Central government and civic services

The suppliers, products and customers in this chapter are in the landscape database, filtered to central government, with known spend and sources. The graph shows how they connect.

Sector overview

Central government IT is the most heterogeneous sector in this survey. It includes the GOV.UK platform itself, the cross-government identity programme, departmental ERP estates, the major operational systems of the largest spending departments (HMRC, DWP, Home Office, MoJ), and a long tail of agency and registry systems (DVLA, DVSA, HMPO, Land Registry, Companies House, the courts).

Parts of this estate were built by GDS on an open architecture and have been cited as examples by other governments. Other parts are long-running programmes led by systems integrators, with overruns of £10m or more and timelines of around a decade. Both kinds of programme exist within the same departments.

Total digital spend

Central-government digital spend is in the region of £14–20 billion per year Med across all departments and agencies. The Government Major Projects Portfolio (GMPP) carries multiple tens-of-billions of pounds in active digital and transformation programmes at any given time. Roughly half of in-year spend is on major operational systems (HMRC, DWP, Home Office); a quarter on cross-cutting infrastructure and shared platforms; the remainder on departmental and agency systems.

Major systems inventory

Lock-in and concentration

Central government has two distinct estates with very different lock-in profiles. The GDS-led shared-platforms estate (GOV.UK, Notify, Pay, Forms, the Design System, parts of One Login) is publicly-owned, mostly open-source, and one of the clearest examples in government of publicly owned, open platforms in production use. The SI-led legacy estate (HMRC’s tax core, DWP legacy benefits, agency mainframes) is deeply locked into a small number of major systems integrators (Capgemini, Fujitsu, IBM, Atos, Accenture) with contract histories spanning decades and switching costs high enough that replacement would need a programme lasting many years.

Recent changes go in both directions. DWP’s 2017 Universal Credit reset, which moved delivery substantially in-house, reduced systems-integrator dependency. HMRC has reduced (but not eliminated) SI dependency through internal capability building. The cross-departmental ERP convergence onto Oracle Cloud increases dependency, placing a critical system for most departments with a single US vendor.

Replaceability assessment

GDS shared platforms D

Already publicly owned and largely open source. The CIC’s role here would be to contribute components and argue for continued openness. GOV.UK Pay and Notify are each used by hundreds of services, which is an existing UK example of the shared-platform approach.

One Login D

Built mostly in-house with significant openness. The sovereignty issues that exist are matters of governance.

HMRC core systems C

Tax-collection systems are at the difficult end of grade C. Replacing CHIEF with the Customs Declaration Service took most of a decade. There is no plausible CIC-aligned replacement target here. The practical measures are procurement, exit terms and capability-building inside HMRC.

DWP estates C

The Universal Credit codebase is publicly owned and built largely in-house, so it is not a replacement target. Modernising the legacy benefits estate is a long internal programme and is not a CIC opportunity.

Cross-departmental ERP C

The convergence onto Oracle Cloud is the largest increase in single-vendor dependency in central government this decade. Replacing it after migration would be very difficult. Procurement terms are the only realistic means of influence.

Registry agencies (DVLA, DVSA, HMPO, Land Registry, Companies House) B

Companies House and parts of Land Registry are already largely in-house with strong open-API surfaces. DVLA and HMPO have heavier vendor dependence and would benefit from the same in-house modernisation pattern that worked for DWP. The CIC’s plausible role is component contributions to the open-API surfaces and, longer-term, helping smaller agencies reduce SI dependency.

Court systems: Common Platform C

Covered in the justice chapter.

The human dimension

Central-government civil servants split into roughly two populations. Those working on or with the GDS shared platforms generally rate their tools highly; the tools are modern, fit-for-purpose, and shaped by user research. Those working with the legacy SI-built estate report the same frustrations as the rest of public-sector users: poor performance, brittle integrations, slow vendor change cycles, and very little ability to change their own systems.

Skills lock-in is significant: the legacy mainframe estates depend on an ageing workforce whose skills few new entrants learn. This is a long-term operational risk regardless of who owns the systems.

Sector summary