Effects of a Single Cash Transfer on School Re-enrollment during COVID-19 among Vulnerable Adolescent Girls in Kenya: Randomized Controlled Trial

"The results demonstrate the potential for such an emergency transfer program and at the same time underscore how lack of resources likely remains a barrier to universal secondary school completion in Kenya."

COVID-19-related school closures in Kenya were among the longest in Africa, putting older adolescent girls nearing the end of secondary school at risk of permanent dropout. This randomised controlled trial (RCT) evaluated a logistically simple cash transfer intervention made to households in densely populated informal urban settlements in Nairobi, Kenya, when schools reopened after prolonged closures.

Calculated to cover average school fees for one term of secondary school, the transfer (16,000 Kenyan Shillings, or ∼US$150) was unconditional, but households were informed that it was intended to help cover the costs of re-enrollment for a specific adolescent girl nearing completion of secondary school. The timing of the SMS (text) message that alerted beneficiaries the transfer had been made reminded them of its intended purpose just as schools were fully reopening. Therefore, Population Council considers the SMS reminder as a nudge that is part of the intervention's messaging package.

For the baseline survey, researchers attempted to interview all 1,912 eligible girls and completed interviews with 1,620 (84.7%), all of whom were included in the study. On December 1 2020, stratified randomisation assigned 813 girls to the COVID-19 cash transfer treatment and 807 to control. Between January 6-8 2021, the transfers were made to bank accounts of the parents of girls assigned to treatment. For 812 of the assigned 813 treatment households, electronic delivery to the designated bank account was verified. In late February 2021, the researchers attempted to re-interview all those interviewed at baseline, completing the follow-up survey for 1,517 (93.6%). Re-interview rates were higher in treatment (96.8%) compared to control (90.5%).

Despite enrollment in the control group being relatively high (88%) after reopening, even higher than before the school closures, beneficiaries were 7.7 percentage points more likely to be enrolled. The main reasons for not enrolling in 2021 were marriage or pregnancy (42%) and being unable to cover school fees (32%). Alongside the effects on enrollment, there was a 9.8 percentage point (13%) increase in expectation of finishing secondary school, relative to 75.5% in the control. The increase in expectations points to the possibility that the effects, measured 6 weeks into the term after full reopening, would persist.

Effects on enrollment are twice as large for older compared to younger girls and 5 times larger for those not enrolled earlier in the year (comprising ∼10% of the sample) compared to those enrolled. These patterns suggest the treatment is more effective for those at higher risk of not completing secondary school, since temporary dropout before completion is likely harder to reverse at higher ages. There are no significant differences in effects on expectations across the subgroups.

"Nevertheless, there are almost certainly other important barriers to increasing female secondary school completion in Kenya - including some that may have been exacerbated by the pandemic - that cash transfers alone cannot resolve. These barriers likely include, for example, low school quality or teen pregnancy. In the longer term, therefore, more comprehensive programming is probably necessary to achieve universal secondary schooling."

Furthermore, "[i]n contrast to the estimated positive effect on enrollment for the targeted adolescent girl, examination of other children in the family (limiting the analysis to the selected sample of those with siblings) raises the possibility that the treatment may reduce average enrollment for her ineligible school-aged siblings. Enrollment for siblings 6-20 years old is 88.6% in control households but 3.4 percentage points lower in treatment....though not robust..., those results point to the importance of considering spillover effects of such interventions..."

In conclusion: "The program effectiveness demonstrates feasibility of the approach and underscores the potential importance of additional resources for schooling during the pandemic, when a large majority of households had suffered income losses."