Do UK employers misreport information about gender pay equity? While many countries have attempted to close gender pay gaps with legislation, such as the 1963 Equal Pay Act in the US and the 1970 Equality Act in the UK, a common criticism of these laws is that enforcement is difficult because firms need not publicly provide or even internally compile gender pay information. In response to these criticisms, in 2017, the UK began requiring employers to disclose detailed information about their gender pay gaps publicly. Politicians and the media lauded the legislation as an example of how Environment, Social, and Governance (ESG) disclosure and “regulation by shaming” could change firms’ pay practices.
However, in our recent study, we find that many UK employers misreport their gender pay gap information. All UK employers with 250 or more employees must disclose the median (and mean) pay gap, defined as the difference in pay between the median (mean) male and female, scaled by the median (mean) male pay. Additionally, each employer must provide a percentage breakdown of the men and women employed in each pay quartile. We document significant irregularities by using the combination of gender statistics reported by individual UK firms and the distributions of reported values across the full population. In 4.8% of disclosures, internal inconsistencies between the metrics render the combination of the reported median pay gap and the quartile breakdowns mathematically impossible. For example, employers may disclose quartile breakdowns indicating the median woman falls in the lower middle quartile of the overall pay distribution and the median man falls in the top middle quartile. If that same employer reports a negative median pay gap, which indicates the median man makes less than the median woman, this combination of reported metrics would be mathematically impossible. We find the rate of impossible disclosures increased from 2017 to 2018, consistent with a lack of discipline for accurate reporting, and persisted in 2019 when the UK’s Equality and Human Rights Commission (EHRC) made reporting voluntary for one year in response to the COVID-19 pandemic.
Our findings raise the question of how such misreporting appears. We consider the role of enforcement in answering this question. Weak enforcement of reporting accuracy may allow some employers to strategically misreport their gender pay gap information. While assessing the intent behind disclosure is difficult, we triangulate employers’ strategic intent using several approaches. We begin by evaluating the prevalence of perfectly balanced average gender pay gaps (0.0%) and perfectly-balanced gender ratios (50.0%/50.0%). These measures are desirable as they indicate the greatest possible diversity, do not risk making either gender feel underrepresented or dominated by the other, are typically seen as “the target” or “ideal,” and are the ultimate goal of gender equality mandates. Consequently, when a firm chooses to misreport, its strongest incentive is to misreport its gender statistics as perfectly balanced. Consistent with widespread strategic misreporting, many employers report 0.0% pay gaps and 50.0/50.0 gender ratios. Based on fitting a smooth curve to the reporting distribution in the neighborhood of the targets, we estimate that one-third to one-half of the employers reporting no mean pay gap are likely misreporting.
To further explore whether a lack of enforcement relates to misreporting, we construct a novel dataset of gender pay gap restatements. We compare restatement rates for impossible disclosures and find when the marquee metric, the median pay gap, is not favorable for the employer (i.e., women earn less than men), the employer is more likely to restate. However, if the impossible disclosure’s median pay gap is more favorable for the employer (i.e., women earn more than men), the employer is relatively less likely to restate their statistics even though they must be erroneous. This asymmetry is further consistent with at least some firms strategically misreporting.
Having established the prevalence of misreporting, we investigate which types of employers are more or less likely to misreport. We find smaller firms, measured by the number of employees, are more likely to misreport. We find no evidence that misreporting relates to firm profitability or whether the firm’s equity is publicly traded. Across all measures of misreporting, firms that include an optional link to a formal gender pay gap report – in which they discuss and contextualize their gender pay performance –are less likely to misreport. In light of the EU’s upcoming CSR Directive and associated audit requirements, we also provide large-sample evidence on whether such audits may effectively combat misreporting. Our evidence is mixed: while CSR audits relate to lower levels of egregious impossible reporting, we find no relation between CSR audits and reporting perfectly balanced gender statistics, consistent with the broader literature that suggests CSR audits are of lower quality and provide only superficial assurance.
We next turn to firms’ potential motives for misreporting. We first examine whether recent involvement in an ESG controversy relates to misreporting, as firms may use their reported gender statistics to sugarcoat their poor performance or distract stakeholders from problem areas. We find firms with poor ESG controversy scores in the Refinitiv database are more likely to report perfectly balanced gender statistics. Given the link between ESG controversies and misreporting, we examine whether employers benefit from misreporting. Because ESG rating agencies consider gender balance, including the information reported via the UK mandate, one potential benefit of reporting perfectly balanced gender statistics is higher ESG scores, which prior work suggests attracts investment and corporate customers. Consistent with the potential benefits of misreporting, employers reporting 0.0% median pay gaps are more likely to enjoy increases in the Social Pillar component of their ESG scores. These findings add to a growing body of evidence that ESG ratings may not be reliable indicators of ESG performance.
Our evidence can inform how regulators can design gender pay gap and ESG reporting mandates. Without accurate reporting, it is unlikely that firms will be held accountable for their gender pay gaps. This is a particularly timely point as the UK gender reporting mandate includes a provision for re-evaluation in 2022, and regulators in other countries are evaluating and/or considering implementing their reporting mandates and enforcement regimes, both for gender pay statistics and workplace diversity statistics more generally. While a complete accounting of the costs and benefits of any regulatory intervention is beyond the scope of any study, our results suggest that in the absence of strong enforcement and oversight, many employers will misreport the required information, undermining the reliability and usefulness of mandated reports. More generally, as regulators worldwide introduce several reporting mandates related to gender pay and ESG performance, misreporting likely undermines the ability of these mandates to improve performance. Our study suggests disclosure alone may not induce real change in firms without sufficient monitoring and enforcement of accurate reporting. Users of self-reported ESG information should exercise caution when using metrics or drawing inferences in the absence of strong enforcement.
McKenna Bailey is a Ph.D. student at the University of Michigan Ross School of Business.
Stephen Glaeser is an Assistant Professor of Accounting at the UNC Kenan-Flager School of Business.
James Omartian is an Assistant Professor of Accounting at the University of Michigan Ross School of Business.
Aneesh Raghunandan is an Assistant Professor of Accounting at the London School of Economics.
This post is adapted from their paper, “Misreporting of Mandatory ESG Disclosures: Evidence from Gender Pay Gap Information,” available on SSRN.
