It’s easier for a Somali asylum seeker to get a residence permit in Norway and Sweden than in Denmark. On the other hand, it’s easier for an asylum seeker from Eritrea to obtain residence in Denmark than in Sweden and Norway.
The figures from 2018 show that Somali asylum seekers had an eight-percent chance of being granted asylum in Denmark, compared with a 34-percent chance in Norway and a 48-percent chance in Sweden. Conversely, 99.5 percent of asylum seekers from Eritrea were granted residence in Denmark, while the figure was 89 percent in Norway and 83 percent in Sweden.
These variations are not unique to the Nordics. In fact, the difference between other European countries is often even more pronounced, and this has led to repeated accusations describing legal decision-making as a “refugee roulette”.
Why are there such significant differences between the Nordic countries, and what factors underlie the national asylum decisions? A Nordic research team from Denmark, Sweden and Norway is working on finding answers to that. Nordic Refugee Determination: Advancing Data Science in Migration Law (NoRDASiL) | NordForsk, funded by NordForsk, is an interdisciplinary project with researchers from the fields of law, medicine and computer science.
To find the answers, the project has obtained a data goldmine:
“What’s unique about our project is that we’ve gained access to such a broad set of data from Sweden, Norway and Denmark. We’ve gained insight into more than half a million legal decisions, which can reveal patterns in connection with the granting of asylum in the Nordic countries. Among other things, we’ve obtained interview transcripts from asylum seekers and specific information about people seeking asylum. This allows us to go a step further than previous research in the area. Existing research has established differences between the countries’ asylum decisions, but it hasn’t been possible to explain the differences more generally and find causal relationships, and that’s what we’d like to do,” says Thomas Gammeltoft-Hansen, project manager and professor of law from the University of Copenhagen.
“It’s important to compare the asylum policies across the Nordic countries,” he explains.
“We have internal border controls between several of the Nordic countries. One reason for this is concerns about irregular asylum flows between the countries. A better understanding of the differences in asylum processing between the Nordic countries can be an important input into that debate. At the same time, we can help put the Nordic countries on the world map. Our data access is relatively unique, and our results have already begun to attract attention among researchers in other countries. In collaboration with those researchers, we can also investigate whether there are any special features of the Nordic asylum procedures that might be worth advocating internationally.”
The Nordic countries have different asylum systems
“When compared to other countries such as the US and Australia, the asylum systems of the Nordic countries appear to be quite similar,” says Thomas. We have a more-or-less common legal culture, which has left its mark on the national authorities of all the countries. Yet there are also significant differences:
“Although we’ve yet to analyse case data from all three countries, our initial mapping shows that there is quite a big difference in how the Nordic countries have structured their asylum systems. Some countries have committees while others have courts, the procedures are different, and the relationship with international rules such as the EU asylum policy also varies between countries,” Thomas Gammeltoft-Hansen explains.
In Denmark, the highest authority is the Danish Refugee Appeals Board, where three (previously five) people are involved in determining asylum applications. Thomas explains that having more people at the table can also be important:
“Many asylum cases are about credibility assessments. Is the applicant’s story assessed to be coherent, consistent and self-perceived? There will always be an intersubjective element in such assessments. Is it an advantage to have more people at the table, and does it matter what professional background the decision-makers have? These are some of the questions that we’re currently trying to find answers to through our data material.”
Asylum decisions are a complex process that not only concerns legislation, but also a large number of other factors. That’s why the asylum process has often appeared as a kind of “black box” for researchers, and we still don’t know very much about how different aspects play into the decision-making process.
Artificial intelligence can determine asylum cases
In the future, the processing of asylum cases will perhaps be partially replaced by computer technology and artificial intelligence (AI). We’re already seeing this in a large number of other areas of law, and it’s also something that is being looked at within the area of asylum in several countries.
It is already happening quite extensively in other parts of the field of immigration, such as in connection with border control, where biometric databases and reports are used to make risk assessments and to try to predict new refugee routes. In Norway, part-automated citizenship case processing has been similarly tested.
According to Thomas Gammeltoft-Hansen, AI is unlikely to completely take over the human assessment in asylum cases:
“We’re very careful not to take the human element out of the asylum process. There must still be an individual assessment of asylum cases, but there’s no doubt that this is an area where we can expect to see some huge developments. Both authorities and NGOs are under great pressure in relation to the ever-changing number of asylum seekers. Here, computer-based tools can help to provide extra capacity and, for example, find previous cases or draft a decision or a case submission. However, it’s absolutely crucial that we’re aware of legal requirements in this process, and that existing biases are not embedded when developing the algorithms. That’s what we hope our research can contribute to,” he concludes.
