Wellcome to Data Feminism — considering our health data projects through an intersectional lens
This article was co-authored by Miranda Marcus & Ekin Bolukbasi
Artist Copyright: © courtesy www.guerrillagirls.com Collection: Tate
With the Pride month coming to an end, a year on from George Floyd’s murder, the 40th anniversary of the first official reported AIDS cases & daily reminders of health and vaccine inequity in today’s pandemic world, it is more important than ever to remind ourselves that sexuality, race, class, age, religion, geography, (dis)ability — and many other demographic factors- intersect with each other to impact an individual’s life experience. When it comes to health science, finding ways to foreground this intersectionality in the way we collect and use data is critical if we want to develop new interventions that work for everyone, not just the minority. This is why in this post we’re reflecting on how Wellcome’s work around data and health match up to the principles of Data Feminism- a guide to intersectional data practice
Data can be a double-edged sword — sometimes creating huge value and sometimes reinforcing systemic injustice. For example, a 2019 study found that an algorithm used in the US to determine which patients should be referred for extra care was racially biased. This meant that black patients were less likely to receive the care they needed compared to equally sick white patients. The way that data is collected, shared and used affects individuals and society in complex ways.
This is not a new issue to the scientific community. But while the harms that data practice can enable are as real as rain, they can be hard to detect when you’re not affected by them. Wellcome sits in a position of huge privilege, the type of privilege that has historically been aligned with white, male, cis-gender, wealthy communities. So, as we are constantly reminded about health and social inequities disproportionally impacting underserved communities, we (Ekin and Miranda) thought it would be a good moment to sense check our work from alternative perspectives.
‘Data Feminism’ is a book written by Catherine D’Ignazio and Lauren F. Klein. It proposes 7 principles to prompt a different way of thinking about data and data ethics. The principles don’t look only at gender- they use the term ‘feminism’ as a shorthand for the diverse, wide ranging projects that challenge racism, sexism, gender discrimination and other forms of inequality and oppression. This perspective goes beyond ‘data ethics’, aiming instead for ‘data justice’ by addressing structural inequality. For example, below is a grid that explains the difference between the two.
The Principles
Principle #1 is to Examine Power. Data feminism begins by analyzing how power operates in the world.
Principle #2 is to Challenge Power. Data feminism commits to challenging unequal power structures and working toward justice.
Principle #3 is to Elevate Emotion and Embodiment. Data feminism teaches us to value multiple forms of knowledge, including the knowledge that comes from people as living, feeling bodies in the world.
Principle #4 is to Rethink Binaries and Hierarchies. Data feminism requires us to challenge the gender binary, along with other systems of counting and classification that perpetuate oppression.
Principle #5 is to Embrace Pluralism. Data feminism insists that the most complete knowledge comes from synthesizing multiple perspectives, with priority given to local, Indigenous, and experiential ways of knowing.
Principle #6 is to Consider Context. Data feminism asserts that data are not neutral or objective. They are the products of unequal social relations, and this context is essential for conducting accurate, ethical analysis.
Principle #7 is to Make Labor Visible. The work of data science, like all work in the world, is the work of many hands. Data feminism makes this labor visible so that it can be recognized and valued.
How our work relates to these principles
- Empowerment and lived experience
One of the underlying principles of the Mental Health strategy is ‘empowerment’. We are making sure the voices of young people with lived experience of anxiety and depression are at the heart of the work we support. Shuranjeet Singh, one of our Lived Experience Consultants, has recently written about the lived experience work going into the Global Mental Health Databank project. Panels of lived experience advisors in all the regions we are working in are feeding into every element of the project, and it is clear what a huge benefit to the project their perspectives are having. However, what is also becoming clear through this work is the importance of building spaces where people who don’t necessarily trust institutions and academia feel their voices are being heard.
Wellcome Data Prizes (WDP) have been designed to make sure participation and involvement is at the centre of the projects we fund. We will listen to and value lived experience, and partner with the communities we’re working with throughout the challenge. For example, the first data challenge on better understanding of youth depression and anxiety will include young people with lived experience from the UK, and South Africa at multiple stages. They will help us refine data prize’s scope and question; they will also participate as solvers and co-create digital tools alongside other mental health experts.
In terms of the Data Feminism principles this commitment to lived experience helps us examine and challenge power (principle #1 and #2), elevate emotion and embodiment (principle #3), rethink binaries and hierarchies (principle #4) and embrace pluralism (principle #5). - Working to promote equity not just equality with the people we engage
Another crucial part of the Wellcome’s mental health strategy is to enable and include research from contexts that are non-WEIRD (Western, Educated, Industrialised, Rich and Developed). The lack of data infrastructure in LMICs means there is less data from those populations. This is a particular issue for mental health research which is so culturally specific. The GMH Databank is working in multiple global regions to explore and demonstrate ways of running international mental health research that is equitable. This means that we need to structure projects so that places without established research data infrastructures can benefit, whilst taking into account the range of cultures data is being collected in. This is no simple task and we don’t have all the answers, but acknowledging structural inequalities in the way we design scientific data infrastructures is crucial.
Our work in health data aims to challenge a landscape where applications of data science in healthcare and research routinely benefit one group of people over another. For example, in the UK, the Covid pandemic has exposed existing health inequalities between different ethnicities and revealed the gaps in patient health records on ethnicity. We have two projects running in parallel to unravel and address these issues. The first project will compare ONS census data and routinely collected health data to understand the impact of missing ethnicity data on research into Covid-19. In a complementary approach Understanding Patient Data (UPD) is conducting a public engagement project to examine and understand the reasons behind data inaccuracy and incompleteness for Black and South Asian people in the UK. On a more global level, our support for Lacuna Fund will help create datasets inclusive of people who might ordinarily not be well represented in health datasets used for Artificial Intelligence (AI) research.
In terms of the Data Feminism principles this commitment to lived experience helps us challenge power (principle #2), rethink binaries and hierarchies (principle #4), embrace pluralism (principle #5) and consider context (principle #6). - Writing about the work openly
We are writing regular updates about what we’re doing. We are also encouraging our partners to do so too. This means we can amplify their voices and perspectives on the work as we learn. It also means we can be challenged as we learn and make decisions. This helps us address the final principle- make labor visible (principle #7).
While it’s exciting to see how some of our work links to the ideas of data feminism, this isn’t just an opportunity to pat ourselves on the back. There is still a very long way to go. But hopefully bringing ideas like Data Feminism into our work will contribute to building more just and equitable science in the future. For example, how do we add context and pluralism to scientific research (principle #5 and #6)? How can we be engaging more with community led data collection? Can we create principles for data equity in science?
