Economics of education in rural China: two experimental studies

This thesis uses experimental methods to study two topics on the effectiveness of school inputs on educational outcomes in rural China. For the first topic, I use unique administrative data from Wugang County Education Bureau (a rural county in Hunan), and a regression discontinuity methodology. I find that selective/elite schools, despite their resource advantages, only have limited effects on raising student educational outcomes. However, magnet classes are effective for the top student group. These findings imply that magnet classes provide benefits at the expense of other students, and in general that concentrating resources in a few elite schools is not an effective way to raise educational outcomes. For the second topic, this time using unique data from Shaoyang County Education Bureau (also a rural county in Hunan), I designed two RCTs to examine the effectiveness of a low cost communication intervention. The intervention used a 12-point assessment form measuring a pupil’s academic work and class behaviour. One RCT communicated these assessment results only to the students (Teacher-Student-Communication, TSC), and the other additionally to the pupil's parents (Teacher-Student-Parent-Communication, TSPC). Test score improvements before and after the 8-month intervention period (with 13 assessments) were the measure of educational effectiveness. I find that the TSPC intervention for maths for left-behind children is particularly important. In addition, TSC helped younger pupils (3d grade) more than old (5' grade), whether left behind or not, showing the importance of early intervention.