Abstract
This paper estimates whether non-union wage supplements associated with unionization are larger for low wage workers. Switching regression estimates determine low and high wage samples and demonstrate larger supplements for most low wage workers. The pattern varies by occupational group.