Different firm size effects on wages for supervisors and workers

We find that: (i) the plant size effect for supervisors is at least double that of non-supervisors, (ii) the difference is statistically significant and (iii) the firm size effect for supervisors is not explained by differential returns on standard variables