Adverse Impact Analysis and Reductions in Force: What Is It and How to Use It Effectively

Transcript

Alitia Faccone:

Welcome to Jackson Lewis' podcast, We Get Work™. Focused solely on workplace issues, it is our job to help employers develop proactive strategies, strong policies, and business oriented solutions to cultivate an engaged, stable and inclusive workforce. Our podcast identifies issues that influence and impact the workplace and its continuing evolution, and helps answer the question on every employer's mind, "How will my business be impacted?"

With leaps in innovation and an uncertain economy, employers are taking steps to right size their organizations. Life sciences companies similarly must contend with the need for workforce reductions amid economic shifts. On this episode of We Get Work™™, we discuss the relevance to life sciences companies of adverse impact analysis and reductions in force, the implications of not acting on the analysis, and how to determine the level of risk associated with a RIF.

Our hosts today are Peggy Strange, a Hartford office principal and co-leader of the firm's Life Sciences Group, and Matt Cammardella, a member of the Life Sciences Group and principal in the Long Island office.

Peggy believes that proactive innovation and support are the keys to a successful client relationship in a dynamic and ever-changing workplace. She looks for ways employers can achieve the same goals as their employees.

Matt assists employers with analyzing reductions in force for adverse impact and assessing the applicability of federal affirmative action laws in the context of mergers and acquisitions.

Peggy and Matt, the question on everyone's mind today is what is an adverse impact analysis and how does that impact my business?

Peggy Strange:

I am here today with Matt Cammardella, my friend and colleague, who is the person to go to for the tips and trends in using adverse impact analysis for decision-making.

So, Matt, you practice in the area of adverse impact analysis, which is way beyond my capability. You're a go-to for me and for a lot of our clients. Tell me a client calls and they say, "Look, we have to do a reduction in force, and in my last job we did this adverse impact analysis. Anyways, my boss wants to know what is this and should we do one?" How do you respond?

Matt Cammardella:

Sure. So basically an adverse impact analysis is just a comparison of selection rates between two different groups. In the employment context, the equal employment opportunity context, it's men and women, members of different racial groups, older workers and younger workers, et cetera. But all things being equal, given enough observations, we would expect to see those selection rates to basically balance out between any two groups, whether it's hiring rates, whether it's promotion rates, whether it's termination rates. So what the adverse impact analysis tells us is if those selection rates are close to what we would expect or whether they are so far apart that we would not expect such a large divergence to occur by chance.

So the analysis and the question about whether or not it should be done, yes. Anytime you have enough employment decisions, if you want to ensure that those employment decisions are being administered in a non-discriminatory and fair and equitable way, yes, you want to do this type of adverse impact analysis and evaluate your procedures.

Peggy Strange:

So yeah, there's technical reasons and there's general good practices. What, in your view, is the purpose of an adverse impact analysis? What are you trying to accomplish in this process?

Matt Cammardella:

Sure. Well, look, the technical purpose of the adverse impact analysis, as I described, is to see if there are meaningful differences in selection rates between two groups. But we can also use it to determine the level of risk associated with something like a reduction in force.

So, Peggy, you're an employment lawyer. This is going to come as no surprise to you. If a plaintiff's going to bring a claim, they're going to bring a claim of discrimination under one of two theories.

First, we have our disparate treatment theory. This is a claim that the employer made an employment decision on the basis of an individual's protected characteristic, whether it be race, gender, disability status, whatever the case may be. In order to prevail under this theory of discrimination, it requires the complaining party, the plaintiff, to prove intent to discriminate against them on the basis of that protected characteristic. So that's what that would require. An employer can, of course, defend against such a claim by providing a legitimate non-discriminatory reason for the employment decision. So that's disparate treatment.

But there's a second theory of discrimination under which a claim can be brought, and that's the disparate impact theory. Under this theory, the claim is that there is a facially neutral selection procedure that has the effect of impacting one group more severely than another.

So Peggy, think for a moment about something like a pre-employment test that involves evaluating an applicant's strength. That test in all likelihood over time would screen out a higher percentage of women than men, because as a general rule, men are bigger and stronger. Not always the case. I wouldn't want to arm wrestle you. But generally speaking, that's what we would observe.

So while the test itself has nothing to do with gender, it's not a gender test, the result of using that test certainly could be a disparate impact against women, but that's not the end of the inquiry. Somebody might be able to plead that case, they might be able to point to the selection procedure and demonstrate that it has a disparate impact, but the employer then can defend against such a claim by demonstrating that the selection procedure is job related, consistent with business necessity, and that it's been applied in a consistent manner.

So evidence of highly improbable differences in selection rates, or what we call statistically significant adverse impact, can support either one of these types of claims, and that's important to understand.

So first, statistics can be used to demonstrate that there was a pattern and practice of discrimination against a certain group by an employer. Typically, this type of statistical evidence would need to be accompanied by some other evidence demonstrating that this was the employer's standard operating procedure. In other words, be able to prove intent. Statistics generally on their own are not going to be sufficient to prove disparate treatment. However, these statistics can also be used to bolster a claim of disparate impact, and show the disparity in selection rates, thus kicking the burden of proof over to the employer to demonstrate that the selection process was job related and consistent with business necessity as well as consistently applied.

So by subjecting the company's selection decisions in a reduction in force to an adverse impact analysis, the employer really has the ability to assess the likelihood and strength of an initial claim that might be brought by a party who claims to be impacted by that procedure itself.

Peggy Strange:

So it's very helpful to have, hopefully. What is the best time... In my scenario, somebody comes to you and says, "I want an adverse impact analysis done," when during this reduction in force process should they consider that? At the beginning? At the end?

Matt Cammardella:

Yesterday. No. The sooner the better, Peggy. As I mentioned, the whole point of conducting the analysis is to assess risk. By performing the analysis, as soon as you get some preliminary decisions and in advance of notifying employees that they're going to be let go, allows the employer to review its selection procedures, review the selection decisions while there's still time to effectuate change. If you're too far downstream, you don't have the ability to change course. In which case, why did you do the analysis?

Peggy Strange:

And why would I do the analysis if everybody can see it? Is there any way to keep it confidential? Can we just stamp it confidential and then so be it?

Matt Cammardella:

Sort of, kind of, Peggy. So look, you're getting at whether or not we can keep the analysis privileged and confidential, and the short answer is probably yes. So if an employer engages counsel to direct the analysis for purposes of providing legal advice, well then the employer can benefit from the attorney-client privilege. And with that privilege, the company can avoid being forced to turn over the analysis results in litigation and avoiding that smoking gun evidence, so to speak, that a plaintiff is always looking for, especially if the employer knew about an issue and didn't address it.

Now, any discussion of privilege would be incomplete unless we threw in a caveat, which is no privilege is absolute. A judge can certainly decide that something is not going to be privileged if the equities of the case mandate that something be turned over. But an employer can help bolster its argument for privilege by taking some simple steps.

So ensuring, to your point, that all communications, all documentation is stamped with some sort of attorney-client privilege language or prepared at the direction of counsel language, something which shows the intent that these documents are meant to be confidential and privileged. The employer also should take steps to restrict access to the analysis to those who have an absolute need to know. This way you're not unknowingly waiving the privilege by sharing it in too broad a circle.

Peggy Strange:

Perfect. So now you're ready to run your adverse impact analysis. What type of information are you going to need to run this analysis?

Matt Cammardella:

Yeah. It's actually a surprisingly short list of things that you need, but we do need several data points. So the starting point is a baseline. That's a list of all of the employees in the workforce who could have been selected. Don't include individuals who were never under consideration. It's just those individuals who could have been selected for the layoff.

For that population, we then need certain additional pieces of information. We need demographic information. So we need their race and ethnicity. We need their gender. We need their age or their date of birth. In addition, we need to know who was selected, who wasn't. And, as we'll talk about in a bit, decisional units. What decisional units do these folks fall into?

Now, a couple of things to note here, Peggy, because there are a couple of issues that pretty regularly trip up clients.

First, missing demographic information. While most employers have date of birth for all of their employees, it's fairly common to come across situations where an employer is missing some race or ethnicity data or some gender data, because as we know, self-identification is voluntary when it comes to this information. So there are times when an employee will elect not to provide it. Now, those companies who are familiar with the EOM reports know that they need complete race and ethnic and gender data in order to report to the government every year. So how do you bridge that gap?

So in that situation, the employer really should be hazarding an educated guess, believe it or not. So a couple different ways to address that. One is through visual observation. One is through reviewing employment records that you have for other purposes. And as sort of a backup, at least on the gender side, oftentimes you can discern gender from the name, although not always.

Look, at the end of the day, there may be still some pockets where we are missing information. In those situations, what we do is we just remove them from the analysis set issue. So if we're missing the race information for a given employee, that person would be excluded from the race analysis that we do.

Peggy Strange:

Perfect. So you mentioned decisional unit. What is the decisional unit and how does that affect the adverse impact analysis?

Matt Cammardella:

So, Peggy, this is one of the most tricky issues for employers. They don't always get it right, and that's because your decisional unit for adverse impact purposes can vary and sometimes can differ from the lists of employees that might go on exhibits to releases in these reduction enforced situations. But at its core, at its essence, the decisional unit represents the most discreet group of employees that the employer looked at together for purposes of deciding who stays and who goes. The decisional unit typically aligns with an individual decision maker in that way.

So, for example, we may see a reduction in force that's targeted administrative support departments in an organization such as, let's say, accounts receivable, accounts payable, and customer service. Each of those departments has its own manager, and that manager was tasked with trimming 10% of his or her payroll costs in his or her department. So those managers then looked at their respective departments, selected those employees for a layoff to meet that goal. So, in those circumstances, each of those departments represents a discrete and separate decisional unit. So we would definitely run an analysis on each of them.

However, over and above those discrete decisional units, we're likely also to roll up all three of those departments into an aggregated sort of analysis, call it administrative support departments, and look at them in totality to see if any trends are emerging. It doesn't change the fundamental decisional unit, but it really is important for an employer to see from a higher level view whether or not any trends are emerging.

Remember, while the analysis itself may not be discoverable in litigation because of the privilege, the underlying data is always going to be discoverable. And one thing we know about plaintiff's attorneys is they're going to use the data in any way, shape, or form that advances their cause. So it's always good, it's always beneficial to perform any different types of analyses, all the different cuts that would allow you to get a perspective of what a plaintiff's attorney would see.

Peggy Strange:

So let's say a client tells you, "Look, we'll do this adverse impact analysis. It'll make me look kind of cool. My boss likes it. But we don't have enough time. We've got our decision making, our selection criteria. It's all done. So thank you for the adverse impact analysis, but we're probably not going to be making any changes to selection decision regardless of what you find in that analysis."

Matt Cammardella:

That's a tough one, Peggy. Look, as a general rule, we're not big fans of undertaking adverse impact analysis if you're not even going to consider the results. So we would generally advise against that because, look, what would be worse than in litigation through discovery it becomes revealed that the analysis was done and completely ignored. That's not a good look for the employer.

However, in all candor, there are some companies it's on their checklist to perform the analysis, they want to know what it says, even if they're not going to do anything about it. And so sometimes we'll do it just so they have a sense of what the numbers look like.

Peggy Strange:

All right, so Matt, sometimes employers will have a reduction in force that's being done in stages, over a certain period of time. What problems develop when you make rolling decisions over time? Can we still do this adverse impact analysis?

Matt Cammardella:

We can, but it certainly does complicate the analysis a bit. Because the review of individual employees is occurring multiple times over discrete periods, you have to tweak your analysis to account for that. But there are statistical techniques that we can deploy, and our statisticians, for example, will do them, that will basically aggregate the probability of being selected when you are looked at multiple times over that rolling period.

So, short answer is yes, you can control for that.

Peggy Strange:

Is there any chance that a reduction of force is simply too small for analysis?

Matt Cammardella:

Too small for statistical analysis? Absolutely. Remember, the adverse impact analysis is designed to identify trends rather than individual discriminatory employment decisions. So, in order to identify trends, you need a fair amount of data. Most statisticians will tell you, get me as much data as you possibly can. They'll tell me they generally want at least 30 or more observations, meaning selections, to get a robust analysis, but certainly we can do it down to 10 or some cases even smaller than that.

And look, while a RIF that only involves a handful of decisions won't necessarily allow you to perform trend analysis, it doesn't mean that you can't do observational analysis. So, for example, if a RIF only chose five people and all five of those folks are 60 or older, you probably want to take a closer look.

And one other call out here is, regardless of what the analysis says, it's always a best practice to look at higher risk decisions anyway, such as those affecting individuals with a disability, those who were on leave or recently took leave, or those who filed a complaint against an employer.

Peggy Strange:

Eyeball those decisions and kick the tires a little bit, right?

Here's the question that always comes up. The results of the adverse impact analysis, they're not good. They're bad. What does a company do? Should they change its decision to eliminate an indicator of adverse impact? What do you do with that information?

Matt Cammardella:

Yep. Yep, great question. So look, remember, if you have an indicator of adverse impact analysis, it does not mean that discrimination has taken place. Adverse impact indicators can be rebutted with the right evidence, as I described earlier.

However, when we see adverse impact, we do want employers to go back and look at the decisions. Go back to the decision makers, review the process by which the decisions were reached. Does everything seem on the up and up? And if so, is it well-documented? Can we produce evidence if we need to? If not, maybe the decisions need to be revisited.

But look, we get this question all the time. A client says, "Hey, I see adverse impact here. How many decisions do we have to change to fix this?" And that's the wrong way to look at it. We only recommend changing selection decisions if after doing that review that I described, we find that we didn't seem to adhere to our process, or there's a lack of documentation or something just doesn't add up.

To make changes purely based on the statistics will subject you to double jeopardy. One, you undermine the whole process by which you were making selections. So if somebody claims that they were discriminated against on the basis of their protected characteristic, well, you know what? You're going to undermine your business rationale for what you've done. And on the other hand, if you've changed decisions, somebody's going to say, "Hey, you know what? I wasn't even part of that impacted group, but I got negatively impacted because I got added to the list when you saw the results of those analyses."

So long story short, don't change decisions based purely on the statistical analysis.

Peggy Strange:

So, Matt, you and I spend a lot of our time in the life science world, and what are some of those unique issues that life science clients see?

Matt Cammardella:

Well, one issue that I see arise pretty frequently, actually, I'm dealing with this issue right now, and it involves oftentimes life science clients who, I don't know, let's say, for example, it's a pharma company and they are reallocating sales and marketing resources from one drug to another. One is sort of being faded out, one is being ramped up. Oftentimes in those circumstances, what we see is that the employer will select folks for elimination, but then reemploy them or redeploy them to roles supporting the new drug, but not everyone.

And if you think about that, that involves two discreet selection decisions; the selection decision to terminate, and then the selection decision to retain from that group of folks who were originally selected for elimination. We can analyze both of those selection decisions, both the negative and the positive.

So that's a unique issue that we see arise pretty commonly in the life sciences space.

Peggy Strange:

Well Matt, thank you very much. This is fascinating, and certainly analysis, statistics, they can tell us a lot. So thank you so much for your time.

Matt Cammardella:

My pleasure.

Alitia Faccone:

Thank you for joining us on We Get Work™. Please tune into our next program where we will continue to tell you not only what's legal, but what is effective. We Get Work™ is available to stream and subscribe on Apple Podcasts, Google Podcasts, Libson, Pandora, SoundCloud, Spotify, Stitcher, and YouTube. For more information on today's topic, our presenters and other Jackson Lewis resources, visit jacksonlewis.com.

As a reminder, this material is provided for informational purposes only. It is not intended to constitute legal advice, nor does it create a client/lawyer relationship between Jackson Lewis and any recipient.