Panel Paper: A New National School Assignment System Strengthening School Choice in Chile: Reducing Inequality By Improving Integration and Access?

Saturday, November 10, 2018
Lincoln 3 - Exhibit Level (Marriott Wardman Park)

*Names in bold indicate Presenter

Ngaire Honey and Alejandro Carrasco, Pontificia Universidad Católica de Chile

Schools in Chile have implemented independent and often restrictive admissions policies, contributing to segregation as selective schools only accept higher SES students, leaving lower SES students in schools of perceived lower quality with open admissions criteria (Bellei 2009, Contreras et al 2011, Carrasco et al 2017). Chile is phasing in a centralized admission system (NSAE), where schools must accept any applicant, and lottery assignment is used for oversubscription. We draw from theories of educational opportunity. Institutional constraints, including restrictive admissions policies, present barriers to equitable participation in school choice, preventing students from certain social classes from participating and leading to further institutionalized inequality (Lubienski, Gulosino, & Weitzel, 2009). Eliminating admissions barriers may contribute to greater equity in school choice participation, and increased integration and educational opportunity.
Research Questions: • Under the NSAE, to what extent are students from distinct subgroups assigned to similar schools?
• To what extent do students from distinct subgroups apply to similar schools?
• To what extent does the NSAE equalize opportunity for students regarding school placement?
• To what extent are differences in characteristics of the school attended by student subgroup smaller than under the previous system?

Data & Methods:
Administrative records provided by the Ministry of Education from the 2016- 2018 school years from the five regions of Chile participating in initial implementation, including the ranked preference of schools for each student, priority status for each school, school assigned, basic student characteristics, and an ID that links to additional administrative datasets with standardized exam scores and survey data. The school and grade level data provide basic school characteristics and an ID linking to additional national datasets.
Descriptive statistics are analyzed for subgroups of students based on SES level and ethnicity. Logistic, mixed logit, and OLS regressions are conducted to examine differences across subgroups in the odds of being assigned one’s preferred school, of selecting (and then of being assigned) a school with certain characteristics controlling for student and school observables.
Question d) utilizes descriptive statistics as well as differences-in-differences. The first year of implementation was for students entering PK, K, 1st, 7th, and 9th grades. Students transitioning into these grades in 2017 are considered treatment eligible, while students transitioning into 2nd, 3rd, 6th, 8th, and 10th grades in provide a comparison group. Access and segregation outcomes are operationalized in multiple ways.
Preliminary findings suggest non-indigenous students, students with higher income, SES, or with educated parents have lower odds of assignment to their preferred school. Students’ SES is reflected in the SES level of the school they are assigned, primarily due to students from different levels of SES selecting schools with different characteristics. Non-indigenous students and students with more educated parents prioritize schools with higher test scores, a higher level of SES amongst peers, and schools with higher demand. We find slight increases in integration, particularly for students residing in urban areas. Changes may be minor due to continued constraints on choice set formation for isolated students and self-selection.