What is the Local News Data Hub?
The Local News Data Hub’s goals are to i) shore up local journalism across Canada by supplying newsrooms with data-informed stories ii) train student data journalists and iii) collaborate with individual journalists and news outlets on local data journalism stories and projects.
The Data Hub’s reporting team, consisting of student data journalists and student data analysts supervised by faculty at Toronto Metropolitan University, identifies government and other data with the potential to generate local stories. Once a dataset is identified and the basic reporting is completed, the Data Hub team writes a story template that is then customized with data and content for specific communities. The local stories are posted on the LNDH website and also shared with the Canadian Press wire service for distribution to CP clients across the country. Find out more about the Local News Data Hub here.
Introduction and background
Our goal with this project was to produce national and local stories examining income inequality and its consequences in different communities across Canada using the 2020 after-tax Gini coefficient as calculated by Statistics Canada. The Gini coefficient is an internationally recognized measure of how income is spread across a society. It is calculated based on Statistics Canada’s after-tax income data, which includes employment earnings, investment earnings such as dividends and interest, and government transfers including social assistance and pensions. It does not include inherited or saved wealth, one-time winnings (e.g. lottery), capital gains or non-financial assets such as real estate.
We looked at inequality data for the country overall, for census metropolitan areas (CMAs), and for census subdivisions (CSDs), which is Statistics Canada’s term for towns, cities and rural municipalities.
We assembled a list of all 418 CSDs that have populations greater than 10,000 people, and then ranked them using the Gini coefficient. A Gini value can range from zero to one, with one representing a population where one household or person is making all of the income in the municipality, while a Gini value of zero means that all people/households within a given community have the same annual income.
We assigned a rank to each municipality. The place with the highest Gini coefficient had the most income inequality/the greatest variation in income and therefore was ranked number one on the list. The municipality that ranked 418th on the list was the place with the least income inequality/least income variation. During the ranking process, we found many municipalities had the same Gini coefficient, so we recorded those as “ties.” Statistics Canada, however, publishes the Gini data with only three decimal places. If more decimal places were provided, it is possible that they may reveal marginal differences between “tied” CSDs.
The stories in this series examine income inequality as it plays out in municipalities that fall within the census metropolitan areas of Toronto, Calgary, Vancouver and Montreal. We produced a national story that explored why income inequality varies so much among larger cities with populations greater than 200,000. The series also includes a story that focuses on St. John’s, N.L. and Atlantic municipalities, as well as stories that look at income inequality in Oakville, Ont. and Canmore, Alta.
Definitions
After-tax income: After-tax income refers to total income less income taxes during a specified reference period, or actual take home pay from all income sources.
Census division (CD): Group of neighbouring municipalities joined together for the purposes of regional planning and managing common services (such as police or ambulance services). Census division (CD) is the general term for provincially legislated areas (such as county, municipalité régionale de comté (MRC) and regional district) or their equivalents. Census divisions are intermediate geographic areas between the province/territory level and the municipality (census subdivision).
Census metropolitan area (CMA): A census metropolitan area (CMA) is formed by one or more adjacent municipalities centered on a population centre (known as the core). A CMA must have a total population of at least 100,000 based on data from the current census, of which 50,000 or more must live in the core based on adjusted data from the previous census.
Census subdivision (CSD): Census subdivision is the Statistics Canada unit of geography used for this study. It is Statistics Canada’s term for a municipality. Please see the Statistics Canada definition for more information.
Gini coefficient: The Gini coefficient measures income inequality, and ranges from zero to one. The closer the Gini value is to zero, the lower the level of inequality. A Gini measure of zero represents perfect equality, where all members of the selected area take home the same income. A Gini measure of 1 represents perfect inequality, where one person makes all of the income in a given area. The Gini coefficient used for this study is based on adjusted after-tax household income per Statistics Canada. It does not take into account other contributors to wealth such as earnings from capital gains, saved or inherited wealth, one-time receipts of lump-sum funds (such as lottery winnings) or non-financial assets such as real estate.
Household: Household refers to a person or group of persons who occupy the same dwelling and do not have a usual place of residence elsewhere in Canada or abroad. The household may consist of a family group such as a census family of two or more families sharing a dwelling, a group of unrelated persons or a person living alone.
Market basket measure (MBM): This is Canada’s official measure of poverty based on the cost of a specific basket of goods and services representing a modest, basic standard of living developed by Employment and Social Development Canada (ESDC). The MBM thresholds represent the costs of specified qualities and quantities of food, clothing, shelter, transportation and other necessities for a reference family of two adults and two children. In the stories, “in poverty” refers to the number of people living in poverty in a given area according to the MBM, while “poverty rate” refers to the proportion of people in a given area that meet the MBM poverty threshold.
Median after-tax household income: The middle number in a distribution of after-tax household incomes among households in a municipality.
P90/P10 ratio: The P90/P10 ratio is a measure of inequality. It is the ratio of the 90th and the 10th percentile of adjusted household after-tax income. The 90th percentile means 90 per cent of the population has income that falls below this threshold. The 10th percentile means 10 per cent of the population has income that falls below this threshold.
Total income: Total income is the income from the workforce as well as dividends and interest from investments, child and spousal support payments, scholarships, and income from government sources including social assistance and pension benefits. It excludes earnings from capital gains, saved or inherited wealth, and one-time receipts of lump-sum funds (such as lottery winnings).
Wealth: A measure of financial wellbeing that takes into account labour market earnings, government transfers, savings, life insurance policies, capital gains, non-financial assets (like real estate), inheritance, and debts to determine overall net worth. For the purposes of this study, wealth captures both income and all non-income funds, including savings, assets and debts that are not captured by Gini values.
Analyzing the data
Data sources
For this series of stories we used the following Statistics Canada tables:
- Table 98-10-0097-01 – Income inequality statistics across Canada: Canada, provinces and territories, census metropolitan areas and census agglomerations with parts (includes Gini Coefficients and P90/P10 data for Census Metropolitan Areas)
- Table 98-10-0096-01 – Income inequality statistics across Canada: Canada, provinces and territories, census divisions and census subdivisions (includes Gini Coefficients and P90/P10 data for Census Subdivisions)
- Table 98-10-0113-01 – Individual Market Basket Measure poverty status by economic family characteristics of persons: Canada, provinces and territories, census divisions and census subdivisions
- Table 98-10-00601 – Household income statistics by dwelling and household characteristics: Canada, provinces and territories, census divisions and census subdivisions
Choosing our variables
On the advice of Xuelin Zhang, a Statistics Canada senior research analyst with expertise in income and poverty statistics, we chose the Gini index for adjusted after-tax household income for this income inequality analysis. The after-tax Gini index is the most accurate reflection of actual take-home pay that Statistics Canada provides on a large scale, as it includes earnings from the workforce as well as dividends and interest from investments, child and spousal support payments, scholarships, and income from government sources including social assistance and pension benefits. It excludes earnings from capital gains, saved or inherited wealth, minus income taxes.
We used the P90/P10 variable to understand the size of the gap between the highest and lowest earning households in a census subdivision. While the Gini coefficient was a useful statistical measure for capturing income variation, it did not reveal any information about the specific patterns of income disparity or the actual incomes of people in the area. The P90/P10, which compares the incomes of the highest-earning and lowest-earning households, provides a more concrete way of understanding income inequality and polarization in a census subdivision.
We also looked at Market Basket Measure data to understand the number of people living in poverty in a given area and the prevalence of poverty.
In the end, we produced:
- A national story that focuses on income inequality in Canada’s largest cities where the disparities tend to be higher because they typically have more heterogeneous populations with more opportunities for very high- or very low-wage work.
- Stories about the divide between high- and low-income earners in municipalities within the larger CMAs of Calgary, Montreal, Toronto, and Vancouver to get a better sense of variations in income inequality within regions.
- Within these CMAs, we examined the following census subdivisions in greater detail: Aurora, Ont.; Brampton, Ont.; Mississauga, Ont.; Oakville, Ont.; Toronto, Ont.; Richmond Hill, Ont.; King Township, Ont.; Calgary, Alta.; Rocky View County, Alta.; Foothills County, Alta.; White Rock, B.C.; City of Vancouver, B.C.; West Vancouver, B.C.; Metro Vancouver A electoral area, B.C.; District of North Vancouver, B.C.; City of North Vancouver, B.C.; Richmond, B.C., Surrey, B.C.; Saint-Amable, Que.; Saint-Lin–Laurentides, Que.; Sainte Sophie, Que.; Mont Royal, Que.; Côte-Saint-Luc, Que.; Montreal, Que.; Beaconsfield, Que.; Westmount, Que.; Mirabel, Que.; and Sainte-Marthe-sur-le-Lac, Que.
- A story about St. John’s, N.L. because it stood out among municipalities in Atlantic Canada, for its higher level of income inequality compared to other East Coast communities. We used St. John’s as the focal point while also providing a snapshot of inequality patterns in the Atlantic region.
- A story about Canmore, Alta. because it stood out among smaller municipalities for its high level of inequality and its high cost of living, and because it illustrates how income inequality can play out in a tourism-oriented town.
- A story about Oakville, Ont. because we wanted an example of how inequality plays out in suburban and commuter cities outside of Toronto.
Applying filters
We began with Statistics Canada’s table of all CSDs (Table 98-10-0096-01), and then applied a filter so that any CSDs with fewer than 10,000 people were excluded. The rationale for this was that in smaller samples the Gini index tends to skew lower than in larger samples, making it more difficult to accurately compare inequality in places with vastly different populations. Another reason income inequality is less pronounced in smaller places is that their local economies tend to be dominated by relatively few industries/companies that offer very few really high or low-paid positions.
We also filtered the table to create a dataset for municipalities with more than 200,000 people to explore differences in income inequality in larger cities in Canada.
Incorporating multiple datasets
We used Statistics Canada census profiles to explore other income-related data including median after-tax income, population totals, and CSD geographical designation (city, town, township, etc.). We used Statistics Canada Table 98-10-0113-01 to incorporate poverty rates as per the Market Basket Measure.
About the Local News Data Hub analysis
Our final dataset included five sheets:
- A sheet displaying all CSDs with more than 10,000 people in Canada, with columns for population and after-tax household Gini coefficients ranked from highest Gini score to lowest.
- A sheet that isolates the 30 most unequal and most equal CSDs, according to the after-tax Gini coefficient, with the province in which each CSD is located, and the P90/P10 ratio for each CSD on the list.
- A sheet listing the Gini coefficient for CSDs with population greater than 200,000 to examine levels of inequality in the larger cities in the country.
- A sheet that includes all of the poverty and income variables for the CSDs with populations greater than 10,000. Columns include:
- CSD/municipality name
- Population in private households
- Gini index for adjusted after-tax household income
- P90/P10 on after-tax household income
- People in poverty as per Market Basket Measure
- Prevalence (per cent) of poverty as per the Market Basket Measure
- Median after-tax income of households in the CSD
- A sheet showing poverty and demographics in highlighted CSDs. This sheet provides more detail for each of the CSDs we planned to write about in our stories. Columns are the same as the previous sheet, but also include:
- Designation: the classification of the geographic area according to Statistics Canada (city, town, township, district municipality, etc.)
- CMA/CD: the larger geographic area to which the municipality belongs
Key findings
Table 1. The CSDs (population greater than 10,000) with after-tax Gini coefficients that ranked the highest in the country
| Inequality ranking | Census subdivision | Province | Gini index using adjusted after-tax household income |
|---|---|---|---|
| 1 (most unequal) | Westmount | Quebec | 0.537 |
| 2 | West Vancouver | British Columbia | 0.51 |
| 3 | Metro Vancouver A | British Columbia | 0.472 |
| 4 | Mont-Royal | Quebec | 0.457 |
| 5 | Rocky View County | Alberta | 0.417 |
| 6 | Saint-Lambert | Quebec | 0.39 |
| 7 | King | Ontario | 0.384 |
| 8 | Foothills County | Alberta | 0.383 |
| 9 | Toronto | Ontario | 0.371 |
| 10 | Oakville | Ontario | 0.366 |
| 11 | Vancouver | British Columbia | 0.361 |
| 12 | Côte-Saint-Luc | Quebec | 0.359 |
| 13 | Oak Bay | British Columbia | 0.354 |
| 14 | Canmore | Alberta | 0.35 |
| 15 | White Rock | British Columbia | 0.349 |
| 16 | Saltspring Island | British Columbia | 0.348 |
| 17 | District of North Vancouver | British Columbia | 0.346 |
| 17 | Beaconsfield | Quebec | 0.346 |
| 19 | Aurora | Ontario | 0.332 |
| 20 | Bromont | Quebec | 0.33 |
| 21 | Niagara-on-the-Lake | Ontario | 0.329 |
| 22 | Calgary | Alberta | 0.327 |
| 22 | Richmond Hill | Ontario | 0.327 |
| 24 | Waterloo | Ontario | 0.324 |
| 24 | City of North Vancouver | British Columbia | 0.324 |
| 26 | Saint-Sauveur | Quebec | 0.323 |
| 27 | Richmond | British Columbia | 0.321 |
| 28 | St. John’s | Newfoundland and Labrador | 0.32 |
| 29 | Montreal | Quebec | 0.318 |
| 30 | Red Deer County | Alberta | 0.317 |
| 30 | Mont-Tremblant | Quebec | 0.317 |
| Canada | National | 0.302 |
Table 2. Ranking for the 30 CSDs (population greater than 10,000) in Canada where the gap between
the highest- and lowest-earning income households was least pronounced
| Inequality ranking | Census subdivision | Province | Gini index using adjusted after-tax household income |
|---|---|---|---|
| 389 | Châteauguay | Quebec | 0.233 |
| 389 | Beloeil | Quebec | 0.233 |
| 389 | Morinville | Alberta | 0.233 |
| 392 | Repentigny | Quebec | 0.232 |
| 392 | The Nation | Ontario | 0.232 |
| 394 | Vaudreuil-Dorion | Quebec | 0.231 |
| 394 | L’Ancienne-Lorette | Quebec | 0.231 |
| 396 | Terrebonne | Quebec | 0.23 |
| 396 | Mirabel | Quebec | 0.23 |
| 398 | L’Assomption | Quebec | 0.229 |
| 398 | Pincourt | Quebec | 0.229 |
| 398 | Bois-des-Filion | Quebec | 0.229 |
| 401 | Saint-Basile-le-Grand | Quebec | 0.228 |
| 402 | Cantley | Quebec | 0.226 |
| 403 | Mercier | Quebec | 0.225 |
| 404 | Mascouche | Quebec | 0.223 |
| 404 | Russell | Ontario | 0.223 |
| 404 | Martensville | Saskatchewan | 0.223 |
| 407 | Lavaltrie | Quebec | 0.221 |
| 408 | Sainte-Julienne | Quebec | 0.22 |
| 409 | Saint-Colomban | Quebec | 0.219 |
| 409 | Sainte-Catherine | Quebec | 0.219 |
| 411 | Sainte-Anne-des-Plaines | Quebec | 0.218 |
| 411 | Marieville | Quebec | 0.218 |
| 413 | Saint-Constant | Quebec | 0.217 |
| 414 | Sainte-Marthe-sur-le-Lac | Quebec | 0.215 |
| 415 | Sainte-Sophie | Quebec | 0.212 |
| 415 | Petawawa | Ontario | 0.212 |
| 417 | Saint-Lin-Laurentides | Quebec | 0.209 |
| 418 | Saint-Amable | Quebec | 0.207 |
