The Passport Delays of Summer 1999

Here you can browse the report together with the Proceedings of the Committee. The published report was ordered by the House of Commons to be printed 12 June 2000.

1. During the summer of 1999, many members of the public encountered great difficulty in obtaining passports from the United Kingdom Passport Agency (the Agency). At the peak of the crisis, maximum processing times in the Agency's regional offices ranged between 25 and 50 days, compared to the Agency's target of 10 working days. By June 1999, around 565,000 applications were awaiting processing, amounting to a backlog of over a month's work. Over 500 people missed their travel dates, many thousands were forced to travel to passport offices to obtain a passport, and many more suffered inconvenience and distress because they could not find out about what was happening to their applications. Public concern about the delays escalated until July 1999, when emergency measures—including free two-year extensions to passports—and a downturn in applications helped to ease pressure on the Agency. Processing times were brought back within the 10 day target by the end of August.[1]

2. The detailed chronology of key costs leading up from 1996 through to the problems of 1999 is shown at Fig 1.[2] In December 1999, the Home Office—the Agency's sponsoring department—announced an increase in passport fees. The increase was intended to put the Passport Agency on a sound financial footing and to fund improvements in customer service. The fee for a standard adult (10 year) passport went up from £21 to £28, and the fee for a child passport (5 year) rose from £11 to £14.80.[3]

3. On the basis of a report from the Comptroller and Auditor General, the Committee considered the events leading up to the crisis and the action being taken to prevent a recurrence. During the course of our inquiry, as an exceptional measure, we invited the former Chief Executive, the Accounting Officer of the Agency at the time, to give evidence. His evidence was of particular assistance to the Committee in gaining an understanding of events leading up to and during the crisis in the summer of 1999.

4. We consider that the Passport Agency's inability to provide an adequate service during the summer of 1999 represented a deplorable departure from the high standards of service the public has a right to expect. We believe this case offers salutary lessons for all organisations providing services direct to the public. The following main points emerged from our examination:


Figure 1: Chronology of key events

July 1996	Passport Agency decided to introduce the digital passport to increase security, replace their existing computer system and improve their efficiency[4]

June/July 1997	Passport Agency awarded contracts to Siemens Business Services and Security Printing and Systems Ltd

5 October 1998	New IT system and outsourced procedures introduced to Agency's Liverpool Office

9 November 1998	Passport Agency Management Board decided to roll-out new system and procedures to the Newport Office

16 November 1998	New system introduced to Newport Office

18 November 1998	Passport Agency Management Board decided to postpone introduction of new system

February/March 1999	Home Office expressed serious concern at developing situation. Action plan agreed

May 1999	Free two-year extensions to recently expired passports offered to public queuing at passport offices

Late June/early July 1999	Emergency measures introduced, including free passport extensions at post offices and call centre established

(i) Although over 500 people missed their travel dates, and thousands more were subjected to much anxiety and inconvenience, the Agency was still within the Home Office target of meeting 99.9 per cent of travel dates. We believe that this target was a poor reflection of the public's reasonable expectations for this service, and look to the Home Office to re-think how these expectations might be better reflected in the new performance standards to be set for the Agency (paragraph 10).

(ii) We pay tribute to the considerable efforts of many junior and middle ranking staff throughout the Agency. Many staff worked long and hard to ensure that the vast majority of travel dates were met, despite the difficult circumstances they faced. The problem in this case was primarily a failure in management at both the Agency and the Home Office, a fact now acknowledged by both organisations (paragraph 11).

(iii) There was a four month delay before testing of the new system was started, and testing of the new system's impact on productivity was not completed before it went live in late 1998. In addition, testing and initial implementation raised questions about the new system's effect on productivity, which were not adequately resolved. We found little evidence of any systematic evaluation of the risks of going ahead once these problems had emerged, or of timely consideration of alternatives. We emphasise the importance of sound risk-management arrangements, especially for projects where mistakes could lead to major costs or disruption for the public (paragraph 27).

(iv) The Agency took the risk of launching the new system in their largest offices, at Liverpool and Newport, which together accounted for half of the Agency's normal processing capacity. The problems at these two pilot sites resulted in the lost production of 400,000 passports in late 1998 and the first half of 1999, and the Agency were never able to make up for this significant loss in processing capacity. The decision to continue the roll-out of the new system to a second office, at Newport, was crucial because it further diminished the Agency's processing capacity when they already needed to make up for shortfalls in production in Liverpool. The Committee emphasises the need for pilot testing of new computer systems should wherever possible begin on a small scale and be rolled-out for testing at larger volumes only when initial tests prove satisfactory (paragraph 28).

(v) The Agency and the Home Office failed to foresee that, when news of the delays became known, the public would take the prudent step of submitting applications early, leading to a rapid rise in demand. Applications received between April and June 1999 were 29 per cent higher than forecast. Departments need to consider whether similar risks attach to other critical public services, and develop contingency plans for managing these risks (paragraph 29).

(vi) The Home Office's forecast of 5.1 million passport applications in 1999 was well short of the 5.6 million actually received in that year. The Department and Agency argued that the outturn might have been inflated to some degree by applications brought forward from future years. However, around half of the difference was caused by 300,000 more child passport applications than was anticipated. The Home Office is expected to act on their commitment to improve their forecasting (paragraph 30).

(vii) The Agency's contingency planning proved wholly inadequate, despite the lessons of the flawed implementation of the Agency's previous computer system in 1989. Although the further roll-out of the new system was halted in mid-November, it was not until the end of February 1999 that the Department was alerted, and a recovery plan instigated. The Home Office authorised recruitment of 300 additional staff in March, but it took time for them to be recruited and trained. Although output increased, it never caught up with the flow of new applications. It was not until the introduction of emergency measures, agreed in late June 1999, that the Home Office and the Agency were able to get on top of the situation and backlogs started to fall. We emphasise the need for adequate contingency plans in key public services, including plans to deal with substantial losses of production capacity (paragraph 31).

(viii) The Agency's business case, on which the decision to go for two parallel private finance contracts was based, assumed 4 million passports a year. However, even though the volume of passports is now expected to be 5 million or more a year—at least 25 per cent higher than was forecast when the contract was let—the unit price agreed with Siemens for each passport processed is fixed irrespective of the number processed. In normal circumstances higher volumes might be expected to bring economies of scale, higher productivity and lower costs. We look to the Agency to explore the scope for modifying the contract so that they can share the savings made by Siemens and pass them on for the benefit of passport applicants (paragraph 41).

(ix) At the time the contract was signed, the Agency had justified the option of running two private finance contracts on the grounds that it would offer a better transfer of risk than the other options considered in the business case. Yet, when serious risks crystallised in the crisis of 1999, the Agency were not able to hold Siemens liable for meeting anywhere near the full costs which arose. Siemens were obliged under the contract to pay only £170,000 of compensation for failure to meet specified standards of performance for the period October 1998 to January 2000. Of the extra costs totalling £12.6 million which the Agency incurred in 1999 as a result of the crisis, Siemens have agreed to pay £2.45 million over a number of years, and even this is more than Siemens consider to be payable under contract terms alone. On Siemens' view of the contract, it is clear that substantial business risk remained with the Agency. In any public-private partnership it is essential that the public sector should understand fully the residual risks which have not been transferred and ensure that these are properly managed (paragraph 42).

(x) In addition to providing additional capacity to prevent a repetition of the problems of 1999, the Agency are taking a number of measures designed to improve their service to the public, such as longer opening hours and a new call centre to handle telephone enquiries. The recent substantial increase in passport fees reflects in part the cost of these improvements, at around £3.50 per passport, as well as the higher costs of production due to the Agency's new system, some £2 per passport. We would have expected some of these costs to be absorbed through efficiency gains under the new arrangements. We therefore look to the Agency to ensure that these gains are realised and passed on to customers (paragraph 48).

(xi) The Home Office and the Agency have assured the Committee that the problems experienced during the summer of 1999 will not recur. We shall hold them to that pledge. We expect the Home Office to report back to the Committee in twelve months' time on the success with which the new system is implemented throughout the Agency's network of offices, and on the progress made towards improving customer service including the Agency's performance against their targets (paragraph 49).

(xii) The recent fee increase also includes a contingency provision of around £1.50 per passport which, if unused, will be used to pay off deficits of around £30 million accumulated by the Agency over the last six years. In large part, those deficits reflect a requirement for the Agency to meet costs incurred by the Foreign and Commonwealth Office on the provision of non-fee-bearing consular protection services and the costs incurred, in previous years, on the issue of passports overseas. We urge the Treasury to reconsider whether significant deficits built up over many years should be recovered from the fees to be paid by future passport applicants (paragraph 50).

(xiii) The Home Office were not alerted to the problems at the Agency until February 1999, almost three months after the decision to halt implementation of the new system. Departments should have arrangements for them to be alerted to significant problems affecting service delivery and quality in their agencies, without becoming involved in day-to-day management. It will be important for the Home Office to examine this issue as part of their quinquennial review of the Agency (paragraph 51).

6. A key source of frustration for many applicants during the first half of 1999 was the difficulty they had in trying to contact the Agency by telephone, seeking information on progress with applications that had been delayed.[5] By March 1999, an early stage in the crisis, over half of all calls were failing to get through to the Agency, compared to the 1.5 million calls that did get through. Between the beginning of March and the end of June, during the height of the crisis, there were around 3.5 million unsuccessful attempts to get through to the Agency by telephone.[6] It was not until early July that the Agency established a call centre to help alleviate the problem.[7] We note that in their recent White Paper "Modernising Government", the government established new standards for the services provided by central government, including a requirement for quick and helpful telephone facilities and clear and straightforward information about services.[8]

7. In addition to problems in contacting the Agency by telephone, many thousands of people had to travel, sometimes long distances and at great inconvenience, to a passport office to receive a passport.[9] Furthermore, over 500 people missed their travel dates. As a result of the crisis, the Agency received around 5,000 valid claims for compensation, including 510 in respect of missed travel dates.[10] For the calendar year 1999, compensation paid to the public totalled £345,000, of which £124,000 was for missed travel dates.[11] The crisis had other impacts on the quality of the Agency's work. Over the period, 2,000 unwanted duplicate passports were issued and the number of passports containing errors rose from 0.4 per cent to 0.6 per cent, as a direct result of the problems encountered by the Agency.[12]

8. Despite the problems, the Agency still managed to issue over five million passports during the calendar year. Ironically, this meant the Agency were still able to meet their target of meeting customers' declared travel dates for at least 99.99 per cent of passports issued.[13] However this target did not capture other key aspects of the service which were also of critical importance to customers.

9. The impact might have been very much worse had it not been for the efforts of many of the Agency's staff. From January 1999 onwards, staff worked significant amounts of overtime in an attempt to catch up on the backlog.[14] In evidence to us, the Home Office paid tribute to the hard work of the Agency's staff and the efforts made to ensure that most people received their passports on time.[15] We asked the Home Office and the Agency whether anyone at a senior level had lost their job because of the Agency's problems. The Home Office had considered the position of the Agency's Chief Executive, but he retired as planned on 2 September,[16] and no one else within the Agency lost their job as a result of this crisis.[17]

10. Although over 500 people missed their travel dates, and thousands more were subjected to much anxiety and inconvenience, the Agency was still within the Home Office target of meeting 99.9 per cent of travel dates. We believe that this target was a poor reflection of the public's reasonable expectations for this service, and look to the Home Office to re-think how these expectations might be better reflected in the new performance standards to be set for the Agency.

11. We pay tribute to the considerable efforts of many junior and middle ranking staff throughout the Agency. Many staff worked long and hard to ensure that the vast majority of travel dates were met, despite the difficult circumstances they faced. The problem in this case was primarily a failure in management at both the Agency and the Home Office, a fact now acknowledged by both organisations.

12. The origin of the crisis was the introduction of a new computerised passport processing system in two of the Agency's largest offices - Liverpool and Newport - in late 1998. The new system was intended to replace an ageing computer system and to introduce a more secure passport. Siemens Business Services, a major IT supplier with contracts with the public sector worth some £1.4 billion, were responsible for developing and providing the new computer system and for undertaking the initial processing of applications. Security Printing & Systems Ltd were responsible for printing and despatch of a new passport with a digital image of the passport holder.[18]

13. In April 1998 the government announced that children under 16 would in time be required to carry separate passports when travelling abroad. The new regulations came into force when the new system went live in Liverpool in October 1998, and were expected to generate up to 800,000 additional passports in the first year.[19] We asked whether it was sensible to introduce a new passport processing system at the same time as child passports. The Home Office told us that although the parallel introduction had produced practical difficulties, it was integral to the design of the new system. In their view, it was good value for money to make these changes together, because otherwise the software would have had to be rewritten soon afterwards to take account of the subsequent introduction of child passports.[20]

14. However, there were indications prior to 1999 that the Agency was finding it more difficult to cope with existing workloads and might struggle to cope with the introduction of a major new system. In the years leading up to 1999, the Agency had achieved substantial efficiency savings but had found it increasingly difficult to meet their targets during peak periods. Between 1991 and 1998 the Agency had achieved cumulative efficiency savings amounting to 33 per cent, including efficiency savings of 9.1 per cent in 1997-98 and 8.9 per cent in 1996-97.[21] However, between 1996-97 and 1997-98, maximum processing times during the peak summer period had increased from 9 days to 13 days (Figure 2).[22] By the Agency's own admission, they had left too little in reserve to respond when things went wrong.[23]

Figure 2: Maximum Processing times for passport applications, 1996-97 to 1998-99

Note (1): The graph for 1998-99 excludes the impact of the introduction of the new computer system from October 1998 onwards.

Maximum processing times during peak periods showed a year on year deterioration in the years 1996-97 to 1998-99

15. The design and development stages of the project were scheduled to take place between July 1997 and May 1998, with testing to start in October 1997 and finish in July 1998, well before the new system went live in Liverpool in early October 1998.[24] The first indication of a potential problem had come during factory-based testing during the summer of 1998. The problems had centred on the interface between the system and its users and its impact on productivity.[25] Further tests on productivity had been planned by the Agency but these were curtailed due to lack of time before the go-live date.[26]

16. One reason for some of the testing being cut short had been the later than expected start to the contract and delays during the design and development stages. The testing stage had started four months later than scheduled.[27] We asked whether any of the parties had requested a postponement to the go-live date, in view of the delays and the results of the initial factory tests on productivity. The Agency told us that they had been aware of the strong arguments for postponing implementation, but were equally aware that there was only a small window of opportunity for introducing the new system during the Agency's quiet season. Delaying the project at this stage would have meant postponing it for probably a year or more.[28] In the Agency's view, the results at the end of the testing programme, although they were not as detailed or completed as early as they had hoped, had suggested they should proceed.[29] With hindsight, Siemens told us that they now believed that more time should have been spent on testing productivity.[30]

17. The Agency had expected to roll-out the new system to all their offices between October 1998 and February 1999 (Figure 3). The Agency's aim was to install the new system before the seasonal rise in applications early in the New Year. The Agency expected each office to run down its existing workload prior to its roll-out date and transfer its incoming work to other offices that would be working normally. The aim was to maintain normal services to the public.[31]

Note: *The London Office deals with counter applications only. The Agency expected the London office to remain open throughout the implementation period and therefore no work needed to be transferred.

The Agency had expected to roll-out the new system to all their offices between October 1998 and February 1999

18. The Agency implemented their new computer system and procedures in the Liverpool office on time, on 5 October 1998. The Agency had expected output levels in the Liverpool office to reach about 30,000 issues per week by mid-November. However, when the new system went live at the Newport office six weeks later, output in Liverpool was just over 8,000 issues per week. It subsequently took nearer to six months, rather than the six weeks anticipated in the Agency's project plan, for output in the two offices to reach the levels achieved under the old system.[32]

19. The Agency had made a contractual commitment to their printers, Security Printing & Systems Limited, guaranteeing a minimum volume of two million new-style digital passports a year. If the volume fell below 2 million, the Agency would have been liable to pay the contractor compensation for unused capacity.[33] We were told that the guarantee had been offered because there was some doubt about the long-term future of the digital passport, and it was expected that the contractor would otherwise seek to charge a higher unit cost.[34] The Agency denied that the guarantee had been a factor influencing their decision to continue the roll-out to Newport, which would have brought the expected volume above the guaranteed level.[35]

20. Liverpool and Newport accounted for half the Agency's production. We asked why the Agency had decided to launch the system at these two offices rather than, for example, the smaller Belfast office. The Agency explained that they had wished to test the system at reasonable volumes, which in their view would not have been possible at the smaller offices. The Liverpool office was also near to the main computer centre where the new passports were issued. It was also where the Agency's IT specialists were based. By starting at the biggest offices the Agency had hoped to be able to increase production on the new system before the busy season started.[36] Siemens did not demur from this plan.[37]

21. Output at Liverpool and Newport increased during early 1999 but the increase was significantly slower than expected. To make up for this lost capacity, from January 1999 the Agency increased the output from their other four offices which were still operating the old system. However, this increase was not sufficient to keep pace with increasing demand. By the end of February, the Agency had serious concerns about their ability to cope.[38] In March, the Agency drew up a recovery plan which included the recruitment of 300 extra staff, based on what they felt was needed to get turn around time back to the target level.[39] However, these measures were still not sufficient to keep pace with a rapid increase in demand. It took until June 1999 for the Liverpool and Newport offices to reach previous production levels, and during this period this resulted in the lost production of 400,000 passports.[40] At the end of June 1999, the Home Office authorised emergency measures to alleviate the problems, including a call centre to deal with telephone enquiries, a further 100 staff and a facility to extend passports at post offices at no charge for up to two years.[41]

22. The Home Office told us that they first became aware of the problem when they received a copy of the Agency's quarterly report in early February 1999, almost three months after the decision to halt the roll-out, and called in the Agency to explain what had happened.[42] Until that point, Ministers had not been informed of the decision taken in the previous November to delay the roll-out.[43] In hindsight the Agency admitted that more might have been done to alert Ministers, but until that point they had thought they would be able to manage.[44]

23. Once the public began to realise that response times were lengthening, however, they prudently began to submit applications earlier. Between April and July 1999, applications exceeded the forecast by 29 per cent.[45] The Agency thought that some applications might have been brought forward from future years by people seeking to benefit from the offer of free two-year passport extensions.[46] The introduction of child passports had also generated some 300,000 passports more than expected.

24. The Agency's annual forecasting of demand is based on work undertaken by the Home Office's Research and Statistics Directorate. In 1999-00, however, there were 5.6 million passport applications compared to the Agency's forecast of 5.1 million.[47] The Home Office told us they are now trying to improve their forecasting, for example by establishing closer links with the travel industry, who also have to make these sort of forecasts.[48]

25. The Agency had experienced problems before, in 1989, when they computerised some of their clerical procedures. On that occasion, the Agency had implemented the new system in their Liverpool office, despite initial problems and lower productivity in their pilot office, Glasgow. The result was severe backlogs of passport applications.[49] The Agency told us that they had been aware of the problems in 1989 and had sought to learn the lessons. When they decided to go ahead with this project they had prepared a risk assessment alongside the business plan.[50] However, both the Home Office and the Agency now acknowledged that the contingency plans for managing those risks were inadequate.[51]

26. In the Agency's view, the key shortcoming in 1999 was the failure to build in sufficient capacity. They could have mitigated some of the risks by building in a higher contingency margin and bringing these contingencies in earlier. However, the Agency had believed that they had to keep the costs of the project to the minimum possible. In retrospect, they now saw that they had tried to complete the project on too tight a budget.[52] The Home Office agreed that too few resources had been set aside to manage the changes within the Agency.[53]

27. There was a four month delay before testing of the new system was started, and testing of the new system's impact on productivity was not completed before it went live in late 1998. In addition, testing and initial implementation raised questions about the new system's effect on productivity, which were not adequately resolved. We found little evidence of any systematic evaluation of the risks of going ahead once these problems had emerged, or of timely consideration of alternatives. We emphasise the importance of sound risk-management arrangements, especially for projects where mistakes could lead to major costs or disruption for the public.

28. The Agency took the risk of launching the new system in their largest offices, at Liverpool and Newport, which together accounted for half of the Agency's normal processing capacity. The problems at these two pilot sites resulted in the lost production of 400,000 passports in late 1998 and the first half of 1999, and the Agency were never able to make up for this significant loss in processing capacity. The decision to continue the roll-out of the new system to a second office, at Newport, was crucial because it further diminished the Agency's processing capacity when they already needed to make up for shortfalls in production in Liverpool. The Committee emphasises the need for pilot testing of new computer systems should wherever possible begin on a small scale and be rolled-out for testing at larger volumes only when initial tests prove satisfactory.

29. The Agency and the Home Office failed to foresee that, when news of the delays became known, the public would take the prudent step of submitting applications early, leading to a rapid rise in demand. Applications received between April and June 1999 were 29 per cent higher than forecast. Departments need to consider whether similar risks attach to other critical public services, and develop contingency plans for managing these risks.

30. The Home Office's forecast of 5.1 million passport applications in 1999 was well short of the 5.6 million actually received in that year. The Department and Agency argued that the outturn might have been inflated to some degree by applications brought forward from future years. However, around half of the difference was caused by 300,000 more child passport applications than was anticipated. The Home Office is expected to act on their commitment to improve their forecasting.

31. The Agency's contingency planning proved wholly inadequate, despite the lessons of the flawed implementation of the Agency's previous computer system in 1989. Although the further roll-out of the new system was halted in mid-November, it was not until the end of February 1999 that the Department was alerted, and a recovery plan instigated. The Home Office authorised recruitment of 300 additional staff in March, but it took time for them to be recruited and trained. Although output increased, it never caught up with the flow of new applications. It was not until the introduction of emergency measures, agreed in late June 1999, that the Home Office and the Agency were able to get on top of the situation and backlogs started to fall. We emphasise the need for adequate contingency plans in key public services, including plans to deal with substantial losses of production capacity.

32. The Agency started the procurement process in 1996, and the two contracts were approved by Ministers in June and July 1997. Over that period, the Agency prepared and updated a business case which considered five options. The Agency concluded that the best option was to run two parallel private finance contracts: one with Siemens to provide a new computer system and the initial processing of applications; and another with Security Printing & Systems Limited for digital printing and despatch of the new passport from a central site. The examination and authorisation of passport issues would continue as an in-house function.[54]

33. The Agency's business case, finalised in July 1997, was based on an assumed volume of 4 million passports a year, below the 5.6 million now expected in 1999-00 and the 5 million or more a year expected thereafter. The Agency tested a range of volumes and found that at volumes above 4.5 million a year it was cheaper to contract out only printing and despatch and to retain in-house all other functions including the development of a new computer system.[55]

34. We asked why the volumes used in the business plan had been so much lower than the volumes now expected. The Home Office told us that the figure of 4 million passports had not included a provision for child passports; the requirement to introduce child passports had not been announced by the government until April 1998, although as early as July 1997 they had been aware that the introduction of child passports was being considered by the government.[56] The Agency told us that at the time they prepared the business case other factors had contributed to uncertainty about future volumes, for example a debate about the introduction of identity cards which, if introduced, might have reduced the demand for passports.[57]

35. The Agency had believed that the option of two private finance contracts would offer a better transfer of risk, even though it was more expensive than an in-house solution at higher volumes.[58] The Agency told us that the difficulty with the in-house solution was that it did not involve the transfer of risk if anything went wrong.[59] They had doubted whether they had the capability to manage such a project in-house. This was a big, complicated IT project, and the Agency had relatively few IT resources and would have been relatively unskilled in tackling such a task.[60] There was also the strong encouragement in the Agency's framework document, from the Home Office, to involve the private sector in running their non-core business.[61] The Agency's former chief executive still believed that the decision to use the private sector was the right one and that the system would ultimately offer value for money.[62]

36. Under the contract, Siemens are reimbursed at a fixed rate, set at £2.72 in October 1999, for each passport processed by the Agency.[63] We questioned why the unit cost agreed with Siemens was fixed when higher business volumes might be expected to bring lower prices. The Agency told us that over the ten year period of the contract there was every possibility that volumes would fluctuate. This was foreseen at the time the contract was let and was a risk Siemens had accepted. However, we note that in the seven years leading up to the introduction of new system passport volumes had steadily increased by over 50 per cent and the volume of applications in future years is now expected to be 5 million or more, at least 20 per cent higher than anticipated in the Agency's business case and the deal made with Siemens. The Agency told us they had no plans to renegotiate the contract, but we note that it makes provision for savings to be shared with the Agency if the contractor achieves improved efficiency through changes to business processes.[64]

37. The cost of the additional measures taken by the Agency to deal with the crisis during 1999 was around £12.6 million. The Agency intend to meet these additional costs from efficiency savings they expect to achieve between 2001 and 2004.[65] The Agency have also announced that Siemens have agreed to pay a total of £2.45 million towards these costs, phased over a number of years.[66] These payments are in addition to compensation, known as service credits, paid automatically by Siemens for failure to meet agreed standards of performance—some £170,000 for the period October 1998 to January 2000.[67]

38. We asked where these efficiency savings were likely to come from. We were told that some savings would arise from the operation of a single system once the new system had been fully rolled out, and some would come from productivity gains as staff became more familiar with the new system. The Home Office and Agency agreed, however, that these were savings which would or should have occurred anyway, whether or not the crisis had happened.[68]

39. We questioned the level of risk transfer noting that the cost of the crisis being borne by the taxpayer is four times greater than that borne by Siemens. The Home Office's view was that the crisis was not due solely to the failure of the new system to deliver the expected levels of productivity. They therefore did not believe they could sustain a position that the IT supplier was wholly responsible.[69]

40. Siemens argued that they had paid the penalty for all the areas of risk that they had accepted in the contract. The cost of developing the new system had been borne by the Company and this cost would not be met until volumes reached a level to enable them to recover their investment. In their view, the technology had been delivered as required at Liverpool, had met all the requirements and was fully functioning.[70] In terms of productivity, they pointed out that responsibility for a large proportion of the issuing process remained with the Agency.[71] Any shortcomings on Siemens' part were, in their view, already covered contractually by the service credit regime.[72]

41. The Agency's business case, on which the decision to go for two parallel private finance contracts was based, assumed 4 million passports a year. However, even though the volume of passports is now expected to be 5 million or more a year—at least 25 per cent higher than was forecast when the contract was let—the unit price agreed with Siemens for each passport processed is fixed irrespective of the number processed. In normal circumstances higher volumes might be expected to bring economies of scale, higher productivity and lower costs. We look to the Agency to explore the scope for modifying the contract so that they can share the savings made by Siemens and pass them on for the benefit of passport applicants.

42. At the time the contract was signed, the Agency had justified the option of running two private finance contracts on the grounds that it would offer a better transfer of risk than the other options considered in the business case. Yet, when serious risks crystallised in the crisis of 1999, the Agency were not able to hold Siemens liable for meeting anywhere near the full costs which arose. Siemens were obliged under the contract to pay only £170,000 of compensation for failure to meet specified standards of performance for the period October 1998 to January 2000. Of the extra costs totalling £12.6 million which the Agency incurred in 1999 as a result of the crisis, Siemens have agreed to pay £2.45 million over a number of years, and even this is more than Siemens consider to be payable under contract terms alone. On Siemens' view of the contract, it is clear that substantial business risk remained with the Agency. In any public-private partnership it is essential that the public sector should understand fully the residual risks which have not been transferred and ensure that these are properly managed.

43. The Home Office and Agency assured us that the problems of summer 1999 would not recur.[73] The Agency plan to expand their office in Peterborough and open a new office in Durham to serve the North East. The latter is expected to require 500 new staff, including staff recruited by Siemens. Along with Peterborough, this expansion is expected to add 25 per cent to the Agency's capacity to process passports.[74] The Agency do not expect the new system to be rolled out to the remaining offices until autumn 2000 at the earliest. The Agency are also planning to introduce extended weekday opening hours and Saturday opening and perhaps ultimately a seven day per week service.[75] The Agency have opened a new telephone call centre in Bristol, to improve their ability to communicate with the public.[76] In the longer term, the Agency hope to make application forms available over the internet and, possibly, allow passports to be renewed by e-mail.[77]

44. As a direct result of the problems at the Agency, the unit cost of producing a passport will rise from the £12 expected in the Agency's business plan to £14, mostly because of extra staff recruited since March 1999 and the cost of running the old system in parallel with the new one. However, unit costs will also rise by £3.45 per passport to pay for the extra capacity and service improvements being introduced as a result of the crisis in 1999. A breakdown of unit costs is shown in Figure 4.[78] These costs are to be passed on to the public in the higher fees for passports from January 2000.

45. The recent fee increase of £7 on the standard adult passport also includes an element of £1.55 per application to meet contingencies, for example the extra costs incurred when dealing with a surge in demand. At an annual volume of 5 million applications this is likely to provide the Agency with a cushion of around £8 million a year. We were told that if this contingency is unused, it will instead be used to pay past deficits, totalling some £30 million, which have been accumulated by the Agency over the last six years. In large part, these deficits are the result of costs incurred by the Foreign and Commonwealth Office on the provision of non-fee bearing consular protection activities and, in previous years, the cost of issuing passports overseas. The Agency acknowledged that in effect new passport applicants in the UK were going to have to pay for not charging enough for passports issued overseas in the past.[79]

46. Although the Home Office have an Advisory Board to assist them in overseeing the work of the Agency, they were unaware of the Agency's difficulties until a late stage. The Home Office believed that this case offered important lessons about how departments should oversee the work of their agencies.[80] The Advisory Board was a strategic body that looked at the annual plans and targets.[81] Ministers were not expected to be involved in the day to day management of the Agency. This approach worked as long as things were going well, but departments needed to be kept fully informed.[82] The Home Office told us that they were examining the need for a closer link between the Agency and the Department, as part of a quinquennial review of the Agency now under way.

47. Our recent report on the Home Office's Immigration and Nationality Directorate's Casework Programme highlighted similar difficulties to those at the Passport Agency, again involving the introduction of a new computer system.[83] The Home Office Accounting Officer told us that he now receives a regular report on all computer contracts in the Department and their agencies. In that report, a system of "traffic lights" showing green, amber and red is used to signal the state of each project. The Home Office's intention is to get all projects to "green" as quickly as possible.[84]

48. In addition to providing additional capacity to prevent a repetition of the problems of 1999, the Agency are taking a number of measures designed to improve their service to the public, such as longer opening hours and a new call centre to handle telephone enquiries. The recent substantial increase in passport fees reflects in part the cost of these improvements, at around £3.50 per passport, as well as the higher costs of production due to the Agency's new system, some £2 per passport. We would have expected some of these costs to be absorbed through efficiency gains under the new arrangements. We therefore look to the Agency to ensure that these gains are realised and passed on to customers.

49. The Home Office and the Agency have assured the Committee that the problems experienced during the summer of 1999 will not recur. We shall hold them to that pledge. We expect the Home Office to report back to the Committee in twelve months' time on the success with which the new system is implemented throughout the Agency's network of offices, and on the progress made towards improving customer service including the Agency's performance against their targets.

50. The recent fee increase also includes a contingency provision of around £1.50 per passport which, if unused, will be used to pay off deficits of around £30 million accumulated by the Agency over the last six years. In large part, those deficits reflect a requirement for the Agency to meet costs incurred by the Foreign and Commonwealth Office on the provision of non-fee-bearing consular protection services and the costs incurred, in previous years, on the issue of passports overseas. We urge the Treasury to reconsider whether significant deficits built up over many years should be recovered from the fees to be paid by future passport applicants.

51. The Home Office were not alerted to the problems at the Agency until February 1999, almost three months after the decision to halt implementation of the new system. Departments should have arrangements for them to be alerted to significant problems affecting service delivery and quality in their agencies, without becoming involved in day-to-day management. It will be important for the Home Office to examine this issue as part of their quinquennial review of the Agency.

The Passport Delays of Summer 1999 - HC 208

28/06/2000