This week I took a course on migration with Hillel Rapoport from the Paris School of Economics. Lots of interesting stuff there: For example, recent empirical research suggests that allowing highly educated persons to emigrate may actually lead to a net increase in human capital in the home country (“brain gain”). The reason: the possibility to emigrate and earn higher wages abroad increases incentives for everyone to get more education and this could outweigh the direct brain drain effect.
Less surprisingly, we also talked quite a bit about the ongoing refugee crisis and an economist’s solution to it, proposed in recent work by Fernandez-Huertas Moraga and Rapoport. Let the EU set quotas for member states according to some initial key (capacity index), let the countries trade quotas on a market similar to the EU’s emission trading system, then use matching theory to allocate refugees to countries until the quotas are filled.
So how would this work in theory? Suppose 1 million refugees need to be allocated to 2 countries, A and B. According to the initial key, country A has a quota of 600,000, country B a quota of 400,000 refugees – say, because A is 50 percent larger than B (in terms of population or GDP). Costs of hosting refugees include both expenses for housing, food and clothing as well as costs arising from potential immigration-related social conflicts. The marginal cost of hosting refugees is increasing: the more refugees you already have, the higher the additional cost of hosting one more. For some reason, at the initial allocation, country A has higher marginal costs of hosting refugees than B, perhaps because popular opinion in country A is more hostile towards foreigners. For concreteness, let initial marginal costs in A and B be 15,000 and 11,000 euros, respectively. In that case, A’s government will be happy to pay up to 15,000 per refugee for the right to transfer refugees to country B. B’s government is happy to accept more refugees for any price above 11,000 euros. Given a competitive market (or a good auctioning mechanism), refugees will be transferred from A to B until marginal costs of hosting refugees are equal across countries ensuring an efficient allocation of asylum seekers. In equilibrium, the price will be somewhere between 11,000 and 15,000 euros per refugee, and country A will host fewer than 600,000, country B more than 400,000 refugees. (Readers familiar with the Coase theorem will notice that the initial allocation of refugees is of no relevance for this result.)
Alright, I can already hear the moral outrage: How dare you put a price tag on refugees?! Well, whether you like it or not, we are already putting a price tag on refugees. 6,000 euros – that’s the compensation the EU Commission is now offering member countries for each relocated asylum seeker. What’s more, under any quota scheme, there have to be some penalties in place for countries that don’t fulfill their quota. And in order for the plan to be incentive compatible, the penalty per refugee not hosted must be higher than the marginal cost of hosting refugees. So at this point, we are merely talking about what the correct price tag (or, equivalently, the correct penalty for not fulfilling the quota) is. There is simply no way to find this out without some kind of market mechanism that elicits the true refugee-hosting costs, as perceived by the member countries’ governments.
There are, of course, problems with that proposal. First and foremost, though the initial distribution of quotas doesn’t matter for the final allocation of refugees across countries, it does matter for who receives how much money in the end. In our example above, the larger the initial quota of country A, the more it will have to pay to country B in equilibrium. This creates a very tricky bargaining situation as each government tries to get as small an initial quota as possible. Judging by the recent experience of intra-EU negotiations, this problem alone might kill the whole thing. But then again, if the EU wants to have any kind of quota system, it has to solve that problem somehow.
Once that problem is solved (I’m assuming the can-opener here), it should be a piece of cake to get the auctioning mechanism going. Also, the algorithm to match refugees to countries is no more complex than the ones used by your average online dating platform.
So, the plan is well within the realm of the technically feasible. It makes great economic sense. Alas, it’s probably a non-starter politically.
PS: Here’s Hillel Rapoports website for additional info.