FAIRWORK · Addressing Ethnic and Gender Inequalities in Labor Markets

Persistent disparities in hiring and wages remain a defining feature of European labor markets, particularly for women and immigrants. These gaps reduce equity and undermine efficiency by causing a misallocation of talent. FAIRWORK investigates two key drivers of inequality and tests scalable interventions to address them: (1) ethnic discrimination at different stages of recruitment and (2) gender differences in job search linked to women’s preferences for jobs providing higher work-life balance. The project pursues three objectives: (i) identifying the determinants of ethnic discrimination at each recruitment stage, (ii) evaluating the role of information frictions in shaping women’s job search behavior, and (iii) testing whether algorithmic screening and work–life balance information can promote both fairness and efficiency in job matching.To achieve these goals, FAIRWORK leverages a unique collaboration with a major online job-matching platform in the hospitality sector, providing access to rich administrative data covering the full recruitment pipeline, from applications to shortlisting, interviews, and final hires. The project combines quasi-experimental methods, surveys, and large-scale randomized controlled trials (RCTs). In WP1, it will test whether algorithmic shortlisting reduces ethnic bias compared to human evaluators under cognitive load, delivering the first causal field evidence on the potential of algorithmic tools to equalize opportunities for disadvantaged workers. In WP2, it will exploit the introduction of work-life balance indicators on job ads and implement an RCT to test whether providing women with information on job amenities increases job finding, retention, and affects gender wage inequality. FAIRWORK's findings will be directly relevant to researchers, policymakers, and European institutions, offering actionable insights into how low-cost, scalable solutions can foster fairer and more efficient labor markets in Europe.