August 16, 2022 | News | Interview
“Using Computational Methods Is Like Wearing 3D Glasses at the Movies”
MPIDR Director Emilio Zagheni and his colleague Emanuel Deutschmann explain in an interview why computational approaches are like a “fresh gust of wind” everyone in the field of migration research should leverage. Their recently published Special Issue “Computational Approaches to Migration and Integration Research: Promises and Challenges” collects seven examples of notable research in this area.
This Special Issue was edited by Emanuel Deutschmann (Europa-Universität Flensburg), Lucas G. Drouhot (Utrecht University), Carolina V. Zuccotti (University Carlos III Madrid), and Emilio Zagheni (Max Planck Institute for Demographic Research) and published in the Journal of Ethnic and Migration Studies.
Professor Zagheni, how are computing and the digital revolution affecting migration and integration research?
Zagheni: If I may, let me address this question with a metaphor. Every summer, around mid-August, there is a big sailing event that takes place in Rostock. It’s called Hanse Sail. If you were at the MPIDR, which is located on the bank of the Warnow river, you would see all sorts of sailboats: long ones, short ones, historical ones, and modern ones. For a week, they sail all the way to the Baltic Sea and back, offering a wonderful display of sailing skills. We can think of the sailboats as sub-areas of migration research, each of them with their own unique characteristics. Computing and the digital revolution can be thought of as a fresh gust of wind that powers all the sailboats, and that in turn accelerates discovery. In order to remain at the forefront of migration research, across basically all sub-fields, it is essential to learn how to leverage this wind and to transform its power into sailing speed. In this Special Issue, that is what interdisciplinary teams of researchers do.
You employed a powerful image to underline the importance of computational approaches. When using these tools, what are the concrete advantages?
Zagheni: The introduction of the Special Issue provides a wide-ranging discussion of the advantages of incorporating computational methods and digital traces into mainstream migration research. I would encourage curious minds to read the paper for a full description.
But I’ll emphasize one key advantage: in many circumstances, there are simply no viable alternatives to address pressing migration questions in a timely manner, at the appropriate level of spatial granularity, and for the purpose of advancing migration theory. Adding computationally-intensive methods and digital data to the toolbox of migration scholars enables researchers to make previously not addressable questions tractable.
Professor Deutschmann, as always there must be some hurdles, too. What are the disadvantages?
Deutschmann: Computational researchers do indeed face a number of challenges. Sometimes, important information that is relatively easy to aquire in a traditional survey or interview, such as citizenship or immigrant generation, is hard to establish based on computational methods and digital trace data. Problems also arise from the fact that a lot of digital data is gathered by platforms like Facebook, Twitter, and Google. Their users are not necessarily representative of the population as a whole, which often makes it difficult to draw generalizable conclusions based on such data. Also, these private companies decide autonomously how much data they make available and with whom they share it. This concentration of power is quite problematic.
Ethical concerns need to be taken very seriously when working with such digital trace data as well…
Deutschmann: Yes. Thorough anonymization, for example, is pivotal as people may not have always considered or actively consented to such secondary examination of their online activities for research purposes. One goal of the Special Issue is to reflect on these challenges and to encourage researchers to weigh the strengths and weaknesses of computational approaches critically.
Why should computational approaches be used in the future?
Zagheni: Computational approaches and digital data are the means to the goal of producing scientific analyses for the benefit of migrants and of host and sending countries and regions. Using computational methods is like wearing 3D glasses at the movies: they add a dimension to amplify and expand our view.
Which specific approach looks most promising to you?
Zagheni: One area where computational methods and digital data have been highly relevant and will continue to be is the measurement of migration phenomena, from migration flows, to migration intentions, to mobility at different levels of geographic and temporal granularity. It has been recognized that traditional data are deficient in this area. Here, adding a data dimension within a solid statistical framework is key.
Is there another?
Zagheni: Yes. A second area is the analysis of integration and segregation. In addition to leveraging new data sources on migration, this area is going to benefit significantly from emerging opportunities in the context of running survey experiments and combining causal reasoning with the opportunities offered by the infrastructure of the digital era.
Let’s talk about “Computational Approaches to Migration and Integration Research: Promises and Challenges” a bit more in depth. What topics does it include?
Deutschmann: The Special Issue covers a number of topics and questions that are central to migration research: Why do people emigrate, where do they migrate to, and what migration patterns emerge between countries globally as a result? They also examine whether refugee movements can be predicted, how host communities respond to the influx of refugees, how people interpret, frame, and discuss the arrivals of refugees, and how spatial segregation patterns of ethnic minorities emerge. Furthermore, one paper examines whether terrorist attacks lead to increased online insults against ethnic minorities. Another one looks at whether living nearby refugee centers correlates with votes for a far-right party.
Why are they important?
Deutschmann: The topics cover some of the central challenges societies around the world are facing today. For example, one article by a team led by Mario D. Molina (New York University Abu Dhabi) examines the link between weather-related factors like the maximum temperature at night and migration decisions in Mexico. Given the rapidly worsening climate crisis, understanding such links is an important issue. Tensions between ethnic groups and violence and discrimination against immigrants and refugees are central problems of our time and are addressed in several articles in the Special Issue, using novel computational methods.
You think the studies can also be important as myth-busters…
Deutschmann: Yes, for example, the above-mentioned study that examined whether living nearby to refugee centers correlates with votes for far-right parties showed that the correlation is actually negative: the more contact with refugees, the lower the vote share of the far-right party.
What might be missing in the collection in terms of approaches or research questions?
Deutschmann: Of course, one Special Issue with just seven empirical articles cannot cover every aspect of migration research, nor all existing computational approaches. For example, it doesn’t include a paper using digitalized historical records, which would be another fascinating source. Smartphone “sensing”—the measurement of spatial moves and social activities produced via smartphone usage—is another blind spot. We hope that this Special Issue will inspire further research in these directions.
Zagheni: Let me add that computational scientists are also well positioned to assess the implications of changing technology on migration behavior. For example, how does the increasing use of information and communication technologies affect migration aspirations, migration outcomes, and the well-being of migrants? These topics are beyond the scope of the Special Issue, but we expect that they will increasingly gain importance in years to come.
Drouhot, L., Deutschmann, E., Zuccotti, C.V., Zagheni, E.: Computational Approaches to Migration and Integration Research: Journal of Ethnic and Migration Studies, Special Issue (2022) DOI: 10.1080/1369183X.2022.2100542