Exploring the impact of data and data-centric technologies on Black communities in the UK
Black communities in the UK are diverse and muli-faceted, so too are their experiences of data-centric digital technologies. That’s why we wanted our exploration of the impact of data and technology on the lives of Black people here to be wide and anchored in real-life experience.
In adopting this approach, we wanted to move away from the high-level aggregate view (peppered with a few individual stories often set in the US) that is often applied to discussions about data-centric technologies. Our suspicion was that the latter tends to present the impact of these new technologies in fairly abstract terms which leave people feeling vaguely unsettled but not empowered to do anything about it.
We also felt that this aggregate view fails to account for the fragmented reality that many of the current crop of data-centric technologies create. In other words, different groups of people experience different versions of reality online. This doesn’t mean that there aren’t commonalities across experiences. However, it does mean the ‘aggregate’ view doesn’t really capture most people’s experience of engaging with data-centric technologies. We suspected that only by first understanding the variety of experiences can we talk about data and technology in terms that most people relate to, can we begin to point out these commonalities. And once that becomes clearer, it becomes easier to see where interests align across groups, and to support group-led efforts for collective action.
For all these reasons, we tried to design our exploration of data and data-centric technologies in ways that are relevant to the lives of the practitioners, organisation and community leaders and others that we were hoping to work with. Findings from our roundtable conversations can be viewed in this presentation and you can learn more about the principles that guided our efforts and what we were hoping to achieve by reading our blog post titled, Who we are, what we value, and the change we want to see.
Now that we’ve explained our approach to, and rationale for this work, let’s outline some of the lessons we learned.
Many of the issues with data-centric technologies are structural
Over the course of the project, we got the chance to have a lot of conversations. It soon became apparent that many of the factors shaping Black communities’ experience of data and data-centric digital technologies and their ability to address emergent issues, are structural. We felt it was important to share the lessons related to these issues with funders and research institutions who have the structural power to do something about them. So we wrote a blog post about some of those lessons:
Recommendations: How to support rather than extract from marginalised communities.
Lessons for the community of impact we’re helping to build
But we are not without agency and there are lessons we can apply to our current ways of working that could make said agency more effective. By ‘we’ I mean the practitioners, community organisers, grassroot organisation leaders, academics and technologists that are part of this Community of Impact (CoI)* that’s working toward uses of data and technology that improve and celebrate life.
So, in no particular order, here are some of the other lessons we learned.
Raising awareness is not enough.
One of our assumptions when we started this work was that many people are largely unaware of the harmful ways that data-centric technologies can affect us. This harm falls especially hard on those of us who are from low income households because economic disadvantage tends to further compound the effects of other forms of marginalisation.
There are good reasons for this lack of awareness. For one thing, the use of data-centric technologies is increasingly pervasive and often undeclared, so if it isn’t your day job, keeping track of trends and developments can be tricky. For another, the effects of these technologies can be diffuse and occur some time after the products and services are deployed. Alix Dunn sums up the negative impact of emergent technologies perfectly when she describes it as ‘slow violence’.
The interviews and discussions we engaged in helped us see that compiling facts about, and highlighting the harms of data-centric technologies on, Black lives aren’t enough to achieve our stated goals. We needed to explain how these harms tangibly affect people’s lives. Also, people need the time and space to discuss, question and reflect on what this means for them, their children and other people they interact with as well as time to figure out what to do about it.
This is why it’s important for us to continue to engage with and support organisations that are invested in marginalised communities and who have the social infrastructure to support these discussions and help people take action. This understanding has been a key driver of our proposal to carry on this work as part of a community of impact.
It’s important to understand who we need to work with to improve the ways in which data and data centric technologies are used in our lives.
Social media has proven an effective way of quickly transforming, otherwise disconnected, individuals into a coalition capable of pushing for a specific change. It works especially well when the change being pursued is relatively simple to define and the desired outcome(s) can be achieved quickly. In these conditions, the coalition’s breadth and longevity don’t matter so much. At the very least, they are manageable challenges.
However, reversing and preempting the kind of ‘slow violence’ we’re looking to tackle with regard to so-called ‘negative externalities’ of data and technology requires different tools. We need networks of people with diverse talents and experiences that are capable of focused, sustained, nuanced and multi-faceted action.
It’s important to foster an environment of psychological safety.
Structural racism allows tech companies (and the institutions that buy and integrate their products) to remain incurious about the impact of their products and services on Black communities and to blithely disregard negative effects when they are pointed out.
For example, the Home Office was aware that the face detection technology they implemented for checking passports photographs online worked sub-optimally for people with darker skin tones. However, having satisfied itself that the photo-app worked sufficiently well for the majority it decided that negative outcomes for ethnic minority users was an acceptable trade-off:
User research was carried out with a wide range of ethnic groups and did identify that people with very light or very dark skin found it difficult to provide an acceptable passport photograph.
However, the overall performance was judged sufficient to deploy.
UK launched passport photo checker it knew would fail with dark skin
Unpicking and analysing the effects of structural racism in yet another sphere of life is emotionally exhausting. Turning what could be a relentlessly draining exercise into something more empowering requires an environment of psychological safety. At our roundtables we encouraged active listening to one another’s experiences. At the start of every session we reiterated our expectation of and commitment to mutual respect.
We discouraged the use of overly broad generalisations and ‘devil’s advocate’ type argumentation as a means of facilitating deeper understanding. Instead we favoured good faith questioning and discussion along with a commitment to shared learning. Outside our roundtables because we were aware of the risks associated with hypervisibility for marginalised groups, we promoted and conducted our discussions off social media.
Challenging tech power requires a whole system approach.
Facilitating and amplifying the agency of organised people was a core driver for us. However, it’s disingenuous to suggest that ‘empowerment’ alone can address the kind of power imbalance that is often at play. It is also misleading to suggest that this is simply down to tech companies (Big Tech or otherwise). It would be impossible for data-centric technologies to be as pervasive as they are without both public- and private-sector organisations embedding their use in their services.
For example, the use of so-called EdTech products and services, created by private sector technology companies, has grown within mainstream education. This can be linked to the Department for Education’s (DfE) policy decision to actively promote the use of such products. It has done this without instituting a robust regulatory framework for assessing and certifying these products and services nor has it created guidelines for schools about how to do so. Further, while encouraging the adoption of these products and services, it hasn’t consistently or comprehensively dealt with the problem of digital exclusion amongst a significant minority of school children in the UK. The blame for any widening of the attainment gap as a result of this policy can’t be placed solely at the feet of tech companies.
It’s also worth noting that without permissive regulators these institutions and corporations would be far more circumspect in their use of such technologies. As it is, investigative journalism, worker mobilisation and legal action (or the threat thereof) have yielded better outcomes for those who have suffered harm as a result of the misuse of data and data-centric technologies, than complaints to the regulator.
Data and data-centric digital technologies also support greater inclusivity.
Poorly designed or inappropriately used data and data-centric products can have big, negative impacts on the lives of individuals. That could be wrongful arrest or being kicked out of a Southern Co-Op store because of unregulated use of biased automated facial recognition. Alternatively, it could be proctoring software wrongly flagging ethnic minorities as potential cheats. With such potentially high stakes it’s easy to focus on the negative impact of these technologies on Black communities. However, homing in on just the negative impact of data-centric technologies denies us the chance to explore opportunities for flourishing. This, in turn, stymies our proactivity, placing us in a permanent reactive state. That’s why we are so inspired by the work of organisations like Afrotech and why we are keen to continue to work on the emergent themes from this work. Our first project is likely to focus on the theme of data about us,[needs to be] for us.
Key Terms
Community of Impact (CoI) – a group of people who work towards a shared passion or cause, and learn how to do the work better through interacting and sharing knowledge. One example would be the Institute of Global Homelessness.