TY - RPRT AU - Orlovaite, Ernesta AU - Chevalier, Arnaud AU - Rastle, Kathy TI - Language Match Is Fixed, Language Understanding Can Be Built: Rethinking Home Language Instruction in Sub-Saharan Africa PY - 2026/Nov/ PB - Institute of Labor Economics (IZA) CY - Bonn T2 - IZA Discussion Paper IS - 19079 UR - https://www.iza.org/publications/dp19079 AB - Nine in ten children in Sub-Saharan Africa cannot read a simple text by the age of ten, and most are taught in a language they do not speak at home. The dominant policy response is home language instruction. Using a unique dataset of 1,718 formerly out-of-school children in an accelerated learning programme in Ghana and Liberia, we compared matched pupils, who spoke the language of instruction at home, with mismatched pupils, who did not. First, understanding of the language of instruction is a better predictor of reading progress across countries than language match. Second, reading is shaped by classroom characteristics: the more classmates shared a mismatched pupil’s home language, the worse that pupil read. Third, language understanding can be built: most mismatched pupils developed oral proficiency in the language of instruction within a single programme cycle, the classroom’s language mix predicted whether they did, and those who did read better. No language policy can match the home language of every child, but every child should be taught to understand the language of the classroom. KW - reading KW - mother tongue KW - language of instruction KW - oral language KW - peer effects KW - Sub-Saharan Africa ER -