Mapping urban socioeconomic inequalities in developing countries through Facebook advertising data

Area of Science:

• Socioeconomic studies
• Urban planning
• Digital data analytics

Background:

• Accurate socioeconomic measurement is crucial for the UN's Sustainable Development Goals.
• Traditional data collection methods for socioeconomic indicators often lack timeliness and spatial detail.
• Novel digital traces offer a complementary data source for enhanced monitoring.

Purpose of the Study:

• To assess the utility of Facebook advertising data for fine-grained socioeconomic mapping in urban areas.
• To evaluate the predictive performance of these digital traces across diverse global settings.
• To explore the potential and limitations of social media data for human development measurement.

Main Methods:

• Collected publicly available, anonymized Facebook advertising audience estimates.
• Utilized these data to predict socioeconomic conditions of urban residents.
• Focused on four diverse cities: Atlanta (USA), Bogotá (Colombia), Santiago (Chile), and Casablanca (Morocco).

Main Results:

• Behavioral attributes from Facebook marketing data accurately mapped urban socioeconomic status at a fine spatial level.
• Predictive performance was consistent across both high- and low-resource urban environments.
• The study confirmed the value of social media data for human development metrics.

Conclusions:

• Facebook advertising data can effectively measure socioeconomic status in urban areas.
• While valuable, generalizing these predictive models across different countries presents limitations.
• Digital data offers a powerful, complementary tool for monitoring global development goals.