The authors conjecture that previous studies have tended to overestimate the productivity of union firms relative to nonunion firms due to inadequate controls for output quality and input usage—important omissions if the higher cost of unionized labor leads to less labor-intensive...