Cities, Climate, Inequalities

Adaptation measures by Quebec municipalities: Progress and determinants according to deprivation level

April 2024

Johann Jacob, Ph.D., Research Professional, Faculty of Education, Université Laval, Observatoire québécois de l’adaptation aux changements climatiques (OQACC) et Pierre Valois, Ph.D., Professeur titulaire, Full Professor, Faculty of Education, Université Laval, Directeur de l’Observatoire québécois de l’adaptation aux changements climatiques (OQACC)

Introduction

Many climate risks are amplified in urban contexts, making climate change adaptation an issue that cities and municipalities cannot ignore. Meanwhile, inequalities make various groups less likely to benefit from adaptation measures. Those affected include low-income individuals, seniors, disabled people, people with health issues, socially isolated individuals, people with reduced mobility, members of ethnic minorities, and people from Indigenous communities (Lager et al., 2023). Amid growing demands for just and equitable forms of climate change adaptation (Fan et al., 2022; Lager et al., 2023; Majumdar & Weber, 2023),

there is a pressing need to develop the capacity to measure and evaluate, using indices whose reliability and validity have been tested, the progress made by Quebec municipalities in adopting measures aimed at: keeping stakeholders informed and preparing them for action; achieving concrete reductions in vulnerability to climate change; as well as improving the adaptation capacity or resilience of human and natural systems. Furthermore, it is essential to describe municipalities’ adaptation efforts in different contexts and to identify why some are more proactive than others in this regard, in order to strengthen their capacity to develop adaptation measures that reduce vulnerabilities faced by specific groups.

After an overview of the relevant literature, this synthesis presents the main methodological aspects of the approach followed to develop and validate measurement indices of municipal adaptation. Second, we present an overview of the adaptation landscape in Quebec municipalities, as rendered through the developed indices, along with key findings with respect to the factors that shape adaptation practices at the municipal level. Finally, these results are analyzed according to the level of social and material deprivation within the municipalities concerned.

Literature

Under the Paris Agreement of 2015 and the Katowice Climate Package that came out of the 2018 United Nations Climate Change Conference (COP24), all parties are responsible for documenting their success in adapting to climate change. As a result, a broad range of initiatives have been launched to develop and implement structured approaches for monitoring progress in climate change adaptation (Berrang-Ford et al., 2019). On the one hand, these initiatives have produced a plethora of discussion documents, guides, and tools. On the other hand, there is little consensus regarding what indicators should be used (Dupuis, 2015). A number of issues associated with measuring adaptation make it especially complicated to assess progress over time. These issues include the persistent lack of clarity surrounding related concepts (vulnerability, resilience, adaptive capacity, etc.) (Siders, 2019), the absence of common indices (ADEME, 2012), the complexity of the socio-economic and environmental processes involved (Bours, McGinn, & Pringle, 2015), and the wide range of possible adaptation practices and behaviours (Ford & Berrang-Ford, 2016).

In response to these difficulties, several indices have been developed and are in use, most of which aim to assess the vulnerability (Delaney et al., 2021; Edmonds, Lovell, & Lovell, 2020; El-Zein, Ahmed, & Tonmoy, 2021), resilience (Feldmeyer et al., 2020; Ferrier et al., 2020; Wu et al., 2020), and adaptive capacity (Siders, 2019) of human systems or ecosystems to natural hazards.

Indices can help simplify complex, multidimensional realities. However, indices are often criticized for the variability in the choices of the dimensions and indicators included, as well as for insufficient validation (Brooks, Adger, & Kelly, 2005; Eriksen & Kelly, 2007; Hinkel, 2011; Klein, 2009). In other words, the indices currently in use suffer from a lack of attention to methodological and psychometric considerations, such as validity and reliability.

Meanwhile, empirical research on climate change adaptation has increasingly relied on psychosocial models that predict individual behaviour, such as the theory of planned behaviour (Ajzen, 1991), to explain what drives people to action. For instance, such studies have looked at how people’s perceptions and beliefs regarding certain measures can pose barriers to adaptation (Clayton et al., 2015; Gifford, 2011; Grothmann & Patt, 2005; Moser & Ekstrom, 2010; Roser-Renouf et al., 2014). In terms of a just transition, various populations vulnerable to the effects of climate change need to be involved in the design and implementation of adaptation measures. Moreover, the beliefs, values, cultures, etc. of these groups need to be considered at all stages of the decision-making process (Lager et al., 2023).

When measuring progress made by municipalities in adapting to climate change, it is therefore necessary to address the psychosocial dimensions of how adaptation practices are adopted. Understanding why some communities are more proactive than others is key to efforts for ensuring municipalities are not only prepared to act, but also equipped to develop and implement adaptation measures capable of reducing the vulnerability of specific groups.

Case, Methods, and Data

This research study had two objectives. To begin with, an index that would help provide an up-to-date picture of how Quebec municipalities have been adapting to heat waves and flooding was designed and validated. The index was intended to account for two aspects (dimensions) of the situation. The first dimension included, among others, practices that municipalities could use to document inequalities, such as by using maps and databases to study the distribution of populations vulnerable to heat or floods based on socio-economic criteria (age, income, sex, address, marital status). The second dimension involved concrete adaptation measures, such as initiatives designed to promote active transportation, along with efforts to assess the impacts of measures that have already been implemented. To assess the index’s reliability and validity, different psychometric analyses were performed, including item analysis, confirmatory factor analysis, concurrent validity analysis, and nomological validity analysis (Jacob, Valois, & Tessier, 2022).

The next stage of research involved using structural equation modelling (Kline, 2015) to identify factors that prompt municipal authorities to adopt adaptation measures regarding these two climate hazards. To tentatively understand the differences between municipalities in their capacity to undertake adaptive measures, additional analyses (chi-square, correlations, effect size) were carried out. Accordingly, we compared adaptation levels and identified the psychosocial factors associated with the adoption of (or refusal to adopt) adaptation practices. In both cases, we considered the overall situation as well as the influence of social and material deprivation (Azevedo Da Silva et al., 2023).

In the summer of 2017, 1,218 urban and land-use planning professionals working at the municipal level in Quebec were invited to answer an electronic questionnaire. 139 responses were received, for a response rate of 11%. Building on a literature review and consultations with municipal stakeholders, the questionnaire was designed to assess, according to a municipality’s respondent, whether the municipality for which she or he works had adopted recommended measures for adapting to heat waves and flooding. We categorized adaptation measures—i.e., any municipal initiative intended to adjust to the impacts of climate change or take advantage of them, as well as efforts to develop or implement such initiatives (Dupuis & Biesbroek, 2013)—according to whether they constituted preparatory (groundwork) action or adaptive action (Lesnikowski et al., 2013). Preparatory action involves efforts to inform stakeholders and prepare them to act. However, it does not involve any actual changes to policies, programs or services. Examples include impact and vulnerability assessments. By contrast, adaptive action involves efforts to reduce the vulnerability of human and natural systems. It involves actual changes to institutions, policies, programs, or the built environment. Furthermore, drawing on the theory of planned behaviour (TPB) (Fishbein & Ajzen, 2010), the questionnaire was designed to collect data on a series of psychosocial variables: intention, attitudes, perception of social norms, and perceived control over practices.

To determine whether the municipalities represented in the survey responses were “advantaged” or “disadvantaged,” we relied on the Material and Social Deprivation Index (MSDI) developed by the Institut national de santé publique du Québec (INSPQ). Because the MSDI assigns deprivation scores by “dissemination area” rather than by municipality, we had to combine each municipality’s dissemination areas to determine its overall score. We weighted each dissemination area according to its share of the municipal population to produce an average deprivation score for the municipality. It is important to note that this approach glosses over internal variations, especially in larger cities that may contain dissemination areas with strikingly different levels of deprivation. Municipalities whose populations ranked in the top three quintiles for both social and material privilege were assigned to the “advantaged” category (n = 24). Those whose populations ranked in the bottom two quintiles for either social or material privilege were assigned to the “disadvantaged” category (n = 110). Five municipalities could not be assigned to either category due to a lack of data.

Results and Discussion

Through statistical analysis, we were able to design an index for measuring adaptation to heat waves and flooding among urban and land-use planners in the municipalities concerned, and to assess its reliability and validity. Table 1 lists the main dimensions (preparatory action and adaptive action) and subdimensions of the index. The subdimensions correspond to categories of action related to overcoming barriers and uncertainty, capacity building, reviewing development plans, assessing impacts and vulnerability, adaptation to heat, adaptation to flood, and monitoring or evaluation of these various types of action (Araos et al., 2016).

The results provide a first multidimensional look at the adaptation efforts of Quebec municipalities from an urban planning and land-use perspective (for full study results, see Jacob, Valois, & Tessier, 2022). However, they need to be treated with caution, given the partial coverage of the target population. The data presented in Table 1 give the overall impression that Quebec municipalities are engaged in adaptation efforts. In fact, they are just as likely to be engaged in efforts to prepare for adaptation as to have launched initiatives for actually developing adaptive capacity and reducing vulnerability. Accounting for level of deprivation, the results show no significant difference between disadvantaged and advantaged communities in terms of the number of preparatory action measures undertaken (0.31 and 0.25 out of 1, respectively): χ²(1, N = 134) = 0.33, p = 0.57; the effect size is zero, with Cramer’s V equal to 0.05. This is not the case for adaptive action measures, with a score of 0.38 out of 1 for disadvantaged municipalities compared to 0.13 out of 1 for advantaged municipalities. In other words, a higher proportion of disadvantaged municipalities (38.2%) than advantaged municipalities (12.5%) have taken adaptive action: χ²(1, N = 134) = 5.82, p = 0.02; the effect size is small, with Cramer’s V equal to 0.21.[1]

As shown in Table 1, only 4% of municipalities represented in the sample have produced vulnerability analyses, 7% have used mapping or databases to study the distribution of populations vulnerable to heat or floods based on socio-demographic criteria, and 24% have assessed the effectiveness of their efforts to reduce heat islands. Given the small proportion of municipalities that have taken such measures, it is vital to look more closely at efforts to document and examine vulnerabilities specific to certain groups. Without a good understanding of these vulnerabilities, municipalities will remain without a clear picture of the climate hazards facing their populations and incapable of understanding the social and spatial inequalities exacerbating the effects of climate change for certain communities and groups. Furthermore, determining whether such inequalities are exacerbated by adaptation practices adopted at the municipal level will require evaluative studies. Up to now, few municipalities have undertaken this kind of research.

Other statistical analysis aimed to understand why certain municipalities have been more proactive than others, and whether these underlying factors vary according to deprivation levels. Specifically, the theory of planned behaviour (TPB) helped us identify factors associated with adaptation practices adopted by Quebec municipalities in response to heat waves and flooding, whether in the form of preparatory action or adaptive action (see Figure 1).

Figure 1: Theory of Planned Behaviour (Ajzen, 1991)

According to the TPB, the adoption of an adaptation practice flows directly from the intention to adopt adaptation practices and the perception of control over the practices. Moreover, the decision to adopt a given practice can be influenced by the perception of control, that is to say the degree to which individuals believe they can adopt or control the practices in question. Insofar as it reflects actual control, perceived control over a practice should moderate the role played by intention. In other words, the intention to adapt should be a good predictor of the adoption of adaptation practices, provided those responsible do not perceive too many barriers (Fishbein & Ajzen, 2010). Based on the TPB, the intention to adopt an adaptation practice should increase if the person concerned has a favourable attitude toward the practice being considered, believes that the practice is supported by significant people (i.e., it is in line with perceived social norms), and has perceived control over the adoption (de Leeuw et al., 2015).

The results of structural equation modelling suggest that attitudes toward adopting adaptation practices (γ = 0.25, p = 0.06) and perceived control (γ = -0.32, p = 0.02) constitute non-negligible determinants of adaptation intention within Quebec municipalities (for full results, see Jacob et al., 2019 as well as Jacob, Valois & Tessier, 2021; 2022). However, the perception of social norms (in the form of sensitivity to other people’s opinions or propensity to conform to their expectations) did not emerge as a significant determinant of adaptation intentions.

We also carried out two series of statistical analyses based on the level of deprivation within each municipality. It is important to recognize the exploratory nature of these analyses, given the small size of our sample (which consists of 24 “advantaged” municipalities and 110 “disadvantaged” ones).

To begin with, where the adoption of adaptation practices was concerned, t-testing and effect size analyses revealed no statistical difference between advantaged and disadvantaged municipalities in terms of attitudes (means of 2.75 and 2.77, respectively) and perceived social pressure (means of 2.25 and 2.41, respectively). The results also showed no statistically significant difference between the perception of control over the adoption of adaptation practices in advantaged municipalities (M = 3.02 out of 4) versus disadvantaged ones (M = 2.80 out of 4): t(114) = 1.94, p = 0.0551. Still, the difference was non-negligible. The effect size, measured by Cohen’s d, was 0.47 (small effect size). The results therefore suggest that advantaged municipalities perceive more barriers to action than disadvantaged municipalities.

Second, to verify whether the reported reasons for taking climate action varied between advantaged and disadvantaged municipalities, we performed two series of Pearson correlation analysis: one on advantaged municipalities and another on disadvantaged one. Thus, the psychosocial precursors to climate action (attitudes, perceived social norms, and perceived control over the adoption of adaptation practices) were correlated with the two main components of our adaptation index (preparatory action and adaptive action).[6]

The results showed that the presence of certain beliefs influencing attitudes and perceived social pressure to take adaptive measures can prompt advantaged municipalities to adopt adaptation practices (see bolded values in Table 2). By contrast, no such beliefs were reported in disadvantaged municipalities, suggesting that perceived potential benefits were not the main drivers of action. A similar dynamic was observed with respect to social pressure. Compared to disadvantaged municipalities, advantaged ones appeared more concerned with expectations regarding the adoption of adaptation practices among various stakeholder groups (see bolded values in Table 3).

Table 2. Correlations between items measuring attitudes, according to a municipality’s level of deprivation (advantaged or disadvantaged) and the type of adaptation practice concerned (preparatory action [PA] or adaptive action [AA])

Table 3. Correlations between items measuring social norms, according to a municipality’s level of deprivation (advantaged or disadvantaged) and the type of adaptation practice concerned (preparatory action [PA] or adaptive action [AA])

Although the presence of certain beliefs influencing attitudes and perceived social pressures were moderately or even highly correlated with adaptation efforts in the case of advantaged municipalities, the role of these factors in prompting action appears to have been offset by the perception of significant barriers. In particular, three types of barriers (including funding issues) appeared to have discouraged advantaged municipalities from adopting adaptation practices, whereas only one (difficulties taking joint action with other regional stakeholders) appeared to have affected disadvantaged municipalities (see the bolded values in Table 4). With respect to funding issues, it is important to note that the disadvantaged municipalities in our sample tended to be larger. Accordingly, they may have had access to more funding than their smaller, advantaged counterparts. The extent to which this explains why funding issues pose a barrier in advantaged municipalities but not disadvantaged ones should be addressed in future studies, including by looking at what share of the budget is allocated to adaptation efforts in different sized municipalities.

Differences in population size could also help explain the variation in the role played by social pressure. In Quebec, the Municipal Powers Act (C.Q.L.R., c. C-47.1) tends to give as much autonomy as possible to the municipalities. It therefore seems plausible that disadvantaged municipalities (i.e., those with larger populations and potentially more resources) do not feel significant pressure from higher levels of government to take action on adaptation, in contrast to advantaged municipalities with smaller populations. As for the greater role of social pressure in advantaged municipalities (correlations of 0.62 compared to 0.34), the closer proximity between municipal officials and residents in less populous towns and cities could provide a potential explanation.

Table 4. Correlations between items measuring perceived control (barriers), according to a municipality’s level of deprivation (advantaged or disadvantaged) and the type of adaptation practice concerned (preparatory action [PA] or adaptive action [AA])

It might seem counter-intuitive that disadvantaged municipalities have been more active in adopting adaptation practices than their more privileged counterparts. In any case, our analysis of the psychosocial factors at play has identified obstacles specific to each context. And by associating such factors with different socio-economic contexts, we can support the development of strategies better adapted to the challenges facing either disadvantaged or advantaged municipalities. Furthermore, in terms of monitoring adaptation efforts, our findings highlight the importance of considering how decision makers understand obstacles to adaptation. Exploring this question will make it possible to assess whether the progress made by certain municipalities on climate action could eventually be threatened by emerging barriers.

Conclusion

From a social and environmental perspective, achieving a just and equitable transition in Quebec’s cities, towns, and regions depends more than ever on being able to monitor the progress made by municipalities in addressing concerns related to inequality and social justice, with a view to ensuring adaptation processes are as fair as possible. For example, such concerns could be addressed through risk and vulnerability assessments that account for the distinct impacts of climate change and adaptation measures on different social groups. Ensuring that vulnerable groups participate in all stages of the municipal decision-making process constitutes another approach to addressing issues of inequality and social justice (Lager et al., 2023). Future initiatives could support the adoption of these approaches by those responsible for climate action at the municipal level through efforts to understand factors that either facilitate or hinder the documentation of vulnerabilities specific to certain groups, to give such groups a voice in municipal affairs, to identify the impacts of adaptation measures, etc. Likewise, the monitoring of adaptation efforts needs to involve documenting, through sample surveys, how members of vulnerable communities experience extreme climate events and their impacts, the perceived risk of climate change, different beliefs regarding adaptation measures being considered, etc. Doing so would make it easier to achieve the desired changes in behaviour and practices thanks to a better understanding of the obstacles facing certain communities that are especially vulnerable to the effects of climate change (Kukowski & Garnett, 2024).

[1] The preparatory action dimension has four subdimensions, whereas the adaptive action dimension has three subdimensions (see Table 1). A binary variable was created for each subdimension to indicate whether the associated measures were either rarely or never adopted (0) or regularly implemented (1). The values for each subdimension were then added together and divided by the number of subdimensions, thereby providing a score between 0 and 1.

[6] This approach was deemed more appropriate than linear multiple regression, given the small number of participating municipalities. Moreover, the small size of our sample also required us to rely more on effect size than on statistical significance to judge the strength of correlations. The Pearson correlation coefficient value serves to estimate the size of the effect (Vacha-Haase & Thomson, 2004). According to Cohen (2013), r values between [0.10 and 0.30], [0.30 and 0.50] and [0.50 and 1.00] are generally interpreted as indicating small, medium, and large effects. Out of caution, we deemed a correlation to be non-negligible if its effect size was medium or large (r ≥ 0.30).

To cite this article

Jacob, J. et Valois, P. (2024). Adaptation measures by Quebec municipalities: Progress and determinants according to deprivation level. In Cities, Climate and Inequalities Collection. VRM – Villes Régions Monde. https://www.vrm.ca/adaptation-measures-by-quebec-municipalities-progress-and-determinants-according-to-deprivation-leveladaptation-measures-by-quebec-municipalities-progress-and-determinants-according-to-deprivation-l/