Estimating gender inequalities in labor-market outcomes using mobile phone data

Abstract

Mobile phone data holds promise for contributing to slow-filling gaps about women and men’s labor. We generated gender-specific predictions of three labor market indicators (employment, unemployment and underemployment) using machine learning models that analyzed digital trace data and geospatial data. While the models correctly predict mobile phone users’ gender in most cases, they predict users’ labor market status much less accurately. With further refinement, we believe the methodology still shows prospects for filling gender data gaps in individual-level labor market statistics.