"This study uses newly disaggregated National Labor Relations Board (NLRB) election data to revisit the theory that sectoral and regional shifts in economic activity contributed substantially to private-sector union decline in the United States. Unlike most studies, which focus on differential employment growth among union and non-union establishments, this article focuses on how such shifts may have affected organizational rates themselves. Improved data permit a shift-share decomposition that indicates that approximately 40% of the decline in union elections is in response to sectoral shifts, the majority attributable to changes within each sector. Moreover, in an update to Dickens and Leonard's 1985 study, the author shows that declining organization rates since 1980 are responsible for a decline in union density of 5.4 percentage points."

Support: Digital