Bridging Data Gaps: Social Media and Traditional Sources in Assessing Migration in Latin America

Victoria Prieto Rosas, Universidad de la República, Uruguay
Julieta Bengochea, Universidad de la República
Mariana Fernandez Soto, Universidad de la Republica
Martín Koolhaas, Universidad de la Republica
Camila Montiel, Universidad de la República
Gabriela Pedetti, Universidad de la República
Martín Pedemonte, Universidad de la República
Alicia Rubini, Universidad de la República

Web social media data provides a real-time alternative for assessing migrant populations, but it has limitations. Establishing a reliable gold standard for comparison is challenging in regions like Latin America, with poor register data and where surveys and census have their own shortcomings. This paper explores the possibilities and limitations of using web social media data to estimate the size and demographic characteristics of migrant populations, comparing metadata with census and household survey data for selected Latin American countries. Using official household surveys and census data, we assess the correlation between absolute numbers of migrants, disaggregated by place of birth and sex, and analogous daily user data from Facebook, Instagram, and Messenger, disaggregated by country of previous residence. This analysis is also conducted for the total population. We focus on countries where recent census and household survey data are available: Argentina (2022), Ecuador (2022), and Mexico (2020). The findings contribute to understanding the applicability and limitations of digital data in migration research for regions while joining a broader conversation on improving migration measurement by combining multiple data sources.

Keywords: International Migration, Big data, Data and Methods, Census data

See extended abstract.