In hiring, termination and promotion cases, the analysis examines each step of the decision making process to determine who should have been equally likely to be selected for the particular employment event. For example, in hiring at each stage you look at who remained eligible to progress to the next stage. Ultimately, the investigation should arrive at the final stage where selection rates of the remaining eligible individuals (the refined pools) are examined to see if the selection rates indicate any statistically significant disparity. If they do, discrimination can be inferred.
Compensation analysis is handled differently. Compensation analysis begins with snap shot data that is examined for disparities of a certain magnitude in pay that correlate with protected class status. Certain factors, such as education, skills and experience, believed to impact pay are then examined in a regression analysis and any unexplained residual disparity is attributed to discrimination. The regression analysis will also include items identified by the contractor as possible explanations for how pay is set.
What does not happen in compensation analysis is an examination of how each person in the snap shot actually came to receive their current pay. When OFCCP refers to compensation history, it only refers to the salary figures over a period of time, not to the story of how those salary figures came about. How the individual came to have that salary can shed significant light on the question of whether discrimination was a factor.
For example, suppose the snap shot data shows that a man and a woman doing the same job under the same supervisor in the same department are being paid different rates. Assume for the sake of this example that the woman is earning $2000 less than the man. OFCCP would analyze the factors it perceives as likely to account for pay disparities and if none of these explained the disparity it would likely conclude a discrimination violation occurred.
However, the actual pay history of these two is that the man and the woman were hired into the same starting position with the same starting salary during a similar time period. Both initially worked for the same supervisor and earned comparable salaries. The man sought a transfer to another supervisor who was known to give out great bonuses and salary hikes. The woman did not make a similar choice because she did not want to relocate. Eventually, operations were consolidated and the two were assigned to work under yet another supervisor who simply continued to pay them at the rates they enjoyed when they were reassigned. The man's salary had been increased by $2000 under the generous supervisor and similarly situated women working under that supervisor had also received similar pay increases. The woman's salary remained lower but was comparable to the salary of similar situated men who stayed with the first supervisor. There was no point in this pay history at which a discriminatory decision was made.
However, unless the contractor was aware of and happened to keep proof of all these facts, it would have a difficult time explaining the disparity in pay using only factors like education, time in grade, market rates at hire or any of the usual components of compensation analysis. In this hypothetical, the woman actually is similarly situated to those who worked under her previous less generous supervisor and the man is similarly situated to those who worked under his immediate previous generous supervisor. They are not actually similarly situated to each other but they look like they are so the contractor will likely be cited. Neither the generous supervisor nor the less generous supervisor made pay decisions based on gender and the new supervisor made no pay decisions at all. The observed disparity in pay is actually the result of the woman making a conscious decision to forego higher pay so that she would not have to relocate. It would be very difficult for a contractor to know all of the decisions made by its employees that impacted their pay.
This example shows why it is so difficult to really isolate who is truly similarly situated in a compensation situation using the enforcement mechanisms currently available. The only gender based compensation decision in this hypothetical will be the one imposed by the agency when the contractor is required to raise the woman's salary by $2000 based solely on her gender.
This is not to say that gender discrimination in pay does not exist. However, the kinds of analyses used to find it do not explore sufficiently whether or not comparators are actually similarly situated. In the hypothetical above, the man and the woman actually had an equal opportunity to obtain a higher salary. The woman could have made the same decision to relocate that the man made and would have kept pace with his earnings.
Compensation analysis as presently constructed is actually focused on equal compensation outcomes not equal opportunities for compensation. Because of this focus on outcomes, in other words, being able to demonstrate no meaningful disparities in pay, companies are incentivized to equalize pay outcomes even where there is no evidence of a discriminatory pay decision or any actual lack of an equal opportunity to earn the higher pay. This means that the enforcement mechanisms of government are actually driving contractors to make gender conscious pay adjustments to avoid being accused of pay discrimination even where no one in the agency or the company can identify any discriminatory pay decision or policy. In a hiring case it is obvious which candidates voluntarily withdrew from the process. It is not nearly as easy in a compensation case to identify which potential affected class members voluntarily chose some other job benefit such as not having to relocate over opportunities to increase their salary and financial benefits.
Given that at present the available analytical tools are ill suited to distinguish situations such as those in the hypothetical above from actual discrimination, what steps can contractors take to minimize false positives in compensation discrimination? One thing contractors can control is how the compensation data is presented to the agency.
Contractors provide a hint as to who is similarly situated by who they group into the same job title. For example, if the contractor presents OFCCP with data that includes a job title "Accountant" in the Financial Department of a company, it is suggesting to the agency that everyone under that title may be similarly situated with respect to pay. This may not turn out to be true but the contractor is creating an impression it will have to overcome. If the contractor pays more for experience and instead of lumping all accountants together, separates the title into Accountant I, Accountant II and Accountant III depending on the level of experience, it is already suggesting to the agency that there is a meaningful difference between levels I, II and III and that not all accountants should be viewed as similarly situated with respect to pay. By including such distinctions in the presentation of the data, the contractor begins to shape the agency's expectations.
Similar distinctions can be made based on the level of responsibility associated with a position. For example, if an employee with supervisory responsibilities has the same title as someone without similar responsibilities, the contractor creates an impression that everyone with the title is similarly situated. If the pay reflects the additional supervisory responsibilities but the agency has no way to know that from the data, the contractor is again making an impression it will have to work to overcome. By simply providing a distinctive title for those with supervisory responsibilities, the contractor begins to shape the agency's expectations. The agency will not find it unusual for employees with supervisory responsibilities to receive higher pay than those without such responsibilities.
Meaningful distinctions in title can go a long way toward creating a more accurate picture of your compensation practices. However, creating separate and distinct titles for every single employee will only impair your credibility and invite the agency to substitute its judgment for yours in determining who is similarly situated.
Coordination and Tailoring
In some enterprises, compensation staff, human resources staff and the front line management all have some influence on pay. Not all of these individuals are equally familiar with what a particular employee actually does from day to day. For example, compensation staff may use market data for a generic accountant position. Human Resources may use a generic position description for accountants. Immediate supervisors may set very different performance standards for certain accountants or they may have generic performance standards but have much more sophisticated work assignments for some accountants on staff. The same principle applies here as to titles, do not lump together under a generic description, positions that you are later going to want to distinguish as warranting significantly different levels of compensation. OFCCP is going to think that everyone with the same position description is similarly situated. The distinctions offered by the front line supervisors as explanations for pay are going to be undermined by the position descriptions and generic market data.
Consistent standards in setting pay that are consistently applied can shape the agency's impression of who is similarly situated. For example, the agency rarely takes a hard look at compensation when the compensation system is established in a collective bargaining agreement. From the agency's perspective, the less discretion and the more uniform the calculus for setting pay, the smaller the likelihood that discrimination could have tainted the process. This principal would also apply in the absence of a collective bargaining agreement, if the formal process for setting pay served a similar purpose.
If you have established criteria for how salaries are set, how bonuses are paid, who is given the opportunity for overtime and the like, it is important to self audit to make sure that the policies and practices are operating the way that you intended. This is because everyone who satisfies your criteria for certain pay increases will be viewed as similarly situated to the others who satisfied the criteria and inconsistencies will raise red flags especially if the inconsistencies disproportionately affect historically disfavored groups. Pay policies also provide an opportunity for the company to restate its commitment to fair pay.
Another reason it is so difficult to determine who is similarly situated with respect to pay is the operation of outside influences on the pay snap shot. For example, a contractor may acquire another company and have to reconcile the pay systems. There may be a period of time when individuals doing the same type of work will not have similar pay because they carried their old salaries from the acquired company. Individuals who came on board as result of the acquisition may not be similarly situated to individuals already on staff at the receiving company. Keeping track of these people separately or somehow annotating the records so that these individuals can be easily sorted for analysis, will make it easier for the contractor to explain that the apparent disparities in pay are not indicative of discrimination but rather a by-product of the acquisition. This way the staff acquired through the acquisition will be compared to similarly situated staff similarly acquired and not to the staff already on board. If there are discriminatory differences in pay that you acquired when you bought the target company, the remedies would be isolated to affected acquired employees and not extended to your original staff since the discrimination is not a result of your compensation practices but rather are a legacy from the acquired company.
If you do not maintain these distinctions, you will likely get a distorted picture of the existence or extent of any compensation discrimination and employees may be included in the affected class that were not actually affected by any discrimination in the acquired company.
Maintaining an accurate picture of who is similarly situated in compensation can be challenging. Currently it is very hard to account for self de-selections, such as electing not to relocate, that may influence pay. However, as noted above there are several steps a contractor can take to avoid exacerbating the problem of determining who is similarly situated in the compensation area. These include not lumping together people you will want to distinguish for compensation analysis into a single title, coordinating and tailoring data that bear on compensation, having clear, objective pay policies and maintaining records distinguishing acquired employees from current employees until their pay has been fully and successfully integrated into your current pay systems.
I hope you find this helpful as you grapple with the area of compensation. Next month I will take up the topic of who is similarly situated in the areas of promotions and terminations.