Monitoring patterns of migration from Venezuela to Colombia using digital trace data

Various studies have proven digital trace data to be a reliable information source to measure short-term displacement, but less evidence exists on the potential for longer term monitoring of forced migrants. As one of the world’s largest forcibly displaced populations, this paper explores whether data derived from Facebook can help identify trends in the sociodemographic composition and mobility patterns of Venezuelan migrants living in Colombia, over the years 2018-2024.

This study uses audience size counts collected repeatedly from the Facebook Advertising Platform between 2018 and 2024 to measure the age, sex and education of Facebook users who originate from Venezuela and live in Colombia, following methods previously developed by Zagheni et al. (2017) and Palotti et al. (2020). These trends are compared to a reference group of Facebook users who live in Colombia but have not previously lived in Venezuela.

Using a beta regression, the study finds significant shifts among Venezuela-originating Facebook users, characterized by i) a greater share of female users, ii) a greater share of users aged 41-64 years and 65+ years, and iii) a falling share of university-educated users. These findings align with secondary sources and suggest that the migrant selection bias seen in the initial years of the crisis has gradually eroded.

The analysis detected a temporary reduction in Venezuelan-originating Facebook users in Colombia during 2022, which may be related to a period of out-migration of Venezuelans leaving Colombia following a lessening of COVID-19 restrictions. Such findings highlight the potential of the Facebook Advertising Platform for long-term monitoring of migrant populations, particularly as a complement to traditional methods.