Short-Term Impacts of a Cash Transfer Program for Girls’ Education on Academic Outcomes: Evidence from a Randomized Evaluation in Malawian Secondary Schools

Abstract

I.IntroductionEducation has been considered as one of the most important determinants of human capital development. Improved education measured by test scores is correlated with increased wages later in life at the micro-level (Blau, and Kahn 2005) and is argued to facilitate the economic growth of a country at the macro-level (Hanushek, and Woessmann 2012). Recent studies emphasize the importance of noncognitive abilities as the underlying mechanisms connecting educational interventions to test scores and later to adulthood outcomes (Glewwe et al. 2013; Wydick et al. 2013). This study contributes to literature by providing a unified evaluation of several academic outcomes and determining educational intervention's effect on a number of intermediate outcome variables (such as non-cognitive skills) that could be potential mechanisms explaining outcomes during adulthood. The ideal way to identify possible underlying mechanism of educational interventions for outcomes in adulthood is a randomized controlled trial (RCT) targeted towards experimentally changing each potential mechanism separately. While the present study was not designed to do this, it is nevertheless important to capture and describe the impact of the educational intervention on these potential mechanisms in the short run.The paper uses experimental data from the cash transfer program for girls' education in Malawi. In 2012, 124 classrooms of grades 9th, 10th, and 11th across 33 public secondary schools in Lilongwe rural areas participated in the cash transfer program and 62 classrooms are randomly selected as the treatment group while the remaining 62 classrooms served as the control. All female students in treatment classrooms received a year of tuition fee assistance directly deposited to the school account, as well as monthly stipends distributed to students. The experimental design evaluates the causal effects of the cash transfer program on schooling outcomes (dropout and absence), test scores, and non-cognitive abilities.In terms of schooling outcomes, dropout rates (both self- and schoolreported) declined. The impact of the program on dropout rates is significant and stronger among 9th grade (starting grade of a secondary school) students, whereas they are insignificant in 10th and 11th grade students. The probability of remaining at the original school in the treatment group (considering both dropouts and transfer students) increased by 6.7 percentage points (63.8%). School absence also diminished among treated students on average by five days per year. For the Junior Certificate Exam (JCE), a national exam at the end of 10* grade in secondary school, the cash transfer program increased the probability of baseline ninth grade students in the treatment group to take the JCE by 15.5 percentage points (24.8%). The probability to pass JCE also rose by 18.7 percentage points (36.7%) and overall exam scores improved by 0.241 standard deviations. However, the effects of the program on the Malawi School Certificate Exam (MSCE), another national exam that secondary school students take at the end of 12* grade, are minimal and statistically insignificant. For non-cognitive abilities, treated students take education more seriously, with higher aspirations for educational achievement.The paper is organized as follows. Section II briefly reviews related literature on the education support program. Section III presents the experimental design and addresses potential biases to the validity of the experiment. In Section IV, the empirical strategy based on simple OLS models is presented, followed by the results and analysis on schooling outcomes, test scores, and non-cognitive abilities in Section V. Finally, Sections VI concludes the paper.II.Brief Literature Review on Education Support ProgramsThis section provides a brief overview of the related literature, limiting the scope to only financial incentives for students from field experiments. …