How’s Life in the City? Life Satisfaction Across Census Metropolitan Areas and Economic Regions in Canada
by Chaohui Lu, Grant Schellenberg and Feng Hou, Social Analysis and Modelling Division, and John F. Helliwell, Canadian Institute for Advanced Research and Vancouver School of Economics, University of British Columbia
Start of text box
This Economic Insights article provides an overview of the life satisfaction expressed by individuals in census metropolitan areas and economic regions across Canada. The results are based on data from the Canadian Community Health Survey and the General Social Survey. The extent to which specific economic and social factors explain variations in life satisfaction across communities and regions is beyond the scope of this article.
End of text box
There is now international support for the measurement of subjective well-being. This includes the adoption of a United Nations resolution in 2011, the establishment of March 20 as International Day of Happiness in 2012, and the release, in 2013, of a set of OECD guidelines (OECD 2013) on the measurement of subjective well-being prepared for the use of national statistical offices. Thirty years ago, Canada was almost alone in collecting survey data on life satisfaction. As of 2014, all but three OECD countries collect some form of life evaluation, with most starting since the release of the OECD guidelines. Since 2005, the Gallup World Poll has been surveying subjective well-being in most countries around the world, thus enabling the preparation of three World Happiness Reports (Helliwell, Layard and Sachs 2015) since 2012. These compare and explain international differences in life evaluations and other measures of subjective well-being.
Among its recommendations for the measurement of subjective well-being, the OECD views life evaluation as the most important and advocates a life satisfaction question as the primary measure, with responses being given on a scale of 0 to 10. For the past several years Statistics Canada has been asking precisely this question on the Canadian Community Health Survey (CCHS) and the General Social Survey (GSS). Together, annual data from these surveys now provide almost 340,000 individual responses—enough to permit, for the first time, the preparation of comparable community-level measures of life satisfaction for 33 census metropolitan areas (CMAs) and 58 economic regions (ERs) across the country.
This article highlights these data by providing an overview of the life satisfaction expressed by individuals in CMAs and ERs across Canada. The article first presents life satisfaction scores across CMAs and ERs on an unadjusted basis; that is, without taking into account the socio-economic characteristics of individuals in those areas. Individual-level socio-economic characteristics are subsequently taken into account, reducing variations in life satisfaction across CMAs only slightly. The extent to which specific economic and social factors explain variations in life satisfaction across communities and regions is beyond the scope of this article. The main objectives here are to document the magnitude of those differences and richness of Statistics Canada data now available to explore them further.
Data for this study are taken from the five cycles of the GSS fielded from 2009 to 2013 and the four cycles of the CCHS fielded from 2009 to 2012 inclusive. CCHS and GSS respondents were asked:Note 1
- Using a scale of 0 to 10, where 0 means "Very dissatisfied" and 10 means "Very satisfied", how do you feel about your life as a whole right now?
Earlier analysis (Bonikowska et al. 2014) shows that survey respondents are able and willing to answer the question,Note 2 that their responses are not influenced by the day of the week or month in which they completed the survey, and that aggregating CCHS and GSS data into a ‘pooled’ sample is a viable way of obtaining enough responses to produce robust estimates of life satisfaction for smaller geographies or population subgroups (Frank, Hou, and Schellenberg 2014; Hou 2014).
This study is based on a pooled sample of almost 340,000 survey respondents aged 15 or older who reside in one of the 10 provinces. A respondent’s place of residence is identified as either one of Canada’s 33 CMAsNote 3 or, for those residing outside of a CMA, as their ER of residence.Note 4 In the smaller CMAs of Guelph, Peterborough and Brantford, sample sizes range from about 1,400 to 1,700, while in Abbotsford–Mission, Kelowna, Trois-Rivières, Greater Sudbury, Barrie and Saguenay sample sizes range from about 1,800 to 2,000. All other CMAs have samples of at least 2,000 respondents.Note 5 Similarly, all of the 58 ERs used for the analysis have samples of at least 1,000.Note 6 The depth of this sample is evident when one considers that the national annual samples for most countries in the Gallup World Poll are approximately 1,000.
Life satisfaction across census metropolitan areas and economic regions
Average life satisfaction from 2009 to 2013 across Canada’s 33 CMAs is shown in Chart 1. It ranges from about 7.8 (on a scale with a maximum value of 10) in Vancouver, Toronto, and Windsor, to around 8.2 in St. John’s, Trois-Rivières and Saguenay. Overall, average life satisfaction varies by 0.44 points across CMAs. This does not take into account any differences in individual-level or community-level characteristics.
An alternative way to view life satisfaction across CMAs is to identify the shares of residents who place themselves towards the top or bottom of the 10-point scale. There are no thresholds over or under which individuals are deemed to be satisfied or dissatisfied; so any such distinction is arbitrary. For illustrative purposes, the shares scoring 9 or 10, or 6 or less, are shown in Charts 2 and 3.
Across CMAs, there is a difference of almost 11 percentage points in the shares of individuals rating their life satisfaction as 9 or 10. The shares are largest in Greater Sudbury, Thunder Bay, St. John’s, Saint John and Saguenay, at 42% to 45%, and smallest in Vancouver, Toronto, Barrie and Edmonton at 34% to 35%. If the analysis is broadened to include individuals rating their life satisfaction as 8 or above (Appendix Table 1), there is a range of almost 14 percentage points across CMAs, with most of the same CMAs located at the top and bottom of the rankings when a threshold of 8 or above, or 9 or above, is used.
At the other end of the scale, there is a 9-percentage-point difference in the shares of CMA residents rating their life satisfaction as 6 or less. This proportion is smallest in Saguenay, Québec and Trois-Rivières, at less than 10%, and largest in Windsor, Toronto, Abbotsford–Mission, and Peterborough, at about 17%.Note 7
A similar range is evident across the 58 ERs considered (Chart 4). Average life satisfaction ranges from about 7.8 to 8.0 in the British Columbia ERs of Northeast, Cariboo, and North Coast and Nechako, the Alberta ER of Red Deer, the Saskatchewan ERs of Prince Albert and Northern, the Manitoba ER of North, and the Nova Scotia ER of Annapolis Valley. At the high end, average life satisfaction is about 8.3 to 8.4 in several ERs in Newfoundland and Labrador and Quebec. Overall, average life satisfaction varies by 0.56 across ERs, again without taking into account any differences in individual-level or community characteristics.
Across ERs, there is a 14-percentage-point range in the shares of residents rating their life satisfaction as 9 or 10 (from 36% to 50%), and a similar range in the shares rating their life satisfaction as 8, 9 or 10 (from 67% to 81%) (Appendix Table 1). Conversely, there is a range of about 9 percentage points in the shares rating their life satisfaction as 6 or less (from 7% to 16%).
Within the research literature it has been shown that differences in life satisfaction across communities within the same country are far smaller than differences across countries and global regions. This is because the supports for high quality of life vary much less within countries than across countries. Hence, it is not surprising that the typical difference across CMAs and ERs in Canada is only one-tenth as large as the typical difference across the 150 countries covered by the Gallup World Poll.Note 8 Nonetheless, the range of about 0.59 in average life satisfaction across CMAs and ERs is similar in magnitude to that observed between individuals who are married and divorced or separated (more on this below). Variations in the percentages of individuals at the lower and higher ends of the life scale are also considerable across CMAs and ERs, at about 10 to 17 percentage points. This raises questions about what accounts for these differences.
Taking individual-level characteristics into account
Individual-level characteristics such as age, employment status and health status have been shown to be correlated with life satisfaction (Boarini et al. 2012)and also vary across CMAs and ERs.Note 9 One question this raises is how much of the difference in life satisfaction across CMAs and ERs remains when the characteristics of their residents are taken into account?
To assess this, the correlations between life satisfaction and a standard set of socio-economic characteristics are first estimated using a multivariate linear regression model. The coefficients in Table 1 show the difference in life satisfaction associated with each characteristic relative to a reference group, net of the other characteristics in the model. The first column shows the coefficients from a base model (Model 1) run on the full sample of GSS and CCHS respondents, while in the second and third columns variables pertaining to community belonging and knowing one’s neighbours are added for respondents who were asked those questions. Overall, the results are consistent with findings in the research literature.
Life satisfaction is slightly higher among women than men, and slightly lower among immigrants than persons born in Canada. The well-documented ‘u-shape’ correlation between age and life satisfaction—with levels lower among individuals in their forties and early fifties than among those in younger and older age groups—is reflected in the age and age-squared variables. Married individuals report higher levels of life satisfaction than those who are divorced or separated, widowed or never married. Model 1 yields a negative correlation between educational attainment and life satisfaction. However, this relationship becomes positive and significant when health status, employment status and/or household income are removed from the model, confirming the now-established view that education affects subjective well-being through its impact on other outcomes. There is a strong positive and monotonic relationship between self-assessed health status and life satisfaction. Individuals rating their health as ‘excellent’ have life satisfaction scores a full point higher than those rating their health as ‘good’, and almost three points higher than those rating their health as ‘poor’. The relationship between unemployment status and life satisfaction is strongly negative, while the relationship between household income and life satisfaction is positive. Finally, life satisfaction is slightly higher among respondents who identify themselves as an Aboriginal person. However, this correlation becomes negative when other variables, such as health status, employment status and/or household income are removed from the model.
Models 2 and 3 confirm a positive relationship between life satisfaction and individuals’ feelings of belonging to their community and whether they know some or most of their neighbours.
To adjust for the individual-level characteristics shown in Model 1 of Table 1, the population characteristics of each CMA and ER are set to the Canadian average and life satisfaction scores are then recalculated.
The adjustment for individual-level characteristics generally results in very small changes in life satisfaction scores within and across CMAs.Note 10 When these characteristics are taken into account, average life satisfaction scores change by less than 0.08 in all 33 CMAsNote 11 and the range of average life satisfaction scores across CMAs decreases by 7% (or by 0.03), from 0.44 to 0.41. Similarly, adjusting for individual-level characteristics changes the share of CMA residents with life satisfaction scores of 9 or 10 by less than 2 percentage points in all 33 CMAs, and reduces the inter-CMA variation in the shares of individuals with such scores by 0.4 percentage points—from 11.3 to 10.9 percentage points—or by about 4%.Note 12 The adjustment for individual-level characteristics plays a larger role in narrowing the inter-CMA variation in the share of respondents with life satisfaction of 6 or less, reducing this from 8.5 percentage points to 7.3 percentage points or by about 14%.
Qualitatively similar results are found within and across ERs.Note 13 When individual-level characteristics are taken into account, average life satisfaction scores change by 0.10 or less in 50 of the 58 ERs, and the range of average life satisfaction scores across ERs decreases by about 16% (or by 0.09), from 0.56 to 0.47. Similarly, the share of ER residents with life satisfaction scores of 9 or 10 is reduced by 2.0 percentage points or less in 51 of the 58 ERs and the inter-ER range in the shares of residents with such scores declines from 13.7 to 13.1 percentage points—or by about 4%. At the lower end of the scale, the inter-ER range in the share of respondents with life satisfaction of 6 or less is reduced from 9.6 to 7.6 percentage points—or by about 21%—when individual-level characteristics are taken into account.
Overall, differences in the socio-economic composition of CMAs and ERs, at least as measured by the variables in Model 1, generally account for about 4% to 16% of the difference in average life satisfaction and ‘high’ life satisfaction across these geographies, and for about 14% to 21% of the difference in ‘low’ levels of life satisfaction.
Looking beyond individual-level characteristics, the results in Charts 1 to 3 appear to suggest that life satisfaction is higher in smaller communities, as most of the CMAs at the top of the rankings have populations under 250,000, while Toronto and Vancouver rank at or near the bottom. Such a relationship is reported in the literature, with Schwanen and Wang (2014, 835) noting that “...a recurrent finding is that life satisfaction and happiness are lower in denser, more urbanized settings.” But when individual-level characteristics are taken into account and smaller, mid-size and larger CMAs across Canada are examined, large within-group differences are evident. Chart 5 shows the share of CMA respondents who rate their life satisfaction as 8, 9 or 10—a broader measure than used in Chart 2—adjusted for differences in individual-level characteristics across CMAs. Across CMAs with populations of less than 250,000, the share of residents rating their life satisfaction ranges from about 65% in Guelph and Barrie to about 76% in Saguenay and Trois-Rivières. Across Canada’s five largest CMAs there is a difference of 6 percentage points between Montréal and Toronto.
Many factors may account for community-level differences in life satisfaction, and there is a growing body of international and Canadian research in this domain. This includes work that examines the role played by the physical characteristics of geographic areas, such as urban size and population density, natural endowments, economic opportunity or deprivation, and access to, and quality of, infrastructure, amenities and services (see Ballas  and Schwanen and Wang  for reviews). The social dimensions of geographic areas are also being explored. For example, using GSS data, Helliwell and Wang (2011) find evidence that the life that matters most to people is local, reflecting the levels of trust and the quality of social connections in their neighbourhoods and workplaces.Note 14 Studies have also considered the importance of social comparisons within areas, such as income relative to one’s neighbours and levels of inequality (e.g., Luttmer , Hou ). Furthermore, analyses of life satisfaction are being done at various levels of geography—across neighbourhoods, communities, provinces and states, and countries.
The extent to which economic or social factors explain geographic variation in life satisfaction appears to vary in terms of the level of geography being considered. The World Happiness Report 2015 uses six main variablesNote 15 to explain about three-quarters of the difference in average life satisfaction evaluations across countries, with income being the most important of these. Within Europe there is a smaller international range in average incomes, and social factors explain a larger share of the cross-national variation in life satisfaction. Likewise, some evidence suggests that social rather than economic factors play a greater role in explaining variations in life satisfaction among individuals and regions within countries (Helliwell and Putnam 2004; Helliwell and Barrington-Leigh 2010). Identifying the factors that account for the inter-CMA and inter-ER variations in life satisfaction shown above lies beyond the scope of this article and are topics warranting further research.
In Canada, rich information on life satisfaction is now available. The five cycles of the GSS and four cycles of the CCHS used for this study provided a sample of almost 340,000 respondents, and the addition of upcoming cycles would increase that to over 450,000. This offers scope for studying life satisfaction among population subgroups or among small geographies. And while this study has looked at life satisfaction across CMAs, it would also be feasible to look more closely at it within CMAs. As well as exposing the variety of life experiences within CMAs, this further disaggregation would increase the total number of geographic areas included in the search for fuller understanding of what community-level characteristics tend to support more satisfying lives.
Ballas, D. 2013. “What makes a ‘happy city’?” Cities 32 (Supplement 1): S39–S50.
Boarini, R., M. Comola, C. Smith, R. Manchin, and F. de Keulenaer. 2012. What Makes for a Better Life?: The Determinants of Subjective Well-being in OECD Countries – Evidence from the Gallup World Poll. OECD Statistics Working Papers, 2012/03. Paris: OECD Publishing.
Available at: http://dx.doi.org/10.1787/5k9b9ltjm937-en.
Bonikowska, A., J. Helliwell, F. Hou, and G. Schellenberg. 2014. “An assessment of life satisfaction responses on recent Statistics Canada surveys.” Social Indicators Research 118 (2): 617–643.
Frank, K., F. Hou, and G. Schellenberg. 2014. Life Satisfaction among Recent Immigrants in Canada: Comparisons with Source-country Populations and the Canadian-born. Analytical Studies Research Branch Research Paper Series, no. 363. Statistics Canada Catalogue no. 11F0019M. Ottawa: Statistics Canada.
Helliwell, J.F., and C.P. Barrington-Leigh. 2010. “Viewpoint: Measuring and understanding subjective well-being.” Canadian Journal of Economics 43 (3): 729–753.
Helliwell, J.F., and H. Huang. 2010. “How's the Job? Well-Being and Social Capital in the Workplace.” Industrial and Labor Relations Review 63 (2): 205–227.
Helliwell, J.F., R. Layard, and J. Sachs, eds. 2015. World Happiness Report 2015. New York: Sustainable Development Solutions Network: A Global Initiative for the United Nations. Forthcoming.
Helliwell, J.F., and R.D. Putnam. 2004. “The social context of well-being”, Philosophical Transactions of the Royal Society B: Biological Sciences 359 (1449): 1435–1446.
Helliwell, J.F., and S. Wang. 2011. “Trust and Wellbeing.” International Journal of Wellbeing 1 (1): 42–78.
Hou, F. 2014. “Keep up with the Joneses or keep on as their neighbours: Life satisfaction and income in Canadian urban neighbourhoods.” Journal of Happiness Studies 15 (5): 1085–1107.
Luttmer, E.F.P. 2005. “Neighbors as negatives: relative earnings and well-being.” The Quarterly Journal of Economics 120: 963–1002.
Organisation for Economic Co-operation and Development (OECD) 2013. How’s Life? 2013: Measuring Well-being. OECD Publishing.
Available at: http://dx.doi.org/10.1787/9789264201392-en
Schwanen, T., and D. Wang. 2014. “Well-being, context, and everyday activities in space and time.” Annals of the Association of American Geographers 104 (4): 833–851.
- Date modified: