81801 OCTOBER 2013 • Number 125 The State of the Poor: Where Are The Poor, Where Is Extreme Poverty Harder to End, and What Is the Current Profile of the World’s Poor? Pedro Olinto, Kathleen Beegle, Carlos Sobrado, and Hiroki Uematsu Although the world witnessed an unprecedented pace of poverty reduction over the last decades, reducing the number of people living in extreme poverty by more than 700 million, approximately 1.2 billion people remained entrenched in destitution in 2010.1 In order to leverage developing country efforts and galvanize the international development commu- nity to exert concerted effort to end extreme poverty, the World Bank has established the twin goals of ending extreme poverty by 2030 and promoting shared prosperity by fostering income growth of the bottom 40 percent of the population in every country. Ending extreme poverty in just one generation is a formidable challenge by all accounts that requires a thorough understanding of the state of the poor. The objective of this note is to analyze some of the diverse by 33 percent between 1981 and 2010. This is due to an in- characteristics of 1.2 billion poor people who are the focus of crease in the number of extremely poor individuals in LICs by the poverty reduction efforts of governments and the interna- more than 100 million, and the stagnant average income tional development community. Despite the impressive prog- among the poor that remained almost as low in 2010 as it was ress in the fight against poverty in the developing world as a back in 1981. whole, the progress has been much slower in Low Income As a share of the GDP of the developing world, the Ag- Countries (LICs). Poverty for middle and high income coun- gregate Poverty Gap is now less than one tenth of what it was tries fell by more than a half since 1981. For LICs, however, 30 years ago. For LICs, the share in 2010 was approximately 8 extreme poverty fell by less than a third. percent of their GDP, down from 24 percent in 1981. Not- The depth of extreme poverty, that is, how far the average withstanding this significant decline, the Aggregate Poverty extremely poor person is from the $1.25 per day poverty line, Gap/GDP ratio in LICs is 16 times larger than the average for has fallen by 25 percent in the past 30 years for the developing the developing world. world as a whole. But most of this drop seems to have hap- The note also features an in-depth profile of extreme pov- pened in China and India. For the rest of the developing erty at a global scale using household survey data collected in world, individuals living in extreme poverty today appear to 73 countries during the 2000s. On the one hand, it offers be as poor as those living in extreme poverty 30 years ago. valuable insights as to where poverty is deeply seated and The aggregate additional annual income needed to lift where stronger efforts are needed: more than three quarters every individual in the developing world out of extreme pov- of those living in extreme poverty are in rural areas and nearly erty (the Aggregate Poverty Gap) has been reduced by more two thirds of the extremely poor earn a living from agricul- than half for the developing world. For LICs, it has increased ture. On the other hand, some results are alarming and dis- 1 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise turbing. More than one-third of the extremely poor individu- 44 percent of LIC citizens lived in complete destitution. This als are children under age of 13, and half of children in LICs is more than twice the average rate for developing countries are in extreme poverty. (21 percent). The global gender gap in education is concentrated Although extreme poverty rates have fallen everywhere, among the poor. Poor women aged 15 to 30, on average, have the number of poor people in LICSs has increased by 103 a year less schooling than poor men of the same age group. For million between 1981 and 2010 (figure 2). While the number the nonpoor, the gender gap is almost half of the gap for the of extremely poor individuals has declined in middle and poor. high income countries (including India and China), in the Access to essential utilities such as electricity, water, last three decades it has increased in LICs. As a consequence, and sanitation are very limited among the poor. The non- after India (33 percent), LICs contain most of the extremely poor are more than twice as likely to have water, and three poor in the world (29 percent in 2010). In 1981 only 13 per- times as likely to have sanitation. While 87 percent of the cent of the extremely poor resided in LICs. nonpoor have electricity, among the poor, just under half As shown in figure 3, both the number of extremely poor have electricity. individuals and the number of people living with incomes The rest of the note is organized as follows. The next sec- above $1.25 per day have increased in LICs. But the number tion discusses trends in poverty rates and poor population in of people living with incomes above $1.25 has increased the developing world between 1981 and 2010, fol- lowed by a detailed analysis on the depth of pover- Figure 1. The Developing World Has Experienced a Large Decline in Extreme ty and average income of the poor. The fourth sec- Poverty Rates tion introduces the concept of the Aggregate 90 Poverty Gap as a proxy for the remaining magni- 84 80 tude of extreme poverty and demonstrates how percent of poor persons this measure has become less important relative to 70 63 the size of economic activities measured by GDP, 60 60 52.1 except for LICs. Results from global poverty profil- 50 ing are discussed in the fifth section. The final sec- 46 40.9 low income, 44 40 34.0 tion concludes. 42.1 India, 33 30 25.0 lower middle, 21 Poverty Trends in the Developing 20 developing world, 20.6 World 10 11 China, 12 At least 721 million fewer people live in extreme upper middle +, 4 0 poverty in the world today than 30 years ago. Pov- 81 84 87 90 93 96 99 02 05 08 10 19 19 19 19 19 19 19 20 20 20 20 erty reduction surpassed expectations—the Mil- lennium Developing Goal 1 of halving extreme Source: World Bank staff estimates based on PovcalNet. poverty between 1990 and 2015 was reached five years ahead of time. But to lift the remaining poor Figure 2. The Number of Extremely Poor People Has Declined by More Than 721 Million between 1981 and 2010 out of deprivation, and to end extreme poverty by 2030, the world will have to work harder and 2,000 1,930 1,898 number of poor persons (millions) 1,800 1,731 smarter as extreme poverty becomes a more diffi- cult problem to solve requiring more targeted and 1,600 835 43% highly effective solutions. To start, we need to 1,400 1,210 know who the extremely poor are, where they live, 1,200 155 13% and where poverty is deepest and harder to end. 1,000 96 5% 48 4% To reach the goal of ending extreme poverty 800 428 394 33% 22% by 2030, the pace of poverty reduction in LICs will 600 260 22% have to increase substantially.2 While extreme 400 322 17% poverty has fallen across the developing world in 200 352 249 13% 29% the last three decades, the pace was considerably 0 1981 1990 1999 2010 slower in LICs (figure 1). Poverty for middle and high income countries (including India and Chi- low income upper middle + China lower middle India total na) fell by more than a half since 1981. For LICs, extreme poverty fell by less than a third. By 2010, Source: World Bank staff estimates based on PovcalNet. 2 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise Figure 3. Population in Low Income Countries Living Above and how far the average poor person is from escaping extreme pov- Below US$1.25 per Day (millions) erty. In the developing world as a whole, the average person extremely poor living in extreme poverty had a higher income in 2010 than it 900 800 had in 1981 ($0.87 per day in 2010 versus $0.74 in 1981, or number of persons (millions) 800 population living with > $1.25/day 700 total population in LICs 18 percent higher). Therefore, the average depth of extreme 627 600 poverty decreased from $0.51 in 1981 to $0.38 in 2010, a 25 499 448 500 257 percent decrease. 397 Most of the drop in the depth of poverty seems to have 400 183 300 147 happened in China and India. For the rest of the developing 200 370 world, individuals living in extreme poverty today appear to 316 352 100 249 be as poor as those living in extreme poverty 30 years ago. For 0 LICs, the average income of a person in extreme poverty in 1981 1990 1999 2010 2010 was not much different from the average income of an Source: World Bank staff estimates based on PovcalNet and World Development extremely poor person in 1981. It increased by only 5 per- Indicators. cent, going from $0.74 per day in 1981 to $0.78 in 2010. The income of the average extremely poor person in middle and much faster (at 3.9 percent per year versus 1.2 percent per high income developing countries also did not increase much year). Also note that the growth rate of the extremely poor between 1981 and 2010. But in India and China, the average population becomes negative at the turn of the century and, extremely poor were much closer to escaping extreme pover- despite population growth, there were 18 million fewer ex- ty in 2010 than the average extremely poor elsewhere in the tremely poor people in 2010 than there were in 1999. world. The income of the average extremely poor in India in- The Depth of Poverty creased by 14 percent, from $0.84 to $0.96. In China, the income of the average extremely poor jumped by 42 percent The depth of extreme poverty, that is, how far the average ex- in 3 decades, from $0.67 to $0.95. tremely poor person is from the $1.25 per day poverty line, has fallen by 25 percent in the past 30 years in the developing The Aggregate Poverty Gap world (figure 4). The poverty rate captures the share of people living below the threshold of $1.25 per day, but it does not The aggregate additional annual income needed to lift every tell us how deep poverty is. The average daily income of the individual in the developing world out of extreme poverty poor compared to the threshold of $1.25 per day indicates (the Aggregate Poverty Gap) has been reduced by more Figure 4. The Income of the Extremely Poor Has Increased, But Not at the Same Rate Everywhere 1.25 poverty line PPP$1.25 1.15 India 2010 average gap average daily income for the poor India PPP$0.29 1.05 gap PPP$ 0.41 India, 0.96 0.95 China, 0.95 developing world, 0.87 0.85 0.84 0.84 lower middle, 0.81 0.80 0.82 low income, 0.78 0.77 upper middle +, 0.77 0.75 0.74 0.74 0.67 0.65 1981 1990 1999 2010 Source: World Bank staff estimates based on PovcalNet. 3 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise Figure 5. The Aggregate Poverty Gap Has Fallen Dramatically, and the Shares gregate income needed to lift all Chinese poor to Have Shifted $1.25 a day represented 33 percent of China’s GDP. Today it represents a tiny fraction of the country’s na- yearly value (billions, US$ 2005 prices) 400 tional product. 361 350 Also striking is how far LICs are from the rest of 298 300 the world in terms the APG/GDP ratio. For LICs, the 178 49% 257 250 share in 2010 was approximately 8 percent of their 200 169 GDP. For all other income groups, this ratio falls be- 16 4% 10% 150 17 5% low 1.1 percent of GDP. The APG/GDP ratio has de- 64 18% 8 41 24% clined substantially from 24 percent since 1981, but 100 57 16% 41 25% the ratio is still considerably larger in LICs than in the 50 46 13% 61 36% rest of the world. 0 1981 1990 1999 2010 While the developing world’s APG/GDP ratio is low income upper middle + China not how much it would cost to end extreme poverty, lower middle India total its sharp decline indicates that the size of the problem relative to the aggregate income of developing coun- Source: World Bank staff estimates based on PovcalNet. tries has likely fallen substantially since 1981. It also indicates that, except for LICs, resources are unlikely than half in the last 30 years (figure 5). Multiplying the to be the main limitation to ending extreme poverty depth of poverty by the number of extremely poor people in in most countries.4 By 2010, the APG/GDP ratio was near or the world, one arrives at the Aggregate Poverty Gap (APG). below 1 percent for all country income groups, except LICs. That is, the aggregate increase in the incomes of the extreme- Thus, the challenge for middle and high income groups is not ly poor needed to lift them all above the extreme poverty so much the amount of resources required by the poor, but line of $1.25 per day.3 In 2010, for the world as whole, this development and implementation of policies and programs figure added to approximately $169 billion (in 2005 Pur- that help redirect those resources to the poor. For LICs, how- chasing Power Parity dollars). This is less than half of what it ever, resources are still likely to be a major constraint to end- was three decades ago. ing extreme poverty. While it has decreased for all other country groups, the Aggregate Poverty Gap for LICs has increased by 33 percent A Profile of the World’s Poor between 1981 and 2010. The contribution of LICs to the Until recently, statistics on extreme poverty have been limit- global APG is by far the highest (36 percent) and it has al- ed to basic ones such as poverty rate, number of poor, and pov- most tripled in the last 3 decades (from 13 percent in 1981). erty gap. With the renewed focus on ending extreme poverty Despite a reduction in the poverty rates, population growth by 2030, there will be an ever increasing demand for more and stagnant incomes of the poor resulted in an increase in detailed and comprehensive information about the 1.2 bil- the APG in LICs from $46 to $61 billion in 1981–2010 lion extremely poor individuals: Where do they live, how do an increase of almost one third. The share of Lower Middle they earn a living, how old are they, and how many of them Income Countries’ APG has increased from 16 percent to 25 have access to basic services? percent of the world’s APG. India’s share of the APG, de- spite both poverty reduction and income gains for the poor, Figure 6. Aggregate Poverty Gap as a Percentage of GDP Has Fallen also increased due to population growth, from 18 to 24 Steeply Everywhere, But Remains High in LICs percent. % of extrteme poverty gap value The world’s APG as a share of the GDP of developing 35 China, 32.9 countries is now less than one tenth of what it was 30 years 30 ago (figure 6). In 1981, the total APG was 5.3 percent of 25 over average GDP 23.7 the GDP of developing countries. In 2010, it was only 0.5 20 percent. That is, if developing countries were able to sud- denly raise the incomes of all extremely poor individuals to 15 $1.25 per day, they would have needed 0.5 percent of their 10 India, 9.8 low income, 7.9 India, 1.1 GDP to do so. Three decades ago, they would have needed 5.4 lower middle, 1.0 5 5.3 1.5 developing world, 0.5 more than 10 times that. The decline of the APG/GDP ra- 0 0.4 2.3 China, 0.2 tio for China is astonishing. It fell from a third of its GDP in 1981 1990 1999 2010 upper middle +, 0.1 1981 to only 0.2 percent in 2010. That is, in 1981 the ag- Source: World Bank staff estimates based on PovcalNet. 4 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise Answers to these questions are called poverty profiles. women aged 15 to 30, on average, have a year less schooling Poverty profiling has commonly been conducted at the than poor men of the same age group. For the nonpoor, the country level, where a country-specific definition of pover- gender gap is less than half. ty is used. Recent development in globally harmonized Poverty rates are highest among children (figure 9). A data5 now allow us to produce profiles of the poor defined third of all poor in the developing world are children 0–12 by the international poverty line of $1.25 a day, thereby al- years, while children are 20 percent of the nonpoor. This pat- lowing us to compare socioeconomic and geographical tern is most dramatic in LICs, where half of all children live in characteristics of the extremely poor across countries. To poverty (figure 10). Because of this demographic pattern, the the best of our knowledge, this is the first attempt to report number of prime-age adults to provide income and support profiles of poor individuals living below $1.25 per day at a per child in nonpoor households is much higher (3 adults) global scale. than in poor households (1.4 adults). Poverty is concentrated in rural areas and the poor are There are large gaps in access to basic services between most likely to earn income in agriculture (figure 7). Over 78 the poor and the nonpoor (figure 11). The nonpoor are more percent of the poor reside in rural area, while the rural popu- than twice as likely to have water, and three times as likely to lation is 58 percent of the developing world. A rural house- have sanitation. While 87 percent of the nonpoor have elec- hold is thus more likely to be poor than an urban one. Not tricity, among the poor, just under half have electricity. surprisingly, a large share of the poor (63 percent) are working Conclusions in agriculture—mostly smallholder farming. The gender gap in education is concentrated among the With the renewed focus on ending extreme poverty by 2030, poor (figure 8). While the poor are equally divided by gender, detailed and comprehensive knowledge of the 1.2 billion ex- the traits of the poor do take on a gender dimension. Poor tremely poor individuals will increasingly be crucial. The ob- jective of the note was to identify where extreme poverty re- Figure 7. Most of the Poor Live in Rural Areas and Work in Agriculture Figure 9. Children Are More Likely to Be Extremely Poor Than 90 extremely poor Others 80 nonpoor 100 total population 5% 8% 70 77.8 percentage of people 80 61 years 60 62.8 or older 47% 58.4 59% percentage 50 53.1 60 19 to 60 years old 40 40.7 40 13% 13 to 18 30 34.9 years old 13% 20 20 34% 12 years or 10 20% younger 0 0 extremely nonpoor rural agriculture poor Source: World Bank staff estimates based on I2D2. Source: World Bank staff estimates based on I2D2. Figure 8. Poor Women Have Lower Schooling Than Poor Men Figure 10. Poverty Rate among Children Exceeds 50 Percent in 10 Low-Income Countries 50 female 9 low-income countries 50% 50% male 9.0 52% developing world 8.6 8 40 years of education 7 42% 6.7 percentage 6 30 5.7 5 32% 20 4 3 2 19% 10 1 0 0 percentage of extremely nonpoor 12 years or 13 years or extremely poor younger older poor education 15–30 years old Source: World Bank staff estimates based on I2D2. Source: World Bank staff estimates based on I2D2. 5 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise Figure 11. The Poor Continue to Lag in Access to Basic Services come by the World Bank, as well as some countries that are no 100 longer Upper Middle Income by the World Bank’s classifica- electricity tion, but have graduated to High Income. As of July 1st, sanitation 2013, countries were classified by the World Bank as Low In- 80 water 87 percentage of population come if their per capita annual Gross National Income (per capita GNI) in 2012 was below $1,035 per year. Lower Mid- with access 60 dle Income are countries for which the per capita GNI is be- 61 56 tween $1,036 and $4,085. Upper Middle Income are coun- 49 tries for which the per capita GNI is between $4,086 and 40 $12,615, and High Income are countries for which per capita 26 GNI is above $12,615. See appendix for the list of countries 20 20 by the latest World Bank classification. 3. In other words, if we had a magic wand and could perfectly 0 target every extremely poor individual, and magically raise poor (extreme) nonpoor their incomes to the $1.25 per day extreme poverty line, in Source: World Bank staff estimates based on I2D2. 2010 the world needed approximately $169 billion per year (in 2005 PPP dollars) to end extreme poverty. The value of mains prevalent and where stronger efforts are called for. The the Aggregate Poverty Gap, however, is not the same as the analyses repeatedly pointed toward LICs where progress in cost of ending extreme poverty. It is the size of the problem poverty reduction has been the slowest over the last three de- which is different from the size (cost) of the solution. cades. Our first attempt to profile extremely poor individuals 4. Another way of interpreting the APG/GDP ratio is the fol- at a global scale revealed alarming living conditions under lowing: Suppose that the real GDP growth for the developing which they live and that more than one third of the poor are world as a whole is 5 percent per year. If 10 percent of this children 12 years old or younger. The analysis presented here, GDP growth accrued to the 21 percent of the developing as well as future work on the state of the poor, should be in- world’s population who are extremely poor, and this 10 per- strumental in enhancing awareness of the issue of extreme cent was distributed in a way that the growth in income of poverty amongst the broader public and facilitate discussion each poor person was exactly his/her distance to the $1.25 on effective poverty reduction policies in the international line, extreme poverty would end in one year. development community to put an end to extreme poverty, 5. The primary data source for the profiles of the poor is the once and forever. International Income Distribution Database (I2D2), a glob- About the Authors ally harmonized database drawn from more than 600 na- tionally representative household surveys. See appendix for Kathleen Beegle is Lead Economist, Pedro Olinto and Carlos Sob- more detail about the I2D2 and how global poverty profiles rado are Senior Economists, and Hiroki Uematsu is Junior Profes- are prepared. sional Officer, all at the World Bank. References Notes Chen, S. and M. Ravallion. 2010. “The Developing World Is Poorer 1. In this note we interchangeably use the terms poor and ex- Than We Thought, But No Less Successful in the Fight Against tremely poor to refer to those living with an income below Poverty.” Quarterly Journal of Economics 125(4): 1577–1625. $1.25 per day in 2005 Purchasing Power Parity dollars as de- Ravallion, M., S. Chen, and P. Sangraula. 2009. “Dollar a Day Revis- ited.” The World Bank Economic Review 23(2): 163–84. scribed in Chen and Ravallion (2011) and Ravallion et al. (2009). While we sometimes refer to those living with in- Appendix comes above $1.25 per day as nonpoor, needless to say, many of them are still very poor, even though not classified as ex- The appendix documents data sources and describes the tremely poor or poor. methodology used to calculate global poverty profiles report- 2. In this note, developing countries are grouped into one of ed in this note. Three data sources are consulted: PocvalNet, the following 5 categories: India, China, Low Income, Lower World Development Indicators (WDI), and the International Middle Income and Upper Middle Income. The Low Income Income Distribution Database (I2D2). Data from PovcalNet and Lower Middle Income groups correspond to the World and WDI are publicly available. Bank's latest income classification as of July 1st, 2013. The PovcalNet Upper Middle Income group, as defined for this note, in- All poverty estimates in this note are from PovcalNet (http:// cludes the countries currently classified as Upper Middle In- iresearch.worldbank.org/PovcalNet /index.htm), an online 6 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise Table A1. List of Countries by Income Classification Low Income Afghanistan Congo, Dem. Rep. Kyrgyz Republic Rwanda Bangladesh Eritrea Liberia Sierra Leone Benin Ethiopia Madagascar Somalia Burkina Faso Gambia, The Malawi Tajikistan Burundi Guinea Mali Tanzania Cambodia Guinea-Bissau Mozambique Togo Central African Republic Haiti Myanmar Uganda Chad Kenya Nepal Zimbabwe Comoros Korea, Dem. Rep. Niger   Lower Middle Income Armenia Guatemala Mongolia Sudan Bhutan Guyana Morocco Swaziland Bolivia Honduras Nicaragua Syrian Arab Republic Cameroon India Nigeria São Tomé and Principe Cape Verde Indonesia Pakistan Timor-Leste Congo, Rep. Kiribati Papua New Guinea Ukraine Côte d’Ivoire Kosovo Paraguay Uzbekistan Djibouti Lao PDR Philippines Vanuatu Egypt, Arab Rep. Lesotho Samoa Vietnam El Salvador Mauritania Senegal West Bank and Gaza Georgia Micronesia, Fed. Sts. Solomon Islands Yemen, Rep. Ghana Moldova Sri Lanka Zambia Upper Middle and High Income Albania Cuba Lebanon Serbia Algeria Czech Republic Libya Seychelles American Samoa Dominica Lithuania Slovak Republic Angola Dominican Republic Macedonia, FYR Slovenia Antigua and Barbuda Ecuador Malaysia South Africa Argentina Equatorial Guinea Maldives St. Lucia Azerbaijan Estonia Marshall Islands St. Vincent & the Grenadines Belarus Fiji Mauritius Suriname Belize Gabon Mexico Thailand Bosnia and Herzegovina Grenada Montenegro Tonga Botswana Hungary Namibia Trinidad and Tobago Brazil Iran, Islamic Rep. Palau Tunisia Bulgaria Iraq Panama Turkey Chile Jamaica Peru Turkmenistan China Jordan Poland Tuvalu Colombia Kazakhstan Romania Uruguay Costa Rica Korea, Rep. Russian Federation Venezuela, RB Croatia Latvia     7 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise poverty measurement tool maintained by the World Bank’s Because availability and frequency of household surveys Development Research Group. PovcalNet provides two types differ significantly across countries, it is infeasible to produce of poverty estimates: country-specific poverty estimates as of global poverty profiles in a given year while maintaining ade- the year of household surveys and so-called “line-up” year esti- quate data coverage. Instead, we choose the latest available mates that allow for comparisons of poverty estimates across survey for each available country in the I2D2, with a cut-off countries in reference years. The current note exclusively uses year of 2000. Thus all the global poverty profiles reported in line-up year estimates as they are the World Bank’s official in- this note should be interpreted as characteristics of the poor ternational poverty estimates. PovcalNet calculates line-up during the 2000s. We use data from as many as 73 countries, year estimates every three years since 1981, with the only ex- although the number of countries included differs across pro- ception being the latest 2010 estimates, two years after the filing variables. previous estimate in 2008. For more details about how line- Household consumption aggregates in the I2D2 are tak- up year estimates are calculated, see Methodology and FAQs en directly from the World Bank regional teams compiling sections in PovcalNet and references therein. and standardizing household surveys in their respective re- gions. It is not constructed by the I2D2 team from the origi- World Development Indicators nal data. Therefore, there are likely to be discrepancies be- Data on population and GDP are obtained from WDI (http:// tween the per capita consumption aggregate in the I2D2 and data.worldbank.org/data-catalog/world-development-indica- the ones in the PovcalNet. For this reason, in each survey data tors) using wbopendata command in Stata (http://data.world- in the I2D2 used in this exercise, we apply the 2010 poverty bank.org/ developers/apps/wbopendata). Note that we use rate from PovcalNet to infer the appropriate poverty line and the latest population and GDP data available in WDI as of the identify poor and nonpoor individuals. time of writing (October 1st, 2013), which results in small To the best of our knowledge, this is the first attempt to discrepancies between our estimates and those in PovcalNet report poverty profiles at the global scale. An exercise such as in terms of poor population and total population. this has not been possible until recently without a globally The International Income Distribution Database (I2D2) harmonized dataset such as the I2D2. Even the I2D2 is not The primary data source for the profiles of the poor is the free from data limitations described above and those widely I2D2, a globally harmonized database drawn from more than documented in the poverty literature, including but not lim- 600 nationally representative household surveys. Poverty pro- ited to the issue of comparability due to different survey de- filing is commonly conducted at the country level, where signs across time and locations and that of Purchasing Power country-specific definition of poverty is used. Our approach, Parity Index used to convert local currencies to the interna- which we call “Global Poverty Profiling,” is unique in that we tional one. While we are aware of these limitations, we believe focus on profiles of the poor defined by the international pov- it is of great importance to make the most of available data erty line of $1.25 a day, thereby allowing us to compare socio- and produce the best available estimates of the state of the economic and geographical characteristics of the extremely poor as part of our renewed mission to end extreme poverty poor across countries. by 2030. The Economic Premise note series is intended to summarize good practices and key policy findings on topics related to economic policy. They are produced by the Poverty Reduction and Economic Management (PREM) Network Vice-Presidency of the World Bank. The views expressed here are those of the authors and do not necessarily reflect those of the World Bank. The notes are available at: www.worldbank.org/economicpremise. 8 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise