Why Dublin "Doesn't Work"

Why Dublin "Doesn't Work"

Publication date:
Blanca Garcés-Mascareñas, Associate Researcher, CIDOB and member of GRITIM, Universitat Pompeu Fabra

E-ISSN: 2013-4428

D.L.: B-8439-2012

The current refugee crisis has shown up the already known, discussed and revised deficiencies in the Dublin system. Thus, on August 27th 2015, after participating in the Western Balkans Summit, which focussed on the refugee crisis, the German chancellor, Angela Merkel, stated in a press conference that the Dublin Regulation “doesn't work” and that we “need a common response for Europe as a whole”. While EU regulations establish that the asylum process should be carried out by the first European Union country that the applicant reaches, Merkel indicated that the European Court of Justice and the German Constitutional Court were blocking the return of refugees to Greece. And she added: “Where are we going to return them to? To Hungary, Austria, Serbia, who all have as many or more refugees than we do?” The German foreign minister, Frank-Walter Steinmeier, expressed the need to reform the Dublin Regulation for “fair distribution” of refugees in Europe.

To understand the criticisms of the Dublin system as well as the slow and, for many, shameful response of the European Union and its member states to what has been called the most serious refugee crisis since the Second World War, it is necessary to understand what Dublin is, what doesn’t work and why, and what are or would be the alternatives. 

What Dublin is

The creation of a European free movement area (the so-called Schengen area) brought the need to harmonise the asylum policies within the European Union to the table. This led to the negotiation of what came to be called the Dublin Convention (1990), which later became the Dublin II Regulation (2003) and Dublin III (2013).

Dublin does not seek to fairly distribute responsibility for refugees between the various member states but to establish the state responsible for processing each application quickly, based on some pre-established criteria. One of the main objectives is to avoid someone seeking asylum in the country of their choice (so-called “asylum shopping) or being present in Europe without any country taking responsibility for examining their request (“orbiting”).

For this, Dublin establishes three principles: 1) asylum seekers have only one opportunity to apply for asylum in the European Union and, if the request is denied, this is recognised by all member states; 2) the member state responsible for examining the application is established by the criteria set out in the Dublin Convention, rather than the preference of the applicants themselves; and, 3) asylum seekers may be "transferred" to the member state to which they have been assigned.

The criteria that define which member state is responsible have been clarified over various revisions of Dublin. Both Dublin II and Dublin III establish family unity as the first criterion: if the applicant has family members who have refugee status or are in the process of claiming asylum the application must be examined in that country. Otherwise, the state responsible is first the one where the applicant holds a residence permit or visa and second that through which the applicant entered the European Union. When none of these criteria apply, the state in which the applicant seeks asylum must take responsibility.

For the functioning of Dublin, in 2003 the EURODAC system came into effect (European Dactyloscopy), which is a centralised database for the handling and storage of the fingerprints of asylum seekers and people detained when crossing the EU's external borders. Using this system, the competent authorities can check whether it is their responsibility to respond to a request for asylum. The applicant may be returned to another EU member state if they are shown to have already applied for asylum in that country or to have crossed its borders to enter the EU. 

What doesn’t work and why not

There are basically three criticisms of the Dublin system. The first is that Dublin doesn't work fairly. Given that the most commonly-used criterion is that of the first country of arrival, the responsibility falls disproportionately (in theory at least) on the border countries. Being registered in the first country of arrival means being unable to seek asylum in other member states, or, in the case of doing so, running the risk of being returned. In 2013, for example, Italy received almost a third of the asylum seekers transferred from another member state.

The second is that Dublin doesn't work efficiently. It is inefficient because, despite the criteria of giving responsibility to the first country of arrival, most applicants seek asylum in a different country to the one in which they arrived. For example, according to Eurostat and Frontex statistics, only 64,625 of the 170,000 irregular arrivals in Italy sought asylum there. In 2013, more than a third of the asylum claims were made by people who had previously applied in another European Union country. Of those, 11% applied in Italy and did so again in Germany, Sweden or Switzerland.

Why do applicants not seek asylum in the country of arrival and, if they do, why apply again in another country? Because the criteria for assigning responsibility under Dublin do not match the preferences of the applicants themselves. According to the then European Commissioner for Home Affairs, Cecilia Malmström (2012), “It should not matter which country you flee to”. This is, in fact, the assumption on which the whole Dublin system is founded. Nevertheless, in practice this is not true. Just because asylum seekers are fleeing their own country, does not mean they are indifferent to where they end up.

The preferences of asylum seekers are often linked to personal concerns − such as the presence of friends and acquaintances in the preferred country and knowledge of the language − but they are also often connected to significant differences between reception countries, above all when it comes to asylum processes, reception conditions, social rights and the chances of finding work. Avoiding seeking asylum in a country that barely recognises refugees and has little reception infrastructure (such as Spain) is the result not of the feared asylum shopping but rather of the need for their most basic rights as refugees to be recognised.

If Dublin is inefficient in preventing asylum applications in a country other than that assigned, it is just as inefficient in fulfilling transfer requests between member states. In 2013 of the total of 76,358 requests, 56,466 were accepted by the reception countries but only 15,938 (20%) were really carried out (Fratzke, 2015: 11). The reasons are multiple. In some cases the decision is appealed by the applicants. It is also common for asylum seekers not to cooperate in their return to other European countries. In other cases, the reception countries do not accept the request because of lack of evidence or, in practice, they do not respond. For example, various member states have complained about communication problems when transferring asylum seekers to Italy.

Another factor that explains the difficulty of transferring asylum seekers to another country (normally that of arrival) is the action of the law courts. In 2011, various judgements from the European Court of Human Rights and the Court of Justice of the European Union banned returns to Greece because of systematic deficiencies in its asylum procedures and reception conditions. Since then, Italy has also been subject to criticisms, with cases in the courts of Austria, Switzerland and the Netherlands. The examples of Greece and Italy reveal the impossibility of assuming that the rights of refugees are guaranteed in all member states. Without this, the Dublin system falls apart. Basically because, as Merkel condemned, there is then nowhere to return them to.

Beyond the implementation difficulties, the member states that send most transfer requests are also those that receive the most. What is more, states often exchange a similar number of requests. For example, in 2013 Germany sent 1,380 requests to Sweden and received 947 back. This led the European Commission to propose a mechanism between states that would allow “redundant” transfers to be cancelled, that is to say, when the sum of those who go and those who come is similar. This mechanism, however, has not been included in either Dublin II or Dublin III. In consequence, the European Union continues to devote a large part of its resources to exchanging asylum seekers without it having a significant effect on the final distribution.

The third and final criticism is that Dublin jeopardises refugees’ rights. As condemned in the report by the European Council on Refugees and Exiles (ECRE, 2013), the fair and efficient examination of asylum applications is not guaranteed in all member states. Additionally, the allocation of responsibility is applied in a very disparate way, for example, by not taking into account the presence of family members, applying the humanitarian clause very restrictively and using the country of first arrival as the main criterion. Other reports, such as those by Fratzke (2015) and Guild et al. (2015) also denounce the way that, even if all the deadlines are met, the asylum seekers that are returned to another country must wait a year or more before their case is examined. Ultimately, all these reports also coincide in pointing out that the detention of applicants prior to transfer to another country is a common practice and that, in most cases, these transfers are made against their will.

As well as being criticised for being unfair, for being inefficient and putting the rights of asylum seekers at risk, Dublin is expensive. The costs of Dublin include the maintenance of EURODAC, the processes related to the transfer requests and the costs of the detention and deportation of those who are ultimately transferred. If Dublin works neither for the member states (for being unfair and inefficient) nor for the asylum seekers (for putting their rights at risk) and is also expensive, why do we carry on with this system?

The persistence of the Dublin system can only be understood if we consider that it is the result of a precarious balance of powers between countries with diverse circumstances and interests that are often opposed. While some ask for an integrated system that shares responsibility more fairly, others insist on maintaining national prerogatives. This tension is not necessarily an East-West division. While it is true that countries such as Poland, Hungary, Slovakia and the Czech Republic have opposed any compulsory sharing out of asylum seekers, it is no less true that countries such as Spain, Portugal and Great Britain have also shown notable reticence. 

The alternatives

Dublin grew out of the need to harmonise the asylum policies within a European area of free movement, but it fell short. First, because it only established some criteria for distributing responsibility. Second, because even as a minimum policy, Dublin does not work. The preference of asylum seekers and the distrust between member states, whether for letting asylum seekers pass or for not guaranteeing certain minimum rights, have left the foundations of the system in tatters. And without Dublin, there is no worthwhile Schengen. In 2011, faced with the arrival of 26,000 Tunisians in Italy, Silvio Berlusconi and Nicolas Sarkozy asked for reform of Schengen to adapt it to the “new pressures”. In 2015, in response to the entry of hundreds of thousands of refugees, the borders have gone up again in countries like Hungary, Slovenia, Slovakia, Croatia, Germany, Austria and Sweden.

If Dublin does not work it is not just a question of poor implementation. As we have seen, the most basic principles fail. Consequently, the alternative is not a new update of the regulations (of Dublin II and Dublin III) but a rethink of how to build a genuine common asylum policy. This means returning to a question as fundamental as how to distribute responsibility.Currently, the criterion of the first country of arrival means the responsibility falls disproportionally on the border countries. This encourages, on the one hand, more restrictive immigration policies by “punishing” with greater responsibility those countries with more open visa policies or less control of their borders. On the other, it leads the border countries to seek to avoid this uneven distribution by not registering the asylum seekers entering through their borders.

Since the start of the 1990s, when Germany took in 460,000 asylum seekers, the question of a fairer distribution of responsibility within the European Union has been on the table. The majority of proposals take into account the GDP of a member state, its number of inhabitants, the size of its territory, its unemployment rate and the number of refugees already there. The latest of these proposals, led by the European Commission to relocate 120,000 asylum seekers from Italy and Greece over the next two years, was approved on September 22nd 2015. Although this is a one-off redistribution of a very limited number of asylum seekers, the criticisms are overwhelming and, in fact, the proposal was approved despite the opposition of Slovakia, the Czech Republic, Hungary and Romania. One of the most controversial issues has been its compulsory nature. As the prime minister of Slovakia, Robert Fico, complained “Slovakia is a sovereign country” and this is a “dictate of the majority” on a very sensitive issue.

Thinking about how to distribute the responsibility means thinking about how to do it in a fairer way but also in a way that takes into account the preferences of the asylum seekers themselves. A distribution system that does not take asylum seekers’ preferences into account is not only ethically reprehensible but also terribly inefficient. On the one hand, many of the applicants avoid seeking asylum in the first country of arrival and, if they do, they then repeat it in another country. In other words, despite Dublin, they decide. On the other hand, the system of transfers between member states does not work either. It is true that the recently approved proposal for distributing 120,000 applicants includes, as a criteria, the “integration potential” of the applicants. Nevertheless, taking issues such as knowledge of the language, family relations and cultural and social ties into consideration is not the same as including them in the decision.

In this direction, diverse proposals have been made. In 2013, a group of German organisations (Pro Asyl, among others) proposed a model of distribution based on the free choice of the applicants. Their proposal consisted, basically, of replacing the country of first arrival with that of the applicant’s first choice. If the applicant had entered through another country, they would be given a permit and financial help to travel within Europe. If this resulted in unequal distribution between countries, compensatory funding would be offered to the countries with a higher number of asylum applications. In another direction, Rapoport and Moraga (2014) propose a system of negotiable quotas that take into account both the preferences of the refugees and those of the countries. At another level, Guild et al. (2015) advocate the free movement of refugees within the European Union, which would allow the minimisation of the importance of the country where the application was formally made.

As well as distributing responsibility, rethinking a common asylum policy also involves approaching how to harmonise standards on both asylum procedures and reception conditions. With this aim, in 2013, the European Asylum Procedures Directive and the Reception Conditions Directive were approved. So far, however, they have been ineffective. Currently, recognition rates remain tremendously diverse: in 2014, according to Eurostat, they ranged from 9% in Hungary and 11% in Croatia all the way up to 94% in Bulgaria and 67% in the Netherlands via 44% in Spain, 42% in Germany and 22% in France. The conditions of accommodation, food, health, employment and medical and psychological attention also vary considerably between the various member states.

The current European Commissioner for Home Affairs, Dimitrios Avramopoulos warned in September 2015 that, “It is of crucial importance that all Member States implement the same rules and the same standards to protect the rights of the migrants but also avoid secondary movements within the EU.” It should also be added that without common rules and standards the transfer of asylum seekers between member states ceases to be legitimate and even − following the intervention of the courts − legal. In summary, without common standards the three basic Dublin principles fail: the applicants seek asylum more than once, they do not necessarily do it in the country assigned according to the Dublin criteria and their transfer is problematic. A report by the European Migration Network (EMN, 2014) points out that to guarantee the application of common standards it is necessary to establish independent mechanisms of control both at national (for example, by an ombudsman) and international levels (through the UNHCR and other organisations).

Nevertheless, to think that by guaranteeing certain common standards in asylum procedures and reception conditions it will “not matter which country you flee to” (in Malmström’s words) is rather naive. Beyond personal concerns, the differences in social rights and above all the labour market are fundamental. In a study on Eritrean asylum applicants, Brekke and Brochmann (2014) show how the majority of refugees in Italy continue to be in a highly precarious situation even after years of residence there, while in Norway they acquire social rights from day one and have better work prospects. In summary, where one seeks asylum is fundamental not only in terms of recognition and primary reception but also when facing starting over again.

Finally, a fundamental question remains: What happens to asylum seekers who are not recognised as refugees? This past October 15th the European Council agreed to intensify return policies with the creation within Frontex of a dedicated Return Office. The German minister of the interior said that it would be possible to “accept and support” people in need of protection, only if those who do not need it stay away or are sent back quickly. Despite the political insistence, the limitations of a returns policy are well known. As well as with the transfers between member states, return is always difficult to carry out: first, because it goes against the desire of the “returnees” and, second, because it does not always count on the cooperation of countries of return. Furthermore, it is a policy that is economically costly and socially controversial. For all of these reasons it is well known that return is never (at least in terms of numbers) the solution.

Why then do we continue with the same (old) policies? When a policy is maintained despite not fulfilling its most basic objectives it is because the real objectives lie elsewhere. It is what, in political science, is called a “symbolic policy”. The current situation is paradigmatic. Although it is known that it is not possible to deal with the present refugee crisis without a genuine common European policy, a policy that is clearly inoperative but which keeps national prerogatives intact is preferred. Although it is known that deportations are never the solution, still the need for returns is insisted on with the objective of appearing “tough” and thereby appeasing a part of the electorate. Nevertheless, insisting on a policy that does not work has a high price: on the one hand, Europe does not respond to its most basic obligations (not only moral but legal) to asylum seekers, on the other, with its lack of a response, Europe justifies and strengthens the most ultranationalist and xenophobic stances.



Brekke, Jan-Paul y Brochmann, Grete. "Stuck in Transit: Secondary Migration of Asylum Seekers in Europe, National Differences and the Dublin Regulation". Journal of Refugee Studies, 2014.

ECRE. Dublin II Regulation: Lives on Hold, 2013.


EMN. The Organisation of Reception Facilities for Asylum Seekers in different Member States, 2014.


Fratzke, Susan. Not Adding Up – the fading promise of Europe's Dublin system. Migration Policy Institute, 2015.


Guild, Elspeth, Cathryn Costello, Madeline Garlick and Violeta Moreno-Lax (2015). "Enhancing the Common European Asylum System and Alternatives to Dublin", CEPS Paper.


Rapoport, Hillel and Moraga, Jesús Fernández–Huertas, "Tradable Refugee-admission Quota: a Policy Proposal to Reform the EU Asylum Policy". EUI Working Paper RSCAS 2014/101.


German Bar Association, AWO Workers' Welfare Association, Pro Asyl, Paritätische Welfare Association, Neue Richtervereinigung, JRS. Allocation of refugees in the EU: for an equitable, solidarity-based system of sharing responsibility, 2013.


Blanca Garcés-Mascareñas, Associate Researcher, CIDOB and member of GRITIM, Universitat Pompeu Fabra