Under the current Common European Asylum System (CEAS), asylum seekers arriving in the European Union must lodge their application in the country where they first arrive. This decentralized, "no-choice" approach places a disproportionate burden on border countries, leads to delays and disputes over responsibility, and incentivizes asylum seekers to undertake illegal secondary movements to reach their preferred destinations. How can we design a better system, one that respects asylum seekers' preferences, member states' priorities and capacities, and the EU's proposed burden-sharing commitments?

This paper answers that question using tools from matching theory and market design. The authors model the asylum system as a "matching with contracts" problem, in which each contract specifies an asylum seeker, a member state, and a waiting time. Member states face two types of constraints: bureaucratic capacities (the number of applications they can process in a given period) and burden-sharing quotas (the total responsibility each state has committed to shoulder). Each asylum seeker is assigned a "burden-size," allowing the model to distinguish between individuals and families applying together.

The paper first identifies a unique choice rule for member states that satisfies three natural properties: feasibility, early filling (minimizing wait times), and respect for member state priorities. Building on this rule, the authors analyze the asylum-seeker-proposing cumulative offer mechanism, a centralized clearinghouse procedure in which asylum seekers propose their most preferred country and waiting time combinations.

When all asylum seekers are treated as having identical burden sizes, the mechanism guarantees both stability (no asylum seeker and member state would mutually prefer to deviate from the assignment) and strategy-proofness (asylum seekers have no incentive to misreport their preferences). However, when burden sizes vary, for example when families count more than individuals, these guarantees break down: stable outcomes may fail to exist, and no stable mechanism can be strategy-proof. Encouragingly, when member states prioritize larger-burden applicants, the mechanism still achieves stability and is only "non-obviously manipulable," meaning no simple, obvious manipulation is profitable.

The findings carry clear policy implications: a centralized asylum matching system is feasible and desirable, but treating individual and family applications equally in the assignment process is essential to preserving its good properties. Burden imbalances arising from family admissions can instead be corrected over time by adjusting member states' future quotas, offering European policymakers a practical roadmap for a centralized asylum assignment system and responsibility-sharing.