Gendered Wage Returns to AI-Supported Non-Routine Tasks and Horizontal Gender Segregation

• In Event: Inequality, Poverty, and Mobility Section Roundtables
  In Refereed Roundtable (60 minute): 10. AI and Technology

Abstract

The diffusion of artificial intelligence (AI) is reshaping non-routine work tasks and is increasingly linked to wage differentials. While prior research documents overall wage premiums associated with AI use, little is known about which AI-supported non-routine tasks yield the strongest wage returns and whether these returns differ by gender across gender-segregated occupational contexts. This study integrates task-based economic approaches with sociological perspectives on gendered occupational structures to examine how AI-supported non-routine tasks relate to wages for women and men. Using data from the 2024 German BIBB/BAuA Employment Survey combined with occupational gender composition from the 2023 German Microcensus, we estimate multilevel mixed-effects models for over 11,900 employees. We analyze wage associations of AI use across IT-related tasks, task complexity, general non-routineness, and leadership roles, while explicitly accounting for horizontal gender segregation. Results show that wage associations with AI are concentrated in highly non-routine and expert-level tasks, particularly expert IT tasks. Across most task domains, women and men exhibit similar wage associations with AI use. Notable gender differences, however, emerge in leadership positions and across occupational contexts. Gender wage gaps are largest in leadership roles without AI support, especially in male-dominated occupations, whereas these gaps largely disappear among AI-using leaders. In female-dominated occupations, women show stronger wage associations with AI in non-leadership roles but lower wages than comparable men in AI-supported functional leadership positions. Overall, the findings indicate that AI-related wage patterns are embedded in gendered task structures and occupational segregation and are associated with context-specific wage differences rather than uniform patterns of gender inequality.

Authors

• Alexandra Wicht, University of Siegen

• Ana Santiago-Vela, Federal Inst for Vocational Education and Training (BIBB)