As the world faces its largest crisis of displaced people since World War II, a new algorithm could help countries resettle refugees in a way that boosts their employment success and overall integration.
Researchers used a machine-learning algorithm to analyze historical data on refugee resettlement in the United States and Switzerland. They found that the refugees’ eventual economic self-sufficiency depended on a combination of their individual characteristics, such as education level and knowledge of English, and where they were resettled within the country. It turned out that refugees with particular backgrounds or skills achieved better outcomes in some locations than others.
In 2016 alone, about 65.6 million people were forced to flee their homes.
The algorithm assigned placements for refugees that they project would increase their chances of finding employment by roughly 40 to 70 percent compared with how the refugees actually fared, according to the new study in Science.
“As one looks at the refugee crisis globally, it’s clear that it’s not going away any time soon and that we need research-based policies to navigate through it,” says Jeremy Weinstein, a professor of political science at Stanford University and a coauthor of the study. “Our hope is to generate a policy conversation about the processes governing the resettlement of refugees, not just on the national level in the United States but internationally as well.”
The group, from Stanford and ETH Zurich, says the algorithm, which could be implemented at virtually no cost, could help resource-constrained governments and resettlement agencies find the best places for refugees to relocate.
Where to go?
In recent years, a record number of people have been displaced as a result of war, persecution, and other human rights violations, surpassing the numbers seen after World War II. In 2016 alone, about 65.6 million people were forced to flee their homes, according to the United Nations’ refugee agency.
Often, countries that resettle refugees in their communities do so either somewhat randomly or according to local capacity of hosting communities at the time of refugees’ arrival. In the United States, refugees who have family members at a particular location are directed to join them there. But refugees without preexisting ties are free to be sent to various locations, and current approaches do not match them to locations where the evidence suggests it would be easiest for them to integrate.
“Our motivation was to bring the best of cutting-edge social science to an area of high policy priority that needs innovation but, because of the limited resources and challenges of navigating large numbers, has not been able to innovate from within,” Weinstein says.
The group developed their algorithm based on socioeconomic data from more than 30,000 refugees, aged 18 to 64, placed by a major resettlement agency from 2011 to 2016 in the United States. The data also included where those refugees were resettled, and their eventual employment status.
Based on this data, the team had the algorithm predict employment probability and optimal locations for a group of refugees who arrived toward the end of 2016 and compared those predictions with how these refugees actually fared in their new homes.
The group found that if the algorithm had selected locations for refugees’ resettlement, the average employment rate among those refugees would have been roughly 41 percent higher.
Jobs in Switzerland
The team went through the same process with data from asylum seekers who had resettled in Switzerland between 1999 and 2013. They predicted the employment rate would have been 73 percent higher among asylum seekers who arrived in 2013 if they had been assigned to the places the algorithm identified as optimal.
In the cantons of Vaud and Zurich, for example, young men from Iraq have good employment prospects. In Vaud, employment prospects are higher for anyone who speaks French. Women from Sri Lanka also have relatively good employment prospects in both cantons; both cantons have comparatively large regional networks from Sri Lanka.
“In a next step, the algorithm could be put into practice; it would be fairly straightforward to integrate it into the existing allocation process,” explains Dominik Hangartner, professor of public policy at ETH Zurich and leader of the Swiss part of the study. He adds that it could also be updated at any time if the composition of refugees and labor market were to change.
“Our goal is to use data and social sciences to propose solutions that improve the asylum process and refugees’ labor market integration. After all, having more refugees employed can also ease the financial burden on the Swiss government, cantons, and municipalities.”
“The employment gains that we’re projecting are quite substantial, and these are gains that could be achieved with almost no additional cost to the governments or resettlement agencies,” adds Kirk Bansak, a lead author of the study and a political science PhD student at Stanford. “By improving an existing process using existing data, our algorithm avoids many of the financial and administrative hurdles that can often impede other policy innovations.”
A simple change
The researchers are not advocating for the algorithm to replace the decision-making of resettlement officials.
“Our approach preserves the ability of policy-makers to set their own parameters and priorities,” the researchers write. “For instance, in a computer-assisted assignment process, the algorithm might provide several recommendations, and placement officers could use their own discretion to determine the final assignment or override any suggestions.”
Yet in contrast to more expensive policy interventions, such as job or language training for refugees, the results of the algorithm, the code of which is available for free to any organization or government, are promising, the researchers say.
“The fact that we are able to generate such significant gains because of a simple change to the resettlement process is a demonstration of just how important it is to bring data-driven insights to policy-making processes,” Weinstein says.
The group says they still need to confirm the algorithm’s predictions through prospective tests that implement this approach in real time. The research team is now developing a number of pilot programs in partnership with governments and resettlement agencies to test the algorithm’s power.