Global Poverty Monitoring Technical Note 20 April 2022 Update to the Poverty and Inequality Platform (PIP) What’s New R. Andres Castaneda Aguilar, Reno Dewina, Carolina Diaz-Bonilla, Ifeanyi N. Edochie, Tony H. M. J. Fujs, Dean Jolliffe, Jonathan Lain, Christoph Lakner, Gabriel Lara Ibarra, Daniel G. Mahler, Moritz Meyer, Jose Montes, Laura L. Moreno Herrera, Rose Mungai, David Newhouse, Minh C. Nguyen, Diana Sanchez Castro, Marta Schoch, Liliana D. Sousa, Samuel K. Tetteh-Baah, Ikuko Uochi, Martha C. Viveros Mendoza, Haoyu Wu, Nishant Yonzan and Nobuo Yoshida April 2022 Keywords: What’s New; April 2022. Development Data Group Development Research Group Poverty and Equity Global Practice Group GLOBAL POVERTY MONITORING TECHNICAL NOTE 20 Abstract The April 2022 update to the newly launched Poverty and Inequality Platform (PIP) involves several changes to the data underlying the global poverty estimates. Some welfare aggregates have been changed for improved harmonization, and the CPI, national accounts, and population input data have been updated. This document explains these changes in detail and the reasoning behind them. Moreover, a large number of new country-years have been added, bringing the total number of surveys to more than 2,000. These include new harmonized surveys for countries in West Africa, new imputed poverty estimates for Nigeria, and recent 2020 household survey data for several countries. Global poverty estimates are now reported up to 2018 and earlier years have been revised. All authors are with the World Bank. Corresponding authors: Christoph Lakner (clakner@worldbank.org) and Minh C. Nguyen (mnguyen3@worldbank.org). The authors are thankful for comments and guidance received from Deon Filmer, Haishan Fu, and Carolina Sánchez-Páramo. We would also like to thank the countless Poverty Economists that have provided data and documentation, and patiently answered our questions. Without them the database of household surveys that underpins the World Bank’s global poverty measures would not exist. This note has been cleared by Umar Serajuddin. The Global Poverty Monitoring Technical Note Series publishes short papers that document methodological aspects of the World Bank’s global poverty estimates. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Global Poverty Monitoring Technical Notes are available at http://iresearch.worldbank.org/PovcalNet/. Contents 1 Introduction .......................................................................................................................................................... 2 2 2020 surveys for countries in Latin America and the Caribbean .........................................................................6 2.1 Argentina ....................................................................................................................................................6 2.2 Bolivia ........................................................................................................................................................ 7 2.3 Brazil .......................................................................................................................................................... 7 2.4 Chile ........................................................................................................................................................... 8 2.5 Colombia ....................................................................................................................................................8 2.6 Costa Rica ...................................................................................................................................................8 2.7 Ecuador ....................................................................................................................................................... 9 2.8 Mexico ........................................................................................................................................................ 9 2.9 Peru ............................................................................................................................................................. 9 2.10 Uruguay ......................................................................................................................................................9 3 New surveys for countries in West Africa ...........................................................................................................9 4 Revision to the recent poverty trend in Nigeria.................................................................................................. 11 5 Changes to welfare aggregates ........................................................................................................................... 14 5.1 Albania 2015-2016 ................................................................................................................................... 14 5.2 Argentina 2017-2019 ................................................................................................................................ 14 5.3 Bolivia 2019 ............................................................................................................................................. 14 5.4 Brazil 2012-2019 ...................................................................................................................................... 15 5.5 Chile 2006, 2017 ...................................................................................................................................... 15 5.6 Costa Rica 1989, 2004-2008, 2010-2019 ................................................................................................. 16 5.7 Ecuador 2012 ............................................................................................................................................ 17 5.8 Mexico 2016 and 2018 ............................................................................................................................. 17 5.9 Pakistan 2001-2018 .................................................................................................................................. 18 5.10 Paraguay 2010-2011 ................................................................................................................................. 19 5.11 Peru 2001-2019 ........................................................................................................................................ 20 5.12 Sao Tome and Principe 2017 .................................................................................................................... 21 5.13 Vanuatu 2010............................................................................................................................................ 21 5.14 EU-SILC ................................................................................................................................................... 22 5.15 LIS ............................................................................................................................................................ 22 6 Changes to survey years ..................................................................................................................................... 23 7 Changes to data coverage type ........................................................................................................................... 23 8 Changes to CPI data ........................................................................................................................................... 23 8.1 China 2017-2019 ...................................................................................................................................... 23 9 Changes to National Accounts Data ................................................................................................................... 24 10 Changes to Population Data ............................................................................................................................... 25 11 Comparability database ...................................................................................................................................... 25 12 Economy-years added/removed ......................................................................................................................... 26 12.1 Economy-years removed .......................................................................................................................... 26 12.1.1 Nigeria 2009/10 HNLSS ................................................................................................................. 26 12.1.2 South Africa 1996 ........................................................................................................................... 27 12.2 Economy-years added............................................................................................................................... 27 13 References .......................................................................................................................................................... 31 14 Appendix 1 – CPI Data sources ......................................................................................................................... 33 15 Appendix 2 – National Accounts Data Sources ................................................................................................. 47 1 1 Introduction The April 2022 global poverty update from the World Bank presents new global poverty estimates for the reference year 2018. It revises the previously published global estimates for the period between 1981 to 2017, as well as the regional estimates from 1981 to 2019. The update includes new surveys that have been received and processed, as well as several changes to the existing data. Some changes reflect improvements in the welfare aggregate based on new harmonization efforts and more available information. This document outlines the changes made to the underlying data by country and explains the reasons why the changes have been made. Table 1 shows revisions to the 2017 regional and global poverty estimates.1 The global poverty headcount at the US$1.90 poverty line in 2017 is revised from 9.3 to 9.1 percent, resulting in a revision in the number of poor from 696 to 685 million. This is mostly driven by downward revisions to the Sub-Saharan Africa estimate, which has been revised from 41.2 percent to 40.0 percent. This change is largely explained by changes to the Western and Central Africa poverty estimate, decreasing from 36.7 to 33.2 percent. This revision is driven by changes to the line-up estimates due to newly available surveys for ten countries in West Africa (see section 3 below), as well as to changes to the poverty trend for Nigeria (see section 4 below). Table 1. Poverty estimates for reference year 2017, changes between June 2021 and April 2022 vintage at different poverty lines $1.90 $3.20 $5.50 Survey Headcount Number of Headcount Number of Headcount Number of Coverage ratio (%) poor (mil) ratio (%) poor (mil) ratio (%) poor (mil) Region (%) Jun Apr Jun Apr Jun Apr Jun Apr Jun Apr Jun Apr 21 22 21 22 21 22 21 22 21 22 21 22 East Asia and 1.4 1.5 29 30 8.4 9.0 174 186 27.6 29.3 571 608 97.6 Pacific Europe & Central 1.3 1.3 6 6 4.6 4.7 23 23 12.6 12.6 62 62 89.5 Asia Latin America & 3.8 4.1 24 26 9.3 10.1 59 63 23.0 24.1 145 152 90.2 Caribbean Middle East & 6.3 6.4 24 24 18.3 18.4 70 70 43.1 43.0 164 164 58.2 North Africa Rest of the World 82.4 0.7 0.7 7 8 0.9 0.9 10 10 1.3 1.3 14 14 South Asia 21.8 n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a Sub-Saharan Africa 87.4 41.2 40.0 433 420 67.3 66.4 707 697 86.2 85.7 906 900 East & Southern 80.6 44.2 44.6 277 279 68.8 69.4 431 435 86.4 86.8 541 544 Africa Western & 97.4 36.7 33.2 156 141 65.1 61.9 276 262 85.9 84.1 364 356 Central Africa World Total 72.7 9.3 9.1 696 685 24.3 24.2 1821 1823 43.5 44.0 3269 3308 Note: Regional poverty estimates are reported if survey coverage is above 50% within a three-years window of the reference year. The global estimate is reported if survey coverage is above 50% and coverage for low- and lower-middle-income countries is above 50%. Sub-Saharan Africa is further divided into East and West Africa, following the World Bank’s regional definition. The number of poor for Sub-Saharan Africa is equal to the sum of the number of poor for East and West Africa. 1The data available at the time of the June 2021 and the April 2022 updates do not offer sufficient population coverage in 2017 for South Asia, so we are unable to publish regional poverty estimates for this region. Survey coverage is assessed within a three-year window either side of 2017, i.e., including surveys that were conducted between 2014 and 2020 (also see Castaneda et al., 2020). The estimate for South Asia is not displayed since this region has a survey coverage less than 50%. 2 This update also revises previously published estimates for 2018 and 2019 (see Table 2 and Table 3). Moreover, improvements in data coverage, due to newly available surveys included with this update, allow us to add new global poverty estimates for 2018. This is due to an improvement in data coverage for low- and lower-middle- income economies. For 2018, the survey coverage for this group of countries is 50.7 percent, while it was only 48.2 percent in June 2021 – below the 50 percent threshold required to report global poverty numbers (Arayavechkit et al., 2021). In 2018, the global poverty headcount rate at the US$1.90 poverty line is 8.6 percent. This marks a decrease of 0.5 percent relative to 2017, corresponding to 28 million fewer poor people. This confirms a continued reduction in extreme poverty at the global level, although at a slower pace in more recent years. In fact, global poverty fell by 2.8 percentage points between 2012 and 2015 (from 12.8 percent to 10.0 percent), and by 1.5 percentage points between 2015 and 2018, confirming the findings published in World Bank (2020). Revisions to the 2018 poverty estimates show a slight increase in the poverty headcount ratio in Latin America and the Caribbean. This is driven by several changes to the household welfare measures used for global poverty measurement (as documented in section 5), and more specifically to revisions in Mexico’s welfare aggregate (see section 5.8). These revisions also affect the regional estimate for 2019, showing an increase in the poverty headcount rates at all values of the poverty line (see Table 3). At higher values of the poverty line, we also observe an upward revision to the estimates for East Asia and the Pacific. In 2018, the poverty headcount rate at the $5.50 poverty line has been revised from 24.7 to 26.4 percent, equivalent to 38 more million poor. The same estimate for 2019 also shows an upward revision but smaller and of about 1 percentage point. This is due to changes to the line-up poverty estimates in China, for which new surveys have been added with this update (see section 12.2). In contrast, estimates for Sub-Saharan Africa show a downward revision in 2018, confirming the pattern in 2017 described above. The poverty rate is 38.9 percent in 2018 (down from 40.4 percent in the June 2021 vintage), decreasing further to 38.3 percent in 2019. At the same time, the number of poor in the region has increased from 420 to 424 between 2018 and 2019, since population growth is outpacing poverty reduction. While improvements in data coverage for countries in Western and Central Africa allows us to report a poverty estimate for Sub-Saharan Africa in 2019, data coverage is insufficient to report the estimate for East and Southern Africa in the same year. Compared to 2017, poverty estimates in 2018 show a decrease in poverty in all regions except for the Middle East and North Africa. The lined-up poverty estimate for the region shows a further increase to 7.1 in 2018. This is largely driven by our estimates for conflict-affected economies. It 3 should be noted that for these countries, survey data is not available in recent years, giving rise to considerable uncertainty. This speaks to a broader issue of data availability in the region which limits our understanding of poverty in 2019. To conclude, the regional or global poverty estimates included in this update stop before the start of the COVID-19 pandemic. However, several country-level estimates for 2020 are available (see Table 19Table 12). Many of these are for countries in Latin America and the Caribbean, but some surveys conducted in 2020 are also available for countries in East Asia and Pacific, e.g., Indonesia, and for several European countries (see Table 19). These surveys show that the pandemic has not only had an impact on poverty, but also on data collection methods and on the methodology used to construct the household welfare aggregate needed for poverty measurement. This document discusses in detail how these changes affect 2020 poverty estimates (see section 2) and their comparability with those from previous surveys (see section 11). 4 Table 2 Poverty estimates for reference year 2018, changes between June 2021 and April 2022 vintage for selected regions and different poverty lines $1.90 $3.20 $5.50 Survey Headcount Number of Headcount Number of Headcount Number of Coverage ratio (%) poor (mil) ratio (%) poor (mil) ratio (%) poor (mil) Region (%) Jun Apr Jun Apr Jun Apr Jun Apr Jun Apr Jun Apr 21 22 21 22 21 22 21 22 21 22 21 22 East Asia and 97.5 1.2 1.2 25 25 7.1 7.4 148 154 24.7 26.4 514 552 Pacific Europe and 89.4 1.1 1.0 5 5 4.2 4.0 20 20 11.9 11.7 58 58 Central Asia Latin America 86.7 3.7 4.0 23 25 9.2 9.9 58 63 22.5 23.7 143 151 & Caribbean Middle East and 50.9 7.0 7.1 27 27 19.9 19.7 77 76 44.4 44.0 172 170 North Africa Rest of the 82.4 0.6 0.6 7 7 0.8 0.8 9 9 1.3 1.3 14 14 World South Asia 21.9 n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a Sub-Saharan 72.6 40.4 38.9 436 420 66.7 65.4 719 705 86.1 85.3 928 920 Africa East & Southern 60.4 43.7 44.0 281 283 68.2 68.8 439 442 86.4 86.7 556 558 Africa Western & 90.6 35.5 31.4 155 137 64.3 60.4 280 263 85.5 83.1 372 362 Central Africa World Total 69.9 n/a 8.6 n/a 656 n/a 23.2 n/a 1760 n/a 42.9 n/a 3259 Note: Survey coverage for low- and lower-middle-income countries for 2018 is: 50.7%. Table 3 Poverty estimates for reference year 2019, changes between June 2021 and April 2022 vintage for selected regions and different poverty lines $1.90 $3.20 $5.50 Survey Headcount Number of Headcount Number of Headcount Number of Coverage ratio (%) poor (mil) ratio (%) poor (mil) ratio (%) poor (mil) Region (%) Jun Apr Jun Apr Jun Apr Jun Apr Jun Apr Jun Apr 21 22 21 22 21 22 21 22 21 22 21 22 East Asia and 95.9 1.0 0.9 20 18 6.3 5.9 131 125 22.6 23.7 473 499 Pacific Europe and 87.4 1.0 1.1 5 5 4.0 4.1 20 20 11.6 11.5 58 57 Central Asia Latin America 86.8 3.7 4.1 24 26 9.2 9.9 59 64 22.5 23.6 145 151 & Caribbean Middle East and 47.9 0.0 0.0 0 0 0.0 0.0 0 0 0.0 0.0 0 0 North Africa Rest of the 82.5 0.6 0.6 7 7 0.8 0.7 9 8 1.3 1.2 14 13 World South Asia 21.9 n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a Sub-Saharan 55.4 n/a 38.3 n/a 424 n/a 64.7 n/a 716 n/a 85.0 n/a 941 Africa East & Southern 32.0 n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a Africa Western & 90.0 34.5 30.5 154 136 63.6 59.4 284 266 85.2 82.4 381 368 Central Africa World Total 66.7 n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a Note: Survey coverage for low- and lower-middle-income countries for 2019 is: 41.0%. 5 2 2020 surveys for countries in Latin America and the Caribbean The COVID-19 pandemic affected both the data collection methodology and, in some cases, the survey questionnaire for the countries in Latin America and the Caribbean with data in 2020. This section summarizes the main methodological changes to the construction of the welfare aggregates. Figure 1 summarizes the main changes and the remainder of this section discusses changes to the welfare aggregate specific to each country. Figure 1 Methodological changes to household surveys in Latin America and the Caribbean, 2019/2020 Note: F2F refers to “face-to-face” data collection. 2.1 Argentina An evaluation of the comparability of data gathered during 2020 by the Argentinian National Statistics Institute (INDEC) is ongoing. After completing their evaluation, INDEC will issue further recommendations on the comparability with previous estimates. In the second quarter of 2020, the survey was collected through phone interviews. The survey is designed as a rotating panel; 25% of the sample constitutes “new” households that interviewed for the first time each quarter. There was no previous contact for those households, so INDEC did not have their phone numbers, which was an obstacle to interviewing most of them. INDEC solved this problem of non-response using a propensity score model to reweight the sample, based on the probability of non-response according to household characteristics (for details, see p. 11 in INDEC, 2020). 6 2.2 Bolivia Users of the poverty and inequality estimates for Bolivia 2020 should be mindful of the period in which the surveys were collected, especially when analyzing the impacts of the COVID-19 pandemic. In the case of Bolivia, the survey was collected at the end of the year. The modules of the survey questionnaire were reduced. In particular, the dwelling module was shorter, and did not contain the homeownership status or housing rent amount. These are the variables that are used by the Socioeconomic Database for Latin America and the Caribbean (SEDLAC) for imputing the rent of owner-occupiers.2 To make 2020 comparable with the historical series, the team developed and tested an imputation model that imputes the structure of the property and the distribution of implicit rent of the own housing of 2019 into the 2020 survey. This methodology is already being used for imputing expected rent throughout the income distribution in the Brazil PNADC 2012-2015 (for more details, see Atamanov et al., 2020). 2.3 Brazil The PNADC data is the main source of information for poverty monitoring in Brazil. The latest annual release included the 2020 data, published by the National Statistics Office (IBGE) in November 2021 and included in this update. The PNADC 2020 was collected throughout 2020, closely following the data collection methodology adopted in previous years. However, due to the health restrictions caused by the COVID-19 pandemic, IBGE had to adapt its data collection strategy from face-to-face to phone interviews. While the questionnaire itself was not changed, changes in the data gathering methodology affected non-response rates. The PNADC follows a rotating panel design. There are five groups of households in the sample, each of which is interviewed five times and using a different questionnaire each time. Interviews are conducted throughout the year. Since 2012, the World Bank has used the first interview. However, for 2020, the fifth interview is used instead because the IBGE has published the social indicators and microdata for the fifth interview only. For more details, see Annex 1 of Lara Ibarra and Vale (2022). Both the first and fifth interviews are conducted throughout the year. The main difference between the first and fifth interviews is that the latter does not include the dwelling module; thus, it does not contain the homeownership status or housing rent amount (variables needed for imputing the rent of owner-occupiers for the SEDLAC datasets). To make 2020 comparable with the historical series, housing ownership and implicit rent is imputed using the 2019 distribution. This is analogous to the approach used in Brazil during 2012-2015 and Bolivia in 2020 (see description on 2020 Bolivia data). 2 SEDLAC is a project by CEDLAS (Universidad de La Plata) and The World Bank, that harmonizes household surveys from that region in an effort to improve comparability. 7 In 2020, there is also evidence of significant under-coverage of the Auxilio Emergencial (AE) program in the survey. AE is a cash-transfer program which was introduced in 2020. According to administrative data, there were over 68 million AE recipients. In the survey, only about 20 million are observed. AE provided monthly transfers that could add up to a maximum of R$4,200 during 2020. For a household in the bottom quintile in 2019, this is equivalent to a 50% increase in their income per capita. Given the magnitude of this program (Lara Ibarra and Vale, 2022) and to better capture the evolution of income and poverty in Brazil, the team has imputed AE beneficiary status in the data. This allowed to complement the observed AE status as reported in the survey.3 Incorporating eligibility criteria from the AE (demographic, employment, and income), the method reached a combined AE population of 42.2 million individuals – leading to a number of program recipients more in line with the administrative records. The imputed AE status is used to construct the household annual income aggregate that underlies the poverty and inequality estimates for Brazil 2020 included in this update. Additional details on the imputation exercise and on the comparability of these estimates with previous rounds of the PNADC are discussed in Lara Ibarra and Vale (2022). 2.4 Chile Changes in the data collection methodology limits comparability between CASEN 2020 (using phone interviews) and previous survey rounds (using face-to-face). Compared to face-to-face interviews, phone surveys present additional challenges that could bias the estimates. While the Ministerio de Desarrollo Social y Familia de Chile (MDSF) adopted several strategies to minimize such bias, it is unlikely that the income and poverty measures are fully comparable between 2020 and the historical series. Therefore, caution must be taken when comparing with previous years. 2.5 Colombia In 2020, the data collection was split in two parts (both parts are included in the 2020 data used in PIP). The first part was collected through phone interviews from March until July. The questionnaire was reduced due to the COVID-19 pandemic. Consequently, the Colombian National Statistics Institute (DANE) imputed social programs using administrative data to identify program recipients and allocate transfers. Between August and December, the second part of the survey was collected using phone surveys and face-to-face visits in rural areas. In terms of household income, the first part of the survey only collected wages and net earnings (65% of aggregate household income). The second part of the survey covered income from the secondary activity, in-kind income, interest, dividends, pensions, retirements, and others. 2.6 Costa Rica Due to COVID-19, data collection changed from face-to-face to phone: 45.17% of households in the sample were visited in person, while 54.83% were contacted by telephone. The survey weights 3 It should be noted that the survey did not explicitly ask for the receipt of AE. The survey only contains a generic question about other transfers. AE recipient status can be inferred from the amounts reported. 8 were calibrated using logistic regressions to minimize bias and ensure that the results for 2020 are comparable with those of 2019 (for details, see p. 27 in INEC, 2020). 2.7 Ecuador This survey was collected in December through face-to-face interviews, and is nationally, urban, and rural representative. Before 2020, the survey was representative at the regional level. 2.8 Mexico The main effects of the pandemic on poverty and inequality were experienced during the second quarter of 2020, when the lockdown measures led to a significant amount of job losses. The ENIGH collects information between August and November. Therefore, the previously used harmonization approach (see section 5.8 below), which constructed the welfare aggregate by using information about the income received in the last month, may not correctly capture the pandemic’s effect on welfare. The new methodology constructs the welfare aggregate by using the average of the income from the last six months, which includes data from the second and third quarters of 2020 and, to some extent, reflect more accurately what happened to poverty and inequality in Mexico during 2020. Earlier years were also revised (see section 5.7 below). 2.9 Peru From mid-March to end-September 2020, the ENAHO was collected through phone interviews and using a reduced questionnaire. In 2020, Peru had extensive cash transfers programs, covering a large part of the population due to the COVID-19 shock. However, there is evidence of significant under-coverage of the recipients of these transfers in the survey. The National Statistical Office (INEI) carried out an imputation exercise using administrative records. In the data used here, a portion of the second wave of one of the cash transfers (Bono Universal), which was erroneously excluded, is imputed in addition. 2.10 Uruguay Due to the COVID-19 restrictions, the 2020 ECH data used new methodologies for data collection and sampling (for more details, see INE, 2020). The data collection was a mix of face-to-face and remote (e.g., phone) interviews. For the remote interviews, a shorter questionnaire and a revised sampling method was used. Instead of the standard cross-sectional sample, a panel-type sampling was used. The results are therefore not strictly comparable with previous versions of the survey. 3 New surveys for countries in West Africa This update includes the first round of household surveys supported by the Harmonized Surveys on Household Living Conditions (PEHCVM) program. This program was launched in 2016 by the World Bank and the WAEMU Commission with the goal of improving the welfare measurement in countries in the West African Economic and Monetary Union (WAEMU). The program 9 provided extensive technical assistance and support to National Statistical Agencies of member countries to implement new household survey instruments and methods to improve poverty measurement, following current international best practices and improving comparability among these countries. The first round of data was collected in 2018/19 (see Table 4). The main feature of the new surveys is that the welfare aggregate is estimated consistently for all WAEMU countries. Some of the main changes to the design of the consumption questionnaire are as follows: • 7-day recall period for food consumption questionnaire: this harmonizes different practices in previous surveys, ranging from diary to 7-days, 30-days or 12-months recall periods. • Imputed rent included in welfare aggregates. This was not the case for any of the previous surveys for these countries. • Expenses in Education and Health recorded at individual levels. Previous surveys collected this information at the household level. While these changes represent improvements in the measurement of household welfare and improve comparability across countries, they also make the new estimates incomparable to previous surveys available for these countries. The new surveys also in many cases produce notably lower poverty rates compared to previous estimates. For example, with the new survey, the poverty rate in Senegal in 2018 is estimated at 7.6% (Table 4). This compares with 38.5% in the previous 2011 survey, which when extrapolated had produced a poverty estimate of 28.8% in 2018. Further details on the comparability issues can be found in the country notes which are forthcoming in the Global Poverty Monitoring Technical Notes series. Table 4 New WAEMU surveys poverty and inequality estimates Economy Year Poverty Poverty headcount rate headcount rate $1.90 (%) $3.20 (%) Benin 2018 19.16 51.26 Burkina Faso 2018 33.66 61.81 Cote d’Ivoire 2018 9.18 34.87 Guinea 2018 23.18 60.39 Guinea Bissau 2018 24.67 59.50 Mali 2018 16.33 49.50 Niger 2018 41.35 75.20 Senegal 2018 7.60 34.03 Chad 2018 33.19 66.44 Togo 2018 24.08 51.83 10 4 Revision to the recent poverty trend in Nigeria This update includes imputed data from three new surveys for Nigeria for 2010/11, 2012/13, and 2015/16. These poverty estimates are the result of survey-to-survey imputations using data from the Generalized Household Survey (GHS) and the 2018/19 Nigerian Living Standards Survey (NLSS) (Lain et al., 2022). The analysis uses consumption data from the 2018/19 NLSS and a set of comparable household characteristics available in both the 2018/19 NLSS and the GHS to impute consumption into three years for which GHS data are available, namely 2010/11, 2012/13, and 2015/16.4 The imputed household consumption data are prepared in three steps; first a set of comparable non-monetary indicators that are available in both the NLSS and GHS are selected. One advantage of using the NLSS and GHS is that many questions on household characteristics and key non- monetary indicators are the same and are asked using the same recall periods. Moreover, the comparability of the indicators can be verified using 2018/19 NLSS and 2018/19 GHS, for which the data collection schedules overlapped. Second, a consumption model using the selected variables and consumption data from 2018/19 NLSS is developed. Variables used in the consumption model include regional dummies, demographics of the household (dependency ratio), household head characteristics (gender, sector of employment category), living conditions (main source of cooking fuel, toilet availability), consumption frequency dummies (food and non-food items), asset ownership (air conditioning, washing machine, cars and other vehicles, generator, microwave, TV set, computer). The model therefore includes variables that capture short-run variation – such as employment and a set of dummy variables for whether certain food and non-food items were being consumed – as well as more stable household characteristics – such as location, demographics, housing amenities, and ownership of assets.5 This model relies on the assumption that these indicators are highly correlated with poverty and that this relationship remains constant across the GHS waves. 4 While household consumption data from the GHS survey are available, they do not meet the quality requirements needed for poverty measurement purposes. First, the GHS data were collected during two relatively short periods to align with the agricultural cycle, in the post-planting and post-harvesting seasons. In contrast, the NLSS consumption data were collected over a period of twelve months. This could also affect the imputed consumption estimate, although the indicators used in the consumption model are less subject to seasonality than household consumption and expenditure. Second, early rounds of the GHS imposed standard units (grams and kilograms) when recording consumption levels when non-standard units may have been more appropriate, potentially biasing direct estimates of the consumption aggregate. The NLSS, by contrast, allowed consumption to be recorded using non-standard units. Third, GHS consumption data are not representative at the state level and the GHS is not used by the Nigeria’s National Bureau of Statistics for official poverty measurement. See Lain et al. (2022) for further details. 5 The analysis adopts a “SWIFT Plus” (Survey of Wellbeing via Instant and Frequent Tracking) approach to conduct the survey-to-survey imputation (Yoshida et al. 2015, 2020). This approach was developed to overcome possible limitations arising in case a large shock or crisis, such as the 2015/16 oil price plunge in Nigeria, occurred between the baseline and target surveys. It consists of including variables that reflect households’ current welfare status, such as employment, in the consumption model used in the imputations. 11 Specifically, the following regression is estimated: ℎ, = ,ℎ, + ℎ, + + ℎ, where ℎ, is the natural logarithm of annual spatially adjusted household consumption expressed in local currency units for household h in time t. ,ℎ, is a vector of household head’s characteristics, are geographical zone-areas dummy variables. The error term is drawn from a normal distribution. Third, using the parameters estimated from this consumption model, consumption is imputed into the GHS and the corresponding poverty rates are calculated. The imputed consumption vector is estimated using 100 imputations. The imputed consumption vector is then converted to US$ 2011 PPP to calculate poverty rates at different international poverty lines. The same exercise is conducted for the GHSs from 2010/11, 2012/13, and 2015/16. The basic results from applying this approach are shown in Table 5. Table 5 Poverty and inequality estimates using survey-to-survey imputations. Poverty rate US$1.90 (%) Poverty rate US$3.20 (%) Gini index 2010/11 GHS 43.54 72.88 35.65 2012/13 GHS 42.49 72.12 35.51 2015/16 GHS 40.75 70.44 35.88 Source: Lain et al. 2022. Note: the table shows the results of the survey-to-survey imputation in each round of the GHS. We develop a consumption model using data on 23 non-monetary indicators available in the 2018/19 NLSS. Using the estimated parameters, we then impute consumption in each round of the GHS. The imputed consumption vector is then converted to 2011 PPPs and poverty estimates are reported at the US$1.90 and US$3.20 poverty lines. Gini coefficients are calculated as the average of 100 estimates resulting from each imputation. The results of the survey-to-survey imputations are validated in three main ways. First, using the same model to impute into the 2018/19 GHS produces a poverty headcount rate that is very close to the direct, official poverty estimate from the 2018/19 NLSS (the difference is less than 3 percentage points). Second, the imputation results are similar to the results of a “backcasting” exercise, which maps macro-data on sectoral GDP growth to micro-data on consumption in the 2018/19 NLSS – through the household head’s employment sector – to estimate consumption in previous years (see Lain et al. 2022 for further details). Despite the two methods having a totally different set of underlying assumptions, they both suggest that poverty reduction in Nigeria stagnated in the 2010s. Third, many non-monetary welfare indicators, including on education and basic infrastructure, from Nigeria’s Demographic and Health Survey (DHS) also showed little improvement in the 2010s. The fact that both survey-to-survey imputations and backcasts suggest that poverty reduction stalled in the 2010s adds further credence to concerns over the quality of the household consumption data from the 2009/10 Harmonised Nigeria Living Standards Survey (HNLSS), which has been removed with this update (see section 12.1.1 below). 12 The interpolated trend in PIP reflects this change and includes the new imputed data from 2010/11, 2012/13, and 2015/16 alongside the poverty estimates from the 2003/04 NLSS and the 2018/19 NLSS (see Figure 2). This results in a substantive revision of the interpolated poverty trend for Nigeria: showing a stagnation in poverty reduction between 2009 and 2019 instead of a 17-percentage points decline implied by the interpolated series previously used for global poverty measurement purposes. It should be noted, however, that the new imputed data from 2010/11, 2012/13, and 2015/16 are not technically comparable with the 2003/04 NLSS nor the latest 2018/19 NLSS estimate available in PIP (see Section 11 on the comparability database for further information). Figure 2 Revisions to recent poverty trend for Nigeria, 2003-2019 60.0% 55.0% 55.9% 50.0% 45.0% 43.5% 40.0% 42.5% 40.7% 39.1% 35.0% 30.0% 2003 2007 2011 2015 2019 Nigeria line-up (Jun 2021 vintage) Nigeria line-up (Apr 2022 vintage) Nigeria survey estimates (June 2021 vintage) Nigeria survey estimates (Apr 2022 vintage) Source: PIP, World Bank 13 5 Changes to welfare aggregates 5.1 Albania 2015-2016 The consumption aggregate for Albania has been corrected. Some consumption categories that are reported weekly were incorrectly assumed to be monthly. Some errors were also due to confusion between expenditures being recorded in old or new Albanian lek, or in monthly or annual terms.6 Finally, only 75% of education expenditures were previously included (since the school year is for 9 months), while the entire education spending is now included. These changes result in higher consumption levels and thus lower poverty than previously estimated and reported (see Table 6). Table 6 Changes to poverty and inequality estimates, Albania HBS 2015-2017 Poverty headcount Poverty headcount Poverty headcount Gini Index $1.90 $3.20 $5.50 Country Year Jun Apr Jun Apr Jun 2021 Apr Jun Apr 2021 2022 2021 2022 2022 2021 2022 Albania 2015 1.13 0.25 8.20 4.69 34.3 24.46 32.91 32.75 Albania 2016 0.86 0.41 9.42 5.46 34.2 23.90 33.71 33.74 Albania 2017 1.27 0.43 8.19 4.28 33.76 23.83 33.17 33.06 5.2 Argentina 2017-2019 The temporal deflator used within each survey round was revised for these years. As the survey is semi-annual, it is necessary to deflate the income received by individuals interviewed in different quarters of the year. In 2017-2019 s2 (second semester) data, this deflation had erroneously used the s1 (first semester) deflator. Table 7 Changes to poverty and inequality estimates, Argentina 2017-2019 Poverty headcount Poverty headcount Poverty headcount Gini Index $1.90 $3.20 $5.50 Country Year Jun 2021 Apr Jun 2021 Apr Jun 2021 Apr Jun 2021 Apr 2022 2022 2022 2022 Argentina 2017 0.93 0.92 3.18 3.10 9.88 9.82 41.15 41.13 Argentina 2018 1.35 1.37 4.00 4.02 12.34 12.38 41.33 41.34 Argentina 2019 1.46 1.30 4.88 4.77 14.40 14.49 42.90 42.91 5.3 Bolivia 2019 A slight revision comes from changes to the temporal deflator used within the survey. It now uses greater precision (a larger number of decimal points). Changes to the poverty and inequality estimates are only detectable at the fourth decimal. 6 The conversion from old to new lek is 10:1. 14 Table 8 Changes to poverty and inequality estimates, Bolivia 2019 Poverty headcount Poverty headcount Poverty headcount Gini Index $1.90 $3.20 $5.50 Country Year Jun 2021 Apr Jun 2021 Apr Jun 2021 Apr Jun 2021 Apr 2022 2022 2022 2022 Bolivia 2019 3.24 3.24 7.82 7.82 19.90 19.90 41.65 41.65 5.4 Brazil 2012-2019 IBGE has updated the sample weights of the PNADC 2012-2019 to reflect better the gender and age group composition of the population. This update was done using the 2010 Census data. See IBGE's technical note 4 (2021). Table 9 Changes to poverty and inequality estimates, Brazil 2012-2019 Poverty headcount Poverty headcount Poverty headcount Gini Index $1.90 $3.20 $5.50 Country Year Jun 2021 Apr Jun 2021 Apr Jun 2021 Apr Jun 2021 Apr 2022 2022 2022 2022 Brazil 2012 3.73 3.89 8.97 9.29 21.08 21.74 53.55 53.44 Brazil 2013 3.08 3.22 7.87 8.11 19.28 19.94 52.75 52.69 Brazil 2014 2.73 2.88 6.99 7.34 17.60 18.36 52.11 52.03 Brazil 2015 3.15 3.37 7.72 8.06 18.63 19.50 51.94 51.93 Brazil 2016 3.87 4.11 8.78 9.26 19.99 21.07 53.28 53.34 Brazil 2017 4.42 4.66 9.07 9.50 20.23 21.16 53.27 53.33 Brazil 2018 4.42 4.66 9.13 9.56 19.82 20.78 53.87 53.87 Brazil 2019 4.61 4.86 9.12 9.55 19.62 20.59 53.43 53.49 5.5 Chile 2006, 2017 Changes have been made to the variable that captures the years of completed education. This change impacts the indicator which flags “coherent” income observations (SEDLAC variable cohh=1).7 Only coherent observations are included in the sample. Table 10 Changes to poverty and inequality estimates, Chile 2006, 2017 Poverty headcount Poverty headcount Poverty headcount Gini Index $1.90 $3.20 $5.50 Country Year Jun Apr Jun Apr Jun 2021 Apr 2022 Jun 2021 Apr 2022 2021 2022 2021 2022 Chile 2006 1.45 1.45 5.36 5.36 19.19 19.19 47.30 47.29 Chile 2017 0.29 0.29 0.68 0.68 3.57 3.57 44.44 44.44 7 In the SEDLAC harmonization, some observations are identified as incoherent. For example, this applies to individual observations that are identified as employed, but record no income in the main occupation. Households are identified as incoherent if the household head or the household member with the highest income or the household member with the highest level of education (if the head does not work) is identified as an incoherent observation. 15 5.6 Costa Rica 1989, 2004-2008, 2010-2019 Several modifications and corrections affected household income: • 1989: Underlying microdata for this year was replaced with the latest version. • 2004-2008, 2014-2019: Changes have been made to imputed rent. The implicit rent imputation model (which imputes the rental value of owner-occupied housing, dwellings received as a gift, usufruct or ceded dwellings) uses the variable “water source”, which was modified throughout the series to improve comparability between years. This impacts imputed rent and, therefore, overall household income. • 2010-2013: INEC released 2010-2013 datasets with an update of the weights. This update is being adopted in the harmonized data and will be used from now on. Table 11 Changes to poverty and inequality estimates, Costa Rica Poverty headcount Poverty headcount Poverty headcount Gini Index $1.90 $3.20 $5.50 Country Year Jun 2021 Apr 2022 Jun 2021 Apr 2022 Jun 2021 Apr 2022 Jun 2021 Apr 2022 Costa Rica 1989 12.33 10.43 22.46 20.44 43.41 41.61 46.69 45.57 Costa Rica 2004 4.42 4.39 9.34 9.36 23.14 23.20 48.35 48.42 Costa Rica 2005 3.22 3.20 7.98 8.02 20.99 20.96 47.47 47.51 Costa Rica 2006 3.19 3.18 7.82 7.89 20.53 20.54 49.44 49.29 Costa Rica 2007 1.76 1.76 5.38 5.37 16.33 16.32 49.29 49.31 Costa Rica 2008 2.26 2.34 5.35 5.39 15.90 15.90 48.64 48.65 Costa Rica 2010 1.58 1.47 4.20 3.87 13.24 12.59 48.16 48.05 Costa Rica 2011 1.74 1.65 4.41 4.29 13.70 13.35 48.72 48.76 Costa Rica 2012 1.69 1.63 4.27 4.08 12.66 12.19 48.62 48.42 Costa Rica 2013 1.62 1.59 4.17 4.03 12.76 12.41 49.27 49.15 Costa Rica 2014 1.49 1.49 4.00 4.00 12.34 12.34 48.63 48.63 Costa Rica 2015 1.54 1.54 4.07 4.07 11.96 11.95 48.38 48.38 Costa Rica 2016 1.27 1.27 3.78 3.78 10.78 10.78 48.71 48.71 Costa Rica 2017 1.06 1.06 2.81 2.81 9.91 9.91 48.35 48.35 Costa Rica 2018 1.46 1.46 3.66 3.66 11.17 11.17 47.97 47.97 Costa Rica 2019 1.01 1.01 3.24 3.25 10.62 10.62 48.19 48.19 16 5.7 Ecuador 2012 Changes have been made to the variable that captures the years of completed education. This impacts the “coherent” income indicator and thus the sample used in the analysis (for more details, see the section on Chile 2006, 2017 above). Table 12 Changes to poverty and inequality estimates, Ecuador 2012 Poverty headcount Poverty headcount Poverty headcount Gini Index $1.90 $3.20 $5.50 Country Year Jun 2021 Apr Jun 2021 Apr Jun 2021 Apr Jun 2021 Apr 2022 2022 2022 2022 Ecuador 2012 4.49 4.49 11.47 11.47 27.57 27.58 46.12 46.12 5.8 Mexico 2016 and 2018 SEDLAC’s harmonization of the income aggregate has changed for 2016 and 2018. The change in the household income aggregate results from considering additional information provided by the questionnaire. All the income questions (both labor and non-labor income) collected by the National Survey of Household Income and Expenditure (ENIGHNS) have information for different reference periods. The survey asks about the income received during each of the last six months. Figure 1 shows a snapshot of the 2016 survey questionnaire where the interviewer asks for the monetary value received from each income source (i.e., each of them in a separate row) during each of the last six months (i.e., each of them reported in a different column). The previous SEDLAC harmonization approach for Mexico used only the information provided by the first column, which corresponds to last month’s income, to create the welfare aggregate. The new harmonization approach exploits the data available from all the columns by computing a welfare aggregate as the average income received over the last six months (all columns). Figure 3 National Survey of Household Income and Expenditure (ENIGHNS) 2016 questionnaire The new harmonization approach has several advantages. First, it allows having a more accurate measure of the effect of the COVID-19 pandemic on poverty and inequality in 2020 (see section 2.8 above). Second, this measure of welfare is aligned with what the NSO reports. Third, measuring the income aggregate using the information from the last six months smooths out 17 seasonality in household income. For information on the comparability of the poverty estimates for Mexico see the comparability dataset discussed in section 11. Table 13 Changes to poverty and inequality estimates, Mexico 2016 and 2018 Poverty headcount Poverty headcount Poverty headcount Gini Index $1.90 $3.20 $5.50 Country Year Jun 2021 Apr 2022 Jun 2021 Apr 2022 Jun 2021 Apr 2022 Jun 2021 Apr 2022 Mexico 2016 2.16 3.23 7.77 10.54 25.44 29.36 46.28 47.68 Mexico 2018 1.73 2.65 6.48 9.24 22.70 26.59 45.38 46.71 5.9 Pakistan 2001-2018 The welfare aggregate for Pakistan has been revised due to the introduction of both a spatial and within-survey temporal deflator. The fieldwork to collect household surveys for Pakistan is typically from August to July (in line to the Pakistan fiscal year cycle). The national poverty measurement methodology (updated in 2015) measures welfare as per adult equivalent expenditure, deflated spatially and temporally within the survey. For the international poverty estimates, per capita expenditure has been used without a spatial deflator and without adjusting for inflation over the survey period. For example, for the survey conducted from August 2018 to June 2019, the average annual CPI of 2018 and 2019 was used to convert the nominal local currency units (LCU) to 2011 PPP values.8 Pakistan experienced high inflation during the second half of 2019 after the fieldwork was completed. This is reflected in the 2018-2019 CPI average, even though it lies outside the period of data collection and should therefore not affect the estimates. With this update, the Pakistan data used for international poverty estimates includes a temporal deflation within the survey period, as well as a spatial deflation. Specifically, and closely following the recommendations by Deaton and Zaidi (2002), the methodology can be described as follows: 1. A Paasche index is used to adjust the nominal consumption aggregate for cost-of-living differences faced by different households in different parts of the country. This approach is also used in the NSO’s aggregate. 2. Unit values obtained from the survey are used as the source of price data. We use median prices by PSU. The PSU contains households in close spatial proximity and interviewed less than a month apart. 3. The Paasche index is computed per PSU, using democratic weights, which give equal importance to every individual within each PSU, independent of household expenditure. This means that the weight of each item is derived as a population-weighted average of the household budget share. 8 This followed the method of using the weighted annual CPI, as described in Lakner et al. (2018). 18 4. The base period is set to January median prices for HIES 2004/05 to 2018/19 (and June for PIHS 2001), which corresponds to the middle of the survey period. Hence, in the case of the 2018/19 survey, the Paasche index transforms the nominal expenditures to real January 2019 prices. The monthly CPI is then used to transform to 2011 PPPs. The approximate Paasche price index at the PSU level can be written as: ln ≈ ∑ ln ( 0 ) =1 where is the PSU’s price for item i (median of the unit value by PSU); is PSU’s budget devoted to item I; 0 is the base price (mean of unit values of the PSUs mostly interviewed in January). Missing unit values are replaced using a hierarchical mechanism: if the unit value is missing for all households in the PSU (because the item was not purchased by any of the households in the PSU), the PSU is assigned the median value in the same quarter and stratum. If it is missing in the same quarter-stratum, it takes the median value in the same quarter and province/region. If missing in the same quarter-region, it takes the national median value in the same quarter. If the latter is unavailable, it takes the overall median value. The trend for the international poverty number remains broadly unchanged, but the level changes slightly. The largest change is observed in the 2018/19 poverty rate. Table 14 Changes to poverty and inequality estimates for Pakistan, 2001-2018   Poverty headcount Poverty headcount Poverty headcount Gini index $1.90 (%) $3.20 (%) $5.50 (%)     Jun 2021 Apr Jun 2021 Apr 2022 Jun 2021 Apr 2022 Jun 2021 Apr 2022 2022 2001 31.0 28.2 74.0 73.5 93.7 93.9 30.4 28.7 2004 19.9 16.7 62.3 60.6 88.4 88.3 32.5 30.9 2005 18.7 16.6 61.5 60.7 88.5 88.8 32.7 31.3 2007 15.0 11.9 57.3 55.3 86.8 87.0 31.6 29.7 2010 8.3 7.4 48.0 47.1 85.1 85.1 29.8 28.8 2011 8.0 7.2 46.6 46.5 83.6 83.9 30.9 29.7 2013 6.2 5.7 40.0 39.5 79.6 80.4 30.7 29.5 2015 4.0 3.8 35.5 35.5 75.9 76.4 32.6 31.3 2018 4.4 3.6 35.7 34.4 76.2 77.6 31.6 29.6 5.10 Paraguay 2010-2011 2010: Changes have been made to households with secondary household members: There was a small error in the definition of secondary household members, and this was corrected. This change affects the household size and hence per capita household income. 19 2011: Changes have been made to imputed rent: The rent imputation model uses the variables “Number of bedrooms in the household” and “dwelling’s construction materials”, modified throughout the series to improve comparability between years. Table 15 Changes to poverty and inequality estimates, Paraguay 2010-11 Poverty headcount Poverty headcount Poverty headcount Gini Index $1.90 $3.20 $5.50 Country Year Jun 2021 Apr Jun 2021 Apr 2022 Jun 2021 Apr 2022 Jun 2021 Apr 2022 2022 Paraguay 2010 5.19 5.19 12.67 12.67 28.75 28.75 50.96 50.98 Paraguay 2011 4.63 4.63 12.29 12.30 27.06 27.06 52.31 52.31 5.11 Peru 2001-2019 Two changes in the data affected household income: • Correction in the coding of the variable that captures non-labor income: Income from people who report it on an annual frequency is now included. This income was mistakenly not added to the non-labor income variable, thus affecting total household income. This has now been corrected. • Changes in the deflator used in the non-labor income variable: The data now uses the temporally and spatially deflated non-labor income as published by the Peruvian National Statistics Institute (INEI). Previously, the deflator for another type of income was used. Table 16 Changes to poverty and inequality estimates, Peru 2001-2019 Poverty headcount Poverty headcount Poverty headcount Gini Index $1.90 $3.20 $5.50 Country Year Jun Apr Jun Apr Jun Apr Jun Apr 2021 2022 2021 2022 2021 2022 2021 2022 Peru 2001 17.38 17.35 33.13 33.06 55.31 55.30 51.32 51.31 Peru 2002 15.16 15.08 29.68 29.67 50.71 50.64 53.59 53.57 Peru 2003 11.99 11.95 28.04 28.01 50.08 50.07 53.08 53.08 Peru 2004 13.60 13.55 28.63 28.58 50.00 49.97 49.89 49.88 Peru 2005 15.45 15.44 30.93 30.93 52.46 52.43 50.45 50.45 Peru 2006 13.46 13.45 26.98 26.94 46.64 46.62 50.33 50.33 Peru 2007 11.12 11.11 23.19 23.18 41.31 41.30 50.02 50.02 Peru 2008 9.09 9.07 19.57 19.55 37.59 37.58 47.47 47.49 Peru 2009 7.07 7.05 17.33 17.30 35.56 35.54 47.01 47.01 Peru 2010 5.54 5.50 14.92 14.90 31.52 31.51 45.55 45.54 Peru 2011 5.23 5.21 13.35 13.34 29.48 29.46 44.67 44.66 Peru 2012 4.81 4.80 12.34 12.33 26.90 26.90 44.45 44.44 Peru 2013 4.36 4.35 11.49 11.48 26.26 26.24 43.89 43.88 Peru 2014 3.73 3.72 10.71 10.70 25.34 25.29 43.15 43.14 Peru 2015 3.64 3.63 10.44 10.43 24.41 24.40 43.36 43.36 Peru 2016 3.53 3.52 10.12 10.09 24.57 24.53 43.65 43.64 Peru 2017 3.42 3.41 9.86 9.85 24.18 24.14 43.30 43.29 Peru 2018 2.69 2.69 8.43 8.45 22.31 22.33 42.37 42.39 Peru 2019 2.19 2.25 7.50 7.52 20.57 20.59 41.51 41.56 20 5.12 Sao Tome and Principe 2017 Data for Sao Tome and Principe are based on the Inquérito ao Orçamento Familiar (IOF). A 2021 quality assessment of the IOF 2017 and the welfare aggregate revealed high and uneven survey non-response rates (25 percent on average, higher in clusters with higher welfare) and unexpected patterns in consumption, attributable to high item non-response in consumption from own production.9 These issues motivated the use of a new methodology to calculate the consumption aggregate. Three changes were made. First, households that did not report any food or non-food consumption were dropped and treated as missing under the assumption that their information was too incomplete to recover. Second, survey weights were adjusted to correct for biased survey non- response rates using the methodology of Korinek and Ravallion (2005) to consider non-random missingness across sampling clusters. Third, a new welfare aggregate was constructed using the variable “total consumption” as the benchmark to estimate household consumption. In the initial estimates published in 2017, the declared purchase value was used as the basis for estimating welfare based on household expenditures. This was complemented with information on own- consumption and gifts, where available. This approach created problems due to high rates of item non-response for goods produced for own-consumption. As a result, the welfare aggregate based on expenditures was lower than the welfare suggested by the total consumption reported earlier in the survey. The new welfare aggregate also includes the periodization of large purchases (i.e. large purchases are spread out over time). These changes led to an increase in consumption for households that had underreported their consumption and consequently to a decrease in poverty and inequality estimates (see Table 17). Table 17 Changes in poverty and inequality estimates for Sao Tome and Principe Poverty headcount Poverty headcount Poverty headcount Gini index $1.90 (%) $3.20 (%) $5.50 (%) Year Jun 2021 Apr 2022 Jun 2021 Apr 2022 Jun 2021 Apr 2022 Jun 2021 Apr 2022 2017 35.64 25.56 65.36 56.97 86.36 82.85 56.32 40.75 5.13 Vanuatu 2010 An error with the sampling weight used in the previous version of the data has been corrected. For a comparison of the poverty headcount rates and Gini index, see the following table. 9 Specifically, food consumption as a share of consumption did not increase in welfare quintile, consumption from own production was a higher share in higher welfare quintiles, and consumption outside of the home was higher for the lowest quintiles. These anomalous patterns do not coincide with other country-specific information (such as the 2010 survey for Sao Tome and Principe) nor results in comparable countries. 21 Table 18 Changes in poverty and inequality estimates for Vanuatu 2010 Poverty headcount Poverty headcount Poverty headcount Gini index $1.90 (%) $3.20 (%) $5.50 (%) Year Jun 2021 Apr Jun 2021 Apr Jun 2021 Apr 2022 Jun 2021 Apr 2022 2022 2022 2010 13.14 13.14 39.45 39.39 72.37 72.43 37.63 37.35 5.14 EU-SILC All historical EU-SILC data have been updated to data released in December 2021. The updates for each country-year are documented on the Eurostat website [CIRCABC → Eurostat → EU- SILC →Library → data_dissemination → udb_user_database]. Further information on EU-SILC data can be found at: https://ec.europa.eu/eurostat/documents/203647/771732/Datasets- availability-table.pdf and https://ec.europa.eu/eurostat/documents/203647/203704/EU+SILC+DOI+2021rel2.pdf 5.15 LIS We continue to use the Luxembourg Income Study (LIS) for the following eight economies.10 Australia, Canada, Germany, Israel, Japan, South Korea, United States, and Taiwan, China. For the countries that use EU-SILC in recent years (typically from the early 2000s), we continue to use LIS data in the earlier years, and the break in comparability (when we switch from LIS to EU- SILC) is indicated in the comparability database. The following country-years have been added with this release (details explained on the LIS website): • Austria: 2016, 2018 • Germany: 1992, 1993, 1996, 1997, 1999, 2017, 2018 • Israel: 2002, 2003, 2004, 2006, 2008, 2009, 2011, 2013, 2015, 2017, 2018 • United Kingdom: 1996, 1997, 1998 • United States: 2019 All LIS data have been downloaded on December 15th, 2022. As before, we use disposable income per capita from the LIS data in the form of 400 bins (see Chen et al., 2018 for more details). 10 The term country, used interchangeably with economy, does not imply political independence but refers to any territory for which authorities report separate social or economic statistics. 22 6 Changes to survey years For the following countries, the survey year was corrected as follows: • Republic of the Congo 2011 corrected to 2011.67 • Iran surveys for 2005, 2006, 2009, 2013, 2014, 2015, 2016, 2017 corrected to 2005.23, 2006.23, 2009.23, 2013.23, 2014.23, 2015.23, 2016.23, 2017.23. • Madagascar 2012 corrected to 2012.73 • Mali 2006 corrected to 2006.25 • Mauritania 2000, 2004, 2008 corrected to 2005.5, 2004.55, 2008.2 • Senegal 2001 corrected to 2001.31 • Solomon Islands 2005 corrected to 2005.5 7 Changes to data coverage type • Uruguay 1999-2005 The coverage for Uruguay 1999-2005 was corrected from national to urban. • Bolivia 1992 The coverage for Bolivia 1992 was corrected from national to urban. 8 Changes to CPI data The baseline source of CPI data has been updated to the IMF’s International Financial Statistics (IFS) as of 1 November 2021. Lakner et al. (2018) provide an overview of the various CPI series that are used in PovcalNet. Table A.1 in the Appendix to this note gives the up-to-date source of the deflator for all countries included in PIP as of the current update. 8.1 China 2017-2019 From 2010-2016, the inflation for rural China was based on the increase in the nominal value of the rural poverty line (Lakner et al., 2018). This practice was adopted in an attempt to reflect the price changes faced by the poor, at a time when food prices were growing significantly faster than non-food prices. With a rural poverty rate of 3.1% in 2017 and 1.7% in 2018 (according to the Chinese national poverty definition using the 2010 poverty standard), the increase in the rural poverty line might only reflect the prices faced by the very bottom tail of the distribution. Hence, we will start using the published rural inflation series from Table 5-1 of the China Statistical Yearbook. This is in line with the practice for urban China, as well as with the preferred approach for the rest of the world where we also default to official CPI series. 23 9 Changes to National Accounts Data The primary source of national accounts data in this update is the January 2022 version of the World Development Indicators (WDI). When WDI data are missing, data from the IMF’s World Economic Outlook (WEO), October 2021 version are used. Supplementary data from the Maddison Project Database (MPD), 2020 version are further used to fill missing observations. For a more complete series, national accounts data are chained on backward or forward using growth rates in WEO data, or MPD data, when WDI data are missing. In addition, the following special economy series are used: • India 2011-2015, after 2015: As before, the reference year estimates for India from 2012 to 2015 are based on a method which adjusts HFCE growth by incorporating findings of a survey-to-survey imputation for 2014.5. Growth rates in national accounts are adjusted to match the poverty estimates from the imputation exercise. The method is described in greater detail in Chen et al (2018) and Newhouse and Vyas (2018). After 2015, growth rates in national accounts are adjusted with a pass-through rate of 67%, as described in Castaneda et al. (2020). • Syrian Arab Republic: WDI data are chained with growth rates from a special national accounts series after 2010 using the following sources: Gobat and Kostial (2016) (2011- 2015) and Devadas et al. (2019) (2016-2019). The national accounts data from WDI are mostly reported for calendar years, but in a few exceptional cases for fiscal years. See Appendix 2 (Table A2.1) for the list of the exceptional countries and the end-dates of their fiscal years. In this update, for the first time, we adjust the national accounts series of these exceptional countries and convert them from fiscal-year data to calendar-year data. Depending on the end-date of the fiscal year, we estimate the calendar-year data as a weighted average of the data from the two relevant fiscal years. Setting = (( – 1) + / )/12, where the fiscal year ends on the th day of the th month and is the number of days in the th month (e.g. if the fiscal year ends on 30th June, = 30, = 6, ⁡⁡= 30), there are two cases of the adjustment. Case 1 ( < 0.5): If the fiscal year ends before the first half of the calendar year (e.g. 31st March for India), the calendar-year data are given as: = −1 + (1 − ) Case 2 ( ≥ 0.5): If the fiscal year ends at the end of, or after, the first half of the calendar year (e.g. 30th June for Australia, 30th September for Haiti), the calendar-year data are given as: = + (1 − )+1 24 where: is GDP/HFCE per capita reported in WDI for year is the GDP/HFCE per capita for calendar year As an example, the 2019 fiscal year in India starts from 1st April 2018 to 31st March 2019. Since a larger share of the 2019 fiscal year is in 2018, WDI reports the 2019 fiscal year data as 2018 data. Thus, our 2019 calendar year estimate for India is estimated as 25% of the data WDI reports for 2018 and 75% of the data WDI reports for 2019. When the fiscal year ends at the end of the first half of the year, the adjustment is more straightforward: for Australia, the 2019 calendar year estimate is given as 50% of the data WDI reports for 2019 and 50% of the data WDI reports for 2020. 10 Changes to Population Data Nearly all population data comes from WDI, which have been updated to the December 2021. Previously, there were only three exceptions to this due to missing population data in WDI; West Bank and Gaza before 1990, for Kuwait 1992-94, and Sint Maarten (Dutch part) before 1998 (Arayavechkit et al., 2021). WDI now has data for Sint Maarten (Dutch part) before 1998 so we exclusively rely on WDI for the population data for this economy. For the other exceptions (West Bank and Gaza; Kuwait), we continue to use the United Nations Population Division’s (UNPD) World Population Prospects 2019 Revision. 11 Comparability database Since September 2019, we provide metadata on comparability of poverty estimates within countries over time. The assessment of comparability is country-dependent and relies on the accumulation of knowledge from past and current Bank staff in the countries, as well as close dialogue with national data producers with knowledge of survey design and methodology (see Atamanov et al. 2019, for more information on reasons that break comparability). With this data update, we have also updated the database to include the new datapoints and made some revisions for previously published datapoints (reflecting new information on comparability). As explained above, COVID-19 affected data collection, so many of the new surveys in 2020 are not comparable to the 2019 data. The updated comparability database can be accessed here: https://datacatalog.worldbank.org/dataset/comparability-over-time-country-level-international- poverty-measures. More information on how to use the database is available in Atamanov et al. (2019), this blog and this replication code. The PIP website also now indicates comparability in its main output. 25 12 Economy-years added/removed 12.1 Economy-years removed 12.1.1 Nigeria 2009/10 HNLSS The 2009/10 Harmonised Nigeria Living Standards Survey (HNLSS) has been removed with this update. At the same time, three new survey years (2010/11, 2012/13, and 2015/16) have been included, based on imputed data (see above). No changes were made to the survey in 2003/04 and earlier. A thorough analysis of the 2009/10 HNLSS consumption data reveals clear anomalies that affect the use of this dataset for poverty calculations (full details can be found in Appendix A of World Bank (2016)).11 These concerns motivated an extensive effort to improve data quality in the following decade, and strongly influenced the changes and improvements adopted in preparation for the 2018/19 NLSS. There are two anomalies that merit particular attention. First, the 2009/10 HNLSS data shows a systematic relationship between measured consumption levels and the month of data collection. Consumption levels in the data collected during the first four months of interviews were systematically above consumption levels in the data collected in the following months, even after controlling for seasonality. This does not match the pattern in consumption observed in the 2003/04 Nigeria Living Standards Survey (NLSS). Much of the sharp decline following the fourth month of data collection comes from purchased food and non-food items, which should be less susceptible to seasonality than own-produced food (see Figure A.1. in World Bank, 2016). Second, comparing the consumption shares in the 2009/10 HNLSS with those in the 2003/04 NLSS contradicts the Engel relationship. Engel’s law states that the food share declines with total consumption. While GDP per capita increased and poverty apparently declined between 2003/04 and 2009/10, the share of consumption devoted to food, especially own-produced food, increased (see Table A.1. in World Bank, 2016). The opposite is typically true. What is more, the share of consumption devoted to food rose more in those regions (zones) where poverty fell more, suggesting the Engel relationship broke down across both time and location. These issues with the 2009/10 HNLSS data are corroborated by new estimates for the poverty trend for Nigeria between 2009 and 2019. Lain, Schoch, and Vishwanath (2022) estimate, using two very different methodologies, a poverty headcount ratio at the US$1.90 poverty line of 46.3 percent in 2009 and of 43.5 percent in 2010/11. This is in stark contrast with the poverty headcount ratio of 56.4 percent obtained using the 2009/10 HNLSS that was previously available in PovcalNet. This difference is consistent with the above discussion on data quality concerns 26 showing that consumption levels are underestimated in the 2009/10 HNLSS, resulting in a higher poverty headcount rate than the level estimated with alternative methodologies. In light of these issues, the 2009/10 HNLSS is removed with this update. 12.1.2 South Africa 1996 After extensive review of available microdata and discussions with the country team, the decision was made to remove the datapoint corresponding to South Africa 1996. This decision was made given the inability to replicate and back up historical estimates with the available microdata. 12.2 Economy-years added The table below gives the list of new economy-years added to the PIP database. A total of 114 new economy-years were added. Table 19 Economies-years added in April 2022 PIP update Economy Reporting Year12 Survey Name Albania 2016 SILC-C Albania 2017 SILC-C Albania 2018 SILC-C Albania 2018 HBS Albania 2019 HBS Argentina 2020 EPHC-S2 Armenia 2020 ILCS Australia 2016 SIH-LIS Australia 2018 SIH-LIS Austria 2019 EU-SILC Belarus 2020 HHS Belgium 2019 EU-SILC Benin 2019 EHCVM Bolivia 2020 EH Brazil 2020 PNADC-E5 Bulgaria 2019 EU-SILC Burkina Faso 2019 EHCVM Chad 2019 EHCVM Chile 2020 CASEN China 2017 CNIHS China 2018 CNIHS China 2019 CNIHS Colombia 2020 GEIH Costa Rica 2020 ENAHO 12 This is the year for which welfare is reported. This is equal to the year variable in PIP. 27 Cote d'Ivoire 2019 EHCVM Croatia 2019 EU-SILC Cyprus 2019 EU-SILC Czech Republic 2019 EU-SILC Denmark 2019 EU-SILC Dominican Republic 2020 ECNFT-Q03 Ecuador 2020 ENEMDU Estonia 2019 EU-SILC Fiji 2019 HIES Finland 2019 EU-SILC Georgia 2020 HIS Germany 1992 GSOEP-LIS Germany 1993 GSOEP-LIS Germany 1996 GSOEP-LIS Germany 1997 GSOEP-LIS Germany 1999 GSOEP-LIS Germany 2017 GSOEP-LIS Germany 2018 GSOEP-LIS Greece 2019 EU-SILC Guinea 2019 EHCVM Guinea-Bissau 2018 EHCVM Hungary 2019 EU-SILC Indonesia 2020 SUSENAS Indonesia 2021 SUSENAS Iran, Islamic Rep. 2019 HEIS Ireland 2018 EU-SILC Israel 2002 HES-LIS Israel 2003 HES-LIS Israel 2004 HES-LIS Israel 2006 HES-LIS Israel 2008 HES-LIS Israel 2009 HES-LIS Israel 2011 HES-LIS Israel 2013 HES-LIS Israel 2015 HES-LIS Israel 2017 HES-LIS Israel 2018 HES-LIS Italy 2018 EU-SILC Kiribati 2019 HIES Kyrgyz Republic 2020 KIHS Latvia 2019 EU-SILC Lithuania 2019 EU-SILC Luxembourg 2019 EU-SILC 28 Malawi 2019 IHS-V Maldives 2020 HIES Mali 2019 EHCVM Malta 2019 EU-SILC Marshall Islands 2020 HIES Mexico 2020 ENIGHNS Moldova 2019 HBS Montenegro 2017 SILC-C Montenegro 2018 SILC-C Netherlands 2019 EU-SILC Niger 2019 EHCVM Nigeria 2010 Imputed GHSP-W1 Nigeria 2012 Imputed GHSP-W2 Nigeria 2015 Imputed GHSP-W3 Norway 2019 EU-SILC Paraguay 2020 EPH Peru 2020 ENAHO Portugal 2019 EU-SILC Romania 2018 HBS Romania 2019 EU-SILC Russian Federation 2014 VNDN Russian Federation 2015 VNDN Russian Federation 2016 VNDN Russian Federation 2017 VNDN Russian Federation 2018 VNDN Russian Federation 2019 HBS Russian Federation 2020 HBS Senegal 2019 EHCVM Serbia 2018 EU-SILC Serbia 2019 EU-SILC Serbia 2019 HBS Slovak Republic 2017 EU-SILC Slovak Republic 2019 EU-SILC Slovenia 2019 EU-SILC Spain 2019 EU-SILC Sweden 2019 EU-SILC Tanzania 2018 HBS Thailand 2020 SES Togo 2018 EHCVM Uganda 2020 UNHS Ukraine 2020 HLCS United Kingdom 1996 FES-LIS United Kingdom 1997 FES-LIS 29 United Kingdom 1998 FES-LIS United States 2019 CPS-ASEC-LIS Uruguay 2020 ECH Vanuatu 2019 NSDP 30 13 References Arayavechkit, Tanida; Atamanov, Aziz; Barreto Herrera, Karen Y.; Belghith, Nadia Belhaj Hassine; Castaneda Aguilar, R. Andres; Fujs, Tony H.M.J.; Dewina, Reno; Diaz-Bonilla, Carolina; Edochie, Ifeanyi N.; Jolliffe, Dean; Lakner, Christoph; Mahler, Daniel; Montes, Jose; Moreno Herrera, Laura L.; Mungai, Rose; Newhouse, David; Nguyen, Minh C.; Sanchez Castro, Diana M.; Schoch, Marta; Sharma, Dhiraj; Simler, Kenneth; Swinkels, Rob; Takamatsu, Shinya; Uochi, Ikuko; Viveros Mendoza, Martha C.; Yonzan, Nishant; Yoshida, Nobuo; Wu, Haoyu. 2021. March 2021 PovcalNet Update : What’s New. World Bank, Washington, DC. © World Bank. https://openknowledge.worldbank.org/handle/10986/35340 License: CC BY 3.0 IGO. Atamanov, Aziz, Joao Pedro Azevedo, R. Andres Castaneda Aguilar, Shaohua Chen, Paul A. Corral Rodas, Reno Dewina, Carolina Diaz-Bonilla, Dean M. Jolliffe, Christoph Lakner, Kihoon Lee, Daniel Gerszon Mahler, Jose Montes, Rose Mungai, Minh C. Nguyen, Espen Beer Prydz, Prem Sangraula, Kinnon Scott, Ayago Esmubancha Wambile, Judy Yang and Qinghua Zhao., “April 2018 Povcalnet update: what’s new”, World Bank Group Global Poverty Monitoring Technical Note, no. 1., April 2018. https://ideas.repec.org/p/wbk/wbgpmt/1.html. Atamanov, Aziz, R. Andres Castaneda Aguilar, Carolina Diaz-Bonilla, Dean Jolliffe, Christoph Lakner, Daniel Gerszon Mahler, Jose Montes, Laura Liliana Moreno Herrera, David Newhouse, Minh C. Nguyen, Espen Beer Prydz, Prem Sangraula, Sharad Alan Tandon and Judy Yang, “September 2019 PovcalNet Update: what’s new”, World Bank Group Global Poverty Monitoring Technical Note, no. 10., September 2019. https://ideas.repec.org/p/wbk/wbgpmt/10.html Atamanov, Aziz; Castaneda Aguilar, R. Andres; Fujs, Tony H.M.J.; Dewina, Reno; Diaz-Bonilla, Carolina; Mahler, Daniel Gerszon; Jolliffe, Dean; Lakner, Christoph; Matytsin, Mikhail; Montes, Jose; Moreno Herrera, Laura L.; Mungai, Rose; Newhouse, David; Nguyen, Minh C.; Parada Gomez Urquiza, Francisco J.; Silwal, Ani Rudra; Sanchez Castro, Diana M.; Schoch, Marta; Vargas Mogollon, David L.; Viveros Mendoza, Martha C.; Yang, Judy; Yoshida, Nobuo; Wu, Haoyu. 2020. March 2020 PovcalNet Update : What's New. Global Poverty Monitoring Technical Note,no. 11;. World Bank, Washington, DC. © World Bank. https://openknowledge.worldbank.org/handle/10986/33496 License: CC BY 3.0 IGO. Castaneda Aguilar, R. Andres; Fujs, Tony; Jolliffe, Dean; Lakner, Christoph; Gerszon Mahler, Daniel; Nguyen, Minh C.; Schoch, Marta; Vargas Mogollon, David L.; Viveros Mendoza, Martha C.; Baah, Samuel Kofi Tetteh; Yonzan, Nishant; Yoshida, Nobuo. 2020. September 2020 PovcalNet Update : What’s New. Global Poverty Monitoring Technical Note;No. 14. World Bank, Washington, DC. © World Bank. https://openknowledge.worldbank.org/handle/10986/34451 License: CC BY 3.0 IGO. Chen, Shaohua, Dean M. Jolliffe, Christoph Lakner, Kihoon Lee, Daniel Gerszon Mahler, Rose Mungai, Minh C. Nguyen, Espen Beer Prydz, Prem Sangraula, Dhiraj Sharma, Judy Yang and Qinghua Zhao., “September 2018 PovcalNet Update: what's new”, World Bank Group Global Poverty Monitoring Technical Note, no. 2., September 2018. https://ideas.repec.org/p/wbk/wbgpmt/2.html. Deaton, Angus; Zaidi, Salman. 2002. Guidelines for Constructing Consumption Aggregates for Welfare Analysis. LSMS Working Paper; No. 135. World Bank. © World Bank. https://openknowledge.worldbank.org/handle/10986/14101 License: CC BY 3.0 IGO.” Deaton, Angus, and Alessandro Tarozzi. "Prices and poverty in India." Princeton, July (2000). Edochie, Ifeanyi Nzegwu, Freije-Rodriguez, Samuel, Lakner, Christoph, Moreno Herrera, Laura Liliana, Newhouse, David Locke, Sinha Roy, Sutirtha, Yonzan. Nishant, 2022. "What Do We Know about Poverty in India in 2017/18?," Policy Research Working Paper Series 9931, The World Bank. Korinek, Anton, Johan A. Mistiaen, and Martin Ravallion. "Survey nonresponse and the distribution of income." The Journal of Economic Inequality 4, no. 1 (2006): 33-55. IBGE, Technical Note 04/2021 “Pesquisa Nacional por Amostra de Domicílios Contínua – PNAD Contínua” https://biblioteca.ibge.gov.br/visualizacao/livros/liv101882.pdf 31 INDEC, 2020, Encuesta Permanente de hogares, November 2020, https://www.indec.gob.ar/ftp/cuadros/menusuperior/eph/EPH_consideraciones_metodologicas_2t20.pdf INE 2020, Metodología de la ECH no presencial, https://www.ine.gub.uy/c/document_library/get_file?uuid=42124b3c-10c3-435f-8061- 3e5cac29b915&groupId=10181 INEC, 2020, Encuesta National de Hogares Julio 2020, https://www.inec.cr/sites/default/files/documetos-biblioteca- virtual/renaho2020.pdf Lain, Jonathan William; Schoch, Marta; Vishwanath, Tara. Estimating a Poverty Trend for Nigeria between 2009 and 2019 (English). Policy Research working paper ,no. WPS 9974, Washington, D.C. : World Bank Group. http://documents.worldbank.org/curated/en/631201647459995911/Estimating-a-Poverty-Trend-for-Nigeria- between-2009-and-2019 Lakner, Christoph, Daniel Gerszon Mahler, Minh C. Nguyen, Joao Pedro Azevedo, Shaohua Chen, Dean M. Jolliffe, Espen Beer Prydz and Prem Sangraula., "Consumer Price Indices used in Global Poverty Measurement ", World Bank Group Global Poverty Monitoring Technical Note, no. 4., September 2018. https://ideas.repec.org/p/wbk/wbgpmt/4.html. Lara Ibarra, Gabriel and Vale, Ricardo “Brazil 2020 data update: adaptations to the methodology to estimate the World Bank’s welfare, poverty and inequality in the country”, World Bank Group Global Poverty Monitoring Technical Note, no. 21., April 2022. Prydz, Espen Beer, Dean M. Jolliffe, Christoph Lakner, Daniel G erszon Mahler, and Prem Sangraula. “National Accounts Data used in Global Poverty Measurement.” Global Poverty Monitoring Technical Note 8. Washington, DC: World Bank. 2019. https://ideas.repec.org/p/wbk/wbgpmt/8.html. The Economist Intelligence Unit (EIU). 2022. Country Report Syria 1st Quarter 2022. http://www.eiu.com/FileHandler.ashx?issue_id=1551743938&mode=pdf Yoshida, Nobuo, A.S. Munoz, C.K. Lee, M. Brataj, and D. Sharma. 2015. “SWIFT Data Collection Guidelines version 2”. http://documents1.worldbank.org/curated/en/591711545170814297/pdf/97499-WP-P149557-OUO9- Box391480B-ACS.pdf Yoshida, Nobuo, X. Chen, S. Takamatsu, K. Yoshimura, S. Malgioglio and S. Shivakumaran 2020. “The Concept and Empirical Evidence of SWIFT Methodology” (Mimeo) World Bank. 2016. “Poverty Reduction in Nigeria in the Last Decade.” World Bank, Washington, DC. https://openknowledge.worldbank.org/handle/10986/25825 License: CC BY 3.0 IGO. World Bank. 2020. Poverty and Shared Prosperity 2020 : Reversals of Fortune. Washington, DC: World Bank https://openknowledge.worldbank.org/handle/10986/34496 License: CC BY 3.0 IGO. World Bank, 2022 “A better future for all Nigerians- Poverty Assessment Nigeria 2022”, Washington DC: World Bank https://documents1.worldbank.org/curated/en/099730003152232753/pdf/P17630107476630fa09c990da7805355 11c.pdf 32 14 Appendix 1 – CPI Data sources Table A1.1 lists the source of CPI used for each economy-year reported in PIP. The columns in the table are defined as follows: • Code: The 3-letter economy code used by the World Bank: https://datahelpdesk.worldbank.org/knowledgebase/articles/906519-world-bank- countryand-lending-groups • Economy name: Name of economy • Year(s): Welfare reporting year, i.e., the year for which the welfare has been reported. If the survey collects income for the previous year, it is the year prior to the survey. • CPI period: Common time period to which the welfare aggregates in the survey have been deflated. The letter Y denotes that the CPI period is identical to the year column. When the welfare aggregate has been deflated to a particular month within the welfare reporting year, the month is indicated by a number between 1 and 12, preceded by an M, and similarly with a Q for quarters. The letter W indicates that a weighted CPI is used, as described in equation 1 in Lakner et al. (2018). • CPI source: Source of the deflator used. The source is given by the abbreviation, the frequency of the CPI, and the vintage; e.g. IFS-M-202111 denotes the monthly IFS database version November 2021. For economy-specific deflators, the description is given in the text or further details are available upon request. 33 Table A1.1. Source of temporal deflators used in PIP April 2022 update Reporting Code Economy Survey CPI period Source Year(s)13 HBS 2000 W IFS-M-202111 AGO Angola IBEP-MICS 2008 W IFS-M-202111 IDREA 2018 W IFS-M-202111 EWS 1996 Y IFS-M-202111 LSMS 2002-2012 Y IFS-M-202111 ALB Albania HBS 2014-2019 Y IFS-M-202111 SILC-C 2017-2019 (prev. year)Y IFS-M-202111 United Arab HIES 2014 W IFS-M-202111 ARE Emirates HIES 2019 Y IFS-M-202111 EPH 1980-1987 Y NSO EPH 1991-2002 M9 NSO ARG Argentina EPHC-S2 2003-2020 M7-M12 NSO EPHC-S2 2007-2014 M7-M12 Private estimates ARM Armenia ILCS 1996-2020 Y IFS-M-202111 HIS-LIS 1981 Y IFS-A-202111 IDS-LIS 1985 Y IFS-A-202111 AUS Australia SIH-LIS 1989-2018 Y IFS-A-202111 SIH-HES-LIS 2004-2010 Y IFS-A-202111 MC-LIS 1987-1995 Y IFS-M-202111 AUT Austria ECHP-LIS 1994-2000 Y IFS-M-202111 EU-SILC 2004-2020 (prev. year)Y IFS-M-202111 SLC 1995 Y IFS-M-202111 AZE Azerbaijan HBS 2001-2005 Y IFS-M-202111 EDCM 1992 Y IFS-M-202111 EP 1998 W IFS-M-202111 BDI Burundi QUIBB 2006 Y IFS-M-202111 ECVMB 2013 W IFS-M-202111 SEP-LIS 1985-1997 Y IFS-M-202111 PSBH-ECHP- BEL Belgium 1995-2000 Y IFS-M-202111 LIS EU-SILC 2004-2020 (prev. year)Y IFS-M-202111 QUIBB 2003 Y IFS-M-202111 EMICOV 2011 W IFS-M-202111 BEN Benin EMICOV 2015 Y IFS-M-202111 EHCVM 2018 W IFS-M-202111 EP-I 1994 W IFS-M-202111 EP-II 1998 Y IFS-M-202111 BFA Burkina Faso ECVM 2003-2009 Y IFS-M-202111 EMC 2014 Y IFS-M-202111 13 This is the year for which welfare is reported. This is equal to the year variable in PIP. 34 EHCVM 2018 W IFS-M-202111 HHES 1983-1985 W WEO-A-202110 HHES 1988-1991 W IFS-A-202111 BGD Bangladesh HHES 1995 W Survey HIES 2000-2016 Y Survey HBS 1989 Y IFS-A-202111 HBS 1992-1994 Y IFS-M-202111 BGR Bulgaria IHS 1995-2001 Y IFS-M-202111 MTHS 2003-2007 Y IFS-M-202111 EU-SILC 2007-2020 (prev. year)Y IFS-M-202111 Bosnia and LSMS 2001-2004 Y WEO-A-202110 BIH Herzegovina HBS 2007-2011 Y IFS-M-202111 FBS 1993-1995 Y IFS-M-202111 HHS 1998-2020 Y IFS-M-202111 BLR Belarus LFS 1993-1999 Y WEO-A-202110 HBS 1995 Y WEO-A-202110 SLC 1996 Y WEO-A-202110 EPF 1990 W IFS-M-202111 EIH 1992 M11 IFS-M-202111 ENE 1997 M11 IFS-M-202111 ECH 1999 M10 IFS-M-202111 BOL Bolivia ECH 2000 M11 IFS-M-202111 EH 2001-2005 M11 IFS-M-202111 ECH 2004 M10 IFS-M-202111 EH 2006-2016 M10 IFS-M-202111 EH 2017-2020 M11 IFS-M-202111 PNAD 1981-2011 M9 IFS-M-202111 BRA Brazil PNADC-E1 2012-2019 Y IFS-M-202111 PNADC-E5 2020 Y IFS-M-202111 BTN Bhutan BLSS 2003-2017 Y Previous WDI/IFS HIES 1985-2002 W IFS-M-202111 BWA Botswana CWIS 2009 W IFS-M-202111 BMTHS 2015 W IFS-M-202111 EPCM 1992 W IFS-M-202111 Central CAF African Republic ECASEB 2008 Y IFS-M-202111 SCF-LIS 1971-1997 Y IFS-M-202111 CAN Canada SLID-LIS 1998-2010 Y IFS-M-202111 CIS-LIS 2012-2017 Y IFS-M-202111 CHE Switzerland SIWS-LIS 1982 Y IFS-M-202111 35 NPS-LIS 1992 Y IFS-M-202111 IES-LIS 2000-2002 Y IFS-M-202111 EU-SILC 2007-2019 (prev. year)Y IFS-M-202111 CASEN 1987 Y IFS-M-202111 CHL Chile CASEN 1990-2020 M11 IFS-M-202111 CRHS-CUHS 1981-2011 Y Special CHN China CNIHS 2012-2019 Y Special EPAM 1985-1988 W IFS-M-202111 EP 1992 W IFS-M-202111 CIV Côte d’Ivoire ENV 1995-2015 Y IFS-M-202111 EHCVM 2018 W IFS-M-202111 ECAM-I 1996 Y IFS-M-202111 ECAM-II 2001 Y IFS-M-202111 CMR Cameroon ECAM-III 2007 Y IFS-M-202111 ECAM-IV 2014 Y IFS-M-202111 Congo, Dem. COD E123 2004-2012 W IFS-M-202111 Rep. ECOM 2005 Y IFS-M-202111 COG Congo, Rep. ECOM 2011 W IFS-M-202111 ENH 1980-1988 Y IFS-M-202111 ENH 1989-1991 M11 IFS-M-202111 COL Colombia ENH 1992-2000 M11 IFS-M-202111 ECH 2001-2005 M11 IFS-M-202111 GEIH 2008-2020 M11 IFS-M-202111 EIM 2004 Y IFS-M-202111 COM Comoros EESIC 2013 Y IFS-M-202111 IDRF 2001 W IFS-M-202111 CPV Cabo Verde QUIBB 2007 W IFS-M-202111 IDRF 2015 Y IFS-M-202111 ENH 1981-1986 Y IFS-M-202111 EHPM 1989 Y IFS-M-202111 CRI Costa Rica EHPM 1990-2009 M7 IFS-M-202111 ENAHO 2010-2020 M7 IFS-M-202111 CYP Cyprus EU-SILC 2005-2020 (prev. year)Y IFS-M-202111 MC-LIS 1992-2002 Y IFS-M-202111 Czech CZE CM 1993 Y IFS-M-202111 Republic EU-SILC 2005-2020 (prev. year)Y IFS-M-202111 DEU Germany GSOEP-LIS 1991-2018 Y IFS-M-202111 EDAM 2002-2013 Y IFS-M-202111 DJI Djibouti EDAM 2017 M5 IFS-M-202111 LM-LIS 1987-2000 Y IFS-M-202111 DNK Denmark EU-SILC 2004-2020 (prev. year)Y IFS-M-202111 DOM ENGSLF 1986-1989 Y IFS-M-202111 36 ICS 1992 M6 IFS-M-202111 ENFT 1996 M2 IFS-M-202111 Dominican ENFT 1997 M4 IFS-M-202111 Republic ENFT 2000-2016 M9 IFS-M-202111 ECNFT-Q03 2017-2020 Y IFS-M-202111 EDCM 1988 Y IFS-M-202111 DZA Algeria ENMNV 1995 Y IFS-M-202111 ENCNVM 2011 W IFS-M-202111 EPED 1987 Y IFS-M-202111 ECV 1994 M6-M10 IFS-M-202111 EPED 1995 M11 IFS-M-202111 ECU Ecuador EPED 1998 M6 IFS-M-202111 (prev. year)M10- ECV 1999 IFS-M-202111 M9 EPED 2000 M11 IFS-M-202111 ENEMDU 2003-2020 M11 IFS-M-202111 HIECS 1990-2012 W IFS-M-202111 Egypt, Arab EGY HIECS 2015 Y IFS-M-202111 Rep. HIECS 2017 W IFS-M-202111 HBS-LIS 1980-1990 Y IFS-M-202111 ESP Spain ECHP-LIS 1995-2000 Y IFS-M-202111 EU-SILC 2004-2020 (prev. year)Y IFS-M-202111 HIES 1993-1998 Y IFS-M-202111 EST Estonia HBS 2000-2004 Y IFS-M-202111 EU-SILC 2004-2020 (prev. year)Y IFS-M-202111 HICES 1981 W IFS-M-202111 ETH Ethiopia HICES 1995-2010 W IFS-M-202111 HICES 2015 M12 IFS-M-202111 IDS-LIS 1987-2000 Y IFS-M-202111 FIN Finland EU-SILC 2004-2020 (prev. year)Y IFS-M-202111 FJI Fiji HIES 2002-2019 W IFS-M-202111 HBS-LIS 1978-2000 Y IFS-M-202111 FRA France EU-SILC 2004-2019 (prev. year)Y IFS-M-202111 Micronesia, CPH 2000 Y IFS-A-202111 FSM Fed. Sts. HIES 2005-2013 Y IFS-A-202111 GAB Gabon EGEP 2005-2017 Y IFS-M-202111 FES-LIS 1969-1998 Y IFS-M-202111 United GBR FRS-LIS 1994-2003 Y IFS-M-202111 Kingdom EU-SILC 2005-2018 (prev. year)Y IFS-M-202111 GEO Georgia HIS 1996-2020 Y IFS-M-202111 GLSS-I 1987 W IFS-M-202111 GHA Ghana GLSS-II 1988 W IFS-M-202111 GLSS-III 1991 W IFS-M-202111 37 GLSS-IV 1998 W IFS-M-202111 GLSS-V 2005 W Survey GLSS-VI 2012 W Survey GLSS-VII 2016 W Survey ESIP 1991 Y WEO-A-202110 EIBC 1994 W WEO-A-202110 GIN Guinea EIBEP 2002 W WEO-A-202110 ELEP 2007-2012 Y IFS-M-202111 EHCVM 2018 W IFS-M-202111 HPS 1998 Y IFS-M-202111 GMB Gambia, The HIS 2003 W IFS-M-202111 IHS 2010-2015 W IFS-M-202111 ILJF 1991 Y IFS-M-202111 ICOF 1993 Y IFS-M-202111 Guinea- GNB ILAP-I 2002 Y IFS-M-202111 Bissau ILAP-II 2010 Y IFS-M-202111 EHCVM 2018 W IFS-M-202111 ECHP-LIS 1995-2000 Y IFS-M-202111 GRC Greece EU-SILC 2004-2020 (prev. year)Y IFS-M-202111 ENSD 1986 W IFS-M-202111 ENSD 1989 Y IFS-M-202111 GTM Guatemala ENIGF 1998 M8 IFS-M-202111 ENCOVI 2000 M6-M11 IFS-M-202111 ENCOVI 2006-2014 M7 IFS-M-202111 GLSMS 1992 W WEO-A-202110 GUY Guyana GLSMS 1998 Y IFS-M-202111 ECSFT 1986 Y IFS-M-202111 EPHPM 1989 Y IFS-M-202111 HND Honduras EPHPM 1990-1993 M5 IFS-M-202111 EPHPM 1994 M9 IFS-M-202111 EPHPM 1995-2019 M5 IFS-M-202111 HBS 1988-2010 Y IFS-M-202111 HRV Croatia EU-SILC 2010-2020 (prev. year)Y IFS-M-202111 ECVH 2001 M5 IFS-M-202111 HTI Haiti ECVMAS 2012 M10 IFS-M-202111 HBS 1987-2007 Y IFS-M-202111 HHP-LIS 1991-1994 Y IFS-M-202111 HUN Hungary THMS-LIS 1999 Y IFS-M-202111 EU-SILC 2005-2020 (prev. year)Y IFS-M-202111 SUSENAS 1984-1999 Y IFS-M-202111 IDN Indonesia SUSENAS 2000-2007 M2 IFS-M-202111 SUSENAS 2008-2021 M3 IFS-M-202111 IND India NSS 1977 W Special 38 NSS 1983 Y Special NSS-SCH1 1987-2011 W Special SIDPUSS-LIS 1987 Y IFS-M-202111 LIS-ECHP-LIS 1994-2000 Y IFS-M-202111 IRL Ireland SILC-LIS 2002 Y IFS-M-202111 EU-SILC 2004-2019 (prev. year)Y IFS-M-202111 SECH 1986 Y IFS-A-202111 Iran, Islamic IRN SECH 1990-1998 Y IFS-M-202111 Rep. HEIS 2005-2019 W IFS-M-202111 M11-(next IHSES 2006 COSIT IRQ Iraq year)M12 IHSES 2012 Y COSIT ISL Iceland EU-SILC 2004-2018 (prev. year)Y IFS-M-202111 ISR Israel HES-LIS 1979-2018 Y IFS-M-202111 SHIW-LIS 1986-2000 Y IFS-M-202111 ITA Italy EU-SILC 2004-2019 (prev. year)Y IFS-M-202111 SLC 1988 M9 IFS-M-202111 M11-(next SLC 1990-1993 IFS-M-202111 year)M3 JAM Jamaica SLC 1996 M5-M8 IFS-M-202111 SLC 1999 M6-M8 IFS-M-202111 SLC 2002-2004 M6 IFS-M-202111 HEIS 1986 W IFS-M-202111 JOR Jordan HEIS 1992-1997 Y IFS-M-202111 HEIS 2002-2010 W IFS-M-202111 JPN Japan JHPS-LIS 2008-2013 Y IFS-M-202111 HBS 1993-2018 Y IFS-M-202111 KAZ Kazakhstan LSMS 1996 Y IFS-M-202111 WMS-I 1992 Y NSO WMS-II 1994 Y NSO KEN Kenya WMS-III 1997 Y NSO IHBS 2005-2015 W NSO KPMS 1998 Y IFS-M-202111 Kyrgyz KGZ HBS 2000-2003 Y IFS-M-202111 Republic KIHS 2004-2020 Y IFS-M-202111 HIES 2006 Y IFS-M-202111 KIR Kiribati HIES 2019 W IFS-M-202111 HIES-FHES- KOR Korea, Rep. 2006-2016 Y IFS-M-202111 LIS LECS 1992 W IFS-A-202111 LECS 1997 W IFS-M-202111 LAO Lao PDR LECS 2002-2007 W Survey LECS 2012-2018 W IFS-M-202111 LBN Lebanon HBS 2011 (next year)M5 IFS-M-202111 39 CWIQ 2007 Y IFS-M-202111 LBR Liberia HIES 2014-2016 Y IFS-M-202111 LSMS 1995 Y IFS-M-202111 LCA St. Lucia SLC-HBS 2016 M1 IFS-M-202111 LFSS 1985 Y IFS-M-202111 HIES 1990 W IFS-M-202111 SES 1995 W IFS-M-202111 LKA Sri Lanka HIES 2002 Y IFS-M-202111 HIES 2006-2012 W IFS-M-202111 HIES 2016 Y IFS-M-202111 HBS 1986 W WEO-A-202110 NHECS 1994 W WEO-A-202110 LSO Lesotho HBS 2002 W IFS-M-202111 CMSHBS 2017 M8 IFS-M-202111 HBS 1993-2008 Y IFS-M-202111 LTU Lithuania EU-SILC 2005-2020 (prev. year)Y IFS-M-202111 PSELL-LIS 1985-1991 Y IFS-M-202111 PSELL-ECHP- LUX Luxembourg 1994-2000 Y IFS-M-202111 LIS EU-SILC 2004-2020 (prev. year)Y IFS-M-202111 HBS 1993-2009 Y IFS-M-202111 LVA Latvia EU-SILC 2005-2020 (prev. year)Y IFS-M-202111 ECDM 1984 W IFS-M-202111 MAR Morocco ENNVM 1990-2006 W IFS-M-202111 ENCDM 2000-2013 W IFS-M-202111 MDA Moldova HBS 1997-2019 Y IFS-M-202111 EB 1980 Y IFS-M-202111 EPM 1993 W IFS-M-202111 MDG Madagascar EPM 1997-2010 Y IFS-M-202111 ENSOMD 2012 W IFS-M-202111 HIES 2002-2009 W IFS-M-202111 MDV Maldives HIES 2016 Y IFS-M-202111 HIES 2019 M11 IFS-M-202111 ENIGH 1984-2014 M8 IFS-M-202111 MEX Mexico ENIGHNS 2016-2020 M8 IFS-M-202111 Marshall MHL HIES 2019 W WEO-A-202110 Islands North HBS 1998-2008 Y IFS-M-202111 MKD Macedonia SILC-C 2010-2019 (prev. year)Y IFS-M-202111 EMCES 1994 Y IFS-A-202111 EMEP 2001 W IFS-M-202111 MLI Mali ELIM 2006-2009 W IFS-M-202111 EHCVM 2018 W IFS-M-202111 40 MLT Malta EU-SILC 2007-2020 (prev. year)Y IFS-M-202111 MPLCS 2015 M1 IFS-M-202111 MMR Myanmar MLCS 2017 Q1 IFS-M-202111 HBS 2005-2014 Y IFS-M-202111 MNE Montenegro SILC-C 2013-2019 (prev. year)Y IFS-M-202111 LSMS 1995-1998 Y IFS-M-202111 HIES-LSMS 2002 W IFS-M-202111 MNG Mongolia HSES 2007 W IFS-M-202111 HSES 2010-2018 Y IFS-M-202111 NHS 1996 W WEO-A-202110 MOZ Mozambique IAF 2002 W WEO-A-202110 IOF 2008-2014 W IFS-M-202111 EPCV 1987 Y IFS-M-202111 EP 1993 Y IFS-M-202111 MRT Mauritania EPCV 1995-2008 W IFS-M-202111 EPCV 2014 Y IFS-M-202111 HBS 2006 W IFS-M-202111 MUS Mauritius HBS 2012-2017 Y IFS-M-202111 IHS-I 1997 W IFS-M-202111 IHS-II 2004 W Survey MWI Malawi IHS-III 2010 W Survey IHS-IV 2016 M4 Survey IHS-V 2019 M4 Survey HIS 1984-1997 Y IFS-M-202111 (prev. year)M7- HIS 2004 IFS-M-202111 (prev. year)M12 MYS Malaysia (prev. year)M7- HIS 2007 IFS-M-202111 (prev. year)M10 HIS 2009 W IFS-M-202111 HIS 2012-2016 Y IFS-M-202111 NHIES 1993 W WEO-A-202110 NAM Namibia NHIES 2003-2015 W IFS-M-202111 ENBCM 1992-2007 W IFS-M-202111 EPCES 1994 W IFS-M-202111 NER Niger ENCVM 2005 Y IFS-M-202111 ECVMA 2011-2014 Y IFS-M-202111 EHCVM 2018 W IFS-M-202111 NCS 1985 W IFS-M-202111 NCS 1992-1996 Y IFS-M-202111 LSS 2003 W IFS-M-202111 NGA Nigeria GHSP-W1 2010 2019 M3-M4 IFS-M-202111 GHSP-W2 2012 2019 M3-M4 IFS-M-202111 GHSP-W3 2015 2019 M3-M4 IFS-M-202111 41 (next year)M3- LSS 2018 IFS-M-202111 (next year)M4 EMNV 1993 M2 NSO EMNV 1998 M6 NSO NIC Nicaragua EMNV 2001 M6 IFS-M-202111 EMNV 2005-2009 M8 IFS-M-202111 EMNV 2014 M8-M10 IFS-M-202111 AVO-LIS 1983-1990 Y IFS-M-202111 NLD Netherlands SEP-LIS 1993-1999 Y IFS-M-202111 EU-SILC 2005-2020 (prev. year)Y IFS-M-202111 IDS-LIS 1979-2000 Y IFS-M-202111 NOR Norway EU-SILC 2004-2020 (prev. year)Y IFS-M-202111 MHBS 1984 W IFS-M-202111 LSS-I 1995 W IFS-M-202111 NPL Nepal LSS-II 2003 W IFS-M-202111 LSS-III 2010 W IFS-M-202111 NRU Nauru HIES 2012 W WEO-A-202110 HIES 1987 Y IFS-M-202111 HIES 1990-1998 W IFS-M-202111 PAK Pakistan IHS 1996 W IFS-M-202111 PIHS 2001 M6 IFS-M-202111 HIES 2004-2018 (next year)M1 IFS-M-202111 EMO 1979-1989 Y IFS-M-202111 PAN Panama EMO 1991 M7 IFS-M-202111 EH 1995-2019 M7 IFS-M-202111 ENNIV 1985 W IFS-M-202111 ENNIV 1994 Y IFS-M-202111 PER Peru ENAHO 1997-2002 Q4 IFS-M-202111 ENAHO 2003 M5-M12 IFS-M-202111 ENAHO 2004-2020 Y IFS-M-202111 PHL Philippines FIES 1985-2018 Y IFS-M-202111 Papua New HIES 1996 Y IFS-A-202111 PNG Guinea HIES 2009 W IFS-A-202111 HBS 1985-1987 Y IFS-A-202111 HBS-LIS 1986 Y IFS-A-202111 POL Poland HBS 1989-2019 Y IFS-M-202111 HBS-LIS 1992-1999 Y IFS-M-202111 EU-SILC 2005-2019 (prev. year)Y IFS-M-202111 PRT Portugal EU-SILC 2004-2020 (prev. year)Y IFS-M-202111 EH 1990 M7 IFS-M-202111 EH 1995 M8-M11 IFS-M-202111 PRY Paraguay EIH 1997 (next year)M2 IFS-M-202111 EPH 1999 M9 IFS-M-202111 42 EIH 2001 M3 IFS-M-202111 EPH 2002 M11 IFS-M-202111 EPH 2003 M9 IFS-M-202111 EPH 2004 M10 IFS-M-202111 EPH 2005 M11 IFS-M-202111 EPH 2006 M12 IFS-M-202111 EPH 2007-2008 M10 IFS-M-202111 EPH 2009 M11 IFS-M-202111 EPH 2010-2020 M10 IFS-M-202111 West Bank PECS 2004-2011 Y IFS-M-202111 PSE and Gaza PECS 2016 W IFS-M-202111 HBS 1989 Y Milanovic (2001) MC 1992 Y IFS-M-202111 HIS 1994-1999 Y IFS-M-202111 ROU Romania IHS-LIS 1995-1997 Y IFS-M-202111 IHS 1998-2000 Y IFS-M-202111 HBS 2001-2018 Y IFS-M-202111 EU-SILC 2007-2020 (prev. year)Y IFS-M-202111 Russian HBS 1993-2020 Y IFS-M-202111 RUS Federation VNDN 2015-2019 (prev. year)Y IFS-M-202111 ENBCM 1984 W IFS-M-202111 EICV-I 2000 W IFS-M-202111 EICV-II 2005 W IFS-M-202111 RWA Rwanda EICV-III 2010 (next year)M1 IFS-M-202111 EICV-IV 2013 (next year)M1 IFS-M-202111 EICV-V 2016 (next year)M1 IFS-M-202111 NBHS 2009 Y IFS-M-202111 SDN Sudan NBHS 2014 M11 IFS-M-202111 EP 1991 W IFS-M-202111 ESAM 1994 W IFS-M-202111 ESAM-II 2001 W IFS-M-202111 SEN Senegal ESPS-I 2005 W IFS-M-202111 ESPS-II 2011 W IFS-M-202111 EHCVM 2018 W IFS-M-202111 Solomon SLB HIES 2005-2012 W IFS-M-202111 Islands HEEAS 1989 W WEO-A-202110 SLE Sierra Leone SLIHS 2003 W WEO-A-202110 SLIHS 2011-2018 Y IFS-M-202111 EHPM 1989 Y IFS-M-202111 M10-(next SLV El Salvador EHPM 1991 IFS-M-202111 year)M4 EHPM 1995-1999 Y IFS-M-202111 43 EHPM 2000-2007 M12 IFS-M-202111 EHPM 2008-2019 M11 IFS-M-202111 SOM Somalia SHFS-W2 2017 Y Special LSMS 2002 Y IFS-M-202111 SRB Serbia HBS 2003-2019 Y IFS-M-202111 EU-SILC 2013-2020 (prev. year)Y IFS-M-202111 NBHS 2009 Y IFS-M-202111 SSD South Sudan HFS-W3 2016 (prev. year)M7 IFS-M-202111 São Tomé IOF 2000 W IFS-M-202111 STP and Principe IOF 2010-2017 Y IFS-M-202111 SUR Suriname EHS 1999 Y IFS-M-202111 MC-LIS 1992-1996 Y IFS-M-202111 Slovak SVK HBS 2004-2009 Y IFS-M-202111 Republic EU-SILC 2005-2020 (prev. year)Y IFS-M-202111 IES 1987-1993 Y IFS-M-202111 HBS-LIS 1997-1999 Y IFS-M-202111 SVN Slovenia HBS 1998-2003 Y IFS-M-202111 EU-SILC 2005-2020 (prev. year)Y IFS-M-202111 LLS-RD-LIS 1967 Y IFS-M-202111 SWE Sweden HIS-LIS 1975-2000 Y IFS-M-202111 EU-SILC 2004-2020 (prev. year)Y IFS-M-202111 SWZ Eswatini HIES 1994-2016 W IFS-M-202111 HES 1999 W IFS-M-202111 HBS 2006 W IFS-M-202111 SYC Seychelles HBS 2013 Y IFS-M-202111 HBS 2018 W IFS-M-202111 Syrian Arab SYR HIES 1996-2003 W IFS-M-202111 Republic ECOSIT-II 2003 Y IFS-M-202111 TCD Chad ECOSIT-III 2011 Y IFS-M-202111 EHCVM 2018 W IFS-M-202111 QUIBB 2006-2015 Y IFS-M-202111 TGO Togo EHCVM 2018 W IFS-M-202111 THA Thailand SES 1981-2020 Y IFS-M-202111 TLSS 1999 Y WEO-A-202110 TLSS 2003-2007 Y Survey TJK Tajikistan HBS 2004 Y Survey TLSS 2009 Y IFS-M-202111 HSITAFIEN 2015 Y IFS-M-202111 TKM Turkmenistan LSMS 1998 Y WEO-A-202110 TLSS 2001 Y WEO-A-202110 TLS Timor-Leste TLSLS 2007-2014 Y IFS-M-202111 TON Tonga HIES 2000 W IFS-M-202111 44 HIES 2009-2015 Y IFS-M-202111 Trinidad and SLC 1988 Y IFS-M-202111 TTO Tobago PHC 1992 Y IFS-M-202111 HBCS 1985 Y IFS-A-202111 HBCS 1990 Y IFS-M-202111 TUN Tunisia LSS 1995-2000 Y IFS-M-202111 NSHBCSL 2005-2015 W IFS-M-202111 TUR Turkey HICES 1987-2019 Y IFS-M-202111 TUV Tuvalu HIES 2010 Y IFS-A-202111 Taiwan, TWN FIDES-LIS 1981-2016 Y WEO-A-202110 China HBS 1991 W IFS-A-202111 HBS 2000 W IFS-M-202111 TZA Tanzania HBS 2007 Y IFS-M-202111 HBS 2011-2018 W IFS-M-202111 HBS 1989 Y WEO-A-202110 NIHS 1992 W WEO-A-202110 UGA Uganda NIHS 1996-1999 W IFS-M-202111 UNHS 2002-2019 W IFS-M-202111 HS 1992-1993 Y IFS-M-202111 UKR Ukraine HIES 1995-1996 Y IFS-M-202111 HLCS 1999-2020 Y IFS-M-202111 ENH 1981-1989 Y IFS-M-202111 URY Uruguay ECH 1992-2005 (prev. year)M12 IFS-M-202111 ECH 2006-2020 (prev. year)M12 IFS-M-202111 CPS-LIS 1974-2001 Y IFS-M-202111 USA United States CPS-ASEC- 2002-2019 Y IFS-M-202111 LIS UZB Uzbekistan HBS 1998-2003 Y WEO-A-202110 Venezuela, EHM 1981-1989 Y NSO VEN RB EHM 1992-2006 M12 NSO VLSS 1992 W WEO-A-202110 VNM Vietnam VLSS 1997 W IFS-M-202111 VHLSS 2002-2018 M1 IFS-M-202111 HIES 2010 Y IFS-A-202111 VUT Vanuatu NSDP 2019 W IFS-A-202111 HIES 2002-2008 Y IFS-M-202111 WSM Samoa HIES 2013 W IFS-M-202111 HIES 2018 Y IFS-M-202111 XKX Kosovo HBS 2003-2017 Y IFS-M-202111 HBS 1998 Y IFS-M-202111 YEM Yemen, Rep. HBS 2005 W IFS-M-202111 HBS 2014 Y IFS-M-202111 45 KIDS 1993 Y IFS-M-202111 HIES 2000 W IFS-M-202111 ZAF South Africa IES 2005-2010 (next year)M6 IFS-M-202111 LCS 2008 W IFS-M-202111 LCS 2014 (next year)M6 IFS-M-202111 HBS 1991-1993 Y IFS-M-202111 LCMS-I 1996 Y IFS-M-202111 LCMS-II 1998 Y IFS-M-202111 LCMS-III 2002 W IFS-M-202111 ZMB Zambia LCMS-IV 2004 W IFS-M-202111 LCMS-V 2006 W IFS-M-202111 LCMS-VI 2010 Y IFS-M-202111 LCMS-VII 2015 Y IFS-M-202111 ICES 2011 Y IFS-M-202111 ZWE Zimbabwe PICES 2017-2019 Y Survey 46 15 Appendix 2 – National Accounts Data Sources This appendix provides details of national accounts data used in aligning estimates to reference years (see Prydz et al, 2019 for methodological details). A complete overview of the sources used is available upon request and will soon be available on the PIP website. Table A2.1. Countries whose national accounts data are reported in fiscal years Code Country Day of month fiscal year ends Month fiscal year ends AUS Australia 30 June BGD Bangladesh 30 June EGY Egypt 30 June ETH Ethiopia 7 July HTI Haiti 30 September IND India 31 March IRN Iran 20 March MHL Marshall Islands 30 September FSM Micronesia 30 September MMR Myanmar 30 September NRU Nauru 30 June NPL Nepal 14 July PAK Pakistan 30 June WSM Samoa 30 June SSD South Sudan 30 June TON Tonga 30 June UGA Uganda 30 June Source: WDI metadata on GDP per capita series (NY.GDP.PCAP.KD) and HFCE per capita series (NE.CON.PRVT.PC.KD) Note: As an example, the 2020 fiscal year in Australia runs from 1st July 2019 to 30th June 2020. 47