Having It Both Ways

ANDREW WELLS-DANG is Washington Representative of the Fund for Reconciliation and Development.

David Dollar and Aart Kraay's claim that "globalization ... has actually promoted economic equality and reduced poverty" is based on a selective use of data and dubious assumptions about causality. Indeed, their evidence remains far short of convincing.

To defend their views, Dollar and Kraay play fast and loose with the economics of inequality. The "mean log deviation" measure they use to claim a global reduction in inequality since 1975 is not a particularly good indicator. All it measures is the relative difference in distribution among the rich compared to the poor. To take an extreme example, a society with half of its members earning $50,000 and half earning $500 would have a mean log deviation of zero. Surely this is not what Dollar and Kraay mean by perfect equality.

The authors then proceed to confuse the issue of inequality with poverty reduction. It is indisputable that absolute poverty has declined dramatically in countries such as China and Vietnam following market reforms, and that millions of people are better off as a result. Nevertheless, relative inequality has risen just as dramatically, creating a host of social problems.

Dollar and Kraay admit from the outset that China and India have a huge effect on global aggregate statistics. In fact, the skew comes primarily from China, as India resembles other developing economies much more than China does. For this reason, other economists, such as the authors of the un Human Development Report, routinely exclude China from aggregate data covering developing nations. Dollar and Kraay, however, include China in their statistics when it suits their purposes—for instance, to show rising global economic growth in the 1990s. Since China had the highest growth rates in the world during the past decade, emerged unscathed by the Asian financial crisis, and contains nearly 25 percent of the world's population, including it obscures what was actually a quite average decade for most developing countries.

When addressing inequality within nations, however, Dollar and Kraay treat China as an exception, noting that "some countries" experience higher inequality with growth and other countries do not. If they were to aggregate this data as they do in other sections of their article, they would likely find that rapid growth tends to increase inequality worldwide. Conversely, if they were to disaggregate links between globalization and growth, they would find that some countries with liberal trade and investment policies grow rapidly, whereas others do not. If China is excluded, the evidence of a correlation becomes substantially weaker.

Finally, as Dollar and Kraay admit, "one needs to be careful about drawing conclusions about causality." Unfortunately, they are not particularly careful. All positive changes, such as economic growth and reduced inequality, are said to be "promoted" and "supported by" globalization, while poverty and increased inequality are the fault of "domestic education, taxes, and social policies." One cannot have it both ways. Trade and investment policies are determined by governments just as social policies are, and governments deserve credit for their economic successes as much as they deserve criticism for their failings. Increased international integration might be as much a result of growth as its cause. If Dollar and Kraay were to take a more comprehensive look at their data, they would see a more complicated picture, in which global trade and investment produce some positive and some negative effects, with neither caused solely by reduced poverty or increased inequality.