The United Nations Development Programme has just released some updated gender data news: the latest Gender Social Norms Index has revealed no improvement in biases against women in a decade. Depressing, but not shocking. Yet this use of gender data reminded me of some of the gaps I see in our more everyday gender data conversations.
The intersections between gender and data are multiple, largely because data is such a fundamental component of power. And the association between algorithmic bias and gender gaps is a discussion that has been well articulated by influential female thinkers like Timnet Gebru, Joy Buolamwini and Caroline Criado Perez.
The first solution posed by many is to call for more sex- and gender-disaggregated data (gender-disaggregated data refers to information that is categorised by gender). By breaking down information based on sex and gender, we might begin to understand and highlight gender differences and disparities - and we can produce reports like that produced by the UNDP above. Certainly, that is a starting point. The disaggregation allows us to make women more “visible”. It sheds light on the lived experience of women in policy conversations that are often blind to our stories. When OpenUp disaggregated municipal level data on youth unemployment, it helped to tell a story not just of how impacted by unemployment young women are, but also of how the work women do engage in domestically, everyday, remains uncounted and undervalued by data collectors and analysts.
But what if we move beyond unpacking the data that is collected, to influencing what data itself is collected? What if it is communities of women that drive the data agenda itself? Data may make visible, but what is it women want to make visible? In other work I have done, I have discussed the concept of “beneficial visibility” - this is the idea that individuals and communities, through their ability to shape the data agenda from the bottom-up, might better control how visible they are to the world - and thus can try and create a ‘level’ of visibility only to the extent necessary to derive beneficial outcomes for themselves and their communities.
This is not just a theoretical idea, but can be embedded in practice. For example, when OpenUp worked with communities to workshop their data agenda and data needs, those communities then collected that data for their own advocacy end. Working with the Witzenberg Justice Coalition, attendees of our data and digital literacy training identified basic and health services as amongst the key challenges faced by the community that they wanted to address using data. They then created a survey, which formed the basis of a community-led data collection exercise. The data was then used for the community’s lobbying activities with the Western Cape Department of Health. This meant the data collection process itself was intrinsically embedded with trust given from whom it was being collected, by whom and for what.
I am suggesting then, in an era obsessed with big data, perhaps some of the most radical approaches to data may be demonstrated in its most localised form: data from a community, for a community. Data from women, for women. This essential distinction ensures the agency of those who the data is being leveraged for, and ensures a sense of data democracy that helps ensure opportunities and outcomes for everyone.
This is important for the conversation about gender equality and data, in the longer-term. Data extractivism dominates the digital landscape. As Noami Klein so wonderfully noted in her recent masterpiece on the hallucinations behind Artificial Intelligence:
“Because we trained the machines. All of us. But we never gave our consent. They fed on humanity’s collective ingenuity, inspiration and revelations (along with our more venal traits). These models are enclosure and appropriation machines, devouring and privatizing our individual lives as well as our collective intellectual and artistic inheritances. And their goal never was to solve climate change or make our governments more responsible or our daily lives more leisurely. It was always to profit off mass immiseration, which, under capitalism, is the glaring and logical consequence of replacing human functions with bots”.
Yet there are still many, many data dark spots. In fact - much of our region remains the dark continent in terms of data and knowledge representation. Perhaps the first acts of resistance to extractivism can be embedded in the gendered practice of data creation and collection?
Over the next few years, OpenUp will be exploring a number of projects to test these ideas. Stay up to date with work here.