Cognitive spillover benefits of early childhood education: Quasi-experimental evidence based on random class assignment from China

This paper investigates the cognitive spillover effects of classmates’ early childhood education (ECE) experience on junior high school students, using data from the first two waves of the China Education Panel Survey (CEPS). To address potential endogeneity in classmates’ composition, we leverage exogenous variation in students’ ECE experience generated by random class assignment upon their entry into junior high school. Employing a value-added model, we find that classmates’ ECE experience significantly enhances students’ cognitive performance. Specifically, a 10-percentage-point (pp) increase in classmates’ ECE enrollment raises students’ cognitive scores by 0.08 standard deviations (SD), while an additional year of classmates’ ECE experience shows insignificant effect. As to underlying mechanisms, the spillover effects are driven by an improved class environment, increased parental homework support, stronger learning efforts and enhanced non-cognitive skills of students, together with the peer interactions within social networks. Among them, students’ behaviors exert the strongest explanatory power of 13 %. Furthermore, the benefits are more pronounced among urban students with ECE experience, those from better-educated families and with moderately below-average baseline cognitive skills.