Paris School of Economics - École d'Économie de Paris

Economics serving society

Hillel Rapoport: “A fair and efficient European response to the refugee crisis”

October 2015

What to do in face of the growing refugee crisis in Europe? According to Hillel Rapoport, “the most recent steps taken by the European Commission go in the right direction” but the measures are too timid to be effective. PSE examines the alternatives that he proposes, including a tradable quota market and a matching mechanism that takes into account the “preferences” of both refugees and countries.

Hillel Rapoport is an Associate member of PSE-Ecole d’économie de Paris and Professor at the University Paris 1 Panthéon-Sorbonne. He is co-director of the G-MonD research group.


Previous articles by H. Rapoport about this Tradable Immigration Quotas model:


This interview was realized in September 2015 (French version) and updated in December 2015 (English version).

When did you start studying the question of migration?

My work on migration dates from the late 1990s. I first studied migrant remittances, the question of “brain drain” and the effects of migration on inequalities within and between countries. From 2008, along with my co-author Jesús Fernández-Huertas Moraga, I started thinking about applying public economics models (the tradable rights system, matching mechanisms) to the question of the distribution of refugees within the European Union, and from 2010, we began to write on this question. Our point of departure was that international migration reduces global poverty, as a number of studies demonstrate. It follows that when a country admits poor migrants originating from poor countries, it contributes to the provision of an international public good: the reduction of global poverty. But this dimension is what economists call an “externality” and as such is ignored in the design of national immigration policies. As a result, the offer of immigration visas is lower than would be desirable from the point of view of host countries taken collectively. The same argument holds for another immigration-related international public good: the protection of international refugees. In our article entitled “Tradable Immigration Quotas”, published in 2014 (1), we developed a complete model of tradable immigration quotas on the one hand, and on the other hand, we detail a “matching” model that takes into account migrant preferences and the preferences of receiving countries in the final distribution. We also propose concrete applications, especially in the context of the refugee crisis in Europe and regarding the case of “climate-change refugees”.

Could you explain your model? And how has it been received by other scholars?

Our model has two distinct but complementary components. First of all, we study the modalities of a “quota market” through which countries can, once the total sum of migrants is known and distributed, reduce or increase their quotas through exchange. It gives to each country the choice of the modalities of its solidarity: welcoming migrants, or financing that welcome by others, at a price that renders compatible the choices of all sides. Once the quotas are known, we specify a model integrating preferences: migrants rank the countries they want to go to, and countries indicate their preferences as to the type of migrants they want to receive. The challenge is to arrive at an ideal “matching”. A final point concerns the case of a country whose quota remains unfilled because it is not listed by the migrants as a desired destination: these countries must pay a penalty equal to the unfilled part of its quota once the price of tradable admission rights has been determined; by definition, this amount is higher than the real costs (including material and social) of welcoming refugees. The sanction thereby constitutes an incentive to become more attractive in the medium term - for example, by improving its conditions of reception.

This kind of model, called the “matching” model is based mainly on the work of Alvin Roth, Nobel prize-winner in economics in 2012, and also closely follows the current debates on the refugee crisis (2). Several applications of these models have been made in the fields of school choice, kidney exchange, or assignment of interns to hospitals. As for the tradable rights markets, their principal application is in environmental remediation and preservation. Our proposition of a mechanism of tradable refugee-admission quotas with matching has been very favourably received by university scholars in a range of seminars, but of course they are more receptive to theoretical arguments than political decision-makers. I hope to advance the debate around our proposals at both academic and policy levels through presentations in workshops (e.g., the “Matching in Practice” workshop held in December 2015 at the Toulouse School of Economics) and policy events (e.g., my recent “DevTalk” presentation at the OECD, also in December 2015).

What are your thoughts about developments in the recent months?

In mid-May 2015, the European Union proposed the introduction of quotas for the reception of refugees by member countries. On the European Agenda on migration, this new obligation was accompanied by a security dimension and an anti-smugglers dimension. We were talking then about 40,000 asylum seekers (mainly Syrians and Eritreans) to be relocated from Italy and Greece, plus 20,000 persons who had obtained refugee status and were still in refugee camps in the Middle East. There were four criteria in the definition of quotas per country: the wealth of the receiving country (GDP), its population, its rate of unemployment and its past efforts in welcoming refugees.

Given that this proposal was on the table, I felt authorized to comment on it and take a stand. I am not sure economists have much to say as to whether such immigration entails a net benefit or cost for the receiving societies, or on whether Europe should admit 100,000 or 1 million refugees. I see this more as a political matter. However, economists can clearly play an expert role when it comes to design fair and efficient ways to allocate refugees – whatever their number – among EU member countries willing to participate in the proposed quota system. Especially, we can provide answers to questions such as: How to minimise the total cost of accepting refugees? How can we integrate the preferences both of the refugees in terms of their desired destinations and of receiving countries in terms of the type of refugees they want to host? Can we design a system which would give to receiving countries incentives to increase, not decrease, their standards for welcoming refugees (instead of the race to the bottom we are witnessing in the current system) and at the same time provide incentives to opt in the quota system, not out?

At the beginning of June, you made a series of proposals during a hearing at the European parliament…

Indeed, I have been invited on the third of June to a public hearing at the European parliament (3), to explain in detail the implications of our model. This hearing, an initiative of the Green Alliance, aimed at rethinking the European asylum system. Until now, it has been governed by the 1990 Dublin Convention, which puts the responsibility for refugees on the country of first entry. As a result, a large number of refugees make the choice to remain clandestine until they reach the country in which they wish to demand asylum. Moreover, this system becomes more obsolete every day with the continuing influx of refugees. This initial contact seemed conclusive to us and the reports from that public hearing made mention of our recommendations – without going any further. At the end of June, just before a European council dedicated to these questions, I published an article (4) on the London School of Economics blog EUROPP – European Politics and Policy – to explain what our model could contribute to improving the political response to the refugee crisis.

In the past several weeks, everything has accelerated: what do you think of the most recent proposals in the Juncker plan?

Honestly, I didn’t think that the European Commission would introduce the quotas. In fact, for some member countries, it is probably really extremely costly politically and socially to accept refugees, for reasons beyond the scope of this interview. However, following the publication in the first days of September of the pictures of a 5-year old Syrian boy, Aylan, lying lifeless on a Turkish beach, an impressive turn-around of public opinions occurred and shifted the balance of power towards the group of countries that were favourable to more openness and generosity. This also triggered a more ambitious and encompassing proposal by the Commission a few days later.

To answer your question more precisely, the steps taken by the European Commission on the ninth of September goes in the right direction. Mr. Juncker presented a range of realistic proposals in the face of the urgent situation. The number of asylum seekers to be allocated was revised to 160,000 and a coercive mechanism was envisioned for the countries that refused to take them in. After justifying themselves before the Commission, these countries would have been obliged to pay a fine in the order of two thousandth percentage points (0.002%) of their GDP. The proposals also mention Europe’s will to take into account the needs of both states and refugees in the general allocation.

What is missing from this plan?

There again, I’m not going to comment on the number of asylum seekers involved. While the notion of monetary compensation is interesting, it presents difficulties in application because of its rigidity, which reduces its pertinence. Simple calculations show that if Spain decided not to conform to the quota policy, the fine that it would have to pay would amount to 1,000 euros per refugee not accepted. For France, the amount would be in the order of 2,000 euros. This is low in comparison to existing European support mechanisms for refugees, which envisage a budget of six to eight thousand euros per refugee. We can only speculate what would be the price determined on a market for refugee-admission quotas but I think it is safe to bet that price would be in a 5 or 6-digit figure. Furthermore, numerous studies show that the best indicator of future integration of migrants is the preference they express for a particular country – unfortunately, this element is for the moment considered secondary and so nothing is done to take it into account in a systematic manner.

The two main weaknesses of the Commission’s proposals rest on the lack of a mechanism for the efficient distribution of refugees, and the fact that it is not taking into account the preferences of migrants, who often prefer being clandestine in their destination of choice over having the status of refugee in a destination imposed on them. Our proposals allow the resolution of precisely these two elements, completely for the first (through the market in tradable admission rights), and partially for the second (thanks to the matching mechanism).

The system put in place by the Commission and adopted in its broad outlines by the European Council on September 23rd (even if the mandatory quotas were replaced by voluntary ones, something which does not fool anybody) is thus a first step in the good direction. But this system can only be effective and durable on condition of transparency; it must therefore be complemented by mechanisms that allows for revealing the true costs of accepting refugees according, reducing these costs through the market for tradable-admission rights and through the possibility for countries to express their preferences over the type of refugees they want to host (e.g., according to family status, country of origin, language spoken, etc.) while at the same time avoiding a race to the bottom in humanitarian standards… Our proposed mechanism does just that.

If your proposal were accepted tomorrow, how would it work?

Our proposal could be put to work on two levels, which correspond to the two components of the model described earlier.

One part is a market for admission rights. Some people will view this as a commodification of human beings, so let’s be very clear: the exchange is about admission rights, not about individuals, and takes place even before the relocation of refugees. The other part is the matching mechanism, which takes into account the preferences of the two sides, refugees and destinations. At the country level, preferences must of course meet certain criteria, defined during collective discussion: languages spoken by the refugees, country of origin, job qualifications, profession, family status. As to the migrants, they would indicate by order of preference the countries in which they would like to settle. From there, we can envision several possible algorithms to do the matching; it is hard to detail their differences without entering into complex technical matters. From our point of view however, so long as a refugee is not sent to a non-desired destination, we have no strong preference for one mechanism or another.

The government of Sweden – at present the country that has taken in the most refugees per head of its population – has invited us to write a report evaluating the different possible workings of our proposed mechanism against the background of the current crisis: which countries would win or lose from the introduction of a tradable admission-rights market? What can we expect in terms of refugee relocation for different candidate matching mechanisms? We hope to answer these questions in the near future and will advocate for an experimental approach allowing to evaluate the performance of the system we propose vis-à-vis the status quo.


(1) « Tradable immigration quotas », Journal of Public Economics - Volume 115, July 2014, Pages 94–108
(2) http://marketdesigner.blogspot.fr/2015/09/refugee-resettlement-long-term-policy.html
(3) “Beyond Dublin: Rethinking Europe’s Asylum System” http://greenmediabox.eu/en/ct/90
(4) “How a tradable refugee-admission quota system could help solve the EU’s migration crisis” http://blogs.lse.ac.uk/europpblog/2015/06/22/how-a-tradable-refugee-admission-quota-system-could-help-solve-the-eus-migration-crisis/


Revue de presse

  • How economics could solve the refugee crisis, Brookings, 16/06/2016.
    Is there an economically grounded solution, with clear benefits for host countries, displaced people, and their home countries, to the current refugee crisis? Consistently growing since 2012 and fueled by the civil war in Syria, the worldwide count of refugees according to ... suite
  • Réfugiés et quotas : il faut laisser à chaque pays le choix des modalités de sa solidarité, Le Monde, 22/10/2015.
    Les tragédies de l’été ont bouleversé les opinions publiques et provoqué un changement radical dans les positions des dirigeants européens, désormais ralliés majoritairement à l’idée de quotas d’admission de réfugiés... suite
  • eBay für Flüchtlinge?, par Jakob Arnim-Ellissen. Format, 06/11/2015.
  • Crise des réfugiés et quotas : des solutions pour mieux répartir les flux, The Conversation, 10/11/2015.
    Les tragédies de l’été et leur retentissement ont à nouveau propulsé la question des réfugiés sur le devant de la scène, suscitant cette fois le retournement des opinions publiques. La plupart des dirigeants européens ont ainsi été conduits à réviser leur position... suite
  • Free Lunch: Economic home truths. Financial Times, 21/06/2016.
    Last month I suggested that a dose of market fundamentalism could improve European refugee policy. A system of tradable refugee quotas, modelled on the existing carbon emissions trading system, could both draw the sting from the perceived attack on national sovereignty... suite

This interview was realized by PSE-Ecole d’économie de Paris and is available via the section "Economics for everyone.