Parametric Estimations of the World Distribution of Income

Maxim Pinkovskiy, Xavier Sala-i-Martin

NBER Working Paper No. 15433
Issued in October 2009
NBER Program(s):   EFG   PR

We use a parametric method to estimate the income distribution for 191 countries between 1970 and 2006. We estimate the World Distribution of Income and estimate poverty rates, poverty counts and various measures of income inequality and welfare. Using the official $1/day line, we estimate that world poverty rates have fallen by 80% from 0.268 in 1970 to 0.054 in 2006. The corresponding total number of poor has fallen from 403 million in 1970 to 152 million in 2006. Our estimates of the global poverty count in 2006 are much smaller than found by other researchers. We also find similar reductions in poverty if we use other poverty lines. We find that various measures of global inequality have declined substantially and measures of global welfare increased by somewhere between 128% and 145%. We analyze poverty in various regions. Finally, we show that our results are robust to a battery of sensitivity tests involving functional forms, data sources for the largest countries, methods of interpolating and extrapolating missing data, and dealing with survey misreporting.

(852 K)

This paper is available as PDF (852 K) or via email.

Acknowledgments

Machine-readable bibliographic record - MARC, RIS, BibTeX