Longitudinal data modeling of the effect of cash transfer on school dropout rates in the Research Initiative to Support the Empowerment of girls (RISE) trial 2016 to 2017

Abstract

School dropout rates, as well as early marriages and pregnancies are high among adolescent girls in rural Zambia. In the quest to fight this, the Research Initiative to Support the Empowerment of girls (RISE) trial has been providing cash transfers and community dialogues to adolescent girls in rural Zambia. The overall goal of the study was to establish the effects of cash transfers on adolescent girls’ school dropout rates in selected provinces of Zambia. The study was nested in the RISE trial which is an ongoing cluster randomized trial being conducted in Central and Southern provinces of Zambia. A total of 3,500 adolescent girls were included in the study. Random intercepts model for binary data was used to model the individual effects estimates taking account of the dependency that was likely to occur due to the repeated measurements and clustering in the study. Girls who were married or cohabiting and girls who had given birth, were significantly less likely to be in school (OR=0.004, 95% CI {0.001-0.02}, p-value<0.0001) and (OR=0.003, 95% CI {0.02-0.04}, p-value<0.0001) respectively. Consistently receiving cash transfers increased the chance of a girl being in school (OR=8.51, 95% CI {4.50-16.08}, pvalue< 0.0001). There was an indication that the combined intervention arm had an increased chance of girls being in school, however, we could not rule out chance finding (OR=0.89, 95% CI {0.59-1.36}, p=0.606). The findings suggest that there is no difference between the two arms (the cash arm and the combined intervention arm) on school dropouts. However, consistently receiving cash transfers was shown to be a protective factor of school dropout rates. The findings also indicate that early marriages and adolescent pregnancies are some of the factors negatively affecting schooling. Key terms: Cash transfer, Cluster Randomised Controlled Trial, Longitudinal