Richard Florida
by Richard Florida
Fri Feb 19th 2010 at 8:15am UTC

What Makes Happy Cities Happy


Earlier this week, I discussed the new Gallup-Healthways Well-Being Index of happy cities. Today, with the help of my Martin Prosperity Institute colleague Charlotta Mellander, we take a look at some of the social, demographic, and economic factors that are associated with the happiness and well-being of cities.

There has been considerable debate on the factors that are associated with happiness and well-being at the national level. The well-known Easterlin Paradox suggested that happiness tends to level off after a certain income threshold. Psychologists, notably Edward Diener, have argued that factors such as health, challenging work, and close social relationships, among others, play a considerable role in happiness. Some have even made the case for instituting a new measure of gross national happiness to supplement conventional metrics like gross national product.

Recent studies by Princeton University’s Angus Deaton and Justin Wolfers and Betsy Stevenson of the University of Pennsylvania’s Wharton School question the Easterlin Paradox and indicate a closer link between happiness and income across nations. Carol Graham raises the enigma of the “happy peasant and the miserable millionaire” as a way to resolve this apparent paradox. Graham suggests that happiness is relative to one’s position in society. Take unemployment for example. Unemployment is crushing for previously employed people in places where gainful employment is the norm. But people in poor countries where unemployment is more the norm find other ways to be happy.

The Gallup-Healthways is the first comprehensive data set we know of that tracks happiness and well-being at the metropolitan level, providing data from a large-scale survey of individuals across 185 metro regions. We look at the associations between the Gallup-Healthways Metro happiness index and key social, demographic, and economic factors. Data-matching reduces the size of our sample to 170 metros – roughly half of all U.S. regions. As usual, we point out that our analysis points only to associations between variables. It does not specify causation or the causal direction of those associations which are questions for future research. Still, the results are interesting across several dimensions.

Income, Wages, and Output: So what is the relationship between metro-level happiness and income, wages, and output? The correlation analysis suggests a moderate relation between wages (.45), income (.4), and economic output per capita. The scatter-graphs below show the relationships are reasonably linear, though there is a better fit for wages and income than for output per capita.


Unemployment: Conventional wisdom and academic studies suggest that a rising unemployment rate would take a big toll on happiness. We find a moderate effect across U.S. metros. The correlation between happiness and the unemployment rate is -.34 and between it and the year-over-year (December 2008 to December 2009) change in unemployment is -.3.



Post-Industrial Economic Structures: In ongoing research, we have been testing the notion that happiness and well-being may be more associated with key features of so-called post-industrial economic structures – namely the shift from physically oriented work to knowledge, professional, and creative occupations and industries – and from lower-skilled to more highly skilled and educated workforces. A large body of research has found a close association between human capital (measured as share of the population with a B.A. and above) and economic development across nations as well as regions; other research has found that human capital levels are becoming more divergent across regions over time. To get at this, we looked at the associations between happiness and human capital, as well as between it and creative-knowledge-professional occupations and blue-collar working class occupations.

Human Capital: Happiness at the city or metro-level is more closely associated with human capital with a correlation of .68 – the strongest correlation of any of the variables we looked at. The scatter-graph below shows a fairly linear relationship.


Creative Class: Happiness is also associated with the creative class, a correlation of .45. The scatter-graph below shows a fairly linear relationship.


High-Tech: Happiness is also associated with locations that have higher concentrations of high-tech industries. We find a correlation of .41 between it and the Milken Institute’s Tech-Pole measure.

Working Class: On the other hand, metro-level happiness is negatively associated with the working class, -.34, a finding which is similar to that for states.

4 Responses to “What Makes Happy Cities Happy”

  1. Mike L. Says:

    Great work, RF! Mentally stimulating!
    The Human Capital plot suggests there are 3 parallel trend lines. Upper, Middle (drawn) and Lower. They are about 2.00 Well-being points or .05 Human Capital points apart.
    Do you have a theory about what distinguishes these three groups substantively?

  2. Colin Says:

    Interesting analysis! Really shouldn’t be done as a bunch of separate regressions though. Have you tried putting them all into one regression to see what the relative effects are and which are significant? Multicollinearity might be an issue but still worth trying.

    If you want me to run it let me know.


  3. Daniel Carins Says:

    Very interesting…

    I’d be suspicious of any self-assessed metric of “well-being” or “happiness”, and also suspicious of any that used referrals to psychiatrists or some other measure of mental health, simply because contemporary, western society seems to place a lot of kudos on being “instable” as a perverse indicator of “being interesting”.

    Happiness as a concept is also fundamentally bogus – as the article points out indirectly, it’s relative and ephemeral by definition. Can you imagine being permanently happy? No, because then it would be ordinary, so you’d need even greater enjoyment to register the difference. I imagine this explains the diminishing returns over a certain income level.

    Fulfilment is a far better aim, but how do you measure that? It may be that indicators of fulfilment – low churn in neighbourhoods and jobs to think of two possible ones – may actually work against existing economic growth models based on mobility of capital and labour.

  4. Leslie M. Saunders Says:

    I’m sure there will be many who debate whether or not it is possible to measure happiness. I’m not one of them because I appreciate the fact that the effort is being made. Measuring the happiness quota of a geographical region may begin with satisfaction regarding income and work fulfillment. Nevertheless, in years to come it may also include following trends related to people’s willingness to move into more diverse community settings; the volunteer contribution people make to the communities-at-large in which they choose to live; measuring random acts of kindness and inclusion compared with random criminal activities, the decrease of stress-related illnesses, etc. Granted, this will sound incredibly hokey to some but measuring the dimensions that indicate whether a city’s spiraling dynamics are moving up, down or not at all seems worthy of further scientific investigation.