House of Commons - Public Accounts

Select Committee on Public Accounts Tenth Report

TENTH REPORT

The Committee of Public Accounts has agreed to the following Report:

INDIVIDUAL LEARNING ACCOUNTS

INTRODUCTION AND LIST OF CONCLUSIONS AND RECOMMENDATIONS

1. The Government introduced Individual Learning Accounts in England in September 2000, to widen participation in learning and to help overcome financial barriers faced by individuals. Accounts were to be available to everyone, including the self-employed, and were to be used to help pay for learning of the learner's choice. At the same time, the Government was keen to target people with particular learning or skill needs; for example, young people without qualifications and in low skill jobs, employees in small firms and those seeking to return to work. The scheme was to be funded from £127.5 million released from the wind-down of the Training and Enterprise Councils together with additional funding of £23 million (subsequently increased to £40 million) and £46 million in 2000-01 and 2001-02 respectively.[1]

2. The scheme was successful in attracting over one million people back into learning. However, in November 2001, a fortnight after announcing the planned suspension of the scheme with effect from December, the Government withdrew it following allegations of fraud and abuse. Total expenditure is likely to exceed £290 million against a budget of £199 million (Figure 1).[2] The scale of fraud and abuse could amount to £97 million, including £67 million fraud.[3]

Figure 1: ILA Expenditure on National Delivery (£ million)
	Income	Expenditure
	DfES Budgets & TEC contributions for ILAs (1)	Incentives for training	Delivery costs (2) - Capita contract	Policy development/ programme support (consultancy/ evaluation)	Total (3)	Forecast for final outturn and overspend.
2000-01	70.0	51.8	15.6	2.6	70.0	70.0
2001-02	129.0	183.1	20.5	1.1	204.7	222.6
2002-03	0.0	0.2	1.5	0.5	2.2	0
Total	199.0	235.1	37.6	4.2	276.9	292.6
Budget	199.0				199.0	199.0
Overspend					77.9	93.6

Notes

(1) Includes the TECs' contributions to the budget for ILAs delivered by the national framework, i.e. after deducting cost of locally delivered ILAs.
(2) In 2000-01 some ILAs were delivered by a unit set up by TECs in the South-East. Figure includes £264,000 in management fees in 2000-01.
(3) Payments at closure of 01-02 Accounts (30 October 2002).

3. On the basis of the Comptroller and Auditor General's Report, the Committee examined the management of risk in designing and implementing the scheme, the effectiveness of monitoring, the Department's relationship with Capita, who operated the scheme, the level of fraud and abuse and the actions taken, and the lessons learned. We draw four main conclusions.

The Department had 5 years to put in place arrangements to implement the Government's commitment to have 1 million individual learning accounts by March 2002. While the Department undertook extensive piloting to test innovative schemes, these did not provide workable solutions, and the scheme implemented was not well thought through or tested and was implemented in too short a time. While the tight timetable was of the Department's own making, after the pilots did not work it should have re-planned the project and ensured full testing before implementation, rather than over-ride sound project and risk management.

Good risk identification and management is essential in major projects, especially those that involve innovative solutions. However, the Department's risk assessment and risk management were not fit for purpose, and were driven more by concerns that the scheme would not attract sufficient new learners. As a result, the Department did not give enough weight to advice received on the risks of fraud and abuse and about quality of training. Had the Department done so, and drawn on the work of this Committee and experience of counter-fraud strategies elsewhere in the public sector, it would have been able to build counter-fraud measures into the design of the scheme rather than react to events.

To encourage new providers and new learners, the Department decided to minimise bureaucracy, including checks on learners, on providers and on the quality of learning. They should have matched this innovation with more rigorous and incisive monitoring downstream, but failed to do so. As a result, they were slow to identify emerging problems, including substantial fraud and abuse.

The Department contracted the operation of the scheme to Capita. However, the contractual arrangements were weak and the Department and Capita did not develop the partnership arrangements necessary for success. This was a major factor in the problems that arose and also meant that the Department bore more of the key risks than planned.

The Accounting Officer was frank in acknowledging the shortcomings of the scheme, and in regretting that they had occurred. He was however less able to assist the Committee in getting at the specific reasons why they had been allowed to occur. Acceptance of responsibility is important, but so is the need for Accounting Officers to provide convincing assurance that weaknesses have been properly analysed and understood so that the necessary improvements can be identified and implemented.

4. Our more specific conclusions and recommendations are as follows.

(ii) Because the Department wanted to encourage people back into learning, and new providers to enter the market, it decided not to impose quality controls. To rely instead on learners and market forces to ensure that inefficient or effective providers were replaced was naïve, given that many of the people the Department was trying to encourage into learning were those least able to assess the quality of the training on offer. It also ignored evidence from the pilots that the scheme was most likely to be successful where new learners had advice from intermediaries such as community groups and trades unions.

(iii) In its current review of adult learning provision, the Department should ensure that potential learners have sufficient information and advice to make reasoned choices about the courses on offer. It should also consider accrediting providers, as the public tend to presume that learning funded under a Government initiative has its endorsement.

(iv) In the absence of quality assurance, the decision to give a positive incentive to providers to recruit learners was fundamentally flawed. It encouraged a substantial number of unscrupulous providers to undertake aggressive marketing, and combined with control weaknesses created the environment for fraud and abuse. Innovative schemes such as this should be more rigorously piloted and tested to identify the risks, and allow effective risk management.

(v) Although the Department commissioned a demand model for the number of accounts, the absence of a business model, sensitivity analyses and contingency arrangements, left it poorly prepared to take action when demand exceeded the funds available. In the event, the reactive steps taken to stem demand were ineffective, as providers found ways around them. The Department should build the need for sound business planning into its procedures for all major projects, especially those including substantial innovation.

(vi) Because of ineffective monitoring, the Department was not aware of unusual patterns of activity, including very large payments to individual providers. Only belatedly did it realise that the apparent "success" of the scheme, as reflected in spending over budget, was an indicator of significant fraud and abuse. As part of revised risk management procedures, the Department should develop its monitoring arrangements, including exception reporting especially on those areas highlighted in risk assessments.

(vii) Capita could have done more to insist that its concerns about risk of fraud and the necessary controls were taken seriously, but felt restricted by the lack of a place on the Project Board. All departments should ensure that private sector partners are integrated effectively into project management arrangements, and that partners can escalate concerns to senior staff, including the Accounting Officer.

(viii) Although the contract involved risk sharing, it did not clearly define areas where risks were shared and in practice more risks remained with the Department than planned. In any future contracts risks and any risk sharing should be clearly identified and effective, drawing on guidance from the Office of Government Commerce.

(ix) Security arrangements were not clearly specified in the contract, concerns raised during procurement were not addressed, guidance was not followed, usage monitoring was inadequate and the robustness of Capita's systems was not independently tested. While there is no evidence of unauthorised access to the systems, a major flaw was the ability of authorised providers to access unused accounts and in one case to offer them for sale. Responsibilities and requirements for IT security should be spelt out clearly in contracts, any concerns raised during procurement should be addressed, and IT procedures should be rigorously tested.

(x) It may be two years before the level of fraud and abuse is clear and action against those involved completed. The Department should follow through this work rigorously, and should also keep its Counter Fraud Strategy and revised counter-fraud measures under review.

(xi) Internal Audit was involved in the project from the outset, as part of the project board. But it did not provide the independent check and assurance that the Accounting Officer needed. In part, this appears to be because it accepted management views about the need for light touch controls, to ensure the effectiveness of the scheme, and did not follow-up concerns raised at various stages about the risks involved. In addition, its planned review of Capita's systems was deferred. The Department should look again at the way it involves Internal Audit in projects. The aim should be to obtain their advice on risks and risk management at key stages on design and implementation, though those involved would need to be independent of subsequent internal control reviews.

MANAGEMENT OF RISK IN DESIGNING AND IMPLEMENTING THE SCHEME

5. The Department tested a number of different ways of implementing Individual Learning Accounts in a series of pilots over more than two years, but then decided that they were unsuitable for a national scheme and implemented a different model.[4] This squeezed the time available to meet the manifesto commitment to deliver 1 million learning accounts by 2002. As a result, despite a significant change of direction at the end of 1999 the project was not re-planned, and there was insufficient testing of the proposed scheme.[5]

6. Risks were highlighted during the design and pilot stages. However, the design was not informed by a formal risk analysis. Although the Department compiled a risk register and reviewed it monthly, it was not put together on a systematic basis and not all issues were actively pursued to ensure risks were minimised. The Department's assessments that risks were low were influenced by the low value of individual transactions, despite past experience of the vulnerability to fraud in franchised and distance provision identified in this Committee's report on Halton College, and advice from the Further Education and other sectors. Neither did the Department draw on experience elsewhere in the public sector on countering fraud. The fraud risk was not re-assessed as high until summer 2001, a few months before the scheme was shut down.[6]

7. The pilots also identified concerns about the lack of quality assurance arrangements, reinforced by the Learning and Skills Council and some providers. However, the Government wanted to encourage more flexible delivery of learning through a wider range of providers and, in particular, those operating in niche markets and those attracting new, non-traditional learners. So, in order to keep the scheme administratively simple, the Department decided not to impose quality controls. Instead it decided to rely on learners and market forces to ensure that inefficient or ineffective providers were replaced. Thus, while the Department required providers to be registered with the Individual Learning Account Centre, it did not subject them to quality assurance. By November 2002, there were 8910 registered providers, some of which were new ventures with no previous involvement in publicly funded education or training. The Department consider that the vast majority of providers were good, or new providers who had the potential to provide different types of training. But there had been a group who abused the scheme.[7]

8. Learners could choose with whom to undertake their learning, but those whom the Department wanted to attract were least likely to be able to compare different providers. In some small-scale pilots, this had been overcome where intermediaries, such as community groups and trades unions, provided the advice learners' needed. However, the Department's decided to concentrate the marketing of learning accounts via learning providers. Research suggests that 45% of account holders first heard of the scheme from providers and many courses were not appropriate to learners' needs. The Department accept that the decision not to have more checking of providers was the fatal flaw in the arrangements.[8]

9. The Department's Internal Audit was involved in the project board overseeing the development of the project. It was instrumental in requesting a project health check in March 2000, which highlighted risks in the arrangements. These included the difficulty in estimating demand and timing, shortage of relevant skills in the Department to manage such a large project, the absence of project quality control procedures, gaps in the risk register and inadequate contingency planning. However, Internal Audit had not escalated any remaining or subsequent concerns to the Accounting Officer. The Accounting Officer told us that this was because Internal Audit had taken into account the Department's decision to take some risks, and have "light touch rules" to get more people into learning, and like others in the Department did not at the time believe that was a wrong decision. He concluded that while he had a lot of confidence in Internal Audit, in this case they had not provided the independent assurance they should have.[9]

MONITORING THE OPERATION OF THE SCHEME

10. The Department targeted its marketing at specific groups, such as people with particular learning or skill needs, young people without qualifications and in low skill jobs, women returners to work and ethnic minorities. However, it did not set quantified targets for them, because the aim was not just to reach these groups but to encourage all types of people to come back into learning. Consequently, while there were some follow-up surveys as the scheme progressed, there was no monitoring of people coming into the scheme.[10]

11. The Department commissioned a detailed demand modelling exercise in Spring 2000, which estimated that 1.3 million accounts would be opened in the first year rising to 1.9 million accounts by 2005. But there was no sensitivity analysis and no business model, and despite the forecast number of accounts, the Department did not develop contingency arrangements if demand exceeded their plan for 1 million accounts. Capita did draw up business plans which included the possibility of higher volumes, but the Department took little notice of them because it expected to have difficulty attracting learners, based on pilots involving Training and Enterprise Councils in 1998 and 1999.[11]

12. As a result, the popularity of Individual Learning Accounts took the Department by surprise. Even though by the end of April 2001, 781,000 accounts had been opened in England, the Department was still considering a marketing strategy to encourage take-up and it was not until early summer 2001 that the Department recognised that the target number of learners, and the budget, would be exceeded.[12]

13. Capita provided weekly, monthly and annual activity reports on the number of accounts opened, expenditure, number of complaints and performance against agreed service targets. But the Department did not employ sufficient resources to review them. Crucially, Capita did not include any "exception reports" highlighting unusual items of activity or particularly large claims. The lack of effective review of management information stemmed from the mistaken assessment of the risk of fraud from the outset. It meant that the Department was not aware of very large payments to some providers (20 providers claimed over £1.5 million and two more than £6 million) until some six months later than it should have been.[13]

14. Concerns about aggressive marketing, such as stopping people in the street and offering them incentives to open accounts, arose as early as December 2000. However, while the Department suspended both providers it did not prosecute on legal advice, because the providers had been exploiting loopholes in the scheme rather than committing fraud. At the time, the Department did not think this was replicated across the rest of the scheme, but it became a feature in 2001, and should have been spotted and dealt with earlier.[14]

15. The Department accepts that as the scheme had been designed to minimise the bureaucracy involved, downstream monitoring should have been stronger. Monitoring was inadequate and key signals had not been picked up, because the Department had been focused on getting to the one million target, and a lot of the risk management that took place was directed towards that. The Department saw the rising number of accounts as a success in stimulating learning rather than a problem. Initially the budget overspend raised the alarm, and only when the Department looked at this alongside the other evidence did it see the other problems growing in the scheme. The action taken to stem the growth of the scheme, including ending the £150 incentive scheme in July 2001, failed as people found other ways to attack the system, particularly the 80% discount arrangements. The Department closed the scheme when it became clear that it could not be stopped in any other way.[15]

THE RELATIONSHIP BETWEEN THE DEPARTMENT AND CAPITA

16. To encourage innovation, the Department adopted a public-private partnership approach for the design and implementation of the scheme. However, by January 2000, after seeking competitive tenders, only one bidder remained. The other bidders dropped out partly because of the limited time available to implement the scheme. Capita told us that they had remained involved because they had the infrastructure to meet most of the Department's requirements, had a good working relationship with the Department, and had a good track record of implementing projects in a speedy way. The Department considered whether they should take action to encourage more bidders to stay in the competition, but concluded that Capita were a competent company that could run the system, and decided to go with them. Not to go with Capita would also have risked delivery of the manifesto commitment.[16]

17. In June 2000, the Department signed a contract with Capita to develop and operate the scheme. Capita was to operate a call centre for enquiries about accounts as well as an administrative centre for registering learners and providers, processing new accounts, maintaining records of learning started and notifying the Department of amounts owing to providers.[17] Project risks were shared between the Department and Capita; for example Capita carried the risks of delivery of the technical solution to tight timescales, and of providing the capacity to support demand levels. But in practice more of the risks remained with the Department than planned, and there was insufficient clarity in areas where risks were shared with Capita.[18]

18. Although the Department envisaged a partnership with Capita, the relationship developed as a more traditional purchaser/supplier arrangement. The Department excluded Capita from the project board, even though Capita asked to be there, because its presence would have restricted open discussions of policy, and there was no sharing of risk assessment and management or discussion at senior levels about what was happening on the scheme. The Department considered that Capita might have shouted louder about emerging issues at various points. The Chief Executive of Capita agreed that he should have escalated issues (for example on the distribution of over eight million blank forms to providers) to the Accounting Officer and even to the Secretary of State, although this was difficult to do in public private partnerships.[19]

19. Capita devised its systems on the understanding that all providers would be accredited or registered with the Further Education Funding Council, Training and Enterprise Councils or awarding bodies, and pre-registered with the learndirect learning opportunities database. However, when the scheme was launched this database was not in place, and was never used for this purpose because of incompatibilities between records, and because the Department was not intending to accredit providers.[20] The specification also required the appointed contractor to carry out checks on the eligibility of the learner and of learning support. But as the learndirect database was still under development, the contract included interim provision for providers to self-certify. When the Department dropped the requirement for providers to be registered with the database, the interim provisions were not updated, leaving the system vulnerable to ineligible claims.[21]

20. Capita's role included developing and testing IT systems and security and its bid acknowledged the need for rigorous procedures to ensure data, programmes and documents were secure from unauthorised access and the importance of making the overall design robust with minimal chance of fraud and collusion. However, Capita did not pursue these points, and although KPMG identified the need to test the robustness of Capita's security arrangements, the Department did not do this and a planned review of Capita's systems by the Department's Internal Audit due in April 2001 was postponed until October 2001. Cap Gemini Ernst & Young's subsequent review of Capita's security system for the Department found that the contract did not include clear mandates on IT security, existing government guidelines had not been followed, usage monitoring was inadequate and no procedures had been established to ensure adherence to the security policy.[22]

21. While there is no evidence of unauthorised access to the system, some authorised learning providers made inappropriate use of it to gain access to unused accounts, some repeatedly during the last days of the scheme so that they could register learning as having started and claim the funding in respect of it. One such provider had offered to sell details of unused accounts and there were allegations of a large number of account details in circulation, and it was this chain of events that led to the decision to close the scheme immediately. The providers involved are under investigation.[23]

22. Following cancellation of the scheme, the contract with Capita continued while they were working with the Department on pursuing fraud and recovering money. The Department was also considering whether to launch a successor scheme and had agreed to work with Capita in developing arrangements for this, but the decision on whether Capita would be the delivery partner was subject to satisfactory progress and the outcome of negotiations with them. In the event, in October 2002 Ministers took the decision to consider any replacement arrangements as part of a wider review of the funding of adult learning and the development of the wider National Skills Strategy.[24]

23. Up to April 2002, the Department had paid Capita some £31.5 million out of the £55 million due under the full contract. Capita told us that it had still made a profit on the contract, although it had suffered a loss of reputation. In January 2003 the Department reached a settlement with Capita on outstanding sums due. This involved a net final payment of £1.5 million, some £1 million less than the estimated contractual entitlement. This takes into account termination charges due to Capita, the return of part of the profits generated by Capita under the contract Benefit Sharing Arrangements, and that the Department has made no payments to Capita since April 2002. The settlement figure was mutually agreed through negotiation to reflect a settlement acceptable to both parties based on Capita's delivery of the ILA system and the shared desire to conclude the wind-down of the original scheme.[25] In addition, at no charge, Capita will continue to provide support on the wind-down of the scheme until April 2003, and the Department will be able to draw fully on the investment in the design and development of the scheme.[26]

LEVEL OF FRAUD AND ABUSE, INCLUDING ACTION AGAINST PROVIDERS

24. In line with police advice, the Secretary of State closed the scheme with immediate effect on 23 November 2001, following the allegations that a large number of accounts had been extracted from the system and offered for sale. Registered learners had used their access to the scheme database to obtain detail of accounts for which they were not authorised. At the time, the Department estimated that if the scheme was not closed immediately, the value of fraudulent claims could run into tens of millions. Latest figures suggest fraud could be as high as £67 million.[27] But the Department does not know the scale of abuse whereby people received poor, low value courses.[28]

25. In addition, although the scheme was successful in generating 2.6 million accounts, only 1.5 million had learning registered by the time the scheme was closed. While a time lag between registering and undertaking learning was to be expected, some of the accounts had been "emptied" by unscrupulous providers. Fraud investigations and compliance visits to providers in 2001 and subsequently showed that 13 providers each registered over 10,000 account holders. This represented a very large number of learners recruited over a short time - more than the number of part-time students studying at all but the largest further education colleges. Two providers had over 30,000 learners. Moreover, there is evidence that a significant number of accounts were opened and incentives claimed without the knowledge or agreement of the account holder, and that a quarter of learners registered as having started learning had not done any.[29]

26. The Department currently has 70 staff working on follow-up investigations, and has someone working with West Midlands Police. Investigations include checking claims from about 700 providers. Some 153 are serious cases, of which 100 are already with the police and 60 people have been arrested, 14 charged, 10 are awaiting court appearances, 10 have been cautioned, and one has been sentenced to 9 months in prison. This is, however, a complicated and time-consuming process since it involves getting evidence that will convince the police and the Crown Prosecution Service.[30]

27. In addition, the Department is pursuing another 400 providers. It is contacting 50,000 registered learners, to ask whether they received learning. This should enable the Department to make estimates of the scale of fraud and abuse overall, and also for those providers which it has concerns about. It has written to 158 providers where there are questions to be answered, where money has been paid improperly or inappropriately, but stopping short of fraud, and where it is seeking recovery of the money paid. In most of these cases, the Department is withholding money until a settlement has been reached. However, it is not yet possible to estimate the amount likely to be recovered.[31]

28. The Department is cross-checking the companies and individuals involved with the Learning and Skills Council, and backwards to the Further Education Funding Council, to identify where else these providers might be providing training. It is also working with other Government Departments with a view to cross-checking with them as well.[32]

29. Of those under investigation, some have been linked to animal rights activity. Charges have been made against 3 providers, but the police do not have evidence to suggest that funds obtained from the scheme have been used to support animal rights activities.[33]

30. At the time of our hearing, about 95% of the money outstanding had been paid. On cancellation of the scheme, the Department froze payments due to providers, amounting to some £15 million, while it carried out checks on legitimacy of claims. The Department did not have total information on the number of providers who had gone out of business as a result of the cancellation, but knew of 13, mainly small, one person businesses. The Department recognises that providers will be more cautious in participating in any similar schemes in future.[34]

LESSONS LEARNED

31. We asked the Department how it was that despite substantial piloting, and advice from various expects and sectors, it came to design a system that was so open to fraud. The Accounting Officer told us that he had not been involved in the decision, but was ashamed of the failures that had taken place. He had looked in detail at what had happened and talked to those involved. There had been a number of key reasons. First, the design, and the decision not to have checks, had been driven by the view that a light touch, non-bureaucratic system was needed to persuade individuals and providers to come into the scheme. Second, there had been a belief that a market driven system would allow individual learners to judge whether the training was of a satisfactory standard and not take up the training unless it was. Third, the Department designed the scheme to provide incentives to encourage providers to find the learners and register them. In addition, the Department's project management, contract and supplier management, and management information was not good enough. With hindsight, he found the decisions inexplicable and accepted that the failure was one of the worst he had come across.[35]

32. The Accounting Officer assured the Committee that the lessons had been learned. For example, he had set in place arrangements to ensure that everybody engaged with supplier and contractor management was properly trained; had set up a unit to support project management; and was undertaking a risk assessment from the bottom to the top of the Department.[36] In addition, the Department has strengthened its existing counter fraud measures, by developing a Fraud Risk Assessment Strategy. Its Special Investigations Unit is to become more proactive in undertaking a programme of inspection work to detect irregularity in those areas most exposed to the risk of fraud, and the results of this work will support the level of assurance given by Internal Audit to the Accounting Officer. The Department has also formed a 'High Level Senior Management Risk Group', to alert the board to key areas of risk, including the risk of fraud.[37] The Department has developed and introduced an improved range of training and development events including specific training programmes, contract awareness sessions, contract surgeries and master classes. Detailed guidance on the financial aspects of contract management including risk assessment and financial monitoring will also be available to staff on the Department's intranet.[38]

1 C&AG's Report, Individual Learning Accounts (HC 1235, Session 2001-02), paras 7-8 Back

2 ibid, paras 1-2, 16; Ev 29 Back

3 Ev 34 Back

4 C&AG's Report, paras 8, 22, 2.2-2.7 (and Report Card 1, p7); Qq 4, 93-94 Back

5 Qq 4-6, 92-93, 125-126, 160-162 Back

6 C&AG's Report, 22-24, 2.14-2.15, 2.2-2.9 (and Report Cards 1-2, pp 7-8); Qq 95, 109-113, 218-219, 222-223 Back

7 C&AG's Report, paras 14, 2.8; Qq 64-65, 157-159 Back

8 C&AG's Report, para 1.12; Qq 129-135, 180-183, 194-199 Back

9 C&AG's Report, paras 2.13, 2.52; Qq 270-279 Back

10 C&AG's Report, paras 18-20; Qq 1-3, 152-155 Back

11 C&AG's Report, para 24 (and Report Card 2, p8), and paras 2.16-2.17; Q 96 Back

12 C&AG's Report, para 2.42 Back