Reading Between The Lines: data and under-representation in TV

Despite having more data available than ever before - understanding the make-up of the UK TV sector’s workforce is actually still quite difficult.

This is for several reasons: different data sets have historically used different terminology; response rates across the UK and from different genres vary; and the language we use to describe and categorise identities is constantly evolving.

For example, the shift from using the acronym B.A.M.E to ‘Ethnically Diverse’ has sparked questions about which ethnic groups are/aren’t included in this particular category of under-representation.

And that’s even before we start to look at things like intersectionality - i.e. how a combination of factors like disability, gender and ethnicity might impact on someone’s access to and progress in the industry.

For the last seven years, the Creative Diversity Network’s Diamond system - the first of its kind - has helped us collect data at scale and spot some potential trends.

As a result broadcasters have refined their commissioning diversity guidelines to prioritise where they think the TV industry should focus its efforts in achieving fair and consistent representation.

Key areas that broadcasters have focused on for off-screen include:

• The lack of representation of people from the Global Ethnic Majority - especially in senior, decision making roles

• The lack of representation of Disabled people across all roles

• Specific genre roles where gender, ethnicity and disability representation is very narrow - i.e. in Scripted it’s writers

• Representation of people from less advantaged socio-economic backgrounds
  

However data sets and surveys from Diamond and organisations like ScreenSkills, Directors UK, Ofcom and BECTU have also highlighted some of the complexities and nuances around representation - how it’s measured and how we interpret the data.

One of the biggest challenges we face is what happens when we ‘lump’ a mix of similar identities into one category - and interpret progress based on the whole group rather than its individual components.

The challenge of grouping data - LGB as an example

Overall the industry data shows that Lesbian, Gay and Bisexual (LGB) individuals as a group are now strongly represented both on and off-screen.

But when we look more closely at the figures - we see that gay men make up the vast majority of the LGB workforce in TV - they are ‘over-represented’ in fact - while there is significant under-representation of lesbian and bisexual individuals.

When you add into the mix the most recent Census data for England and Wales, the under-representation of lesbian and bisexual females is even more striking. The 2021 Census suggests that lesbian or bisexual females actually make up a greater percentage of the population than gay or bisexual men.

The power and evolution of language and meaning

Another area where we need better and more consistent effort concerns the trans community. The decision to de-couple the ‘T’ from LGB when collecting data has hugely helped us to understand trans representation (or lack of) on an off-screen.

As a result, Diamond has collected specific and more detailed data about trans people working in TV. However the responses have often needed to be redacted, as the sample groups have been too small for the data to be published without compromising people’s anonymity. What we can conclude from Diamond is that we are not attracting or retaining many people at all who identify as trans in our workforce.

Could the language we use and the culture we operate in be a significant part of the problem? It’s currently Pride Month, and arguably the strongest message being shared amid the rainbow flags and supportive hashtags, is that we do not live in a trans-inclusive society.

Recently at Gritty Talent we decided to remove from our inclusion survey tool the ‘tick box’ question asking if someone was a ‘trans male’ or ‘trans female’, in order to act on the negative language and stigma around trans identities that often presents itself ‘neutral’ in these types of mass communications.

There is no consensus yet on the best way to describe and analyse data from this under-represented group. But we followed the advice from Stonewall to include questions around whether someone is living as the same or a different gender to the one they were assigned at birth, in order to make the data collecting more sensitive and real-world.

Even this is far from perfect - we recognise that the NHS and many trans people prefer the term ‘gender affirmation’ to ‘gender reassignment’. Despite this - the Equalities Act 2010 - which sets out the protected characteristics and makes discrimination unlawful - still uses the term ‘gender reassignment’. This in a way speaks to our point: the law, public bodies, official documents and data sets often feel out of step with the march of societal progress.

Outside of the TV industry we also see challenges in the way public bodies collect and categorise top level data that might help us understand under-representation. I.e. there is a lack of specific census data on the working age population of lesbian, gay, and bisexual individuals. Also in England and Wales the governments collect data on multiple gender identities - but Scotland and Northern Ireland do not - they opt for the binary Male or Female.

The result - if we want to compare gender or LGB stats from the TV workforce to the most recent UK working population datasets - we are going to struggle to make the words and numbers align.

Race, ethnicity and roles in TV

Similar inconsistencies and gaps exist when trying to understand data about the TV industry’s Global Ethnic Majority talent. Again we have to exercise caution in presuming that everyone included under these categories have had the same or similar experiences, and/or face the same barriers.

Take the Global Ethnic Majority - we know that this group is made up of multiple non-white ethnicities, who’s heritage, culture and experiences are all different and varied.

On first glance the data from Diamond Data might indicate that people belonging to the GEM are strongly represented on-screen (20.9% vs 14% of the working population) and fairly well represented off-screen (12.9%). However once you break down this category, it becomes clear that there is still considerable work to do - with most data from East Asian groups needing to be redacted due again to the low numbers in the response group.

We also see that people from Black, South Asian and East Asian groups are represented nowhere near the UK working population figures, and this is acutely the case when you cross-cut the data to examine Senior vs Non-senior roles.

The most prominent is that Global Ethnic Majority talent seems to be clustered in Non-Senior roles. People from mixed-ethnicity groups appear to have slightly better representation - but as a whole, people from GEM groups are not well represented in decision making and management roles. This has a huge implication for culture and inclusion - if the voices and perspectives of around 18% percent of the UK population are not around the boardroom table.

We also need to ask ourselves difficult questions about the unconscious racial biases and prevalence of harmful stereotypes that might be blocking the progress of talent with darker skin tones.

Evolving our understanding, fine tuning our data

So how do we as a sector start to collect better data sets and act more effectively on the patterns they reveal?

It’s done in small, closely monitored steps.

Diamond has committed to disaggregating race and ethnicity data in future reports where possible, so that we might understand how people from different racial and ethnic groups are represented.

At Gritty Talent, we’ve recently conducted what we believe is a first of its kind R&D project to ‘stress test’ the current diversity questions that are asked of people in Equal Opportunities Monitoring Forms and EDI surveys - in order to come up with a best practice set of questions and descriptions which reflect the multicultural, multi-identity and emotionally intelligent society that we are. It’s early days, but having had a sneak peak at the revised wording - it’s exciting to see what can be achieved when you ask a critical mass of people to tell you how they wish to identify and be described. But we’re just one organisation. It seems clear that to create a truly inclusive and representative TV industry, we must all level up.

Level up how we collect data; how we categorise less-advantaged groups; how we perceive and react to these protected characteristics. By looking at the terminology we use, and considering the implications of the existing datasets we have at our finger-tips, we can make better decisions about the data we collect in the future, and the way we’ll use it to drive representation forward.

And once we learn from trying out different ways to describe, delineate and act on data - we need to take these learnings and apply them to areas where we’re still at square one. I’ve deliberately not examined disability or socioeconomic groups in this article because these categories present their own unique and nuanced set of challenges regarding language, stigma and accurate data collection. But we know that the sector is working on it. And for now, a work in progress is better than nothing - and proof that the commitment is there to be better at representing and including everyone.