Solving school segregation with the matching mechanisms of schools

The aim of matching mechanisms is to collect information from market participants – schools and children – as ordered preferences, which are called priorities in the case of schools, and find an allocation that is safe and efficient for them.

This approach has recently been very fruitful in many real-life cases that explore the allocation of school places to students in primary, secondary and upper-secondary schools. Although generally speaking, no mechanism incentivises all participants to be truthful, there are some useful design principles. For example, if priorities are based on objective criteria like distance, the GS mechanism (also known as deferred acceptance) is safe and incentivises the true revelation of preferences for families. There is also evidence of ‘bad’ design, e.g. immediate acceptance known as the Boston mechanism, and it has been mostly replaced with the GS mechanism.

The number of existing applications is growing (see below) with the number of urban areas facing various social problems related to parental free school choice. These problems are related to urban segregation, which in turn feeds into school segregation; the distances children have to travel to schools and urban transport problems related to it, siblings in different schools and the ability of disadvantaged families or special education children to exercise school choices and more.

Therefore, designing the school priorities (what criteria to consider first), in addition to the allocation algorithm, is key to a good policy design. There is increasing literature to face new challenges of encouraging diversity with the use of quotas or priority classes that in many cases can enforce social justice, e.g. affirmative action approaches.

Table: Examples of some matching designs in Europe

Table is not representing a complete list of mechanisms; it contains the most well-known. Mechanism properties might change year by year. SEN = special educational needs, DA = deferred acceptance.

• ISCED 1: A=Amsterdam, P=Portugal, L=London, S=Stockholm, N=Nijmegen.
• ISCED 3: A=Amsterdam (before 2015 used version of Boston with randomisation in breaking ties); H=Helsinki (5 preferences and sequential version of Gale Shapley); Bu=Budapest (schools have to submit strict rankings); P=Paris (maximum 10 preferences can be listed only within districts); Br=Brussels (10 schools can be listed and matching is in two stages, Boston and DA).
• ISCED 6: G=Germany (DA from 2012, voluntary participation by universities); E=Estonia (centralised application environment, but no matching coordination, 3 preferences within one university); B=Belgium (all applicants are accepted); F=France; H=Hungary (all above the threshold of the grade score are accepted).

The table above shows that most urban mechanisms address either school-home distances or siblings in their designs and that the algorithm used is increasingly GS. However, the problems that are difficult to disentangle with market design are manyfold, such as the concern that high-stakes exams, distance priorities or the ability of middle classes to use their social class privileges are putting children of disadvantaged families in worse-off positions, especially in systems with early tracking.

Authors: Põder, Kaire; Veski, Andre; Lauri, Triin

 References

• Abdulkadiroğlu, A. and Sönmez, T. (2003). School Choice: A Mechanism Design Approach. American Economic Review, Vol. 93, No. 3, pp. 729–747.
• Abdulkadiroğlu, A., Che, Y.-K., and Yasuda, Y. (2011). Resolving Conflicting Preferences in School Choice: The Boston Mechanism Reconsidered. American Economic Review, Vol. 101, No. 1, pp. 399–410.
• Abdulkadiroğlu, A., Pathak, P. A., and Roth, A. E. (2009). Strategy-proofness versus Efficiency in Matching with Indifferences: Redesigning the NYC High School Match. American Economic Review, Vol. 99, No. 5, pp. 1954–1978.
• Abdulkadiroğlu, A., Pathak, P. A., Roth, A. E., and Sönmez, T. (2006). Changing the Boston School Choice Mechanism. National Bureau of Economic Research, Working Paper Series, 11965.
• Gale, D. and Shapley, L.S. (1962). College Admissions and the Stability of Marriage. The American Mathematical Monthly, Vol. 69, No. 1, pp. 9–15.
• Roth, A.E. (2008). Deferred Acceptance Algorithms: History, Theory, Practice, and Open Questions. International Journal of Game Theory, Vol. 36, No. 3-4, pp. 537–569.