Advances in Management and Applied Economics

The Effects of Distributional Assumptions on the Full-time and Part-time Wage Differentials

• [ Download ]
• Times downloaded: 4

• Abstract

   

  The extent to which the distribution of the disturbance term in the estimated wage equation affects the wage differential between full-time and part-time workers is examined in this paper. Adopting a switching regression model with known sample selection, I found that the normality assumption generates larger wage estimates than the estimates of the non-normal distributions. The results indicate that the Normal distribution produces the larger wage differentials than the Non-normal distributions. Also, regardless of distributional assumption, differences in full-time and part-time characteristics account for a larger portion of the full-time and part-time wage differentials. The empirical message derived from this study is that, studies that rely solely on the normality assumption may not provide a true picture of the size of the estimated wage gap between full-time and part-time workers. In general, the study seems to suggest that the estimated wage differential between groups such as male-female and white-black under the normality assumption may be overstated.

   

  JEL classification numbers: J31, J22, J23, J20.

  Keywords: Distributional assumptions, Switching regression, Normal and Non-Normal distributions, Full-time/Part-time wage differentials.