THE EVOLUTION OF GENDER WAGE DIFFERENTIALS IN THAILAND: 1991–2007 — AN APPLICATION OF UNCONDITIONAL QUANTILE REGRESSION

Abstract

Using unconditional quantile regression and an approach which allows the identification of the individual contributors to the gender wage gap and its over-time change, we study the gender wage differentials over the whole distribution in Thailand from 1991 to 2007. We also propose an extension of the single quantile decomposition to explore the origins of the distributional changes in gender wage differentials over time. A persistent sticky floor effect and declining gender wage gaps over time are observed. Aggregate decompositions using a reweighting approach show that most of the gap and changes in gaps are due to (changing) coefficients effect. Detailed decompositions show that within the (changing) characteristics effect, education contributes to the narrowing of the gap over the whole distribution, and within the coefficients effect age and the constant term are the most influential factors but with opposite effects.