Workplace Diversity: A Process Not a Problem

Workplace Diversity: A Strategic Process, Not a Problem

By Roger Philby, CEO of The Chemistry Group

As a talent strategy consultancy we are often asked about workplace diversity; why it matters and how to achieve it. And for the last decade, our response has been the same. On the ‘benefits’ of having a diverse workplace, there is no debate. It is imperative. The ‘benefits’ if you wish to frame it so, cannot be over-extolled.

To the other question, how to achieve it, first you must interrogate your intention, for diversity isn’t a problem looking to be solved, it’s existential. Genuine workplace diversity can neither be engineered from below or ‘delivered’ from above. Nor is it a task to be laid at the door of the talent department. Instead, it is a way of working, an operational mindset that should permeate every aspect of your business.

One of the most influential global D&I experts Asif Sadiq, presently Head of Equity & Inclusion at WarnerMedia has demonstrated repeatedly how embedding diversity and inclusion in all elements of a business results in ‘innovative, creative and inclusive products and services.’ He is also very clear that it is ‘a long-term road map.’ Those that don’t regard it as such and adopt a tick-box approach will find their impact to be little more than tokenistic.

At The Chemistry Group we must practice what we preach. Our founding principle, that we believe that everyone has the opportunity to be brilliant at work, has always resulted in Chemistry being a diverse workplace by default. And by using our hiring tools, diverse workforces are often the outcome of the processes we put into place with our clients. That is to say, we have gone to great lengths to ensure that the tools we use within these processes do not engrain bias within an organisation.

How do we ensure the fairness of a selection procedure? Our Head of Science, Diarmuid Harvey, explains:

An unfair selection procedure is one which is known to cause adverse impact to a protected minority group for reasons which have no relevance to the outcome you are trying to achieve. Only including job-relevant and performance-predicting factors supports the tool’s legitimacy in a selection context. However, at Chemistry we want to go further to ensure we are not disparately impacting any groups as we strongly believe that everyone should have the opportunity to be brilliant. Therefore, we employ a number of approaches to test for indicators of adverse impact- this can be done in several ways.

Our sampling approach and model builds will work to ensure any biases present in your existing population are highlighted and accounted for. By drawing on information derived from your incumbent population and our existing norm groups we then utilise a number of statistical methods to test the assessment for any expected differences between groups. Finally, we conduct regular post-launch monitoring of the tool’s performance. Any indication of adverse impact can be adjusted for in the assessment by tweaking aspects of the WGGL profile [our trademarked approach that establishes the benchmark for your talent], which means that only minimal impact to the tool’s performance will be experienced.

In respect of bias, we see it everywhere. Unintended bias usually operates at the unconscious level and is caused by social and cognitive factors. It can be demonstrated in hiring methods (through the biases of any hiring manager); performance appraisals (particularly those reliant on ‘softer’ metrics like manager ratings); salary determination, etc – essentially, and unsurprisingly, anything that has a human judgment or decision-making component.

Its consequences are far-reaching. Not only is a system allowed to take root that treats people disparately because of their immutable characteristics, it leads to a poorly defined work culture, a lack of productivity – or at least a lack of focus on the factors that truly drive performance – and a lack of diversity in the incumbent population. And it is this, the lack of focus and engagement with the factors that truly drive performance, that is so damaging. It’s why we continually hear the same gripes from companies, that they never quite feel they’ve nailed their hiring or got the right people in the right roles, and the reason why is because they have never truly gotten to grips – or analysed – what drives performance. Because if they had, the chances are they would have a very diverse workplace. It must be noted, however, that not all of these disparities are driven by determined (if unconscious) action on the part of the company (or its agents) and can be caused also by factors outside of the control of the company.

As an example, a lack of gender diversity in many STEM industries is likely in part due to immutable gender differences in interests, and therefore it isn’t always the case that absolute differences are indicative of an unintentional bias. A company in this industry is unlikely to have a balance between males and females across roles because of a self-selection bias acting at the attraction stage (where far fewer females will apply for these roles), and therefore contextualisation is very important here. However, disparate pay between the genders for individuals of similar experience and abilities is definitely an unfair bias that can be seen in these industries, potentially due to these environments being male-orientated and therefore rewarding characteristics more commonly seen in males.

We take care to ensure we are assessing for bias in many ways. When we design a sampling strategy we always want to ensure we have a representative sample, and that biases baked into some of our variables (e.g., performance ratings) are not being reproduced through our analysis and recommendations.

We can also test for biases which are less obvious, such as managers who are more punitive/lenient in their performance evaluations which can have consequences for individuals due to the characteristics of the assessor rather than the assessed. Through being constantly aware of the potential existence of these biases, and testing and accounting for them as much as we can – both in our modeling and our recommendations – we can ensure that these unintended (or otherwise) biases are not continually reproduced in our clients’ processes.

As leaders, we must do more. Actively review the processes within your organisation to discover barriers to advancement. Become cognizant of educational institutions that foster social justice and mobility. In the UK, for example, QMU London would be one such institution I’d be interested to interview graduate candidates from – why? Because they’ve recently been named the best university in the country for social mobility.

Reviewing your processes will lead to some very frank conversations, and potentially some uncomfortable revelations. But, crucial to creating an inclusive work culture is to foster a culture of belonging. Employees should be able to exchange stories of their lived experience and there is a wealth of evidence to prove that high-performing teams operate in safe spaces, where they feel they can offload to each other. It’s important that leaders and managers from every echelon partake in these conversations because workplace diversity is not an exercise or a training day here and there. It’s a way of working, a way of thinking, employed every day, every minute by everyone within an organisation.