This page in:

School girls and school boys smiling at camera in the class room.
School girls and school boys smiling at camera in the class room. Avatar_023 /

In October 2019, the World Bank and the UNESCO Institute for Statistics (UIS) launched a new metric, Learning Poverty, which highlighted that 53% of all children in low- and middle-income countries were not able to read an age-appropriate text with comprehension by age 10 (Figure 1). This figure included those who were tested in schools and found unable to read with comprehension an age-appropriate text, along with out-of-school children. Because schools have numerous other functions that contribute to children’s health and well-being —such as promoting safety, nutrition, and socialization, and facilitating parents' labor market participation—and at the macro level schooling can help build social cohesion, democracy, and peace. All those complementary functions mean that schooling has value over and above the measured cognitive learning that it leads to, and they justify including schooling deprivation in the concept of learning poverty. The new metric is already focusing action to turn the tide on low levels of reading comprehension. It’s being used to rally the international community to prevent a lost generation of children (in this call by dozens of world leaders, through the Save Our Future initiative), and to motivate the tracking of COVID’s impact on education (e.g., in the Joint Survey carried out with UNESCO, UNICEF, and OECD and a new collaboration with UNICEF and Johns Hopkins). It is also helping to explain the need to focus on foundational learning (e.g., by these global and national foundations).  In a new paper, Will Every Child Be Able to Read by 2030?, we provide the full technical background and main results for the learning poverty metric, together with robustness tests, heterogeneity analysis, tests of external validity, and extensions.

Figure 1 Learning Poverty around the World (hover to see country numbers)

<script type='text/javascript' src=''></script><div class='tableauPlaceholder' style='width: 100%; height: 746px;'><object class='tableauViz' width='100%' height='746' style='display:none;'><param name='host_url' value='' /> <param name='embed_code_version' value='3' /> <param name='site_root' value='&#47;t&#47;EdStats' /><param name='name' value='lpv-covid-simulation&#47;dsh_lpv_map' /><param name='tabs' value='no' /><param name='toolbar' value='yes' /><param name='showAppBanner' value='false' /></object></div>

Why did the Bank and the UIS create the learning poverty measure, and what have we learned so far about the progress the world is making toward ending learning poverty?  First, the rationale. The Sustainable Development Goals (SDGs), which every UN member signed onto in 2015, embody high global aspirations for education.  SDG 4 makes this commitment: by 2030, the signatories will “ensure inclusive and equitable quality education and promote lifelong learning opportunities for all.” The various targets under this goal cover the educational landscape, starting with universal access to quality ECD and preschool and extending to equal access to affordable university education. But the very first of these commitments is Target 4.1, which is to “ensure that all girls and boys complete free, equitable, and quality primary and secondary education leading to relevant and effective learning outcomes.” In other words, the world has committed to achieve universal completion of both primary and secondary school for all youth — and with meaningful learning — by 2030. However, given the depth of the learning crisis in many low- and middle-income countries and the observed rate of progress in the recent past, we show this target is not feasible. A target has to be ambitious, but it must also be attainable. Otherwise, it is useless as a motivator to propel the required actions. This was true even before the COVID-19 pandemic hit, and it is even more so now, as the school closures and global recession triggered by the pandemic interrupt student learning and reduce attachment to schooling.

The learning poverty metric is designed to spotlight two fundamental challenges at the core of the SDG 4 aspirations:  ensuring that all children can read by age 10 with at least a minimum level of comprehension (SDG 4.1.1b) and that all children are enrolled in school (SDG 4.1.4). This simple learning poverty indicator can resonate with any stakeholder—parent, educator, community leader, employer, politician—and it is also technically sound. Ensuring that all students are in school and reading with comprehension is essential to achieving the ambitious SDG targets and to building human capital. Children need to learn to read so that they can read to learn. Those who do not become proficient in reading by the end of primary school often cannot catch up later, because the curriculum of every school system assumes that secondary-school students can learn through reading (Figure 2). Reading is a gateway to all types of academic learning.  A target of “every child reading by age 10” is a fair development objective. All children reading is a human right. The rate of learning poverty should be zero—just as the rate of extreme monetary poverty or hunger should. In high-income countries, the learning poverty rate is 10%, with only a small share not learning to learn to read with comprehension before the end of primary school; and for the highest-performing countries, the figure is 3% or less. While it may take decades to build up an entire high-quality education system, ensuring that all primary school-age children are attending schools which can teach children to reach a minimum proficiency in reading should require much less time. Reading proficiency can serve as a good proxy for (contemporaneous) foundational learning in other subjects, particularly at the level of the educational system. This indicator help focuses attention on equity. To eliminate learning poverty, countries have to give all children good foundational skills; they can’t just try to improve the mean by focusing on the elite students.  This broad foundation of skills sets up the entire cohort for success in secondary school and beyond. Finally, the learning poverty measure is more responsive to targeted policy interventions than other measures such as the Learning Adjusted Years of schooling (LAYS). It is easier to make tangible short-term progress on Learning Poverty, because it requires improving only the learning and schooling of the youngest cohort, and we have proven techniques for doing that.  This is a key step toward the crucial longer-run goal of increasing LAYS.   

Figure 2. Learning Poverty vs. Learning Deprivation (share of students below min. proficiency) in secondary school, for 25 LICs and MICs

Learning Poverty vs. Learning Deprivation (share of students below min. proficiency) in secondary school, for 25 LICs and MICs

To develop the learning poverty estimates, we combine data from 100 countries that account for 81% of children worldwide and 80% of children in low- and middle-income countries, using internationally comparable learning thresholds produced by the Global Alliance to Monitor Learning (GAML), which is led by the UIS. The results of that effort, documented in detail in our working paper and summarized briefly in World Bank (2019), are sobering:

  • More than half of children in low- and middle-income countries have not achieved minimum levels of proficiency in reading by age 10, or in most cases by the end of primary.  An estimated 53% of children at or near the end of primary school age are not yet able to read a short, age-appropriate story with comprehension. 
  • At the rates of progress seen so far this century, the goal of ensuring that all children can read by 2030—in other words, reducing the rate of learning poverty to zero—is far out of reach.  While the share of children who are “learning-poor” has been declining, the pace of progress is far too slow to ensure that all children will be able to read by 2030.  We estimate that under a “business as usual” scenario —that is, with progress at the rate we saw during 2000-17—44% of children in 2030 will still be unable to read at age 10.  This indicator is an early warning that all the education-related SDGs are in jeopardy, and that grounding aspirations requires a more plausible medium-term target. 
  • Even if countries reduce their learning poverty at the fastest rates we have seen in recent decades, the world will not come close to attaining the “every child reading” goal by 2030.  If every country were to reduce learning poverty like the top performers over the 2000-15 period—meaning that they matched the rates achieved by countries at the 80th percentile of the regional distribution of gains—the global learning poverty rate can be reduced from 53% in 2015 only to 27% in 2030.  Stated differently, if every low- and middle-income country ramped up its efforts to address learning poverty and doubled or tripled its historical rate of progress, it would be possible to cut the global learning poverty rate in low- and middle- income countries by nearly half.  In October 2019, the World Bank announced a corporate commitment to support countries to “by 2030, reduce by at least half the share of children in low- and middle- income countries who cannot read by age 10.” The goal of this target was to promote tangible progress toward the SDGs and improved Human Capital, by focusing on medium-term learning goals and immediate action to improve foundational skills. Our analysis shows that this intermediate goal was already highly ambitious when announced, requiring a global rate of progress to increase nearly triple its 2000-2017 rate.

All of these findings are based on data from before the COVID-19 pandemic—meaning that the situation is now even worse than indicated by these estimates. While the data sources used in this analysis are not yet available for 2020, nor will they be available for at least another year or two, there is no question that the levels of learning poverty are now higher, and the recent trends worse than reported here. Initial simulations suggest that in a pessimistic scenario, learning poverty in low- and middle-income countries could increase from 53% to 63% due to COVID-19. The effects of COVID will make the learning target even harder to achieve, but they also underline how important it is to have a summary indicator like learning poverty.  By keeping our eyes on that indicator, together we can build more effective and equitable basic education systems in the wake of the crisis.


Silvia Montoya

Silvia Montoya Director, UNESCO Institute for Statistics UIS

Reema Nayar

Practice Manager, Latin America and Caribbean, Education Global Practice

Halsey Rogers

Lead Economist, Education Global Practice

Brian Stacy

Data Scientist, Development Data Group, World Bank