"Two well-documented empirical findings are that unionized employees typically receive substantially higher compensation than their non-union counterparts and that union representation in the United States has declined over time. Some observers have hypothesized a causal link between these two phenomena: the achievement of a union wage premium, they argue, hastened union decline by inducing employers to adopt various union avoidance strategies. Using data from the Current Population Survey and the Census Bureau's Census of Construction, the authors test this hypothesis for the construction industry. Even in estimations that allow for a lagged response and incorporate a variety of controls, they find no evidence that high union/nonunion wage ratios in construction in the 1970s or 1980s resulted in lower union membership in 2000"

Support: Paper