Predicting frequency of political discussion at work (0-4 scale): Labor market power as predictor
Notes: All variables standardized to run from o to 1. Industry and occupation fixed effects included but not graphed. Survey weights and robust standard errors applied.