Private tutoring and its impact on students' academic achievement, formal schooling, and educational inequality in Korea

Over the last two decades, private tutoring has emerged as an important issue in education as its demand has been growing around the world. However, the evidence of the effectiveness of private tutoring is still mixed. Using the Korean Educational Longitudinal Survey, this dissertation explores the causal impact of private tutoring in Korea on three outcomes: students' academic achievement, the quality of the learning environment in formal schooling, and educational inequality. The first set of empirical analyses explores how private tutoring in secondary schools affects students' academic achievement in both short-term and long-term aspects using Ordinary Least Squares, Instrumental Variable, and Propensity Score Matching methods. The results suggest that private tutoring in middle school, on average, has positive short-term effects on students' academic achievement in middle school, but minimal long-term effects on the university entrance examination scores. By subject area, English and math tutoring are effective in improving academic achievement in middle school, whereas verbal tutoring is not. Moreover, private tutoring in grade 7 is most beneficial for students in middle school. In terms of private tutoring in high school, only math tutoring is beneficial for high school students in improving scores on the university entrance examination. The second set of the analyses employs Ordered Logit, Propensity Score Matching, and Difference-in-Differences methods to estimate the impact of private tutoring on the quality of formal schooling. The quality of the learning environment in formal schooling is measured by students' attention to lessons in class. On average, private tutoring shows a positive influence on students' attention to lessons in grade 8, but the magnitude of its influence is not substantial. However, when differentiating the results by ability group, positive effects are detected mostly in the low-ability group, which means that low achievers pay more attention to lessons in schools if they participate in private tutoring. These results imply that private tutoring improves the overall learning environment in formal schooling, which in turn increases the overall quality of schooling. The third set of the analyses uses Quantile Regression, Two-Stage Least Absolute Deviation estimator, and Propensity Score Subclassification to estimate the heterogeneous effects of private tutoring between ability groups, which provides implications on educational inequality based on academic achievement. The overall results suggest that private tutoring in middle school exacerbates educational inequality between high and low achievers, which implies a widening of the achievement gap. In addition, enrolling in tutoring at an earlier grade level results in greater heterogeneity between high and low achievers in academic performance than enrolling in tutoring during later grade levels. However, private tutoring in high school contributes to reducing the achievement gap; low achievers benefit more from private tutoring in high school compared to high achievers. Moreover, three years of cumulative math tutoring and receiving a single year of math tutoring in grade 12 contribute to narrowing the achievement gap between low and high achievers in the university entrance examination scores.