17th October 2026
The Government promised to save £83 million by handing civil service pension administration to Capita. Instead, pensioners have faced delays, backlogs and uncertainty over money they depend on. Now the taxpayer must ask a simple question: how did a contract designed to save money go so badly wrong?.
The transfer went ahead in December 2025 despite missed transition milestones and concerns about the readiness of the new system. The National Audit Office has since highlighted weaknesses in how the Government assessed and monitored the promised benefits.
The £83 million was a forecast, not a saving already delivered. With millions of pounds in support loans paid to affected members and the long-term future of the service under review, the full cost of the failure remains to be established.
This is more than a pension administration problem. It raises serious questions about how Whitehall awards contracts, responds to warning signs and holds private companies to account when essential public services fail.
An £83 million saving was promised
The Cabinet Office awarded Capita a seven-year contract worth £239 million in November 2023. The new arrangement was intended to replace the previous administration provided by MyCSP, with the transfer taking place on 1 December 2025.
The Government's financial case included projected savings of £83 million over the life of the contract compared with continuing the previous arrangement.
Those savings were expected to come partly from improved technology, automation and more efficient administration. In principle, these are reasonable objectives. A modern pension service should be able to process routine transactions efficiently, provide accurate information and allow members to access their records without unnecessary delays.
But an important distinction must be made. The £83 million was a forecast of savings, not money already saved and independently verified.
The National Audit Office (NAO), which scrutinises public spending, identified weaknesses in how the expected benefits were to be measured. In particular, there were no agreed milestones sufficient to establish whether the promised improvements and savings would actually be delivered.
That raises an obvious question: if the financial benefits were central to the decision to outsource, why was there not a sufficiently robust system for measuring whether they were achieved?
Warning signs before the transfer
The problems did not emerge entirely without warning.
Before the service transferred, Capita had missed three important transition milestones. The Cabinet Office withheld £9.6 million in transition payments because agreed requirements had not been met.
Nevertheless, the transfer proceeded on schedule, with a simplified IT system and some functionality delayed.
Capita also said it inherited a backlog of 86,000 cases and more than 15,000 unread emails. These figures help explain the scale of the challenge facing the new provider, although they do not settle the question of who was responsible for each problem or how the transition should have been managed.
The previous administration was not without difficulties either. The NAO recorded 4,780 complaints about the service in 2024/25, the highest number since at least 2016/17.
The choice was therefore not between a perfect existing system and an obviously inferior replacement. The Government had a service with problems that needed addressing and a new contract intended to improve matters. The issue is whether it adequately assessed the risks of moving a complex pension operation to a new provider before the replacement system was ready.
When a service deals with people's retirement income, meeting a contractual transfer date should not take precedence over ensuring that the service can operate properly.
The human cost of administrative failure
For a pensioner, a delayed payment is not merely a disappointing performance statistic. It can mean difficulty paying rent, heating bills, council tax or other essential expenses.
Some affected members have required financial assistance while waiting for their pension arrangements to be resolved.
By August 2026, £22.1 million had been paid in transitional support loans to more than 3,900 members, according to the NAO.
These loans have helped people manage immediate hardship, but they should not be confused with compensation or with the final cost of the failure. They also do not answer the question of how many people have experienced stress, lost time or financial difficulty while trying to resolve their cases.
The ultimate cost to taxpayers is also still a wider question. It could include the expense of government intervention, repairing systems, managing complaints and backlogs, withholding contractual payments, and potentially transferring the service again.
A proper assessment must count all these costs rather than concentrate on the original tender price.
Who is accountable when outsourcing fails?
Private contractors can bring specialist skills, investment and technology to public services. Outsourcing can work well when the service is clearly specified, competition is effective and performance is properly monitored.
But a contract does not remove the Government's responsibility to the people who rely on the service.
The Cabinet Office chose the procurement route and remained responsible for overseeing the arrangement. Capita was responsible for delivering the contracted service. Those responsibilities are different, and both need to be examined.
It is not enough for officials to say that a contractor has failed to meet its obligations. If warning signs were visible before the transfer, the Government must explain why the transition was allowed to proceed and whether a delay would have been less damaging than moving to a system that was not fully ready.
Equally, the contractor must be held to the standards and obligations it agreed to meet.
Accountability should not stop at identifying a technical problem or blaming a backlog inherited from the previous provider. It should establish what went wrong, who made the relevant decisions, what action was taken when problems emerged and how similar failures will be prevented.
Is the Government bringing the service back in-house?
The Government has been considering the longer-term future of the scheme, including whether it should return to public-sector administration.
As of early October 2026, the position was still under review, with further evidence expected from an independent remedial adviser and a technical audit. That means it would be premature to describe a return in-house as a completed decision.
The Government has also announced a broader policy change. From April 2027, contracts worth more than £1 million are to be assessed for whether delivery in-house could be more effective before renewal.
That is a potentially useful development. But bringing a service back into government is not automatically a solution. It requires staff, technical expertise, sound systems and a carefully managed transition. A poorly planned return could create another period of disruption for the people the service exists to support.
The decision should therefore be based on evidence about the costs, risks and quality of both options, not on a desire to demonstrate that one model is ideologically preferable to the other.
The lessons for other public services
The Capita case should prompt a more searching examination of how government departments make outsourcing decisions.
Before awarding a contract, departments should publish a credible comparison of the full cost of outsourcing against in-house delivery. That comparison should include contract management, technology, staff, transition costs and the possible expense of bringing a service back into government.
Projected savings should be treated as targets to be tested, not benefits to be assumed. Contracts should contain measurable milestones, independent checks and meaningful consequences when agreed standards are not met.
For essential services, there should also be clear rules about readiness. If a new system cannot demonstrate that it can handle the workload, preserve records and make accurate payments, the transfer date should not be treated as an immovable deadline.
Finally, government should retain enough expertise to challenge its contractors effectively. It cannot manage a complex service properly if it has lost the knowledge required to understand what the provider is doing or to intervene when performance deteriorates.
These principles apply well beyond pensions. They are relevant to outsourced IT, benefits administration, public appointments, local authority services and other areas where a failure can affect large numbers of people.
Savings must be real, not just promised
The civil service pension affair is not proof that outsourcing never works. Nor does it establish that the previous provider would necessarily have delivered a better service.
It does, however, expose the risks of treating an attractive financial forecast as sufficient justification for a major change in a critical public service.
If the Government cannot demonstrate that outsourcing delivers the promised savings while maintaining or improving service quality, it must be prepared to reconsider the arrangement.
The £83 million forecast should now be tested against the full financial and human consequences of the transfer. The public deserves a clear account of what has been saved, what has been spent addressing the problems and what the long-term solution is expected to cost.
There is also a lesson for the wider public sector. Outsourcing should be a decision based on evidence, not an automatic preference for private delivery. Equally, in-house provision should not be assumed to be better simply because it is public.
The real test is whether the service works, whether the public receives value for money and whether those who depend on it are treated properly.
For civil service pensioners, that test is immediate. For taxpayers, the bigger question is how many other contracts have been judged by the savings promised at the beginning rather than the benefits actually delivered.
That is a question the Government should answer before awarding or renewing more contracts, not after another essential public service has run into trouble.