Early economists and philosophers, such as Adam Smith and Jeremy Bentham, were serious about the study of happiness. Yet with the rise of quantitative methods in economics calling for more parsimonious definitions of welfare, happiness fell out of fashion, and utility became synonymous with income. Over a century later, in the mid-1970s, Richard Easterlin revisited the relationship between happiness and income. His findings uncovered what seemed to be a paradox; average happiness levels did not increase over time as countries grew wealthier, nor was there a clear relationship between average per capita GDP and average happiness levels across countries, once they achieved a certain minimum level of per capita income. This puzzle is known as the Easterlin paradox.
In recent years there has been a renewed debate about whether the Easterlin paradox holds, not least because an increasing number of economists have started to use happiness surveys to explore all kinds of questions. Recent studies by Betsey Stevenson and Justin Wolfers, and by Angus Deaton, based on new data from the Gallup World Poll, find a consistent log-linear, cross-country relationship between income and happiness, directly challenging Easterlin’s findings (see Stevenson and Wolfers 2008, and Deaton 2008). This has resulted in a heated and, at times, even acrimonious debate among economists.
Two right answers
Rather ironically, both sides of the debate may be correct. There are two reasons why:
- One reason is substantive. On the one hand, it makes sense that people in richer countries are happier than those in destitute ones, whereas on the other hand, many things other than income contribute to people’s happiness, regardless of their level of income. Many of these things – like freedom, stable employment, and good health – are easier to come by in wealthier countries. Still, there is plenty of variance in the availability of these things even across countries with comparable income levels.
- The other reason is methodological. The later studies use new data from the Gallup World Poll, which includes many more (un-weighted) observations from small poor countries in Africa and from the transition economies than did Easterlin’s original studies (as well as his more recent ones). The transition countries in particular have relatively low levels of happiness, in part as a result of happiness levels falling markedly with the painful structural changes that accompanied the collapse of centrally planned economies. Meanwhile, some of the sub-Saharan African countries have had flat or even negative rates of growth over time. Thus rather than a story of higher levels of income pulling up happiness at the top, it may be one of falling or volatile income trajectories pulling down happiness at the bottom.
How happy are you with your life?
There are also differences in the questions that are used to measure happiness. Easterlin’s work is based on the World Values survey, the US General Social Survey, and the Eurobarometro survey, among others, all of which use open-ended happiness or life satisfaction questions. Generally speaking, these questions ask, “how happy are you with your life?” or “how satisfied are you with your life?”, with possible answers ranging from “not at all” to “very” on a 4 or 5 point scale. The Gallup World Poll uses Cantril’s best possible life question, which is “please imagine a ladder with steps from zero to ten, if the higher the step, the best possible life, on which step of the ladder to you personally feel you stand?”
Both sets of questions are reasonable gauges of happiness, broadly defined, and both correlate in a similar manner with the usual variables. Research based on all of these questions finds that, on average, stable marriage, good health, and enough income are good for happiness (with how much income is enough varying across countries), and that unemployment, divorce, and economic instability are bad for happiness. Age and happiness have a remarkably consistent U-shaped relationship, with the turning point in the mid to late forties, at which point happiness increases with age, as long as health and partnerships stay strong. Indeed, I have studied this relationship in countries as diverse as the UK, Chile and Afghanistan, and it holds in all of them, with modest differences in the turning point. Among other things, this relationship reflects an alignment of expectations and reality as people “grow up”.
At the same time, there is some variance in the findings based on different questions. The best possible life question is more framed than the open-ended happiness questions, providing respondents with a relative component when they are asked to assess their lives. Mario Picon, Soumya Chattopadhyay, and I tested the questions against each other in the Gallup World Poll for Latin America, a region for which we had both sets of questions in the same survey. We found that the answers to the best possible life question correlate more closely with income – both across and within countries – than open-ended happiness questions (Graham, Chattopadhyay, and Picon in Diener et al. forthcoming).The difference is greater across countries than within them.
Thus, simply due to methodology – what sample of countries and which happiness questions are used – it is possible to come to different conclusions about the Easterlin paradox. The substantive question of what – beyond income – makes people happy is an additional and more complicated part of the story. Figure 1 presents findings from my previous research with Stefano Pettinato, based on an open-ended happiness question and a very simple linear specification of income. While the richer countries are, on average, happier than the poorer ones, there is no clear income and happiness relationship within each set of countries, making it impossible to draw a clear conclusion about the Easterlin paradox.
The figure drums home the point that wealthier countries are, on average, happier than destitute ones, but after that, the story becomes more complicated. Country level averages are influenced, among other things, by cultural differences in the way that people answer surveys, and these cannot be controlled for in the cross country comparisons in the way they are when we assess happiness across large samples of individuals within and across countries.
New research on people’s capacity to adapt
My more recent research on happiness around the world throws another monkey wrench into the equation (see Graham 2010). While the research confirms the stable patterns in the determinants of happiness worldwide, it also shows that there is a remarkable human capacity to adapt to both prosperity and adversity. Thus, people in Afghanistan are as happy as Latin Americans – above the world average – and Kenyans are as satisfied with their healthcare as Americans. Crime makes people unhappy, but it matters less to happiness when there is more of it; the same goes for both corruption and obesity. Freedom and democracy make people happy, but they matter less when these goods are less common. The bottom line is that people can adapt to tremendous adversity and retain their natural cheerfulness, while they can also have virtually everything – including good health – and be miserable.
One thing that people do have a hard time adapting to is uncertainty. For example, my latest research, with Soumya Chattopadhyay and Mario Picon based on daily records of around 1,000 Americans from January 2008, shows that average happiness in the US declined significantly as the Dow fell with the onset of the crisis. Average happiness fell 11% from 6.94 (on an 11 point scale) prior to the onset of the crisis, to a low of 6.19 on November 16, 2008. Yet when the market stopped bottoming out and some semblance of stability was restored in late March 2009, average happiness recovered much faster than the Dow. By June 2009 it was higher than its pre-crisis level: 7.15 on June 21 – even though living standards and reported satisfaction with those standards remained markedly lower than they were prior to the crisis. Once the period of uncertainty ended, people seemed to be able to return to previous happiness levels, while making do with less income or wealth (see Figure 2).
Indeed, people seem to be better at adapting to unpleasant certainty than they are to uncertainty. It is surely a good thing that most Americans have been able to adapt to the economic costs of the crisis and return to their natural happiness levels. And even better that the average person in Afghanistan can maintain cheerfulness and hope despite the situation they live in. While this capacity to adapt may be a good thing from the perspective of individual psychological welfare, it may also result in collective tolerance for conditions that would be unacceptable by most people’s standards. This may help explain why different societies tolerate such different norms of health, crime, and governance, both within and across countries. Without understanding these norm differences, it is very difficult to craft policies to improve health, living conditions, and governance structures.
Happy peasants and miserable millionaires
This capacity to adapt – and the mediating role of norms and expectations – poses all sorts of measurement and comparison challenges, particularly in the study of the relationship between happiness and income. Can we really compare the happiness levels of a poor peasant in India, who reports to be very happy due to low expectations or due to a naturally cheery character, with those of a successful and very wealthy CEO, who reports to be miserable – due to his or her relative rankings compared to other CEO’s, or to a naturally curmudgeonly character? This is something that I have called the “happy peasant and miserable millionaire problem” (or the happy peasant and frustrated achiever problem). On one level it suggests that happiness is all relative. On another it suggests that some unhappiness may be necessary to achieve economic and other sorts of progress. The examples of migrants who leave their home countries – and families – to provide better futures for their children, or revolutionaries who sacrifice their lives for the broader public good, come to mind. This also raises more difficult questions, such as whether we should tell a poor peasant in India how miserable they are according to objective income measures in order to encourage that peasant to seek a “better” life; or whether we worry more about addressing the millionaire’s misery or increasing the peasant’s happiness.
This happy peasant and miserable millionaire paradox also raises the question of the appropriate definition of happiness. What makes happiness surveys such a useful research tool is their open-ended nature. The definition of happiness is left up to the respondent, and we do not impose a US conception of happiness on Chinese respondents, or a Chinese definition on Chilean ones. The open-ended nature of the definition results in the consistent patterns in the basic explanatory variables across respondents worldwide, in turn allowing us to control for those variables and explore variance in the effects of all sorts of other things on happiness, ranging from crime rates to commuting time to the nature of governing regimes.
Fulfilment or contentment?
At the same time, as we think about happiness as a measure of welfare with relevance to policy – something that is increasingly in the public debate – then the definition does matter. Are we thinking of happiness as contentment in the Benthamite sense, or as a fulfilling life in the Aristotelian sense? There is still much room for debate. My studies of happiness around the world suggest that respondents’ conceptions of happiness vary according to their norms such as expectations and ability to adapt. Our priors as economists and policymakers likely suggest that some conceptions of happiness – such as the opportunity to lead a fulfilling life – are worth pursuing as policy objectives, while others – such as contentment alone – are not. Yet that choice entails normative judgements and a debate which we have not yet had.
At the very least, this conundrum will give economists food for thought – about happiness and income, and beyond – for several years to come. Moreover, despite the difficulty it poses for both method and economic philosophy, it will also force us to think deeply about what measures of human well being are the most accurate benchmarks of economic progress and human development.
Deaton, Angus. (2008), “Income, Health, and Well-Being around the World: Evidence from the Gallup World Poll”, Journal of Economic Perspectives, Vol. 22, 2.
Easterlin, Richard. (1974), “Does Economic Growth Improve the Human Lot? Some Empirical Evidence”, in Paul Delvin and Melvin Reder, Nations and Households in Economic Growth: Essays in Honor of Moses Abramowitz. New York: Academic Press.
Graham, Carol and Stefano Pettinato (2002), Happiness and Hardship: Opportunity and Insecurity in New Market Economies, Washington, DC: The Brookings Institution Press).
Graham, Carol. (2010), Happiness around the World: The Paradox of Happy Peasants and Miserable Millionaires, Oxford: Oxford University Press.
Graham, Carol, Soumya Chattopadhyay, and Mario Picon (2010), “Does the Dow Get You Down? Happiness and the U.S. Economic Crisis”, mimeo, The Brookings Institution, Washington, DC, January.
Graham, Carol, Soumya Chattopadhyay, and Mario Picon. (forthcoming), “The Easterlin and Other Paradoxes: Why Both Sides of the Debate May Be Correct” in Ed Diener, John Helliwell, and Daniel Kahneman, International Differences in Well-Being, Oxford: Oxford University Press.
Stevenson, Betsey and Wolfers, Justin. (2008), “Economic Growth and Subjective Well-Being: Re-Assessing the Easterlin Paradox”, Brookings Panel on Economic Activity, April.