The Home Office's UK Border Agency took risks with public money to build new systems on time to support the Labour government’s new point-based immigration programme.
But the report by the National Audit Office, which is published today, is not particularly critical of the IT programme.
“The [implementation] programme had a chequered history, starting poorly but finishing well,” said the NAO.
The NAO found that officials paid Fujitsu £4m for developing applications that were not used; and that Fujitsu was appointed after only limited competition.
Fujitsu ended up trying to design a “complex system whilst legal, policy and business requirements for the system were still being resolved”. Officials at the UK Border Agency told the NAO that Fujitsu underestimated the complexity of the project and did not have the right staff in place for a fast start-up. Fujitsu disputes this.
Conservative MP Richard Bacon, a member of the Public Accounts Committee, said the IT was introduced in a rush:
“The UK Border Agency’s Points Based System struggled to get off the ground because of poor management of new computer systems and tight deadlines.
“The rush to deliver the Points Based System on time caused the Agency to start development work before the overall scheme had the Home Office’s approval.
“Computer firm Fujitsu began designing new software before the UKBA had decided what sort of system it needed and, in order to save time and money, the Agency stripped out important parts of the sponsor management system.
“Unsurprisingly, this decision has weakened the UKBA’s ability to manage sponsors.”
Bacon said the UK Border Agency needs to “vastly improve its IT systems so that they will provide much better management information, particularly the identification of individuals on expired visas”.
The Home Office plans a new casework IT programme - but the department has a history of failure of IT-based caseworking programmes.
The NAO concludes:
"The Points Based System doesn’t deliver its full value for money potential, while poor data and monitoring hampers the removal of failed applicants."
These are some of the main IT-related points in the NAO report:
“To meet the requirements of the [points-based immigration] System, in 2005 the Department set up a programme to develop supporting IT systems, caseworking processes, rules, guidance and forms at a cost of £35m.
“The programme had a chequered history, starting poorly but finishing well. The original business case forecast savings of £224m over 15 years, based on the assumption that all existing caseworking systems, home and abroad, would be replaced.
“When, in 2007, caseworking systems were taken forward as a separate programme on a later delivery schedule, forecast savings fell to £65m. The Office of Government Commerce gave an amber rating at its first assessment in 2005.
“In late 2006, the scope of the programme was unclear while the Department decided whether to delay the caseworking element of the programme.
“To try to deliver on time, the Programme Management team decided to limit competition for the IT development work to its three existing IT suppliers and to move commercial arrangements and software development ahead of formal approval.
“In April 2007, when the Home Office General Investment Group finally gave approval for the programme, it did so on the basis that development costs were reduced by a third, leading to a further reduction in scope.
“The IT elements of the programme were delayed from the start mainly because the winning supplier (Fujitsu) was trying to design a complex system whilst legal, policy and business requirements for the system were still being resolved.
“The Agency also believes that Fujitsu underestimated the complexity of the project and did not have the right staff in place for a fast start-up, although this is disputed by Fujitsu.
“To save time and money, the Agency stripped out planned functionality from the System; for example: its capability to flex the points requirements and to extract management information. These changes increased its reliance on Fujitsu to make changes to the system.
The Agency paid Fujitsu £4m for work it had carried out in 2007 developing applications which were not used.
The Office of Government Commerce said in June 2008:
“The programme suffered from serious failures in IT supply in the Autumn of 2007 and although pragmatic measures have been taken to re-phase the project which achieved a successful (but de scoped) launch in February 2008, the programme is still in recovery with many aspects requiring remedial work, proper integration and firmer control.”
"The Agency worked hard to meet tight deadlines for other aspects of implementation, including preparation and piloting of application forms and guidance and staff training. Policy guidance to staff has improved in clarity since the System was launched.
“ The Agency has not taken enough systematic action to ensure, where it can, that migrants leave the UK when they are no longer entitled to remain. This has been due partly to a lack of exit checks, making it difficult to identify overstayers, and to IT systems which cannot identify individuals needing to renew their visas.
“The Agency estimates there may be up to 181,000 migrants in total (not just entering through the System) in the UK whose permission to remain has expired since December 2008.
“It expects to revise this estimate downwards, however, following matching with new data being provided by its e-Borders project. In March 2009, its case ownership team for the North East, Yorkshire and the Humber region started to contact refused applicants to explain that they should leave. The team reports that it has encouraged around 2,000 migrants to leave voluntarily since then. The Agency is now rolling this out to other regions.
“ The Agency’s handling of migrant applications made in the UK is inefficient, mainly because of poor legacy IT. Currently, migrant application caseworkers in the UK have to access many different databases adding, on average, an extra 20 per cent to the time taken to decide an application.
“A new casework IT programme is planned to improve the handling of migrant applications both in the UK and at overseas posts. The Agency’s new sponsor licensing software is also not fully efficient or fully accessible to staff outside Sheffield (where the sponsor licensing unit is based), but there are no definite plans for an upgrade.”
It cannot be said with credibility that central government doesn't take risks with public money. Building a new IT system in a rush to meet a new government policy is analogous to planning for a disaster.
An innovative approach here might have involved trial and error but in the end might have cost a lot less than £35m, and perhaps done a better job of helping to keep track of applicants.
After all the system is not handling millions of applications: between February 2008, when a phased implementation of the points-based system began, and December 2010, the Agency handled only 445,000 migrant applications. Does that justify a £35m IT spend - especially as the £35m is not future-proofed?
It's time the civil service did things differently, taking into account the excellent suggestions in the recent report of the Institute for Government, System error: Fixing the flaws in government IT.
Immigration - the Points-based system - NAO report