Abstract

The aim of this article is to assess the impact of armed conflict on the number of years of education of primary school pupils in Mali. To do this, we formed two cohorts from the Enquete Harmonisée sur les Conditions de Vie des Ménages (EHCVM), adopting an identification strategy similar to that of Dabalen and Paul (2012). A first cohort (12-22years) or young cohort is composed of all children likely to be directly affected by the conflict (being in primary school over the period 2012-2019) and a second cohort (23-32years) composed of those no longer in primary school at the time of the conflict. We then merged this database with the conflict database (ACLED) to identify affected individuals on the basis of the occurrence of a conflict within the department in which they reside. By adopting various identification strategies (binary or continuous treatment) and evaluation modes (Difference in Difference, Propensity Score Matching) and taking individual heterogeneities into account, it emerges that the affected cohort has fewer years of education than the unaffected cohort. Affected individuals have fewer years of education than their unaffected peers, and this conflict has more pronounced impacts in rural areas (value) and among girls (value). The main mechanisms highlighted in this study are the deterioration of the learning environment (insecurity and absenteeism) and the erosion of household purchasing power. The main findings tend to sound the alarm about the need for policies to (i) reduce the intensity of conflict, (ii) set up training programs to enable children to continue schooling even in conflict zones, (ii) set up intensive schooling programs in the worst-affected areas to reduce disparities within departments, place of residence (rural and urban) and by gender (girls and boys).