The Measure of Human Happiness

A theme at many sessions at this year's AIF has been happiness--what it is, how you advance it, how you measure it. Fascinating. Justin Wolfers and Robert Frank had an interesting exchange on this earlier in the week, and I'm continuing to turn their arguments over in my mind.

Wolfers tore into the "Easterlin Paradox", which is the claim that happiness does not rise with income beyond a certain point. That finding (see Richard Easterlin: Does Economic Growth Improve the Human Lot?) gave rise to the popular view that, for rich countries at least, economic growth is a treadmill. People are struggling to improve their status, and feel happier if they succeed, but the race for status goods is zero-sum. Growth in absolute income cannot raise everybody's relative position. It allows higher consumption but expands desires at about the same rate. The gain in happiness, if any, is small. For a rich country, the obsession with growth in GDP is an error.

Wolfers walked through his impressive array of data--derived from a remarkable project of international comparisons undertaken by Gallup--and argued that the Easterlin claim is simply and unambiguously false. Higher incomes make people happier. It takes ever larger increases in absolute income to yield a given improvement in happiness (happiness rises with the log of income), but there is no point of saturation. Within countries, richer people are happier than poor people. Globally, rich countries are happier than poor countries. He examined some of the statistical evidence which is said to point the other way, and showed it was wrong. Economic growth does what it is supposed to.

He also made a particularly telling observation, I thought, against the unqualified Easterlin/Frank view that growth is nothing but an arms race for positional goods. On the arms race theory, it is better to have a middling income in a middle-income country than to live in a rich country with a higher income--if that higher income leaves you, relatively speaking, dirt poor. We have a "beautiful experiment", as Wolfers put it, to test this idea. On the arms race theory, Americans should be sneaking across the border into Mexico rather than the reverse. So much for the arms race theory.

Frank rightly argued--and Wolfers agreed--that status matters too. Frank is also right, I think, to point to the negative externalities that flow from conspicuous consumption. But in the discussion about which matters more, income or status, Wolfers easily got the best of the debate.

Maybe too much so. Frank let him get away with a couple of things.

Frank had argued, for instance, that McMansions show what Fred Hirsch called the social limits to growth: in the struggle for status, people build houses so pointlessly big that they become a nuisance to their owners. Wolfers' riposte was to ask, in that case, why a tiny house in Aspen--a marker of low status relative to the neighbours--was so ridiculously expensive. Wouldn't it be better to lord it over the Joneses down valley in Basalt or Carbondale? Status, evidently, is not that important. Frank offered no rebuttal, but the answer he could have made is so obvious. Tiny houses in Aspen are ridiculously expensive precisely because they signify high status. ("I live in Aspen. Just a small place, you know, but it's Aspen. Where do you live? Ah. Carbondale. Nice.")

Wolfers also asked why, if status matters so much, parents strive to send their children to good schools--where a child of given ability will rank lower, because the other pupils will tend to be brighter. Again, no rebuttal. But of course it is quite likely better for the child, on a lifetime status basis, to be a middling performer in an excellent school than a star in a bad one. (I went to Oxford. I should know.) That is a judgment a parent might very well make.

I came away thinking the Easterlin Paradox was a smoking wreck, and that pursuit of economic growth remains a worthy objective even for rich countries. But if I were Wolfers, I might stick to the Mexico experiment when illustrating the argument.