A comment on the editorial (25 October) that critiqued the World Bank methodology for estimating the number of poor in the world. This response aims to refute the four major arguments made in the editorial.
DISCUSSIONnovember 8, 2008 EPW Economic & Political Weekly78‘How Many Poor in the World?’: A Reply Martin RavallionMartin Ravallion (Mravallion@worldbank.org) is with the Development Research Group of the World Bank, Washington DC.A comment on the editorial (25 October) that critiqued the World Bank methodology for estimating the number of poor in the world. This response aims to refute the four major arguments made in the editorial.The editorial “How Many Poor in the World?” (25 October) claims that the World Bank’s latest estimates of the extent of global poverty “present a cri-sis of credibility”.1 The editorial gives four reasons for this rather negative assess-ment. Let me address them in turn. First, it argues that our international poverty line is “little more than a destitution line”. To recap, our proposed standard for measuring global poverty is $1.25 per person per day at 2005 purchasing power parity (PPP), using the PPP exchange rates for household consumption from the 2005 round of the International Comparison Pro-gram(ICP), which is a massive price data collection project spanning 150 countries. The $1.25 line is obtained as the average of the national poverty lines in the poorest 15 countries in a compilation of 75 national poverty lines for developing countries (though the line is similar for the poorest 10 or poorest 20 countries, given that rela-tive poverty effects only start to emerge amongst middle-income countries). We find that 25% of the developing world’s population lived below this line in 2005.There is no doubt that $1.25 a day is a frugal line; most middle income countries have higher lines. The average of all ex-cept the poorest 15 countries is $2.50 a day, which is not reached by more than half the developing world’s population. Towards the other extreme, the average official poverty line in the US is $13 per day in 2005; 96% of those living in the develop-ing world are poor byUS standards. However, if by a “destitution line” one means that a life-threateningly low level of living is implied, then not one of the 15 national poverty lines that go into form-ing the $1.25 line can be considered a destitution line. Each of the national lines meets “industry standards” in its construc-tion, in that it uses a food bundle that attains standard nutritional requirements by prevailing diets and it includes an allowance for non-food spending. (The linesvary, some being more generous thanothers in both the food and non-food allowances; so we take an average.)Also note that India’s official poverty line is about $1.00 a day using the same PPPs; (see my article in the same issue of EPW). So if $1.25 marks “destitution” what are we to make of India’s national poverty line? In truth, neither the $1.25 line nor India’s (lower) official line represent “destitution”. Second, the editorial points out that the standardPPP calculations reflect a broader set of goods than typically consumed by the poor. This is correct, and we have long recognised this point (going back to our first “dollar a day” poverty calculations for the 1990World Development Report). How-ever, the editorial fails to point out (as I also noted in my EPW article) that we have implemented a new set of “PPPs for the poor” using the 2005 ICP data. These re-weight the prices to accord with the consumption patterns of people near the poverty line. We have published results in testing robustness of our new poverty line to this new PPP and alternatives (also referred to in my article). Third, the editorial claims that our method is inconsistent for inter-temporal comparisons. Our method – let us call it method A – entails converting the inter-national poverty line to local currencies using the PPP at the benchmark year for which the new ICP price surveys were done (now 2005) and then adjusting over time using the country’s consumer price index (CPI) to get to the prices prevailing in the relevant survey year. The editorial claims that it would only be a “fluke” if our method gave the same result if instead we had used the obvious alternative, method B, of converting all country-distributions to $ and then made the poverty calcula-tions adjusting for inflation in the US. However, it is readily verified that if the calculation is done correctly, then it would be no fluke at all; methods A and B will always give exactly the same answer given the same data. The mistake made in the editorial (and by one or two other critics) is to not realise that if one uses method B then thePPP for the non-benchmark year
DISCUSSIONEconomic & Political Weekly EPW november 8, 200879must be calculated consistently with the differential rates of inflation between the US and the country in question. But that is exactly how PPPs are routinely updated for non-benchmark years. Granted, there may well be an incon-sistency if one compares thePPPs obtained from different benchmark years, for the consistency of the PPPs with the differen-tial rates of inflation from the nationals CPIs is not assured (given differences in the quality of goods used and the weights). This is well known. We follow standard practice in other international compari-sons of only doing the PPP conversion for the benchmark year, and then making inter-temporal comparisons using the national data; internal consistency between methods A and B is then automatic. Thus our inter-temporal comparisons of poverty in India follow essentially the same methods used by the Planning Commission; the main difference is that we use a 25% higher poverty line, as I have already noted. Note also that the consistency of the PPPs for different benchmark years with the CPIs is a moot point given that the changes inICP data and methods between benchmark years clearly swamp these other differences. For example, the differ-ences between the 1993 and 2005 ICP surveys – entailing a very substantial improvement in the quality of the price surveys and methods of imputation – make any attempt to compare PPPs across different benchmark years highly dubious. Finally, the editorial argues that the lower the poverty line “the more flattering is the resulting trend decline in poverty” and that this “suggests that the trends...are not so much a reflection of what is actually happening to poverty on the ground as a verification of the consistency of arithmetic”. It is true that in the limit, as the poverty line goes to a very high level, the rate of poverty reduction goes to zero. That is also true in the limit at very low lines; when the poverty rate is near zero it cannot fall much. But we are not near either limit. The fact that we find that the rate of progress against poverty is slightly higher for $1 per day than $2 per day hardly indicates that our results are arithmetic artefacts. We readily acknowledge that the data are not ideal; in this respect, measuring poverty is no different to any other avenue of social or economic research. But we have to make the most of the data availa-ble. Thankfully, there have been great im-provements in the data – from both the household survey data and the price level data from the ICP – since we started our global poverty monitoring effort 20 years ago (and it would be fair to say that the World Bank has played a key role in these improvements). The latest revisions reflect that progress. Note 1 The estimates are reported in Shaohua Chen and Martin Ravallion, “The Developing World Is Poorer Than We Thought, But No Less Successful in the Fight against Poverty”, Policy Research Working Paper 4703, World Bank, 2008. ReferenceRavallion, Martin (2008): “A Global Perspective on Poverty in India”,Economic & Political Weekly, 25 October, pp 31-37.SAMEEKSHA TRUST BOOKSInclusive GrowthK N Raj on Economic DevelopmentEssays from The Economic Weekly and Economic & Political WeeklyEdited by ASHOKA MODYThe essays in the book reflect Professor K N Raj’s abiding interest in economic growth as a fundamental mechanism for lifting the poor and disadvantaged out of poverty. He has also been concerned that the political bargaining process may end up undermining growth and not provide support to those who were excluded from access to economic opportunities. These essays, many of them classics and all published in Economic Weekly and Economic & Political Weekly, are drawn together in this volume both for their commentary on the last half century of economic development and for their contemporary relevance for understanding the political economy of development in India and elsewhere.Pp viii + 338 ISBN 81-250-3045-X 2006 Rs 350Available fromOrient Blackswan LtdMumbai Chennai New Delhi Kolkata Bangalore Bhubaneshwar Ernakulam Guwahati Jaipur LucknowPatna Chandigarh Hyderabad Contact: info@orientblackswan.com
DISCUSSIONnovember 8, 2008 EPW Economic & Political Weekly78‘How Many Poor in the World?’: A Reply Martin RavallionMartin Ravallion (Mravallion@worldbank.org) is with the Development Research Group of the World Bank, Washington DC.A comment on the editorial (25 October) that critiqued the World Bank methodology for estimating the number of poor in the world. This response aims to refute the four major arguments made in the editorial.The editorial “How Many Poor in the World?” (25 October) claims that the World Bank’s latest estimates of the extent of global poverty “present a cri-sis of credibility”.1 The editorial gives four reasons for this rather negative assess-ment. Let me address them in turn. First, it argues that our international poverty line is “little more than a destitution line”. To recap, our proposed standard for measuring global poverty is $1.25 per person per day at 2005 purchasing power parity (PPP), using the PPP exchange rates for household consumption from the 2005 round of the International Comparison Pro-gram(ICP), which is a massive price data collection project spanning 150 countries. The $1.25 line is obtained as the average of the national poverty lines in the poorest 15 countries in a compilation of 75 national poverty lines for developing countries (though the line is similar for the poorest 10 or poorest 20 countries, given that rela-tive poverty effects only start to emerge amongst middle-income countries). We find that 25% of the developing world’s population lived below this line in 2005.There is no doubt that $1.25 a day is a frugal line; most middle income countries have higher lines. The average of all ex-cept the poorest 15 countries is $2.50 a day, which is not reached by more than half the developing world’s population. Towards the other extreme, the average official poverty line in the US is $13 per day in 2005; 96% of those living in the develop-ing world are poor byUS standards. However, if by a “destitution line” one means that a life-threateningly low level of living is implied, then not one of the 15 national poverty lines that go into form-ing the $1.25 line can be considered a destitution line. Each of the national lines meets “industry standards” in its construc-tion, in that it uses a food bundle that attains standard nutritional requirements by prevailing diets and it includes an allowance for non-food spending. (The linesvary, some being more generous thanothers in both the food and non-food allowances; so we take an average.)Also note that India’s official poverty line is about $1.00 a day using the same PPPs; (see my article in the same issue of EPW). So if $1.25 marks “destitution” what are we to make of India’s national poverty line? In truth, neither the $1.25 line nor India’s (lower) official line represent “destitution”. Second, the editorial points out that the standardPPP calculations reflect a broader set of goods than typically consumed by the poor. This is correct, and we have long recognised this point (going back to our first “dollar a day” poverty calculations for the 1990World Development Report). How-ever, the editorial fails to point out (as I also noted in my EPW article) that we have implemented a new set of “PPPs for the poor” using the 2005 ICP data. These re-weight the prices to accord with the consumption patterns of people near the poverty line. We have published results in testing robustness of our new poverty line to this new PPP and alternatives (also referred to in my article). Third, the editorial claims that our method is inconsistent for inter-temporal comparisons. Our method – let us call it method A – entails converting the inter-national poverty line to local currencies using the PPP at the benchmark year for which the new ICP price surveys were done (now 2005) and then adjusting over time using the country’s consumer price index (CPI) to get to the prices prevailing in the relevant survey year. The editorial claims that it would only be a “fluke” if our method gave the same result if instead we had used the obvious alternative, method B, of converting all country-distributions to $ and then made the poverty calcula-tions adjusting for inflation in the US. However, it is readily verified that if the calculation is done correctly, then it would be no fluke at all; methods A and B will always give exactly the same answer given the same data. The mistake made in the editorial (and by one or two other critics) is to not realise that if one uses method B then thePPP for the non-benchmark year 
DISCUSSIONEconomic & Political Weekly EPW november 8, 200879must be calculated consistently with the differential rates of inflation between the US and the country in question. But that is exactly how PPPs are routinely updated for non-benchmark years. Granted, there may well be an incon-sistency if one compares thePPPs obtained from different benchmark years, for the consistency of the PPPs with the differen-tial rates of inflation from the nationals CPIs is not assured (given differences in the quality of goods used and the weights). This is well known. We follow standard practice in other international compari-sons of only doing the PPP conversion for the benchmark year, and then making inter-temporal comparisons using the national data; internal consistency between methods A and B is then automatic. Thus our inter-temporal comparisons of poverty in India follow essentially the same methods used by the Planning Commission; the main difference is that we use a 25% higher poverty line, as I have already noted. Note also that the consistency of the PPPs for different benchmark years with the CPIs is a moot point given that the changes inICP data and methods between benchmark years clearly swamp these other differences. For example, the differ-ences between the 1993 and 2005 ICP surveys – entailing a very substantial improvement in the quality of the price surveys and methods of imputation – make any attempt to compare PPPs across different benchmark years highly dubious. Finally, the editorial argues that the lower the poverty line “the more flattering is the resulting trend decline in poverty” and that this “suggests that the trends...are not so much a reflection of what is actually happening to poverty on the ground as a verification of the consistency of arithmetic”. It is true that in the limit, as the poverty line goes to a very high level, the rate of poverty reduction goes to zero. That is also true in the limit at very low lines; when the poverty rate is near zero it cannot fall much. But we are not near either limit. The fact that we find that the rate of progress against poverty is slightly higher for $1 per day than $2 per day hardly indicates that our results are arithmetic artefacts. We readily acknowledge that the data are not ideal; in this respect, measuring poverty is no different to any other avenue of social or economic research. But we have to make the most of the data availa-ble. Thankfully, there have been great im-provements in the data – from both the household survey data and the price level data from the ICP – since we started our global poverty monitoring effort 20 years ago (and it would be fair to say that the World Bank has played a key role in these improvements). The latest revisions reflect that progress. Note 1 The estimates are reported in Shaohua Chen and Martin Ravallion, “The Developing World Is Poorer Than We Thought, But No Less Successful in the Fight against Poverty”, Policy Research Working Paper 4703, World Bank, 2008. ReferenceRavallion, Martin (2008): “A Global Perspective on Poverty in India”,Economic & Political Weekly, 25 October, pp 31-37.SAMEEKSHA TRUST BOOKSInclusive GrowthK N Raj on Economic DevelopmentEssays from The Economic Weekly and Economic & Political WeeklyEdited by ASHOKA MODYThe essays in the book reflect Professor K N Raj’s abiding interest in economic growth as a fundamental mechanism for lifting the poor and disadvantaged out of poverty. He has also been concerned that the political bargaining process may end up undermining growth and not provide support to those who were excluded from access to economic opportunities. These essays, many of them classics and all published in Economic Weekly and Economic & Political Weekly, are drawn together in this volume both for their commentary on the last half century of economic development and for their contemporary relevance for understanding the political economy of development in India and elsewhere.Pp viii + 338 ISBN 81-250-3045-X 2006 Rs 350Available fromOrient Blackswan LtdMumbai Chennai New Delhi Kolkata Bangalore Bhubaneshwar Ernakulam Guwahati Jaipur LucknowPatna Chandigarh Hyderabad Contact: info@orientblackswan.com
