Can Money Buy Happiness? A macroeconomic multinational perspective on the Easterlin Paradox
Harvard College ‘13
In 1978, Richard Easterlin suggested that wealth corresponded to happiness within nations, but not between them. This idea, known as the Easterlin Paradox, was supported through time series measurements from several European and North American countries. In this paper, we re-examine the paradox using the most recent macroeconomic data. Our analysis, which includes almost all countries, shows that contrary to predictions made by the Easterlin Paradox, income equality does not impact happiness while GDP per capita and median household income do. Our results also show that government economic policies have little impact on the population’s happiness. Finally, we use the concept of GDP per hour worked to help explain why the Easterlin paradox might be observed in a handful of countries.
The field of economic psychology was relatively stagnant during the mid-20th century. This was primarily due to the rigidity of long-standing macroeconomic models and the inflexibility of well-established economic methods at the time. Indeed, few developments, aside from improved statistical techniques, were witnessed then. Economists faced with the challenge of assessing the efficacy of public policies or quantifying changes after economic interventions defaulted to established economic measures as a proxy for public well-being.
It is no surprise then that the introduction of happiness measures into economics was hailed as a great addition to the analytical tools economists had at their disposal (Di Tella & MacCulloch, 2006). These new measures offered economists the ability to judge the well-being of society directly rather than rely on indirect—yet more objective—economic measures such as GDP per capita, Gini, etc.
This curious intersection between economics and psychology enabled both disciplines to utilize the tools of the other to better understand, and refine, their own models. For instance, these measures were instrumental to the establishment and refinement, of Martin Seligman’s Positive Psychology (Seligman, Steen, Park, & Peterson, 2005; Fordyce, 2005). In the field of economics, happiness measures were used to better appreciate how economic indices relate to social well-being beyond the theoretical framework of economics—for example, the extent of the applicability and usefulness of the Human Development Index (HDI) to society (Leigh & Wolfers, 2006).
The study of how happiness changes with social or economic factors has been slowly gaining traction within the scientific community since the 1980s. Several factors in addition to money have been investigated for a possible connection to happiness. For instance, the link between years of education, intelligence, and heath on one hand and happiness on the other was investigated in a limited 1998 Dutch study, where an intermediate level of non-vocational schooling was found to maximize happiness. IQ was reported to impact health but not happiness (Hartog & Oosterbeek, 1998). The variation in all these factors was found to be slight. Similarly, in a 2002 longitudinal Swedish study of over 5000 individuals, happiness, unsurprisingly, correlated positively with health, income, and education (Gerdtham & Johannesson, 2002). As for income inequality, measured through the Gini coefficient, Johns & Ormerod (2008) found that it had little effect on US well-being. In fact, while the Gini index has been climbing since the 1970s, signaling an ever increasing gap between the rich and the poor, happiness indicators and reports of subjective well-being have fluctuated around the same baseline level.
Other factors, such as average temperature (Rehdanz & Maddison, 2005), age (Cameron, 1975), and marital status (Kamp Dush & Taylor, 2012), have been investigated as well, with varying effects on well-being. None of those effects, however, were as prominent or as controversial as that of money on happiness as manifested by the Easterlin paradox.
The concept of Easterlin paradox was presented early on in the introduction of happiness economics. This concept was named after Richard Easterlin, whose research found happiness to be significantly correlated with income within countries but not between them (Easterlin, 1974). In other words, the richer you are relative to those around you, the more likely you are to report being happier. Being in a rich country, Easterlin argued, will not necessarily increase your happiness.
Based on those findings, Easterlin (1994) proposed that increasing the income of everyone in society would not, on its own, increase the overall happiness level in that society. The best reforms, he argued, are those concerned with income disparity within society. In 2001, Easterlin expanded his theory to an individual level by suggesting that individual happiness does not increase linearly with income; rather, there is a significant correlation at the lower income bracket that levels off at higher incomes. He attributed this effect to increasing aspirations that match increasing income, offsetting any increase to reported well-being or happiness (Easterlin, 2001). A number of studies have come to support Easterlin’s conclusions. For instance, a 2011 longitudinal study by Oishi et al. of Americans between the years 1972-2008 found that wealth inequality, as measured by the increasing Gini index, indeed corresponded with a decreasing level of disatisfaction among Americans.
Easterlin’s findings were not without critics. In an extensive review, Rafael Di Tella (2003) from Harvard Business School looked at well-being data from the 1970s to the 1990s for European countries as well as the United States. The results indicated that macroeconomic factors influenced the reported subjective national well-being reports although the influence was weak at best (0.209 for GDP per capita and 0.281 for benefit replacement level). Hagerty & Veenoven (2003) came to similar conclusions from time-series analysis of 21 countries (mostly European nations and the United States). They found that happiness did increase with increasing national income and with increasing growth in most of those countries. It might be worth noting tangentially that there are also those who are at odds with both camps, contending that happiness is way too complicated to be measured in a meaningful way for practical utility. Johns (2007) reasserts this view by demonstrating that happiness cannot be predicted from public spending, violent crime, unemployment, and other factors. For the purposes of our analysis, we will assume that happiness measures, which are discussed in the next section, are good approximations of a country’s subjective well-being.
How is happiness measured?
Quantifying happiness, or well-being, in an individual is a difficult task. Happiness is extremely hard to define—even broadly—and is, by its very own nature, subjective. Asking respondents how they feel today might not reflect how they felt that week, that month, or even their general mood that year. Their overall concerns might not change drastically over their lifespan (e.g. financial and familial stability), but their daily circumstances will, creating wild fluctuations in their life satisfaction evaluations. Within the fields of psychology and economics, there have been attempts at addressing the subjective nature of this problem by proposing “objective” well-being measures. The most famous examples of those measures can take one of the following two forms. In his book, Kahneman (2003) suggests the concept of objective happiness derived from a record of point-instant experiences. This moment-based approach eliminates biases of end effect and peak effect, giving a less subjective evaluation of happiness. The other approach is derived from economics. It usually resorts to calculating objective well-being from other measures such as health and wealth indices (Fox, 2012).
Both of these approaches to measuring happiness have inherent problems that render them impractical or even meaningless for our purposes. A moment-based approach is difficult with a single individual, let alone with a multinational large sample. An objective measure derived from other indices will be mechanical at best and would positively correlate with wealth, defeating the purpose of examining the paradox.
Nevertheless, large sample surveys address of all these problems. Such surveys take a happiness snapshot of a large number of people that, theoretically, averages out to the population’s collective well-being. This method has the attractive advantage of being relatively accurate and very fast to conduct, providing a wealth of information effortlessly through a simple and short survey.
The main question we are interested in is whether the Easterlin paradox still applies today on an international scale. We set out to include all countries in our analysis—something that, to the best of our knowledge, has not been done before. We postulate that should the paradox apply to a modern landscape, a greater income inequality will lead to decreased happiness within a population (or a decreased segment of the population expressing happiness). A result in which the Gini index shows no correlation with happiness would cast doubt on the Easterlin Paradox as an international phenomena.
As “wealth” can take more forms than simply the gross domestic product and the Gini coefficient, we also consider how government economic policy (government spending, savings, inflation and purchasing power parity, etc.) impacts the population’s happiness.
Economic and Social Methods
The measures we are interested in are:
• The Gini index: a measure of income inequality, where a score of 100 corresponds to the country’s wealth concentrated in a single person, and a score of zero means everyone in the country is equally wealthy;
• GDP per capita: the total amount of goods and services produced in a country per year divided by that country’s mid-year population. A per capita GDP of $15,000 or more is usually considered that of a developed nation. The world’s average is about $11,000;
• Gross national savings: an estimate of the country’s total income minus total expenditure;
• Unemployment rate: the number of those without jobs divided by the number of those of working age;
• Current government account balance: the sum of all cash transfers to and from a government, including investments, payments, and trade;
• Average consumer price inflation: inflation on consumer product prices;
• Implied PPP conversion rate: comparison of the prices between countries to obtain the same goods and a measure of how expensive a country is relative to others;
• Working hours: the total number of working hours per year for an average worker in a given country;
• Median household income: the average income made by a household in a given year after tax deductions.
To examine the latest data, we referred to the most recently available International Monetary Fund (IMF) database to acquire the following measures: GDP per capita, implied PPP conversion rate, gross national savings, average consumer price inflation, unemployment rate, current government account balance and population size.
The median household income and the average working hours were obtained from the OECD database.
The Gini coefficient index data was obtained from the World Bank database.
The happiness index scores were acquired through an extensive Gallop poll that covered 155 countries and thousands of respondents. The respondents were asked to reflect on their life satisfaction, as well as how they felt the previous day. The answers to both questions were given on a scale of 0-10. The data provided the percentage of those thriving (i.e. those who provided consistently high scores), percentage of those suffering (those who consistently provided low scores), as well as the average satisfaction rate.
All data were matched temporally, by year.
To test the idea that income inequality, measured through the Gini index, leads to decreased happiness, we regressed the percentage of people who said they were happy in a country with the Gini coefficient for that country, matched by year (in this case, 2010). The results suggest that there is no clear relationship between these two variables (R2 << 0.01). In other words, some countries with high-income inequality (and high Gini coefficient) still have very broad segments of the population responding that they were thriving (Figure 1). Therefore, while it is true that Denmark, the happiest country on earth according to Gallup’s data, has a very low Gini score, Japan, the Ukraine, and Belarus also have comparably low Gini scores but have less than 40% of the population indicating that they are thriving.
Another way to test the Easterlin Paradox is to examine the relationship between GDP per capita and happiness. This relationship has been examined multiple times before in other studies, but we decided to test whether it still applies to the newest set of data. Our results indicated that, consistent with previous findings, GDP is still a very good predictor of national happiness (R2 = 0.57, Figure 2).
The following regression equation is moderately robust in predicting the percentage of respondents who indicated they are thriving based on the country’s GDP per capita alone:
% happy = -51 + 21 ∙ log(GDP) .
In other words, for every 10-fold increase in GDP (e.g. $1,000 to $10,000), the percentage of respondents who indicate they are thriving increases by 21%.
In a rebuttal by Easterlin (2005) to Hagerty and Veenhoven (2003), Easterlin asserts that GDP per capita is a poor predictor of the populations’ income, and therefore, one needs to look more closely at what individual units (be it persons or households) are making to witness the paradox. In other words, Easterlin argued that GDP per capita poorly translates into actual income for individuals. To test this hypothesis, we looked at the Organization for Economic Co-operation and Development (OECD) countries (n = 34). Although these countries tend to be wealthier on general, the trend is still generalizable. Our analysis of those countries revealed a positive correlation between happiness and median household income (R2 = 0.42, Figure 3a) but a negative correlation between happiness and number of total work hours per year (R2 = 0.31, Figure 3b).
Examining all the other macroeconomic indicators yielded statistically insignificant results with the exception of unemployment rate, which was a moderately good predictor of the percentage of people who responded that they were thriving (Table 1).
The results from our analysis clearly show the Gini coefficient is an extremely poor predictor of happiness. Indeed, if the Easterlin paradox holds true, it is expected that increased inequality (and therefore increased Gini score) would lead to decreased percentage of people reporting they are happy. This is not the case, however, in our cross-country dataset. Moreover, our analysis of the GDP per capita confirms what other studies have found so far: a positive correlative relationship exists between income and happiness (Stevenson & Wolfers, 2008).
To test Easterlin’s 2005 rebuttal to Hagerty and Veenhoven, we looked at the median household income for OECD countries. Results mirror what we found using GDP per capita alone with similar correlation coefficients, i.e. that increasing income leads to increasing happiness.
Having demonstrated that the existence of the Easterlin Paradox is doubtful on a global scale, we are faced with the problem of trying to explain how results from certain countries appear to support the paradox. In these countries (e.g. US, China, Japan), an increase in wealth does not correspond to increasing happiness from time series measurements (Easterlin, 2001). Those countries have been used extensively by Easterlin to support his theory. For this, we have to turn to an interesting and statistically significant result from our analysis: the inverse correlative relationship between the number of working hours and happiness. While a positive relationship exists between income and working hours, and money does seem to bring people happiness, too much work can lead to increasing dissatisfaction. Put differently, countries that make the most money with the least number of working hours are the happiest. This prompted us to suggest the introduction of a new variable to capture this quality: GDP per hour worked, which we shall refer to as macroeconomic work efficiency (MWE). We defined MWE as:
MWE = GDPpc ∙ (whrs)-1 ,
where GDPpc is the GDP per capita in USD and whrs is the average number of working hours per annum. Since GDPpc = Yt/Nt, where Yt is the Gross National Product (GDP), and Nt is the population size at mid-year, the MWE can be expressed as:
MWE = Yt /(Nt ∙ whrs) .
Nt∙whrs is equivalent to the total number of hours worked in a certain country. Therefore, MWE can be expressed as the ratio:
MWE = Yt / (∑whrs) .
This new measure, the macroeconomic equivalent of the average hourly pay, had a modestly high correlative relationship with happiness (R2 = 0.58 compared to R2 = 0.46 for a linear GDP relation). It measures the amount of money earned per unit time worked. The country that had the highest MWE in our dataset, the Netherlands, was also one of the happiest; its MWE was $34/hr.
Consequently, we attribute observations supporting the Easterlin paradox to a stagnating MWE. In other words, the return per unit time worked has remained relatively stable over time in countries where happiness does not seem to keep up with income alone. Further research is needed to support this hypothesis.
Looking at other factors, we found slightly significant relationships with happiness: unemployment rate (R = 0.275), current government account balance (R = 0.168), and government debt (R = 0.142). From these results, we have the counterintuitive outcome that government spending and saving—or economic policy in general—have little to do with a nation’s happiness. The only way governments can have a meaningful impact, from an economics standpoint, is to inject more money into the system. Juxtaposing this result with the extent to which people overestimate the impact of their governments’ economic policies on their well being (a la protests on Greece, Italy, and Portugal austerity measures) further underscores the counterintuitive nature of this conclusion.
Cultural complexities, which render more involved happiness calculations difficult, may attribute to this result. Indeed, what is considered generous government spending in one country might be considered dismal in another. Additionally, government spending or saving as well as its fiscal responsibility might not directly translate into a perceptible impact on citizens, further clouding the relationship between the two.
On an international scale, the Easterlin paradox does not seem to apply. Basic predictions made by the paradox regarding GDP and income inequality are patently violated from the most recent macroeconomic and social data.
Notwithstanding these results, government policies seem to have very little effect on the happiness of the population from our results of several common economic policies’ outcome measures.
Finally, factoring in working hours into the happiness model made it a better predictor of happiness. These results confirm Pouwelsa’s (2008) suggestion that money’s utility comes at a cost that needs to be considered. We also believe that further research into working time patterns is needed to explain why the Easterlin Paradox, as a local anomaly rather than a global pattern, applies to a handful of countries.
Cameron, P. (1975). Mood as an Indicant of Happiness: Age, Sex, Social Class, and Situational Differences. Journal of Gerontology.
Di Tella, ,. R., MacCulloch, R. J., & Oswald, A. J. (2003). The Macroeconomics of Happiness. The Review of Economics and Statistics.
Di Tella, R., & MacCulloch, R. (2006). Some Uses of Happiness Data in Economics. The Journal of Economic Perspectives.
Easterlin, R. A. (1974). Does Economic Growth Improve the Human Lot. Nations and Households in Economic Growth: Essays in Honor of Moses Abramovitz.
Easterlin, R. A. (2005). Feeding the Illusion of Growth and Happiness: A Reply to Hagerty and Veenhoven. Social Indicators Research.
Easterlin, R. (2001). Income and Happiness: Towards a Unified Theory. The Economic Journal.
Easterlin, R. (1994). Will raising the incomes of all increase the happiness of all? Journal of Economic Behavior and Organization.
Fordyce, M. (2005). A Review of Research on the Happiness Measures: A Sixty Second Index of Happiness and Mental Health. Social Indicators Research Series.
Fox, J. (2012). The Economics of Well-Being. Harvard Business Review .
Gerdtham, U.-G., & Johannesson, M. (2002). The relationship between happiness, health, and socio-economic factors: results based on Swedish microdata. Journal of Socio-Economics.
Golden, L., & Wiens-Tuers, B. (2006). To your happiness? Extra hours of labor supply and worker well-being. Journal of Socio-Economics.
Hartog, J., & Oosterbeek, H. (1998). Health, wealth and happiness: why pursue a higher education? Economics of Education Review.
Johns, H., & Ormerod, P. (2007). Happiness, Economics and Public Policy. Social Science Research Network.
Kahneman, D., Diener, E., & Schwarz, N. (2003). Well-Being: The Foundations of Hedonic Psychology. Russell Sage Foundation.
Kamp Dush, C. M., & Taylor, M. G. (2012). Trajectories of Marital Conflict across the Life Course: Predictors and Interactions with Marital Happiness Trajectories. Psychological Bulletin.
Leigh, A., & Wolfers, J. (2006). Happiness and the Human Development Index: Australia Is Not a Paradox. The Australian Economic Review.
Oishi, S., Kesebir, S., & Diener, a. E. (2011). Income Inequality and Happiness. Psychological Science, 22 (9), 1095-100.
Pouwelsa, B., Siegersa, J., & Vlasblom, J. D. (2008). Income, workinghours, and happiness. Economics Letters.
Rehdanz, K., & Maddison, D. (2005). Climate and happiness. Ecological Economics.
Seligman, M. E., Steen, T. A., Park, N., & Peterson, C. (2005). Positive Psychology Progress: Empirical Validation of Interventions. American Psychologist.
Stevenson, B., & Wolfers, J. (2008). Economic Growth and Happiness: Reassessing the Easterlin Paradox. Brookings Papers on Economic Activity, 1-87.