Moving Pictures publicly through the Humanitarian Data Exchange. Response to the work has been overwhelmingly positive, particularly from academic peers who are intrigued by the methodology and from international government organisations which are interested in how this type of data can enhance their decision-making. “Government statisticians often express frustration that private companies hold vast amounts of data on population movement that they can’t access,” said Professor Abel. “I hope this collaboration demonstrates what’s possible when one of the world’s largest internet platforms opens up its data in a way that enables researchers to generate meaningful migration estimates across a broad range of countries.” The study, which was published in Proceedings of the National Academy of Sciences, also prompted an opinion piece in The New York Times titled ‘To Understand Global Migration You Have to See It First’, which praised the research for providing the first such clear picture of humanity in motion and making it public. Professor Abel is currently working with colleagues to develop a deep learning method that integrates the international migration flow estimates from the Facebook data with a wide range of other migration sources and related measures. “We are producing annual estimates of global migration over a much longer time horizon than what was possible using the Facebook data alone,” he said. “This will help build a more comprehensive, historically grounded understanding of global migration dynamics and serve as a model for combining digital and traditional data sources in migration research using machine learning methods.” The study captured and analysed migration flows across numerous countries during a period of exceptional global disruption in 2022. The results demonstrated that digital traces – such as anonymised, aggregated location data – can reveal shifts in migration long before they appear in official statistics. This provides an early lens on population movements during events – such as the COVID-19 pandemic at the time – and regional conflicts, underscoring the potential of digital data as an early warning system for changes in shifting migration trends. The findings estimated that 39.1 million people migrated internationally in 2022 (0.63 per cent of the population of the countries in the study sample). Migration flows significantly changed during the COVID-19 pandemic, decreasing by 64 per cent before rebounding in 2022 to a pace 24 per cent above the precrisis rate. The researchers also found that migration from Ukraine increased 10-fold in the wake of the Russian invasion. To support research and policy interventions, they released these estimates “Improved migration data can better inform governments’ responses to migrant crises, support the development of more effective policies, and help researchers uncover deeper insights into migration trends and dynamics.” “Traditional international migration statistics are often published with delays, inconsistent definitions and, in many countries, not published at all,” said Professor Abel. “This results in a fragmented and imprecise picture of global migration patterns. Our approach opens new avenues for evidence-based migration policymaking and humanitarian response, particularly in rapidly evolving situations where timely data is crucial.” The work demonstrates that it is possible to measure migration accurately and in an expeditious manner across a wide range of countries. “Since migration flow data is often unavailable – and is only consistently reported in predominantly wealthy Western nations – our approach provides a rare window into movement patterns in many developing countries, placing those patterns within the broader global migration network,” said Professor Abel. Early warning system Delays and inconsistencies Our approach opens new avenues for evidencebased migration policymaking and humanitarian response, particularly in rapidly evolving situations where timely data is crucial. Professor Guy Abel The study took the groundbreaking approach of leveraging online data from Meta to provide the most comprehensive and timely estimates of migration’s true global sweep to date. The co-author of the research, Professor Guy Abel from the Department of Sociology, explained that the idea originated from a shared interest in developing new methods to understand international migration in near real-time. “Meta’s Data for Good initiative uses data from its platform to support research that advances public knowledge and informs policymaking,” he said. “At the outset of their global mobility project, Meta contacted me based on my prior work on indirect methods for quantifying international migration. “Their aim was to estimate the migration of the entire population – not just Facebook users – which meant they needed to develop techniques to weigh the data appropriately and validate the results against official statistics. That initial outreach laid the foundation for this collaborative effort, in which researchers from Harvard University also participated.” What makes the work pioneering is that using anonymous, aggregated location data from a near-global digital platform enabled the researchers to observe migration flows at a much finer temporal scale, across a broader geographic range, and with a consistent definition of migration that is not feasible with existing data sources. Significantly, the study adopted the United Nations’ guidelines for defining a migrant which include timing criteria aimed at excluding business travellers and tourists, and including only people who spend the majority of successive years in two different countries. An innovative study takes a novel approach to measuring migration flows using online data, thereby providing a clear picture of a global population that is on the move. HKU Bulletin | Nov 2025 Research 24 25
RkJQdWJsaXNoZXIy ODI4MTQ=