Decomposing temporal changes in covariate contributions to wage inequality

Abstract

This article provides counterfactual extensions to the Fields inequality decomposition framework and at the same time uses the nonparametric methodology in estimating bootstrap SEs and constructing confidence intervals. For methodological illustration, we decompose changes in schooling's inequality contribution into changes in attributes and returns using a simple counterfactual measure that combines return to schooling in a given year with the distributional characteristics of schooling in another year. Using data from the Philippine Labour Force Surveys (LFSs) in 1988 and 1995 and replication methodologies from Racine (2010), the results show that schooling's contribution to inequality in the Philippines is robust but has declined in 1995 relative to 1988. The decline in the said measure is due to the reduced variability of schooling achievement in 1995 but is mitigated by the rise in schooling returns in 1995.

Keywords:

• inequality decomposition
• wage inequality
• Philippines
• bootstrap

JEL Classification:

• C14
• D63

Notes

¹ For this article, we simply consider the quadratic Mincerian specification wherein the log of wages is linear in schooling and quadratic in potential experience.