The Negative Peer Effect: Why do peers' success have a negative effect? (Japanese)

Recent research in the field of economics has placed more emphasis on the importance of the "peer effect," which is as defined as the influence from peers within the same class/cohort, as a determinant of student achievements. However, in previous literature, there is no consistent result about the peer effect, especially in Japan, in part due to data limitation. This study thus takes advantage of using a micro dataset of a representative survey that tracks students in primary and secondary schools in Saitama prefecture from 2015-2016, exploits year-on-year changes in the cohort composition within the same school, and identifies a causal peer effect in the framework of value-added educational production function. We find that the peer effects are negative and statistically significant, regardless of grade, gender, and subjects examined. Conditional on individual, class, and school characteristics, if the mean score of the class (measured by the item response theory (IRT)) increased by 1, the student's score decreased by a range of 0.09 to 0.23 in Japanese, and 0.16 to 0.27 in math. It is possible to point out that having a high-achieving classmate in the same class provides a self-awareness to the student suggesting that his/her rank within the class is low and could lose motivation for further learning. It is important to encourage them to have high level of expected return to education.