uาFTong WangiThe University of Chicago Harris School of Public Polic
y, Ph.D. candidatej
่FA Welfare Comparison of School Choice Mechanisms
In China, Chinese parallel mechanism is widely used for school admissions. Theoretically, Chinese parallel mechanism is less manipulable than Boston mechanism, yet it is argued that Boston mechanism is more efficient under certain conditions. Using a survey of high school choice participants, we recover students' preferences over schools directly. Merging the survey data with the administrative data allows us to directly measure the students' welfare and also measure the "magnitude" of students' manipulation of their preferences. Since we have the administrative data on students' choices over the years which covered the period when Boston mechanism was used and also when Chinese parallel mechanism was used, by assuming students' relative preference over high schools is stable, we directly compare students' welfare under the two mechanisms, both of which are manipulable. The result shows that Boston mechanism leads to lower average welfare than Chinese parallel mechanism and Boston mechanism hurts students with lower social economic status. The mechanism change leads to little change in equity, from Gini coefficient of 0.15 under BM to 0.148 under Chinese parallel mechanism. Students rank the last choice in the first block more aggressively under Chinese parallel mechanism than Boston Mechanism. Evidence shows that Chinese parallel mechanism gives higher test score students less chance and lower score students more chance to elaborate their ability to game the mechanism, which can partially explain the welfare performance of the two mechanisms.