EU states agreed on 23 September to implement a refugee quota system which will distribute 120,000 refugees across the EU, despite four member states – the Czech Republic, Hungary, Romania and Slovakia – voting against the proposal. Luc Bovens and Anna Bartsch write that regardless of the wider debate over whether a quota system is justified or not, it is vital that the ‘distribution key’ determining how many refugees are assigned to each state is fair. They argue that the distribution key proposed by the European Commission is ill-conceived and regressive, and that a fairer system with recalculated quotas may go some way toward convincing the dissenting states to support the system.
In an emergency meeting on 23 September, the leaders of the EU endorsed a proposal from the European Commission to distribute 120,000 people who are likely to be in need of international protection and are currently held up in Greece, Italy and Hungary. This will apply to nationals fleeing from countries who have recognition rates in the EU larger than 75 per cent – currently Syria, Eritrea and Iraq. The applicants will be spread between all other states, except for the UK, Ireland and Denmark who are exempt on grounds of the Lisbon Treaty.
Hungary, the Czech Republic, Romania and Slovakia voted against the proposal, whereas Finland abstained. According to a recent BBC report the Czech government argues that quotas deny states their sovereign right to decide their own asylum policies, encourage more people to come, and cannot be enforced since refugees will just move around anyway.
How fair is the refugee quota system?
Irrespective of the wider question of whether a quota system is desirable or not, if we are to have a quota system then the distribution key should be well motivated and fair in its implementation. The distribution key in the proposal of the Commission is neither.
The distribution key determines how many applicants each state should take. It is based on the size of the population of each state (40 per cent) and on its GDP (40 per cent). The remaining 20 per cent are ‘corrective factors’ based on the average number of asylum applications received in the last four years (10 per cent) and the unemployment rate (10 per cent). Less will be expected from states who have taken in a substantial share of applicants in the last four years and who have higher unemployment rates. The numbers in the Commission’s Press Release are shown in column A of Table 1 below.
Table 1: Distribution keys for the refugee quota system
Note: The intention of the quota system is to relocate 120,000 refugees from Greece, Hungary and Italy (although Hungary was later removed from the proposal), so these countries are not shown in the table. Denmark, Ireland and the UK are also not shown as they are not participating in the system. For a full explanation of the Commission’s proposal, see the accompanying press release. The data from the European Commission is shown in column A and World Bank data 2014, except for Luxembourg and Malta 2013 is shown in column B and C. Those states that would have higher shares of refugees under the allocation used in column C are highlighted in red. Cyprus is highlighted for the reasons outlined below.
Let us bracket the corrective factors: They make little difference since they often balance each other out and are not heavily weighted to begin with. We will focus on the two main determinants, population size and GDP. The Commission’s reasoning is exceedingly simple. Clearly more is expected from states with larger populations, all other things being equal, and more is expected from richer states, all other things being equal. So, the reasoning goes, quotas should be proportional to the weighted sum of population size and GDP. When we bracket the corrective factors and assign equal weights of 50 per cent to population size and 50 per cent to GDP then we get the quotas in column B. These are quite close to the quotas of the Commission with the corrective factors in column A.
But is an additive formula with equal weights the correct way to set the distribution keys? We do not think so. Let us think about another case of responsibility sharing: when we need to pay for a public good. In this case we look at each state’s capacity to make a contribution. This capacity is determined by how rich the state is. The GDP per capita is a crude measure of the riches of the average citizen in each state. If the average Luxembourger is (roughly) eight times richer than the average Croatian then a principle of equal proportional sacrifice would require that the average Luxembourger pay eight times more than the average Croatian. But there are also eight times more Croatians than Luxembourgers. So Luxembourg and Croatia should pay roughly the same towards the public good. The average German is five times richer than the average Romanian and there are 4 times more Germans. So Germany should pay roughly 20 times more towards the public good.
The reasoning is analogous for responsibility sharing for the reception of applicants for refugee status. Considering their capacity to take in applicants, as measured by relative riches, the average Luxembourger should shoulder eight times more applicants for refugee protection than the average Croatian. And considering that there are roughly eight times more Croatian than Luxembourgers, both states should take in roughly equal numbers. This is in line with the principle of equal proportional sacrifice in which the capacity to make a contribution is measured by relative riches and population size is given due consideration.
We calculate the quotas following this new distribution key in column C. Notice how the principle of equal proportional sacrifice asks substantially more from the richer states highlighted in red and substantially less from the poorer states that are not highlighted. In other words, the Commission’s distribution key offloads the responsibility from the backs of richer states onto the backs of poorer states. The distribution key is strongly regressive and unfair towards poorer states.
The richer Luxembourg takes on less than half of the applicants of poorer Croatia whereas it should take on roughly the same number. The richer Germany takes on seven times more applicants than the poorer Romania whereas it should take on 20 times more. Overall, the principle of equal proportional sacrifice would substantially ease the obligations of the poorer Eastern and Southern members. (We ignore Cyprus because it is unclear whether the population and GDP data of the Commission and of the World Bank include the Northern part of Cyprus.)
We grant that the capacity to contribute to the collective effort of receiving applicants for protection is not only determined by riches. There is a range of factors that are relevant. For instance, a state’s capacity to take in applicants for refugee protection is greater not just when it is richer, but also when it (i) has unmet labour needs, (ii) has an ageing population, and (iii) has existing reception facilities and know-how so that start-up costs are minimal. If we construct a composite index that takes into account all these factors with proper weights assigned to them, then that will tend to take even more weight off the shoulders of poorer states.
Furthermore, there are economies of scale in refugee protection—receiving n times more applicants is less than n times more costly. And one might argue that fairness requires that the rich take on more than a proportional share, as in progressive taxation. Both considerations would prompt us to put in place a progressive distribution key. And this will once again shift the load from poorer states to richer states.
Hence, adding more complexity should make quotas even more favourable than the quotas we proposed. The distribution key needs serious rethinking and quotas need to be recalculated. Some opposing states will still object to any quotas in principle but they may become more willing to cooperate when they are being presented with a well-motivated distribution key yielding fair quotas.
Note: This article gives the views of the authors, and not the position of EUROPP – European Politics and Policy, nor of the London School of Economics. Featured image: Robert Fico, Prime Minister of Slovakia, Credit: European Council (CC-BY-SA-3.0)
Shortened URL for this post: http://bit.ly/1j2X97L
Luc Bovens – LSE
Luc Bovens is a Professor in LSE’s Department of Philosophy, Logic and Scientific Method, and Coordinator of the MSc Programme in Philosophy and Public Policy. He is currently a Laurance S. Rockefeller Fellow at the Center for Human Values at Princeton University.
Anna Bartsch – LSE
Anna Bartsch is currently a BSc. Philosophy and Economics undergraduate and works as a research assistant at the LSE. Her interests include public policy and applied economics.