1. NIRS 2-the National Insurance Recording System-is a large and complex computer system which supports the administration of the national insurance scheme. It holds details of some 65 million individual national insurance contribution records. This information is fundamental to the accurate calculation of contributory social security benefits, such as retirement pension and underpins payments to pension schemes in respect of contributors with contracted out personal pensions. In 2000-2001, the Inland Revenue collected over £50 billion in national insurance contributions and the Department of Social Security (now the Department for Work and Pensions) paid out £46 billion in contributory benefits, based on records held on the system.1
2. NIRS 2 was developed by Accenture (then Andersen Consulting) in 1995. The contract covered the replacement of NIRS 1, transfer of data, development of the system to implement legislative changes arising from the Pensions Act 1995, and the operation of the new system until 2004. The original contract was valued at £45 million for operational services with provision for software enhancements increasing that amount to £76 million. We examined the difficulties that were experienced in delivering NIRS 2 on three occasions between 1998 and 2000.2
3. In 1998 the Government proposed significant changes to pensions and national insurance legislation, for example to introduce stakeholder pensions and pension sharing on divorce. The Inland Revenue, who had taken over responsibility for NIRS 2 in April 1999 with the transfer of the Contributions Agency, negotiated an extension to the contract to cover the work needed to support these legislative changes. The estimated value of the extension is between £70 million and £144 million, depending on the amount of work ordered over the remaining life of the contract.3
4. On the basis of a Report by the Comptroller and Auditor General4 we looked at the need for the new work, the reasons why the Inland Revenue chose to extend the contract with Accenture rather than conduct a fresh competition, and Accenture's performance so far.
5. In the light of this examination, the Committee draws three overall conclusions.
• The original contract for NIRS 2 showed the impact of setting too tight a deadline in statute for implementation of the Pensions Act 1995. The Department of Social Security repeated this mistake by not properly assessing the aggregate impact on NIRS 2 of the later proposed changes to pensions and national insurance legislation, or the risks and extra costs involved. Our Report on Ensuring Policies Deliver Value for Money has already emphasised the need for policies to be implementedwith sufficient flexibility and for the risks faced to be identified, assessed and regularly monitored.5
• In non-competitive situations, Departments need to have in place rigorous methodologies for assessing the reasonableness of the prices on offer, and arrangements to avoid contractors earning excessive profits. In the case of the NIRS 2 extension, the Inland Revenue's benchmarking arrangements suggested that Accenture's proposals offered value for money. The Inland Revenue are sharing the benefits of improved productivity, and "super profits", but Accenture have outperformed their target productivity levels by a wide margin. On the basis of experience so far, the prices agreed appear to be very generous for a non-competitive contract, where, in practice, the Inland Revenue had little option but to use Accenture because of the high break costs of the original contract.
• The Inland Revenue face a fresh challenge in 2004, when their contracts with Accenture and with EDS (their strategic supplier of other information technology systems) come up for renewal. Although they are taking steps to generate competition, the barriers may be too great, especially in terms of learning and set up costs for these large and complex systems in a deal likely to be worth over £4 billion. The methodologies they use to estimate and benchmark proposals will need to be rigorous.
6. Our more specific conclusions and recommendations are as follows.
(i) The major welfare reforms planned by the new Government after May 1997 led to significant changes to NIRS 2, well beyond those that could reasonably have been foreseen when the initial contract was placed in 1995. While the Department of Social Security considered the technical feasibility and costs of each change, they did not assess whether the changes could be delivered within the existing contract. Failure to do so restricted the Inland Revenue's options when it became clear that significant additional work was required, because implementation dates were already set in legislation. The Inland Revenue should explore with the Office of Government Commerce how to build additional flexibility into future contracts, for example by inviting the bidders to propose a separate pricing structure for major enhancements as part of the initial tendering process.
(ii) The Inland Revenue took advantage of the contract extension to introduce incentives for delivery and productivity, and to address many of the weaknesses the Committee of Public Accounts identified in its earlier reports on NIRS 2 and other IT projects that encountered difficulties. For example, there is now a clear framework for acceptance testing.
(iii) The Inland Revenue negotiated arrangements for sharing the benefits of any productivity improvements achieved by Accenture, and to share any "super profits" above agreed margins. The productivity actually achieved by Accenture on the first two developments (3.4 compared with 7.5 days per function point) and the fact that Accenture appear to have made "super profits" in year two of the contract, do raise serious questions about how rigorous the original estimates and benchmarking were. The Inland Revenue should consider the scope to renegotiate the target productivity rate for the rest of the contract in line with performance to date.
(iv) Both parties attributed improvements in performance under the contract to stronger and more robust partnership arrangements. In our Reports on Improving Construction Performance and Non-Competitive Procurement in the Ministry of Defence6 we highlighted the potential benefits of partnering, including incentivisation of performance; transparency; minimising the risk of disputes; replicating lessons learned on earlier projects and a sensible allocation of risks.
___________________________________________________________________________________________________
1 C&AG's Report, paras 1, 1.2
2 Committee of Public Accounts: 46th Report, The Contract to Develop and Update the Replacement National Insurance Recording System (HC 472, Session 1997-98); 22nd Report, Delays to the new National Insurance Recording System (HC 182, Session 1998-99); 31st Report, National Insurance Fund 1998-99 and Wider Issues of Fraud and Error in Benefits Paid by the Department of Social Security (HC 350, Session 1999-2000).
3 C&AG's Report, para 2
4 C&AG's Report, NIRS 2: Contract Extension (HC 355, Session 2001-02)
5 Committee of Public Accounts: 49th Report, Ensuring that Policies Deliver Value for Money (HC 541, Session 2001-02)
6 Committee of Public Accounts: 2nd Report, Improving Construction Performance (HC 337, Session 2001-02); 29th Report, Non-Competitive Procurement in the Ministry of Defence (HC 370, Session 2001-02)