Do we have a pay equity problem? Using data to find out.

Having a data analytics background, I often get caught being asked for help with accessing data or twist a PivotTable. Now, because I am a nice guy, I generally try to help. But more and more I am trying to help people really understand their problem rather than just explaining how to use the vlookup.

Let's work through a case study. 4 November was Equal Pay day (EPD). This is the day represents the difference between men’s average wage and women’s average wage. According to the EPD website, the current pay gap is 17.9%. This seems like a lot!

Our hypothetical company has a look and finds that they are at 12%. How do I find out whether this is problem? Or maybe if it is a symptom of something else?

We think out problem might be that “women get paid less than men”. 

First thing we want to find out is whether our 12% is a big number. And if it is, are some areas of the company worse than others. 

As we look through the different workforce segments, we find that this number reduces when you look at like-for-like roles. It doesn’t look so bad across the company. Especially compared to 17.9%. This analysis also allows us to more easily see and address individual issues. Does that spell the end of the problem?

Digging further, we see that 25% of our senior leaders are female. This may explain why the overall company looks worse than when we look at like-for-like roles - the higher wages in this male dominated space distorts the overall average. But maybe it uncovers another problem. Do we have role equity issues?

To find that out, we need to look at our talent pipelines. Is our recruitment and succession planning well balanced? Perhaps we are battling the legacy of our industry but are showing equity in our process? Maybe it is something systematic and we need to look at our culture? Lets have a look at work life balance. Do we see part time or flexible work practices being taken up across the board? Is it limited to team members or do leaders take it up? Perhaps our males are breaking traditional ‘rules’ and making it acceptable to be more flexible regardless of gender or the level in the organisation.

You can see that as we step through the problem and use data to test and tease our hypotheses, we can confirm whether we are not he right track. We can also use that this thinking to stretch the initial proposition to understand what is at the root of the issue. It may be something completely different to what we first thought.

Here are a few steps that you can use to unpack the data (other problem solving techniques not withstanding)

Identify and measure the problem

The first thing to do is to define the problem. Here we said women get paid less than men. We also can measure that by dividing the average female wage by the average male wage.  This is the first step of finding the right evidence to solve our problem. 

Is that a small number or a big number? Understand the metric

Once we have the measurement, we can then see if it is a problem. Our company’s result might be 12%, which might seem big, but when we look at the Australian result, it is not that bad.

It is also really important to understand the metric. In this case we use an average wage. Because of the way that wages are set, higher paying leaders can increase the average. If there leadership cohort has a gender imbalance, then the overall metric can be distorted by the nuance of the definition and the comparator groups.

Segment the data to understand it better

Looking at different workforce segments can tell a different story. Do different organisational units, or maybe geographical areas have different rates? these differences can hide or distort the overall metrics.

In this case, different pay grades will show much closer average wages because they are more similar. This doesn’t mean that it is not an issue, just that it isn’t as bad as it might seem.

What can other metrics tell us about the problem.

When you look at different metrics together, then can really add to the story. It might give some different context, or perhaps highlight a different problem altogether.

When we look at pay equity with the females in senior leader roles metric, we can find the reason why the overall pays look like a problem. The actual problem may be role equity, but it is shown out in pay equity. 

Refine the problem and measure that.

Once we have a new problem, we can go through a similar process to know whether this is the real problem, or just a symptom of something else.

In this case, we delved deeper to see if it was a cultural issue around the acceptance of balancing work and home. Here we found new metrics that can help us measure any transformation once we put a solution in place.