1.On the basis of a report by the Comptroller and Auditor General, we took evidence from HM Treasury and HM Revenue & Customs (HMRC).
2.Tax reliefs reduce the tax an individual or business owes. In October 2019, the UK had 1,190 tax reliefs. There are two broad categories of tax reliefs: structural tax reliefs that are largely integral parts of the tax system and define the scope and structure of tax (such as the income tax personal tax allowance); and non-structural tax reliefs where government opts not to collect tax to support social or economic objectives. The UK had 362 of these non-structural tax reliefs in October 2019. Some of these tax reliefs reflect policy decisions to support a particular group or sector, such as the housing market. Others are designed to incentivise the behaviour of individuals or businesses by making a choice less costly. These include tax reliefs on pension contributions and reliefs on research and development expenditure by business.
3.Ministers propose the introduction or amendment of tax reliefs, and Parliament scrutinises proposals as part of the Budget process. Ministers depend on the exchequer departments—HM Treasury and HMRC—to work in partnership and oversee tax reliefs and provide advice. HM Treasury leads on the design of tax reliefs and monitors their value for money and relevance. HMRC implements tax reliefs, monitors their use and cost, and evaluates them.
4.HMRC is responsible for estimating and reporting on the cost of tax reliefs. In October 2019, HMRC reported the cost of 111 of the tax reliefs which are designed to support the government’s economic and social objectives. These tax reliefs had a combined forecast cost of £155 billion in 2018–19. In May 2020, HMRC reported provisional costs for 47 more tax reliefs, which had a combined annual cost of £4 billion. Aggregating the cost of tax reliefs gives a sense of their scale, but it does not reflect the amount of tax that would be generated if they were removed as some taxpayers would respond by changing their behaviour and there may be wider economic impacts. In 2018–19, the ten largest tax reliefs had a combined forecast cost of £117 billon, which was around 5% of the UK’s Gross Domestic Product. The ten largest included reliefs on employer and employee pension contributions. The cost of these two reliefs combined had a total forecast cost of £38 billion.
5.In March 2015, we examined the management of tax reliefs by HMRC and HM Treasury. We concluded that HMRC did not systematically evaluate the effectiveness of all tax reliefs intended to change behaviour and found that some reliefs which had existed for many years had not been evaluated. We welcomed HMRC’s acceptance that it should be accountable for assessing, monitoring and evaluating tax reliefs. We recommended that it should regularly evaluate and report on whether tax reliefs designed to influence behaviour were achieving the desired impact and were doing so in a cost effective manner. In its response, HMRC disagreed with the recommendation, contending at the time that it was not practical to evaluate all reliefs with behavioural objectives on a regular basis. Nevertheless, the Department also stated that it had a substantial number of analysts preparing evaluative work in-house to inform tax policy development.
6.In the five years since March 2015, HMRC published evaluations of the behavioural impact of just 13 of the approximately 150 reliefs that aim to change behaviour. None of the ten largest tax expenditures were covered by these evaluations and only four of the tax reliefs costing over £1 billion a year were covered. Of these four, the evaluations found that the research and development relief for small- and medium-sized enterprises (annual cost £2 billion) had a positive impact on behaviour and gift aid (£1 billion) had some impact on behaviour. The other two reliefs—entrepreneurs’ relief and employment allowance (both costing £2 billion)—had limited impact on behaviour.
7.We were dissatisfied at the fact that none of the ten largest tax reliefs had been properly externally reviewed by HMRC. We asked HMRC why it had not evaluated any of these reliefs, such as pension reliefs. HMRC explained that cost was only one factor it took into account in selecting which reliefs to evaluate. It told us that some large reliefs, such as VAT relief on food, were difficult to evaluate because they are in a sense structural reliefs. It explained that it also took into account how likely a tax relief was to achieve its intended impact. This was particularly the case for those reliefs which were designed to achieve a specific behavioural economic change, which HMRC considered were easier to evaluate because their effects could be more marginal. HMRC also told us that it needed to consider which tax reliefs “politicians might be interested in reforming” as there was little point in spending money on evaluating a tax relief in an area where there was no appetite to reform. HMRC told us that it was keen to move towards an increasingly systematic approach for deciding which tax reliefs to evaluate, prioritising the largest reliefs that seek to incentivise behaviours. It explained that it would also apply other criteria to this, including strategic fit, priority and urgency, and the likely impact of the research.
8.The need for greater monitoring of the impact of tax reliefs was also raised with us by key stakeholders in the sector. We heard from the Chartered Institute of Taxation, which raised concerns about “the almost total lack of attention, at least so far as is visible to the outside world, as to how effective those measures prove over time.” It also contended that a key part of HMRC’s policy maintenance responsibility should be making sure that tax reliefs are achieving their objectives at reasonable cost. The Institute argued that this should be undertaken as part of a programme of regular reviews which monitors their take up, the cost of the relief, and whether it is having the desired impact on behaviours.
9.Evaluations typically cost between £50,000 and £250,000. The NAO estimated that HMRC had spent around £2 million on evaluating tax reliefs since 2015. HMRC has an annual central research budget of £2 million per year to fund evaluations of tax reliefs and other research to inform its wider business and HM Treasury priorities. We asked HMRC why it spent so little on evaluations compared to the value of taxes. It explained that its £2 million budget was for external research only and that it also undertook internal evaluations and analysis. However, it had no central record of the research it had conducted internally. HMRC was also unable to tell us how much it had spent on internal research, and in written evidence submitted after the hearing stated that it did not record activity spent on reviewing reliefs internally. HMRC accepted that it was important for it to focus on bringing together its internal analysis and publishing whatever information it could.
10.Tax reliefs on pensions contributions, designed to encourage people to save for their own personal pensions, are among the largest tax reliefs. HMRC forecast that the gross cost of these reliefs totalled £38 billion in 2018–19. We were concerned by claims that the tax relief was not being taken up and was also potentially not encouraging savings in real terms. HM Treasury asserted that take-up of pension tax relief had “grown over time” and that there was some evidence that “amongst some people it is quite a popular product,” but did not provide details on this.
11.We asked how the exchequer departments could be sure of the impact that pension tax reliefs were having if they had not evaluated them. HMRC asserted that pension reliefs had been subject to extensive evaluative attention, with “virtually no stone left unturned” as part of work to strengthen the incentive to save in 2015. However, it also accepted that it had not commissioned an external evaluation of the relief. HM Treasury added that it had undertaken substantial open consultation and research, including with focus groups, as part of its 2015 review, and that the Government had decided not to change the relief as a result. We encouraged HMRC and HM Treasury to commission a full external evaluation of the relief and to publish the information it held on the relief in order to encourage and support debate on this issue.
12.HMRC collects and reports data on who benefits from some tax reliefs. For example, HMRC reports annually on the number of people gaining from entrepreneurs’ relief and how much they gain. In March 2020, as part of the 2020 Budget the Chancellor announced that he was reforming entrepreneurs’ relief on the basis that the relief was expensive, ineffective and unfair. The relief cost over £2 billion a year and nearly three quarters of the cost benefited just 5,000 individuals.
13.We asked HMRC what assessment it had made to determine who benefited from another large tax relief, the VAT relief on the construction of new dwellings. HMRC forecast that the relief cost £15 billion in 2018–19. We asked whether the relief would distinguish between the benefit of someone spending large sums of money to build a single house, compared to another spending the same amount of money to build many more houses for low-paid families. HMRC told us that this tax relief was intended to incentivise the construction of new dwellings and encourage housebuilding. It explained that it considered the test of whether this was a success to be based on whether the relief reached the intended target part of the economy rather than who benefited.
14.We also asked who benefited from pension tax reliefs, and the split between different types of pensions. HMRC told us that its aim was to be fully transparent with all the information that it held, and referred to the work that had been undertaken to support the government’s 2015 review of pension tax reliefs. In September 2019, HMRC published data on the cost of different pension tax reliefs between 2012–13 and 2017–18. HMRC’s data did not show which groups or sectors benefited from these reliefs, or how the reliefs were used by those working in the public and private sectors, or by those with defined contribution or defined benefit schemes.
15.We were concerned that the lack of available information on some large reliefs had meant that it was not possible for HMRC and HM Treasury to know whether all of those who were expected to benefit had been able to do so and whether a relief is working. We heard from Standard Life Aberdeen, who told us that lower-paid workers in specific groups, who most needed tax relief on their pensions contributions, were at most risk of missing out on this tax relief. The financial services company Royal London estimated that around 1.75 million low-paid and part-time workers, auto-enrolled into employer pensions, were missing out on tax relief on their pension contributions. Around three quarters of these workers are women.
16.We asked the exchequer departments when they would take action on the issue of workers not receiving pension tax relief. HM Treasury said the Government recognises the different impacts of the two systems of paying pension tax relief on pension contributions for workers earning below the personal allowance. It referred to the Government’s announcement in Budget 2020 that a call for evidence would be published in spring 2020 on this subject. The call was due to ask for views on how to address the different pension outcomes for lower earners, depending on whether their employer’s pension scheme used the net pay or relief at source method of tax relief on their pension contributions. HM Treasury told us that in the light of COVID-19 the Government was considering the publication of this and other Government documents on a case by case basis. It planned to provide more information on the timeframe for publication of this call for evidence in due course.
1 C&AG’s report, The management of tax expenditures, Session 2019–20, HC 46, 14 February 2020
2 C&AG’s report, paras 1, 3 and Figure 1
3 C&AG’s report, para 4
4 C&AG’s report, paras 9–10, 1.5–1.6
5 HMRC, Non-structural tax reliefs –Additional cost estimates, 20th May 2020.Costs are not reported for the other 204 tax reliefs with economic and social objectives.
6 C&AG’s report, para 3 and Figure 3; HM Treasury, , April 2020
7 Committee of Public Accounts, The effective management of tax reliefs, Forty-ninth Report of Session 2014–15, HC 892, March 2015
8 HM Treasury, Treasury Minutes Government responses on the Thirtieth, the Thirty Fifth, the Thirty Seventh, and the Forty First to the Fifty Third reports from the Committee of Public Accounts: Session 2014–15, July 2015
9 C&AG’s report, paras 1.5, 3.3, 3.6, Figure 14
10 Q 52
11 Qq 46–47
12 Qq 40, 63
13 Written Evidence MTE0002 – Mr Richard Wild, Chartered Institute of Taxation, published 10 June 2020
14 C&AG’s report, paras 3.4 and 3.5
15 Qq 40, 63 C&AG’s report, para 3.3, footnote 39
16 Q 41; Letter from Jim Harra, Chief Executive and First Permanent Secretary, HMRC, 26th June 2020
17 Q 63
18 Qq 32, 47; Written Evidence MTE0003 – Management of tax reliefs, Ruari Grant (Senior Public Affairs Associate, Standard life Aberdeen), and C&AG’s Report, para 9 and Figure 3
19 Qq 33, 47–48
20 Qq 47–48, 51
21 HMRC, , August 2019, Table 4
22 Budget Speech 2020, available at
24 HMRC, , October 2019
25 Written Evidence MTE0003 – Management of tax reliefs, Ruari Grant (Senior Public Affairs Associate, Standard life Aberdeen) published 10 June 2020
26 Royal London, , April 2019
28 Net pay is where the employer takes pension contributions from an individual’s pay before it is taxed. Relief at source is where the employer takes pension contributions after tax and National Insurance has been taken from an individual’s pay.
29 Q 49, Letter from Beth Russell, Director General, Tax and Welfare, HM Treasury, 26th June 2020 and HMRC, , March 2020; and HMG
Published: 20 July 2020