Skip to main content

A Comprehensive Approach to School Choice and Education

Periodic Reporting for period 3 - CompSCHoice (A Comprehensive Approach to School Choice and Education)

Reporting period: 2017-10-01 to 2019-03-31

School choice is one of the most hotly debated policies in education. Advocates argue that school choice allows equal access to high quality schooling for all. High-income families have always had more choice, either through residential choice or through enrollment in private schools. Therefore increased choice should also improve equity by allowing minority and low-income students to choose too. On the other hand, school choice critics suggest that school choice can increase sorting of students between schools based on their socio-economics status, suggesting high-income families benefit more from these policies. The effects of school choice on schools and individual performance are also key to this debate.

Three different and disconnected literatures in economics provide different and often contradicting answers to these questions. This project proposes a unified theoretical framework that merges these three literatures and allows for a comprehensive analysis on school choice design and its impact on actual choice, outcomes and segregation in schools and neighbourhoods. The work is organised along several subprojects and unique and newly constructed data sets from Barcelona and Catalunya are used to address novel empirical challenges. The data constructed for Barcelona shall become one of the largest and most comprehensive data sets not only on school choice but also on public education worldwide.

It has become evident in recent years that the impact the educational policies may have shall not be only measured through test scores. In education has been shown to impact not only cognitive skills, but also non-cognitive skills. The problem is that we do not have adequate measures for these skills. This project aims to initiate of getting behavioral measures of those traits, which shall not only allow us to evaluate policies in terms of those traits, but will also set the target for educational agents.
The following describe the progress made on different fronts.

1) School choice and sorting in school. We have worked on a first draft of the paper that is now R&R at the Economic Journal. We are now revising the model to generalize the model and present it in a more pedagogical manner so that a broader audience can read it. The paper is currently named “Sorting in Public School Districts under the Boston Mechanism”. In the paper we show that the widely used Boston Mechanism (BM) fosters ability and socioeconomic segregation across otherwise identical public schools, even when schools do not have priorities over local students. Our model includes an endogenous component of school quality determined by the peer group and an exogenous one. If there is an exogenously worse public school, BM generates sorting of types between a priori equally good public schools: an elitist public school emerges. A richer model with some preference for closer schools and flexible residential choice does not eliminate this e§ect. It rather worsens the peer quality of the nonelitist school. The existence of private schools makes the best public school more elitist, while reducing the peer quality of the worst school. The main alternative assignment mechanism, Deferred Acceptance, is resilient to such sorting effects.

Within this line of research there is the paper “Catchment Areas and Access to Better Schools” where we compare popular school choice mechanisms in terms of children's access to better schools (ABS) than their catchment area school, in districts with school stratification and where priority is given for residence in the catchment area of the school. In a large market model with two good schools and one bad school, we find that both the Boston Mechanism and Deferred Acceptance, the most popular assignment mechanisms, convey a non-negligible possibility that catchment area priority determines the final assignment to a large extent regardless parents’ preferences. Top-Trading Cycles is an alternative that provides more access to better schools than DA. This paper is under revision at the Journal of the European Economic Association.

2) Structural Estimation of Preferences and Counterfactual Analysis of Mechanisms. We have now submitted the fourth revision of the paper to the Journal of Political Economy. We model household choice of schools under the Boston mechanism (BM) and develop a new method, applicable to a broad class of mechanisms, to fully solve the choice problem even if it is infeasible via the traditional method. We estimate the joint distribution of household preferences and sophistication types using administrative data from Barcelona. Counterfactual policy analyses show that a change from BM to the Deferred Acceptance mechanism would decrease average welfare by 1,020 euros, while a change to the top trading cycles mechanism would increase average welfare by 460 euros.

Within this line also the paper “Priorities in School Choice: The Case of the Boston Mechanism in Barcelona” is in its second revision at the Journal of Public Economics.

The Boston mechanism is a school allocation procedure that is widely used around the world and has been criticized for its incentive problems. In order to resolve overdemands for a given school, most often priority is given to families living in the neighborhood of the school. Using a very rich data set on school applications for the case of the Boston mechanism in Barcelona, we exploit an unexpected change in the definition of neighborhood. This change allows us to identify that a large fraction of families systematically ranks first high priority schools, neighborhood schools in this case. Additional data on school enrollment decisions and census data shows that some seemingly unsophisticated parents are high income families that can rank hard-to-get schools because they can afford the outside option of a private school in case they do not get in. This sheds light on important inequalities beyond parents’ lack of sophistication found in the literature.

3) Government Preferences and Priorities. The purpose of this part of the project is to understand the government’s objectives when designing school choice procedures. On this front we have made great progress in accessing the data of all the school zones in the 900 municipalities in Catalonia.

4) Month of birth and performance in school. On this topic we have worked on two different topics. The first is directly looking at the impact of the month of birth on performance in school. The paper is called “Maturity and School Outcomes in an Inflexible System: Evidence from Barcelona” and is R&R at the Journal of the Spanish Economic Association. Having a unique cut-off to determine when children can access school induces a large heterogeneity in maturity to coexist in a classroom. We use very rich data on performance in internal and external evaluations in public schools in Catalonia to show that: 1) Relatively younger children do significantly worse both in internal and external evaluations and experience greater retention, although the effect is decreasing as children become older; 2) Younger children in our data exhibit higher dropout rates and chose the academic track in secondary school less often; 3) Heterogeneity analysis: the effect is homogenous across SES and significant across the whole ability distribution; 4) Younger children are more frequently diagnosed with learning disabilities.

On the other hand we have used the month of birth as an instrument for peer quality to understand how teachers grade students in schools on relative terms, that is, by comparing their performance to that of their peers. We call this phenomena grading on a curve. Student access to education levels, tracks or majors is usually determined by their previous performance, measured either by internal exams, designed and graded by teachers in school, or external exams, designed and graded by central authorities. We say teachers grade on a curve whenever having better peers harms the evaluation obtained by a given student. We use rich administrative records from public schools in Catalonia to provide evidence that teachers indeed grade on a curve, leading to negative peer e ects. This puts forth a source of distortion that may arise in any system that uses internal grades to compare students across schools and classes. We find suggestive evidence that school choice is impacted only the year when internal grades matter for future prospects.

5) Measuring non-cognitive traits. The objective of this part of the project is to provide novel measures of non cognitive skills that shall help us evaluate school systems on these very relevant dimensions. We have run a pilot in some schools using the methods in psychology and neuroscience and working together with teachers to adapt the measures to the activities that can be run within a classroom.
We are making substantial progress in understanding how the design of school choice and of different assignment processes can have on the efficiency and fairness of the resulting assignment in neighbourhoods and schools.

We are starting to make some progress in the assessment of non-cognitive skills.

We are providing new empirical and methodological tools to provide new insights on policy design that shall have a large impact on policy design, but also on how we evaluate policies in the future.