Tallinn University of Technology

Relocating refugees through algorithms, assures employment, but potentially creates the socially segregated spaces. A recent research published by scientists from Tallinn University of Technology and University of Tartu highlights the shifts in algorithmic governance, where the positions of data subjects and social context are to be considered.

As large populations continue to migrate, authorities are looking for new ways of governing this mobility often utilizing innovative technological approaches. The use of such technologies may have its pros and cons.

Professor Anu Masso of TalTech and Tayfun Kasapoglu, a Ph.D. candidate from the University of Tartu recently published an article ‘Understanding Power Positions in a New Digital Landscape: Perceptions of Syrian Refugees and Data Experts on Relocation Algorithm.’ The article explores the perspectives of Syrian refugees and data experts on the use of a relocation algorithm that matches refugees with cities based on their skills and labor needs. 

While the use of algorithms becomes more commonplace in every field of life, the potential social outcomes of these developments may be ignored at times. Especially social groups that are more likely to be direct subjects of such practices seem to not have a voice in these discussions according to the article. The recently published article examined the perspectives of refugees who can be potentially affected by a relocation algorithm and also the data experts who are making important decisions based on/with the help of algorithms that may affect other people’s lives in significant ways.   

Based on interviews conducted with data experts and refugees in Estonia and Turkey, the article highlights that both refugees and data experts are aware of the potential benefits of using a relocation algorithm, especially for employment. However, refugees point out cultural concerns such as religious dietary restrictions, places of worship and social features like personal networks which may be difficult for algorithms to consider in relocation decisions. Data experts, on the other hand, focus on technical aspects of the issue such as transparency, accountability, and ethics. Data experts also know that an unfair decision could diminish the trust in state institutions. The main point both refugees and data experts agree is that in relocation decisions, refugees’ concerns and wishes about where to be relocated should be prioritized.

According to the authors, the main contribution of the study is the awareness that in order to come up with fair algorithmic solutions, the positions of data experts who develop algorithms, the positions of data subjects who are directly affected by algorithmic decisions and the algorithms themselves have to be understood very well. The algorithmic governance entails understanding and considering this triple relation – (un)intended power of algorithms, the experts designing these solutions, and refugees. In this triple relation, the power of algorithms should be considered an extension of state power and its migration policy inscribed into algorithms.

In the conclusion, the article suggests that rather than experimenting with algorithms on very sensitive issues and creating further layers of inequality and lack of transparency, the authorities should consider human diversity expressed in the individual’s life while making important decisions about people such as where they should be living. 

Currently, Professor Anu Masso and Tayfun Kasapoglu and their research group are carrying out further studies to explore the social power of algorithms for forced migratory groups and the potential effect of different social contexts with a special focus on Estonia and Turkey.