Difference of opinion of population living on less than $"per day Chen Ravallion 25 20 Sala-i-M: 10 1⊥上⊥0 075808590 Population living on less than $1*per day, bn Chen Ravallion 1.2 Sala-i-Martin 197075 80 85 9598 Source: Authors But this is not at all the picture that emerges from the second, and far more widely cited, set of estimates. Calculations by the World Bank, using direct surveys of households, carry the official imprimatur of the UN, which uses them in monitoring progress towards its millennium Development Goal on poverty. and they seem to show relatively little reduction in poverty over recent decades a paper by shaohua Chen and Martin Ravallion of the bank lays out the thinking behind the ank's estimates. The authors put the proportion of people living on less than a dollar a day at 28% in 1987-far higher than the corresponding figure according to Mr sala-i-Martin's work. By 1998, the proportion in poverty had in fact fallen(something which you might not guess if you listened only to those who deplore the wickedness of global capitalism), but only to 24% Compare that with Mr sala-i-Martin' s estimate of just 7% That discrepancy draws attention to the danger of focusing too much on the dollar-a-day threshold That is a crowded part of the global income distribution For this reason alone, switching from one data source to another or moving the official poverty line from one level to another, is apt to have a large effect on the figures. this underlines the importance of not regarding any of these numbers as holy writ Still, the question remains, why are the differences so big? Several factors are at work. The World Bank attempts to measure consumption poverty as opposed to income poverty".To the extent that poor people manage to save, their consumption will be less than their income, and so there will be more poor people on the bank's definition. Also, the Bank expresses its poverty ratios as proportions of population in the developing countries Mr Sala-i-Martin, for instance, uses global population the effect is to make Mr Sala-i-Martin ' s estimates, other things equal, smaller than the Bank's. Country samples also vary from study to study. And onBut this is not at all the picture that emerges from the second, and far more widely cited, set of estimates. Calculations by the World Bank, using direct surveys of households, carry the official imprimatur of the UN, which uses them in monitoring progress towards its Millennium Development Goal on poverty. And they seem to show relatively little reduction in poverty over recent decades. A paper by Shaohua Chen and Martin Ravallion of the Bank lays out the thinking behind the Bank's estimates. The authors put the proportion of people living on less than a dollar a day at 28% in 1987—far higher than the corresponding figure according to Mr Sala-i-Martin's work. By 1998, the proportion in poverty had in fact fallen (something which you might not guess if you listened only to those who deplore the wickedness of global capitalism), but only to 24%. Compare that with Mr Sala-i-Martin's estimate of just 7%. That discrepancy draws attention to the danger of focusing too much on the dollar-a-day threshold. That is a crowded part of the global income distribution. For this reason alone, switching from one data source to another, or moving the official poverty line from one level to another, is apt to have a large effect on the figures. This underlines the importance of not regarding any of these numbers as holy writ. Still, the question remains, why are the differences so big? Several factors are at work. The World Bank attempts to measure “consumption poverty”, as opposed to “income poverty”. To the extent that poor people manage to save, their consumption will be less than their income, and so there will be more poor people on the Bank's definition. Also, the Bank expresses its poverty ratios as proportions of population in the developing countries; Mr Sala-i-Martin, for instance, uses global population; the effect is to make Mr Sala-i-Martin's estimates, other things equal, smaller than the Bank's. Country samples also vary from study to study. And on