One of the key elements of the EU’s model for sharing the burden of the migration crisis is a so called ‘distribution key’ which calculates the proportion of refugees each EU state should accommodate. However, as Philip Grech writes, the precise construction of this formula has been the subject of criticism, with the inclusion of factors such as unemployment figures skewing the distribution in some cases. He states that while the distribution key has been reformed from its original configuration, there is still some way to go before it can be considered a genuinely fair way to distribute refugees.
While asylum application numbers in the EU are currently slightly lower than during the peak of October last year, the refugee crisis is far from over. The Syrian Civil War keeps raging, the continuation of the EU-Turkey deal is uncertain, and EU member states are still divided on the degree of solidarity that should be shown in addressing the matter.
One of the central tools that has been put forward as a way for sharing the responsibility for the crisis is a so called ‘distribution key’ that assigns each EU member state a share of the refugees entering the EU, based on objective, measurable criteria. The EU used a distribution key in September 2015 for the relocation of 160,000 asylum seekers from Italy and Greece to the other EU member states – a process which is far from complete. The same key has also been suggested for a permanent relocation mechanism across the EU/EFTA. Unfortunately, a closer look at its construction reveals a serious logical flaw.
Flaws in the distribution key
The distribution key is based on the following four factors: a country’s population size (which is given a weight of 40% in the calculation); a country’s GDP (40% weight); the per capita average number of asylum applications received by a country over the previous five years (10% weight); and a country’s level of unemployment (10% weight). Using these four elements, each country can subsequently be assigned a given proportion of those refugees entering EU territory. The process for doing this involves computing four separate intermediate distribution keys (one for each factor) from which a weighted average is calculated – though in practice the actual procedure is a little more complicated than this as a cap on the effect of asylum applications and unemployment is introduced additionally.
The intention of including the last two factors (the asylum applications received by a country and unemployment rates) is to make the key fairer by reducing the burden on those states that have existing unemployment problems or presently receive large number of applications due to their geographic position. In an accompanying press release upon presenting this formula, the European Commission stated that ‘the corrective factors for the average numbers of asylum applications and unemployment rates are applied inversely, meaning that high existing asylum application numbers and a high unemployment rate would result in fewer individuals being relocated to a member state.’
However, this is not what happens in reality because the corrective factors do not have the intended effect. A comparison with how the distribution key would assign refugees without these two ‘corrective’ factors (i.e. if it were simply based on a country’s population size and its GDP) reveals a paradoxical effect, as illustrated in Table 1 below. While Cyprus has higher unemployment rates and more past applications per capita than Germany, its refugee share according to the EU distribution key is in fact higher than if the corrective factors were not taken into account. At the same time, the very opposite is true for Germany: the incorporation of the corrective factors actually reduces its refugee share, despite being better off than Cyprus regarding both past asylum applications and unemployment. In other words, the corrective factors induce a correction – but in the wrong direction.
Table: Comparison of Germany and Cyprus assuming refugees are distributed among all 32 EU/EFTA countries (click to enlarge)
Note: Data taken from Eurostat. For more information, see the author’s accompanying journal article.
Notably, the reasons for these distorting shifts are not political but purely mathematical and have to do with the particular way the EU distribution key is constructed. The short explanation is that the intermediate keys for the above two ‘corrective factors’ are simply not meaningful.
To illustrate this, imagine there are only two hypothetical countries, Country A and Country B, with unemployment rates of 9% and 11%, respectively, among which all refugees are being distributed. Based on that information alone, which of the two countries should take on a higher share? The intermediate key for unemployment would assign Country A 55% of refugees in this scenario, with Country B taking on the other 45%.
So far so good, but what if Country A has 49 million inhabitants, while Country B has only one million inhabitants? It would clearly be unfair if this 55-45 split was used in this scenario. If the distribution were purely based on these shares of population, then Country A would receive 98% of the refugees, with only 2% of refugees going to Country B. Leaving GDP and applications per capita to one side, if we used a similar weighting to the EU’s distribution key (20% unemployment rate and 80% population), this would leave Country A with 89.4% of refugees, and Country B 10.6%. Thus, despite the fact that Country A has a lower unemployment rate than Country B, it actually benefits from its incorporation (89.4% with vs 98% without unemployment). The converse conclusion holds for Country B.
And with respect to the actual EU distribution key, countries in the EU indeed vary much less with respect to unemployment, and asylum applications for that matter, than with respect to population size or GDP – and hence so do the corresponding intermediate distribution keys. The reason for the difference in variation is not accidental: While population and GDP are both measures of the size of a country, the two ‘corrective factors’ are not. Therefore, the latter cannot be used to determine a refugee share in absolute terms, e.g. in the form of the above intermediate keys. Rather, they can be used to modify an already given share in relative terms.
The European Commission has since, in a new proposal to address the refugee crisis, reverted to the uncorrected distribution key which is based on population size and GDP only. While it is welcome that this avoids the described paradoxical effects, a more sophisticated, well-grounded, distribution key would be desirable: the incorporation of unemployment effects and past asylum efforts per se is a good idea. There are several ways in which this could be done, as proposed, for instance, here and here. A distribution key along those lines would have greater credibility and would thus be better suited for the difficult upcoming negotiations within the EU in regards to the refugee crisis.
For more on this topic, see the author’s accompanying journal article in European Union Politics
Note: This article gives the views of the author, and not the position of EUROPP – European Politics and Policy, nor of the London School of Economics. Featured image credit: mammovies (CC-BY-SA-2.0)
Shortened URL for this post: http://bit.ly/2bsA0x3
Philip Grech – ETH Zurich
Philip Grech is Chair of Negotiation and Conflict Management at ETH Zurich.