The learning-poverty measure combines schooling and learning data to describe a system-level condition, not the performance of an individual education provider.
What the learning-poverty indicator measures
Learning poverty measures the share of children unable to read and understand a simple text by the end of primary-school age, combining children out of school with those below minimum proficiency in school. The update estimates a severe global increase following pandemic disruption. The indicator is designed for system action rather than individual diagnosis.[S1]
Learning poverty combines two components: children who are out of school and children in school who have not reached minimum reading proficiency by late primary age. This construction makes exclusion and weak learning visible in one system-level indicator. It does not classify an individual child, determine the quality of one school or identify the cause of a country result.[S1]
Definition, estimates and the RAPID framework
The update reports a pre-pandemic baseline of 57 percent in low- and middle-income countries and a simulated post-disruption estimate of 70 percent. It presents the RAPID framework for reaching children, assessing learning, prioritizing fundamentals, increasing instructional efficiency and developing psychosocial health and wellbeing.[S1]
The 2022 update explains the methodology, presents global and regional estimates and links recovery to the RAPID framework: Reach every child; Assess learning levels; Prioritize the fundamentals; Increase instruction efficiency; and Develop psychosocial health and wellbeing. The indicator establishes scale, while the framework organizes policy response. Evidence of implementation still needs to show which learners were reached and whether learning changed.[S1]
A recovery programme can preserve the connection from diagnostic evidence to instructional response. The record identifies the assessment used, groups not represented, priority learning content, instructional time and follow-up result. Where the assessment changes, a linking method or explicit break is needed. Otherwise, an apparent gain may arise from a different instrument rather than improved reading proficiency.[S1]
Using a global signal for local diagnosis
The estimate concerns children in low- and middle-income countries and focuses on reading at a defined stage. It does not cover the full curriculum, older learners or every dimension of foundational learning. Provider use needs direct, age-appropriate assessment and enrolment information.[S1]
The measure concerns minimum reading proficiency at the end of primary-school age and combines data from schooling and learning sources. It is not a general index of every subject, age or education level. Institutional use requires a locally valid baseline, a defined learner cohort and an assessment aligned with the learning question. Out-of-school learners cannot be inferred from a school's enrolled population.[S1]
Uncertainty, coverage and causal limits
The post-disruption figure is modelled rather than a complete contemporaneous census. Cross-country assessment coverage and definitions vary. An aggregate rate cannot identify which learner needs which support, and repeated testing alone is not a recovery intervention.[S1]
Country coverage, assessment comparability and modelled estimates affect precision. Pandemic disruption also changed who was assessed and under what conditions. A movement in the indicator can reflect access, learning or data changes. Local action therefore needs disaggregated information and a record of assessment conditions. A global percentage cannot establish that a particular programme caused improvement.[S1]
Learning-poverty evidence in ICEQC analysis
ICEQC records learning poverty as external system evidence. It is not an ICEQC threshold and cannot determine provider conformity without the published requirements and direct evidence. No later ICEQC clause is attached to the earlier article.
ICEQC may use learning-poverty evidence to identify a material education-quality question, while keeping the external indicator separate from certification criteria. The World Bank and partner organizations are not represented as approving ICEQC. No subsequent ICEQC clause is applied to the earlier record.