Eighth-grade students in Amsterdam face an unfair secondary school admissions process that rewards parents who lie about their preferences, according to doctoral research by Mayesha Tasnim at the UvA Civic AI Lab.
Under the Central Lottery and Matching system, students submit a list of twelve schools and receive a random number that determines the order of placement. If all twelve choices are full, students are placed manually through a placement guarantee using reserved spaces.
How strategic choices distort placements
The system creates incentives for families to game the system rather than list their genuine choices. Parents can list their true top schools first, then fill the remainder with oversubscribed schools, raising the odds of triggering the manual placement guarantee to secure a top choice.
Families who understand how to manipulate the process gain an advantage over those who submit truthful lists. The tactic only works if a limited number of families do it, and parents who try it risk landing further down their lists if the strategy fails.
The issue is unusual because Amsterdam schools do not evaluate or rank applicants based on exams or interviews, unlike admissions models used in cities like New York.
Calls to overhaul the matching algorithm
In 2017, the process left some children assigned to their sixteenth, seventeenth or eighteenth choices. An algorithmic update introduced in 2018 intended to fix those mismatches instead created the loopholes for strategic selection.
Tasnim argues that the school-board umbrella organisation Osvo could switch to a fully random lottery without the placement guarantee, or adopt an optimised algorithm designed to place children within their top six or seven choices.
City officials defended the setup to the research lab, stating it relies on a Nobel Prize-winning design. Tasnim argues the original model was altered heavily to work in Amsterdam, where schools have no say over which pupils they receive.

