Two more complaints about attempts to measure the relationship between income and emotional well being.

Continuing yesterdays discussion about happiness surveys (Interesting but not correct), I want to raise two additional issues, income expectations and government transfers. The first concern is that the survey design on emotional response  uses questions about current income (“The GHWBI asks individuals to report their monthly family income in 11 categories.”). However, it isn’t this month’s income that matters nearly as much as the households current wealth and lifetime remaining income in determining consumption and consumption is likely the principle mechanism by which income generates happiness (since status is a zero-sum game).   Deaton is a major contributor to our understanding of the permanent income hypothesis, so I’m not sure why he omitted this.

If income consists of a permanent and a transient component, then conditionally on observing an extreme income you are likely experiencing a temporary income shock. Many of the high income people in their data are those with normal permanent income but  experiencing positive income shocks. We know that households tend to adjust consumption to permanent  shocks and not to temporary ones. Therefore, a large fraction of the high income are not rich, they are only temporarily high income and we won’t expect them to have much higher consumption and therefore higher happiness than the lower income groups. This means that their statistical test is averaging the happiness / emotional experiences of two groups the permanently and the temporarily rich. And the temporarily rich may well be experiencing periods of high stress due to busy times in their professional life. Depending on the relative sizes of these groups you could get almost anything. Their results (High income improves evaluation of life but not emotional well-being) largely bear this out.

Table 2. Tests for income satiation of life evaluation and emotional well-being
Positive affect Blue affect Stress Ladder
Top vs. second 0.0035 0.0013 0.0055 0.2264
t value (1.9) (0.6) (1.5) (19.4)
Second vs. third 0.0082 −0.0131 0.0016 0.2268
t value (4.4) (5.7) (0.4) (19.7)
Observations
Top group 72,744 73,104 73,109 73,068
Second group 40,136 40,291 40,301 40,283
Third group 88,887 89,278 89,290 89,245
The coefficients reported are the differences inmean outcomes between the
two indicated income categories. The top category is >$10,000/mo, the second
category is $7,500–$9,999/mo, and the third category $5,000–$7,499/mo. SEs are
corrected for spatial clustering within zip codes.

As the title suggests, people evaluate their life as better at all groups of higher income observed. But on the other indicators they get an ambiguous effect. The rich appear to have more worries (lower blue) and slightly less stress (not statistically significant or economically so) .

“We defined positive affect by the average of three dichotomous items (reports of happiness, enjoyment, and frequent smiling and laughter) and what we refer to as “blue affect”—the average of worry and sadness. Reports of stress (also dichotomous) were analyzed separately (as was anger, for which the results were similar but not shown) and life evaluation was measured using the Cantril ladder. The correlations between the emotional well-being measures and the ladder values had the expected sign but were modest in size (all <0.31).”

But again, we think happiness depends on consumption which is proportional to permanent income (P) not total income (P+T). The more variance in T the closer the regression coefficient of emotional response to P+T (the estimates in this table)  will shrink to zero regardless of the relationship between emotional response and P.   That attenuation effect combined with the composition effect (high income can occur at times of emotional stress) may explain why Blue Affect and Stress give insignificant, small, and or even wrong signs. But the well being ladder doesn’t suffer from this. We know when we are in the busy season that incomes are high but this will eventually end and we can relax to spend our windfalls. The ladder measure allows the conscious mind to opine on how good things are for the the present and future in a way that the more reptilian emotional measures cannot.

My second complaint is an absence of recognition of government transfers. By asking about current income they omit that we live in a system with a steeply progressive tax structure with extensive transfers to the poor and elderly. There are also a large number of public good and government services that the government provides at no marginal cost (say public school education and national defense). In light of this, there is no wonder that they find diminishing marginal return to income. There are two ways that this works. One, if you make minimum wage with a spouse and child in a state without sales tax your marginal tax rate is on the order of -15% because of the EITC. If you are single and childless and make $120,000 a year in a high tax state you might have a tax rate as high as 50%. Those differences alone explain 2/3 differences in the marginal utility of income, not from the curvature of the utility function but just from taxes. So maybe the lesson here is that taxes can have a big impact on the marginal utility of income.

Add in government services and transfers and that may explain much of the rest. But if public goods depend on average national income and public and private goods are compliments in utility (clean air and wilderness vacations, radio stations and radios) , then private income could have a diminished role in individual happiness even  as it has a huge role in aggregate happiness. Indeed, the WSJ had an article today (The Rise of Consumption Equality) discusses how consumption is much more equitable than incomes and that is possible only because of general social wealth. Similarly, sub-Saharan has experienced a jump in quality of life and life expectancy without increasing incomes because of global public goods like the invention of antibiotics and vaccines funded by rich countries elsewhere. This idea teaches a slightly different lesson, that you personally shouldn’t worry too much about money if you can may a decent (say high 5 figures) wage as long as you get to live in an affluent nation.