Insights on Canadian Society
The intersection of flooding and deprivation: A study of neighbourhoods
Text begins
Introduction
Flooding, triggered by heavy rainfall, coastal storm surges and other weather events, has become a growing reality for many Canadians. Costing $2 billion in structural damages to homes per year, flooding is the number one natural disaster in Canada – more common than earthquakes, tornadoes and even wildfires. Furthermore, the risk of flooding is expected to continue to grow in the coming decades, largely due to the effects of climate change.
The Intergovernmental Panel on Climate Change explains that climate change can intensify the water cycle, resulting in increased heavy rainfall and associated flooding.Note 1 Indeed, rising sea levels and heavier precipitation mean a larger area of land in Canada is now at risk of flooding. In 2024, 80% of Canadian cities were located on flood plains.Note 2
Depending on the extent of damage, the aftermath and recovery from a flood can take weeks or even months. For individuals and families, it often means assessing damage, contacting insurance, replacing destroyed belongings, repairing property, and general clean-up (including managing mold). Dealing with flooding takes both time and money – and may be particularly difficult for socio-economically vulnerable populations. In these cases, full recovery may not be possible since the financial impacts of recovery are long-term, resulting in lasting socio-economic damage. To add to the risk of long-term harms, vulnerable populations may also be more likely than other Canadian families to live in flood-prone areas.
To examine the issues of flood exposure and socioeconomic disparities at a neighbourhood level, this article uses the Canadian Index of Multiple Deprivation (CIMD) (see textbox),Note 3 as well as Census neighbourhood-level income data. It seeks to answer the following two questions: 1) what are the demographic and socio-economic characteristics of areas that have experienced flooding over the past four years, and 2) have these flooded areas seen changes in levels of socio-economic deprivation between 2016 and 2021?
Flooding extent maps were obtained from Natural Resources Canada’s Emergency Geomatics Services (EGS),Note 4 for the years 2017 to 2020 (i.e., the between-census period). For consistency, only provinces identified in the flood maps were included in this study (i.e., British Columbia, Alberta, Manitoba, Ontario, Quebec, and New Brunswick) (Map 1).
Description for Map 1
This visual contains five maps that display the flood extent in Northern Alberta, Southern British Columbia and Alberta, Southern Manitoba, Eastern Ontario and Western Quebec, and Southern New Brunswick. The red represents areas that experienced flooding between 2017 and 2020, the blue represents water, and the green represents the provinces. For example, this map displays flooding in Eastern Ontario and Western Quebec that occurred along the provincial borders. Although the maps display smaller areas in Manitoba and New Brunswick, these maps highlight considerable flooding in Manitoba, where 32.7% of neighbourhoods were affected by flooding within the study period, and New Brunswick where 17.6% of neighbourhoods were affected by multiple years of flooding. This map does not specify the number of floods experienced or proportion of floods by region.
Source: Natural Resources Canada, Flood Extent Maps, 2017 to 2020.
What is the Canadian Index of Multiple Deprivation?
The Canadian Index of Multiple Deprivation (CIMD) was created in 2016 to better understand inequalities through various measures of social well-being, including health, education, and justice.
The CIMD measures area-based deprivation across four socio-economic dimensions: residential instability, ethno-cultural composition, economic dependency, and situational vulnerability. The data are aggregated at the Dissemination Area (DA) level, which is a small, relatively stable area of census geography comprised of about 400 to 700 people – approximately corresponding to a neighbourhood in most cities.
Description for Figure 1
The first dimension is residential instability which includes the following indicators: proportion of dwellings that are apartment buildings, proportion of persons living alone, proportion of dwellings that are ownedFigure 1 Note 1, proportion of movers within the past 5 years, proportion of the population that is married/common-lawFigure 1 Note 1, and median 2021 household incomeFigure 1 Note 1.
The second dimension is ethno-cultural composition which includes the following indicators: proportion of population that is foreign born, proportion of the population self-identified as visible minority, proportion of population with no knowledge of either official language (linguistic isolation), average number of persons per room, and proportion of the population which are recent immigrants (arrived in five years prior to Census).
The third dimension is economic dependency which includes the following indicators: proportion of the population participating in the labour force (aged 15 and older)Figure 1 Note 1, proportion of the population who are aged 65 and older, ratio of employment to populationFigure 1 Note 1, dependency ratio (population aged 0-14 and population aged 65 and older divided by population aged 15-64), and proportion of population receiving government transfer payments.
The fourth dimension is situational vulnerability which includes the following indicators: proportion of the population identified as Indigenous, the proportion of homes needing major repairs, proportion of the population aged 25-64 without a high-school diploma, the proportion of single parent families, and median dollar value of dwellingFigure 1 Note 1.
- Note 1
-
This indicator is reverse-coded, meaning it was coded opposite of the measure. For example, proportion of population that is married or common-law becomes proportion of population that is single, divorced, separated or widowed.
Notes: The dimensions are ordered such that the dimension on the left explains the highest percentage of the variance of the data and the dimension on the right explains the lowest percentage. Excludes the territories.
Source: Statistics Canada, Canadian Index of Multiple Deprivation, 2021.
Floods were more common in rural areas than urban centres
Between 2017 and 2020, just under 1 in every 10 neighbourhoods included in the study (i.e., 9.3% of Dissemination Areas, or DAs) were affected by flooding (Table 1). This translates into 4,829 flooded neighbourhoods over a three-year period. These same neighbourhoods were often repeatedly affected by flooding, with 41% of flooded areas experiencing two or more floods within the study period.
| Total | Rural and urban areas Table 1 Note 2 | Province | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Rural area | Small or medium population centre (1,000 to 99,999) | Large urban population centre (100,000 or greater) | British Columbia | Alberta | Manitoba | Ontario | Quebec | New Brunswick | ||
| number | ||||||||||
|
||||||||||
| Total Dissemination Areas Table 1 Note 1 | 52,014 | 9,565 | 10,833 | 27,650 | 7,848 | 6,203 | 2,223 | 20,465 | 13,805 | 1,470 |
| percentage | ||||||||||
| No floods | 90.7 | 77.8 | 91.3 | 95.2 | 92.3 | 97.8 | 67.3 | 96.0 | 84.2 | 74.9 |
| At least one flood between 2017 and 2020 Table 1 Note 3 | 9.3 | 22.2 | 8.7 | 4.8 | 7.7 | 2.2 | 32.7 | 4.0 | 15.8 | 25.1 |
| 1 flood between 2017 and 2020 | 5.5 | 13.2 | 5.5 | 2.3 | 7.7 | 2.2 | 20.2 | 2.2 | 7.9 | 7.5 |
| 2 or more floods between 2017 and 2020 | 3.8 | 9.0 | 3.2 | 2.4 | 0.0 | 0.0 | 12.5 | 1.8 | 7.9 | 17.6 |
Among the most susceptible to flooding were neighbourhoods in rural areas, with almost half (48.4%) of flooded areas lying outside of population centres. These areas include small towns and villages; rural parts of Census Metropolitan Areas (e.g., estate lots); and more remote lands. Looking at it another way, more than one-fifth (22.2%) of rural areas were affected by floods, compared to 4.8% of areas in large urban centres. Rural areas were also more likely to experience multiple years of flooding – with 9.0% of rural Dissemination Areas affected for more than one year during the study period.
Based on the flood extent maps, there was some variation in the number of communities that experienced floods across provinces. For example, about one-third (32.7%) of neighbourhoods in Manitoba were affected by floods within the study period – 3.5 times higher than the national average of 9.3%. In addition, some provinces reported multiple years of flooding, such as New Brunswick, where 17.6% of neighbourhoods were affected by multiple years of flooding.
Neighbourhoods affected by flooding were more likely to have greater levels of economic dependency and situational vulnerability
Household income alone was not related to flood susceptibility, with no significant differences in flooding across income quintiles. Specifically, 8.9% of neighbourhoods in the lowest income quintile (i.e., before-tax household incomes ranging from $7,982 to $39,569) and 7.3% in the highest income quintile (i.e., mean neighbourhood household incomes above $69,338) were affected by flooding (Table 2). Over time, experiences of flooding generally did not have a notable impact on before-tax income. There were two exceptions: Alberta and Manitoba. Flooded communities in these two provinces saw drops in before-tax income between 2016 and 2021. That said, income falls short in capturing true socio-economic diversity. It cannot fully capture financial vulnerability or stability, as it does not take into account types of income sources, housing conditions, as well as assets and debts.Note 5
| Total Dissemination Areas Table 2 Note 1 | No floods | At least one flood | |
|---|---|---|---|
| number | percentage | ||
Sources: Natural Resources Canada, Flood Extent Maps, 2017 to 2020. Statistics Canada, Census Boundary Files, 2016; Canadian Index of Multiple Deprivation, 2016. |
|||
| Income Quintile | |||
| 1 - lowest | 10,711 | 91.1 | 8.9 |
| 2 | 10,555 | 88.8 | 11.2 |
| 3 | 10,136 | 89.8 | 10.2 |
| 4 | 9,138 | 91.8 | 8.2 |
| 5- highest | 8,681 | 92.7 | 7.3 |
| Missing rural/urban designation | 2,793 | ... not applicable | ... not applicable |
| Canadian Index of Multiple Deprivation (CIMD) Quintiles | |||
| Economic dependency | |||
| 1 - least deprived | 9,404 | 93.0 | 7.0 |
| 2 | 9,843 | 91.7 | 8.3 |
| 3 | 9,946 | 90.7 | 9.3 |
| 4 | 9,715 | 90.3 | 9.7 |
| 5 - most deprived | 9,140 | 88.6 | 11.4 |
| Missing rural/urban designation | 3,966 | ... not applicable | ... not applicable |
| Ethnocultural composition | |||
| 1 - least deprived | 8,695 | 82.4 | 17.6 |
| 2 | 9,161 | 86.0 | 14.0 |
| 3 | 9,688 | 91.9 | 8.1 |
| 4 | 10,164 | 94.6 | 5.4 |
| 5 - most deprived | 10,340 | 97.7 | 2.3 |
| Missing rural/urban designation | 3,966 | ... not applicable | ... not applicable |
| Residential instability | |||
| 1 - least deprived | 9,558 | 89.5 | 10.5 |
| 2 | 9,445 | 89.0 | 11.1 |
| 3 | 9,494 | 89.7 | 10.3 |
| 4 | 9,471 | 92.2 | 7.8 |
| 5 - most deprived | 10,080 | 93.9 | 6.1 |
| Missing rural/urban designation | 3,966 | ... not applicable | ... not applicable |
| Situational vulnerability | |||
| 1 - least deprived | 9,763 | 93.0 | 7.0 |
| 2 | 9,863 | 91.4 | 8.6 |
| 3 | 9,821 | 90.2 | 9.8 |
| 4 | 9,670 | 89.8 | 10.2 |
| 5 - most deprived | 8,931 | 89.9 | 10.1 |
| Missing rural/urban designation | 3,966 | ... not applicable | ... not applicable |
Taking a broader approach, flooding did disproportionately affect neighbourhoods with greater levels of economic dependency. In particular, communities who were the most economically dependent,Note 6 defined as reliance on the workforce, government transfers (e.g., social assistance, employment insurance, Old Age Security), along with a high dependency ratio (children and seniors) were 1.6 times as likely to experience at least one flood in 2016 as the least economically dependent quintiles (Table 2).
Another indicator of deprivation is situational vulnerability, where homes are in greater need of repair with lower market values, along with population compositions associated with financial vulnerability (e.g., lower levels of education, single parents). Again, communities with high levels of situational vulnerability were most likely to face a flooding event, where the proportion affected by one or more floods (10.1%) was higher than the proportion for communities with lower levels of situational vulnerability (7.0%) (Table 2).
In contrast, communities characterized by residential instability, such as those with high residential mobility, and greater number of renters to owners, were not at greater risk of flooding. In fact, the opposite was true. Communities with low residential instability were more likely to be affected by a flood during the study period: 10.5% versus 6.1% of communities with high levels of residential instability (Table 2).
Post-recovery period: What happened to the socio-economic composition of flooded communities?
In the post-recovery period, flooded communities, regardless of pre-flood economic deprivation, saw evidence of economic improvements. This was most pronounced in large urban areas, where deprivation related to economic dependency decreased, with average deprivation quintiles falling slightly, from 2.94 in 2016 to 2.78 in 2021. In comparison, the average economic dependency quintile increased from 2.76 in 2016 to 2.79 in 2021 for urban areas that were not affected by flooding (Chart 1.1).
This could be somewhat attributed to government financial assistance and insurance following disasters, that can provide Canadians with support as they recover and adapt for the future. Canada’s National Adaptation Strategy considers how vulnerable communities require access to additional resources and recovery support.
Despite improvements in economic dependency in flooded areas, these neighbourhoods saw increases in levels of vulnerability in other dimensions. Notably, although residential instability was not associated with an increased risk of flooding, it was found to increase after a flooding event, primarily among rural areas. Specifically, rural areas and small or medium population centres affected by flooding saw greater increases in residential instability over the study period (2.19 to 2.38 and 3.24 to 3.36 respectively), compared to those communities that did not experience flooding (2.19 to 2.31 and 3.29 to 3.36 respectively). For urban communities, residential instability did not change after a flooding event (Chart 1.2).
Communities that experienced flooding saw minimal changes in levels of deprivation related to situational vulnerability between 2016 and 2021. Rural communities observed improvements over time with regards to this dimension of deprivation, but this was consistent regardless of having experienced flooding.
Data table for Chart 1
| 2016 | 2021 | |||||||
|---|---|---|---|---|---|---|---|---|
| Average quintile | 95% confidence interval | Standard error | Average quintile | 95% confidence interval | Standard error | |||
| lower bound | upper bound | lower bound | upper bound | |||||
| quintile | ||||||||
| Notes: 1 - least deprived, 5 - most deprived. Error bars represent the 95% confidence intervals.
Source: Statistics Canada, Canadian Index of Multiple Deprivation, 2016 and 2021. |
||||||||
| 1.1 Canadian Index of Multiple Deprivation - Economic dependency | ||||||||
| No flood | ||||||||
| Total | 2.96 | 2.95 | 2.98 | 0.01 | 3.01 | 3.00 | 3.02 | 0.01 |
| Rural area | 3.27 | 3.24 | 3.31 | 0.03 | 3.35 | 3.32 | 3.38 | 0.03 |
| Small or medium population centre (1,000 to 99,999) | 3.29 | 3.26 | 3.31 | 0.03 | 3.33 | 3.30 | 3.36 | 0.03 |
| Large urban population centre (100,000 or greater) | 2.76 | 2.74 | 2.77 | 0.02 | 2.79 | 2.78 | 2.81 | 0.02 |
| At least one flood | ||||||||
| Total | 3.20 | 3.16 | 3.24 | 0.04 | 3.06 | 3.02 | 3.11 | 0.04 |
| Rural area | 3.27 | 3.21 | 3.33 | 0.06 | 3.16 | 3.10 | 3.22 | 0.06 |
| Small or medium population centre (1,000 to 99,999) | 3.42 | 3.33 | 3.51 | 0.09 | 3.24 | 3.15 | 3.33 | 0.09 |
| Large urban population centre (100,000 or greater) | 2.94 | 2.86 | 3.01 | 0.07 | 2.78 | 2.70 | 2.86 | 0.08 |
| 1.2 Canadian Index of Multiple Deprivation - Residential instability | ||||||||
| No flood | ||||||||
| Total | 3.05 | 3.04 | 3.06 | 0.01 | 3.03 | 3.02 | 3.04 | 0.01 |
| Rural area | 2.19 | 2.17 | 2.21 | 0.02 | 2.31 | 2.29 | 2.33 | 0.02 |
| Small or medium population centre (1,000 to 99,999) | 3.29 | 3.27 | 3.32 | 0.03 | 3.36 | 3.34 | 3.39 | 0.03 |
| Large urban population centre (100,000 or greater) | 3.20 | 3.18 | 3.22 | 0.02 | 3.11 | 3.09 | 3.13 | 0.02 |
| At least one flood | ||||||||
| Total | 2.75 | 2.71 | 2.79 | 0.04 | 2.87 | 2.83 | 2.91 | 0.04 |
| Rural area | 2.19 | 2.15 | 2.24 | 0.04 | 2.38 | 2.33 | 2.42 | 0.04 |
| Small or medium population centre (1,000 to 99,999) | 3.24 | 3.16 | 3.32 | 0.08 | 3.36 | 3.28 | 3.44 | 0.08 |
| Large urban population centre (100,000 or greater) | 3.31 | 3.23 | 3.39 | 0.08 | 3.32 | 3.24 | 3.40 | 0.08 |
Conclusion
Extreme weather events and natural disasters, such as floods, can be extremely disruptive for individuals, families, and communities. This analysis examines levels of deprivation among neighbourhoods that experienced flooding, compared to areas not affected by floods to explore potential socio-economic disparities.
The recovery process following a climate-related disaster can have a range of consequences for communities. This study suggests that the post-flood socio-economic landscape is mixed, underscoring the importance of various dimensions, notably urban/rural and provincial differences. For instance, residential instability worsened following a flood, most notably in rural areas. However, improvements to levels of economic dependence were observed for communities that experienced flooding.
Future research could further delve into these intersections, as well as examine whether other extreme weather events, such as wildfires and earthquakes, follow similar patterns in risk and recovery.
Data sources, methods, definitions and limitations
This study uses data from the 2016 and 2021 Census of Population (population quintiles calculated at the Dissemination Area scale) and the Canadian Index of Multiple Deprivation (CIMD). The CIMD is derived at the Dissemination Area (DA) level, which is a small, relatively stable area of census geography composed of about 400 to 700 people. Lower scores and associated quintile rankings correspond to areas that are least deprived, while higher scores and quintile rankings relate to areas that experience greater levels of deprivation.
Publicly available flood extent maps based on satellite imagery were obtained from Natural Resources Canada’s Emergency Geomatics Services (EGS), for the years 2017 to 2020 (i.e., the between-census period). ArcGIS PRO (ESRI) was used to determine which DAs overlapped with flooded areas.
This analysis focuses on deprivation related to economic dependency, which relates to reliance on the workforce and uses indicators such as the dependency ratio (the population aged 0 to 14 and population aged 65 and older divided by the population aged 15 to 64), and the proportion of the population not participating in the labour force.
Limitations
The CIMD 2021 is a cross-sectional index and is based on 2021 Census of Population microdata. Caution must be used when interpretating comparisons between 2016 and 2021 data. Since slightly different methodologies were used to create indices for the 2016 CIMD, quintiles were used to compare across census years. As with the Census, the CIMD is subject to the same limitations, response rates, and non-response bias of the Census, and non-sampling errors.
Another limitation of the flood data is that the data are based on satellite flood extent, rather than self-reported information about whether a dwelling actually experienced a flood. This approach may not capture flooding that occurs at a below-surface level, such as flooding to a dwelling following a heavy rain period.
References
Charkraborty, Liton et al. 2022. “Exploring spatial heterogeneity and environmental injustices in exposure to flood hazards using geographically weighted regression.” Environmental Research. Vol. 210, no. 112982.
Government of Canada. 2023. Canada’s National Adaptation Strategy.
Intergovernmental Panel on Climate Change. 2021. Climate change widespread, rapid, and intensifying.
Johnston, Craig et al. 2024. “Flood risk and residential lending.” Bank of Canada. Financial Systems Hub.
Natural Resources Canada. 2024. “Floods and river ice break-up.”
Sparling, Erik et al. 2024. “Turning the tide: How flood risk transparency can drive equitable outcomes in Canada.” Canadian Climate Institute.
Statistics Canada. 2024. The Canadian Index of Multiple Deprivation: User Guide.
- Date modified: