Assessing Biases of Facebook Data to Nowcast Migrant Stocks in the United States

Esther Denecke, Max Planck Institute for Demographic Research (MPIDR)
Monica Alexander, University of Toronto
Emanuele Del Fava, Max Planck Institute for Demographic Research (MPIDR)
Emilio Zagheni, Max Planck Institute for demographic Research

Data on migration are often insufficient or released with a large delay. While the delayed release of migration statistics is a general problem, the Covid-19 pandemic further emphasizes the need for timely data to assess the impact of discontinuities on migration. New and innovative data sources, such as social media data, could help monitoring and estimating migrant stocks and flows in real time. However, these data are highly biased. Only if this bias is properly accounted for, social media data may supplement existing representative data sources such as surveys. Often, it is assumed that this bias is constant over time which may not be tenable as both platforms and user behavior underlie frequent changes. We revisit the assumption of constant biases using data from the Facebook Ads Manager assessing the variability of the Facebook data over time.

Keywords: Digital and computational demography, Big data / Social media, International migration, Bayesian methods / estimation

See extended abstract.

  Presented in Session 199. Augmenting Census and Other Data to Better Understand Spatial Population Distributions