The economics of happiness (or "stated well-being") has enjoyed a decade-long surge, including recent articles in top general-interest journals, cover stories in Time and The Economist, surveys in the Journal of Economic Literature (Frey and Stutzer 2002) and the Journal of Economic Perspectives (Kahneman and Krueger 2006), and dozens of books for both academic and mainstream audiences. Most of the attention has focused on questions about what makes people happy (money, marriage, employment, etc.). Recently, however, economists have attempted to use happiness surveys to address a central question of public policy – valuing public goods.
One basic purpose of government is to provide goods that market forces will not, and a central task of economics is to place monetary values on those goods so that governments can prioritize projects. That task, however, requires knowing how much people will be willing to pay for these non-traded goods. Methods have included travel cost models, hedonic housing price estimations, and more recently, and controversially, contingent valuation or stated-preference surveys. Now a new tool has entered the field – happiness economics.
The fundamental idea is straightforward. Survey people as to their incomes, how happy they are, say for example on a scale from one to seven, and any other demographic data available. Collect data, perhaps from other sources, on the amount of the local public good present for each respondent. Then estimate respondents' happiness as a function of their incomes, the public good, and other observable data. Presumably happiness will increase with income and with the quantity of the public good, all else equal. The estimated function tells us exactly how much income, on average, people would be willing to give up (perhaps in higher taxes) in order to receive more of the public good.
The whole approach can be summarised in one paragraph of math. First estimate
where H is an individual's response to a question about happiness, P is the quantity of the local public good, Y is the respondent's income, X is a vector of other characteristics of the respondent and the locality, and ε is an error term. Then totally differentiate the estimated function, set dH=0, and solve for the marginal willingness to trade income for the public good.
This approach has been used to value airport noise reduction (van Praag and Baarsma, 2005), flood control (Luechinger and Raschky, 2009), terrorism protection (Frey et al., 2009), income inequality (Luttmer, 2005), climate (Rehdanz and Maddison, 2005), and air quality (Welsch, 2009; Luechinger, 2009). Similar approaches have been used to study the tradeoff between unemployment and inflation (Di Tella et al., 2001), whether cigarette taxes make smokers better off by forcing them to smoke less (Gruber and Mullainathan, 2005), and whether sick people have higher or lower marginal utilities of income, an important question for health insurance (Finkelstein et al., 2008).
Not surprisingly, this new approach to valuing non-traded goods is fraught with problems. One of the biggest involves the fact that happiness, income, and the public good are determined simultaneously and may each depend on one another. For example, while income may make people happy, happier people may also earn higher incomes. Only two papers address this (Luttmer, 2005; and Powdthavee 2009). Both use instrumental variables for income and find that the income coefficient is much larger once they account for its endogeneity, suggesting that the approach overstates the value of the public good. If the income coefficient is larger, less income will be necessary to compensate for a decrease in the public good.
Perhaps more troubling, few studies account for the fact that the level of the public good may be determined simultaneously with happiness. People desiring a particular public good may relocate to areas with high levels of it or lobby their local governments to provide more of it. Afraid of crime? Live in a gated community, or vote for higher police expenditures. Asthmatic? Move to a place with clean air, or vote for tough pollution control laws.
A related concern is that people may become habituated to whatever level of public good they currently experience. Many of us were not terribly bothered by smoke in restaurants until smoking in public places was banned, and now a tiny whiff of smoke makes us unhappy. We were habituated to the smoke, and now we are habituated to its absence. If asked to compare a restaurant today to its smoky equivalent years ago, each would make us equally happy. But if asked to compare a restaurant today to its smoky equivalent next door, the smoke-free restaurant would make non-smokers happier, as they have become accustomed to the lack of smoke.
This poses an enormous problem for the valuation of public goods using happiness data, because by definition everybody in the same region experiences the same level of the public good at the same time. Every one of the studies cited above regresses individuals' happiness on regional annual averages of the public good (terrorism, flood control, airport noise, and pollution). If people in a particular region locate there because of its public good, lobby their governments to provide the good, or become habituated to the local level of the good, the public good coefficient will be underestimated, and the valuation using this approach will be too small.
Using individual variation in public good quality
In a recent working paper (Levinson, 2009), I try to address this problem by using daily local air quality. This is a special case of a public good that varies daily, for reasons external to any particular respondent, and presumably too quickly for anybody to become habituated. I use the General Social Survey (GSS), which each year asks several thousand people throughout the US "taken all together, how would you say things are these days? Would you say that you are very happy, pretty happy, or not too happy?" I obtained from the GSS the confidential information about the US county or city in which each respondent was surveyed and the date on which the respondent was surveyed. I used that information to match responses to EPA data on the air pollution on that date in that place, and NOAA data on the temperature and rainfall on that date in that place.
I then estimate equation (1), and use equation (2) to calculate that, on average, people are willing to forego about $40 of annual income for a one-standard-deviation reduction in the particulate matter in the air for one day. I try numerous specifications, with various control variables and interactions, and the point estimates range from about $20 to about $60. How large is this change? A one-standard-deviation change in pollution amounts to 14.4 micrograms per cubic meter (μg/m3) of particulates. The average particulate reading in the sample is 30.4 μg/m3, so the change represents a 50% increase (or decrease) in pollution, which corresponds to a move from an average county in the US to one of the worst-polluted counties, Riverside or San Bernardino, CA, or approximately the reduction in particulates attributed to the 1970 and 1977 Clean Air Acts.
In the end, we should probably not put too much faith in the $40 estimate. It makes strong assumptions about preferences by comparing the stated happiness of different individuals. It is based on temporary differences in pollution, while happiness does not seem responsive to longer-term differences in pollution. And it may overstate the value by largely ignoring the simultaneous determination of income and happiness.
Nevertheless, this new approach to valuing public goods contains promise. Its shortcomings differ from those of the typically used approaches, so that if nothing else, it serves as a useful point of comparison. It comes from nationally representative surveys, and so can be used to assess how willingness to pay for public goods varies over time and by region, age, income, education, and current public good levels.
Perhaps most importantly, the results add further evidence that self-reported happiness captures something meaningful about the people's environments, and in this case support a sizeable willingness to give up income in return for improved environmental quality.
Di Tella, Rafael, Robert MacCulloch and Andrew Oswald. 2001. "Preferences over Inflation and Unemployment: Evidence from Surveys of Happiness" American Economic Review 91(1): 335-341.
Finkelstein, Amy, Erzo F.P. Luttmer, and Matthew J. Notowidigdo. 2008. "What Good is Wealth without Health? The Effect of Health on the Marginal Utility of Consumption" NBER Working Paper 14089.
Frey, Bruno S. and Alois Stutzer. 2002. "What Can Economists Learn from Happiness Research?" Journal of Economic Literature Volume 40, Issue 2, June 2002.
Frey, Bruno, Simon Luechinger and Alois Stutzer. 2009. "The Life Satisfaction Approach to Valuing Public Goods: The Case of Terrorism" Public Choice, forthcoming.
Gruber, Jonathan and Sendhil Mullainathan. 2005. "Do Cigarette Taxes Make Smokers Happier?" Advances in Economic Analysis and Policy 5(1)
Kahneman, Daniel, and Alan B. Krueger 2006. "Developments in the Measurement of Subjective Well-Being." Symposium: Happiness Economics, Journal of Economic Perspectives, 20(1): 3–24.
Levinson, Arik. 2009. "Valuing Public Goods Using Happiness Data: The Case of Air Quality" NBER Working Paper 15516.
Luechinger, Simon. 2009. "Valuing Air Quality Using the Life Satisfaction Approach" Economic Journal 119(536):482-515.
Luechinger, Simon and Paul A. Raschky. 2009. "Valuing Flood Disasters Using the Life Satisfaction Approach" Journal of Public Economics, forthcoming.
Luttmer, Erzo F.P. 2005. "Neighbors as Negatives: Relative Earnings and Well-being" Quarterly Journal of Economics 120(3): 963-1002.
Rehdanz, Katrin and David Maddison. 2005 "Climate and Happiness" Ecological Economics 52(1): 111-125.
van Praag, Bernard M. S. and Barbara E. Baarsma. 2005. "Using Happiness Surveys to Value Intangibles: The Case of Airport Noise" The Economic Journal 115(500): 224-246.
Welsch, Heinz. 2009. "Implications of happiness research for environmental economics" Ecological Economics, in press.