Exploring Pay Equity: More Insights from Dr. Brian Marentette

And we are starting our part two of our conversation with Brian about pay equity. Guys, do you mind if I just go ahead and just jump right in? Brian Marentette, PhD: Let's do it So Brian, We talked a lot about pay equity analysis last time, some basic principles. And as I thought about this, I have a couple questions.

If there seems to be perhaps two different ways that, that one might conduct a pay equity analysis in their organization. They could either do it software driven. Right, or they could hire a third party consultant. Can you talk a little bit about that? Brian Marentette, PhD: Sure, so I am on the consultant side of things, so I will try not to be too biased in my discussion of this, but we too use software. We all have to use software at the end of the day.

And ultimately those pieces of software are really just tools that allow us to run the analysis. Now, within the last several years several companies have made the software more user friendly, more available to, HR analysts and compensation analysts to be able to do some of these pay equity types of analyses. by themselves, which is great. The more that you can introduce this concept to the world and have more of it going on, the better, right?

We're all looking to address pay equity and make it more equitable and fair. Now there's obviously some caveats there.

What, what comes with, Software typically is limited support, right? You do off offer customer service and different technical support resources and things. But the level of support that, we provide in on the consulting side is pretty extensive.

We have gone through years of schooling and years of research and experience on how to do these analyses and how to overcome obstacles and challenges and resolve data issues that, if you're just using software and have not been trained in this space are going to be brick walls for you. It's going to be extremely difficult. So to an experienced analyst that has been doing pay equity work that. Maybe understands the fundamental steps that are involved and has the ability to work with the data.

Certainly, they could be using software and doing it effectively. And that's great to somebody that is new to the space has not been trained in statistics or, understands how to set up statistical models for analyzing compensation. Generally, that's where, consulting and the service side is going to be more beneficial. There's a lot of complexity and nuance in some of the data that we use and how we set up the analysis that again, can be confusing.

Multiple routes that you can take and how do you decide which way to set this up and how to run it? And then most importantly on the backend. Once you run everything, you get your results. What does it all mean? How do you interpret these findings? What is it really telling you? Cause it's easy to just set something up with a dataset, hit run, get your results, and then start putting money towards potential problem areas that you've got.

And in our experience, we've seen a few clients that have gone from using software internally to then coming to us.

When they've got a claim or they're being audited by the department of labor Saying we're being told we have pay equity problems and we've been doing it for years and we don't see anything in our results Okay, let us give it a shot and we run it and then we see a completely different story than what they've told us so yeah to the experienced user and analyst of course using software would be a viable route for everybody else They might find more benefit in a Consultant

relationships that guides them through it. The personal example that immediately popped into my head was trying to do my income taxes. Do I use this software? And I go, Oh, I have no idea what that question means. Or do I go to an accountant who understands the law? Brian Marentette, PhD: Yeah, exactly. Yeah. And I, for years, I was able to do my own taxes because like back in grad school, I was broke. I didn't make any money. I didn't have anything to really do.

So yeah, like a smaller employer. And we had this example earlier, which was, if you have 10 employees, can you really be doing pay equity? Sure. And guess what? You don't need software. You won't even open a software program. You might look at Excel once to see what everybody's making, but in that situation, you're gonna be doing what's called like a cohort analysis, which is literally looking at each person and saying. What are they making? Okay. How does that job compare to this other person?

Because all 10 of your employees are probably doing different things. Do we value that work that much more than this? And does that person have that much more experience that we should be paying them that much higher? And it's really like a pretty manual review. That's considered a pay equity analysis. Now it's on a very small sample or small head count, but yeah.

When you get to be a much more complicated organization, I can promise you, Companies like Google or Microsoft, they are not doing their pay equity analysis through an HR analyst, internally on software, things get much more complicated. I'm sure. I would think too, that having a third party consultant involved means a less of a likely that you're going to be biased in your own findings. Brian Marentette, PhD: Potentially.

Yeah. And I think, yeah, there's a potential there for somebody internally to maybe steer away, even unintentionally steering away from setting an analysis up that might produce, disparity or a red flag in their findings. But I think the, Probably one of the larger benefits of having it done externally is the ties into the transparency piece. Hey, we're having our books audited by an external firm.

They're coming in, they're running the analysis, they're telling us what the problems are, they're telling us how much we need to adjust salaries to resolve them. It's not done internally. This is all done externally and they have no vested interest in how we do it. We operate. So

Great question there on the point with legal counsel and privilege because a lot of what, pay equity analysis, even if you're not doing it for legal purposes, even if you're on more of a proactive side of the house, you're going to be surfacing. What we refer to as statistically significant differences. When we do these pay equity analyses, we're looking for differences that are large enough to be meaningful, not just due to random chance.

So of course there's going to be differences along, gender or race lines and things, but are they big enough to really be meaningful? Do we know that they're not just due to random occurrence and. Through that exercise, you are producing results that could be used against you in the court of law. If you're not doing the work under attorney client privilege could be sued.

Any employer could be sued by an individual employee or a class action, a group of employees and whatever analyses you've done internally, if they are not protected. They can say, Oh, let's bring those in as evidence. You found disparities. You didn't address all of them. And you let that work, go on you're in a bad position there as an employer. Do you get hired by attorneys then?

How does that relationship get triggered such that some information is attorney client privileged or can in house counsel hire you? And now all of a sudden, that, that sort of veil is in place. I just don't, I don't know anything about that. Brian Marentette, PhD: I'm not an attorney. I don't pretend to be one on TV either. And or podcasts or anything. Yeah. But it, there's some gray area there.

I'd say the safest route is using external legal counsel who coordinates through internal counsel, external engages us to do the analysis, advise external counsel so that they can advise the company on interesting legal system. That's the safest route. I do remember being in a conversation with you and I don't know what it was a year ago, maybe. And so this is going back to the soft software versus the consultant driven. And I hope I. I characterize this question correctly.

And if not, if I don't, hopefully you understand what I'm referring to. But I remember that you were suggesting that there can oftentimes be a tendency within The results that are given by the software program to have this big series of overcorrections to the left or, up and down within classifications or positions that actually end up causing more problems.

And that's another reason why you might want to have a consultant engage and look at the results, whether they're software results from you. Your program or software results from some platform. Did I say that anywhere near correctly? Brian Marentette, PhD: Yeah, I think so. Yeah. Yeah. Yeah. No, you're on track. And it goes back to setting up your analysis.

Of course, if you analyze everything by job title if you're looking at individual jobs, see a gap and it's controlling for relevant factors, you can feel pretty confident that you need to address that. Now, not every job title has enough people to be able to analyze them as a standalone group. You need lots of head counts really to run these like 30 employees in a particular grouping. So if you don't hit that that group doesn't get analyzed.

Now to address that people doing this on their own might start to aggregate groups. Of jobs together. They might even run it company wide and maybe try to control for things like the department that somebody is in or, the job group that they're in or your job level. And so you get these big, we call them like big models, right? There are hundreds, maybe thousands of people in them and you run it and you find, gosh, we have a gap of 1. 2 million. Females are impacted by 1. 2 million.

Now. They didn't set that up correctly. And so they take 1. 2 million to close that gap and they distribute it to these females. And then they run it again and it looks good. But what they didn't realize is that for all these jobs that are within that, everybody's pretty comparable. But they didn't really control for the job that somebody's doing. And so they give all this money to people in all these jobs that are female. And then now when you look at it by job title.

Females are, you get that huge overcorrection, right? So in all these jobs, now females are making more than their male counterparts. And you've basically created a pain, another problem, inequity. So yeah, you went from probably having not very much of an issue to now you actually have a major issue that's swung the other direction. And so that's where the complexity, can lead people down the wrong path without them really even realizing it.

You can run all these analyses in a software program and it'll tell you how much money you need. Get that bad result to go away, but that bad result is due to a faulty analysis, really. And again, that's in our experience when we get called in. It's because that exact thing has happened. Interesting. Thank you.

Even if it's a consultant coming in The goal of the pay equity analysis is to really reflect how they make pay decisions. So ensuring that your statistical model reflects the factors that you use to set pay. So tenure, maybe the time of the job, maybe performance, or if you have prior experience all those things that you use internally as a comp team to set pay need to be factored into your analysis. If you leave some of those out you're really, you've got an incomplete picture.

And that can also lead to that problem of, kind of like, flagging false positives. Really of, red hotspots in your results. And so making sure that you are, you understand your pay decisions, how do you set pay and then making sure you've got data that reflects that. And if you don't have data in your HRS, that ties to some of those things like prior experience be ready to go, pull your sleeves up when you find problem areas. And you think it might be due to prior experience.

You got to go look and maybe pull resumes potentially and see if prior experience plays a factor there. So that's one piece is making sure you understand your compensation practice so that the analysis that's intended to account for that can really do that. The other thing is knowing your pay variables. Like what do you want to analyze? Those different types of pay, your base salary of courses is probably the place to start. That's the most important really it's hardest to change.

It's the the largest sum of money, usually that's the place to start. But then looking at other forms of pay and do you know what influences those elements of pay. So if you want to look at like overtime do you really want to look at overtime? Is there any opportunity for discretion or bias to enter the equation of who's earning overtime? Government agencies might look at it, but from a, more of a pay equity standpoint, maybe over time is not as much of a concern potentially.

Bonuses, if you have bonuses that are tied to a formulaic kind of bonus plan, okay it's going to be the result of the formula that you've instituted versus discretionary bonuses. Maybe we want to look really more at just those end of year discretionary lump sums. So that's the other thing is understanding your pay and which forms of pay you want to analyze and how should they be analyzed what factors influence them. And the last piece is the grouping of how do you want to group your employees?

Like I mentioned with job title, that's the most straightforward, it's the easiest. To look at everybody within the same job should be making roughly the same amount. But then are there other what we would refer to as pay analysis groups that you could look at, and then how do you interpret those outcomes as well? So maybe you want to analyze all employees within a single pay grade, right? You have grades one through 12. Maybe we use pay grade as our means of grouping. That. legal concerns.

It's outside of the Title VII Civil Rights Act framework, but it will tell you another look at whether you might have differences that, again, are falling on race or gender lines and that'll include larger groups of people. Almost all of your pay grades are going to have at least 30 people that you can use to run the analysis. So that's one set. I'd say that's the starting place for a pay equity analysis. Those are things you need to pay attention to of, how do you set up your model?

What pay are you going to look at? How are you going to group your employees? Now within each of those, there's other more nuanced. Discussion points and things that should be worked out with you and the consultant or within the software. But those are the three big categories to pay attention to. So on that third question, how do you want to group your employees? How do you even decide that? Brian Marentette, PhD: Oh, good question there.

Questions that are being asked, what's the, what are you seeking out of this pay equity analysis? Is it strictly a legal review? Okay, we can limit it to job title. No need to look any further. That's where really the legal standard would stop. Typically more of a DNI lens. Okay. Maybe we go and run it by departments and see that I talked about the pay gap, let's look at it without controlling for levels, not factoring in what job level somebody's at.

Just look at it by department, what departments are showing. the largest spread there, the largest gap after we factor in performance and some of these other variables. And so it's really incumbent upon the client or the end user of the pay equity analysis to define what do we want out of this? If it's really just making sure at a, job title. We have equal pay for equal work. That's great. If you want to know some of those other broader questions do we have talent acquisition issues?

Do we have an uneven distribution of men and women throughout our hierarchy or our pay pay grades, then we can look at broader lenses and that will certainly open up that can for you.

The daycare workers were mostly women and you were contending that, the, maybe the more difficult job was the daycare center operator. And so they were clearly not similar job titles. Yet they. When you compare them and maybe you could make a case for one of them being, a more difficult, more challenging job. And the women who predominantly worked there were getting paid less versus the other one that was on the other end of the spectrum.

That's clearly something that human has to be able to sort out. How do you even start diving into something like that? Because that seems it seems like finding pay equity issues within job titles is going to be relatively easy. Identifying these is going to be much more difficult and probably the bigger problem. Brian Marentette, PhD: Yeah.

Yeah. So a lot of it from our perspective, like we can assist clients with that if they have what we refer to as a job architecture in place. Job architecture, meaning, they've got more than just job titles and like divisions, or departments, they might have job families. They might have job levels like individual contributor one, two, three. Supervisor one, two, three manager, one, two, three, et cetera.

And when you have that, now you can start to look at Hey, we know this job family, IT and finance are our two highest paid job family. They have the largest market demand, highest salaries. We can maybe we can analyze them together and look for our our P1s, our supervisor ones compared to each other. And we just put them all in together and we analyze it. Just by job level with this grouping.

And then you can start to see jobs that, again, if they're at the same level and they're in like similar complexity job families that's where you can start to highlight some of that. And again, from our seat as a consultant, we can look at the job architecture and say this is how you potentially could start combining things and let's run it. We'll see. How do things look if we see wildly disparate results, then it's okay, let's step that back and maybe break this out a little bit more.

But it's an iterative process and working with the client on, okay, using the data you've got, how could we potentially run this? And let's give it a shot. Or you can leave it incumbent upon the organization. If they don't have that documented architecture in place, it's incumbent on them to say okay, we know these jobs are all within like an entry level contributor role. We could probably start looking at them together over here. Yeah. That sort of thing. It's a lot more work.

It occurs to me that with the, everybody's got a human capital management system now that's in place. Most every organization who's got any sizable number of employees.

And it's, it occurs to me that potentially those could be used to add certain amounts of metadata to the structure so that you did, now you would group entry level service, entry level professional, and you could add areas of expertise or degrees of expertise or degrees of education or degrees of experience, and you could start looking at. All sorts of job titles along those sort of lines, rather than, some of the more traditional job, family, job title kinds of aspects.

Is that anything, has anybody ever tried to really slice and dice HR differently, because everything's all database now? Or would that be helpful? Brian Marentette, PhD: Yeah, absolutely. I think some providers are doing that within the HRIS, they have some of that architecture built into it so that as you're structuring, as you're integrating into that database you have the ability to map jobs to certain levels and that sort of thing. And we're always advocates of data and management of data.

Yeah. Organization database, the better and easier it will be for us as consultants or as your analysts to yeah, be able to parse out and explain things that otherwise require the human to go in and qualitatively review. That was a bit of a rabbit trail. I'm sorry about that, but it just, it did seem to so maybe make sense and it could be something that could, maybe potentially solve some analysis issues or make some analyses easier over time and in your space.

We're all here because of the computer and because of the database, all being ever present in HR. If that hadn't happened, back in the early nineties, it wouldn't have created all of these aspects of this cottage industry that have come about your work, even your service work, our software work, it's all there because of that. So it's a fun, exciting territory. Brian Marentette, PhD: Yeah. Yeah. I don't know how the world existed without computers, reams of paper. Brian Marentette, PhD: Yah

If I, or an HR professional wanted to learn more about pay equity just to be aware of what laws to pay attention to, just being able to recognize a fair pay structure. What resources would you suggest that they go to? Brian Marentette, PhD: Gosh there's a lot on the EEOC's website I don't have it handy to share to you, but there's really, gosh, at a federal level, it's going to be your civil rights act, 1964, Title VII. There's also the Equal Pay Act.

You have a state level, a number of different states particularly California and Illinois currently. That require you as an employer to submit employee level compensation data to the States for regulatory purposes. The best place to go is going to be our website, frankly, and I will follow up with you to give you the pertinent information.

There's just a lot of information out there and, there's some associations, obviously things like SHRM and World at Work, which is a compensation specific organization but really there's again, probably too much information out there when it comes to pay equity in that it's it's not like a relatively new field by any means. We've been looking at compensation for decades.

But in, in terms of how people go about doing pay equity analyses and, how they're approaching it, even companies that are based in Europe come over to the U S and offer pay equity services. It's fundamentally different than how we might operate as more of a U S centric, EEOC compliance.

So I'd say, first of all, maybe consider contacting legal counsel to ask them what legal requirements they have given the state that they're in or how large their organization is, whether or not they have a contract with the federal government to do business.

If they are, under the Department of Labor and the Office of Federal Contract Compliance Programs, the OFCCP there's a number of factors that might influence your decision of how you might go about doing a pay equity analysis given who you are. And of course, a conversation with me, I can work through that with you if you shoot me an email.

It's been really fun to have you on here and to give us some basic information about what pay equity looks like and did a few deep dives. And so Mike, any other final thoughts you'd like to share?

And so it's been a really great time and look forward to, doing it again sometime down the road. So thank you very much for being here with us very much. Appreciate it. Brian Marentette, PhD: Thank you. I appreciate the kind words and it's been a pleasure speaking with both of you. I'm happy to come back anytime I was going to say, check me out on LinkedIn. Jenny, you are reading my mind.