The measurement of happiness has become vogue in economics and public policy. But is there any real use to it, or is it just a gimmick?
In May 2006, David Cameron, then recently-elected leader of the Conservative Party, told a Google-sponsored conference that: “it’s time we focused not just on GDP, but on GWB – general well-being”. Since 2010, the UK’s Office of National Statistics has followed through on this idea with regular surveys of the nation’s wellbeing wherein self-reported life satisfaction (strongly related to but not synonymous with ‘happiness’) has played an important part. So too have a number of other nations: USA, France, Canada and so on.
I can see the rationale. Whereas more objective measures such as GDP per capita and the like have long been traditional goals of public policy, they can be seen not as ends in themselves but rather a means to creating a contented and satisfied population – the truly valuable ‘output’ of an economy. So if happiness is the end surely it is best to measure it? Not least because, in one number, it can encapsulate the benefits of a large number of difficult-to-measure intermediate ‘goods’ – e.g. security, cultural richness, nice environment and the like.
The positive news for such an approach is that in many ways studies on happiness are consistent with some of the traditional economic measures in a way that makes intuitive sense. For instance, the results of a Gallup World Poll of life satisfaction indicate that, “high-income countries have greater life satisfaction than low-income countries” even if, for the richest countries there is little relationship between income and happiness. [1]
Positive news also is that measures of happiness are in line with aspects of society that are often seen as subjectively ‘good’ in their own right. For instance, a study of the effect of democratic participation on happiness conducted in Switzerland (where the large number of decisions put to popular vote and the measureable differences in ability to vote among different cantons allow quantitative analysis) shows that “direct democracy…and federal structure…systematically and sizeably raise self-reported individual well-being”. [2] More controversially, given its political implications, is a purported link between equality and health and happiness. Kate Pickett and Richard Wilkinson’s 2010 book ‘The Spirit Level’, was a best-selling presentation of this idea.[3]
Despite some of these links, I think it’s clear that using happiness as a measure of wellbeing and thus as a tool for policymaking has some serious weaknesses. For one thing, any measure of happiness is, by definition, subjective. It is therefore not possible to disentangle cultural norms from the answers received: in short, if an Englishman tells a survey that he is ‘fairly happy’ is it possible that he is feeling the same emotion as a Swede who is ‘very happy’? Unless I’d just won the Euromillions rollover on the day West Ham signed Lionel Messi, ‘not too bad’ is the best you’ll usually get from me.
One puzzle that underlines this difficulty is the issue of suicide. In criminology, homicide rates are often used as a benchmark of crime because – regardless of culture and of time – it is one offence that is always reported and recorded. By analogy, suicide could be seen as a benchmark of unhappiness – no matter what a subjective survey answer says about happiness, it is certain that an individual who commits suicide is not happy. But there is very little relationship between high reported levels of happiness and low suicide rates – indeed “the happiest places have the highest suicide rates” [4]
In a less dramatic way, reported happiness levels show some other odd characteristics. For instance, looking at data from the USA, the level of happiness at a point in time for a cross section of the population is positively related to income: high income means high happiness. However, if a cohort (a group of individuals selected for study) is followed though time, its satisfaction levels do not grow despite that fact that its income grows with time (as individuals become more successful and senior within their chosen employment). What’s more, when people are asked to compare their life satisfaction at the time of the survey with their satisfaction in the past they claim they are happier than they were. Likewise, when asked to predict how they will feel in the future they respond that they will be even happier. Taken together, these three observations – positive effect of income at a point in time; constant satisfaction for a cohort moving through time; and an upwards-sloping curve of expectation and recollection of satisfaction – are completely inconsistent. [5] It’s a paradox.
The economist Richard Easterlin proposed a conceptual framework to explain this inconsistency. [6] If individuals have a set of material aspirations (desire to own a hi-fi, own a car, own a home etc.) then, as more items on the list are ‘crossed off’, the more satisfaction as person could be expected to feel. If people start off with similar aspirations (an assertion that he provides evidence for by referencing a study showing the almost identical ‘wish list’ of young Americans) then higher income would lead to higher satisfaction. It would also explain the observation about past versus present, and present versus expected future satisfaction: compared to an individual’s current state and current aspirations, the past may well have been worse and, if the individual expects a higher future income, the future would be better.
But what of the observation that a cohort’s actual satisfaction does not change over time? To answer this, Easterlin suggests that, as individuals become wealthier, their aspirations alter. It takes more to make them happy: instead of any car, they need a fast car; instead of any home, they need a bigger one in a better neighbourhood. If shifting aspirations keep pace with additional income, then, Easterlin proposes, over time a cohort’s satisfaction will remain unchanged. Anyone who has ever worked in the City or on Wall Street will be entirely familiar with this feeling. My first ever bonus (which caused unbridled joy at the time) would have met with a disrespectful sneer by the time of my retirement.
Seen on a macro scale, this framework would also explain aspects of countries’ self-reported happiness: after a certain point, extra national wealth does not lead to extra satisfaction (and may erode it slightly). Happiness levels in most rich countries have flat lined despite rising wealth. [7] In my view, these insights relate to the work in the field of psychology by Maslow – in particular his ‘Hierarchy of Needs’ wherein he theorised that “the appearance of one need usually rests on the prior satisfaction of another more [predominant] need”. [8] Poorer countries are fulfilling lower level needs as they get wealthier but, after a certain point, material needs are fundamentally satisfied and Easterlin’s ‘shifting aspirations’ model takes over thus reducing the connection between income and satisfaction. It would also explain the link between democratic participation and happiness if this is a higher order need for an already rich country.
Overall though, the use of happiness as a macro measure of well-being and as a guide to policy seems badly flawed, at least once a country reaches a certain level of material development: it is simply too unresponsive to changes in other variables and lacks explanatory power for some social ills (e.g. suicide). But there is one way that measuring happiness may be useful. Micro studies that explore how various life events affect happiness may have some use. For example, Blanchflower and Oswald’s study of wellbeing in the UK and the US shows that the happiness impact of unemployment and divorce “are large” (unsurprisingly). The dollar income needed to create the same happiness as that of “a lasting marriage…is estimated to be worth $100,000 a year”. [9] Sounds a bit low in my case; but since there is a chance my wife might read this, I am rather axed to say so.
But what studies like this do suggest is that focusing public policy on strengthening the institution of marriage or on using public funds to avoid unemployment (despite the fact that such a measure could possibly reduce GDP growth) could make sense if happiness maximization in a policy goal. It is in this ‘neo-utilitarian’ way, in my view, that indicators like ‘happiness’ may be of use – to provide some quantitative backing to policy debates, which, currently, are based solely on handwaving qualitative or ideological arguments.
I think we might all be happy with that.
Screaming and Shouting 
Leave a Reply