Who leaves school prematurely in Uganda: Do predictors vary by place of residence?

This paper examined the predictors for school dropouts in Uganda following the introduction of universal primary education close to 30 years ago. We used 10 % of the 2014 Uganda Population and Housing Census dataset (albeit relatively old) to examine the predictors of school dropouts in; i) Rural Uganda ii) Kampala (the capital) and iii) other urban areas. We specifically applied the logistic regression model to predict the probability that a child aged 6–16 dropped out of school (versus not) in 2014. While the socioeconomic status of households, household size, composition and age of the child were the strongest predictors for dropouts, the latter were found to vary by place of residence, due to variations in; geography, population composition, nature of economic activity, supply of education and levels of commitment of local government authorities in school supervision. Action points to redress dropouts in Uganda cannot be implemented ubiquitously but ought to take into consideration peculiar circumstances pertaining in; i) Rural areas ii) Kampala and iii) other urban areas. Policy should especially engender fiscal reforms to increase local revenue and redress the various predictors for dropouts in the differing rural and urban spaces.