School drop out in Bangladesh: insights using panel data

Abstract

This paper examines the relative strength of different factors associated with school drop out using data collected between 2007 and 2009 in Bangladesh. A sample of 9046 children, aged 4-15, was selected across six districts for a household survey focusing on children's school access and experiences. Two groups of children were identified: those who were enrolled in school in both 2007 and 2009 and those who dropped out by 2009. Using a multivariate logit model, results show age and gender, together with financial constraints, such as lack of income and school expenditure, as the top predictors of school drop out. Two other important predictors are lack of parental support for children's school work and school absenteeism. This paper identifies some possible interventions that could reduce school drop out. These include campaigns to reduce late entry and overage enrolment, reductions in direct costs of schooling, and more support for low achieving children to compensate for lack of educational support at home.