Monday, 2 May 2016

The 'efficient' allocation of refugees

Some time back I promised one of my students I would write about the refugee crisis in Europe. So here goes: If we had free movement of people, then refugees would simply move to their preferred location (which may or may not be a Western country). From an overall global welfare perspective, this should be the preferred solution (if you want an explanation why, Michael Clement argues persuasively that there are trillion dollar bills being left on the sidewalk as a result of restrictive immigration policies in Western countries).

However, there isn't free movement of people, which means that from an economic perspective one of the interesting aspects of the crisis is how 'best' to allocate refugees between countries. Thinking about European countries that are facing the brunt of the wave of refugees, the current solutions are clearly not working. Open Europe has good coverage of the problems here. In short though, the 'Dublin regulation system', whereby refugees apply in the country where they first arrive and are returned there if they move elsewhere, has failed with peripheral European countries like Greece simply shepherding migrants through to the next country in.

An alternative solution was developed in the form of a €3bn deal with Turkey, whereby migrants are returned to Turkey, only appears to cover migrants in the thousands (compared to the 1.25 million refugees who entered Europe last year). This was essentially a Coasean bargain between the European countries and Turkey. The Coase Theorem tells us that, if private parties can bargain without cost over the allocation of resources, they can solve the problem of externalities on their own (i.e. without government intervention). In this case the private parties are the EU governments and the Turkish government. This bargaining solutions would work provided the payment to Turkey is more than enough to compensate for the cost of hosting the refugees, and provided the payment is less than the alternative cost (of dealing with the refugees) for the EU governments. Also, it would work provided the transaction costs (the costs of arranging the agreement) are low, the cost of monitoring and enforcing the terms of the agreement are low, and there are no free riders (EU countries that would benefit from the agreement, but refuse to contribute their share of the €3bn). In theory the Coasean solution would work, but its failure was already being discussed when it came into force, mainly because it still fails to address the allocation of refugees between countries.

Earlier, at the end of January, Dalibor Rohac wrote an interesting op-ed on the allocation issue in the New York Times. Rohac writes:
Europe’s current refugee crisis is often presented as a quantity problem: There are simply too many migrants for the European Union to absorb. But this situation is not without historical precedent. Europe has accepted large numbers of immigrants before. The issue this time is political. It has little to do with the absolute numbers of asylum seekers. The problem lies with the European Union’s dysfunctional asylum system, which encourages countries to pass refugees on like hot potatoes, and places the burden of registering and processing asylum seekers on a small number of countries on the Union’s border...
But there are ways out of this seemingly desperate situation.
For one, the quota system, proposed by the European Commission, could be made flexible. In 1997, the Yale University legal scholar Peter Schuck proposed a system of tradable refugee quotas. The European Union would still have to agree on the total number of migrants to whom it is willing to grant asylum, and on how they would be distributed among the member states. But the quota market would allow countries such as Slovakia or Hungary, whose leaders refuse to accept any refugees, to “bribe” others to carry their obligations on their behalf, putting a concrete price tag on the unwillingness of Central Europeans to help.
Essentially the allocation issue can be solved with some sort of quota. The quota system creates a set of 'obligations' for EU countries - the obligation to accept a given number of refugees each year. In ECON110 we talk about four criteria of an efficient property rights system, and effectively these obligations are a form of property rights (albeit, negative property rights) [*]. An efficient system should have obligations (or rights) that are: (1) universal; (2) exclusive; (3) transferable; and (4) enforceable.

Universality in this context means that all refugee flows would need to be covered by the obligations system, and all EU countries would be obligated to take refugees. The system would start to break down if there were additional flows of refugees that were not covered, or where countries were able to opt out of the system, for instance. Exclusivity in this context means that all of the costs of the refugee flows should be borne by the country that is accepting that group of refugees. This means that there can be no free riders. Transferability means that the obligations can be freely traded between countries. If the Netherlands wants to accept fewer than their quota of refugees, they might be able to trade the obligation to Sweden, presumably in exchange for something that Sweden wants. Finally, enforceability means that there must be some form of penalties (presumably from the EU) for countries that refuse to comply with their obligations. A system of obligations meeting these four criteria would be an efficient way of allocating refugees among European countries.

However, Rohac also notes an alternative solution: explicit market in refugee quotas is not the only possible fix to the current crisis, according to two researchers at the University of Oxford, Alex Teytelboym and Will Jones. To bring the chaotic influx of refugees under control, the European Union could also create a centralized “matching system,” which would involve none of the cash payments that are often seen as repugnant.
[In the matching system] [a]pplicants would rank European Union countries by order of preference and submit that ordering to a central clearinghouse.
Some countries, such as Germany or Sweden, would likely remain oversubscribed. But because applicants would be submitting a complete ordering of European Union countries they are applying for, they could still be matched with, say, their second, or third choice, instead of being rejected outright.
The European Union member states would in turn specify how many and what refugees they are willing to accept. 
The problem with the proposed matching system is that no country would be obligated to take refugees, so the system would lack enforceability, and countries could easily opt out, leaving us back where we started. So, while matching might appeal to those who are squeamish about the transferability of obligations between countries, a quota system that obeys simple criteria would work much better in practice. This is provided an initial allocation of obligations could be agreed, and of course all participating countries would have an incentive to ensure that their initial allocation was as low as possible, if only so that they could obtain concessions from other countries after the system comes into force. The joys of politics!

[Update]: Reflecting on this overnight, I wrote this post as if the matching system and a system of tradeable obligations are somewhat mutually exclusive. Of course, they are quite complementary. Once a country knows how many refugees they are obligated to accept, there needs to be some mechanism to select which refugees they take, which is where matching could contribute. Similarly, a matching system will tell which refugees are most compatible with each country, but not how many each country should take (e.g. at what 'level' of compatibility should the cut-off for acceptance be?), which is where the system of tradeable obligations becomes helpful.


[*] Please note that I am explicitly not referring to refugees as property here. I am simply linking the concept of a system of obligations to that of a system of property rights, because the efficiency of both systems relies on the same four criteria.

[HT] For the Open Europe blog, Marginal Revolution, which incidentally has been following the refugee issue over the past few months (see here).

No comments:

Post a Comment