Apologies for missing a post last week, last week I was at the Association of Public Policy and Management (APPAM) annual meeting, and this week I am in South Africa (hopefully some comments on South African inequality issues soon). Below are some impressions from one of the APPAM sessions, “The Measurement and Evaluation of Happiness and Subjective Well-Being.”
In an excellent post, Ben summarized the state of our knowledge about the relationship between income and subjective wellbeing (loosely called happiness). For many years it was believed that a society’s economic development and its average level of happiness were not correlated (the so called “Easterlin paradox”). But there is now consensus that income and self-rated wellbeing are robustly associated both across and within countries. Stevenson and Wolfers (2008), for example, find that averaged across a variety of settings, increasing logged income by 1 point increases subjective wellbeing by 0.4 points.
Does more income cause happiness? We can easily imagine why the causal arrow might instead run from happiness causing income instead of the other way around. People with more cheerful or optimistic personalities are more likely to be successful in the labor market. Or, some other variable, such as the quality of a person’s childhood, their personal appearance, or their culture may simultaneously cause happiness and income.
At APPAM, several papers attempted to disentangle the money-wellbeing relationship using natural experiments or natural variation.
Debt Relief in India
Christopher Robert examines whether heavily indebted farmers in India that received a one-time government bailout were more likely to report higher levels of life satisfaction. In 2008, the government took on about $14.6 billion worth of agricultural debt and tens of millions of Indian farmers benefited. Average relief amounted to roughly half the annual consumption expenditure of the average rural household. Small landholders (with parcels less than 5 hectares) received 100% debt relief, while those with parcels larger than 5 hectares received only 25% debt relief, and were required to repay the remaining 75%. This sharp discontinuity permits Robert to compare the subjective wellbeing of households on either side of the boundary, as well as to compare households by total overall debt relief (since some families had much larger debts wiped out than others).
Using a method that converts debt into an equivalent value in income, Robert is able to estimate the change in the amount of subjective wellbeing produced by more income on the margin. Robert presents a range of estimates based on a number of assumptions about the value of debt income in monetary terms, and most of these are large, even in comparison to the consensus estimate of 0.4 from Stevenson and Wolfers (mentioned above). Most of the observable difference in levels of life satisfaction occurs in the group that received 100% debt relief, as shown in the figure below.
Elderly Pensions in South Korea
Erin Hye-Won Kim examines whether life satisfaction increased among Korean elders before and after the implementation of a substantial non-contributory government pension. Prior to the implementation of the pension program, Korean elders mainly relied on family members to provide financial support, and average levels of financial support were very low on average: in 2005 Korean elders were among the poorest elderly in the OECD and were much poorer than non-elderly Koreans. Perhaps indicating the dire circumstances of Korean elders, the suicide rate in this population has been more than ten times higher than non-elderly Koreans and is staggeringly higher than the rest of OECD elders.
The government pension program was passed in 2007 and enacted in 2008. Kim uses longitudinal data with interviews from 2004, 2006, and 2008. Kim finds that when the government reformed its elderly pension scheme, average monthly income increased among most elders (although there was some crowdout of family giving, so that every dollar of pension provided resulted in an average of 30 cents less from family). Average life satisfaction also increased significantly among elders eligible for the pension (the comparison group had income that was too high to qualify). Kim does not provide standardized estimates of the average difference in life satisfaction, so comparison with Wolfers and Stevenson’s estimate is not possible.
Big Picture Thoughts
It was exciting to see some quasi-experimental work by fellow graduate students that tries to unravel the happiness income puzzle. (Here’s another nice paper from the session that exploits within twin pair differences from China). It is important for policy purposes to understand how much cash transfer policies would be expected to increase happiness, life satisfaction or other dimensions of wellbeing. But before we promote such policies, it is also important to understand the underlying psychology of happiness: why do life satisfaction and experiential happiness not respond in the same ways to income? What are the likely points of satiation, and do they vary across societies? When do non-monetary policies do more to promote happiness than cash? (Important to know for birthdays, holidays, and anniversaries too).