|Author Name||ONOZUKA Yuki (Fellow (Policy Economist, RIETI))|
|Creation Date/NO.||October 2020 20-J-039|
|Download / Links|
This paper examines the relationship between college selectivity and the role of recommendation-based admission. First, I use college admission data and show the relationships between college characteristics, such as institution type, selectivity, and major fields of study offered, and the spread of recommendation-based admissions programs. Then, I use student-level data to analyze the differences between college students admitted through written exam-based admission and those through recommendation-based admission from three aspects: characteristics during high school, attitude and activities inside and outside college, and performance at college. Since college selectivity is not available in my dataset, I focus on heterogeneity across high school rank in this descriptive analysis. The results suggest that the reasons why students use recommendation-based admission vary by high school rank. Furthermore, students who were admitted through recommendation-based admission from high-ranked high schools tend to be desirable in terms of autonomy in both study habits at college and social activities, and they tend to recognize their large skill growth at college. This might mean that recommendation-based admission at selective colleges admit students with desirable characteristics that cannot be measured by written exams. However, regardless of the rank of the high school of origin, students admitted through recommendation-based admission are more likely to study seriously in class and to be satisfied with their college, compared to those admitted through written exam-based admission from high schools of the same rank. Moreover, no evidence supports their inferior performance at college. Negative perception of recommendation-based admission among the public probably comes from the fact that the admission is widespread among less selective colleges.