This paper provides new evidence on the impact on educational attainment of an international child support (ICS) program, implemented in ten primary schools located in the peri-urban districts of the city of Goma (Democratic Republic of Congo). Using original micro data from a sample of 309 children (treated, 121; control, 188), we explore, through a difference-in-differences approach, whether the ICS program, which reduces the structural uncertainty faced by the targeted pupils and their families, impacts on a broad set of alternative educational outcomes—namely, performance scores in four different subjects (civic education and religion, French languages, national languages and mathematics), total score, failure and school drop-out rates. The results show that sponsored children report lower drop-out and failure rates in comparison to their control peers and that they succeed in catching up in all subjects in two school years. Results are robust to the implementation of a coarsened exact matching procedure, exploiting the data structure to produce unbiased estimates along with bounded ex-post balancing.

中文翻译：

---

国际儿童赞助改善学校表现：来自戈马 (DRC) 的证据

本文提供了有关国际儿童支持 (ICS) 计划对教育程度影响的新证据，该计划在戈马市（刚果民主共和国）城郊地区的 10 所小学实施。我们使用来自 309 名儿童样本（治疗组，121 名；对照组，188 名）的原始微观数据，通过差异化方法探索 ICS 计划是否可以减少目标学生及其家庭面临的结构不确定性，影响一系列广泛的替代教育成果——即四个不同科目（公民教育和宗教、法语、国家语言和数学）的成绩、总分、失败和辍学率。结果表明，与对照同龄人相比，受资助儿童的辍学率和失败率较低，并且他们在两个学年内成功地赶上了所有科目。结果对于粗化精确匹配过程的实施是稳健的，利用数据结构产生无偏估计以及有界事后平衡。