Enlightenment Now: The Case for Reason, Science, Humanism, and Progress

Among the countries that punch below their wealth in happiness is the United States. Americans are by no means unhappy: almost 90 percent rate themselves as at least “pretty happy,” almost a third rate themselves as “very happy,” and when they are asked to place themselves on the ten-rung ladder from the worst to the best possible life, they choose the seventh rung.²⁹ But in 2015 the United States came in at thirteenth place among the world’s nations (trailing eight countries in Western Europe, three in the Commonwealth, and Israel), even though it had a higher average income than all of them but Norway and Switzerland.³⁰ (The United Kingdom, whose citizens place themselves at a happy 6.7 rungs up from the worst possible life, came in at twenty-third place.)

Also, the United States hasn’t gotten systematically happier over the years (another decoy that led to the premature announcement of the Easterlin paradox, because the United States is also the country with happiness data that stretch back the farthest). American happiness has fluctuated within a narrow band since 1947, deflecting in response to various recessions, recoveries, malaises, and bubbles, but with no consistent rise or fall. One dataset shows a slight decline in American happiness from 1955 to 1980, followed by a rise through 2006; another shows a slight decline in the proportion saying they are “very happy” starting in 1972 (though even there the sum of those who say they are “very happy” and “pretty happy” has not changed).³¹

The American happiness stagnation doesn’t falsify the global trend in which happiness increases with wealth, because when we look at changes in a rich country over a few decades we’re peeping at a restricted range of the scale. As Deaton points out, a trend that is obvious when you look at the effects of a fiftyfold difference in income between, say, Togo and the United States, representing a quarter-millennium of economic growth, may be submerged in the noise when you look for the effects of, say, a twofold difference in income within a single country over just twenty years of economic growth.³² Also, the United States has seen a greater rise in income inequality than the countries of Western Europe (chapter 9), and its growth in GDP may have been enjoyed by a smaller proportion of the populace.³³ Speculating about American exceptionalism is an endlessly fascinating pastime, but whatever the reason, happyologists agree that the United States is an outlier from the global trend in subjective well-being.³⁴

Another reason it can be hard to make sense of happiness trends for individual countries is that a country is a collection of tens of millions of human beings who just happen to occupy a patch of land. It’s remarkable that we can find anything in common when we average them, and we shouldn’t be surprised to find that as time passes, different segments of the population go in different directions, sometimes jerking the average around, sometimes canceling each other out. Over the past thirty-five years African Americans have been getting much happier while American whites have gotten a bit less happy.³⁵ Women tend to be happier than men, but in Western countries the gap has shrunk, with men getting happier at a faster rate than women. In the United States it has reversed outright, as women got unhappier while men stayed more or less the same.³⁶

The biggest complication in making sense of historical trends, though, is one that we came across in chapter 15: the distinction between changes across the life cycle (age), in the zeitgeist (period), and over the generations (cohort).³⁷ Without a time machine, it’s logically impossible to disentangle the effects of age, cohort, and period completely, to say nothing of their interactions. If, for example, fifty-year-olds were miserable in 2005, we couldn’t tell whether the Baby Boomers had a hard time dealing with middle age, the Baby Boomers had a hard time dealing with the new millennium, or the new millennium was a hard time to be middle-aged. But with a dataset that embraces multiple generations and decades, together with a few assumptions about how quickly people and times can change, one can average the scores for a generation over the years, for the entire population at each year, and for the population at each age, and make reasonably independent estimates of the trajectory of the three factors over time. That in turn allows us to look for two different versions of progress: people of all ages can become better off in recent periods, or younger cohorts can be better off than older ones, lifting the population as they replace them.