Happiness: A measure of cheer

‘The welfare of a nation can … scarcely be inferred from a measure of national income.’
So the US Congress was warned in 1934 by Simon Kuznets, who thus continued a long tradition of pointing out that there is more to life than money. But the economist’s comments broke particular ground: they were attached to the first serious attempts to produce national income accounts – the tally of all that a country produces and earns – for the US. Kuznets and his small research team had built them, and he knew their limits.
Where Kuznets led, others have followed. From the upper echelons of the administration of President Barack Obama to the offices of Nicolas Sarkozy, his French counterpart, to David Cameron, UK prime minister, the goal of gauging a nation’s well­being has captured the imagination of policymakers. They join less likely countries such as Bhutan, whose mission to measure “gross national happiness” has made the Himalayan mountain kingdom a trendsetter.
Mr Cameron was the most recent to take up the cause, saying Britain needed to look for alternative measures that would show national progress “not just by how our economy is growing, but by how our lives are improving; not just by our standard of living, but by our quality of life”. While some analysts suspect each politician has his own motives – appearing nice to electors, flattering the economy and so on – the result has been to create a sense of momentum behind happiness economics.
Embracing happiness is one thing; measuring it another. There are, broadly speaking, three approaches to measuring a country’s wellbeing. One is to use the long-established framework of national income accounts and adjust it to better reflect national welfare. The second is to collect data on objective measures that are plausibly related to wellbeing: anything from life expectancy to crime, suicide rates to income inequality. The third – and the three are not mutually exclusive – is to try to measure national welfare directly by asking people how they feel: the equivalent of measuring wealth by asking: “On a scale of one to 10, how rich are you?”
Intellectually the project of measuring national welfare seems to belong on the centre-left. National income accounts themselves – like unemployment statistics – rose to prominence in the 930s as Franklin Roosevelt tried to lead the US out of the Great Depression and the government recognised it understood very little about how the economy was really doing – beyond very badly. The second world war and the central planning mentality that came with it merely intensified the focus on the government being able to understand how the economy was functioning.
The countervailing view was most pithily expressed by John Cowperthwaite, the laisser-faire financial secretary of Hong Kong in the 1960s, who claimed he refused to collect economic statistics because it would only provide ammunition to the planners. His views have not prevailed. Kuznets won the Nobel Prize in economics in 1971 – only the third year in which it was awarded – and it is hard to find serious economists who think that collecting national income accounts is pointless. The question is whether there is a benefit in trying to supplement those with accounts of national welfare – or, to put it cutely, an index of national happiness.
Mr Cameron says that there is. “You cannot capture happiness on a spreadsheet any more than you can bottle it,” he declared last month. The happiness index project, he added, had a practical purpose: to help the government understand, “with evidence”, the best way of improving people’s wellbeing. In short, Mr Cameron believes happiness indices can help his Conservative-Liberal Democrat coalition government plan its way to a happier nation in the same way Kuznets’ national accounts helped policymakers respond to the Depression and the outbreak of war.

Before understanding what governments might do with measures of happiness, it is worth asking how to measure it. Although “neuro-economists” are not averse to sticking people into brain scanners with a view to gleaning their inner thoughts, the easiest way to find out whether somebody is happy is to ask them. This might seem uncontroversial; after all, it seems almost definitional that if you’re happy then you know it.
Yet even here there are two quite distinct approaches. The first, and best known, is to collect information about “life satisfaction”, asking people how satisfied they are with their lives, taken as a whole, on (for instance) a scale of one to 10, eliciting a once-and-for-all answer to the question of how life is working out.
The alternative is to zoom in on a particular set of events and ask what feelings they produced. One example is the “day reconstruction method” produced by researchers including Norbert Schwarz, a psychologist, the economist Alan Krueger (until recently chief economist of the US Treasury), and Daniel Kahneman, a psychologist and another winner of the Nobel memorial prize in economics. The DRM asks people to recall, episode by episode, the previous day’s events and the most prevalent accompanying feeling – stress, peace, exhaustion, elation.
In short, one approach measures life satisfaction and the other mood – and the two concepts are quite different ways to think about happiness. One survey comparing women in Rennes, France, and Columbus, Ohio, found that the American women were twice as likely to say they were very satisfied with their lives, but the Frenchwomen spent more of their day in a good mood. “We have tended for too long to use a single word to refer to a wide variety of things,” says Prof Kahneman. “In particular there’s a real need to distinguish between life satisfaction and mood or experienced happiness. They are quite distinct and they have different causes and consequences.” If politicians are to incorporate the measurement of happiness into the fabric of national statistical frameworks, this is a distinction they will have to start taking seriously.
One obvious objection to the new-found emphasis on indicators of national welfare is that it is not, in fact, new-found at all. The most famous (albeit contested) finding in happiness economics is the Easterlin Paradox – that money buys happiness in any given society but richer societies seem to be no happier than middle-income ones – which dates back to the 1970s.
Still, data on subjective wellbeing have been improving ever since. The UK has published Social Trends for four decades, an almanac for modern times that includes information on crime and the fear of it, attitudes to marriage, satisfaction with the National Health Service, calls to child protection numbers, air pollution, North Sea fish stocks and much else. Other countries do likewise.

This puts the wellbeing agenda in a strange place. Those who believe in it know that the data already exist: Andrew Oswald of Warwick University and a leading advocate of the use of well­being data points to the UK’s Labour Force Survey, which already includes questions on depression and anxiety. Prof Oswald wants to see more such questions and have more attention paid to them, but this is a request to do more rather than to do something profoundly different.
But professors Kahneman and Krueger, and others, do have plans for a more radical departure. They propose the publication of “time accounting” measures alongside regular national accounts. The US Bureau of Labour Statistics has, funded by the National Institute on Aging, begun to implement some of their proposals.
Time-use surveys use representative samples to see how a nation spends its time – whether cooking, commuting or watching television. They are a well-established part of the US statistical apparatus. The happiness element comes in when the surveys are combined with the DRM. Because this asks survey respondents to rank the emotions they felt while spending time in various activities, it can be used to produce a measure of how long people spend in a predominantly unpleasant state of mind – as well as the kinds of things they are doing at the time. (For the curious, the particularly miserable activities include commuting and working. Sex, lunch and dinner are rarely unpleasant.)
The Kahneman-Krueger team have developed a stripped-down version of the DRM, which can be conducted as a regular telephone survey, and the Bureau of Labour Statistics has already conducted its first survey with the new method. Pilot studies seem to produce results broadly faithful to the DRM; the first data from the BLS work are due early in 2011.
Will national time accounts – or some alternative measure of welfare – give politicians a tool of practical importance? Or will Cowperthwaite be proved correct: that the data will only encourage politicians to meddle? Prof Krueger thinks not. “This approach is particularly amenable to evaluating public policy interventions, as investing in roads and high-speed rail can affect commuting time, overtime laws affect work hours, and playgrounds affect leisure activities.”
National time accounts will do more than that, because they sidestep an insuperable problem of the life satisfaction approach: that one person’s five out of 10 might be another person’s eight out of 10. Up to a point this problem can be shrugged off, especially when looking at large sample sizes, but if there are systematic differences between the way people express these judgments – differences between urban and rural, white and black, men and women, Danish and French – the data cannot be used to make comparisons across these groups. (In one Eurobarometer survey, 64 per cent of Danes described themselves as “very satisfied” and only 16 per cent of French. It is tempting to question how much the survey really tells us about the relative wellbeing of France and Denmark.)
National time accounts measure how much time people spend engaged in activities where the strongest of many emotions is a negative one. This greatly lessens the interpersonal comparability problem (although it cannot remove it entirely). In the end it may produce controversial findings – imagine, for instance, if it were shown that women spent far more time doing things they disliked than men did, or that the elderly lived blessed lives compared with the young.
The UK Office for National Statistics will continue its consultation until mid-April. By then the first American time-use accounts should be available. If Mr Cameron is serious about happiness, he will be casting an optimistic glance across the Atlantic.

FT Analysis 28 December 2010
Full version (graphs, sidebar) available at FT.com