Using social media advertising data to estimate migration trends over time

Abstract

Understanding migration patterns and how they change over time has important implications for understanding broader population trends, effectively designing policy and allocating resources. However, data on migration movements are often lacking, and those that do exist are not produced in a timely manner. Social media data offer new opportunities to provide more up-to-date demographic estimates and to complement more-traditional data sources. Facebook, for example, can be thought of as a large digital census that is regularly updated. However, its users are not representative of the underlying population, thus using the data without appropriate adjustments would lead to biased results. This chapter discusses the use of social media advertising data to estimate migration over time. A statistical framework for combining traditional data sources and the social media data is presented, which emphasizes the importance of three main components: adjusting for non-representativeness in the social media data; incorporating historical information from reliable demographic data; and accounting for different errors in each data source. The framework is illustrated through an example that uses data from Facebook’s advertising platform to estimate migrant stocks in North America.