Mother-tongue instruction and later labour market outcomes: evidence from a natural experiment in Ethiopia

This paper offers empirical evidence on the effect of mother-tongue instruction in primary school on students’ later labour market outcomes. Since Ethiopia has adopted mother-tongue instruction after the 1994 Ethiopian education reform, many students who have attended primary school after 1994 are exposed to mother tongue instruction, resulting in a variation in exposure to mother-tongue instruction by birth cohort. In Amhara state, Amharic is adopted as medium of instruction both before and after the education reform whereas other states in Ethiopia have changed the medium of instruction after the 1994 reform. The duration of students’ exposure to mother-tongue instruction, however, varies depending on the state in which they have attended primary school since states in Ethiopia mandate a switch from mother-tongue to English instruction either in grade 5, 7, or 9. Exploiting these two plausibly exogenous sources of variations (across states and birth cohorts) and using data from the 2013 Ethiopian Labour Force Survey, we estimate difference-in-differences model. Estimates from our preferred specifications suggest that mother-tongue instruction in primary school improves later labour market outcomes, but the size of its effect decreases with the number of years an individual was exposed to mother-tongue instruction in primary school.