A Quantile Regression Modelling Approach to Study Gender Wage Gap in India

Samapriya Trivedi, Shambhavi Mishra
Statistika, 105(1): 102–122
https://doi.org/10.54694/stat.2024.18

Abstract
The Indian labour market exhibits significant gender wage disparities, particularly among regular/salaried employees and casual workers. To study these disparities comprehensively, we present a dual-methodological approach by combining Quantile Regression (QR) and Melly-Machado-Mata (MMM) decomposition. Using secondary data from India's Periodic Labour Force Survey (PLFS) 2020–21, the study highlights the intricate interplay of various demographic, personal, and occupational characteristics on wage distributions. The findings highlight the persistence of the gender wage gap across different quantile levels for both employment types. The decomposition results reveal that discrimination significantly contributes to the wage gap, particularly at lower income levels, indicating a "sticky floor" effect for regular/salaried employees. Conversely, casual workers face a consistent wage gap across all quantiles, with discrimination remaining a crucial factor. This research highlights the robustness and precision of QR modelling and decomposition, providing a comprehensive framework for scientifically assessing the gender-based wage gap and exploring policy interventions to address these inequalities.

Keywords
Quantile regression, decomposition, Oaxaca-Blinder, Melly-Machado-Mata, PLFS