PAC: Government Flagship Digital Identification System failing its users

REPORT SUMMARY

GOV.UK Verify (Verify) was heralded as a flagship digital programme for government, but three years after it went live, the programme has not delivered value for money and members of the public using the system have been hampered by a catalogue of problems, including difficulty signing up and accessing multiple government services.

Verify clearly demonstrates many of the failings we see all too often on large government projects: expectations were over-optimistic from the start, key targets have been badly missed and results simply not delivered. 

Only 19 government services have adopted Verify, fewer than half the number expected; and only 3.9 million people have signed up as Verify users, less than one-sixth of the forecast 25 million users by 2020. Some of the most vulnerable people using the system– such as those applying for Universal Credit – are among the worst affected.

Despite over 20 internal and external reviews, the Government Digital Service (GDS) and the Cabinet Office have failed to get the programme on track. Six months after announcing that public funding would stop in March 2020, GDS and the Cabinet Office have not resolved major uncertainties about how Verify will operate beyond that date.

COMMENT FROM PAC CHAIR MEG HILLIER MP

“Three years after GOV.UK Verify was introduced, the system is failing its users and struggling to meet key targets.

“Key government departments do not want to use the system and members of the public are facing problems signing up.  

“Once again, the Government has not delivered on a project that was over-ambitious from the start. This is a verdict the Public Accounts Committee are making all too often on large government projects.

“The Government has now decided to stop public funding to Verify in 2020. Before then, it has a duty to get this programme working properly for existing users, such as people claiming Universal Credit, and set out a plan of action for when public funding ends.”

CONCLUSIONS AND RECOMMENDATIONS

1 GDS has failed to meet any of its original performance targets for Verify and vastly overestimated the benefits it could achieve. GDS set over-optimistic performance expectations for Verify. It forecast that 25 million people would be using Verify by 2020, when in fact only 3.9 million were using it as of March 2019 – less than one sixth of the target. GDS also predicted that at least 46 government services would adopt Verify by March 2018, but at March 2019 only 19 services were using it to verify people’s identities. Poor performance has directly affected the level of expected benefits, which are based on anticipated take-up. Compared to its original estimate of £2.5 billion of benefits over 10 years, GDS is now estimating £366 million of benefits over Verify’s lifetime. Not only is this a striking reduction, we are also sceptical about the validity of benefits claimed, given that the National Audit Office could not replicate the benefits using data supplied by GDS. An unknown amount of costs have been shunted onto other government departments, such as the costs of reconfiguring their systems to use Verify and additional manual processing costs for people unable to use Verify.

Recommendation: The Cabinet Office should write to us before the summer recess setting out the lessons it has learned from the failure of the Verify programme, and what steps it is taking to prevent similar failures in future.

2 People using Verify have been badly served by an onerous system that is not fit for purpose. Only around half of people who attempt to sign up to Verify succeed in doing so in a single attempt. It is even lower for people trying to apply for Universal Credit through Verify, since they are often less likely to have the documents required such as passports. Despite Universal Credit being Verify’s biggest customer, just 38% of Universal Credit claimants can successfully use Verify when applying for the benefit. As well as being unfit for purpose, this is incurring extra costs for the taxpayer: the Department for Work & Pensions expects to spend around £40 million over 10 years on processing applications for Universal Credit manually. In addition, signing up as a Verify user does not guarantee that people will be able to access the government service they wanted in the first place. Verify users can be ‘locked out’ if pre-existing data held on them by the service does not match their Verify details. 

Recommendation: GDS should, by the 2019 summer recess, write to us setting what changes are being made to Verify to better support people claiming Universal Credit.

Recommendation: The Cabinet Office, GDS and Department for Work & Pensions should agree and set specific targets for the number of people they aim to get applying for Universal Credit through Verify.

3 GDS’s inability to get buy-in from departments ultimately led to Verify’s failure. GDS failed to develop a product that departments wanted to use, and its later efforts to increase government usage (for example, by offering Verify to local authorities and the NHS) floundered. The Cabinet Office Permanent Secretary acknowledged that GDS did not convince enough departments to adopt Verify to make it work. HM Revenue & Customs, for example, felt that Verify did not meet its needs, and continued to use and develop its own Government Gateway system; only around 4% of its customers use Verify to access its services. We see this as another case of the Cabinet Office coming unstuck after seeking to impose a central initiative on unpersuaded departments – shared services being a notorious past example.

Recommendation: The Cabinet Office needs to secure the commitment of departments to cross-government programmes. In the initial business case for such programmes, it should outline how it will get buy-in from departments and other key stakeholders, and what action they will take should departments go back on their commitments. We expect to see such plans in place when we examine cross-government programmes in future.

4 Verify was characterised by poor decision making by the Cabinet Office and GDS, compounded now by their failure to take proper accountability. The programme lacked strong leadership and oversight, despite being subjected to over 20 internal and external reviews. The Cabinet Office Permanent Secretary said he was sorry the Cabinet Office did not intervene earlier. However, even now, the Cabinet Office and GDS are not taking proper accountability for the programme’s shortcomings. The witnesses did not take seriously enough their responsibilities to explain and account for why the programme failed to meet its original goal of providing a single identity assurance service across government. They chose instead to emphasise the work currently being done to develop a wider market for identity verification services. It is simply not good enough to rewrite the programme’s aims after the event to try to explain away underperformance. This Committee’s previous report on accountability for taxpayers’ money makes clear that Accounting Officers have a primary duty to ensure projects are a good use of public money and consequently must raise value for money concerns at the appropriate time.

Recommendation: For its projects at risk of failure, the Cabinet Office should ensure Accounting Officer (AO) assessments are conducted at the proper time. It should provide us with an update on how many AO assessments have been undertaken in the last 12 months, and what actions it has taken as a result of these.

5 The Cabinet Office and GDS have no meaningful plan for what will happen to Verify post-2020. Major uncertainties remain about how Verify will operate once public funding stops in 2020. We are seriously concerned that the assurances the Cabinet Office gave us for the future were rooted in what, at this stage, are unknowns. The uncertainties include how the market for verifying people’s identities will develop and the price that will be charged for verification services. This is not a strong foundation for the future of a flagship digital programme. There is a clear risk that government services using Verify could face large and unaffordable cost increases from 2020 if the market price is significantly higher than the current contract price. GDS expects the Crown Commercial Service to negotiate an attractive bulk rate for government with providers. But government is essentially ‘betting’ on a private sector market emerging that can provide services more cheaply than anything it could build or buy for itself.

Recommendation: Alongside its Treasury Minute response, the Cabinet Office and GDS should write to us by the summer recess setting out the detailed plan for how Verify’s services will be maintained after 2020, including how government services using Verify will be protected from unaffordable cost increases.

6 The Cabinet Office and GDS have not protected taxpayers’ interests in securing Verify’s intellectual property. Verify’s intellectual property includes the Verify brand, its public portal and the infrastructure ‘hub’ that links Verify’s users, providers and government services. The Cabinet Office and GDS seem to have given little thought to the value of this intellectual property, and how taxpayers’ investment in it would be recouped should private providers secure substantial profits from Verify in the future. They have not yet had discussions about what will happen to Verify’s intellectual property, including whether it will be sold or given to providers.

Recommendation: The Cabinet Office and GDS should take urgent action to clarify the value of Verify’s intellectual property, to protect the interests of taxpayers. They should detail this in the plan we have requested by the summer recess.