Interview: Using data 'brilliantly' to deliver levelling up

Tom Smith, Chief Data Officer at the Department of Levelling Up, Housing and Communities (DLUHC), and Director of the department’s Spatial Data Unit, speaks to Government Transformation magazine about how his organisation is optimising the use of data to help deliver the levelling up agenda.

A public service calling

Smith joined DLUHC in March 2022 from the Office for National Statistics (ONS), where for half a decade he was Managing Director of the Data Science Campus – delivering data science and capability programmes for the benefit of the UK and globally.

During the pandemic, he had a six-month stint at the UK Health Security Agency to help build the Joint Biosecurity Centre and improve the country’s response to Covid-19. However, Smith’s career started in academia and industry start-ups. A physicist by training, he later did a PhD in AI and neuroscience “making robots play football”.

Smith’s public sector call came with the realisation of the potential that the vast amounts of data held by the administration could do to improve people’s lives.

“I got very interested in what you can do using data held by government and government agencies to design and target public programmes, and evaluate impact and monitor progress.”

Although he is conscious of the challenges that government faces with recruiting data and digital specialists, Smith says that Civil Service and public sector work has the unique proposition of offering high-impact social value at government scale, and the clear recognition that what he and his team are doing is helping to the UK “in its mission of levelling up”.

Using data at DLUHC

DLUHC and its predecessor departments have a long history of using data to inform policy and support delivery programmes. An example can be found in the Index of Multiple Deprivation – a long-standing tool that measures relative deprivation by area in England. The department also launched the first open data portal across government, then working with the Government Digital Service to build this into Data.gov.uk.

“Much more recently, there’s been work on linking longitudinal data around homeless communities and cohorts to better understand the challenges faced by those groups over time, and to target and support intervention to improve outcomes,” adds Smith.

Behind Smith’s appointment as DLUHC’s first Chief Data Officer is the department’s goal to continue building and strengthen its internal capability to allow data to be applied across all delivery, programme and policy areas using “world-class data tools”.

Supporting this work is the newly established Spatial Data Unit, of which Smith is Director. The directorate was created following the government’s Levelling Up White Paper and its role is to support the levelling up agenda with high-quality local data and analysis. In Smith’s words, “the spatial data unit has a very simple role: use data brilliantly to help deliver levelling up across the UK.”

“Talent is spread equally across the UK, but opportunity is not,” Smith continues. “Spatial data is key to understanding in detail the problems you're trying to solve and what solutions and interventions can help.”

The directorate relies heavily on geospatial data – time-based data relating to specific location – and geospatial data science to produce innovative analysis that can support the levelling up delivery.

The unit works closely with local government and partners to ensure that it has the right data to support levelling up’s 12 missions. This often means improving and producing “better data” at a local level that can help assess progress and challenges in different areas of the country to understand what interventions and local programmes are most appropriate. It also means operating across government, working with partners in other departments such as the ONS.

A big challenge for Smith comes with the trade-off of responding to short-term priorities while building long-term capabilities: “This is not a new challenge, lots of groups face this. But it’s a useful challenge and a brilliant learning curve.”

“That trade-off is a powerful experience for the team because you need to successfully deliver on the short-term and longer-term as you build from start-up. We must be able to support and respond to the urgent and high-priority requests coming in from ministers and senior decision-makers.

“In the early days, that response will be very reactive, very hand-to-mouth. But longer-term, we put in-place the tools and capabilities so we can respond rapidly on a routine basis – and using the short-term commissions as signals of what capability is needed for the longer-term.”

Creating a “data bridge” to unlock data value

To make sense of the huge amounts of data being produced at DLUHC, Smith says that there is a need for a “data bridge” able to link the data available with the necessary skills and tools to make sense of it.

“You need to create a partnership through that data bridge,” he explains. “You need strong domain understanding, analysts and researchers with background in the subject, policymakers who really understand and have detailed in-depth knowledge of the issues that you’re looking to tackle and improve. And you need to partner that with the data skills and expertise in the technical sense of what can be done with particular types of data.”

The Spatial Data Unit plays a vital role in enabling this data bridge, hence it being embedded in the levelling up mission. Instead of sitting in the corporate centre, the unit is physically sat within the levelling up groups that look at research and analysis and support policymaking.

“We feed in and discuss with those groups and are part of those conversations on a daily basis,” Smith says.

Smith works within a highly skilled team that has the technical brilliance and expertise necessary to unlock data value, including data engineers and scientists able to apply the tools and methods necessary to extract features and produce analysis out of text data or fast-moving network data, for instance.

But one skill that’s essential for Smith and his group is curiosity: “What is it that I can use to tell me something new about the world that we don't currently know or that we haven't yet fed into this policymaking discussion?”

Working closely with policymakers and researchers means that Smith’s team must be clear about how they present their data, whether through visualisations, dashboards or other tools, so they are fed into the decision-making process by ministers and senior officials at DLUHC and the wider government.

He adds: “We need to be really smart at understanding how and where to insert our data into the discussion so it informs live debates, live discussions, live decisions.”

The future of data in government

Central government has made leaps and bounds in data interoperability, something that was widely seen during the pandemic. Smith acknowledges the positive collaboration between departments when it comes to data-sharing, but he also thinks that much more can be done.

To achieve this successfully, continuous data-sharing between government organisations will be essential. However, this, Smith says, “is rightfully hard and rightfully is treated really seriously”. Data ethics and safeguards are paramount, and Smith highlights the role that the Centre for Data Ethics and Innovation - a government expert body that ensures the ethical use of data and AI – as well as other organisations are doing to ensure that.

But Smith also says that along with considering the safeguards of data-sharing, government must also consider the risks of missing out on using data.

“Citizens expect us, as government and the public sector, to make really good use of data for the benefit of the UK,” Smith says. “The UK expects us to run programmes, deliver policy and create legislation that has real benefit and real value – and you need good data to underpin that.”

As an example of an innovative use of data-sharing for the public good, Smith mentions an initiative by the Ministry of Justice’s (MoJ) Justice Data Lab to reduce reoffending rates. Using highly sensitive data from MoJ, the Data Lab helps charities and delivery providers improve their support to ex-offenders and tracking progress.

“It’s a great example of using sensitive data that government holds to support better outcomes - in this case helping people with skills and support so they are less likely to end up back in prison - without releasing or disclosing that sensitive data,” Smith continues.

“I want us to be doing more examples of creating high value with public data while respecting and following safeguards and data ethics.”

There is an increasing number of Chief Data Officers across Whitehall departments and public sector organisations – in Smith’s view a positive step that is setting the direction of travel and will help build the necessary infrastructure and capabilities to do better with data. But he thinks that there is still a potential role for a greater whole-government Chief Data Officer able to look at the big picture and create lasting impact.

Smith concludes: “Any Chief Data Officer for government needs to look at and work with the chief data officers in departments and think hard about systemic change that can support better use of data across government.”