Choices, preferences and priorities in a matching system for refugees

Will Jones and Alexander Teytelboym

We propose a 'matching system' that simultaneously gives refugees some choice over where they seek protection and respects states' priorities over refugees they can accept.

Syrians fleeing the current conflict have been repeatedly told that they cannot ‘choose’ the state in which they seek long-term protection. In Australia, the idea that asylum seekers are ‘shopping’ for the best sanctuary forms a persistent part of the rhetoric around keeping them out. In these and other cases, the premise is that it is unjustifiable for refugees to be allowed some choice over where they seek protection. The consequence enshrined in the Dublin Regulation is that refugees may apply for asylum in only one European Union country.

From the perspective of states, refugee flows are chaotic, unpredictable and widely regarded as socially disruptive and destabilising. Everyone recognises that the Dublin Regulation, which seeks to address this by placing the obligation to render asylum on the first EU country an asylum seeker reaches, is not fit for purpose. In parallel, there is an urgent need to design systems to overcome the political deadlock among European states over asylum.

The ‘Refugee Match’

We propose a system which can both give refugees choices over where they are to be protected and enable states to manage the sharing of responsibility for granting asylum in a way which is equitable and efficient.[1] The way in which we allocate students to schools, junior doctors to hospitals and kidneys from living donors to recipients is by ‘matching’ the two sets. Refugees need to be ‘matched’ to states in precisely the same way in order for them to be protected. Furthermore, we want a system which participants on both sides will want to participate in, which will best satisfy their preferences and desires, and which will do so in a manner that is equitable and transparent. It could even give states currently unwilling to share responsibilities additional incentives to get involved.

Concretely, in our proposal, states and refugees submit their preferences – about which refugees they most wish to host or which state they most wish to be protected in – to a centralised clearing house which matches them according to those preferences. Refugees, in principle, could submit their preferences from anywhere, saving them the risk of a dangerous journey and the extortion of people smugglers. This system involves no payment, works where there are quotas or other constraints, and can be made to work so that it is:

(1) comprehensive – all refugees within the system are hosted somewhere (with quotas agreed by participating states adding up to the total number of refugees seeking places ‘in the marketplace’)

(2) stable – refugees and countries do not end up dissatisfied with their choice and wanting to ‘re-match’ by undertaking secondary movements

(3) efficient – no refugee can be made better off without making at least one other refugee worse off.

Finally, it can be made ‘safe’ for states and refugees to honestly reveal their true preferences.

Beyond this, there is a lot to be determined. It would be for the designers of the system to decide which refugees the match would apply to, and what sorts of preferences states and refugees were allowed to express. For example, the system could be designed to allow states to identify priority categories based on skills gaps; this might be useful in persuading states in Eastern Europe with labour shortages to participate. There may be some reason why states would wish to decide in advance that the refugee populations they take must meet some ‘distributional’ requirements. For example, states could collectively pre-commit to taking a diverse population of refugees and this feature can be built into the system. The designers of the system would face many choices in order to meet whatever set of goals was decided upon. Our claim is only that, whatever those goals are, a matching system will deliver these goals better than the current system.

It is very unlikely that all states will have the same preferences. Even if all states ended up ranking refugees in the same way, the clearing house would still be an improvement on the status quo, as the preferences of the refugees themselves would become the deciding factor in determining who went where.

However, there are a variety of principles which could be used in trying to determine who will be prioritised, given practical and political limits on how many refugees can be taken in. For example, the UK government has stated that its priorities are determined by greatest need and an assessment of where the UK can singly make the greatest difference. On the other hand, the governments of Slovakia, Poland and the Czech Republic have all signalled a willingness to take more refugees but only if they are Christians. Whether these principles are seen as discriminatory or as largely uncontroversial, the point is that different states are already free to rank these principles differently, and other states might rank the same refugees differently; for example, it would be eminently reasonable for Brazil and France to prefer Lusophone and Francophone refugees respectively.

Similarly, refugees will have a variety of preferences. There are abundant reasons to believe that the preferences of refugees are as heterogeneous as they themselves are. Currently, refugees must prioritise reaching the location where they feel they are most likely to be protected. In consequence, we know relatively little about the choices refugees would make if they knew they were guaranteed protection somewhere. We would like to find out.

Using a matching system per se does not dictate which principles states are allowed to use in ranking refugees, and the clearing house could permit or forbid the use of any criteria. Just as matching for doctors should not allow hospitals to engage in racist hiring practices, the clearing house would only allow states to rank refugees based on criteria which are compatible with the principles and goals of the 1951 Convention, and maybe other sets of principles.

Of course, in order to actually solve refugee crises, states would have to accept enormous inflows of refugees and find a way to resolve the ongoing conflagration in Syria and elsewhere. Matching systems, such as the one we propose here, are never the total solution to the various issues they seek to address. They are merely a substantial improvement on the status quo within the constraints of what is politically palatable, and may give states incentives to relax these constraints. Although matching mechanisms cannot make states behave morally, they will nonetheless improve the situation for refugees, whether or not states can be made to act in accordance with their legal and moral obligations. This is therefore a pragmatic proposal in the spirit of those who argue that states will contribute towards efforts to protect refugees when they recognise a relationship between the rights of refugees and their own interests.

The Refugee Match is a realistic, pragmatic, quickly implementable and just improvement on much of the current international refugee regime. A matching system, which respects the preferences and choices of refugees and the priorities of states, can better protect the human rights of the vulnerable, and increase the likelihood that states will participate in sharing responsibilities for the international protection of refugees. Any system which genuinely upheld the rights of refugees would have to start by respecting their choices. Asylum seekers ought to be able to choose the states where they want to spend their lives. The Refugee Match would be a good start.


Will Jones

Departmental Lecturer, Refugee Studies Centre, University of Oxford


Alexander Teytelboym

Research Fellow, Institute for New Economic Thinking at the Oxford Martin School, University of Oxford

[1] Jones, W and Teytelboym, A ‘The Refugee Match’, under review. Available on request from the authors.


FMR 51
January 2016


Opinions in FMR do not necessarily reflect the views of the Editors, the Refugee Studies Centre or the University of Oxford.
FMR is an Open Access publication. Users are free to read, download, copy, distribute, print or link to the full texts of articles published in FMR and on the FMR website, as long as the use is for non-commercial purposes and the author and FMR are attributed. All articles published in FMR in print and online, and FMR itself, are licensed under a Creative Commons Attribution-NonCommercial-NoDerivs (CC BY-NC-ND) licence. Details at