobilization of local societal groups—shape the practice of local politics in Canadian cities (Horak et al., Reference Horak, Lucas and Taylorforthcoming). However, there are good reasons to believe that formal structures of decision-making authority have an important influence on the practice of local politics and that it is therefore important to understand the shape that they take and how it varies across Canadian municipalities.

Patterns of Authority Concentration in Canadian Urban Municipalities

What can we learn using our index about patterns of authority concentration in Canadian municipalities? Before we present the composite index scores, let us explore the results on individual indicators. Figure 1 shows the raw indicator values on the geographical, executive and legislative dimensions. Each indicator is shown separately, and each is equally weighted. The municipalities are arranged in the table by province from top (west) to bottom (east) and from most to least populous within each province. Lower-tier municipalities in two-tier local government systems are designated with an (L); rural municipalities with a large land area and no single dominant population centre are designated with an (R).

Figure 1. Geographical, executive and legislative concentration of authority

Note: In this figure, as well as in Figures 2 and 3, (L) denotes lower-tier municipalities in two-tier local government systems. (R) denotes rural municipalities that cover a large land area with no single dominant population centre.

Legislative

Finally, political parties, as might be expected, exist only in Quebec and British Columbia—all of our Quebec cases have them, but only the four most populous municipalities in British Columbia currently do, for a total of 14 out of 65.

In aggregate terms, the clearest pattern in Figure 1 is the contrast between British Columbia and Quebec, on the one hand, and the rest of the country, on the other. The common wisdom of British Columbia and Quebec being “different” is thus supported. However, as we have already seen, Quebec and British Columbia are also substantially different from each other. In British Columbia, power is geographically concentrated, but there are no executive committees—in part because small (sometimes partisan) councils elected at-large probably already behave somewhat like executive committees might. In Quebec, by contrast, many cities appear—whether by design or not—to counterbalance the combination of party systems and strong executive committees, which produce what some have called a “Westminsterization” of local politics (Mévellec and Tremblay, Reference Mévellec and Tremblay2013), with a decentralization of authority to submunicipal borough councils.

The aggregate picture across the other provinces is one of relatively low authority concentration on our first three dimensions. Outside of British Columbia and Quebec, only Medicine Hat in Alberta has more than two of the five indicators of concentrated authority. In Ontario, the recent introduction of “strong mayor” powers has increased the concentration of authority in big cities. Perhaps unsurprisingly, the six large rural municipalities in our sample have dispersed authority on all three dimensions shown in Figure 1. Medicine Hat aside, a couple of idiosyncratic cases worth mentioning are Toronto, the only city outside Quebec that has a “Quebec-style” combination of community/borough councils, mayoral executive powers and an executive committee, and Hanover, Manitoba, the most deconcentrated municipality in our entire sample.

We now turn our attention to the extent to which governing authority over service delivery is concentrated in municipal hands. Figure 2 aggregates the raw scores for all 10 services based on the scoring criteria outlined in the previous section. Compared to the scores on our first three dimensions, the patterns are not as clear. The most immediately apparent patterns are that lower-tier municipalities in two-tier systems—noted (L)—control fewer services directly and that less populous municipalities in some provinces tend to score lower than more populous ones. We can, however, use our data to test Lucas’ (Reference Lucas2017) finding that service delivery models vary notably by policy field. Lucas’ finding is based on a longitudinal study of six municipalities, whereas CMAP allows us to examine patterns on a much larger cross-sectional sample.

Figure 2. Concentration of authority over service delivery

Table 2 shows the distribution of service delivery models across the 10 services included in our index. All nonmunicipal models, including private, voluntary, provincial and federal delivery, are bundled together in the right-most column. Clearly, service delivery does vary systematically by policy field across our sample. Fire, water, sewer, parking and waste management services all tend to be delivered directly by municipal departments. For transit, police and library services, service delivery is split relatively evenly between municipal department, dependent SPB, and nonlocal models. Meanwhile, electricity and EMS are most often delivered by a nonlocal body. Taken together, these patterns provide clear evidence of the importance of property services as a core component of what Canadian municipalities do, since four of the five services that are usually delivered by municipal department—fire, water, waste management and sewer—are also direct services to property, whereas services that are not as directly connected to property (transit, libraries, police, EMS) are much less frequently under direct municipal control.

Table 2. Distribution of Service Delivery Model by Service

Now we will combine all four dimensions of our index of authority concentration, reweighting each dimension equally (see above) to generate a composite authority concentration score. Figure 3 presents the results. The scores range from 0 (1 case) to 7 (2 cases) out of a maximum possible score of 8, with a mean score of 3.37 and a median of 3. Clear interprovincial differences are evident, underlining the importance of different provincial contexts and legislative frameworks as factors that shape the institutional structure of local authority. For instance, in Quebec the minimum score is 5; outside Quebec and British Columbia, the maximum score is 5, which is found in only one municipality (Medicine Hat). Table 3 presents mean scores by province and dimension. The composite mean score is highest in Quebec, driven by a high concentration of legislative, executive and service delivery authority (by contrast, Quebec has the lowest scores of any province on the geographical dimension), and lowest in the Atlantic provinces and Manitoba. The continued prevalence of institutionally “weak” mayors and nonpartisan elections in most provinces is reflected in the prevalence of low scores on the executive and legislative dimensions. The results corroborate the picture of significant variation by province in the institutional structure of authority in Canada.

Figure 3. Composite index of authority concentration

Table 3. Mean Authority Concentration Scores by Dimension and Province

Our index can also be used to explore whether the concentration of authority varies by municipal population. As the visualization in Figure 4 shows, the concentration of authority does vary in systematic and largely intuitive ways across our sample. The geographical concentration of authority decreases in more populous municipalities, which are more likely to have ward elections and borough or community councils; however, larger population is associated with more concentrated authority on the legislative, executive and service delivery dimensions, as well as with a higher overall concentration of authority across our four dimensions.

Figure 4. Authority concentration index score by population quintile