The well-being of nations

Andrew Oswald considers recent moves in economics, famously the most dismal of sciences, to take the happiness and psychological health of the population as seriously as a country's GDP

My subject - economics - is becoming cheerier. You may have noticed from the newspapers that there is significant interest in the issue of how to measure human "happiness", and that the prime minister, David Cameron, has asked government statisticians and economists to stop focusing so intently on gross domestic product (roughly, a simple count of how rich a country is) and to start collecting persuasive measures of national well-being and happiness.

Cameron has set us a complicated task. But social scientists have recently been doing a great deal of research on the topic. In my judgement, we have been getting somewhere. This research was not driven by a concern for "impact" in today's sense. It was spurred by intellectual questions about how humans are, rather than any practical desire to alter our world - although the results may do just that.

If you have ever sat through the graphs and equations of a first-year undergraduate lecture in economics, you may feel that the words "economics" and "happiness" do not greatly deserve to go together. Or you may even have the instinctive reaction that happiness is something to be debated by philosophers and not quantified. Such reactions are reasonable ones.

Indeed, I have been in a number of emotive, and unfortunately sometimes not well-informed, lectures in which speakers wish to attack this new intellectual trend. However, this does not make people's intuitive criticisms right.

How do we do this kind of research work and what is being discovered? We begin with random samples of people, sometimes quite large ones - millions of citizens from nations around the world. Then we try to understand, as a statistical matter, what explains the pattern of happiness or mental health across different sorts of individuals and, second, what it is that explains the levels of happiness and mental health across different sorts of nations - Britain, Russia, India and so on. I think these are vital issues in social science and public policy - I hope most people would agree with that - although they are undoubtedly difficult ones. Nobody pretends that measuring the well-being of a person or society is painlessly easy. Ultimately, runs the politicians' concern, can we learn how to make whole countries happier?

Back in the 1970s a man called Richard Easterlin, who is still going strong in his mid-80s in California, helped to start the field of the economics of happiness, although at the time almost nobody would listen. At the start of the 1990s, a group of youngish British economists revivified the dormant field. The impetus for their work was the availability of new data sources that were unimaginably big by previous standards, combined with an instinct that not all was ideal in conventional economics.

The first conference on the topic of the economics of happiness was held in London in the autumn of 1993. It is no secret that it was a failure. Only eight people attended, despite the 100 chairs and the posters we put out on the day; nevertheless, it was a start. Scholars such as Andrew Clark and David Blanchflower made important intellectual contributions to the field early on when it was unfashionable to work on the topic.

Things have not greatly looked back, although of course there will be swings in sentiment to come. I imagine we will see a retrenchment, a turning-against, in this field, followed some years later by a revival, and ever onwards and upwards, in the usual cycle traced out by the warfare of ideas. If you want a good modern introduction to the research, you might try Nick Powdthavee's 2010 book The Happiness Equation: The Surprising Economics of Our Most Valuable Asset. It is beautifully written.

So, standing back a bit, how do we measure human happiness and well-being? The simplest approach began with the General Social Survey (GSS) of the US, which began in 1972 and still today asks randomly sampled individuals, "Taken altogether how would you say that things are these days? Do you think of yourself as very happy, pretty happy or not too happy?" One third of Americans tick the box saying they feel very happy with their life, and 10 to 15 per cent tick the "I'm not too happy with my life" box.

In the past 10 to 20 years, we have moved on to other indicators that are closer to measures of psychological health or mental strain. One I work with a lot is the General Health Questionnaire (GHQ) score. This is a mental-health indicator and it is also used by epidemiologists and doctors around the world. It is a string of questions: "How well have you been sleeping?"; "Have you been worrying?"; "Have you been thinking of yourself as depressed ... or not contributing?", and so on.

The latest work blends subjective scores such as these with physiological measures and other objective indicators. Articles on such topics have appeared in the past year or two in journals including Science and the Proceedings of the National Academy of Sciences of the United States of America.

Researchers take well-being data and then they estimate so-called regression equations. What this means is that we take the data points on different people - giant clouds of dots, you might say, if you want to think of a graph on a computer screen - and find the best-fitting lines through those points. In that way, we aim to uncover hidden patterns. We want the deep-down, implicit relationships between income, education, gender, having children, and other things, and people reporting whether or not they are happy with life and whether they seem, on standardised scores, to have good mental health. It is fair to ask if it is possible to do this in a systematic way.

One thing we know is that if you look at magnetic resonance imaging "slices" through brains, emotions such as happiness and sadness do show up in different parts; so, at some deep physiological level, we know something about what happiness and its opposite look like. It is also beginning to be understood that certain genes appear to be correlated with greater happiness scores on standardised measures.

We know too that if you ask people about how happy they are and then ask their friends, their family and their partners, that the answers those others give about whether Freda Jones is a happy person are correlated with what Freda Jones says, again lending credence to the idea that such answers should be taken seriously.

We have discovered that happiness scores are correlated with the observed number of smiles that a person gives during the day; we know that they are correlated with the quality of memory, blood pressure and how fast your heart beats - recently my colleagues Blanchflower and Nick Christakis and I have been doing a lot of work on heart rate in large UK samples.

We are aware that, intriguingly, when you control for everything else statistically, for every extra €50,000 I earn a year, my heartbeat seems to be one beat a minute slower. And heart rate correlates with coronary heart disease, and with the stress hormone, cortisol. We know too that well-being scores are predictive of whether your marriage will last, whether you will quit your job and so on.

On happiness and hypertension, it has recently been demonstrated that there is apparently, if you look across countries, quite a powerful relationship between countries where more people say they are happy and countries with less hypertension (that is, lower blood pressure). Hence again, at a deep level, these subjective responses of people given in surveys are correlated with objective well-being criteria.

In Western countries, people are fairly happy with their lives. Roughly 15 per cent of British people say "I'm completely satisfied with my life", which I tend to think shows a certain lack of imagination. But there we have it. One of the good things about well-being data is that the numbers allow humans to say what they feel. That is liberal, in the true sense of the word.

Plainly life is a mixture of ups and downs. One of the reasons why I got interested in these kinds of data back in the early 1990s is that these happiness scores are highly correlated, in a sensible way, with all the good and bad events we would expect. There is also a strong life-cycle pattern. If you are a typical person you will slide down - and then up - in a giant U-shape of happiness through your life. The accompanying graph (see below) is based on a random sample for the UK, but researchers have replicated the pattern over and over. Good luck, wherever you are on this curve (I am going up).

When we look at well-being scores, what do we discover? We find, holding other things constant, that women report higher happiness with their lives. People with lots of friends say the same. There is a strong correlation between being married or having a steady cohabitation and describing your life as a happy one, and that is consistent with a lot of data we have on mental health.

If you have a PhD (even in economics), then congratulations; you arguably did the right thing, because it is correlated with high well-being. And, more broadly, education is associated with greater well-being - maybe because it gives you more security in life.

People on higher incomes, holding everything else constant, are happier with their lives. Money helps.

What are the very negative things in life? Unemployment, serious illness and a new divorce or new separation are the giant negatives we see in our happiness equations. These are equivalent in dollar values to the loss of enormous sums of money, be it the negative effect of being unemployed or the negative effect of splitting up with your wife or husband.

Finally, if you are thinking of investing in children, I am afraid on the statistical evidence I cannot advise you to do so. On the personal level, yes, I can, particularly if you can have daughters like mine. But although parents often find it hard to believe, the evidence, holding other things constant, shows no statistically significant positive effect on happiness from having children.

If you are an environmentalist or an environmental economist, happiness economics is potentially particularly valuable for you. This is not yet widely realised by the Green movement. Just in the past few years we have managed to put dollar, euro and pound values on the happiness from clean air, lack of noise, lack of chemical additives and so on. Brilliant papers have been written by Simon Luechinger, a young researcher at the Institute for Empirical Research in Economics at the University of Zurich, who has traced people's happiness levels back to the levels of sulphur dioxide emissions in plants far from their homes. Some of the most exciting work is now a remarkable blend of geographical information system data and information from well-being surveys.

There is no doubt that economists knew something or other when they taught their students from time immemorial that higher income allows people to have higher "utility", or higher happiness. But is there really good causal evidence? Recently we realised that longitudinal information on people before and after they won the lottery would count as proper evidence: this is studying windfalls, of course. We get quite convincing scientific evidence that, in this rather controlled setting, money makes people happier. Using lottery wins is the closest we can get to running a real scientific experiment. We compare people who played the lottery and won merely a tiny amount (the "control group") with the "treatment group" of those who win large sums.

Health matters. There is a great deal of research done by epidemiologists, partly by people like me and partly by psychologists, to show that there seem to be very deep, powerful links between the quality of your mind - what is going on in your head - and how well your body works. We do not understand the mechanisms at all well. Researchers including Andrew Steptoe at University College London are doing terrific work here.

Many of us will go through divorce, or we have done so already - something like 40 or 50 per cent of us in Western society. So, in the long term, is divorce a cause of greater happiness, or is it irrational? Jonathan Gardner and I published a paper on this subject in a statistical journal and for some time it was its most downloaded paper. I suppose it was because word got around and people clicked on it to learn about their own situation.

Overall, the data show that, two years after divorce, happiness does rise. Hence divorce is, typically, rational. I should hasten to add that I am not recommending that you go home tonight and say: "Darling, we need to talk." But the finding seems important when divorce is so common. In a sense, it's reassuring to a social scientist, and perhaps to you, that although divorce is extraordinarily painful and nobody could recommend it, it does seem to work for those who take that step. We see similar kinds of happiness recovery - bounceback - in many other spheres of life.

What about nations themselves? Denmark, Switzerland and the Netherlands always do noticeably well in happiness league tables. If we look at a scatter plot across countries, with GDP per capita on the x axis and a standardised life satisfaction score on the y axis, there is a positive association. On average, the greater a nation's real (ie, inflation-adjusted) GDP, the greater its mental well-being. So there's a simple association between having plenty of dollars in a well-off country and feeling happy.

However, that is somewhat misleading. The biggest problem for economic policy in Western society is now called the Easterlin Paradox. If you take random sample surveys (as University of Southern California economist Easterlin did for the US) and ask people "How happy are you with your life?" the answers just run flat through the years. It does not matter how rich the country gets. Some nations are even trending down slightly in their reported levels of well-being.

So GDP shoots up and up through the decades, but happiness scores do not rise. There is either something wrong with this Easterlin Paradox, or it is of fundamental importance to policymakers and politicians. At the moment we are not certain which it is, but I must tell you that, on the balance of the evidence, Easterlin's Paradox is true, alive and well, and it is probably alive and well in your country, wherever you are reading this.

This paradox - that countries do not get happier as they get richer - is intellectually, even personally and emotionally, a threat, at least in the eyes of many policymakers, bankers, business owners and politicians. By and large, most people, if they have been involved in policy or if they have had conventional economics training, do not want to hear it, and if they are willing to hear it they do not want to believe it. The Easterlin Paradox is unsettling.

We really do not know what goes wrong as a country gets richer, but there are some clues. One thing is that, in a deep sense, we are animals of comparison. We cannot help it. What I want is to have three BMWs and for my friend to have an old Ford. That is what I want deep down, even though I cannot face up to that fact, and cannot say it to him or admit it to myself.

The problem is one of inescapable relativities. For well-off countries, the tide of economic growth lifts all the boats and unfortunately, where having three BMWs was unusual, eventually it becomes the norm. A kind of generalised neutrality, a sort of washing-out effect, seems to be produced. Second, researchers think that people habituate to money. In other words, I initially enjoy my Ferrari: I go zooming down the high street in Warwick but, you know, the 28th time that I do it, the novelty has rather worn off. I've habituated. Finally, maybe human beings make bad choices as they get richer and do things that do not make them happier.

Economists have resisted this idea for a long time; we have not listened to psychologists such as Harvard University's Daniel Gilbert. Yet there is evidence that people expend too much time and energy on things that lead to cash. We depress, as it were, our personal lives, and we do that to try to make more money, and to achieve greater success in our work. I am not preaching here. I have been a workaholic pretty much the whole of my adult life, and it is very tricky to change that. The bottom line is that many of us are conceivably making the wrong choices, even for our own "utility" functions, as the jargon goes.

There is another important application of the economics of happiness. One day, in my judgement, it will be used a lot in the courts. Many court actions - negligence actions, for example - are about injuries to people that do not come with automatic price tags attached. Judges have to adjudicate, and juries have to dole out sums of money for complex human tragedies.

Our methods have found the first way to do this more systematically. If you estimate happiness equations, and then back out the damages numbers - in other words if, sad to say, your partner was killed going home tonight by a car - we could calculate how much money it would take in principle to compensate you. That is a horribly black calculation and perhaps only an economist would write those kinds of sentences, but unfortunately judges are in the business of trying to adjudicate on such real issues.

In conclusion, human happiness is on the agenda and economics is changing. Economists have probably been wrong to believe that economic growth makes (rich) societies happier. This is unsettling and still debated; there is a small amount of evidence against, and many economists will not change their minds whatever data are presented to them about it. But the weight of the evidence is in line with Easterlin's Paradox.

More broadly, and partly for that reason, it seems that policy in the coming century will need to concentrate more on non-materialistic goals. This does not mean the end of economics as we know it, merely the need for a little more flexibility in the standard economist's mind. And that is happening.

One day, towards the end of my lifetime, BBC TV news headlines will likely begin not with "economic growth was up 1.2 per cent last month" but rather "the country's composite psychological well-being index rose 0.8 per cent". By then, such a thing will seem natural. People will look back and puzzle to themselves why it was not done that way in the early part of the 21st century.

Smiling? The ages of contentment


Smiling? The ages of contentment
Credit: Andrew Oswald and Nick Powdthavee
Middle-age slump: Individuals' measures of happiness follow a strong life-cycle pattern in Britain

  • Print
  • Share
  • Save

Related images

  • THE cover, 19 May 2011
  • Print
  • Share
  • Save
Jobs