Union Effects on Productivity: Evidence from Western U.S. Sawmills

Abstract

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 products and techniques. To avoid those problems, this study examines a fairly standardized commodity, lumber, and controls for detailed product attributes and inputs. An analysis of data from a survey administered by the authors shows that unionized sawmills were between 12% and 21% less productive than nonunionized mills in fiscal year 1986. As predicted, when product quality and raw material usage are not included in the analysis, the estimate of union productivity is biased upward.