Econometric Analysis of Private and Public Wage Determination for Older Workers Using A Copula and Switching Regression

A. Wichian, S. Sriboonchitta

Abstract

This paper aims on applying the copula approach to an endogenousswitching regression model to determine the public and private sector wages forolder workers in Thailand. The copula approach to endogenous switching modelsnot only allows for exibility in the specication for the margins but also considersemploying dierent families of copulas to measure the dependence between the dis-turbance terms in the switching equation ("s) and those in the two wage equations("0 and "1). This paper demonstrates that based on the log{likelihood value andthe criterion of BIC, all of the copula{based models perform better than the stan-dard model, for this context, especially with the Frank-Gaussian (L{t{t) model.Furthermore, these results show the presence of signicant negative dependencyof unobservable factors between the switching regression and the wage regression,which implies that those older workers who are engaged in the public sector havesuered more from the wage penalty than those in the private sector. Thus, theconcerned policy makers should run campaigns that encourage the public sectorto retain the older workers with high wages, especially on those individuals withhigher education and skills.