"Money doesn't buy happiness" is a cliche for a reason. The Nobel laureate psychologist/economist Daniel Kahneman and Princeton economist Angus Deaton have found that "emotional well-being" (that is, what emotions people report themselves as having felt the previous day) maxes out at around $75,000 of annual income, even though peoples' evaluations of how well their lives are going rise indefinitely with income. "We conclude," Kahneman and Deaton write, "that high income buys life satisfaction but not happiness."

But this kind of analysis can be problematic when it's used to compare whole societies, rather than individual people. In his Project Syndicate column on Friday, the economic historian Robert Skidelsky cited the work of Robert Easterlin in comparing happiness in rich and poor countries. "Easterlin found no correlation between happiness and (gross national product) per head," Skidelsky writes. "In other words, GNP is a poor measure of life satisfaction." Skidelsky is being loose by conflating "happiness" (which researchers generally use to refer to "emotional well-being") and "life satisfaction."

The only problem: Easterlin apparently was wrong. Justin Wolfers and Betsy Stevenson — then at the University of Pennsylvania, now at University of Michigan — took a stab at the question in 2008, and found that, contrary to Easterlin's findings, overall well-being always rises with income. They looked at data from numerous polls that ask about peoples' emotional well-being and life satisfaction — including the First European Quality of Life Survey, the Eurobarometer survey, the Gallup World Poll, the World Values Survey and the Pew Global Attitudes survey — and find that there's a strong, statistically significant relationship between real GDP per capita and overall well-being. Here are the Gallup results for life satisfaction:

Source: Stevenson and Wolfers

Same deal if you look at happiness:

Source: Stevenson and Wolfers

The GDP values here are a "log scale." That means that the relationship isn't linear: a $1 increase in GDP per capita doesn't always buy the same amount of increased life satisfaction. It buys less as the country grows richer. But even in the richest of countries, a rise in GDP still improves perceptions of happiness and life satisfaction.

This also holds over time. Here's how happiness in the United States, as measured by the General Social Survey, compares with the "output gap" — that is, the gap between where the economy stands now and where it would be if it were firing at all cylinders:

Source: Stevenson and Wolfers

When the economy was doing well, people were happy. When it was doing poorly, people were sad. It makes intuitive sense.

The fact that the relationship isn't linear suggests that economic growth is less important to national happiness for rich countries than it is for poor countries. But that doesn't mean it's unimportant in rich countries, and as Kahneman and Deaton have found, for the majority of Americans making under $75,000, happiness still depends strongly on income. Pursuing happiness and pursuing growth aren't alternatives. They go hand in hand.