The Causal Effect of Unionization on Firms’ Earnings Management : Evidence from the Regression Discontinuity Design

This study explores the causal effect of unionization on firms’ earnings management. By comparing firms that just pass union elections by a small number of votes to those that barely loss elections, the regression discontinuity design (RDD) estimations suggest significantly lower earnings management for unionized firms. Further, this decrease is only significant in states without right-to-work legislation where unions are more powerful. These findings are consistent with unionized firms having incentives to report lower earnings to prevent unions from demanding higher wages and benefits. Among unionized firms, I do not find evidence that unions can undo the effect of lower reported earnings when negotiating on future pension and retirement benefits, and employee expenses, i.e. firms reduce employee expenses to some extent by strategically doing lower earnings management