This paper examines the effect of maternal education on child mortality in Bangladesh by exploiting quasi-experimental variations in the duration of exposure to a school stipend project for identification. Results from the instrumental variable estimation indicate that an additional year of maternal schooling reduces both under-five and infant mortality by about 20 percent. The findings are statistically significant and robust to a number of model specifications, including survival models controlling for right censoring of child mortality. Analysis of potential mechanisms suggests that maternal education reduces child mortality through greater wealth and literacy, positive assortative mating, lower fertility, delayed marriage and childbearing, greater health-related knowledge, better health-seeking behaviors, and female empowerment, but not through female employment.

中文翻译：

---

孟加拉国孕产妇教育对儿童死亡率的影响

本文通过利用学校助学金项目持续时间的准实验变化来研究孕产妇教育对孟加拉国儿童死亡率的影响，以进行识别。工具变量估计的结果表明，孕产妇教育多一年可将五岁以下儿童和婴儿死亡率降低约 20%。该研究结果具有统计学意义，并且对许多模型规范具有稳健性，包括控制对儿童死亡率进行正确审查的生存模型。对潜在机制的分析表明，孕产妇教育通过提高财富和识字率、积极的择偶交配、降低生育率、推迟结婚和生育、增加与健康相关的知识、改善寻求健康的行为和赋予女性权力来降低儿童死亡率，