Conducting a thorough pay equity audit is an important step in addressing compensation inequality among employees. It is an essential component of any diversity, equity, and inclusion (DEI) program and, when done correctly, can increase employee satisfaction, enable an organization to proactively mitigate inequity, and reduce legal liability. 

However, one of the most complex issues to address during a pay equity audit is the preparation of workforce data. The data must be accurate before beginning the analysis.

This article is second in the three-part series, Designing a Successful Pay Equity Program, and focuses on:

  • Why data quality management is critical
  • How data quality impacts the success of a pay equity audit
  • What employers should do to ensure accurate data 

Pay Equity Auditing Starts With Reliable Data

Accurate data is essential for any thorough analysis and is necessary for drawing precise conclusions. When it comes to a pay equity audit, workforce data is required. Many organizations store their workforce data across numerous systems (benefits administration platforms, HRIS systems, payroll platforms, time and attendance systems, financial management platforms, applicant tracking systems, performance review systems, etc.) and this creates a myriad of potential data quality issues. 

To overcome these challenges, organizations need to leverage a well-thought-out extraction, transformation, and loading (ETL) process.

The first step is to collect and aggregate all the data from each system. This is known as the extraction process. The data then needs to be reviewed to identify any potential data quality issues. This is where the data transformation process begins. If flaws are identified and are not corrected before loading the data into pay equity auditing analytics software, then an organization may be addressing non-existent problems while simultaneously overlooking the true issues driving pay inequity. 

The challenge then becomes, how do organizations assess their workforce data quality? The data must be vetted and checked through a series of lenses referred to as the five dimensions of data quality.

Five Dimensions to Achieving Data Quality

When it comes to analyzing the quality of data, there are five dimensions that are essential for identifying and correcting data quality errors. 

The first dimension of data quality is completeness. Completeness of data refers to whether all the necessary information has been collected. For example, some HR and payroll systems may only support one hire and termination date per employee. For employees that are rehired, this lack of data could lead to inaccurate employee tenure and could make an employee appear overpaid. Anytime there is a lack of complete data, it can have serious implications on the pay equity audit results. 

The second dimension is validity. This deals with how data is entered and stored in a system. It is important for all employee data to be valid so the final results aren’t skewed and accurate comparisons can be made between individuals or groups of employees. An employee ID, for example, must be entered correctly in a data system or it could lead to associating the wrong pay data with a particular employee, which would further lead to inaccurately drawing comparisons between workers’ compensation.

The third and fourth dimensions relate to consistency. Ensuring that the data is consistent across both platforms and fields is critical. First, let’s review an example of data consistency across platforms. If an employee was promoted to a new job title with a new set of responsibilities in the middle of a pay period and this change is reflected in the HR system, but not in the payroll system or vice versa, the data is inconsistent across the two platforms. If the data is then sourced from the system with inconsistent data for use in the pay equity audit, a mistake will have been made. 

Next, let’s consider an example of data consistency across data fields. If the wrong hire or termination date is entered into the payroll system and it doesn’t align when compared to the employee’s actual wages and hours worked, then the data across fields is inconsistent. If data lacks consistency across platforms and fields, inaccurate conclusions can be made when comparing employee earnings during the pay equity analysis.

The fifth dimension of data quality is anomaly detection, which is the process of identifying the most extreme values in the data. These extreme values indicate highly unlikely or impossible scenarios. Here’s an example: If the payroll or time and attendance system show that an employee worked 600 hours in a given month, the value is likely an impossible anomaly, since there are less than 200 working hours in a given month. Flagging this data value as an anomaly is essential, as failing to do so could skew the employee’s earnings.

It should be noted, however, that not all anomalies are invalid. Anomalies can also reflect accurate data that can have a substantial influence on pay equity audit results, such as highest or lowest compensation and longest tenure. This type of data needs to be reviewed for accuracy and data anomalies must be addressed so false conclusions are not made during a pay equity analysis.

These five dimensions of data quality must be repeated until there is 100% confidence that the data can be trusted for use in the pay equity audit. The implications are significant, as pay equity audits yield compensation recommendations and could have real financial implications to both the company and the employees if done incorrectly.

Building Trust in Your Data for Routine Pay Equity Auditing

Making the commitment to achieve pay equity and getting it right requires recognizing the importance of data quality.

Since workforce data is constantly changing as employee data is created, updated, and deleted, it’s important that the information be routinely checked. To ensure data accuracy, organizations must incorporate a regular data quality process that identifies any new issues and addresses any repeat issues that appear. While there is no set time frame for how often data quality should be addressed, a general rule of thumb is once a month. Ongoing monitoring of the information ensures all workforce data is as accurate as possible.

As we have discussed, pay equity auditing is complex and if the results of the analysis are built off low-quality data and then relied upon as guidance in adjusting workers’ compensation, the negative impacts can be significant. Organizations that proactively audit and monitor their systems for high-quality data are setting themselves up for success and reducing their legal liability. More importantly, organizations that actively work towards achieving pay equity are upholding the Equal Pay Act and are giving back to society by improving workers’ livelihood.

Learn More About Designing a Successful Pay Equity Program

This article is the second in a series of three that explains how organizations can design a successful pay equity program. The final piece will explore how to conduct a pay equity audit, the importance of attorney-client privilege, how to implement a chosen solution and act upon pay equity audit findings. Together, this series will provide insight into how organizations can create a sustainable framework for achieving pay equity. To learn about how organizations approach creating a pay equity program, read part one of the series.

Organizations looking to disclose pay equity, diversity, and inclusion data information should do so within an ESG reporting framework. Download our white paper, DEI in ESG Reporting to learn about the different standards you can leverage for sharing your progress.

Leave a Comment