A Comparison of JMP Wage Decomposition and Quantile Regression Methods in Wage Inequality Assessment

Nazmi Y. Yağanoğlu, Hakan Ercan


The decomposition technique of Juhn, Murphy and Pierce (1993) and quantile regression are two of the main tools of wage inequality analysis. JMP technique has the advantage of decomposing the change in wages into three components, and showing residual inequality easily. Quantile regression has the advantage of showing a detailed picture of wage distribution at different quantiles. We apply both techniques to March Current Population Survey (CPS) data of the US Bureau of Labor Statistics (BLS) to analyze the changes in wage inequality in the US during the 1967-2005 period. We compare the results to see which technique produces more useful results in response to the research question at hand. We find that it is a good idea to check the quantile regression results before concluding on JMP values since if quantile regression coefficients are very different from OLS coefficients (meaning the wage distribution is quite different from a normal one), results of two methods differ greatly and the application of JMP is problematic.


Wage Inequality, US, JMP, Wage Decomposition, Quantile Regression

Full Text:

Full Text

Comments on this article

View all comments

Contact info:
ODTÜ İktisadi ve İdari Bilimler Fakültesi
A Binası 06800 Çankaya / Ankara
E-mail: metusd@metu.edu.tr
Tel: +90 312 210 2006
Powered by Open Journal Systems.
Copyright METU Studies in Development 2010-2012.