Learning Poverty Monitoring Series Technical Note 1 1 Learning Poverty Updates and Revisions What’s New? July 2021 João Pedro Azevedo, Silvia Montoya, Maryam Akmal, Yi Ning Wong, Laura Gregory, Koen Martijn Geven, Marie-Helene Cloutier, Syedah Aroob Iqbal, Adolfo Gustavo Imhof, Natasha de Andrade Falcão, Cristelle Kouame, Mahesh Dahal, Tihtina Zenebe Gebre, and Maria Jose Vargas Mancera Abstract The July 2021 release of learning poverty estimates involves several changes to the data underlying the country-level learning poverty figures. This document provides details of the key changes made. Some country-level estimates have changed or become available for the first time due to new data from recent assessments: TIMSS 2019, PASEC 2019, and SEA-PLM 2019. In cases where new assessment data call for a change to the learning poverty estimates, the corresponding enrollment data used for learning poverty calculations have also been updated so that the enrollment year is as close as possible to the assessment year, depending on data availability. In the latest release, country-level estimates of learning poverty are available for 120 countries. Acknowledgments We thank Marguerite Clarke, Noah Yarrow, Paul Andres Corral Rodas, Sachiko Kataoka, Shabnam Sinha, and Tanya Guyatt for helpful feedback. The findings, interpretations, and conclusions expressed in this report 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. 2 Table of contents Introduction .................................................................................................................................................. 5 How is learning poverty measured? ............................................................................................................. 5 New assessment programs used to measure learning deprivation ............................................................. 6 TIMSS 2019 ............................................................................................................................................... 6 SEA-PLM 2019 ........................................................................................................................................... 6 PASEC 2019 ............................................................................................................................................... 6 Changes to learning poverty, learning deprivation, and schooling deprivation estimates at the country level ............................................................................................................................................................... 7 East Asia & Pacific ..................................................................................................................................... 7 Addition of new assessment data ......................................................................................................... 7 Replacement of prior assessment data ................................................................................................ 8 No changes to assessment data............................................................................................................ 9 Europe & Central Asia ............................................................................................................................. 10 Addition of new assessment data ....................................................................................................... 10 Replacement of prior assessment data .............................................................................................. 10 No changes to assessment data.......................................................................................................... 11 Latin America & the Caribbean ............................................................................................................... 16 Addition of new assessment data ....................................................................................................... 16 Replacement of prior assessment data .............................................................................................. 16 No changes to assessment data.......................................................................................................... 17 Middle East & North Africa ..................................................................................................................... 19 Addition of new assessment data ....................................................................................................... 19 Replacement of prior assessment data .............................................................................................. 19 No changes to assessment data.......................................................................................................... 19 South Asia ............................................................................................................................................... 22 Addition of new assessment data ....................................................................................................... 22 Replacement of prior assessment data .............................................................................................. 22 No changes to assessment data.......................................................................................................... 22 Sub-Saharan Africa .................................................................................................................................. 23 Addition of new assessment data ....................................................................................................... 23 Replacement of prior assessment data .............................................................................................. 23 No changes to assessment data.......................................................................................................... 25 3 North America......................................................................................................................................... 27 Addition of new assessment data ....................................................................................................... 27 Replacement of prior assessment data .............................................................................................. 27 No changes .......................................................................................................................................... 27 Additional details ........................................................................................................................................ 27 Country-level estimates of learning poverty, learning deprivation, and schooling deprivation. ............... 28 References .................................................................................................................................................. 34 4 Introduction The release of new learning assessment results, TIMSS 2019, SEA-PLM 2019, and PASEC 2019, offers more recent learning data and calls for changes to the learning poverty numbers. Significant changes are seen in some country-level estimates where international and regional learning assessments replace national ones. The learning poverty estimates have been updated and revised for the first time in July 2021 since the initial launch of the measure in October 2019 by the World Bank and UNESCO. This note only covers the changes to country-level estimates of learning poverty, and does not include changes to the regional and global estimates, which are planned for Fall 2021, and will include the forthcoming LLECE 2019 results and results for a number of countries that are currently participating in policy linking workshops to enable usage of national assessments to report against SDG 4.1.1. The process for changing country-level estimates is decoupled from the process for changing global and regional estimates, as they have different protocols and timelines: • Country-level updates or revisions* can be done biannually during March and September. • Global or regional revisions can be done once a year by September. This is an opportunity to change previously published aggregates reflecting changes in income group classification, revisions of the underlying school enrollment data, or any validated requests to revise national learning deprivation estimates. • Global or regional updates are done every two years in September. We need a minimum number of new assessments to update a global/regional number. *Update refers to a new number for a country-year with no previous data, and revision refers to a new number for a country-year with previous data. The following sections provide a comprehensive overview of the latest country-level learning poverty estimates and flags significant changes for all countries (Part I and Part II) that have a prior or new learning poverty estimate. Changes to the country-level estimates are reflected in the latest two-pager country briefs, which can be accessed here. In October 2019, country-level estimates were produced for 114 countries, and the estimates for 62 low- and middle-income countries were used to calculate the global and regional estimates (World Bank, 2019). In July 2021, country-level estimates are produced for 120 countries. Note that the country-level briefs contain learning poverty estimates for all countries with any available assessment data that can be used to calculate learning poverty, regardless of assessment year and whether it falls inside the new reporting window. Therefore, country numbers that were previously included in the October 2019 release and are not revised or updated in the current round will continue to be available in the country-level database. However, the Fall 2021 global and regional update will only use assessment data for countries that fall inside the new reporting window, anchored around 2017 instead of 2015, to calculate the aggregated regional and global estimates. How is learning poverty measured? The learning poverty (LP) indicator combines the share of primary-aged children out-of-school who are schooling deprived (SD), and the share of pupils below a minimum proficiency in reading, who are 5 learning deprived (LD). By combining schooling and learning, the indicator brings into focus both “more schooling�, which by itself serves a variety of critical functions, as well as “better learning� which is important to ensure that time spent in school translates into acquisition of skills and capabilities. 𝑳𝑷 = 𝑺𝑫 + [(� − 𝑺𝑫) ∗ 𝑳𝑫] where LP is Learning Poverty; LD, Learning Deprivation, is the share of children at the end of primary below minimum proficiency, as defined by the Global Alliance to Monitor Learning (GAML) in the context of the SDG 4.1.1b monitoring; SD, Schooling Deprivation, is the share of primary-aged children who are out-of-school, and is linked to SDG 4.1.4. New assessment programs used to measure learning deprivation Data for three international and regional assessments were released in December 2020. This new learning data has enabled updates and revisions to some country-level learning poverty estimates. The new assessments are: TIMSS 2019 TIMSS 2019 is an international assessment covering 64 countries and 8 benchmarking systems in multiple regions across the world (TIMSS 2019). The 2015 round covered 57 countries (TIMSS 2015). TIMSS tests children in mathematics and science in grades 4 and 8. We use results for grade 4 science for calculating learning poverty measures.1 The field work for the assessment started in 2018 and ended in 2019 (IEA 2019). We use TIMSS 2019 results for calculating learning poverty measures for 10 countries in the latest release that do not have an international (for example, PIRLS, which is generally prioritized for calculating learning poverty when available) or regional assessment in reading: Albania, Armenia, Croatia, Cyprus, Japan, Korea Rep, Montenegro, North Macedonia, Serbia, and Turkey. SEA-PLM 2019 SEA-PLM 2019 is a regional assessment covering 6 countries in Southeast Asia. It tests children in the subjects of reading, writing, mathematics, and global citizenship in grade 5 (SEA-PLM and SEAMEO 2019). We use results for grade 5 reading for calculating learning poverty measures. The field work for the assessment started in 2019 and ended in 2019 (SEA-PLM 2019). The first round of the assessment done in 2019 is used for the learning poverty update. We use SEA-PLM 2019 results for calculating learning poverty measures for 6 countries in the latest release: Cambodia, Lao PDR, Malaysia, Myanmar, Philippines, and Vietnam. PASEC 2019 PASEC 2019 is a regional assessment covering 14 countries in Sub-Saharan Africa. The 2014 round covered 10 countries (Confemen 2014). PASEC tests children in the subjects of language, mathematics, and reading in grades 2 and 6 (Confemen 2019). We use results for grade 6 reading for calculating learning poverty measures. The field work for the assessment started in 2018 and ended in 2019 (PASEC 2019). We use PASEC 2019 results for calculating learning poverty measures for 14 countries in the latest release: Benin, Burkina Faso, Burundi, Cameroon, Chad, Congo Dem Rep, Congo Rep, Cote d’Ivoire, Gabon, Guinea, Madagascar, Niger, Senegal, and Togo. 1 The only exception is Jordan where we use TIMSS math instead due to the unavailability of TIMSS science. 6 Additionally, more recent national learning assessment data is available for some countries, such as Bangladesh. In case a country has multiple assessments, the following hierarchy is followed for choosing an assessment: Table 1. Hierarchy for selecting among multiple assessments in each country Priority rank 1 International assessment in reading For example, global assessments such as PIRLS, or regional assessments such as LLECE, PASEC, or SEA-PLM with reading results Priority rank 2 International assessment in a subject other than reading For example, TIMSS results for science Priority rank 3 National learning assessment (interim reporting) Notes: (1) In addition to the assessment hierarchy, the “reporting window� also determines what assessment is used for a country. For example, under the new reporting window of 2013-2021 centered around 2017, TIMSS 2015 is prioritized over PIRLS 2011 for Indonesia and Croatia in the latest release because PIRLS 2011 falls outside the reporting window. (2) The use of national learning assessment is a temporary measure in the absence of a regional or international assessment. (3) Despite the availability of TIMSS 2019 data for Pakistan, reading data from the national learning assessment 2014 is used to calculate learning poverty in the new release due to on-going efforts to potentially include the 2019 national learning assessment in a policy linking exercise, which would enable reporting on SDG 4.1.1, and inform the learning deprivation dimension. Changes to learning poverty, learning deprivation, and schooling deprivation estimates at the country level This section provides a country-by-country overview of changes to the learning poverty, learning deprivation, and schooling deprivation estimates in light of new assessment data. East Asia & Pacific Addition of new assessment data The learning poverty estimates for the following countries have been computed for the first time due to the availability of new assessment data. Lao PDR Lao PDR did not have a prior country-level estimate for learning poverty. In the new country- level update, SEA-PLM 2019 data for reading is used to compute an estimate for learning poverty. Enrollment data for 2017 is used. The new release contains gender disaggregated learning poverty estimates. Myanmar Myanmar did not have a prior country-level estimate for learning poverty. In the new country- level update, SEA-PLM 2019 data for reading is used to compute an estimate for learning poverty. Enrollment data for 2017 is used. The new release does not contain gender disaggregated learning poverty estimates due to missing gender disaggregated enrollment data. 7 Replacement of prior assessment data The learning poverty estimates for the following countries have changed compared to the estimates in the October 2019 release due to the availability of new assessment data. Cambodia (with different assessment) Cambodia had a prior country-level estimate for learning poverty using national learning assessment data 2013 for reading. There is new data for Cambodia from SEA-PLM 2019 for reading. This new data is used for the new country-level learning poverty estimates. Enrollment data for 2017 instead of 2012 is used. The new release contains gender disaggregated learning poverty estimates, unlike the previous release. Indonesia (with different assessment) Indonesia had a prior country-level estimate for learning poverty using PIRLS 2011 data for reading. There is more recent (not new) data for Indonesia from TIMSS 2015 for science, which was not used previously. The more recent data is used for the new country-level learning poverty estimates. Enrollment data for 2014 instead of 2011 is used. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Japan (with same assessment) Japan had a prior country-level estimate for learning poverty using TIMSS 2015 for science. There is new data for Japan from TIMSS 2019 for science. This new data is used for the new country-level learning poverty estimates. Enrollment data for 2016 instead of 2015 is used. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Korea, Rep. (with same assessment) Korea, Rep. had a prior country-level estimate for learning poverty using TIMSS 2015 for science. There is new data for Korea, Rep. from TIMSS 2019 for science. This new data is used for the new country-level learning poverty estimates. Enrollment data for 2016 instead of 2015 is used. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Malaysia (with different assessment) Malaysia had a prior country-level estimate for learning poverty using national learning assessment data 2017 for reading. There is new data for Malaysia from SEA-PLM 2019 for reading. Enrollment data for 2017 is used, as in the previous release. This new data is used for the new country-level learning poverty estimates. The new release contains gender disaggregated learning poverty estimates, unlike the previous release. Philippines (with different assessment) Philippines had a prior country-level estimate for learning poverty using TIMSS 2003 data for science. There is new data for the Philippines from SEA-PLM 2019 for reading. This new data is used for the new country-level learning poverty estimates. Enrollment data for 2016 instead of 8 2003 is used. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Vietnam (with different assessment) Vietnam had a prior country-level estimate for learning poverty using national learning assessment data 2011 for reading. There is new data for Vietnam from SEA-PLM 2019 for reading. This new data is used for the new country-level learning poverty estimates. Enrollment data for 2013 instead of 2011 is used. The new release does not contain gender disaggregated learning poverty estimates due to missing gender disaggregated enrollment data, as in the previous release. No changes to assessment data The learning poverty estimates for the following countries have not changed, as they continue to be calculated using the same round of the same assessment, as done in the October 2019 release. Australia (same round of same assessment) Australia will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. China (same round of same assessment) China will use national learning assessment data 2016 for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release does not contain gender disaggregated learning poverty estimates due to lack of microdata or availability of relevant information in assessment reports, as in the previous release. Hong Kong SAR (same round of same assessment) Hong Kong SAR will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2013 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Macao SAR, China (same round of same assessment) Macao SAR, China will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Mongolia (same round of same assessment) Mongolia will use TIMSS 2007 data for science for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment 9 data for 2007 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. New Zealand (same round of same assessment) New Zealand will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Singapore (same round of same assessment) Singapore will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Thailand (same round of same assessment) Thailand will use TIMSS 2011 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2009 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Europe & Central Asia Addition of new assessment data The learning poverty estimates for the following countries have been computed for the first time due to the availability of new assessment data. Albania Albania did not have a prior country-level estimate for learning poverty. In the new country- level update, TIMSS 2019 data for science is used to compute an estimate for learning poverty. Enrollment data for 2017 is used. The new release contains gender disaggregated learning poverty estimates. Montenegro Montenegro did not have a prior country-level estimate for learning poverty. In the new country-level update, TIMSS 2019 data for science is used to compute an estimate for learning poverty. Enrollment data for 2017 is used. The new release contains gender disaggregated learning poverty estimates. Replacement of prior assessment data The learning poverty estimates for the following countries have changed compared to the estimates in the October 2019 release due to the availability of new assessment data. Armenia (with same assessment) 10 Armenia had a prior country-level estimate for learning poverty using TIMSS 2015 for science. There is new data for Armenia from TIMSS 2019 for science. This new data is used for the new country-level learning poverty estimates. Enrollment data for 2017 instead of 2015 is used. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Croatia (with different assessment) Croatia had a prior country-level estimate for learning poverty using PIRLS 2011 for reading. There is new data for Croatia from TIMSS 2019 for science. This new data is used for the new country-level learning poverty estimates. Enrollment data for 2016 instead of 2011 is used. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Cyprus (new round of same assessment) Cyprus had a prior country-level estimate for learning poverty using TIMSS 2015 for science. There is new data for Cyprus from TIMSS 2019 for science. This new data is used for the new country-level learning poverty estimates. Enrollment data for 2015 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. N. Macedonia (with different assessment) N. Macedonia had a prior country-level estimate for learning poverty using PIRLS 2006 data for reading. In the new country-level update, TIMSS 2019 data for science is used to compute an estimate for learning poverty. Enrollment data for 2015 instead of 2006 is used. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Serbia (with same assessment) Serbia had a prior country-level estimate for learning poverty using TIMSS 2015 for science. There is new data for Serbia from TIMSS 2019 for science. This new data is used for the new country-level learning poverty estimates. Enrollment data for 2017 instead of 2015 is used. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Turkey (with same assessment) Turkey had a prior country-level estimate for learning poverty using TIMSS 2015 for science. There is new data for Turkey from TIMSS 2019 for science. This new data is used for the new country-level learning poverty estimates. Enrollment data for 2016 instead of 2015 is used. The new release contains gender disaggregated learning poverty estimates, as in the previous release. No changes to assessment data The learning poverty estimates for the following countries have not changed, as they continue to be calculated using the same round of the same assessment, as done in the October 2019 release. 11 Austria (same round of same assessment) Austria will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Azerbaijan (same round of same assessment) Azerbaijan will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Belgium (same round of same assessment) Belgium will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Bulgaria (same round of same assessment) Bulgaria will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Czech Republic (same round of same assessment) Czech Republic will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Denmark (same round of same assessment) Denmark will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Finland (same round of same assessment) Finland will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. France (same round of same assessment) 12 France will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Georgia (same round of same assessment) Georgia will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Germany (same round of same assessment) Germany will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Greece (same round of same assessment) Greece will use PIRLS 2001 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2001 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Hungary (same round of same assessment) Hungary will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Iceland (same round of same assessment) Iceland will use PIRLS 2006 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2006 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Ireland (same round of same assessment) Ireland will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Italy (same round of same assessment) Italy will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment 13 data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Kosovo (no learning poverty estimate) Kosovo has no learning poverty estimate in the current or previous release. While Kosovo participated in TIMSS 2019, there is no learning poverty estimate due to missing enrollment data. The new release does not contain gender disaggregated learning poverty estimates due to missing gender disaggregated enrollment data, as in the previous release. Kazakhstan (same round of same assessment) Kazakhstan will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Kyrgyz Republic (same round of same assessment) Kazakhstan will use national learning assessment 2014 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2014 is used, as in the previous release. The new release does not contain gender disaggregated learning poverty estimates due to lack of microdata or availability of relevant information in assessment reports, as in the previous release. Latvia (same round of same assessment) Latvia will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Lithuania (same round of same assessment) Lithuania will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Luxembourg (same round of same assessment) Luxembourg will use PIRLS 2006 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2006 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Moldova (same round of same assessment) Moldova will use PIRLS 2006 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2006 is used, as in the previous release. The new release does not contain gender 14 disaggregated learning poverty estimates due to missing gender disaggregated enrollment data, as in the previous release. Netherlands (same round of same assessment) Netherlands will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 1997 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Norway (same round of same assessment) Norway will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Poland (same round of same assessment) Poland will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Portugal (same round of same assessment) Portugal will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Romania (same round of same assessment) Romania will use PIRLS 2011 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2011 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Russian Federation (same round of same assessment) Russian Federation will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Slovak Republic (same round of same assessment) Slovak Republic will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. 15 Slovenia (same round of same assessment) Slovenia will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Spain (same round of same assessment) Spain will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Sweden (same round of same assessment) Sweden will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Ukraine (same round of same assessment) Ukraine will use TIMSS 2007 data for science for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2007 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. United Kingdom (same round of same assessment) United Kingdom will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Latin America & the Caribbean Addition of new assessment data The learning poverty estimates for the following countries have been computed for the first time due to the availability of new assessment data. NA Replacement of prior assessment data The learning poverty estimates for the following countries have changed compared to the estimates in the October 2019 release due to the availability of new assessment data. NA 16 No changes to assessment data The learning poverty estimates for the following countries have not changed, as they continue to be calculated using the same round of the same assessment, as done in the October 2019 release. Argentina (same round of same assessment) Argentina will use LLECE 2013 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2013 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Belize (same round of same assessment) Belize will use PIRLS 2001 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2001 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Brazil (same round of same assessment) Brazil will use LLECE 2013 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2013 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Chile (same round of same assessment) Chile will use LLECE 2013 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2013 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Colombia (same round of same assessment) Colombia will use LLECE 2013 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2013 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Costa Rica (same round of same assessment) Costa Rica will use LLECE 2013 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2006 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Dominican Republic (same round of same assessment) 17 Dominican Republic will use LLECE 2013 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2013 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Ecuador (same round of same assessment) Ecuador will use LLECE 2013 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2013 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. El Salvador (same round of same assessment) El Salvador will use TIMSS 2007 data for science for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2007 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Guatemala (same round of same assessment) Guatemala will use LLECE 2013 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2013 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Honduras (same round of same assessment) Honduras will use LLECE 2013 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2013 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Mexico (same round of same assessment) Mexico will use LLECE 2013 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2013 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Nicaragua (same round of same assessment) Nicaragua will use LLECE 2013 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2010 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Panama (same round of same assessment) Panama will use LLECE 2013 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment 18 data for 2013 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Paraguay (same round of same assessment) Paraguay will use LLECE 2013 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2012 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Peru (same round of same assessment) Peru will use LLECE 2013 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2013 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Trinidad and Tobago (same round of same assessment) Trinidad and Tobago will use LLECE 2013 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2010 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Uruguay (same round of same assessment) Uruguay will use LLECE 2013 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2013 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Middle East & North Africa Addition of new assessment data The learning poverty estimates for the following countries have been computed for the first time due to the availability of new assessment data. NA Replacement of prior assessment data The learning poverty estimates for the following countries have changed compared to the estimates in the October 2019 release due to the availability of new assessment data. NA No changes to assessment data The learning poverty estimates for the following countries have not changed, as they continue to be calculated using the same round of the same assessment, as done in the October 2019 release. 19 Algeria (same round of same assessment) Algeria will use TIMSS 2007 data for science for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2007 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Bahrain (same round of same assessment) Bahrain will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Egypt, Arab Rep. (same round of same assessment) Egypt, Arab Rep. will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Iran, Islamic Rep. (same round of same assessment) Iran, Islamic Rep. will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Israel (same round of same assessment) Israel will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Jordan (same round of same assessment) Jordan will use TIMSS 2015 data for math for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2004 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Kuwait (same round of same assessment) Kuwait will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Malta (same round of same assessment) 20 Malta will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release does not contain gender disaggregated learning poverty estimates due to missing gender disaggregated enrollment data, as in the previous release. Morocco (same round of same assessment) Morocco will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Oman (same round of same assessment) Oman will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Qatar (same round of same assessment) Qatar will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Saudi Arabia (same round of same assessment) Saudi Arabia will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2014 is used, as in the previous release. The new release does not contain gender disaggregated learning poverty estimates due to missing gender disaggregated enrollment data, as in the previous release. Tunisia (same round of same assessment) Tunisia will use TIMSS 2011 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2011 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Yemen (same round of same assessment) Yemen will use TIMSS 2011 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2010 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. United Arab Emirates (same round of same assessment) 21 United Arab Emirates will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. South Asia Addition of new assessment data The learning poverty estimates for the following countries have been computed for the first time due to the availability of new assessment data. NA Replacement of prior assessment data The learning poverty estimates for the following countries have changed compared to the estimates in the October 2019 release due to the availability of new assessment data. Bangladesh (with same assessment) Bangladesh had a prior country-level estimate for learning poverty using national learning assessment 2015 for reading. There is new data for Bangladesh from national learning assessment 2017 for reading.2 This new data is used for the new country-level learning poverty estimates. Enrollment data for 2017 is used, as in the previous release. The new release does not contain gender disaggregated learning poverty estimates due to lack of microdata or availability of relevant information in assessment reports, as in the previous release. No changes to assessment data The learning poverty estimates for the following countries have not changed, as they continue to be calculated using the same round of the same assessment, as done in the October 2019 release.3 Afghanistan (same round of same assessment) Afghanistan will use national learning assessment 2013 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2017 is used, as in the previous release. The new release does not contain gender disaggregated learning poverty estimates due to lack of microdata or availability of relevant information in assessment reports, and due to missing gender disaggregated enrollment data, as in the previous release. India (same round of same assessment) 2 While the 2015 and 2017 tests were put on a common scale by using a methodology linking test items, the comparability of the 2017 national learning assessment may be affected by the fact that the test administration was carried out in January 2018 instead of November 2017 (Directorate of Primary Education, 2018). 3 The only exception in India where we use the same assessment results as used in the October 2019 results, but learning poverty estimate has changed due to a change in the enrollment data. 22 India will use national learning assessment 2017 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2017 instead of 2013 is used. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Pakistan (same round of same assessment) Pakistan will use national learning assessment 2014 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2014 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, unlike the previous release. Sri Lanka (same round of same assessment) Sri Lanka will use national learning assessment 2015 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2015 is used, as in the previous release. The new release does not contain gender disaggregated learning poverty estimates due to lack of microdata or availability of relevant information in assessment reports, as in the previous release. Sub-Saharan Africa Addition of new assessment data The learning poverty estimates for the following countries have been computed for the first time due to the availability of new assessment data. Guinea Guinea did not have a prior country-level estimate for learning poverty. In the new country-level update, PASEC 2019 data for reading is used to compute an estimate for learning poverty. Enrollment data for 2016 is used. The new release contains gender disaggregated learning poverty estimates. Replacement of prior assessment data The learning poverty estimates for the following countries have changed compared to the estimates in the October 2019 release due to the availability of new assessment data. Benin (with same assessment) Benin had a prior country-level estimate for learning poverty using PASEC 2014 for reading. There is new data for Benin from PASEC 2019 for reading. This new data is used for the new country-level learning poverty estimates. Enrollment data for 2017 instead of 2014 is used. The new release does not contain gender disaggregated learning poverty estimates due to missing gender disaggregated enrollment data, as in the previous release. Burkina Faso (with same assessment) Burkina Faso had a prior country-level estimate for learning poverty using PASEC 2014 for reading. There is new data for Burkina Faso from PASEC 2019 for reading. This new data is used 23 for the new country-level learning poverty estimates. Enrollment data for 2017 instead of 2014 is used. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Burundi (with same assessment) Burundi had a prior country-level estimate for learning poverty using PASEC 2014 for reading. There is new data for Burundi from PASEC 2019 for reading. This new data is used for the new country-level learning poverty estimates. Enrollment data for 2017 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Cameroon (with same assessment) Cameroon had a prior country-level estimate for learning poverty using PASEC 2014 for reading. There is new data for Cameroon from PASEC 2019 for reading. This new data is used for the new country-level learning poverty estimates. Enrollment data for 2017 instead of 2014 is used. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Chad (with same assessment) Chad had a prior country-level estimate for learning poverty using PASEC 2014 for reading. There is new data for Chad from PASEC 2019 for reading. This new data is used for the new country-level learning poverty estimates. Enrollment data for 2016 instead of 2013 is used. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Congo, D.R. (with same assessment) Congo, D.R. had a prior country-level estimate for learning poverty using PASEC 2010 for reading. There is new data for Congo, D.R. from PASEC 2019 for reading. This new data is used for the new country-level learning poverty estimates. Enrollment data for 1999 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, unlike the previous release. Congo, Rep. (with same assessment) Congo, Rep. had a prior country-level estimate for learning poverty using PASEC 2014 for reading. There is new data for Congo, Rep. from PASEC 2019 for reading. This new data is used for the new country-level learning poverty estimates. Enrollment data for 2012 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Cote d’Ivoire (with same assessment) Cote d’Ivoire had a prior country-level estimate for learning poverty using PASEC 2014 for reading. There is new data for Cote d’Ivoire from PASEC 2019 for reading. This new data is used for the new country-level learning poverty estimates. Enrollment data for 2014 is used, as in the 24 previous release. The new release does not contain gender disaggregated learning poverty estimates due to missing gender disaggregated enrollment data, as in the previous release. Gabon (with same assessment) Gabon had a prior country-level estimate for learning poverty using PASEC 2006 data for reading. There is new data for Cote d’Ivoire from PASEC 2019 for reading. This new data is used for the new country-level learning poverty estimates. Enrollment data for 1997 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, unlike the previous release. Madagascar (with different assessment) Madagascar had a prior country-level estimate for learning poverty using PASEC 2015 for reading. There is new data for Madagascar from PASEC 2019 for reading. This new data is used for the new country-level learning poverty estimates. Enrollment data for 2003 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, unlike the previous release. Niger (with same assessment) Niger had a prior country-level estimate for learning poverty using PASEC 2014 for reading. There is new data for Niger from PASEC 2019 for reading. This new data is used for the new country-level poverty estimates. Enrollment data for 2017 instead of 2014 is used. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Senegal (with same assessment) Senegal had a prior country-level estimate for learning poverty using PASEC 2014 for reading. There is new data for Senegal from PASEC 2019 for reading. This new data is used for the new country-level learning poverty estimates. Enrollment data for 2017 instead of 2014 is used. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Togo (with same assessment) Togo had a prior country-level estimate for learning poverty using PASEC 2014 for reading. There is new data for Togo from PASEC 2019 for reading. This new data is used for the new country-level learning poverty estimates. Enrollment data for 2017 instead of 2014 is used. The new release contains gender disaggregated learning poverty estimates, as in the previous release. No changes to assessment data The learning poverty estimates for the following countries have not changed, as they continue to be calculated using the same round of the same assessment, as done in the October 2019 release. Botswana 25 Botswana will use PIRLS 2011 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2012 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Comoros (same round of same assessment) Comoros will use PASEC 2008 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2007 is used, as in the previous release. The new release does not contain gender disaggregated learning poverty estimates due to lack of microdata or availability of relevant information in assessment reports, as in the previous release. Ethiopia (same round of same assessment) Ethiopia will use national learning assessment 2015 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2015 is used, as in the previous release. The new release does not contain gender disaggregated learning poverty estimates due to lack of microdata or availability of relevant information in assessment reports, as in the previous release. Mali (same round of same assessment) Mali will use PASEC 2012 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2012 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Mauritania (same round of same assessment) Mauritania will use PASEC 2004 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2004 is used, as in the previous release. The new release does not contain gender disaggregated learning poverty estimates due to lack of microdata or availability of relevant information in assessment reports, as in the previous release. Mauritius (same round of same assessment) Mauritius will use PASEC 2006 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2006 is used, as in the previous release. The new release does not contain gender disaggregated learning poverty estimates due to lack of microdata or availability of relevant information in assessment reports, as in the previous release. South Africa (same round of same assessment) South Africa will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2015 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. 26 Uganda (same round of same assessment) Uganda will use national learning assessment 2014 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2013 is used, as in the previous release. The new release does not contain gender disaggregated learning poverty estimates due to lack of microdata or availability of relevant information in assessment reports, as in the previous release. North America Addition of new assessment data The learning poverty estimates for the following countries have been computed for the first time due to the availability of new assessment data. NA Replacement of prior assessment data The learning poverty estimates for the following countries have changed compared to the estimates in the October 2019 release due to the availability of new assessment data. NA No changes The learning poverty estimates for the following countries have not changed, as they continue to be calculated using the same round of the same assessment, as done in the October 2019 release. Canada (same round of same assessment) Canada will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. USA (same round of same assessment) USA will use PIRLS 2016 data for reading for the new learning poverty estimates. The same results were used to calculate learning poverty estimates in the previous release. Enrollment data for 2016 is used, as in the previous release. The new release contains gender disaggregated learning poverty estimates, as in the previous release. Additional details This section provides further information about changes to the learning poverty estimates. Data on school participation used to measure schooling deprivation The out-of-school adjustment in learning poverty relies on enrollment data. Our preferred definition is the adjusted net primary enrollment as reported by the UNESCO Institute for Statistics (UIS). This data 27 relies both on the population census and the EMIS (Education Management Information System). We use the same year of school participation as the preferred learning assessment for each country, depending on availability, which means enrollment data for countries with new assessments may have been updated in the calculation of their learning poverty estimate. Details of these updates related to the enrollment year can be found in the earlier section, and in the accompanying data base and learning poverty country briefs. In the case of India, the total net enrollment rate for 2017 from India’s National Institute of Educational Planning and Administration and verified by the government of India is used in the latest release for the out-of-school adjustment in the learning poverty calculation, even though there are no changes to the assessment data. Previously, adjusted net primary enrollment, as reported by UIS for 2013, was used to calculate India’s learning poverty estimates. Estimating learning scores from microdata Score estimates from raw assessment data are computed using the repest command in Stata as documented by Francesco Avvisati and François Keslair (2014) in the new release. In the old release, score estimates are computed using pv command as documented by Kevin Macdonald (2008). This shift in methodology makes negligible difference to the scores calculated. Country-level estimates of learning poverty, learning deprivation, and schooling deprivation. Table 1 provides a summary of the country-level changes in learning poverty. The highlighted rows indicate countries where the learning poverty estimates have changed due to inclusion of new assessment data. It is important to be careful when comparing changes in learning poverty estimates over time, as the estimates will not be comparable for countries where different assessments are used to calculate the estimates across the two rounds. Note that in the latest accompanying data base for learning poverty, we only includes estimates for each country that are temporally comparable. As a result, prior results from different assessments that are not temporally comparable to the current estimates are removed. For example, in countries in East Asia that switched from national learning assessments to SEA-PLM 2019, only the SEA-PLM results are available because the national learning assessments are not temporally comparable to the regional assessment and were only used for interim reporting. Others countries, such as Croatia, N. Macedonia, and Indonesia, switched from PIRLS to TIMSS, in which case the TIMSS results are shown for reasons related to temporal comparability. Table 1. Comparison of country-level learning poverty estimates across the July 2021 and October 2019 releases July 2021 release October 2019 release Country LP LD SD Assess- Assess- Enroll- LP LD SD Assess- Assess- Enroll- ment ment ment ment ment ment estimate estimate estimate estimate estimate estimate used year year used year year Afghanistan 93.4 87.0 49.6 NLA 2013 2016 93.4 87.0 49.6 NLA 2013 2016 Albania 16.5 14.0 2.9 TIMSS 2019 2017 4.4 None 2015 Algeria 67.9 66.5 4.1 TIMSS 2007 2007 67.9 66.5 4.1 TIMSS 2007 2007 American None None Samoa 28 Andorra None None Angola 22.5 None 2011 22.5 None 2011 Antigua and 17.5 None 2017 18.5 None 2015 Barbuda Argentina 53.9 53.6 0.6 LLECE 2013 2013 53.9 53.6 0.6 LLECE 2013 2013 Armenia 26.1 20.0 7.6 TIMSS 2019 2017 35.0 30.0 7.2 TIMSS 2015 2015 Aruba 0.7 None 2014 0.7 None 2014 Australia 8.6 5.5 3.2 PIRLS 2016 2016 8.6 5.5 3.2 PIRLS 2016 2016 Austria 2.4 2.4 0.0 PIRLS 2016 2016 2.4 2.4 0.0 PIRLS 2016 2016 Azerbaijan 23.3 19.2 5.0 PIRLS 2016 2016 23.3 19.2 5.0 PIRLS 2016 2016 Bahamas, 11.6 None 2016 11.6 None 2016 The Bahrain 32.1 30.6 2.1 PIRLS 2016 2016 32.1 30.6 2.1 PIRLS 2016 2016 Bangladesh 58.1 56.0 4.9 NLA 2017 2017 57.2 55.0 4.9 NLA 2015 2017 Barbados 9.6 None 2017 7.5 None 2015 Belarus 4.0 None 2017 5.4 None 2015 Belgium 6.4 5.1 1.3 PIRLS 2016 2016 6.4 5.1 1.3 PIRLS 2016 2016 Belize 76.4 74.8 6.5 PIRLS 2001 2001 76.4 74.8 6.5 PIRLS 2001 2001 Benin 55.8 54.5 3.0 PASEC 2019 2017 78.2 77.3 3.6 PASEC 2014 2014 Bermuda 16.5 None 2012 16.5 None 2012 Bhutan 18.1 None 2017 15.4 None 2015 Bolivia 7.5 None 2017 10.1 None 2015 Bosnia and 22 TIMSS 2019 None Herzegovina Botswana 48.3 44.3 7.2 PIRLS 2011 2012 48.3 44.3 7.2 PIRLS 2011 2012 Brazil 48.4 46.9 2.7 LLECE 2013 2013 48.4 46.9 2.7 LLECE 2013 2013 British Virgin 3.6 None 2016 1.0 None 2015 Islands Brunei 3.6 None 2017 3.7 None 2015 Darussalam Bulgaria 11.7 5.2 6.8 PIRLS 2016 2016 11.7 5.2 6.8 PIRLS 2016 2016 Burkina Faso 74.6 67.0 23.0 PASEC 2019 2017 85.4 78.6 31.7 PASEC 2014 2014 Burundi 95.6 95.5 2.7 PASEC 2019 2017 92.9 92.7 2.7 PASEC 2014 2017 Cabo Verde 13.7 None 2017 13.0 None 2015 SEA- Cambodia 90.0 89.0 9.4 2019 2017 51.1 49.8 2.6 NLA 2013 2012 PLM Cameroon 71.2 69.8 4.8 PASEC 2019 2017 77.2 75.9 5.2 PASEC 2014 2014 Canada 4.3 4.3 0.0 PIRLS 2016 2016 4.3 4.3 0.0 PIRLS 2016 2016 Cayman None None Islands Central African 31.6 None 2012 31.6 None 2012 Republic Chad 94.3 92.4 25.5 PASEC 2019 2016 97.7 97.0 21.1 PASEC 2014 2013 Channel None None Islands Chile 36.8 30.3 9.3 LLECE 2013 2013 36.8 30.3 9.3 LLECE 2013 2013 China 18.2 18.2 0.0 NLA 2016 2016 18.2 18.2 0.0 NLA 2016 2016 29 Colombia 48.6 44.7 6.9 LLECE 2013 2013 48.6 44.7 6.9 LLECE 2013 2013 Comoros 86.0 82.3 20.8 PASEC 2008 2007 86.0 82.3 20.8 PASEC 2008 2007 Congo, Dem. 96.6 90.8 63.2 PASEC 2019 1999 86.0 62.0 63.2 PASEC 2010 1999 Rep. Congo, Rep. 70.7 66.4 12.8 PASEC 2019 2012 85.1 82.9 12.8 PASEC 2014 2012 Costa Rica 32.5 31.7 1.1 LLECE 2013 2006 32.5 31.7 1.1 LLECE 2013 2006 Cote d'Ivoire 82.6 78.0 21.1 PASEC 2019 2014 82.3 77.6 21.1 PASEC 2014 2014 Croatia 4.3 2.0 2.4 TIMSS 2019 2016 4.0 1.0 3.0 PIRLS 2011 2011 Cuba Curacao 0.0 None 2013 0.0 None 2013 Cyprus 10.0 8.0 2.2 TIMSS 2019 2015 16.2 14.3 2.2 TIMSS 2015 2015 Czech 3.0 3.0 0.0 PIRLS 2016 2016 3.0 3.0 0.0 PIRLS 2016 2016 Republic Denmark 3.6 2.6 1.0 PIRLS 2016 2016 3.6 2.6 1.0 PIRLS 2016 2016 Djibouti 44.9 None 2017 44.4 None 2015 Dominica 1.8 None 2016 1.8 None 2016 Dominican 80.7 79.4 6.6 LLECE 2013 2013 80.7 79.4 6.6 LLECE 2013 2013 Republic Ecuador 62.8 62.1 1.9 LLECE 2013 2013 62.8 62.1 1.9 LLECE 2013 2013 Egypt, Arab 69.6 69.2 1.4 PIRLS 2016 2016 69.6 69.2 1.4 PIRLS 2016 2016 Rep. El Salvador 55.0 53.0 4.2 TIMSS 2007 2007 55.0 53.0 4.2 TIMSS 2007 2007 Equatorial 55.8 None 2015 55.8 None 2015 Guinea Eritrea 62.3 None 2017 57.4 None 2015 Estonia 6.3 None 2016 6.4 None 2015 Eswatini 24.1 None 2016 23.7 None 2015 Ethiopia 90.3 88.7 14.0 NLA 2015 2015 90.3 88.7 14.0 NLA 2015 2015 Faroe Islands None None Fiji 0.1 None 2016 2.0 None 2015 Finland 2.6 1.7 0.9 PIRLS 2016 2016 2.6 1.7 0.9 PIRLS 2016 2016 France 7.1 6.3 0.9 PIRLS 2016 2016 7.1 6.3 0.9 PIRLS 2016 2016 French 0.6 None 1996 0.6 None 1996 Polynesia Gabon 30.4 23.7 8.7 PASEC 2019 1997 36.8 30.8 8.7 PASEC 2006 1997 Gambia, The 21.4 None 2017 26.0 None 2015 Georgia 13.8 13.5 0.4 PIRLS 2016 2016 13.8 13.5 0.4 PIRLS 2016 2016 Germany 5.7 5.5 0.2 PIRLS 2016 2016 5.7 5.5 0.2 PIRLS 2016 2016 Ghana 14.9 None 2017 9.9 None 2015 Gibraltar None None Greece 10.6 5.5 5.4 PIRLS 2001 2001 10.6 5.5 5.4 PIRLS 2001 2001 Greenland None None Grenada 3.2 None 2016 2.2 None 2015 Guam None None Guatemala 67.3 63.6 10.1 LLECE 2013 2013 67.3 63.6 10.1 LLECE 2013 2013 Guinea 82.5 77.8 21.1 PASEC 2019 2016 19.7 None 2014 30 Guinea- 28.1 None 2010 28.1 None 2010 Bissau Guyana 4.4 None 2012 4.4 None 2012 Haiti 42.7 None 1997 42.7 None 1997 Honduras 74.7 69.4 17.1 LLECE 2013 2013 74.7 69.4 17.1 LLECE 2013 2013 Hong Kong 3.2 1.4 1.9 PIRLS 2016 2016 3.2 1.4 1.9 PIRLS 2016 2016 SAR, China Hungary 5.9 2.9 3.1 PIRLS 2016 2016 5.9 2.9 3.1 PIRLS 2016 2016 Iceland 9.3 6.8 2.7 PIRLS 2006 2006 9.3 6.8 2.7 PIRLS 2006 2006 India 56.1 53.7 5.1 NLA 2017 2017 54.8 53.7 2.3 NLA 2017 2013 Indonesia 53.4 49.4 8.0 TIMSS 2015 2014 35.4 33.8 2.4 PIRLS 2011 2011 Iran, Islamic 35.7 35.1 0.9 PIRLS 2016 2016 35.7 35.1 0.9 PIRLS 2016 2016 Rep. Iraq 7.7 None 2007 7.7 None 2007 Ireland 2.3 2.3 0.0 PIRLS 2016 2016 2.3 2.3 0.0 PIRLS 2016 2016 Isle of Man None None Israel 11.7 9.0 2.9 PIRLS 2016 2016 11.7 9.0 2.9 PIRLS 2016 2016 Italy 3.5 2.1 1.4 PIRLS 2016 2016 3.5 2.1 1.4 PIRLS 2016 2016 Jamaica 7.7 None 2004 7.7 None 2004 Japan 3.7 2.0 1.8 TIMSS 2019 2016 2.2 1.0 1.2 TIMSS 2015 2015 Jordan 52.0 50.0 4.0 TIMSS 2015 2004 52.0 50.0 4.0 TIMSS 2015 2004 Kazakhstan 2.2 1.9 0.3 PIRLS 2016 2016 2.2 1.9 0.3 PIRLS 2016 2016 Kenya 16.9 None 2012 16.9 None 2012 Kiribati 3.5 None 2017 3.1 None 2015 Korea, Dem. People’s 3.4 None 2009 3.4 None 2009 Rep. Korea, Rep. 4.4 1.0 3.5 TIMSS 2019 2016 3.0 0.3 2.7 TIMSS 2015 2015 Kosovo 41.0 TIMSS 2019 None Kuwait 51.0 49.4 3.3 PIRLS 2016 2016 51.0 49.4 3.3 PIRLS 2016 2016 Kyrgyz 64.5 63.8 1.9 NLA 2014 2014 64.5 63.8 1.9 NLA 2014 2014 Republic SEA- Lao PDR 98.1 98.0 6.7 2019 2017 4.4 None 2015 2015 PLM Latvia 4.0 0.8 3.2 PIRLS 2016 2016 4.0 0.8 3.2 PIRLS 2016 2016 Lebanon 11.5 None 2017 16.4 None 2015 Lesotho 18.4 None 2017 19.9 None 2015 Liberia 63.2 None 2016 62.3 None 2015 Libya 0.0 None 2006 0.0 None 2006 Liechtenstei 0.7 None 2016 0.2 None 2015 n Lithuania 3.0 2.7 0.3 PIRLS 2016 2016 3.0 2.7 0.3 PIRLS 2016 2016 Luxembourg 3.0 1.2 1.7 PIRLS 2006 2006 3.0 1.2 1.7 PIRLS 2006 2006 Macao SAR, 3.7 2.4 1.3 PIRLS 2016 2016 3.7 2.4 1.3 PIRLS 2016 2016 China Madagascar 95.1 93.7 21.9 PASEC 2019 2003 96.7 95.8 21.9 PASEC 2015 2003 Malawi 3.2 None 2009 3.2 None 2009 31 SEA- Malaysia 42.8 42.0 1.4 2019 2017 12.9 11.7 1.4 NLA 2017 2017 PLM Maldives 0.5 None 2017 3.5 None 2015 Mali 90.5 85.7 33.0 PASEC 2012 2012 90.5 85.7 33.0 PASEC 2012 2012 Malta 28.6 26.8 2.4 PIRLS 2016 2016 28.6 26.8 2.4 PIRLS 2016 2016 Marshall 21.5 None 2016 21.6 None 2015 Islands Mauritania 94.9 92.9 28.1 PASEC 2004 2004 94.9 92.9 28.1 PASEC 2004 2004 Mauritius 40.5 38.0 4.0 PASEC 2006 2006 40.5 38.0 4.0 PASEC 2006 2006 Mexico 43.2 42.5 1.2 LLECE 2013 2013 43.2 42.5 1.2 LLECE 2013 2013 Micronesia, 16.0 None 2015 16.0 None 2015 Fed. Sts. Moldova 11.0 8.7 2.5 PIRLS 2006 2006 11.0 8.7 2.5 PIRLS 2006 2006 Monaco None None Mongolia 39.5 38.1 2.3 TIMSS 2007 2007 39.5 38.1 2.3 TIMSS 2007 2007 Montenegro 27.8 25.0 3.7 TIMSS 2019 2017 5.2 None 2015 2015 Morocco 65.8 63.8 5.4 PIRLS 2016 2016 65.8 63.8 5.4 PIRLS 2016 2016 Mozambiqu 12.5 None 2017 10.4 None 2015 e SEA- Myanmar 89.3 89.0 2.3 2019 2017 3.8 None 2015 2014 PLM Namibia 2.2 None 2017 9.4 None 2013 Nauru 15.6 None 2016 12.7 None 2014 Nepal 5.2 None 2017 3.2 None 2015 Netherlands 1.6 1.3 0.3 PIRLS 2016 1997 1.6 1.3 0.3 PIRLS 2016 1997 New None None Caledonia New Zealand 11.4 10.0 1.5 PIRLS 2016 2016 11.4 10.0 1.5 PIRLS 2016 2016 Nicaragua 69.8 69.3 1.6 LLECE 2013 2010 69.8 69.3 1.6 LLECE 2013 2010 Niger 90.4 85.6 33.2 PASEC 2019 2017 98.7 97.9 38.9 PASEC 2014 2014 Nigeria 34.1 None 2010 34.1 None 2010 North 43.1 38.0 8.3 TIMSS 2019 2015 39.7 34.2 8.3 PIRLS 2006 2006 Macedonia Northern Mariana None None Islands Norway 6.0 5.8 0.2 PIRLS 2016 2016 6.0 5.8 0.2 PIRLS 2016 2016 Oman 41.8 40.9 1.5 PIRLS 2016 2016 41.8 40.9 1.5 PIRLS 2016 2016 Pakistan 74.5 65.0 27.3 NLA 2014 2014 74.5 65.0 27.3 NLA 2014 2014 Palau 0.6 None 2014 0.6 None 2014 Panama 66.6 64.1 7.1 LLECE 2013 2013 66.6 64.1 7.1 LLECE 2013 2013 Papua New 22.3 None 2016 22.3 None 2016 Guinea Paraguay 74.4 71.3 10.8 LLECE 2013 2012 74.4 71.3 10.8 LLECE 2013 2012 Peru 55.7 53.7 4.2 LLECE 2013 2013 55.7 53.7 4.2 LLECE 2013 2013 SEA- Philippines 90.5 90.0 4.5 2019 2016 69.5 66.3 9.6 TIMSS 2003 2003 PLM Poland 6.3 2.0 4.4 PIRLS 2016 2016 6.3 2.0 4.4 PIRLS 2016 2016 Portugal 6.5 3.0 3.6 PIRLS 2016 2016 6.5 3.0 3.6 PIRLS 2016 2016 32 Puerto Rico 24.0 None 2016 18.5 None 2015 Qatar 35.3 33.8 2.2 PIRLS 2016 2016 35.3 33.8 2.2 PIRLS 2016 2016 Romania 20.0 14.1 6.9 PIRLS 2011 2011 20.0 14.1 6.9 PIRLS 2011 2011 Russian 3.3 0.9 2.4 PIRLS 2016 2016 3.3 0.9 2.4 PIRLS 2016 2016 Federation Rwanda 6.1 None 2017 4.1 None 2016 Samoa 3.8 None 2017 2.7 None 2015 San Marino 6.9 None 2012 6.9 None 2012 Sao Tome 3.0 None 2017 8.6 None 2015 and Principe Saudi Arabia 38 .3 36.7 2.5 PIRLS 2016 2014 38.3 36.7 2.5 PIRLS 2016 2014 Senegal 69.1 58.9 24.8 PASEC 2019 2017 74.1 65.2 25.7 PASEC 2014 2014 Serbia 9.3 8.0 1.4 TIMSS 2019 2017 8.1 7.4 0.8 TIMSS 2015 2015 Seychelles 4.5 None 2005 4.5 None 2005 Sierra Leone 0.8 None 2016 5.0 None 2015 Singapore 2.8 2.7 0.1 PIRLS 2016 2016 2.8 2.7 0.1 PIRLS 2016 2016 Sint Maarten None None (Dutch part) Slovak 8.5 6.6 2.1 PIRLS 2016 2016 8.5 6.6 2.1 PIRLS 2016 2016 Republic Slovenia 5.8 3.7 2.2 PIRLS 2016 2016 5.8 3.7 2.2 PIRLS 2016 2016 Solomon 30.5 None 2017 29.4 None 2015 Islands Somalia 76.5 None 2007 76.5 None 2007 South Africa 79.8 77.9 8.4 PIRLS 2016 2015 79.8 77.9 8.4 PIRLS 2016 2015 South Sudan 67.8 None 2015 67.8 None 2015 Spain 4.9 3.4 1.5 PIRLS 2016 2016 4.9 3.4 1.5 PIRLS 2016 2016 Sri Lanka 14.8 14.0 0.9 NLA 2015 2015 14.8 14.0 0.9 NLA 2015 2015 St. Kitts and None None Nevis St. Lucia 5.5 None 2007 5.5 None 2007 St. Martin None None (French part) St. Vincent and the 1.7 None 2017 1.2 None 2015 Grenadines Sudan 37.9 None 2017 42.7 None 2015 Suriname 9.8 None 2017 6.6 None 2015 Sweden 2.3 1.9 0.4 PIRLS 2016 2016 2.3 1.9 0.4 PIRLS 2016 2016 Switzerland 0.2 None 2016 0.4 None 2015 Syrian Arab 32.8 None 2013 32.8 None 2013 Republic Tajikistan 1.5 None 2017 4.9 None 2015 Tanzania 6.6 None 2007 6.6 None 2007 Thailand 23.5 21.9 2.0 TIMSS 2011 2009 23.5 21.9 2.0 TIMSS 2011 2009 Timor-Leste 19.2 None 2017 15.0 None 2015 Togo 82.2 80.6 8.4 PASEC 2019 2017 85.6 84.2 8.5 PASEC 2014 2014 Tonga 0.1 None 2015 0.1 None 2015 33 Trinidad and 20.7 19.7 1.3 PIRLS 2016 2010 20.7 19.7 1.3 PIRLS 2016 2010 Tobago Tunisia 65.3 65.1 0.4 TIMSS 2011 2011 65.3 65.1 0.4 TIMSS 2011 2011 Turkey 15.0 10.0 5.6 TIMSS 2019 2016 21.7 17.6 5.0 TIMSS 2015 2015 Turkmenista 11.6 None 2014 11.6 None 2014 n Turks and Caicos None None Islands Tuvalu 2.5 None 2016 0.3 None 2014 Uganda 82.8 81.1 9.0 NLA 2014 2013 82.8 81.1 9.0 NLA 2014 2013 Ukraine 27.9 18.3 11.8 TIMSS 2007 2007 27.9 18.3 11.8 TIMSS 2007 2007 United Arab 34.3 32.4 2.8 PIRLS 2016 2016 34.3 32.4 2.8 PIRLS 2016 2016 Emirates United 3.4 3.2 0.2 PIRLS 2016 2016 3.4 3.2 0.2 PIRLS 2016 2016 Kingdom United 7.9 3.9 4.1 PIRLS 2016 2016 7.9 3.9 4.1 PIRLS 2016 2016 States Uruguay 41.7 41.4 0.5 LLECE 2013 2013 41.7 41.4 0.5 LLECE 2013 2013 Uzbekistan 1.5 None 2017 2.9 None 2015 Vanuatu 13.3 None 2015 13.3 None 2015 Venezuela, 13.8 None 2017 8.0 None 2015 RB SEA- Vietnam 19.6 18.0 1.9 2019 2013 1.7 1.1 0.6 NLA 2011 2011 PLM Virgin 0.0 None 1993 0.0 None 1993 Islands (U.S.) West Bank 6.1 None 2017 7.7 None 2015 and Gaza Yemen, Rep. 94.7 93.5 18.9 TIMSS 2011 2010 94.7 93.5 18.9 TIMSS 2011 2010 Zambia 14.0 None 2017 12.1 None 2013 Zimbabwe 14.8 None 2013 14.8 None 2013 Note: See accompanying data base for detailed data on learning poverty, learning deprivation, and schooling deprivation, including gender disaggregated estimates where available. References Boston College, TIMSS & PIRLS International Study Center. 2016. “TIMSS 2019 User Guide for the International Database.� https://timss2019.org/international-database/ Boston College, TIMSS & PIRLS International Study Center. 2021. “TIMSS 2019 User Guide for the International Database.� http://timssandpirls.bc.edu/timss2015/international-results/Confemen. 2019. “PASEC 2019: Qualité des Systèmes Éducatifs en Afrique Subsaharienne Francophone. Performances et Environnement de L’enseignement-Apprentissage au Primaire.� https://www.confemen.org/wp- content/uploads/2020/12/RapportPasec2019_Web.pdf Confemen. 2014. “PASEC2014: Education System Performance in Francophone Sub-Saharan Africa. Conférence des ministres de l'Éducation des États et gouvernements de la Francophonie Competencies and Learning Factors in Primary Education.� https://www.pasec.confemen.org/wp- content/uploads/2015/12/RE_Pasec2014_GB_web21.pdf 34 Directorate of Primary Education, Monitoring and Evaluation Division. 2018. “The National Student Assessment 2017 Grades 3 and 5.� http://dpe.portal.gov.bd/sites/default/files/files/dpe.portal.gov.bd/publications/24ec04b1_3064_41d3_ 9ac7_9b778a02e17b/The%20National%20Student%20Assessment%20%202017%20Book%20(1).pdf Francesco Avvisati and François Keslair. 2014. "REPEST: Stata module to run estimations with weighted replicate samples and plausible values," Statistical Software Components S457918, Boston College Department of Economics, revised 06 Jan 2020. IEA. 2019. “Trends in International Mathematics and Science Study 2019.� https://www.iea.nl/studies/iea/timss/2019 Kevin Macdonald, 2008. "PV: Stata module to perform estimation with plausible values," Statistical Software Components S456951, Boston College Department of Economics, revised 03 Feb 2019. PASEC. 2019. “PASEC2019 international assessment.� http://www.pasec.confemen.org/evaluation/evaluation-internationale-pasec2019/ SEA-PLM. 2020. “Process Timeline.� https://www.seaplm.org/index.php?option=com_content&view=article&id=40&Itemid=275 UNESCO. 2021. “Technical Cooperation Group on the Indicators for SDG 4.� http://tcg.uis.unesco.org/policy-linking/ UNICEF & SEAMEO. 2020. “SEA-PLM 2019 Main Regional Report: Children’s learning in 6 Southeast Asian countries.� Bangkok, Thailand: United Nations Children’s Fund (UNICEF) & Southeast Asian Ministers of Education Organization (SEAMEO) – SEA-PLM Secretariat. https://www.unicef.org/eap/media/7626/file/SEA-PLM%202019%20Main%20Regional%20Report.pdf World Bank. 2021. “Classification of Countries.� https://elibrary.worldbank.org/doi/full/10.1596/978-1- 4648-0484-7_classification_of_countries World Bank. 2019. “Ending Learning Poverty : What Will It Take?.� World Bank, Washington, DC. © World Bank. https://openknowledge.worldbank.org/handle/10986/32553 License: CC BY 3.0 IGO. 35 Learning Poverty Monitoring Series Technical Note 1 Learning Poverty Updates and Revisions What’s New? worldbank.org