Happiness and Inequality Revisited
I was correctly chided for my earlier post on the connection between inequality and happiness, and thanks for the comments. My thinking was also clarified by hearing Wilkinson deliver a fabulous lecture to the Ottawa Economics Association this week.
Wilkinson and Pickett’s central argument is that the connection between income inequality and a wide range of measures of well-being – from health to fear of crime – runs through stress and anxiety. People in more equal and solidaristic societies enjoy better outcomes, and might also be expected to be more satisfied with their lives. Objective and subjective well being should overlap to a considerable degree.
I find on a cursory examination that this indeed seems to be broadly true for Canada.
The Tables below present data on income inequality and on happiness by province. The former series – taken from HRSDC social indicators – is the ratio of the top to bottom quintile of after tax family income. Provinces are ranked from most to least equal. The latter – taken from the recent CSLS study – is the percentage of persons saying they are very satisfied or satisfied with their lives.
The second Table shows that the order of provinces ranked from most to least equal corresponds fairly closely with the order of provinces ranked from most to least happy. It is notable that the two pole provinces are the same – PEI and Newfoundland are the most equal and the most happy – while Ontario and BC are the least equal and the least happy.
Some enterprising person with more time and methodological talent than I should use the HRSDC social indicators to do Wilkinson and Pickett style gradients for the Canadian provinces linking income inequality to health, crime and other outcomes.
| Inequality | Happiness | |
| PEI | 6.2 | 94.08 |
| NL | 7.9 | 93.44 |
| NS | 7.9 | 91.70 |
| NB | 7.6 | 93.28 |
| Que | 8.1 | 92.13 |
| Man | 8.1 | 91.95 |
| Sask | 8.8 | 92.23 |
| AB | 8.9 | 92.11 |
| ON | 9.3 | 90.18 |
| BC | 10.1 | 90.57 |
| Rank Order | Inequality | Happiness |
| PEI | 1 | 1 |
| NL | 2 | 2 |
| NS | 3 | 8 |
| NB | 4 | 3 |
| Que | 5 | 7 |
| Man | 6 | 6 |
| Sask | 7 | 4 |
| AB | 8 | 5 |
| ON | 9 | 10 |
| BC | 10 | 9 |
Good idea, Andrew, starting out a new project soon looking at the declining middle class. I do think the various outcomes could be grouped in a multivarIate dimension and reclassified then compared back to these measures.
Very illuminating data, Andrew. I will definitely use this in a McGill social work entitled Poverty and Inequality, that I teach. I am sure you are also aware of Linda McQuaig and Neil Brooks new book on poverty and inequality, The Trouble with Billionaires. If anyone is interested, here’s the link to a short review I have written of it: http://www.montrealgazette.com/entertainment/Linda+McQuaig+Neil+Brooks+take+economic+inequality+Trouble+with/3918953/story.html?id=3918953
I’m reluctant to let politicians appropriate the goal of happiness – they will almost certainly just talk more nonsense. But if you really want to cite the CSLS survey, you might want to cite their conclusion:
“We find that evidence from Canada suggests that mental health status, sense of belonging, physical health, and stress level are more
significant determinants of happiness than household income.”
Not much grist for the economic mill.
not sure of RCP has ever heard of intervening variables, but I certainly think income has a root causation in each of the variables mentioned, and secondly we are talking about inequality.
And hey, only a progressive would even think of trying to measure happiness, so if we can measure the notion of profit down to ever bean counting molecule of cost, then why can we not at least try and measure happiness. (armies and armies of accountants and lawyers measuring cost, but no one anywhere measuring happiness, too bad we have such poor design to our social institutions.)
Thanks, Paul, but I do know some stats. Since the CSLS study is a regression study, I think you might be hinting at some collinearity, and I agree that could be the case. Nonetheless, what I quoted is what they concluded.
And I am really, really dubious about designing social policy based on average self-reported happiness rather than something a little more objective like the proportion of people below the low-income cutoff. You may disagree.
I think the csls conclusion might be modified to say the distribution of income is more important to subjective well being than the average level.
RCP – I prefer objective indicators as well, and note that people in very hierarchical societies may well be happy – as in Brave New World
Andrew: I was thinking of “The Ones Who Walk Away From Omelas”, but Brave New Wold is a good example too.
income is the root, now we could all live in poverty and be equal in terms of happiness, ie.e unhappiness. But at some point along the income scale, the others kick in to contribute to happiness, and then the rest drives off of income. So no, I would not say collinear, I would say intervening and then would display some kind of correlation. I do not think it is actually linear, although I do accept we all wanna think in straight lines, which is fine i guess unless one is living somewhere along that lineear where it is not actually linear.
I agree low income cut offs are good bit more objective.
However if we truly had a society that actually focused on people, we would have a measure that indeed would go a long ways more into defining happiness, free from hunger, angst, stress, shelter and some other measures that rarely make the national spreadsheet. during these times of economic decline, they are more important than ever.
licos are not sufficient