Up Front

Web Chat: The Economics of Happiness

Carol Graham

Times are tough for many Americans, complete with job losses and financial stress at home and the conduct of two wars abroad. How does this turmoil affect the overall happiness and well-being of Americans? Carol Graham, an expert on inequality and subjective well-being, notes that while people across the globe have a surprising ability to adapt to adverse conditions, uncertainty and doubt are much more difficult for people to deal with. This is evidenced by the fact that average happiness in the United States – defined by a set of data that measure general well-being – dropped significantly with the onset of the 2008 financial crisis. Once stability was restored and uncertainty ended, happiness levels recovered quickly.

Carol Graham took your questions in a live web chat to explore how measures of happiness may be more novel barometers for well-being than more easily defined variables such as income or wealth. Seung Min Kim, assistant editor at POLITICO, moderated the discussion.

The transcript of this chat follows.

Seung Min Kim: Hi everyone. This week’s chat deals with the economics of happiness — how do the job losses and financial stresses, as well as two wars overseas, affect the well-being of Americans? Brookings’s Carol Graham, an expert on inequality and subjective well-being, is here to talk about this topic.

12:31 [Comment From Shawn: ] Are people with more money generally happier?

12:31 Carol Graham: Money does matter to happiness, but it’s not the whole story. Being destitute is bad for happiness and richer people are, on average, happier than poor people. However, after people have “enough” money (and enough varies across societies), all sorts of other things mediate the income and happiness relationship, such as how your income compares to your neighbors.

12:31 [Comment From Jennie: ] What are some common ingredients to a happy population?

12:34 Carol Graham: In the same way that money matters to individual happiness, societies that are wealthier are, on average, happier. But there are large differences between them as there are a lot of things that come with higher levels of income, such as better public goods, better health, and better quality government, that vary across countries of similar income levels. For the most part, countries with higher levels of freedom and democracy and more social capital are happier than those with less of these things. But there are also important cultural differences across societies that are linked to happiness levels.

12:34 [Comment From Rebecca: ] How has the economic crisis affected the happiness of Americans?

12:38 Carol Graham: The economic crisis had major effects on the happiness of Americans. Average national happiness levels are typically very steady; after the crisis set in, they dipped down a lot and fell steadily – almost tracking the fall in the Dow – until the end of March 2009, the point at which the bottom stopped falling out of the markets. At that point, while markets recovered sluggishly, happiness recovered much faster and by July of 2009 was higher than at pre-crisis levels. The same respondents still reported that their economic situation was worse than before. Thus they were happier with less income but only after the uncertainty surrounding the crisis abated. Typically, people are better able to adapt to unpleasant certainty than they are to uncertainty, and this crisis is a clear example.

12:38 [Comment From Eric: ] How do Americans tend to define happiness?

12:41 Carol Graham: When we use happiness questions in surveys, we do not define happiness for our respondents. The questions are open-ended (with possible answers running from very happy to not at all happy) and the first in the survey, so they are not biased by other questions. That is what makes happiness surveys such a good research tool and lets us compare the happiness of respondents across countries and cultures (and why we find such consistent patterns in the basic correlates). That said, people in different cultures conceive of happiness differently. While there is no one definition of happiness for Americans, as a society we typically place a lot of emphasis on individual effort and opportunity, and our conception of happiness likely emphasizes those objectives more than social harmony or the collective good, as some other societies might.

12:41 [Comment From Emily: ] What implications do measures of happiness have on policy decisions?

12:47 Carol Graham: At this juncture, measures of happiness have not yet directly entered the policy arena, nor is it clear that they should (we do not know yet). These measures – and the information therein – can surely inform policy and produce better decisions. Surely understanding that health, stable relationships and employment, and friendships are very important to happiness, and that uncertainty is bad for happiness can help us think about the appropriate emphasis that we place on, for example, economic growth versus other objectives. But directly making policy decisions based on happiness surveys is potentially risky for two (among other) reasons. The first is intertemporal problems: there are many policies that may reduce happiness today in order to increase aggregate welfare tomorrow (or for our children), such as reducing fiscal deficits or reforming the health system. Happiness surveys will tell us that they are reducing happiness today, but its not clear we should avoid the policies because of that. The second is that we have not had any sort of public discussion about what conception of happiness we care most about. Happiness as contentment versus happiness as the opportunity to lead a fulfilling life, for example, are two very different definitions of happiness which would each have very different policy implications.

12:49 [Comment From Chris: ] Are there validated ‘near real time’ metrics for population scale happiness or well-being?

12:53 Carol Graham: Danny Kahneman and Alan Krueger at Princeton have conducted extensive time-use studies in the U.S. and France, which measure the happiness of individuals as they go through their daily activities. They find that commuting time and time with children are some of the least happy times, while time socializing with friends and sex are the times of the day that people report to be happiest. That said, and as I noted above, different conceptions of happiness might place different weight than time use surveys would. Surely most people value the role of their children in the big picture of their lifelong happiness; that does not mean that every trip to a soccer practice during rush hour is a happy experience on a daily basis, for example.

12:53 [Comment From Chris: ] Have any major government statistical or economic agencies begun to include these in their suite of ongoing monitoring tools alongside more conventional metrics such as producer prices, labor force participation or inflation?

12:56 Carol Graham: A number of government statistical agencies are at least exploring the prospect. The British have an office for national well being indicators, and French President Nicholas Sarkozy led a major international research effort (the Sarkozy Commission) which has called for all countries to adopt broader measures of well being as indicators of progress. The CDC has called for the incorporation of well being measures into its health indicators in this country, and Bhutan has replaced GDP with GNH or Gross National Happiness. While all of these efforts are nascent ones, it suggests that broader measures of well being – many but not all of them based on happiness surveys – are being taken seriously and are entering the policy domain and discussions.

12:57 [Comment From Eric: ] How can something as subjective as “happiness” be quantified, so that one can make meaningful comparisons between countries?

1:03 Carol Graham: While conceptions of happiness vary across countries and cultures and people, we find that the answers to general happiness or life satisfaction surveys are remarkably similar across hundreds of thousands of people across countries (even of different development levels) and over time, at least in terms of the most basic variables determining happiness, such as income, health, stable employment, and stable partnerships. And with increasing econometric innovations, we are also able to control for unobservable differences across individuals that could bias responses and their relationship to these variables. Psychologists also find validation in the way people answer these surveys. Thus while there may be differences in the way people conceive of happiness when they answer these surveys, the patterns in terms of the variables that happy and unhappy responses are linked to are very consistent. Because of the large size of the samples that we have, we have confidence in quantitative analysis of the results. That does not mean, however, that we can really add cardinal value to, for example, the difference between reporting to be very happy versus just being happy. We can, however, talk about percentage changes and probabilities of moving from one category to the other.

1:04 [Comment From Aruna: ] I am a student of International Development and I have been looking at slum economies – about how people seek an ecosystem of happiness and social connections in some of the most squalid conditions imaginable. They also have profitable small industries within slums. This is another layer of perception. Don’t you think?

1:06 Carol Graham: I have also done a lot – in fact most – of my work in developing countries. On the one level, the standard variables linked to happiness are remarkably similar. On the other, within those societies, people are remarkably able to adapt to extreme levels of adversity, to high levels of crime and corruption, and to worse norms of health, among other things. Thus the patterns in the very basic variables across societies and across people within them are very similar, regardless of development levels, but the effects of environmental and other conditions can vary a great deal, depending on what people’s norms and expectations are.

1:07 [Comment From Terry: ] How do you persuade policy-makers to take happiness as a serious indicator or barometer instead of the more traditional economic indicators?

1:08 Carol Graham: I am not sure that actively persuading policymakers is the right approach. This is a new approach in economics and it is yielding incredibly interesting and policy relevant insights into what determines human welfare and well being. And policymakers are starting to notice it and see the utility in it. I think that is a more effective route than attempting to push the usage of an approach that is still a work in progress or a nascent science.

1:08 [Comment From Mark, Greenbelt, MD: ] Don’t they have some sort of happiness index in Brunei? have you studied that? What’s that about? I guess it helps that they’ re one of the richest nations in the world.

1:12 Carol Graham: I have not looked at Brunei, but I have looked closely at Bhutan, which has GNH as a national indicator. One might worry that the focus on happiness rather than economic growth and progress would divert a country from development, but during the time Bhutan has had GNH they have reduced poverty, dramatically increased literacy, democratized, and opened up somewhat to the global economy. Its an amazing experiment. But they also achieved all of these goals under an enlightened monarchy. One could imagine a much less enlightened regime using the concept of gross national happiness to keep people without much information and with low expectations “content” without progress. The development process, for example, can be destabilizing and produce frustration and unhappiness in the short term.

1:12 [Comment From Sally: ] I’ve seen some research on the happiness levels of Democrats vs Republicans. Are there real differences? Or is that media hype?

1:15 Carol Graham: Indeed, many studies including my own research finds that Republicans are happier than Democrats, after controlling for income, religion, and a host of other things. Thus I ascribe this to very different philosophies: Republicans are more likely to believe that income and success are just rewards for hard work, while Democrats are more likely to believe that the system gives some people advantages over others and that there are injustices that need to be fixed. Having the former philosophy is, not surprisingly, linked to slightly higher happiness levels. What we do not know is the direction of causality – whether happier people are more likely to select such a view of the world, or whether having that view of the world makes them happier.

1:16 [Comment From Tom: ] Do you think it’s a good or a bad idea for governments to track the happiness levels of their citizens?

1:17 Carol Graham: Tracking the happiness levels of citizens and how they vary across time – and across domains, for example as in the relative weight of health versus income versus marriage, has the potential to provide very useful information which can be compared within and across countries, as GDP data is. This does not suggest that the information should replace GDP data but it does suggest that it could provide useful complements to it.

1:19 [Comment From Jennifer Seiden: ] They always say ‘money can’t buy happiness,’ but I think now that I’ve been poor, I’d be really good at being rich – and happy. Do you see differences in people who achieve wealth and those who are born with it?

1:21 Carol Graham: There are no direct studies that I know of that cover the effects of inherited versus achieved wealth. But we do know that many people derive great happiness from the process of achieving – the process of being immersed in work – consistent with Aristotle’s concept of eudaemonia. We also know that the happiness affects of a change in status, such as a promotion, are much longer lasting than those of an income gain. All of this suggests that there is an additional and perhaps more lasting effect of the achievement component of income than of the owning component.

1:21 [Comment From Bill S.: ] What do you say to people who would argue this is frivolous research? What practical, real-world applications does this aid?

1:26 Carol Graham: If used inappropriately, for example in polling people about what policy today would make them happy, then I think surely the research could be accused of being frivolous. But we do not directly ask people what makes them happy. We take a measure of their general happiness in the first survey question and then see how variance in these measures correlate with a host of other variables of interest, ranging from income to employment status to commuting time. There are a host of questions that traditional approaches, which are based on the analysis of revealed consumption choices, cannot answer. In particular, we cannot measure the welfare effects of institutional arrangements that individuals are powerless to change (like economic volatility or inequality) with traditional approaches; we can with happiness surveys. We also cannot explain behaviors that are driven by norms, addiction, or self control problems, such as low expectations among the very poor or obesity or smoking, with optimal choice, revealed preferences approaches. We can measure the welfare effects of these at times perverse choices via happiness and well being surveys. That can hardly be a frivolous enterprise.

1:26 [Comment From Abbey Brumberg: ] If you had unlimited funding, what areas of happiness research would like to pursue at the moment?

1:28 Carol Graham: I am very interested in the well being effects of inequality and of macroeconomic volatility; in the effects of different health conditions on happiness (and in particular I am now studying obesity); and in the broader question of the implications of individual adaptation (for example to high levels of crime and corruption) for societies as a whole. I am trying to conduct research in all of these areas. Unlimited funding would be a great help!

1:29 [Comment From Timothy: ] Are there ways that happiness research could be used to inform strategies for ‘winning the hearts and minds’ of citizens in countries dealing with terrorism? Could happiness studies be used to gain, build and maintain the support of citizens in these countries?

1:31 Carol Graham: I do not think we can use happiness surveys to directly win hearts and minds. But we surely can use happiness surveys to understand what makes people happy or, more relevant, unhappy and frustrated, across and within societies. WE can explore the role of religion, conflict, income differentials, and a host of other things, which would in turn help us understand how and why insurgent groups might be supported by some parts of particular societies. I most recently did the first study of happiness in Afghanistan and it yielded a number of novel and at times counter-intuitive insights.

1:32 Seung Min Kim: Thanks everyone for chiming in this afternoon. Great conversation. Have a wonderful week!