This paper analyzed the effect of labor unions on wage premium and wage inequality using establishment-worker linked panel data which combines workplace panel survey from 2014 to 2019 with employment insurance data.
 First, an unconditional quantile regression methodology was used to estimate how much union wage premiums differed over the entire sample during the period. The result shows that the union raises the wage premiums of the 40th quartile wage group the most and decreases the effect toward the low wage group and especially the high wage group. It seems that union represent the economic interests of the middle and lower wage groups most.
 Estimating the relationship between the inequality index by business establishment and the unionization rate using the fixed effect, the unionization rate showed a statistically significant negative(-) relationship with all inequality indexes such as the Gini coefficient, the Tile index, the Palma ratio, and the p90/p10 quantile.
 The above estimation results imply that labor unions have had a statistically significant impact on inequality reduction in the labor market of at least the sample (business establishment with 30 or more employees), and it is necessary to consider progress in unionization as a way to improve the unequal labor market situation in the future.