JMIR Publications

ABSTRACT

Background:

The Ukrainian refugee crisis lacks reliable data about refugee flows, demographic structure, and integration. But those data are necessary for the UNHCR and governments in preparing high-quality projections for emergencies and the conditions for the integration of refugees who intend to stay in immigration societies. Although Facebook, Instagram and YouTube are social platforms with the most users, very few studies have been written about their potential for migration studies and integration.

Objective:

The objective was to test the usefulness of Big Data insights from social network platforms to gain first demographic insights into refugees’ age and gender structure, migration flows, and integration trends.

Methods:

The primary methodological concept of our approach is to monitor the so-called “digital trace” of refugees left on social networks Facebook, Instagram and YouTube and their geo-locations. We focus on users that use social network platforms in Ukrainian and Russian language. We sampled the data before and during the war outbreak, standardised the data and compared it with the first official data from UNHCR and national governments. We selected specific keywords, i.e. migration and integration-related queries, using YouTube insight tools. Using FB and Instagram, we collected our own data archive because Meta offers data only for the present day with the ability to compare this day with the average of the past 12 months. In the next step, we collected signals that indicate integration willingness.

Results:

Our approach shows that the number of Facebook and Instagram users is increasing rapidly in Ukrainian neighbouring countries and Germany after the war outbreak in Ukraine. Testing performed matches the trend of immigration of Ukrainian refugees in Poland and Germany, as well in the cities and the German Bundesländer. The tested correlation between the number of Ukrainian refugees in Poland and FB and Instagram users in Ukrainian in Poland shows that the increase in frequency index is correlated with stepped-up emigration from Ukraine. R2 is 0.1324 and shows a positive correlation, and a p-value is statistically significant. The analysis of the FB group of Ukrainian in the EU shows that those groups can be a valuable source for studying integration. Ukrainians are increasingly expressing interest in learning the German language, which indicates integration willingness. One of the contributions of the second used method, YouTube insights, is that it shows that by searching for video material on the YouTube platform, the intention of users to migrate, or in this case, to flee from Ukraine, can be estimated.

Conclusions:

The usefulness and the main advantage of this approach are enabling first insights into integration willingness and identification of trends in the movement and intentions of refugees when there is no official data. On the one side, this method allows governments to estimate how many refugees intend to enter the labour market and integrate into the immigration society and, on the other, better respond to the recent humanitarian crisis. Despite its beneficiaries, this approach has many limitations, and there is a need for many more studies to perfect this method.