The refugee crisis — match us if you can
‘However many refugees we decide to resettle, there’s no excuse for doing the process wastefully’
Writing in the 1930s, Lionel Robbins, head of LSE’s economics department, defined economics as “the science which studies human behaviour as a relationship between ends and scarce means which have alternative uses”. It’s the study of who gets what and why.
That typically means that economists study conventional markets: how prices work, how people respond to them and how the whole system might function or malfunction. But sometimes a market simply will not do. We don’t allocate children to state school places based on their parents’ willingness to pay. Most countries don’t sell passports to the highest bidder. We do not have a legal market in iced kidneys.
Whether we like it or not, the problem of who gets what and why remains. Sometimes it is grubbily resolved by the emergence of parallel markets — for example, children can be placed in desirable schools at taxpayer expense if their parents buy or rent expensive homes in the right areas.
Over the past few decades a small group of economists — most notably Nobel laureate Alvin Roth, author of Who Gets What — and Why (2015) — has been designing “matching mechanisms” to address allocation problems without resorting to traditional markets. A typical problem: matching teaching hospitals with trainee doctors. The doctors want good hospitals and the hospitals want good doctors. Each side will also have a focus on a particular field of medicine, and the doctors may have preferences over location. Some doctors may be dating fellow medics, who are themselves searching for a teaching hospital.
A good matching mechanism tries to satisfy as many of these preferences as possible. And it ends the need for people to second-guess the system. Bad matching mechanisms reward people who say that a compromise option is really their top preference. Such mind-games are alienating and unfair; in a well-designed matching system, they can be eliminated.
Roth and a growing number of his students and colleagues have designed matching mechanisms for schools and hospital placements, and even mechanisms to ensure the best match for donated kidneys. In each case a market is socially unacceptable but ad hoc or lottery-based allocations are also poor solutions. Nobody wants a random kidney, or to be assigned a place on the whim of a well-meaning bureaucrat who doesn’t really understand the situation.
By balancing competing demands, good matching mechanisms have alleviated real suffering in school systems and organ donation programmes. Now two young Oxford academics, Will Jones of the Refugee Studies Centre and Alexander Teytelboym of the Institute for New Economic Thinking, are trying to persuade governments to use matching mechanisms in the refugee crisis.
Most popular discussions of the crisis focus on how many refugees we in rich countries should accept. Yet other questions matter too. Once nations, or groups of countries, have decided to resettle a certain number of refugees from temporary camps, to which country should they go? Or within a country, to which area?
Different answers have been tried over the years, from randomly dispersing refugees to using the best guesses of officials, as they juggle the preferences of local communities with what they imagine the refugees might want.
In fact, this is a classic matching problem. Different areas have different capabilities. Some have housing but few school places; others have school places but few jobs; still others have an established community of refugees from a particular region. And refugee families have their own skills, needs and desires.
This is not so different a problem from allocating trainee doctors to teaching hospitals, or children to schools, or even kidneys to compatible recipients. In each case, we can get a better match through a matching mechanism. However many refugees we decide to resettle, there’s no excuse for doing the process wastefully.
There is no perfect mechanism for matching refugees to communities — there are too many variables at play — but there are some clear parameters: housing is a major constraint, as is the availability of medical care. Simple systems exist, or could be developed, that should make the process more efficient, stable and dignified.
One possibility is a mechanism called “top trading cycles”. This method invites each refugee family to point to their preferred local authority, while each local authority has its own waiting list based on refugee vulnerability. The trading cycles mechanism then looks for opportunities to allocate each family to their preferred location. The simplest case is that, for example, the family at the top of the Hackney waiting list wants to go to Hackney. But if the family at the top of Hackney’s list wants to go to Camden, the family at the top of Camden’s list wants to go to Edinburgh, and the family at the top of Edinburgh’s list wants to go to Hackney, all three families will get their wish.
Right now, the UK is a promising candidate to pioneer the use of one of these matching mechanisms to place refugees. The government has pledged to resettle 20,000 Syrian refugees now in temporary camps. Local authorities have volunteered to play their part. But to make the best possible matches between the needs of the refugees and the capabilities of these local authorities, it’s time to deploy a little economics.
Written for and first published at ft.com.