House of Commons - Public Accounts

Select Committee on Public Accounts Thirty-First Report

INLAND REVENUE: EMPLOYER COMPLIANCE REVIEWS

THE IDENTIFICATION OF NON-COMPLIANT EMPLOYERS

5. The Department carried out more than 54,000 compliance reviews in 1995-96. They detected irregularities in over 41,000 cases (76 per cent) and identified tax and national insurance contribution underpayments amounting to £290 million, including interest and penalties. The detection rate in the Department's ten Regional Executive Offices varied from 71 per cent in Scotland and London to 81 per cent in South Yorkshire. Within regions there were wider variations between individual Local Employer Compliance Units. In Scotland, for example, the proportion of compliance reviews where an irregularity was detected ranged from under 40 per cent to over 90 per cent.[3]

6. When asked in evidence why there were significant local variations in the detection of non-compliant employers,[4] the Department told the Committee that there were a number of factors which produced variations including the size of employers and the nature of their business, the type of review being carried out and the number of trainees in each Unit. For example, in Scotland several Units concentrated on Schedule E only and undertook reviews of financial institutions in the City which typically provided generous remuneration packages, including benefits in kind. In these high risk cases there was more opportunity to detect non-compliance. By contrast, other Units in Scotland would have little opportunity to examine high-risk cases. The Department were, nevertheless, concerned about the variations and, in January 1997, had set up a review which was expected to report shortly. This review would undoubtedly include recommendations on best practice. The Department had also asked their Regional Executive Offices to monitor variations and to take corrective action by next year, if necessary.

7. Staff in the Department's Local Employer Compliance Units do not currently have a database of employers to help them select cases to review. In 1995, the Department decided to develop a computer system to support their compliance work which would include a database of employers. They expected this to be available from April 1998. An earlier trial with a simple database in two Units in 1992 and 1993 helped staff reduce the time spent on selecting potentially non-compliant employers for review and increased the tax yield. One Unit's yield increased by £315,000 (139 per cent) and the other Unit's yield rose by £132,000 (50 per cent), compared with a 25 per cent increase nationally. The Department estimated that it would have cost roughly £20,000 to provide a similar database for all their Units but decided against this on the grounds that the system would have been inconsistent with their planned information technology strategy and would have diverted resources from higher priority work.[5]

8. Asked whether their decision not to provide the database to their other Units represented good value for money,[6] the Department said that, with hindsight, they should have examined more rigorously the potential offered by the local databases and apologised for not doing so. They considered, however, that their decision had to be seen in the wider context of their computer strategy. Their overwhelming priorities were to provide computer support for Self Assessment and to introduce a fully computerised employer compliance system. They had a limited resource of skilled people available to develop these systems.[7] Had they gone ahead with the local databases it would have inevitably taken people off these tasks.[8] The Department accepted that the existence of a strategic programme did not necessarily preclude an interim solution but believed it would have been a distraction for the people developing the new systems at a time when they were under great pressure.[9]

9. With hindsight, the Department believed that their estimate of the cost of introducing the local database was unrealistic, as it excluded setting-up and training costs, some hardware costs, and the costs of regularly updating the databases.[10] They believed that it was many orders of magnitude too small but could not now provide a revised estimate for extending the database to the other Units.[11]

10. The Department told the Committee that the system would have provided each Unit with a database of all their employers and a facility to search the system for details of employer size by number of employees; address; postal code; and trade classification.[12] It would also have provided a facility to record additional information such as dates of previous employer compliance reviews. It would not, however, have provided any link to existing computer facilities. While the Department were always looking for ways of improving their targeting of non-compliance, they had to balance their priorities.[13] They reached a view on the likely effectiveness of two different tools and it was on that basis that the decision was reached.[14]

11. When asked how much tax had potentially been lost by not introducing the local databases,[15] the Department told the Committee that it was extraordinarily difficult to make an estimate. The results from one of the two Units had been affected by one very large case. The Department also said that it was possible that their new employer compliance computer system would identify the same employers for investigation in the future and, if necessary, they could go back six years, and in more serious cases 20 years, to recover unpaid tax. In those circumstances, there would be no tax loss—simply deferred collection of tax. They were confident that a great deal of any tax they might have lost would come in later.[16]

12. The Inland Revenue's financial case for the introduction of the new system showed estimated implementation costs of £6.9 million and running costs of £2.4 million. Gross savings were estimated to be £14 million, giving a net cost saving of £4.7 million. Subject to the availability of additional funds, the Department may enhance the system to incorporate a facility for creating profiles of high risk employers, based on predetermined risk rules. The risk module would achieve its full potential two to three years after the introduction of the system in April 1998, when a store of comparative information was available.[17]

13. We asked the Department whether the business case for the new system included an estimate of the additional tax that might be collected.[18] The Department said that the business case referred to the potential increase in tax yield from introducing the new system and took it into account but it could not be reliably quantified. They added that the implementation costs of the new system had risen to just over £11 million and estimated running costs had risen to £3.6 million. The estimated gross savings were now more than £18 million, giving a net cost saving of nearly £3.6 million.

14. The Contributions Agency refer cases to the Department's Local Employer Compliance Units to review if, after visiting employers to check their national insurance records, it appears that income tax may have been unpaid. The Department's investigations led to the identification of unpaid tax in only 65 per cent of the cases referred in 1995-96. If staff resources devoted to unproductive referrals from the Agency in 1995-96 had been deployed on other employer compliance reviews, the Units might have identified an additional £6 million of unpaid tax.[19]

15. The Department were asked what they and the Agency were doing to improve the quality of referrals.[20] They told the Committee that this was an area of major concern both to them and the Agency. They would be providing education and training for Agency staff to assist them to recognise tax irregularities when they came across them to improve the quality of referrals. The Department also said that they were going to review all the Agency's referrals in the current year where only national insurance liabilities had been found in order to establish the reason why no tax was recovered and to take corrective action, where necessary, by June 1998. In addition they planned to seek assurance that the correct decision had been made in cases which were not referred to Units by the Agency. The Government have recently announced that the Contributions Agency will be transferred to the Inland Revenue in April 1999.[21]

Conclusions

16. The Committee is concerned at the wide disparity in the success of the Department's Local Employer Compliance Units, which review smaller businesses, in identifying non-compliant employers. In Scotland, for example, the proportion of compliance reviews where an irregularity was detected ranged from under 40 per cent to over 90 per cent. The Department have set up a review to look into these differences, and we look to them to implement recommendations on best practice.

17. We also expect the Department to exercise close oversight over their Regional Offices' work to monitor such variations in performance, and we expect them to take prompt corrective action where these differences are due to shortcomings in the way Units select and conduct cases.

18. In spite of a successful trial with a simple employer database in two of their Local Employer Compliance Units which yielded increased tax of £315,000 in one Unit (139 per cent) and £132,000 (50 per cent) in the other, compared with a 25 per cent increase nationally, the Department decided that extending the database to all Units would have conflicted with their information technology strategy and would have diverted staff from higher priority work We consider that the Department should have examined a lot more thoroughly the costs and benefits of extending the database, pending the introduction of their new employer compliance computer system, and in particular the potential loss of tax from not doing so.

19. The Department argued that a great deal of any tax that might have been lost from not extending the simple database to all Units could be recovered later because they might identify the same employer for review in the future and recover unpaid tax for earlier years. In our view, however, this remains to be demonstrated, and we consider that the database could still have been a worthwhile interim measure with a significant payback in terms of additional tax yield.

20. It is disturbing, especially in the light of the Department's decision not to extend the experimental database, that their new employer compliance computer system will not be fully effective until around 2001, that costs have increased by some £5 million and that net savings are now around £1 million less than originally anticipated.

21. It is also surprising that the Department's business case for the new employer compliance computer system did not include at least a broad estimate of the amount of additional yield that might be expected. The Committee recognises the difficulty of making such estimates but expect the Department to include them wherever possible in future investment appraisals.

22. The Committee are concerned at the potential loss of tax, possibly £6 million in 1995-96, resulting from unproductive referrals from the Contributions Agency. We note the recent decision to transfer the Contributions Agency to the Inland Revenue and expect the Department to use the new opportunities provided by the transfer to build on the work currently being carried out by both organisations to improve the quality of referrals and increase the tax yield.

3 C&AG's report (HC51 of Session 1997-98), paras 1.2, 2.5 and Figure 3 Back

4 Q4 Back

5 C&AG's report (HC 51 of Session 1997-98), paras 2.6-2.9, 2.12 Back

6 Q2 Back

7 Q29 Back

8 Q2 Back

9 Q16 Back

10 Q2 Back

11 Qs 14-16 Back

12 Q10 Back

13 Q12 Back

14 Q49 Back