The Undercover Economist

What Is “Income per Natural”?

How a new way to count national income could change how we think about immigration and development.

Which nation produces the richest people in the world? You might think that is an easy question to answer: Just grab the latest figures from the International Monetary Fund, and you’ll see that the answer is Luxembourg ($102,000 gross domestic product per head in 2007). The United States is in ninth place ($46,000) and the U.K. in 11th ($45,000).

There are some methodological wrinkles to iron out: what exchange rate to use, for instance. And for the poorest countries, such as Liberia ($200 dollars per person in 2007) or Burundi ($130), the numbers involve some guesswork. But overall, these are not controversial statistics—unless you are Lant Pritchett or Michael Clemens.

Pritchett, of Harvard’s Kennedy School, and Clemens, of the Center for Global Development, a Washington, D.C., think tank, argue that my opening question should be answered in a radically different way. Rather than measuring the income of people who are now residents of Liberia, Clemens and Pritchett have produced a research paper estimating the income earned by people who were born in (say) Liberia, regardless of where they now live—what Clemens and Pritchett call “income per natural” of Liberians.

For Luxembourg—or any other rich country—there is a trivial difference between income per natural and more conventional measures of national income. But for Liberia, the difference is anything but trivial: The Liberian-born make 50 percent more than Liberian residents. Nor is Liberia unique: Clemens and Pritchett estimate that the income of the Samoan-born is nearly twice the income of the Samoan resident, and the Guyanese-born are more than twice as well-off as residents of Guyana.

These dramatic differences have a simple explanation: Many poor people became richer by leaving their country of birth. Clemens and Pritchett estimate that “two of every five living Mexicans who have escaped poverty did so by leaving Mexico; for Haitians it is four out of five.”

There is a point to this exercise: Clemens and Pritchett want to call attention to the fact that migration has made a lot of migrants richer. Traditional measures of income tend to mask this fact. In rich countries, we usually ask whether migrants improve the lot of existing residents, not whether migration improves the lot of migrants. Meanwhile, the welfare of migrants rarely figures in debate in developing countries or in development institutions such as the World Bank, because the migrants have gone.

Simply because of the way the discussion is framed, the benefits to migrants tend to be ignored. Imagine a man who moves from earning 10,000 euros in Poland (an above-average wage) to 15,000 pounds in the U.K. (a below-average wage). Simple arithmetic says that he has reduced the average income of both countries; that could be true even if he has impoverished nobody and enriched himself a great deal.

The “income per natural” statistic is the latest in a long line of alternatives to gross domestic product, the standard measure of an economy’s size. Others—variously championed by Nobel laureates such as Amartya Sen, Daniel Kahneman, Joseph Stiglitz, and the late James Tobin—try to adjust GDP to account for the depletion of natural resources, or to incorporate measures of health and education, or even (in Kahneman’s case) to start from scratch with time-weighted accounts of happiness.

I sometimes wonder if these alternative measures make a difference to the way policy is conducted. After all, no government ever tried to maximize GDP, anyway, so why try so hard to measure something else? But Pritchett is convinced that the way the discussion is framed does make a difference; once we formally recognize that people can escape poverty by moving as well as by sitting still, we may start to regard migration as a legitimate source of economic development.

That, at least, is what Pritchett hopes. His next paper, with Michael Clemens, will argue that the pay gaps caused by racial and sex discrimination, while very real, are miniscule compared with the pay gaps caused by national boundaries. That should raise a few eyebrows. Do not expect this subject to go away while Pritchett is on the case.