Magnet Schools and Student Achievement: Evidence from a Randomized Natural Experiment in China

Abstract: This paper examines the impact of attending a magnet school on student achievement using school admissions lotteries in China. Although lottery winners were more likely to attend magnet schools that appear better in many dimensions, including peer achievement, we find little evidence that winning a lottery improved students' performance on the High School Entrance Exam or their enrollment status at elite high schools three years later. Magnet school popularity, measured by either the competitiveness of the admission lottery or the take-up rate of lottery winners, is highly positively correlated with the average student achievement, but largely unrelated to the treatment effect on test scores that we estimate for each school. This evidence suggests that parents value peer quality beyond its effect on achievement gains, or confuse average student achievement with value added. The finding that magnet schools are sought mainly for their observed superiority in average student achievement rather than for their academic value added casts doubt on the potential of school choice to improve student achievement, at least in this context.
Curriculum Vitae
Research
Teaching

Peer Effects on Student Achievement: An Instrumental Variable Approach Using School Transition Data

Abstract: In the empirical analysis of peer effects on student achievement, most econometric attention has been directed at issues of the endogenous peer group formation and the simultaneous peer interactions. I argue in this paper that the impact of measurement error is also important. Ammermueller and Pischke (2009) show that measurement error in contextual peer characteristics leads to inconsistent estimates of peer effects. I extend the analysis of measurement error to models that control for both lagged individual and peer group outcomes. I show that measurement error between lagged individual and peer group outcomes is likely to be positively correlated because of the continuing presence of a student's former peers in her current peer group, which leads to a negative bias in the estimates of peer effects. I propose an empirical strategy to correct this bias by using the lagged outcomes of new peers to instrument for the lagged outcomes of all peers. To illustrate these points, I conduct a Monte Carlo study and also present an analysis of peer effects in middle school in China using a panel data set of school transition. The results from both exercises show that the standard estimates of peer effects have a substantial negative bias, and that the proposed IV estimates can correct the bias.