Fair360, formerly DiversityInc’s Chris Parker, director of research and data analytics, delivered a talk on how companies can measure their human capital metrics to understand whether their strategies are producing the results they are aiming for. He addressed the guests before the unveiling of the 2020 Fair360, formerly DiversityInc Top 50 Companies for Diversity list.
Diversity and inclusion make organizations more creative and innovative; better at retaining top talent; and better at anticipating and responding to customer needs. Parker said for companies to address whether their D&I strategies are yielding these results, they should examine how well their talent pipelines are working and how they stack up against competitors. All of these questions can be answered by assessing human capital metrics and benchmarking data that Fair360, formerly DiversityInc provides.
To illustrate his points, Parker used data that companies submitted for this year’s Top 50 competition. He took data from companies that performed in the top and bottom 25% in terms of their overall human capital management (not in terms of the official Top 50 Companies for Diversity list). To measure human capital management, Fair360, formerly DiversityInc relies on data regarding the racial and gender makeup of all workforce levels and programs within an organization.
“We’re getting a huge amount of data, and all of that together is being used to identify the companies that are best in terms of representation, and in the bottom 25%, with respect to representation,” Parker said.
The data Parker presented addressed men of color, women of color, white men and white women in respect to representation ratios and percentages in management. He said Fair360, formerly DiversityInc also looks at those metrics within talent management processes such as hiring, promotions and losses.
“When we look at the percentage of white men, we see a clear difference for a higher percentage within the bottom 25% relative to the top 25% … And that leaves a bigger slice of that representation pie, if you will, to be shared among people of color and white women.”
The presence of a benchmark in assessing these percentages is crucial.
“It almost means nothing if you just look at one set of these data, for example, in isolation,” Parker said. “I don’t really know whether the data I’m seeing for the top 25% is really good, or is there a lot of room for improvement? How does it stack up?”
To help compare, companies can relate their data back to population or workforce demographics. Another benchmark a company can use in its assessment of representation in management is representation within its own workforce. In other words, does the management reflect the diversity in lower levels of the organization?
“We could look at what percentage of a particular group is in overall management and look at that as a ratio compared to what percentage is in the overall workforce,” Parker said.
“Ideally, a company’s management would be perfectly represented in the workforce [with a] ratio of 1-to-1. However, as the workforce becomes more diverse, looking at the ratio between management and the rest of the workforce can make it look as if a company is failing, when indeed it is making progress.”
“You might have greater overall diversity, but that’s just not being reflected in your management yet,” Parker said.
The analysis then needs to delve into how well diverse management talent pipelines are working. Once a company assesses its data and areas to work on, it needs to assess how it brings different people from different talent pools into leadership positions.
To begin, a company should assess its representation of new hires and internal promotions into management positions, as well as representation in attrition.
For example, in the bottom 25% of companies, about 43% of external hires were white males. Companies in the top 25% of the data set excel in hiring and promoting people of color and white women, compared to the bottom 25%.
However, unexpectedly in the top 25%, there were actually more people of color and white women leaving.
“We’d certainly like to understand why. Is there something going on within the organization in terms of issues related to perceptions of unfairness or feeling lower levels of inclusion or belonging?” Parker said. “Or, possibly, is it a result of just increased demographics? If I have a … more diverse pool of management in the first place, it makes sense that I would be losing talent from those groups in higher proportions.”
Ratios are also key to answering questions like these. In the top 25%, the ratios of turnover are close to one, indicating that the representation of the losses of those groups are proportional to the overall representation in management. In the bottom 25% of companies, there is higher attrition of women of color compared to that top 25% benchmark.
Putting all talent processes — new hires, promotions and regrettable losses — it becomes clear that while both groups are promoting and hiring more women of color into management, companies in the bottom 25% are also losing women of color to a higher degree than the top 25%.
Working to rehire and promote these people is estimated to cost 200 times greater than the salaries of that group, so focusing on what is driving regretted losses and working to mitigate it is in companies’ best interests.
“It really makes no sense to emphasize hiring and promoting women of color [only to] then just let them walk out the door because you’re not treating them well or because there are other issues in the workplace that are driving lower levels of inclusion or belonging,” Parker said.
In the end, Parker said it is important for companies to measure their own human capital metrics against either internal or external benchmarks. Looking at hiring, promotions and attrition, as well as a robust and diverse pipeline can help a company assess the strength of its human capital practices. Deep diving into certain areas to better understand what’s behind the numbers can help companies understand their diversity strengths and weaknesses holistically.
