Rethinking Measurement of Pay Disparity and its Relation to Firm Performance

I develop measures of firm-level pay disparity and examine the relation between these measures and firm accounting performance. Using comprehensive compensation data for a large sample of firms in the S&P 1,500, I do not find a statistically significant relation between the ratio of CEOto-mean employee compensation and accounting performance. I next create empirical models that allow me to separate the components of CEO and employee compensation explained by economic factors from those that are not, and use them to estimate explained and unexplained pay disparity. After validating my estimate of unexplained pay disparity by documenting that it is negatively related to measures of employee satisfaction, I find robust evidence of a negative (positive) relation between unexplained (explained) pay disparity and future firm performance. Additional tests show that the negative relation between unexplained disparity and firm performance is more pronounced for firms with weak corporate governance and those where employee turnover is more prevalent and more likely. These results provide support for the predictions of several theories on the relation between pay disparity and firm performance, and offer a roadmap for investors to interpret the CEO-to-median-employee pay ratio that publicly traded firms must disclose beginning in 2018.