The barriers to a National Data Strategy are falling: it’s time to move forward

The government is well aware that it needs a National Data Strategy – and after many months of delay, the conditions are emerging for it to make progress.

In June 2018, then-culture secretary Matt Hancock promised a strategy to “unlock the power of data across government and the wider economy” – then events intervened. That April, the technophile Hancock had persuaded Theresa May to pass the data policy brief from the Government Digital Service to his Department for Digital, Culture, Media and Sport (DCMS). But a month after announcing the Strategy, he was moved to the health department – leaving the agenda without a clear champion. Focusing on Brexit priorities, DCMS lost a year before launching a call for evidence – whose latter stages were derailed by the general election campaign. Meanwhile, the Cabinet Office tried to hasten progress by recruiting a permanent secretary-level Government Chief Digital Information Officer (GCDIO), but couldn’t fill the role. By December, the government looked set to miss its self-imposed 2020 deadline for delivering the strategy.

In February 2020, the picture looks quite different. Late last year, DCMS began rebuilding its team by appointing a data policy lead.. And in this reshuffle, Oliver Dowden – the former Cabinet Office minister who’d tried to hire a GCDIO – took the helm at DCMS: the department now has a minister who understands digital issues and the need for rapid, cross-departmental action.

The pull factors

So the barriers to progress are evaporating. And there are powerful reasons for central government to get moving on a National Data Strategy, unlocking a huge range of benefits for service quality and operational efficiency. To take a few examples…

The ability to share data across government to common standards will vastly improve the experience of citizens accessing digital services, supporting seamless cross-departmental services built around users’ needs. And it would dramatically boost civil service efficiency, minimising duplication, stripping out routine data management tasks, and improving officials’ tools and roles.

Improving data quality will provide the means to strengthen case management decision-making, improve the targeting of resources, and use reliable evidence to shape strategic decisions on service design, business planning and infrastructure investments.

Citizens will only permit departments to make full use of their personal data if they believe their privacy and rights are being respected. Implementing a data strategy will bolster and unify transparency, reporting and data management standards across government, demonstrating that government is a reliable custodian of citizens’ data ­– and thus improving public confidence and departments’ license to innovate in the field.

Setting out cross-government data standards would expand and stimulate the market for data management products and services, prompting the tech sector to respond. Just as the Government Digital Service’s creation of technology and design standards fostered an ecosystem of cheaper, better products, common data standards could help expand choice and reduce costs for departments.

So implementing a strategy could help transform government’s capabilities, improve public services, reduce waste – and give civil servants the tools they need to deliver on the government’s priorities. But there’s another reason for getting on with the job: the continued delay is imposing a huge opportunity cost.

The push factors

In the absence of a strategy, each department and agency is developing its data systems and policies in isolation. Organisations are setting rules on data formats and definitions – but these are not compatible across government. They’re building internal networks and data-sharing protocols – but without the means to ensure that they’ll support inter-departmental exchange. They’re commissioning data services and buying new tools – but, lacking the economies of scale and quality standards produced by central buying frameworks, many are investing in tools that may be rendered obsolete when the strategy lands. And the longer this divergent progress continues, the greater the challenge of aligning their data operations when the government does provide a clear vision of its goals. Every day we delay, the task of implementing a data strategy grows a little greater.

Every day we delay, the task of implementing a data strategy grows a little greater.

Civica has helped many civil service bodies to create data strategies, building our own understanding of how best to promote their development, ease their implementation and maximise their impact. And as we await the next chapter in the development of a national strategy, this seems a good time to share some of what we’ve learned.

The key elements of a data strategy

The starting point is an organisation’s business strategy: data leaders should pick out its over-arching goals, and identify the datasets and capabilities required to realise them. Not all data is created equal, and you can’t do everything: organisations must concentrate on the datasets most essential to developing policy, targeting resources and strengthening services, building a picture of the connections and processing capabilities required to pull out those insights. This latter piece is key to success: by bringing together existing datasets and building links to other departments’ resources, many public bodies could develop a far clearer picture of how their services work in practice – and thus how to improve them.

Next, develop your approach to five key topics. The first is data architecture: how and where data is held, and the standards governing language, definitions and formats. Documented in data dictionaries or conceptual data models, these standards ensure that datasets are compatible and can easily be combined for analysis or use in service delivery.

With those compatibility issues addressed, the data integration agenda covers the standards and protocols required to transfer information between different systems and organisations easily and securely.

Then there’s data quality: the standards used to evaluate a dataset across dimensions including accuracy, validity and timeliness, and to record how it’s been gathered, handled and processed. This aspect of the strategy should also set out the processes for addressing data quality issues at source, while allowing civil servants to prioritise implementing standards within those datasets that are most important to the organisation.

Master data management governs the organisation’s approach to developing a ‘golden record’ for critical data points, such as citizen identity or location. By eliminating duplicates, capturing the best available information and sharing it across the system, this process underpins data managers’ ability to combine datasets and build services around users’ needs.

Finally, data governance concerns the allocation of responsibilities and accountabilities. Who decides which datasets must reach a quality standard, and whether data is fit for a specific purpose? Who makes a ruling when different teams or organisations have conflicting requirements? Who ensures that people are following the rules when, for example, data is shared or accessed from many points?

These are the core elements of an effective strategy. But a strategy can present great ideas on all these fronts, while collapsing at the implementation stage: successful delivery depends on excellent stakeholder engagement throughout the development process, the strategy’s ability to flex around local needs, and its potential to help staff across the organisation realise their own goals.

The key principles in developing a strategy

Genuine consultation is crucial to understanding which reforms would most assist policymakers and delivery staff in their work; scouting the risk that planned changes might cause new problems; ensuring that staff will have the necessary tools and resources; and building connections with the people who’ll be tasked with delivery. The leaders and teams who’ll implement the strategy must play a key role in designing it; and if you’re bringing in external support to help you develop the plan, you should ensure they’ll stick around to help roll it out. Development should be an iterative process, taking on board feedback at every stage until its launch – and extending throughout implementation, so the strategy keeps on evolving as conditions change and lessons are learned.

Another essential principle is that of flexibility. Most strategies will include a path to introducing central standards, buying frameworks, shared tools and performance metrics – and these have huge value in promoting interoperability and exchange, cutting costs and monitoring progress. But every public body has its own priorities, and needs enough autonomy to apply central processes in ways that support its core mission and recognise its unique delivery environment.

Genuine consultation is crucial to understanding which reforms would most assist policymakers and delivery staff in their work

So, for example, data quality policies and standards should provide a common framework for measuring data quality, while allowing departments to determine how far each dataset must comply to meet business needs. Buying frameworks should present organisations with a list of approved, interoperable tools, rather than insisting they all buy the same system. And performance metrics should permit them to decide which capabilities are developed first – enabling the centre to ensure that every department is making good progress, while allowing departments to prioritise the reforms that will address their goals.

In general, central systems should operate more like the safety and emissions standards applied to new cars than the blueprints created by manufacturers for new models: the goal is to ensure that each organisation can use the same infrastructure, while building data vehicles that suit the unique journey they need to make.

I’d add that, in general, strategies should focus on goals and capabilities rather than specific technologies: provide standards and tools that will help departments to achieve their business goals, rather than setting out exactly how that is best achieved. And they should be carefully integrated with other agendas, such as those around civil service tools and workplace reform – or you risk setting up tensions that will undermine delivery at every level.

The time is now

Producing and implementing a National Data Strategy is a big task – but the longer it’s delayed, the more problems we’re storing up for the future. Without central direction, progress across government will remain faltering and divergent – and as departments develop their own systems, the task of retro-engineering them to promote compatibility and exchange is growing ever greater. Meanwhile, the lack of common standards and data management practices leaves government exposed to security threats and the inadvertent infringement of citizens’ data rights.

And a well-built strategy could give the country so much – supporting economic growth, improving public sector efficiency and enhancing public services. Currently, civil servants working to realise the value of government data must struggle to access datasets, plough through onerous reconciliation processes, and find ways around a host of technical and governance challenges – only to face a new set of problems around data quality, compatibility and maintenance. Given a strong strategy, within a few years we could offer them the tools to seamlessly connect high-quality data resources within an agreed governance and management framework, producing invaluable insights into their organisations’ operations and clearing the path to genuinely responsive, user-focused public services.

Many of the UK’s civil servants understand the potential of government data, but they have little hope of realising it under the status quo. It’s time we gave them the tools, resources and support to tap into this enormous asset – producing huge benefits for civil service staff and organisations, taxpayers and the public.

You can read more about Civica's approach to trusted data analytics here.