show image

Jeni Tennison

VP and chief strategy adviser, Open Data Institute

Why, now more than ever, we need to fix our data blind spots

Covid-19 has exposed the gaps in our national data infrastructure, and highlighted that our data culture, even motivated by a crisis, is still far from effective. Data has the power to paint a picture of what’s really going on around us. But where people who need it don’t have access to it, and in some cases data isn’t even collected, it can have serious consequences. It can skew our perceptions, lead to ineffective and potentially harmful interventions, or cause significant and deadly trends to remain hidden, with (at worst) fatal consequences.

For example, when the NHS does not share data about the impacts of Covid-19 on people of particular backgrounds or ethnicity, it prevents anyone from understanding the scale of its differential impact on minority communities or the ability of agencies to address it. When the Department of Health and Social Care and the Care Quality Commission don’t even collect data on the capacity of care homes and the experience of those within them, we are blind to their fragility.

Data is the defining resource of the current digital revolution. But there are huge questions about how collecting it, granting access to it and using it should be managed and governed. Our current ways of thinking about and regulating data are insufficient to the task.

Data brings with it the power to understand the world and the reasoning for acting on that understanding. The data we collect determines what we see and how we see it; how it is shared determines who gets to benefit from the insights and innovation it can bring.

When those developing digital public services, such as the contact tracing app, don’t monitor the characteristics – even sensitive and protected ones – of people using those services, how can they understand whether they are reaching people equally? If we don’t know about these gaps, we cannot put in place alternatives – whether digital or not – that reach missing communities. When testing and symptom tracking data is not broken down by geographic area or shared beyond central government, we prevent devolved decision making and local initiatives. Responses are delayed, communities disempowered, and businesses and governments are held back from making interventions suited to their local context.

These blind spots and failures in safely sharing and granting access to data are not just a problem of data collected by the public sector. When companies keep to themselves the insights they glean from the detailed data they collect as a side effect of us going about our daily lives – till receipts, insurance claims, phone connections, location histories – we are left reliant on lagging survey-based statistics. We are deprived of the opportunity to be quickly informed of the current state of the world, to act now on current problems rather than those we were experiencing a month or a quarter ago.

Data can give us the power to act responsively, in informed ways. But it can also be misused, and harm us. The very fact that we are tracked, that data about us is collected and stored, that others have knowledge about our lives, makes us modify our behaviour. We have the right to be included, our stories told, and to benefit from the insights data about us can give us, whatever our ethnicity, age or location. But we also have the right to privacy and to autonomy, and to be anonymous.

Squaring these circles – giving people, communities and small businesses voices through data, and the power to act based on it, while protecting them from the harms that its collection and use can bring – requires us to explore different ways of managing and regulating this powerful resource.

One approach that people around the world are researching, experimenting with and building are data institutions: organisations whose purpose includes stewarding data on behalf of others, to maximise its benefits and minimise its harms.

New kinds of data institutions come in many different forms. Commons-based data institutions enable people to band together to maintain open data assets, such as the maps maintained by OpenStreetMap. Other data institutions provide independent governance over sensitive data, such as Social Science One which manages research access to data held by Facebook. Data trusts assign the authority to make decisions about data collection, use and distribution to trustees, bound by fiduciary responsibility to act in the best interests of those affected by the use of that data. Data cooperatives and unions where workers bring together the data they hold to create collective insights to support negotiation with the people they work for.

Organisations that collect use and share data need to recognise their broader responsibilities around the data they collect and the communities they operate in. They need to examine their blind spots, consider how their hoarding of data constrains others’ abilities to act and innovate. They need to engage much more widely with the people whose lives they impact. Whether it’s responding to a crisis, or building a more prosperous and sustainable future, we need data to work for everyone.

Jeni Tennison is vice president and chief strategy adviser at the Open Data Institute