unlike in other regions. Despite the increase in coverage, there are continued concerns about lags in data availability and public access, comparability over time and across countries, and underreporting and selective compliance.

Ravallion explained that market exchange rates generally are not useful in assessing real incomes in developing countries, in that they tend to equate purchasing power with traded goods. Therefore the International Comparison Project has developed more realistic estimates of purchasing power parity exchange rates. The latest ICP data were released in 2005 with a new one scheduled for release in 2011. Despite improvements in the ICP process since the 1970s, serious issues remain in that they tend to be urban biased under representing rural areas, which is a particular concern in China, where only 11 cities were surveyed.

Ravallion also discussed the arguments in favor of relative poverty lines. He noted that many Organization for Economic Cooperation and Development countries have relative poverty lines, that is, a fixed proportion of the country’s mean or median income. He suggested that one way to allow for relative poverty is to have a poverty line that is constant at very low incomes?representing absolute poverty—and then increasing at somewhat higher incomes.

Ravallion concluded by noting that in most instances there is no need to form a single composite poverty index that includes data on nutrition, child mortality, schooling, and violence, as these indicators are too disparate to combine into a single measure. The ultimate goal should be to create a set of multiple indexes that looks at these non-consumption factors and can usefully guide policy makers, rather than develop a single multidimensional index.


James E. Foster, The George Washington University

James Foster described the recently released Oxford Multidimensional Poverty Index (MPI). He explained that this type of index was needed because conventional measures of poverty capture only an income or consumption dimension of poverty, when in fact there are many other aspects of poverty. He suggested that such a measure must be understandable and easy to describe as well as conform to common-sense notions of poverty. It must also be technically solid and data must be available. While such indexes have been talked about for some time, one impetus for the new index came from Mexico, where a government law mandates that poverty be measured multidimensionally.

The new index provides a dual cutoff approach to measuring poverty. Within each dimension there is a deprivation cutoff, and then across the dimensions there is a poverty cutoff. That is, if someone is deprived in enough dimensions or in enough breadth, they are considered poor. Foster described the approach as being intuitive, transparent, and flexible. He emphasized its use in country applications, where one can target and evaluate policies. He also said that it was participatory, in that country stakeholders could determine cutoff and weights rather than having a one-size-fits-all index. The three specific dimensions included in the MPI are education, health, and standard of living, and there are 10 associated indicators (Box I 2-1).


2 The presentation is available at http://sites.nationalacademies.org/PGA/sustainability/foodsecurity/PGA_060826, presentation by James Foster (February 16, 2011).

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement