The "updated" estimates of global poverty by the World Bank do not address the basic problem with the past and current estimates, which is the lack of a clear criterion for identifying the poor. There is no basis to conclude that the new set of purchasing power parity rates employed to generate the new poverty estimates are closer to the "truth". We can only conclude that they are differently distorted than the earlier ones. There is, however, a feasible alternative method that would place at its core an identification criterion for the poor based on elementary human capabilities. The careful coordination of household surveys and poverty line construction across countries so as to capture whether individuals have the resources necessary to achieve these capabilities can ensure enduring comparability of poverty estimates over time and space.
WORLD BANK’S NEW POVERTY ESTIMATESoctober 25, 2008 EPW Economic & Political Weekly44Differently Distorted: The World Bank’s ‘Updated’ Poverty EstimatesSanjay G ReddyThe “updated” estimates of global poverty by the World Bank do not address the basic problem with the past and current estimates, which is the lack of a clear criterion for identifying the poor. There is no basis to conclude that the new set of purchasing power parity rates employed to generate the new poverty estimates are closer to the “truth”. We can only conclude that they are differently distorted than the earlier ones. There is, however, a feasible alternative method that would place at its core an identification criterion for the poor based on elementary human capabilities. The careful coordination of household surveys and poverty line construction across countries so as to capture whether individuals have the resources necessary to achieve these capabilities can ensure enduring comparability of poverty estimates over time and space.The World Bank released late in the summer what it referred to as “updated” global poverty estimates.1 Its new assessment of the extent, distribu-tion and trend of global income poverty was necessitated by the results of a new worldwide survey of prices undertaken by the International Comparison Program (ICP) a multinational initiative of statistical agencies. The World Bank has established a new basic international poverty line (IPL) of $1.25 2005 purchasing power parity (PPP), which replaces earlier basic poverty lines (of $1.08 1993PPP and $1.00 1985 PPP, both widely referred to as “$1 per day”) corresponding to earlier base years. The revised figures purport to estimate world poverty for a range of years since 1981, and thus crucially affect our under-standing of the world over the last quarter century of economic globalisation and widespread adoption of liberal economic policies. In the World Bank’s words, the new poverty estimates “reveal” that “1.4 billion people in the developing world (one in four) were living on less than its current basic poverty line in 2005, down from 1.9 billion (one in two) in 1981”. These figures contrast with the World Bank’s own earlier claim that 969 million people in the devel-oping world (fewer than a fifth) were living on less than its previous basic poverty line in 2004, down from 1.47 billion (somewhat more than one in three and less than two in five) in 1981 [Chen and Ravallion 2007]. The World Bank conceives of the similarity between the two poverty lines as lying in their being in both cases allegedly repre-sentative of poverty lines employed in poor countries. However, as we shall see, this is not the case in a meaningful sense. Moreover, the changes it has made have generated inconsistency and incoherence.Many aspects of the global order, such as the movement toward freer trade, as well as national institutions and policies, are frequently defended by referring to their effect on the poor, not least in the corridors of power. The World Bank’s poverty estimates are thus central to their assessment. Moreover, the first Millennium Development Goal (MDG) is defined in terms of these estimates, making this revi-sion of great importance for determining whether the world is on track to reduce poverty by the amount required.Can the World Bank’s new estimates be trusted? Can they be trusted more than its own earlier greatly lower poverty esti-mates, which they are intended to replace? Unfortunately, the World Bank’s new esti-mates are based on the same methods it used earlier and are undermined by the same problems as the earlier estimates. A Conceptual and Practical MireTwo problems are foremost, as noted in the critique by Sanjay Reddy and Thomas Pogge (forthcoming). The first is that the World Bank’s chosen international poverty line is far too low to cover the cost of purchasing basic necessities. As we docu-mented, a human being could not live even minimally adequately in theUS on $1.25 a day in 2005 (or $1.40 in 2008), nor therefore on an amount that is equivalent in purchasing power elsewhere, contrary to the World Bank’s claims. Since the IPL is defined in purchasing power equivalent units, meant to capture a constant level of purchasing power across countries, semantic consistency requires that the absolute poverty line must suffice to meet basic human requirements in the base country if it is deemed to do so elsewhere. The World Bank insists that its poverty line does so in developing countries, even though it most evidently does not in the US, but if these are both the case then the poverty lines cannot correspond to the same level of purchasing power, as initially presumed. This incoherence is difficult to overcome. The World Bank’s claim that its poverty line is sufficient to capture the require-ments of avoiding absolute poverty in other countries, despite its being insufficient in the US draws attention to the second problem: that the World Bank uses inap-propriate PPPs to convert its poverty line across currencies. Consider the question of how many rupiahs are needed in Jakarta I would like to thank J Harrington, Thomas Pogge and S Subramanian for their invaluable comments and suggestions. The text of this article overlaps partially with that of one published inChallenge.Sanjay G Reddy (sr793@columbia.edu) is at Barnard College, Columbia University, US.
WORLD BANK’S NEW POVERTY ESTIMATESoctober 25, 2008 EPW Economic & Political Weekly46base year comes a new set of PPPs. The new set ofPPPs provides a new set of relative prices deemed to exist betweencountries. The old list (or vector) ofPPPs is not merely shifted up or down. Rather the individual elements in the list are diversely raisedor lowered. Thisconstitutes a twisting of the original PPP vector rather than a simple rescaling. This is why it is not possible to determine “global inflation” within this framework.2 There is merely relative infla-tion and deflation of prices in one country (relative to those in another). This is also why it is impossible to compare straight-forwardly the IPLs of different base years. The relative magnitude of two IPLs de-fined in the units of distinct base years is dependent on the change in the PPP for the country between base years as well as the level of domestic inflation which has taken place in the country, and thus the ratio of the IPLs varies from country to country. It becomes impossible to “update” theIPL according to global inflation (since there exists no such thing within the framework being used) and so it must be fixed anew.However, the procedure used in the calculation must be deemed to be consistent in some relevant sense, lest the concepts of poverty applied in the different base years be conceded to be different (which would imply that the estimates associated with at least one of the base years do not adequately capture absolute poverty). What might that sense be? The World Bank claims that the procedure it uses is consistent because for each base year it fixes its international poverty line so that it is always “representative” of the domes-tic poverty lines used in poor countries. The World Bank’s understanding of repre-sentativeness has two levels, which it ap-pears to view as compatible. At a funda-mental level, it claims that when countries’ poverty lines are lined up according to their per capita income, the resulting curve is kinked – it is relatively flat for poorer countries (implying that poor countries’ poverty lines are relatively close together) up to some threshold level of per capita income, after which it begins to rise. It claims that its IPL is representative of poor countries in the sense that it cor-responds to the flat portion of this kinked curve. At an operational level, the World Bank also seeks to rationalise itsIPL ac-cording to certain calculation methods (such as, most recently, by equating the IPL to the mean of the poverty lines of the15 poorest countries in a sample it has assembled). The World Bank refers both to its (unchanging) fundamental criterion and its (changing) operational criteria in justifying its choice of IPL.The claim that the IPL is consistently representative of the domestic poverty lines used in poor countries is dubious for at least four reasons. First, the World Bank employs shifting conceptions of represent-ativeness at the level of operational criteria. In its first major exercise of global poverty assessment, it claimed that when converted into common units using PPPs (for the 1985 base year) eight domestic poverty lines of poor countries were “close” to the amount it fixed as theIPL. In its second exercise, it fixed theIPL as the median of the 10 lowest domestic poverty lines (when they were converted into common units using PPPs of the 1993 base year). In its third and most recent exercise, it adopted the mean of the domestic poverty lines of the 15 poorest countries (when they were converted into common units using PPPs of the 2005 base year) to fix theIPL. Em-ploying any one of these operational criteria consistently would have led to different results or to no result at all. For instance, the cluster of eight domestic poverty lines around a “most typical” level referred to in the original exercise was no longer to be found when the PPPs were “updated”. Second, the set of poor countries whose domestic poverty lines the World Bank consulted has shifted from one exercise to the next. Third, many of these domestic poverty lines were in fact defined by the World Bank itself and, in any case, may have been informed by irrelevant consid-erations as well as relevant and irrelevant considerations which were discrepant from country to country. Fourth, whether a particular set of poverty lines of poor countries with similar incomes appears to be in a close range (or to constitute a flat portion of the curve relating poverty lines and incomes) depends on the PPPs used to convert them into common units, and these PPPs are hopelessly inappropriate for poverty assessment. When different PPPs are used, the apparent flatness of the curve, and the level at which it is flat are both affected [on which see Reddy and Pogge, forthcoming]. The effort to “update” the base year in which the international poverty line is defined thus ineluctably draws the World Bank into a conceptual and practical mire.Comparing New and OldThe new estimates of the proportion of the world’s population in poverty suggest that the number of poor is almost 50 per cent more than the World Bank had previously proposed. Although the rate of poverty reduction since 1990 is almost the same under the new estimates as the old, this appears to be a fluke and plausibly attri-butable to data errors. If the final year of the comparison is moved backward by just three years to 2002, for instance, the rate of reduction of world poverty ap-pears notably less favourable under the new estimates. Moreover, much of the apparent poverty reduction since may be a figment of the imagination, resulting from the attribution of aggregate growth in the intervening years to the poor, with-out any survey evidence to prove they have experienced real growth in con-sumption. The only region that appears to have had a faster rate of poverty reduc-tion under the new estimates, regardless of whether the period is taken to begin in 1980 or in 1991, is Latin America. Europe and central Asia as well as the Middle East and north Africa have much lower rates of estimated poverty reduction ac-cording to the new estimates than they did previously. The estimated proportion of the popu-lation that is currently poor in Latin America has barely changed under the new esti-mates, whereas it has risen by between 20 per cent and 30 per cent in sub-Saharan Africa and in south Asia and by multiples elsewhere (by a factor greater than four in Europeand central Asia, greater than three in the Middle East and north Africa and almost two in east Asia). The enor-mous fluctuations in the World Bank’s poverty estimates, of a kind which would be unacceptable for most economic statistics, make them unfit for use. The World Bank has already undertaken two revisions of the base year (and the associatedIPL) and wreaked havoc to itspovertyestimates
WORLD BANK’S NEW POVERTY ESTIMATESEconomic & Political Weekly EPW october 25, 200847each time, as it has changed its conception of the IPL that is allegedlyrepresentative of those in poor countries as well as the manner in which it converts this poverty line into the currencies of those countries. Does it intend to continue on the same path? The next global price survey which will inevitably necessitate sucha revision is scheduled for 2011. The World Bank can at that point choose between pulling the rug from underneath itself yet again, continuing to usePPPs from an ear-lier base year despite their growing appar-ent irrelevance, or admitting that its method is wholly wrong. What is the real number of poor people in the world? The scandal is that nobody knows, and attempts to “guesstimate” thenumber on the basis of the existing in-adequate data and methods are currently but fools’ errands. A new and more serious approach to collecting the required data is the only answer. What about the trend? There is reason to believe that there has been enormous poverty reduction in China. However, whether this has led to poverty reduction globally, and if so, how much, depends on how much poverty we think there was originally elsewhere, and what the trend of poverty reduction has been there. In Latin America and the Caribbean, for instance, the more well-justified (although still highly imperfect) estimates of poverty by the UN’s Economic Commission for Latin America and the Caribbean suggest substantially higher absolute poverty estimates than those provided by the World Bank, possibly raising the estimated level of global povertyin each year and lowering its rate of reduction. Estimates of poverty in India vary greatly depending on the method used but there is good reason (as suggested, for instance, by a recent report of the National Commission for Enterprises in the Unorganised Sector) to take a more expansive view of poverty and vulnerability than has traditionally been done in India, in which case reasonable estimates of thepoverty headcount ratio mayrange as high as four-fifths of the population (although this number is quite under-standably controversial). Rates of ReductionTheNCEUS notes that higher poverty lines in India have been associated in the recent period with a lower rate of reduction of the headcount ratio of poverty [Table 1.2 in NCEUS 2008]. This has also been true globally, both within the World Bank’s earlier estimates and its recently revised ones. The table reports the global rate of reduction of poverty for different possible poverty lines.It is evident that the estimated rate of reduction of poverty appears less impres-sive when the IPL is increased. Conceptu-ally, an increase in theIPL is equivalent to scalar multiplication of the list of PPPs by a common factor. However, it may be that the poverty lines used should be increased or decreased differently in different countries in order to be made individually more appropriate (while in-sisting that they possess a common mean-ing in terms of the requirements to achieve elementary human capabilities). As a result, this type of variation is in-completely informative of the implica-tions of employing alternative methods. Nevertheless, the dependence of the global headcount ratio on the IPL provides one example of why one cannot be cavalier about the choices involved.An Available AlternativeThe World Bank’s failure to take criticisms seriously and to develop new poverty esti-mates in a transparent manner mean that the excuse that it is doing the best that canbe done is increasingly flimsy. There exists a feasible alternative method. This approach places at its core an identification criterion for the poor based on elementary human capabilities. The careful coordina-tion of household surveys and poverty lineconstruction across countries so as to capture whether individuals have the re-sources necessary to achieve these capa-bilities can ensure enduring comparability of poverty estimates over time and space, invalidating the need to use ever shifting PPPs as in the current “money metric” approach.3 Such an effort would be along the lines of the coordination of national accounts – a previous crowning achieve-ment of the United Nations and its member countries.4Notes 1 See ‘World Bank Updates Poverty Estimates for Developing World’, World Bank Research Depart-ment web site. 2 Contrary to what is claimed in Bhalla (2002). 3 It is interesting to note in passing that the recent debates over the validity of Indian poverty esti-mates turn on a closely related issue. The “updat-ing” of the 1973-74 poverty line previously estab-lished for India by using available price indices leads to different results than does the replication of the original procedure used to establish that line. This problem stems from the failure to specify the initial poverty line in terms of an adequately detailed conception of elementary capabilities. If this had been done, a price index corresponding to this conception would have suggested itself and the updating of the poverty line over time need not have led to results widely discrepant from those produced by the replication of the original procedure (as at present). 4 Ironically, such an exercise is regularly completed by both the private sector and the international civil service, which produce estimates of the relative cost of living for executives in different cities. In light of this, the claim of infeasibility is almost risible. For further details, see Reddy and Pogge (forthcoming, op cit) and Pogge and Reddy (2006).References Bhalla, S (2002):Imagine There’s No Country: Poverty, Inequality and Growth in the Era of Globalisation, Peterson Institute for International Economics, Washington DC.Chen, Shaohua and Martin Ravallion (2007): ‘Absolute Poverty Measures for the Developing World, 1981-2004’, Proceedings of the National Academy of Sciences, October 23, Vol 104, No 43.Heston, Alan (2008): ‘The 2005 Global Report on Pur-chasing Power Parity Estimates: A Preliminary Review’,Economic & Political Weekly, March 15.NCEUS (2008): Report on Conditions of Work and Promotion of Livelihoods in the Unorganised Sector, New Delhi.Pogge, Thomas (2008): ‘Where the Line Is Drawn: A Rejoinder to Ravallion’, UNDP International Pov-erty Center One Pager 68, October.Pogge, Thomas and Sanjay Reddy (2006): ‘Unknown: Extent, Distribution and Trend of Global Income Poverty’,Economic & Political Weekly, June 3.Reddy, Sanjay and Thomas Pogge (forthcoming): ‘How Not to Count the Poor’ in J Stiglitz, S Anand and P Segal (eds),Debates in the Measurement of Global Poverty, Oxford UniversityPress,Oxford.Table: MDG Achievement at Different International Poverty LinesIPL Level in 2005-dollars 1990 Baseline 2015 Target Annual Reduction Actual How Is the World Doing (Millions of Poor) Reduction Reduction Needed to be Reduction in Regard to MDG-1? of 27.5% Needed to “on Track” Achieved (100% = Exactly on Track) (Millions) Reach Target in 2005 1990-2005 (%)(Millions)(Millions) $1.00/day 1,303.2 358.4 1.28 228.7 424.2 185%; much ahead of schedule$1.25/day 1,817.5 499.8 1.28 318.9 417.9 131%; ahead of schedule$2.00/day 2,753.6 757.2 1.28 483.2 155.8 32%; much behind schedule$2.50/day 3,076.6 846.1 1.28 539.9 -63.6 -12%; regressingSource: Pogge (2008).
WORLD BANK’S NEW POVERTY ESTIMATESoctober 25, 2008 EPW Economic & Political Weekly44Differently Distorted: The World Bank’s ‘Updated’ Poverty EstimatesSanjay G ReddyThe “updated” estimates of global poverty by the World Bank do not address the basic problem with the past and current estimates, which is the lack of a clear criterion for identifying the poor. There is no basis to conclude that the new set of purchasing power parity rates employed to generate the new poverty estimates are closer to the “truth”. We can only conclude that they are differently distorted than the earlier ones. There is, however, a feasible alternative method that would place at its core an identification criterion for the poor based on elementary human capabilities. The careful coordination of household surveys and poverty line construction across countries so as to capture whether individuals have the resources necessary to achieve these capabilities can ensure enduring comparability of poverty estimates over time and space.The World Bank released late in the summer what it referred to as “updated” global poverty estimates.1 Its new assessment of the extent, distribu-tion and trend of global income poverty was necessitated by the results of a new worldwide survey of prices undertaken by the International Comparison Program (ICP) a multinational initiative of statistical agencies. The World Bank has established a new basic international poverty line (IPL) of $1.25 2005 purchasing power parity (PPP), which replaces earlier basic poverty lines (of $1.08 1993PPP and $1.00 1985 PPP, both widely referred to as “$1 per day”) corresponding to earlier base years. The revised figures purport to estimate world poverty for a range of years since 1981, and thus crucially affect our under-standing of the world over the last quarter century of economic globalisation and widespread adoption of liberal economic policies. In the World Bank’s words, the new poverty estimates “reveal” that “1.4 billion people in the developing world (one in four) were living on less than its current basic poverty line in 2005, down from 1.9 billion (one in two) in 1981”. These figures contrast with the World Bank’s own earlier claim that 969 million people in the devel-oping world (fewer than a fifth) were living on less than its previous basic poverty line in 2004, down from 1.47 billion (somewhat more than one in three and less than two in five) in 1981 [Chen and Ravallion 2007]. The World Bank conceives of the similarity between the two poverty lines as lying in their being in both cases allegedly repre-sentative of poverty lines employed in poor countries. However, as we shall see, this is not the case in a meaningful sense. Moreover, the changes it has made have generated inconsistency and incoherence.Many aspects of the global order, such as the movement toward freer trade, as well as national institutions and policies, are frequently defended by referring to their effect on the poor, not least in the corridors of power. The World Bank’s poverty estimates are thus central to their assessment. Moreover, the first Millennium Development Goal (MDG) is defined in terms of these estimates, making this revi-sion of great importance for determining whether the world is on track to reduce poverty by the amount required.Can the World Bank’s new estimates be trusted? Can they be trusted more than its own earlier greatly lower poverty esti-mates, which they are intended to replace? Unfortunately, the World Bank’s new esti-mates are based on the same methods it used earlier and are undermined by the same problems as the earlier estimates. A Conceptual and Practical MireTwo problems are foremost, as noted in the critique by Sanjay Reddy and Thomas Pogge (forthcoming). The first is that the World Bank’s chosen international poverty line is far too low to cover the cost of purchasing basic necessities. As we docu-mented, a human being could not live even minimally adequately in theUS on $1.25 a day in 2005 (or $1.40 in 2008), nor therefore on an amount that is equivalent in purchasing power elsewhere, contrary to the World Bank’s claims. Since the IPL is defined in purchasing power equivalent units, meant to capture a constant level of purchasing power across countries, semantic consistency requires that the absolute poverty line must suffice to meet basic human requirements in the base country if it is deemed to do so elsewhere. The World Bank insists that its poverty line does so in developing countries, even though it most evidently does not in the US, but if these are both the case then the poverty lines cannot correspond to the same level of purchasing power, as initially presumed. This incoherence is difficult to overcome. The World Bank’s claim that its poverty line is sufficient to capture the require-ments of avoiding absolute poverty in other countries, despite its being insufficient in the US draws attention to the second problem: that the World Bank uses inap-propriate PPPs to convert its poverty line across currencies. Consider the question of how many rupiahs are needed in Jakarta I would like to thank J Harrington, Thomas Pogge and S Subramanian for their invaluable comments and suggestions. The text of this article overlaps partially with that of one published inChallenge.Sanjay G Reddy (sr793@columbia.edu) is at Barnard College, Columbia University, US.
WORLD BANK’S NEW POVERTY ESTIMATESoctober 25, 2008 EPW Economic & Political Weekly46base year comes a new set of PPPs. The new set ofPPPs provides a new set of relative prices deemed to exist betweencountries. The old list (or vector) ofPPPs is not merely shifted up or down. Rather the individual elements in the list are diversely raisedor lowered. Thisconstitutes a twisting of the original PPP vector rather than a simple rescaling. This is why it is not possible to determine “global inflation” within this framework.2 There is merely relative infla-tion and deflation of prices in one country (relative to those in another). This is also why it is impossible to compare straight-forwardly the IPLs of different base years. The relative magnitude of two IPLs de-fined in the units of distinct base years is dependent on the change in the PPP for the country between base years as well as the level of domestic inflation which has taken place in the country, and thus the ratio of the IPLs varies from country to country. It becomes impossible to “update” theIPL according to global inflation (since there exists no such thing within the framework being used) and so it must be fixed anew.However, the procedure used in the calculation must be deemed to be consistent in some relevant sense, lest the concepts of poverty applied in the different base years be conceded to be different (which would imply that the estimates associated with at least one of the base years do not adequately capture absolute poverty). What might that sense be? The World Bank claims that the procedure it uses is consistent because for each base year it fixes its international poverty line so that it is always “representative” of the domes-tic poverty lines used in poor countries. The World Bank’s understanding of repre-sentativeness has two levels, which it ap-pears to view as compatible. At a funda-mental level, it claims that when countries’ poverty lines are lined up according to their per capita income, the resulting curve is kinked – it is relatively flat for poorer countries (implying that poor countries’ poverty lines are relatively close together) up to some threshold level of per capita income, after which it begins to rise. It claims that its IPL is representative of poor countries in the sense that it cor-responds to the flat portion of this kinked curve. At an operational level, the World Bank also seeks to rationalise itsIPL ac-cording to certain calculation methods (such as, most recently, by equating the IPL to the mean of the poverty lines of the15 poorest countries in a sample it has assembled). The World Bank refers both to its (unchanging) fundamental criterion and its (changing) operational criteria in justifying its choice of IPL.The claim that the IPL is consistently representative of the domestic poverty lines used in poor countries is dubious for at least four reasons. First, the World Bank employs shifting conceptions of represent-ativeness at the level of operational criteria. In its first major exercise of global poverty assessment, it claimed that when converted into common units using PPPs (for the 1985 base year) eight domestic poverty lines of poor countries were “close” to the amount it fixed as theIPL. In its second exercise, it fixed theIPL as the median of the 10 lowest domestic poverty lines (when they were converted into common units using PPPs of the 1993 base year). In its third and most recent exercise, it adopted the mean of the domestic poverty lines of the 15 poorest countries (when they were converted into common units using PPPs of the 2005 base year) to fix theIPL. Em-ploying any one of these operational criteria consistently would have led to different results or to no result at all. For instance, the cluster of eight domestic poverty lines around a “most typical” level referred to in the original exercise was no longer to be found when the PPPs were “updated”. Second, the set of poor countries whose domestic poverty lines the World Bank consulted has shifted from one exercise to the next. Third, many of these domestic poverty lines were in fact defined by the World Bank itself and, in any case, may have been informed by irrelevant consid-erations as well as relevant and irrelevant considerations which were discrepant from country to country. Fourth, whether a particular set of poverty lines of poor countries with similar incomes appears to be in a close range (or to constitute a flat portion of the curve relating poverty lines and incomes) depends on the PPPs used to convert them into common units, and these PPPs are hopelessly inappropriate for poverty assessment. When different PPPs are used, the apparent flatness of the curve, and the level at which it is flat are both affected [on which see Reddy and Pogge, forthcoming]. The effort to “update” the base year in which the international poverty line is defined thus ineluctably draws the World Bank into a conceptual and practical mire.Comparing New and OldThe new estimates of the proportion of the world’s population in poverty suggest that the number of poor is almost 50 per cent more than the World Bank had previously proposed. Although the rate of poverty reduction since 1990 is almost the same under the new estimates as the old, this appears to be a fluke and plausibly attri-butable to data errors. If the final year of the comparison is moved backward by just three years to 2002, for instance, the rate of reduction of world poverty ap-pears notably less favourable under the new estimates. Moreover, much of the apparent poverty reduction since may be a figment of the imagination, resulting from the attribution of aggregate growth in the intervening years to the poor, with-out any survey evidence to prove they have experienced real growth in con-sumption. The only region that appears to have had a faster rate of poverty reduc-tion under the new estimates, regardless of whether the period is taken to begin in 1980 or in 1991, is Latin America. Europe and central Asia as well as the Middle East and north Africa have much lower rates of estimated poverty reduction ac-cording to the new estimates than they did previously. The estimated proportion of the popu-lation that is currently poor in Latin America has barely changed under the new esti-mates, whereas it has risen by between 20 per cent and 30 per cent in sub-Saharan Africa and in south Asia and by multiples elsewhere (by a factor greater than four in Europeand central Asia, greater than three in the Middle East and north Africa and almost two in east Asia). The enor-mous fluctuations in the World Bank’s poverty estimates, of a kind which would be unacceptable for most economic statistics, make them unfit for use. The World Bank has already undertaken two revisions of the base year (and the associatedIPL) and wreaked havoc to itspovertyestimates 
WORLD BANK’S NEW POVERTY ESTIMATESEconomic & Political Weekly EPW october 25, 200847each time, as it has changed its conception of the IPL that is allegedlyrepresentative of those in poor countries as well as the manner in which it converts this poverty line into the currencies of those countries. Does it intend to continue on the same path? The next global price survey which will inevitably necessitate sucha revision is scheduled for 2011. The World Bank can at that point choose between pulling the rug from underneath itself yet again, continuing to usePPPs from an ear-lier base year despite their growing appar-ent irrelevance, or admitting that its method is wholly wrong. What is the real number of poor people in the world? The scandal is that nobody knows, and attempts to “guesstimate” thenumber on the basis of the existing in-adequate data and methods are currently but fools’ errands. A new and more serious approach to collecting the required data is the only answer. What about the trend? There is reason to believe that there has been enormous poverty reduction in China. However, whether this has led to poverty reduction globally, and if so, how much, depends on how much poverty we think there was originally elsewhere, and what the trend of poverty reduction has been there. In Latin America and the Caribbean, for instance, the more well-justified (although still highly imperfect) estimates of poverty by the UN’s Economic Commission for Latin America and the Caribbean suggest substantially higher absolute poverty estimates than those provided by the World Bank, possibly raising the estimated level of global povertyin each year and lowering its rate of reduction. Estimates of poverty in India vary greatly depending on the method used but there is good reason (as suggested, for instance, by a recent report of the National Commission for Enterprises in the Unorganised Sector) to take a more expansive view of poverty and vulnerability than has traditionally been done in India, in which case reasonable estimates of thepoverty headcount ratio mayrange as high as four-fifths of the population (although this number is quite under-standably controversial). Rates of ReductionTheNCEUS notes that higher poverty lines in India have been associated in the recent period with a lower rate of reduction of the headcount ratio of poverty [Table 1.2 in NCEUS 2008]. This has also been true globally, both within the World Bank’s earlier estimates and its recently revised ones. The table reports the global rate of reduction of poverty for different possible poverty lines.It is evident that the estimated rate of reduction of poverty appears less impres-sive when the IPL is increased. Conceptu-ally, an increase in theIPL is equivalent to scalar multiplication of the list of PPPs by a common factor. However, it may be that the poverty lines used should be increased or decreased differently in different countries in order to be made individually more appropriate (while in-sisting that they possess a common mean-ing in terms of the requirements to achieve elementary human capabilities). As a result, this type of variation is in-completely informative of the implica-tions of employing alternative methods. Nevertheless, the dependence of the global headcount ratio on the IPL provides one example of why one cannot be cavalier about the choices involved.An Available AlternativeThe World Bank’s failure to take criticisms seriously and to develop new poverty esti-mates in a transparent manner mean that the excuse that it is doing the best that canbe done is increasingly flimsy. There exists a feasible alternative method. This approach places at its core an identification criterion for the poor based on elementary human capabilities. The careful coordina-tion of household surveys and poverty lineconstruction across countries so as to capture whether individuals have the re-sources necessary to achieve these capa-bilities can ensure enduring comparability of poverty estimates over time and space, invalidating the need to use ever shifting PPPs as in the current “money metric” approach.3 Such an effort would be along the lines of the coordination of national accounts – a previous crowning achieve-ment of the United Nations and its member countries.4Notes 1 See ‘World Bank Updates Poverty Estimates for Developing World’, World Bank Research Depart-ment web site. 2 Contrary to what is claimed in Bhalla (2002). 3 It is interesting to note in passing that the recent debates over the validity of Indian poverty esti-mates turn on a closely related issue. The “updat-ing” of the 1973-74 poverty line previously estab-lished for India by using available price indices leads to different results than does the replication of the original procedure used to establish that line. This problem stems from the failure to specify the initial poverty line in terms of an adequately detailed conception of elementary capabilities. If this had been done, a price index corresponding to this conception would have suggested itself and the updating of the poverty line over time need not have led to results widely discrepant from those produced by the replication of the original procedure (as at present). 4 Ironically, such an exercise is regularly completed by both the private sector and the international civil service, which produce estimates of the relative cost of living for executives in different cities. In light of this, the claim of infeasibility is almost risible. For further details, see Reddy and Pogge (forthcoming, op cit) and Pogge and Reddy (2006).References Bhalla, S (2002):Imagine There’s No Country: Poverty, Inequality and Growth in the Era of Globalisation, Peterson Institute for International Economics, Washington DC.Chen, Shaohua and Martin Ravallion (2007): ‘Absolute Poverty Measures for the Developing World, 1981-2004’, Proceedings of the National Academy of Sciences, October 23, Vol 104, No 43.Heston, Alan (2008): ‘The 2005 Global Report on Pur-chasing Power Parity Estimates: A Preliminary Review’,Economic & Political Weekly, March 15.NCEUS (2008): Report on Conditions of Work and Promotion of Livelihoods in the Unorganised Sector, New Delhi.Pogge, Thomas (2008): ‘Where the Line Is Drawn: A Rejoinder to Ravallion’, UNDP International Pov-erty Center One Pager 68, October.Pogge, Thomas and Sanjay Reddy (2006): ‘Unknown: Extent, Distribution and Trend of Global Income Poverty’,Economic & Political Weekly, June 3.Reddy, Sanjay and Thomas Pogge (forthcoming): ‘How Not to Count the Poor’ in J Stiglitz, S Anand and P Segal (eds),Debates in the Measurement of Global Poverty, Oxford UniversityPress,Oxford.Table: MDG Achievement at Different International Poverty LinesIPL Level in 2005-dollars 1990 Baseline 2015 Target Annual Reduction Actual How Is the World Doing (Millions of Poor) Reduction Reduction Needed to be Reduction in Regard to MDG-1? of 27.5% Needed to “on Track” Achieved (100% = Exactly on Track) (Millions) Reach Target in 2005 1990-2005 (%)(Millions)(Millions) $1.00/day 1,303.2 358.4 1.28 228.7 424.2 185%; much ahead of schedule$1.25/day 1,817.5 499.8 1.28 318.9 417.9 131%; ahead of schedule$2.00/day 2,753.6 757.2 1.28 483.2 155.8 32%; much behind schedule$2.50/day 3,076.6 846.1 1.28 539.9 -63.6 -12%; regressingSource: Pogge (2008).
