"Using data from the Survey of Income and Program Participation, the author estimates the impact of union coverage on training in the United States, comparing the construction industry with other broad industry sectors. The author finds no statistically significant union training effect in construction or other private-sector industries for employer-paid training. For a more broadly defined job-training measure, however, a large union effect is found in construction, and a smaller, yet statistically significant, effect is found in other private-sector industries. The results are not entirely consistent with either the standard human capital model or models of imperfect competition. "

Support: Paper