"Does money buy happiness?" is the kind of open-ended question that inspires great novels and Baz Luhrmann movies, but it also has spawned quite a research program in economics.
The University of Southern California's Richard Easterlin kicked the discussion off in 1974 when he proposed his now-famous "Easterlin paradox". He compiled happiness surveys, both within and between countries, and found that both types of data suggest that reported happiness does not actually increase as people or countries get richer. Here, for instance, is a scatterplot of real gross national product per capita (the then-current measure of how rich countries were) versus average self-reported happiness levels* in 1960.
There's basically no relationship. West Germany is close to 20 times richer than Nigeria, and no happier for it. Hence, the paradox: how can it be that such a large wealth disparity wouldn't show up in the happiness data at all? And if that really is the case, should we be focusing as much on increasing growth?
The answer, economists Betsey Stevenson and Justin Wolfers found, was that Easterlin was just wrong. There is a clear upward relationship between income and happiness. It's just logarithmic: the happiness value of the next dollar you earn is always worth less than the one you earned before it.
Further, there's a difference between happiness (the answers people to give to questions like "Taking all things together, would you say you are: 'very happy,' 'quite happy,' 'not very happy,' 'not at all happy?'") and life satisfaction (the answer they give to questions like "All things considered, how satisfied are you with your life as a whole these days?"). Easterlin treated the two measures as roughly synonymous (the graph above uses life satisfaction, for instance), but they're not. Stevenson and Wolfers find a stronger correlation between income and life satisfaction than they do between income and happiness, for instance.
Stevenson and Wolfers also found that happiness fluctuations pretty perfectly tracked changes in the economy. That shouldn't happen if income doesn't buy happiness.
Recent research has sided with Stevenson and Wolfers against Easterlin. The Nobel laureate psychologist/economist Daniel Kahneman and Princeton economist Angus Deaton found that "emotional well-being" (or happiness) maxes out around $75,000 of income in the U.S. Any more than that, and the benefits are negligible. But they found that life satisfaction keeps rising. "We conclude," Kahneman and Deaton write, "that high income buys life satisfaction but not happiness."
Now Stevenson and Wolfers are back to show that their thesis has held up quite well during the recession. GDP per capita is still strongly associated with countries' overall life satisfaction levels.
And within countries, people with more money are more satisfied. These curves show the relationship between income and reported life satisfaction within a number of countries. In each country, the relationship is logarithmic but real.
Most intriguingly, they found that the General Social Survey's happiness question, which asks, "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?" supports the theory that happiness rises constantly with income.
This is somewhat in tension with Deaton and Kahneman's finding, though Stevenson and Wolfers note in their conclusion that Deaton and Kaheman's conclusions are "based on very different measures of well-being, and so they are not necessarily in tension with our results."
What does this imply for policy? Well, it suggests that boosting GDP per capita is very important, for one thing. But the fact that the relationship is logarithmic, not linear, also provides solid evidence for the theory, common among fans of redistributive policies, that poor people get more out of the next dollar they receive than the rich do.
Mark Schmitt, now a senior fellow at the Roosevelt Institute, once made a memorable case for progressive taxation along these lines. "Progressive taxation is based on ability to pay, which increases logarithmically," he wrote. "It's patently obvious that the hardship of paying 15 percent for someone earning $20,000 is far greater than the burden of the same percentage for someone earning $200,000."
It may have been patently obviously, but Wolfers and Stevenson show it's empirically supportable too.
* The data here is the result of what Easterlin calls "Cantril-style" happiness questions, which produce "a rating by each individual of his personal standing on a scale from 0 (the worst possible life) to 10 (the best possible life), where 'worst' and 'best' are defined by each person for himself." These are basically life satisfaction measures, as we'll get into above. Read the paper for a more detailed explanation.