In Have We Already Met the Millennium Development Goal for Poverty? the World Bank’s main poverty analyst Martin Ravallion (hereafter Ravallion) has raised several issues pertaining to my recent book Imagine There’s No country: Poverty, Inequality, and Growth in the Era of Globalization (hereafter Imagine) published by the Institute for International Economics, Washington, DC. In addition, the World Bank’s recent publication, Global Economic Prospects (World Bank GEP 2003), raises some of the same questions as Ravallion.
In Imagine, I had offered an alternative to the World Bank method of determining the magnitude of poverty and its change over the 50-year period, 1950–2000. The years 1980 to 2000 were singled out for special analysis as the globalization that has taken place over this period has been the subject of much controversy. Among the conclusions reached were that world inequality and poverty had declined during the globalization period; indeed, world poverty had declined by a magnitude greater than any other 20- year period in history. The Imagine results also indicated that one of the principal Millennium Development Goals (MDG)—that of halving the incidence of world poverty from 30 percent in 1990 to 15 percent by 2015—had already been reached in 2000. These conclusions are questioned by the World Bank (GEP and Ravallion) on the grounds of bad data or misinformed analysis, a critique not unlike that offered by Milanovic (2002) against Sala-i-Martin’s (2002a, b) study on world inequality—a study which, incidentally, reaches conclusions very similar to Imagine.
The focus of attention in this reply are the poverty numbers produced by the World Bank and Imagine. While Imagine analyzes the level and trend in poverty over each of the last 50 years, the World Bank has presented estimates for only selected years between 1987 and 1998. The major differences with the World Bank results lie in the estimates for the 1990s; in particular, for the period 1987 to 1998. The World Bank estimates suggest that poverty (between 1987 and 1998, $1.08 poverty line, 1993 PPP prices) in the nonindustrialized world declined by only 4.8 percentage points—from 28.3 to 23.5 percent. In Imagine, from a higher initial poverty level of 29.3 percent in 1987, the poverty decline was reported to be 13.2 percentage points—more than two and a half 2 times higher than the World Bank, and a poverty level equal to 16.1 percent in 1998, and 13.1 percent in 2000.
Thus, according to the World Bank, about a quarter (23.5 percent) of all residents in the nonindustrialized world—or about 1.2 billion people—lived in poverty in 1998. In sharp contrast, according to Imagine, just 800 million people were poor—and this number had been further reduced to 670 million in 2000. These differences have important policy implications, particularly with regard to the role of economic growth in poverty reduction. The question for all of us is which estimate is closer to the truth; the answer to this question should determine the nature of the policy response of governments and international organizations.
The two “competing’ methods yield radically different results for essentially the same phenomenon—that is, by what percentage has poverty declined and what is the level of poverty today. This paper evaluates the methods in detail and attempts to put the similarities and differences in perspective. It is important to stress that the debate is not about whether there is any or little poverty in the world but its magnitude according to the conventionally accepted definition of a-dollar-a-day poverty line (in 1985 prices). The debate is not about whether there was any poverty reduction during the globalization era or even whether growth is important for poverty reduction. Rather, the debate is about whether the “observed” reduction in poverty of only 5 percentage points was commensurate with the growth in expenditures and incomes that occurred in the developing world. If the poverty level today is too high (relative to expectations) or the reduction in the past too little (e.g., only 5 percentage points), then the demand by international organizations for more aid and search for a “new development order” may be justified. If, on the other hand, the past two decades of high growth have witnessed a rapid reduction of world poverty (13 percentage points between 1987 and 1998 and 31 percentage points between 1980 and 2000), then the correct policy response is to understand what brought about growth and to apply this knowledge to regions with high poverty, such as sub-Saharan Africa.