Labor market sorting and the gender pay gap revisited

Area of Science:

• Labor Economics
• Econometrics
• Sociology of Work

Background:

• Gender pay gaps are a persistent issue in labor markets worldwide.
• Existing estimation methods may not fully account for labor market structures like occupational segregation.
• Switzerland provides a relevant case study due to its distinct labor market segmentation.

Purpose of the Study:

• To investigate the impact of gender segregation on the estimation of gender pay gaps.
• To propose and evaluate methods for more accurate pay gap estimation in segregated labor markets.
• To quantify the differences in estimated unexplained gender pay gaps based on methodological choices.

Main Methods:

• Analysis of labor market data from Switzerland to identify segments with perfect gender sorting.
• Application of econometric models, including flexible semi-parametric estimators.
• Comparison of standard Blinder-Oaxaca decomposition with methods accounting for lack of overlap and enforcing ex ante comparability.

Main Results:

• Sizable labor market segments exhibit perfect gender sorting, preventing direct non-parametric identification of pay gaps.
• Methodological choices regarding ex ante comparability and estimator selection significantly influence pay gap estimates.
• A flexible semi-parametric estimator explained up to 38% more of the raw gender pay gap, reducing unexplained gaps by up to 44% compared to standard methods.

Conclusions:

• Gender segregation poses a significant challenge to accurately estimating gender pay gaps.
• Standard estimation techniques may underestimate or misattribute the causes of gender pay disparities.
• More sophisticated methods accounting for labor market structure are crucial for reliable gender pay gap analysis.