Amyas Morse, the Comptroller and Auditor General, has today issued a report on the 2015-16 accounts of HM Revenue & Customs (HMRC).
Tax revenues and spending in 2015-16
HMRC raised £536.8 billion of tax revenues this year, an increase of £19.1 billion (3.7%) on 2014-15 and paid out £40 billion in benefits and credits (approximately one-fifth of the government’s total benefit expenditure). The taxes that contributed to most of this increase were Income Tax and National Insurance Contributions which together increased by £10.3 billion (3.8%); Corporation Tax which increased by £4.1 billion (9.9%); and VAT which increased by £2.1 billion (1.8%). Capital Gains Tax and Insurance Premium Tax also recorded significant increases, by 28.1% (to £7.3 billion); and 27.6% (to £3.7 billion) respectively. The annual cost of running HMRC was £3.2 billion in 2015-16 (£3.1 billion in 2014-15).
HMRC’s other key performance indicator is its compliance yield which measures the effectiveness of its compliance and enforcement activities. HMRC’s estimate of compliance yield in 2015-16 was £26.6 billion against a target of £26.3 billion. Compliance yield is not simply a cash figure. It draws on a range of different measures of revenue generated or losses prevented all of which involve a degree of estimation and uncertainty. The NAO recommends that HMRC should work to provide further explanation so that readers are better informed about the estimations and assumptions that underlie HMRC’s reporting of its performance.
Transforming tax administration
HMRC has begun to implement its plans to transform how it administers tax. Its vision is to have “the most digitally advanced tax system in the world”. By 2021, it expects to employ 16% fewer staff, substantially rationalise its estate and automate more of its processes. In the past year HMRC has made plans to invest more than £2 billion on its transformation in the next five years; launched digital accounts for individuals; announced plans to close 137 offices and the location of 13 new regional hubs; and secured agreement for its plans to replace its IT services contract, Aspire, which it has revised to reduce the risk of carrying out too much change too quickly.
HMRC’s approach looks credible and proportionate to the scale of the risks involved, and it has worked closely with the Treasury and Cabinet Office to develop and refine its plans. It is too early to evaluate how well its approach is working but one of the most critical tests will be how management responds when things do not go as expected. NAO have identified two areas of risk:
- Optimism bias in key assumptions – in the last Parliament, HMRC was over-optimistic about how much change it could deliver all at once, and how fast it could reduce demand for telephone contact in particular. This resulted in a collapse of its service to personal taxpayers in 2014-15 and the first half of 2015-16. HMRC has since recovered the quality of its service to personal taxpayers by recruiting more staff and has adjusted its future resource plans in the light of this experience.
- Understanding the costs and benefits to taxpayers – HMRC has yet to estimate the costs for individual taxpayers or businesses of making the transition to online services or to quantify the benefits they can expect. Over the next year, it plans to develop a fuller picture of what it will cost taxpayers to use the new systems. Most business customers will be required to update HMRC quarterly rather than annually about their tax affairs, and some may need to purchase new software that works with the new systems. Some businesses are sceptical of HMRC’s evaluations of the costs and benefits of previous changes to the tax system.
Since tax reliefs reduce the amount of tax due, they can be the focus of tax avoidance. Reports by the NAO in 2014 and the Committee of Public Accounts in 2015 concluded that HMRC needed to improve how it monitors and reports on tax reliefs. HMRC has subsequently made important improvements to the way it administers some reliefs, has identified and disseminated emerging good practice, and made more information about the costs of reliefs publicly available. However, HMRC’s monitoring of tax reliefs is not yet systematic or proportionate to their value or the risks they carry. There remain a large number of reliefs performing a wide range of different functions and HMRC’s practice in administering reliefs varies considerably. For example, while HMRC closely manages some low value reliefs for businesses it does not always do so for higher-value personal tax reliefs, such as principal private residence relief, worth £18 billion in 2015-16. While the cost of a tax relief is only one dimension of the risk to tax revenue, HMRC could not show us that it had a consistent approach to assessing the degree of risk that each relief carries.
The NAO recommends that HMRC should make clear that its good practice guidance for staff administering tax reliefs is compulsory to give it assurance that oversight for each relief is suitable. It should also publish all relevant information on the cost and impact of tax reliefs in a way that makes it more accessible so that Parliament can understand whether reliefs are working as intended.
Tax Credits and Child Benefit error and fraud
The C&AG has qualified his regularity audit opinion on the 2015-16 Resource Accounts because of material levels of fraud and error in the payments of Personal Tax Credits. He has qualified the accounts on these grounds every year since tax credits were introduced in 2003-04. HMRC’s central estimate of error and fraud in 2014-15 (the most recent available) is £1.37 billion overpayments (4.8% of total spending on Personal Tax Credits) and £0.19 million underpayments (0.7% of total spending on Personal Tax Credits).
HMRC has maintained the levels of error and fraud within Personal Tax Credits by designing and implementing interventions that enable it to identify and target high-risk cases. HMRC analysis shows that during 2014-15 it continued to reduce losses caused by children being incorrectly included in claims and by undeclared partners. However, their most recent analysis suggests it has not continued to reduce losses in other categories, particularly relating to earnings. HMRC should continue to use its analysis of losses within risk categories to inform where best to focus intervention activity.
HMRC estimates that overpayment of Child Benefit due to error and fraud was £170 million in 2015-16, equivalent to 1.4% of total spending on Child Benefit. The vast majority of this relates to customers not replying to requests for information during testing and who HMRC assumes to be non-compliant. HMRC should develop a more rigorous approach to its testing and evaluation of error and fraud in Child Benefit to identify the true level of losses, the root causes of these and to identify the appropriate actions to reduce error and fraud.