Secondary school placement problem in Turkey

Abstract

In this paper, I study the placement mechanism, a centralized student placement via standardized test, which is used for matching eight grade students and high schools, in Turkey within a many-to-one matching framework. The placement mechanism used is the two-stage segmented system with multi-category serial dictatorship. I show that this system is fair, but fails to satisfy non-wastefulness, strategy-proofness, efficiency and respecting improvements. I further show that, under the constraint of placing students to private and state schools in separate markets, there exists no fair and non-wasteful placement mechanism that satisfies strategy-proofness, efficiency and respecting improvements. I, then introduce two restrictions on the students’ preference profiles; blocked preferences and common preferences. I show that; in those restricted matching environments using Gale and Shapley Student Optimal Deferred Acceptance Algorithm instead of multicategory serial dictatorship makes the system be the best placement mechanism among all stable matching mechanisms.