Summary

• GDS has identified people, organisations, and places as foundational data domains repeatedly used across UK public services.
• Its proposed approach relies on shared identifiers and “reference spines” rather than combining government information into one central database.
• The work exposes an implementation problem behind joined-up digital government: organisations must invest in common data infrastructure even when benefits fall outside their own boundaries.

The Government Digital Service is developing a common approach to the basic data used repeatedly across public services, tackling a problem that has frustrated successive attempts to build joined-up government: different organisations frequently maintain their own versions of information about the same people, businesses, and places.

A discovery team inside Government Digital Service has spent the past year identifying what it calls foundational data — core facts and identifiers repeatedly required by national departments, local authorities, and devolved public bodies. The initial work concentrates on people, organisations, and locations.

GDS wants those datasets to become easier to find, understand, and reuse across organisational boundaries, reducing duplicated record keeping and repeated validation of information another part of government may already hold. It is not proposing a central database containing all of that material, instead favouring a federated model in which existing organisations continue maintaining their own authoritative records.

The proposed connective layer is built around what the team calls “reference spines”: consistent ways to establish that separate systems refer to the same person, organisation, or place. In practice, that means agreeing identifiers and standards that allow information to be connected where there is a lawful reason to do so without requiring every dataset to be physically combined.

Government still identifies the same thing differently

Location data illustrates both the opportunity and the difficulty. The UK already has the Unique Property Reference Number, which provides a persistent identifier for properties and locations, but GDS says its use across public services remains inconsistent. Where systems identify the same address differently, organisations have to reconcile records before they can connect services or analyse demand reliably.

The position is less mature for organisations. Companies House numbers provide a common identifier for incorporated companies, but government deals with charities, schools, partnerships, public bodies, community organisations, and other entities for which no single identifier is consistently used across every relevant service.

Information about people presents the hardest governance problem because the benefits of linking records sit alongside legitimate constraints on how personal data should be combined and reused. GDS says the aim is not a central population database and that connections should occur only with an appropriate lawful basis, public benefit, transparency, safeguards, and oversight.

Keeping datasets with the organisations best placed to maintain them preserves clearer ownership, although a federated architecture also makes interoperability dependent on those organisations agreeing and implementing common standards. The technology is less difficult than the institutional coordination required to make it routine.

Shared infrastructure creates uneven incentives

The economics are awkward because the organisation paying to improve a dataset may not capture most of the resulting benefit. A council investing in cleaner address information, for example, could make services easier for health bodies, central departments, emergency services, and regulators as well as its own teams.

Departmental budgeting can struggle to reward infrastructure whose value appears somewhere else, and GDS acknowledges that organisations are understandably focused on delivering their own priorities even where investment in shared data assets could generate wider gains. The problem resembles other common digital infrastructure: everyone benefits from dependable foundations, while each individual organisation has an incentive to wait for somebody else to fund them.

Better foundational data could reduce routine administrative work that accumulates across government. Citizens are repeatedly asked for the same information, while staff spend time checking addresses, establishing which organisation a record belongs to, or reconciling slightly different versions of a person’s details.

Those inconsistencies also constrain more ambitious technology programmes. Artificial intelligence systems, automated eligibility checks, proactive public services, and cross-department analytics all depend on establishing which records correspond with sufficient confidence. Adding more sophisticated software above inconsistent identifiers does not remove the inconsistency.

The work consequently sits beneath several wider GDS programmes, including GOV.UK, GOV.UK One Login, the National Data Library, artificial intelligence services, geospatial data, and the National Underground Asset Register. Each relies in different ways on information being sufficiently standardised to move between systems.

GDS points to Denmark and Estonia as examples of governments with stronger mechanisms for discovering and reusing foundational data, although transferring those approaches into the UK means working across a larger and more institutionally fragmented public sector. Devolved administrations, councils, agencies, and central departments have different systems, legislation, responsibilities, and technical histories.

The discovery work will feed into a Strategic Data Roadmap due from the government’s Chief Data Officer in the autumn. GDS is also seeking feedback on an alpha Person Data Standard, giving the programme an immediate standards route rather than leaving it at the level of architectural principles.

Whether those standards become useful infrastructure will depend on adoption. Shared identifiers produce little value when only a handful of organisations implement them, while mandating common approaches without funding migration can leave older services dependent on local workarounds.

The government’s wider digital strategy assumes that services can increasingly be organised around events in people’s lives rather than departmental structures. Before that can work consistently, the systems involved need a dependable way to establish that they are referring to the same person, organisation, or place. GDS is tackling a basic data problem, but one that sits underneath almost every attempt to make government software operate as a connected system rather than a collection of separate services.