Econometric Analysis of Older Workers' Hours of Work Using A Copula and Sample Selection Approach

A. Wichian, S. Sriboonchitta

Abstract

This paper aims at a copula approach to a sample selection modelto determine the factors aecting the labor force participation and working hoursof older workers in Thailand. Several of the copula functions, such as the Gaus-sian, and Archimedean copulas are compared to the standard Heckman's methodwhich is restricted to linear dependence. This paper demonstrates that the copulaapproach to the sample selection model, which allows for exibility in dependencestructure and relaxing the joint normality assumption, works in this context andperforms better than the BVN (bivariate normal model) especially, the Frank (L{t)copula model. Furthermore, the results obtained show the presence of signicantpositive dependency of unobserved factors between two error terms, which impliesthat the selectivity bias exists and also that the unobserved factors that increase(decrease) the propensity to participate in the labor force of older workers alsoincrease (decrease) the working hours. In addition, the policy recommendationis that the stakeholders should consider on having a program that can increaseworking hours of some worker groups such as female workers, or workers who arenot single or workers who do not stay in the Bangkok province.