Artificial Intelligence and the Gender Pay Gap in Labour Markets: Cross-Country Evidence and the Role of Gender-Sensitive Reforms
Artificial Intelligence and the Gender Pay Gap in Labour Markets: Cross-Country Evidence and the Role of Gender-Sensitive Reforms

Summary/Abstract: This paper investigates whether the diffusion of artificial intelligence (AI) is associated with gendered labour market outcomes at the macro level and whether gender-sensitive institutional reforms mitigate gender inequality. We constructed a cross-country dataset covering up to 50 countries over the period 2012–2022. Given the rapid acceleration of AI adoption after 2022, the estimates should be interpreted as evidence for the 2012–2022 period and may not reflect the most recent wave of AI diffusion. We then estimate three complementary econometric models. First, using 500 promotion-related observations pooled across countries, we estimate a logistic regression to test whether AI adoption and an AI bias proxy are associated with women’s promotion probabilities (RQ1; H1a–H1b). Second, using a balanced panel of 10 countries observed over 10 years (N = 100), we estimate country fixed-effects models for female labour force participation and the gender pay gap (RQ2; H2a–H2c). Third, we employ a difference-in-differences specification on the same panel to assess the impact of gender-sensitive institutional reforms on the gender pay gap (RQ3; H3). Across the logit and fixed-effects models, coefficients on AI adoption and the AI bias indicator are small, statistically insignificant, and imprecisely estimated, providing no support for the hypotheses that AI diffusion widens gender gaps in promotions, participation, or wages. By contrast, the reform indicator in the difference-in-differences model is negative, sizable, and highly significant, indicating that treated countries experience an average reduction of about 6.4 percentage points in the gender pay gap relative to non-treated countries. The findings suggest that AI adoption, as currently measured, is not a detectable structural driver of gender inequality in labour markets at the country level, whereas gender-sensitive institutional reforms are empirically associated with improved wage outcomes for women.

• Details
• Contents