Graphs and charts are, suddenly, everywhere.
Whether they’re explaining votes, predictions and probabilities in the US presidential election, or being used to justify public-health decisions on Saturday night TV, these lines and arrows are an attempt to bring to life information that has been collected, processed and analysed. And whether or not you can get your head around the small print, a data-driven decision is probably being made about you right now.
These decisions might be small-seeming, tactical ones about the kinds of ads you see or the content that is recommended to you; they might be personal ones, that affect your day-to-day life, such as your credit score or ability to access a public service; or they might be policy decisions that affect you and your family - anything from traffic calming in your neighbourhood to the number of hospital beds available in your community.
Data about you might be contributing to the global effort to develop a coronavirus vaccine, or it might be fuelling decisions being made about how to engineer an economic recovery. It could be helping to prove who you are, getting you paid, and sending your good thoughts to a loved one - all at the same time. It is everywhere, and it is as various and varied as the people and things it describes.
The UK government’s National Data Strategy is open for consultation until 2 December. This is a huge, ambitious and complex programme of work that will affect everyone in the UK, and inform and enable our dealings with the rest of the world. It’s not really a technical document, but a decision-making one - and as such, it deserves the widest audience it can possibly get.
What makes it so important is that it sets out a wide-ranging and ambitious plan for the way data is used in this country - not just in government but in business, public services and charities, and it proposes lowering regulatory barriers, and making it easier to share more information about people and things between sectors. It covers everything from energy switching to epidemiology, from roads and bridges to working with victims of domestic violence and preventing suicide ideation. Its broad sweep includes many of the most sensitive and complex aspects of human existence and relationships, yet it looks like a technical document and it will largely be read by technical people.
In particular, DCMS have asked for responses that explain the impact of the National Data Strategy on people with protected characteristics. In UK law, the protected characteristics are age, disability, gender reassignment, marriage and civil partnership, pregnancy and maternity, race, religion or belief, sex, and sexual orientation. And just like in the physical world, people who don’t represent the mainstream are often not fairly treated or represented in the datafied one.
One of the biggest issues is that data is often missing and unrepresentative; another is the assumption that all data is neutral and true, and that everything can be equally measured.
For instance, police arrest data does not necessarily represent violent crime; instead it shows what researchers Kristian Lum and William Isaac call “some complex interaction between criminality, policing strategy, and community - police relations”. It is a record of what has happened, often without context, and the things that are captured are not an objective description of reality. After all, even a body-worn camera projects a specific point of view.
Moreover, things that look important in one bit of the data universe might be missing in another.
For instance, until very recently, information about ethnicity was not captured on death certificates, but it has long been a part of arrest records. The truth of that is not just unpleasant, it points to structural racism: a society that associates ethnicity with crime but not with care, and makes decisions about public services on that basis, is neither equal nor representative.
The UK’s approach to innovation must not hinge on deepening social inequalities; it should focus on uniting and strengthening us.
Poor data management did not create the hostile environment policy, but the destruction of landing cards for members of the Windrush generation did lead to the deportation of 83 people and the denial of healthcare and legal representation for hundreds more. A failure to update Excel led to the loss of thousands of Covid test results at a critical time for controlling the spread of the virus, and poor implementation of a data-driven solution meant that thousands of children from historically disadvantaged schools were marked down in their A level results.
The people who lose out when these decisions go wrong are more likely to be the people who fall outside of the datafied mainstream, not just because of protected characteristics but because of economic disadvantages and institutional bias. A data-driven Britain must put safeguards to ensure that innovation benefits everyone equally. For that reason, I’m asking civil society leaders to use their response to the National Data Strategy to call for specialist Data Commissioners, who can champion minoritised communities by scrutinising and improving the collection and management of data, understanding potential social implications, and ensuring transparency around data-driven decisions.
The consultation for the National Data Strategy closes on 2 December.
If you would like to co-sign a co-ordinated response, please email email@example.com by 27 November giving your name and affiliation. A more detailed breakdown of the social implications of the strategy is given in this Policy Briefing.