Predicting School Dropout with Administrative Data

Abstract

School dropout is a growing concern across Latin America because of its negative social and economic consequences. Identifying who is likely to drop out, and therefore could be targeted for interventions, is a well-studied prediction problem in countries with strong administrative data. In this paper, we use new data in Guatemala and Honduras to estimate some of the first dropout prediction models for lower-middle income countries. These models correctly identify 80% of sixth grade students who will drop out within the next year, performing better than other commonly used targeting approaches and as well as models used in the United States.