Switzerland to test algorithmic asylum placement
Switzerland is to start using a new algorithm to distribute asylum seekers among the cantons in such a way that they can find work more easily, the State Secretariat for Migration (SEM) has told Swiss public television, SRF.
The confirmation, broadcast in the current affairs programme 10vor10External link, comes a few months after researchers from the Federal Technology Institute Zurich (ETHZ) and Stanford University revealed that their algorithm could increase employment of asylum seekers by up to 30%.
+ More on the algorithm for most efficient asylum placements
Based on big data gleaned from tens of thousands of people from past and present – focussing on gender, nationality, age, and other criteria – the algorithm calculates in which canton an asylum seeker would be most likely to find work.
It also continuously learns and updates itself according to the situation and placement success of new arrivals, as well as the constantly shifting labour markets in the various cantons.
Currently, Switzerland distributes asylum seekers randomly among its cantons, according to a quota system. However, as SEM spokesman Daniel Bach told the 10vor10 program on SRF, a pilot project is placing 1,000 asylum seekers according to the ETH/Stanford algorithm.
The aim is to see a significant increase in employment success over the next two to three years, said Bach. A control group of equal size will be distributed at random to compare success rates.
Switzerland is the first country ever to test such a data-based placement system. With other countries struggling to manage refugee influxes, it could be an attractive model to replicate. “Interest from various countries is very high,” said ETH professor Dominik Hangartner.
In compliance with the JTI standards
More: SWI swissinfo.ch certified by the Journalism Trust Initiative
Contributions under this article have been turned off. You can find an overview of ongoing debates with our journalists here. Please join us!
If you want to start a conversation about a topic raised in this article or want to report factual errors, email us at email@example.com.