Let me confess at the outset to being a grouch about happiness, or rather about the new discipline of ‘happiness studies’. Although economists have driven forward research into the question of what it is that increases our well-being, my suspicion is that it is people who dislike the rigour of economics who are the most enthusiastic advocates of gearing public policy towards maximising happiness rather than economic growth. If growing richer makes us no happier, then instead of worrying about incentives and enterprise, we can – well, insert whatever paternalistic policy you like. One popular candidate is a highly progressive income tax, or perhaps a tax on luxuries, to force people out of the money-grubbing rat-race which is making them miserable. 

This marvellous, sharp pamphlet by Helen Johns and Paul Ormerod, surveying the relevant research, pinpoints the flaws in the analysis which leads some happiness gurus to this kind of conclusion. Take the main piece of evidence offered by those who would downgrade economic growth as a policy priority: the absence of a positive correlation between GDP per capita and happiness, at least above a level of income of about $20,000 a year. As the authors point out, this is at odds with other data which indicates that in any society at a given time, rich people are happier than poor people. In the longitudinal rather than the time series data, there is a positive correlation between money and happiness. 

They go on to demonstrate that it is the way the happiness ‘data’ are constructed which generates the absence of any correlation in the time series. Happiness is measured by the proportions of respondents to sur-veys who describe themselves as ‘very happy’, ‘pretty happy’ or ‘not too happy’. On this 1-3 scale the average for the UK is 2.2 and has remained round that level for decades even as GDP per capita has increased. However, as the authors point out, the most the happiness index could ever increase from its current level (if all respondents became ‘very happy’) is 35 per cent. Data constructed in this way simply have different time series properties to an economic variable such as GDP. Unsurpris-ingly, then, happiness is not correlated over time with lots of other variables, such as real public spending per capita, income inequality, or crime rates. 

More weight can be placed on the kind of panel data evidence (reported by Professor Andrew Oswald amongst others) which shows happiness is positively related to marriage, good health, religious faith, community bonds and having a job. Alas, it is much harder for happiness gurus to recommend sweeping policy changes arising from these results. This is not to say that policy should just ignore our well-being. One could reasonably regard it as valid for the government of a wealthy society to have a broader set of aims than growth alone, and to use the evidence from psychological and economic research to assist them in setting policy. But this is a long way from the enthusiasm in some quarters for policies that will force us to be happy whether we like it or not. 

Enlightenment Economics