Estimating the World Distribution of Income

National Bureau Of Economic Research Inc, Cambridge MA

Investigators

Abstract

One of goals of economic development is to reduce poverty and income inequality. To assess whether progress, we need good empirical estimates of the number of poor. Unfortunately, all the estimates we have of world poverty come from a single institution: the World Bank. The problem is that the World Bank methodology is obscure and often controversial. Such obscure procedures and data make it very hard for independent researchers to check the accuracy of the World Bank numbers. This research proposes a new, simple and clear methodology to estimate the number of poor citizens in the world by estimating the world distributions of income and consumption and calculating the fraction of the distribution that lies below a certain poverty line such as the $1/day line. We propose to build on the PI's previous work by combining micro survey data and aggregate GDP per capita data to estimate the income shares for five quintiles for every country for every year between 1970 and 2000. The individual distributions are then integrated to estimate the world distribution of individual income using kernel density functions. The PI has estimated world income distributions from 1970 to 1998. The proposed research will: expand the data to 2000, include countries from Eastern Europe and the former Soviet Union, assign regional income shares to the countries for which there are no surveys, extend the income analysis to consumption, and check the robustness of the approach by extending the non-parametric kernel approach to parametric estimates of individual country distributions using different distributions before integrating them to estimate the world distribution. The results of the income distribution data will be used to estimate yearly poverty rates for each country, each region, and the world as a whole. The poverty rates can be used, for example, to see what policies and/or institutions favor the increase in income of the poorest citizens of the economy. The proposed research will have broader impact in three areas---policy, research, and general public. The major outcome of this research will be a new methodology to estimate the world distribution of individual income and consumption. The data will make easier to conduct research into the determinants of poverty. The data will also make it possible to craft as well as evaluate poverty-reducing programs. This could have enormous consequences for human welfare all over the world.