December 19, 2005
By William Cline
Three years ago the World Bank said that freeing international trade of all barriers and subsidies would lift 320 million people above the $2 a day poverty line by 2015. But new World Bank projections emphasizing $1 a day poverty and based on new data and methods put the number at just 32 million people. CGD/IIE Senior Fellow William R. Cline, author of Trade Policy and Global Poverty, has been examining the Bank's new calculations and argues that the first estimate was closer to the truth.Q: Why should these estimates matter?A: The estimates are important both because of their inherent significance for global policy and more immediately because they could give an easy argument to the opponents of globalization by implying that even the World Bank doesn’t think global free trade would reduce poverty by much. This is a windfall for the anti-globalists.Q: You have estimated that full trade liberalization would lift roughly 440 million people above the $2 a day poverty line. How can economists using similar tools come to such different conclusions?A: The World Bank's first estimate was not that different: 320 million people. Its new estimate is 95 million people for the same concept ($2 per day poverty, total static and dynamic effects). This is because of lower estimates of protection in a new dataset, and because of application of lower measures than before for the percent change in poverty given a 1 percent change in income. I am still examining both of these changes, but at least initially they seem exaggerated to me.Q: How so?A: The new dataset says agricultural protection is much lower than in the previous dataset, but the methodological changes responsible may be questionable. One example is a shift to ignore high “above quota” tariffs entirely unless the quota is 90% filled, instead of averaging the low within-quota tariff and the high above-quota tariff. In addition, there is a seeming change in the World Bank’s decisions on which concepts to emphasize. In some prominent publications of its new estimates, the World Bank authors do not even mention the 95 million figure and instead emphasize 32 million people lifted out of poverty, because they choose to focus on a $1 a-day definition of poverty and limit their attention only to static effects rather than including dynamic effects.Q: Your book was published in 2004. Has anything happened since then that would cause you to revise your estimates?A: Not much has happened that would warrant a change in the estimates. It is really a question of whether the earlier datasets and methodologies were right. Although some adjustments may be appropriate, my sense is that the new World Bank estimates have gone too far. It should be kept in mind, moreover, that even the earlier estimates did not include gains to developing countries from liberalization of trade in services.*Read the World Bank response to this critique.
Three years ago the World Bank said that freeing international trade of all barriers and subsidies would lift 320 million people above the $2 a day poverty line by 2015. But new World Bank projections emphasizing $1 a day poverty and based on new data and methods put the number at just 32 million people. CGD/IIE Senior Fellow William R. Cline, author of Trade Policy and Global Poverty, has been examining the Bank's new calculations and argues that the first estimate was closer to the truth.Q: Why should these estimates matter?A: The estimates are important both because of their inherent significance for global policy and more immediately because they could give an easy argument to the opponents of globalization by implying that even the World Bank doesn’t think global free trade would reduce poverty by much. This is a windfall for the anti-globalists.Q: You have estimated that full trade liberalization would lift roughly 440 million people above the $2 a day poverty line. How can economists using similar tools come to such different conclusions?A: The World Bank's first estimate was not that different: 320 million people. Its new estimate is 95 million people for the same concept ($2 per day poverty, total static and dynamic effects). This is because of lower estimates of protection in a new dataset, and because of application of lower measures than before for the percent change in poverty given a 1 percent change in income. I am still examining both of these changes, but at least initially they seem exaggerated to me.Q: How so?A: The new dataset says agricultural protection is much lower than in the previous dataset, but the methodological changes responsible may be questionable. One example is a shift to ignore high “above quota” tariffs entirely unless the quota is 90% filled, instead of averaging the low within-quota tariff and the high above-quota tariff. In addition, there is a seeming change in the World Bank’s decisions on which concepts to emphasize. In some prominent publications of its new estimates, the World Bank authors do not even mention the 95 million figure and instead emphasize 32 million people lifted out of poverty, because they choose to focus on a $1 a-day definition of poverty and limit their attention only to static effects rather than including dynamic effects.Q: Your book was published in 2004. Has anything happened since then that would cause you to revise your estimates?A: Not much has happened that would warrant a change in the estimates. It is really a question of whether the earlier datasets and methodologies were right. Although some adjustments may be appropriate, my sense is that the new World Bank estimates have gone too far. It should be kept in mind, moreover, that even the earlier estimates did not include gains to developing countries from liberalization of trade in services.*Read the World Bank response to this critique.Disclaimer
CGD blog posts reflect the views of the authors, drawing on prior research and experience in their areas of expertise. CGD is a nonpartisan, independent organization and does not take institutional positions.
