Ok, so before we begin, I need to tell you I’m a self-confessed data geek and believe presenting it in a compelling way is a powerful motivator for change.
Let’s start with pay gap reporting, the limitations and evolution.
The biggest pay gaps
Here in the UK, the primary focus of assessing inequity so far is to report on the differences in pay between women and men. The UK Equal Pay Act was introduced in 1970, meaning a woman and a man in the same employment, performing the same work, must receive equal pay.
Marvellous – and the result of women campaigning hard.
Then in 2017, the British government introduced further legislation, Gender Pay Gap reporting. This requires all organisations with 250 or more employees to publicly report their own mean and median pay and bonus gaps, for all women and all men.
Before we go any further, here are some simple definitions:
- Mean – otherwise known as the average, is the sum of all the numbers, divided by the total number of numbers
- Median – midpoint of all numbers when stacked highest to lowest.
Across the UK the median hourly earnings reported for women and men showed a difference of 27.5% back in 1997, meaning the mid-point for women’s salaries was 27.5% lower than the mid-point for men’s salaries. The mid-point gap has steadily reduced to 14.9% in 2022.
It’s great to see pay gaps closing, but we need to go beyond gender, and I’m not convinced the way through that is to introduce separated reporting for ethnicity pay gaps, disability pay gaps etc.
When you consider socioeconomic status, often measured as a combination of education, income and occupation, you’ll find that along with all the women, and people from underrepresented groups who face barriers, there are White men who aren’t having a great time in life either. Maybe because they live in an area where education and infrastructure to provide jobs or transport to get to work, has been underinvested. But you would never find this by looking at gender or ethnicity pay gaps separately.
So, you need to look at pay gaps for all demographics with an ‘advantage lens’ to understand who benefits most and least. And consider the standard of living being offered.
Track CEO to average worker pay ratio
According to the World Inequality Report 2022, the richest 10% of the global population own 76% of all wealth. And The Economic Policy Institute found that CEOs at America’s largest public firms were paid 399 times as much as a typical worker in 2021. They also found that CEO pay had rocketed 1,460% since 1978. In contrast, compensation of the typical worker grew by just 18.1% from 1978 to 2021.
I encourage every organisation to report on their CEO to average worker pay ratio and use that insight to influence future wage policy.
Culture
To see if you have an inclusive culture, you need to find out if people from underrepresented groups are having the same good experience as people from overrepresented groups.
Disaggregate engagement scores
In engagement surveys you might be pleased to see an overall score of 83% and feel confident no action is needed. But to understand inclusion you need to break that 83% down by characteristic, eg ethnicity. You will likely find your largest population, let’s say White people, are having a better time and their responses have generated a rating of 84%. And the ratings generated from your smaller populations of Black, South Asian, Mixed-race people etc are only 59%, 62% and 65%. Their scores are being hidden because of the relative weighting of the largest population. And these scores paint a very different picture of how inclusive the organisation is.
Use better questions
Since 2020 there has been a surge in engagement survey providers adding diversity and inclusion questions to meet demands from clients. Trouble is, I’ve yet to see anything that’s adding real power, because the questions seek a rating of ‘perception’ rather than personal experience.
Instead of adding questions like ‘This organisation is committed to diversity and inclusion,’ or ‘People here are treated fairly regardless of religion, disability’ etc., use questions (you may already have) that describe the feeling of inclusion; being valued, heard and involved.
Questions should start with ‘I’ to centre the individual, for example:
- ‘I have good opportunities to learn’
- ‘If I’m treated unfairly, I’d be comfortable speaking up’
- ‘I have enough say in how to do my job’
Then when you get the results, break them down by demographic and repeat for each department if they’re big enough to protect anonymity. Leaders seeing gaps in scores for their area of the organisation generates personal awareness and responsibility for the differing experiences people may be having. I’ve seen gaps of over 20 points inspire leaders to create a much fairer and welcoming environment for their teams.
Diversity
Now let’s get into hiring, progression and retention.
Report on diversity at every level
Start by dividing your organisation into layers of hierarchy and report the percentages of each demographic for each layer. Then compare those percentages to the local, national or global population depending on the nature of your organisation. This will highlight any areas of over- and under-representation. And if you have large enough departments replicate the format for each team.
This one-page report can help guide your approach to setting targets, and it doesn’t need to be the same in every team. You might want more women in tech, operations and engineering, and focus on career progression for people of colour in HR, legal and marketing.
Report on the recruitment process
Recruiters often get a hard time for not bringing in enough diverse candidates, when data often shows it’s the decisions made by hiring managers that’s limiting progress. To see what’s happening, report on candidate demographics at the application, shortlist and interview stages.
You don’t necessarily need to report on every demographic, but if, for example, you’re looking to increase the mix of ethnicities, you should report on the ethnic diversity at each stage of the recruitment process.
Report on people process outcomes
The most powerful step in exploring day-to-day bias is to report on how the outcomes of all your usual people processes vary for different demographic groups.
At least once a year, you should look at all the oftenindividual decisions made across the organisation, and review the outcomes as a total organisation for:
- Performance ratings
- Pay increases
- Recognition awards
- Promotions
- Secondments
- Resignations
- Dismissal
- Redundancies
- Settlement agreements
This data can be confronting and provide valuable information to update process documentation and guide people to review for bias in the moment when conducting these activities in the future.
Collect good data
There are some universal challenges with diversity data, including incompleteness, trust and what information can be collected in different parts of the world. You’ll need to make sure you use data in accordance with local data protection legislation and protect people’s private information.
Here in the UK, we comply with the European General Data Protection Regulation (GDPR). When it landed in 2018 many organisations stopped collecting diversity data (uh-oh!) because they weren’t using it. The good news is that you can legally collect most data for the purpose of ‘monitoring equality’, so long as you tell people how you use it and have appropriate governance in place.
The level of insight you can obtain will depend on the systems and data insight capability your organisation has invested in. Your reporting is likely to start off basic (which helps to avoid overwhelm), and get more sophisticated as capability and confidence grow.
How to report with gaps in your data
Even when your data is not complete, you can report results for the ‘knowns’ alongside the ‘unknowns’ to get the full picture of people’s experience. So, at the highest level for disability you would report scores for people who:
- are living with a disability (known)
- have no disability (known)
- haven’t answered the question (unanswered)
- have responded ‘prefer not to say’ (prefer not to say)
If you have a high percentage of ‘prefer not to say’ responses, you’ll need to build trust that you’re using that data to monitor a fair experience at work.
No matter how good your data is, seek to understand who is benefitting the most and who is benefiting the least. This is how to truly report on inclusion and make sustainable progress so that the world of work works better for everyone.
Because unless you’re consciously including people you’re almost certainly unconsciously excluding people.
Catherine Garrod, pictured below, is the author of Conscious Inclusion: How to ‘do’ EDI one decision at a time and founder of Compelling Culture. She works with organisations to determine whether people from underrepresented groups are having the same good experience as people from overrepresented groups. The People Space will be running a more in-depth discussion with Catherine for our Work’s Not Working… Let’s Fix It! podcast next month
