Monitoring migrations is not an easy task. While in today’s economy, survey data about economic confidence or public opinion are collected on a daily basis, that is not the case for migration statistics, which come from Censuses, population registers, and, occasionally, ad-hoc surveys–and are often outdated and inconsistent across countries. Problem is, this lack of information does not allow policy makers to build up timely and consistent strategies to manage migratory flows.
That’s why researchers are turning to social networks, using the zillion of data produced by users as a tool to monitor recent trends in migration. In the last few years, there have been quite a lot of studies focusing on this methodology. One of the latest is contained in a paper called “Inferring International and Internal Migration Patterns from Twitter Data“, published earlier this year by scholars from NYC’s Queen College, Doha’s Qatar Computing Research Institute and Stanford University. In it, scientists tried to develop an indicator of geographic mobility in OECD countries, based on publicly available geolocated tweets.
Downloading tweets coming from 500,000 Twitter users, mapping them to countries (until they obtained geolocated tweets for about 3,000 users in each country considered), and then considering the place from where they were posted as a signal of the users’ location at a certain moment in time, after having refined the sample with a bit of statistical tricks (I’m over simplifying a lot here, of course, you may want to read the complete paper for a detailed explanation), they were able to estimate movements in and out of the countries considered.