Human migration has become something far beyond what the authors of the United Nations 1951 Refugee Convention could have ever imagined. In 2018, twenty people are forcibly displaced from their homes every minute of every day (UNHCR). There are more refugees in the world than ever before.
The problem is not unsolvable. Data experts have developed migration algorithms to help streamline the process of resettling displaced people into new communities. But as is often the case with big data, there are upsides and downsides.
Refugee resettlement isn’t just about giving migrants a roof over their heads. People who’ve been forcibly displaced from their homes need to rebuild their lives. This can include learning a new language, adapting to culture shock, obtaining an education, or finding a new career. Stanford University’s Immigration Policy Lab has developed a migration algorithm to help determine where refugees will be most likely to succeed.
How does a migration algorithm work?
Well, in this case, data scientists collect data on both refugee characteristics and the conditions in potential host communities. The reality is that there is no one perfect place for everyone to live. Different environments will suit people with different skills, talents, and traits. IPL’s migration algorithm tracks how well resettled refugees do in their new homes and matches refugees with similar characteristics to locations where people like them have succeeded. Displaced people can rebuild their lives, and the host communities get new members who contribute things the area previously lacked. It’s win-win.
Algorithmic assignment holds the potential to simultaneously improve outcomes for refugees and the communities in which they are resettled.
~Jeremy Ferwerda, assistant professor of government at Dartmouth College
The needs of refugees — and the communities that host them — aren’t static. And we need to be careful about how we determine what success looks like. One of the primary measures used to measure how displaced people are doing in their new communities is how quickly they find work. Seems reasonable, right? Gainful employment is a big part of building a life.
But there are complicating factors we need to consider:
- Employment odds — and labour supplies — are not static. Migration algorithms must keep up with shifting community needs and trends. Otherwise, some areas could become oversaturated with some skill-sets, backgrounds, etc.
- Finding a job isn’t the same as finding the right job. If we measure success with job placement, are we just driving refugees into dead-end or minimum wage employment? It’s important to regularly review what success looks like.
- Refugees are still people, not numbers. It’s important to steer clear of the idea that only those capable of finding a career in their new home are worthy of assistance.
Many of us are excited to apply the power of migration algorithms to help our fellow man. But altruism isn’t the only way data is used for immigration programs.
Early in 2017, the government of New Zealand launched a pilot program aimed at using visa application data to determine which immigrants should be deported. Reports indicate the country is using data that includes age, gender, and ethnicity to identify those deemed likely to cost taxpayers or commit crimes.
“It looks at people who place the greatest burden on the health system, people who place the greatest burden on the criminal justice system, and uses that data to prioritise those people (for deportation).”
Officials have defended the program, noting that using migration algorithms ensures the decision-making process is less biased than if humans were selecting those for deportation. But we already know that algorithms, wielded irresponsibly, can perpetuate racism, introduce bias, and make life harder for the very people who most need our support. It’s a simple matter of bias in, bias out. The data used for this program is only as neutral as the people who recorded it in the first place.
Want to Know More?
At Datassist, working to help marginalized populations — especially refugees and displaced populations — with data is a huge part of what we do. Want to know more? Do you need help using data to help people displaced by conflict, politics, or natural disaster? Get in touch with our team now.