Estimation of Gender Wage Gap in the University of North Carolina System

2025-05-29

AI Summary

This study investigates the gender wage gap among tenure-track faculty in the University of North Carolina system. Employing a causal inference framework, it pioneers the systematic integration of propensity score matching (PSM) and causal forests—methods previously underutilized in higher-education compensation research—to rigorously control for confounding variables including institutional type, discipline, rank, years of service, H-index, and academic service load. This approach isolates the net causal effect of gender on salary. Results indicate that, conditional on comparable qualifications, female full professors earn approximately 6% less than their male counterparts—a statistically significant and robust finding across multiple sensitivity analyses. By moving beyond conventional descriptive regression, the study delivers stronger causal evidence on gender-based pay inequity in academia, offering a methodologically rigorous foundation for policy interventions aimed at achieving equitable faculty compensation.

Technology Category

• Reasoning under Uncertainty: Causality
• Machine Learning: Causal Learning
• Knowledge Representation and Reasoning: Action, Change, and Causality

Application Category

• Economics, Online Markets and Human Computation: Fairness, privacy, and diversity in economic environments
• Social Networks and Social Media: Fairness and bias in social network and social media analysis
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Gender pay equity remains an open challenge in academia despite decades of movements. Prior studies, however, have relied largely on descriptive regressions, leaving causal analysis underexplored. This study examines gender-based wage disparities among tenure-track faculty in the University of North Carolina system using both parametric and non-parametric causal inference methods. In particular, we employed propensity score matching and causal forests to estimate the causal effect of gender on academic salary while controlling for university type, discipline, titles, working years, and scholarly productivity metrics. The results indicate that on average female professors earn approximately 6% less than their male colleagues with similar qualifications and positions.

Problem

• Estimating gender wage gap among UNC faculty
• Applying causal inference methods for wage disparity analysis
• Controlling variables to isolate gender's effect on salary

Innovation

• Used parametric and non-parametric causal methods
• Applied propensity score matching technique
• Employed causal forests for salary analysis