We examine how firms estimate CEO-employee pay ratios in response to the CEO pay ratio rule, the first mandated pay inequality disclosure for U.S. firms. Our findings reveal that firms disclose lower pay ratios when they use more complex methods to identify the median employee. The relation between the estimation method and the disclosed pay ratio is stronger for firms headquartered in states with a greater societal aversion to income inequality and is weaker when CEO pay in the prior year is lower. Firms’ estimation choices do not merely represent selection bias or potentially omitted variables such as firm size, industry, compensation design complexity, or workforce composition – including the presence of foreign, temporary, seasonal, or highly paid employees. Although firms that use more complex methods to identify the median employee disclose lower pay ratios, we find no evidence of real changes in pay inequality among these firms. Our results suggest that some firms use estimation discretion to appear to conform to stakeholders’ preferences instead of taking real actions. These practices call into question the informativeness of CEO pay ratio disclosures, highlighting a potential cost of granting discretion in mandatory ESG disclosures.