The Analysis of the Impact of Panel Attrition on Estimation of Regular-Irregular Worker Wage Gap in the KLIPS
The aim of this paper is analyzing the effect of panel attrition on estimation of regular-irregular worker wage gap using KLIPS (Korean Labor and Income Panel Study). Using two wave sub-panels of KLIPS, we first analyze the characteristics of attritions. We find that the nonrandom attrition has occurred and it causes the underestimation of regular-irregular worker wage gap. Second, we decompose the attrition bias into ‘ability bias’ and ‘distortion bias’. And third we develop the estimation strategies to reduce the bias. We have found that the bias is not negligible although it has been attenuated by change of job of workers.