Towards a fairer distribution of asylum seekers

The European Commission has worked out a distribution key, described in a press release of 9 September 2015, to spread the responsibility for receiving refugees (or, more accurately, migrants who claim to be in need of international protection from sending countries that have refugee recognition rates in the EU greater than 75 per cent—currently Syria, Eritrea and Iraq).

The key stipulates quotas that are proportional to a measure consisting of population size (40%) and GDP (40%), plus two "corrective factors": unemployment rate (10%), and the number of refugees received in the last four years (10%). In the Council decision of 22 September 2015, the key is applied to roughly 66,000 refugees who are held up in Greece and Italy to be distributed between the remaining states in the European Union, except for Denmark, Ireland and the UK, who are exempt on grounds of the Lisbon Treaty. The Commission’s quotas can be found in column J of Table 2.

The Commission brings in population size at the wrong point in the calculation. It notices that various factors matter, including population size, and it assigns weights to these factors to determine what would be a reasonable willingness to absorb refugees for each state. What it overlooks is that population size matters in a very different way from the other factors.
First we determine what would constitute a reasonable willingness to absorb refugees for a typical citizen of each state. We propose that the following three factors are relevant.

First, this willingness is a function of riches. Citizens of richer states are in a better position to carry the financial burden of the reception of refugees than citizens of poorer states. Second, it is a function of the labour market. The influx of refugees who are typically eager to work is more of a benefit for a labour market with more job vacancies and hence a state with a higher job vacancy rate should be more willing to absorb refugees. And third, it is a function of demographics. The influx of refugees who are typically younger than the local population is more of a benefit for an ageing population, and hence a state that has more of an ageing population should be more willing to absorb refugees. Riches make it easier to accept the financial costs of absorbing refugees, whereas job vacancies and an ageing population make the absorption of refugees more of a benefit (or at least less of a cost).

A reasonable measure of riches is GDP per capita at Purchasing Power Parity since the cost of receiving refugees will be affected by the cost of living in the receiving member state. See Column B of Table 1.

What is relevant about the labour market is job vacancy rates rather than the unemployment figures which the European Commission uses. What matters is whether refugees will be able to find work and job vacancies are a better measures than unemployment figures, since the latter are also affected by the nature of the welfare state. In a generous welfare state with high education levels, high unemployment figures can coexist with a high demand for manual labour. See column C of Table 1.

A simple measure of the extent to which the population is ageing is the proportion of over 65 year olds in the population of each state. See column D of Table 1.
Unlike the European Commission, we did not include the number of applicants received in the last four years in our measure. We grant that states that have opened their doors in the past should be given a break. But on the other hand, economies of scale enter in.

States who have opened their doors in the past already have institutions for reception in place whereas it is costly to set up such institutions for states who only had minimal numbers of refugees knocking on their doors. Given that these considerations pulls in both directions we decided not to include this factor. In any case, the factor only has small weight in the Commission’s calculation.

We then construct a multi-dimensional measure. For each factor, we construct an index between 0 and 1 that is a linear function of the values of the factor, so that the highest value on each factor maps onto 1 and the lowest value maps onto 0. This normalises the data. See columns B, C and D of Table 2. Subsequently a decision needs to be made about weights. In the absence of any principled argument that one variable should have more weight than another, we propose to work with equal weights, but this can easily be adjusted. We now have a measure for the reasonable absorption willingness per capita for each state. See column E of Table 2.

It is at this point that we can bring in population sizes. We list population sizes in column E of Table 1. We multiply the willingness to absorb per capita by the population size of each state yielding the willingness to absorb per state in column F of Table 2. We then spread the load of refugees proportionally to this willingness. Column G lists the percentages and column I lists the numbers for a pool of 66,089 people. Columns H and K are the percentages and the numbers on the Commission’s key.

Let us focus on the member states who are substantially affected, say whose quotas increase with more than 20 per cent (in yellow) or decrease with more than 20 per cent (in blue). To understand these comparisons, note that on our scheme, GDP per capita is slightly less important, but job vacancies and an ageing population have come in as equally important co-determining factors.

Germany increases with almost 50 per cent because of a high job vacancy rate and an ageing population. The Czech Republic increases with more than 30 per cent and Malta with more than 60 per cent for similar reasons, and in addition their low GDP per capita matters less.

Luxembourg decreases with more than 50 per cent because of low job vacancy rates, a youthful population, and its high GDP matters less. Cyprus, Poland and the Slovak Republic decrease with 55 to 65 per cent because of low job vacancy rates and youthful populations. France and Spain decrease with 20 to 25 per cent because of their low job vacancy rates.

There are various changes one could make to our key. One could take into account other factors to determine the willingness to absorb per capita. One might take out outliers in calculating the indices (e.g. Luxembourg on GDP per capita). One might want to change the weights of the factors. Or one might want to use a progressive rather than a proportional assignment of refugees relative to the willingness to absorb. We invite our readers to do so and have made our Excel file available for this purpose.