Happiness and the Easterlin paradox

Romesh Vaitilingam interviews Richard Easterlin for Vox

January 2009

Transcription of an VoxEU audio interview [http://www.voxeu.org/index.php?q=node/3439]

Romesh Vaitilingam: Welcome to Vox Talks, a series of audio interviews with leading economists from around the world. My name is Romesh Vaitilingam, and today's interview is with Professor Richard Easterlin from the University of Southern California. Richard and I met at the American Economic Association's annual meetings in San Francisco in January 2009, where we spoke about his pioneering research on the economics of happiness. I began by asking him to explain the idea to which he has given his name – the Easterlin Paradox.

Richard Easterlin: The paradox is the seeming contradiction between the cross-section evidence on the relation between income and happiness and the time-series evidence.

The cross-section evidence is that happiness and income are positively related. That's true on comparisons at a point of time among income groups within a country. It's also true of comparisons at a point in time of richer and poorer countries. The paradox is that when you look at what happens within a country over time, as income goes up, happiness does not rise the way one would expect it to, on the basis on the cross-section.

Romesh: OK. You've looked over data over many countries and over long periods of time to provide evidence in support of this paradox.

Richard: Well, the original statement was based simply on evidence for the United States where - when I first did this work - that was the only country for which there was a fairly substantial time series available. Now over the past 20 years, basically, there has accumulated much more evidence on happiness in countries all over the world. And what I've been doing the last two years is looking at that evidence to see to what extent the paradox still prevails when you look at countries more generally.

Romesh: Now, some people have done research claiming that the paradox doesn't exist and happiness does increase with income. But you still hold to it, you think the evidence is still very much on the side of what you originally found.

Richard: Yes. Let me respond in two segments: one is what I think the evidence shows, and second, why I think the argument that the paradox doesn't hold is incorrect.

With regard to what the evidence as a whole shows, I've done now studies with my colleagues at the University of Southern California of three groups of countries. The developed countries – and there we have about 17 countries, mostly European non-transition countries plus the United States, Australia, and Canada. For those 17 countries, we have time series that average, as I recall, about 20 years in length.

If you compare the rate of economic growth as the independent variable, let's say, with the rate of change in happiness, the expectation based on the cross-section would be the higher the rate of economic growth, the greater the improvement in happiness. When in fact you look at the data, the relationship is nil. There is no significant relation between the growth rate of GDP per capita and the rate of improvement in subjective wellbeing or happiness. That's for the developed countries.

The second group of countries is the developing countries, for which the data are much more fragmentary and require a lot more effort to assure comparability, because the question changes, the context to questions change, the survey coverage changes. So, we've spent a lot of time trying to adjust or correct. When we do that, we end up basically with only nine developing countries, but they are countries like China, Brazil, Chile, Nigeria, the Union of South Africa. So, we have a pretty substantial cross-section of countries. Turkey is another one.

There, I think, we've been able to get time series that average about 16 years in length. And when you look at that, the same relationship holds that we find for the developed countries, that the higher the rate of economic growth, there is no significant relationship to the rate of improvement in happiness.

The third group of countries is the transition countries of Eastern Europe. There the time series pattern that's associated with the transition is a V-shaped pattern. There is a very substantial collapse of GDP and recovery of GDP per capita. And that's accompanied by a similar movement in happiness, or subjective wellbeing. So, you get a V-shaped pattern of happiness in those countries, paralleling the V-shaped happiness of the movement of GDP per capita.

So, what you want to be sure to do is be able to fit a trend line that is not simply capturing the downward movement or the upward movement, but what the long-term trend is, which is whether happiness is greater at the present - 2005, which is the most recent date we have - compared with as early a transition data as we can get, which is typically around 1989 or '90.

There we have, I think, about 11 transition countries that range from East Germany to Russia, the Russian Federation. Again, there is no significant relationship between the long-term growth rate in the economy in those countries and the rate of improvement of happiness.

If we pool all of these three sets of countries, so we get about 35 countries or something like that, there is no relationship between the rate of economic growth and the improvement of happiness. So, that's what the evidence shows.

Now, the question is how do people arrive at different answers to that question? There are several reasons, but the principal one is very simple. It has to do with the patterns in the transition countries. As I said, in the transition countries both GDP per capita and happiness follow V-shaped patterns. So if you take the contraction period - the downward movement along the V - both GDP per capita and happiness are declining. If you take the upward movement of the V, both GDP per capita and happiness are improving.

Now, those are short-term changes, and the rates of GDP per capita change, the annual rates of GDP per capita change are obviously much higher than is true of the long-term rates of change. The long-term rates average out the short-term fluctuations.

So if you throw those data for the contraction phase and the expansion phase of the transition countries in with the long-term data, what you get is the transition country data for the short phases are at the extremes: high rates of recovery and high improvement in life satisfaction, high rates of collapse and high rates of collapse in happiness. And then you produce this positive relationship that others are saying contradicts the paradox. But it doesn't contradict the paradox, because what they're looking at are short-term relationships not the long-term trends.

Romesh: So you're reading of the data is very much that the paradox still holds. What are your feelings…

Richard: It's stronger than that. I mean the original statement of the paradox was about a rich country, the United States. And most of the earlier evidence that I or others were able to come by subsequently was for European countries and other developed countries. No one had really done anything on the developing countries, and very little had been done on the transition countries.

So, and when I approached this, I really was not seeking to find the results that came out. I was interested in the simple question of whether, with regard to the transition countries, you really see an improvement in subjective well-being associated with the shift from socialism to free-market society. And in the developing countries, which are moving from much lower levels of income, whether you see an improvement in happiness associated with the rates of economic growth there, some of which of course are quite high, as in China.

So, I didn't know what the outcome would be in advance. But when I looked at the data, it turned out that the paradox holds across the board, not just for the developed counties, which is what most people have assumed heretofore. Most people have assumed that in the developing countries, higher rates of economic growth would raise subjective wellbeing.

Romesh: Dick, what's your explanation of what's going on? What underlies the paradox? Why is it that there isn't this clear relationship between rising income and rising happiness?

Richard: I think that there are two factors at work, but here, of course, we're getting more speculative. I think there is some evidence supporting this. But one factor is that as people's incomes go up, their material aspirations rise. So if their aspirations stayed the same, and their incomes went up, they'd closer to achieving what they considered to be the good life in material terms. But when their aspirations go up, that undercuts the effect of the actual improvement of incomes. So, they end up no happier than before.

There is no reason to suppose that this effect on aspirations is confined to higher-income countries, that lower-income countries somehow are immune to the rise in aspirations. And indeed if you look at the sort of anecdotal evidence in newspapers and so on, you see - in countries like India, China and so on - how people's aspirations for automobiles and so on are just escalating enormously.

So aspiration is part of the story I think. The other part of the story I think has to do with the domains of happiness other than the material domain. What happens with regard to people's family life circumstances? What happens with regard to people's health circumstances? What happens to their job opportunities?

So let me take the transition countries as a fairly good example, you see. Because in the transition countries, what's clearly happened is on the economic side, their average living levels in real terms have improved. Where there used to be no goods on the shelves, there is now a greater availability of goods.

So on the material side, if you ask people about their satisfaction with their standard of living or their income and so on, you generally get an improvement over the 15 to 20 year period - from the start of the transition to kind of the current time. On the other hand, if you ask them about their satisfaction with things like their work, with their childcare, with their health, these are domains in which their satisfaction is generally diminished, because these used to be conditions that were assured.

So what people have done in the transition from the socialist to free-market society is essentially trade off sources of security, with regard to having a job and having the children provided for and educated, their health and so on. Those have gotten worse, you see, and tended to offset any positive effect from the effect of greater income.

Now in the developing countries, I think there is something analogous that goes on. Let me take China as an illustration. In China, it's well-recognized that income inequality between urban and rural areas has increased enormously. Presently, it's on the order of three times greater income in the urban than the rural areas. So, based simply on an economic comparison, you see, you would say well, people must be happier in the urban areas. In fact when you compare happiness in the rural and urban areas, there is almost no difference. The rural may be possibly a little better off.

Well, why is that? Again I think it has to do with - well, part of it is the people in the urban areas are the ones much more exposed to what's going on with regard to new goods and so on. So their aspirations are rising more than in rural areas. But the other thing is that the movement from rural to urban areas sort of tears apart the social network that prevailed in rural areas - family support, community support systems and so on. People are much more isolated in urban areas and thrown more on their own.

In China, particularly, it's much harder now for young men and women to form families, because there is such a surplus of men. So, China is a little exceptional in that case. Basically, I think there are changes that occur in the other domains that are important for happiness that tend to offset the effect of the improvement of goods. Not just that aspirations tend to rise, but also people feel much more anxiety about their work, about their family ties, neighbours, and so on. They just don't have the kind of connections they used to have.

So that's what I think is going on and what the evidence is basically showing - it's telling us something about the importance of domains for happiness that economists tend to disregard and simply say well, if income's better off, then people must be better off

Romesh: Final question, Dick. You're a lone voice working in this area - or at least when you started off over 30 years ago. Many more people have come into this area, happiness economics seems to be booming and it's piqued the interest of policy-makers, certainly in some countries. What's your impression of how the field has progressed? Where have we made progress in understanding the drivers of happiness? And what are the opportunities for the future, do you think?

Richard: Well, I think we have made a lot of progress. It's been very encouraging to see the growing volume of research. A lot of what I'm saying draws upon the findings of people like Andrew Oswald, Andrew Clark, Richard Layard, and others, who have I think tended to identify the importance of these other factors in people's levels of happiness and also, to some extent, identified the role that aspirations are playing, drawing upon the psychological literature.

Romesh: And now for the future, you see this as a field that will continue to grow?

Richard: Well, I certainly hope so. I think what it does is put us much more in touch with what's really important in people's lives. The nice thing about the happiness measures is each person gets one vote…whereas when it comes to GDP per capita here, if you have the higher income, you're getting more votes than the lower-income person.

Romesh: Richard Easterlin, thank you very much.

Richard: My pleasure. Thanks a lot.