Association for Public Policy Analysis & Management

Since the 1980s, college entrance examination (CEE) has been the only way for students to get into colleges in China, and the one-time CEE score has been the golden standard to determine whether a student can be admitted by a college or not. A certain amount of students admitted by colleges through CEE, however, have been questioned to be good at taking exams only.

In 2003 the Independent Admission Policy（IAP）was implemented in some top universities in China, aiming at selecting students with comprehensive abilities and academic potentials. These top universities are entitled to organize additional written tests and interviews besides CEE. Students who perform really well in these tests are able to obtain bonus points, which increase the probability of being admitted by these universities even if their CEE scores are not high enough to meet the admission cutoff score. Up to now, IAP has become an important admission method and almost 100 universities have been authorized by the Ministry of Education to organize IAP in China.

Empirical researches on evaluating the effectiveness of IAP have been gradually emerging since 2003, however, whether IAP is effective in selecting outstanding students is still an open question because the endogeneity issues have not been addressed by rigorous causal inference methods. That is to say, it is not convincing to draw the conclusion that IAP is effective even if students admitted through IAP perform better in colleges than students admitted through CEE because the former students could have been admitted through CEE if there weren’t IAP.

This paper focuses on evaluating the effectiveness of IAP by using a combined dataset of representative students’ survey data and administrative data in a top university in China. First, the paper examines whether there is difference in students’ performance in colleges between CEE students and IAP students by using OLS. PSM is also used to further compare students with comparable observable characteristics. Next, the paper selects IAP students with CEE scores below the admission cutoff scores and compares their academic performance with CEE students and IA students with CEE score above admission score. Last, OLS with different bandwidths is applied to simulate regression discontinuity methods to control the endogenous issue.

The main findings are as follows: overall IAP students have better academic performance in colleges while their CEE scores are significantly lower compared with students admitted through CEE. Such difference in academic performance still exists between CEE students and IAP students with comparable CEE scores. Furthermore, there is no significant difference in academic performance in colleges between IAP students with CEE scores above and below the admission cutoff scores. Specifically, even IAP students with lower scores have better academic performance than CEE students. These results indicate the effectiveness of IAP is effective in students’ selection to some extent. Some more policy implication is discussed at the end including the relationship of effectiveness, equity and the cost-effectiveness of IA policy.