HMRC Performance inquiry

The way HMRC measures the tax gap and the impact of its efforts to close it remain unclear. There has been a downward trend in the difference between the amount of tax that should, in theory, be collected by HMRC, and what is actually collected (the tax gap) over the last 10 years. While HMRC’s tax gap estimates fluctuate year to year the figure has fallen from 8.3% in 2005–06 to 6.5% (£36 billion) in 2014–15. But we are concerned that HMRC may be painting too rosy a picture of its success in reducing the tax gap. HMRC’s Annual Report does not set out clearly what impact its compliance work (tackling those who seek to avoid or evade their tax liabilities), which yielded an estimated £26.6 billion in 2015–16, has on reducing the tax gap. HMRC’s measure of compliance yield also contains a high degree of estimation and projection and it is not clear whether these estimates are realised in terms of extra tax being collected. Recommendation: HMRC should report each year on the effect its work to generate compliance yield is having on its efforts to reduce the tax gap. As part of this, HMRC should assess how accurate its compliance estimates turn out to be in practice.

Despite our repeated recommendations, HMRC still does not make tax reliefs sufficiently visible to support parliamentary scrutiny and public debate about areas where the UK chooses not to collect tax. We welcome HMRC’s agreement to improve the accessibility of the information it currently publishes on tax reliefs, which is scattered over a number of different publications but not included in its annual report. However, HMRC needs to do more. It should monitor whether the use and cost of tax reliefs is in line with Parliament’s intentions when the reliefs were enacted. While HMRC publishes cost data for 180 tax reliefs, more than many other countries, this only accounts for some 15% of existing UK tax reliefs. HMRC does publish cost data where it has them, but it is not clear how it decides which reliefs to collect data on from taxpayers through their tax returns. HMRC still does not provide sufficient information to explain the impact that tax reliefs are having on behaviour. The need to review the impact of tax reliefs was illustrated by the National Audit Office’s analysis of entrepreneurs’ relief which showed that this was costing three times more than expected. Recommendation: HMRC should include an analysis of tax reliefs and their costs in its annual report to improve accountability about the areas where government has chosen not to collect tax. HMRC should make clear why it has decided to collect data only for a small minority of tax reliefs.

HMRC is staking a great deal on the success of its plans to digitise the tax system, but once again it lacks an adequate plan if demand for its call centres does not reduce as quickly as it hopes. HMRC faces an enormous challenge to maintain services, while delivering spending cuts, restructuring its business, replacing the Aspire contract (through which it has procured almost all its IT), at the same time as re-locating almost all its staff and dealing with the implications of Brexit. HMRC has lost its Chief Digital and Information Officer as a result of market pressures and losing further key staff would damage HMRC’s capability to deliver transformation and manage the risks when its IT contract ends. HMRC has committed to reducing 6 HM Revenue & Customs performance in 2015–16 staff in its personal tax service by a third in the next three years. When HMRC made some 5,600 staff reductions in 2014–15 customer service for personal taxpayers collapsed and HMRC had to recruit 2,400 additional staff in the following year to stabilise services. We note HMRC’s assurances that it will seek additional resources if digital services do not reduce demand for personal tax services in line with its expectations, but we are concerned that it has not agreed a contingency plan for this eventuality with HM Treasury. Recommendation: By March 2017, HMRC should demonstrate to the Committee that it has a credible plan to make savings without damaging customer service, and that it has agreed a contingency plan with HM Treasury should its projections prove to be inaccurate in practice.

HMRC receives too many complaints and the vast majority of those referred to the tax adjudicator are upheld in part or in full. Our recent report on quality of service to personal taxpayers highlighted that taxpayers spent more than 4 million hours waiting for HMRC to answer the telephone in 2015–16.1 We welcome HMRC’s statement that it has reduced average call waiting time to less than five minutes and that it aims to get it below three minutes. However, while the speed of telephone answering appears to have improved, the number of complaints has increased. In 2015–16, HMRC received 80,400 customer complaints, compared to 74,400 in 2014–15. Furthermore, 85% of the 1,808 complaints referred to the tax adjudicator in 2014–15 were upheld in part or in full. HMRC accepts that there is something wrong with the way it currently handles complaints and is reviewing its procedures. Recommendation: HMRC should review the complaints it receives, to identify what action it can take to reduce the number raised, and its complaints procedures to minimise the number referred to and upheld by the tax adjudicator.

The poor quality of service received by tax credit claimants under the Concentrix contract resulted in unnecessary hardship and suffering. In 2014, HMRC signed a contract with Concentrix to provide additional capacity to tackle fraud and error in the tax credit system on a payments-by-results basis. We raised concerns about the contract and the treatment of claimants in June 2016 and were given assurances by HMRC in July that things were improving. However, in August 2016 customer service levels for tax credit claimants collapsed. HMRC told us that Concentrix had failed to cope with an increase in call numbers following a letter it had sent to a large number of claimants about their claims. For example, in the third week of August the basic levels of customer service provided by Concentrix deteriorated to the point that less than 10% of phone calls were being answered within five minutes. HMRC considered that such an increase in call volumes had been entirely predictable. HMRC told us that it would not renew the contract and was dealing with all claimants directly. Subsequently HMRC announced that it had cancelled the contract, cleared all of the 181,000 incomplete cases that it took over from Concentrix and had completed around 28,500 of the roughly 32,500 requests for review of Concentrix decisions (known as mandatory reconsiderations). HMRC also noted that as a result of the contract ending, around 250 Concentrix staff had 1 Committee of Public Accounts, Quality of service to personal taxpayers and replacing the Aspire contract, Thirteenth Report of Session 2016–17, HC 78, July 2016, Part 1, paragraph 13 HM Revenue & Customs performance in 2015–16 7 transferred to HMRC. We expect to return to this subject in the New Year when the results of an investigation into this matter being undertaken by the National Audit Office are available. Recommendation: HMRC must ensure it has clear customer service standards whether a service is delivered by HMRC or one of its contractors. HMRC should identify quickly the lessons to be learned from this episode and act upon them, particularly given the concerns raised months before HMRC chose to act.

HMRC and HM Treasury need to make the tax affairs of large multinational companies more transparent to increase the pressure on them to pay their fair share of tax. This Committee has repeatedly called for the tax affairs of multinational companies to be made more transparent. We therefore welcome the fact that HMRC will soon receive automatically information on the global activities of large multinational companies that do business in the UK (through what is known as country-by-country reports). However, these country-by-country reports will be supplied on a confidential basis and will thus not contribute to increasing transparency over the tax affairs of these companies. We also welcome the cross-party consensus that now exists on the need for public country-by-country reporting, which will require international agreement to make a reality. The power to require public country-by-country reporting was set out in UK law following work by members of the Committee to promote an amendment to the Finance Bill

The country-by-country reporting that has been agreed is for information to be supplied on a confidential basis, as this was the only basis on which some countries would agree to adopt the new initiative. The Finance Act 2016 gave the Government the additional power to introduce a public country-by-country reporting requirement for large multinational companies.20 This power was adopted after work by members of the Committee to promote an amendment to the Finance Bill on the issue.21 The Government supports the view that country-by-country reporting should be public but believes that it should be implemented on an international basis or, failing that, on a multilateral basis. The Government is committed to working with international partners to reach agreements on a multilateral basis.22 Ideally, the Department would like to see public country-by-country reporting agreed internationally as a global standard.23 13 Qq 50, 52 14 Q 52 15 Qq 59, 62 16 Qq 69–71 17 Q 71 18 Q 72 19 Qq 72–74 20 Finance Act 2016, section 161 and Schedule 19, paragraph 17 (6) 21 Amendment 145 to consideration of Finance Bill 2016 (Report Stage), 5 September 2016 22 Q 74 23 Q 75 10 HM Revenue & Customs performance in 2015–16 2 Plans to improve tax services

HMRC (the Department) needs to transform its whole organisation to achieve its aim of becoming one of the most digitally advanced tax administrations in the world, while also making significant changes in parallel. By 2021, it expects to employ 16% fewer staff who will mostly be working in 13 regional centres. This will require a substantial rationalisation of the HMRC estate resulting in the closure of 137 of its existing locations (90%). Most of HMRC’s processes will be automated, and a higher proportion of its staff will undertake specialist work to challenge those taxpayers who seek to avoid or evade their tax liabilities.24 HMRC expects its digitisation of the tax system to promote and enable better compliance, which in due course will be reflected in a reduction in the size of the tax gap.25

Under its spending review settlement, HMRC is committed to realising £98 million of savings through a reduction by approximately a third in the number of those staff that deal with customer services.26 HMRC expects that, as it introduces more online services, a channel shift will occur resulting in a reduction in the demand for phone calls as more people will be able to meet their needs through online services. We pointed out that in the past customer services had suffered when HMRC had cut staff before any significant reduction in the number of telephone calls had been realised. The Department recognised this risk.27 It told us that 6.2 million people had opened a personal tax account and were able to interact to some degree with HMRC online. However HMRC accepted that it was possible that being more transparent, and providing more information to taxpayers in a digital way, would stimulate more demand for HMRC services, including those provided by phone.28

When HMRC made some 5,600 staff reductions in 2014–15 customer service for personal taxpayers collapsed and HMRC had to recruit 2,400 additional staff in the following year to stabilise services.29 We asked HMRC about its contingency plans to ensure customer service levels did not deteriorate, as they had in the past, if its assumptions about reductions in demand for telephone services proved to be unrealistic.30 We also raised a number of factors that could damage HMRC’s capability to deliver its transformation programme and manage the risks when its Aspire contract ends. HMRC’s Chief Digital and Information Officer recently left the Department to join the private sector.31 The prospect of Brexit also increases the level of uncertainty and challenge that HMRC faces in managing its change programme. HMRC told us that it expected Brexit to have an impact on: customs and excise duties; social security; administration of the VAT regime, state aid; information exchange with other EU countries; ongoing litigation; and 24 C&AG Report, para 2.3 25 Qq 62–65 26 Qq 112, 114 27 Qq 112, 115–117, 144 28 Q 117 29 Committee of Public Accounts, Quality of service to personal taxpayers and replacing the Aspire contract, Thirteenth Report of Session 2016–17, HC 78, July 2016, Conclusions and Recommendations, paragraph 2 30 Q 144–145 31 Qq 149–151 HM Revenue & Customs performance in 2015–16 11 on businesses and their relationship with the Department.32 HMRC noted that if demand did not reduce in the way it expected it would have to ensure that adequate levels of staff were in place to meet taxpayers’ needs.33

In the last Parliament, HMRC had made over-optimistic assumptions about how much change it could make all at once, which led to significant deterioration in the quality of its customer services for some 18 months.34 Our recent report highlighted that taxpayers spent more than 4 million hours waiting for HMRC to answer the telephone in 2015–16.35 We challenged the Department as to whether it had enough evidence, particularly estimates of the cost to individual taxpayers interacting with the tax system, to support its assumptions as it pushes on with its planned changes.36 HMRC told us that it did not yet have an estimate of the costs to individual taxpayers of complying with their tax requirements. It only had estimates of the impact on businesses, currently estimated to be £11.5 billion a year, with the Department committed to a target to reduce it by £400 million a year by the end of this Parliament.37

HMRC acknowledged that customer time was not a free good. It said it would look at whether it could quantify the cost of customer time.38 The Department considered that it should focus on providing customers with a “good basic level of service” and with further options for how they could interact with HMRC “in a simple and transparent way”.39 The Department told us that it had reduced average waiting times to below five minutes and aimed to get the average to below three minutes.40 HMRC explained that a key indicator it used to gauge customers’ experiences was their level of satisfaction with the services HMRC provides, rather than monitoring the direct financial costs incurred by customers. In the case of digital accounts the monitoring of satisfaction levels takes place in real time as there is an exit survey as customers leave.41

In 2015–16, HMRC received 80,400 customer complaints, compared to 74,400 in 2014–15.42 We asked why the number of complaints had increased, particularly as the Department for Work and Pensions, which is also telephony based, had managed to reduce its number of complaints.43 HMRC could not explain the increase and accepted that it may have some lessons to learn from DWP’s approach.44 We expressed our concern that 85.3% of the 1,808 complaints that were escalated to the Adjudicator’s Office in 2014–15 were subsequently upheld in full or in part. HMRC acknowledged that its handling of complaints was not good enough noting that there was something wrong with the system if complaints could go through two different processes in HMRC and yet HMRC’s decisions were being overturned by the tax adjudicator.45 32 Qq 91, 93, 102 33 Qq 145–147 34 Q 109; C&AG Report, para 2.19 35 Committee of Public Accounts, Quality of service to personal taxpayers and replacing the Aspire contract, Thirteenth Report of the Session 2016–17, HC 78, incorporating HC 79, 27 July 2016 36 Qq 104, 107 37 Qq 104–106, 110 38 Q 119 39 Q 118 40 Qq 112, 121, 144 41 Q 120 42 HM Revenue & Customs, Annual Report and Accounts 2015–16, HC 338, July 2016 43 Q 124 ; Department for Work & Pensions, Annual Report and Accounts 2015–16, HC 331, July 2016 44 Q 124 45 Qq 125-132 and HM Revenue & Customs, Annual Report and Accounts 2015–16, HC 338, July 2016, Fig. 13 12 HM Revenue & Customs performance in 2015–16

In 2014, HMRC signed a contract with Concentrix, a private sector company, to provide additional capacity to tackle fraud and error in the tax credit system on a payments- by-results basis. We raised concerns about the contract and the treatment of claimants in June 2016 and were given assurances by HMRC in July that things were improving.46 In September 2016 HMRC announced the it would not renew its contract with Concentrix.47 In November 2016 HMRC announced the contract had been terminated early.48 The operation of this contract has caused unnecessary distress and hardship to HMRC’s tax credit customers.49 HMRC told us that in the third week of August the basic levels of customer service provided by Concentrix “deteriorated significantly” to the point where less than 10% of phone calls were being answered within five minutes. From mid-August onwards there had been a “fundamental failure of basic customer service”.50 HMRC told us that Concentrix simply had not put enough people on the phones to cope with the high volume of calls it received after it had written to a large number of claimants about their claims.51 HMRC told us that this increase in call volumes had been entirely predictable.52 HMRC told us that it was now dealing with all claimants directly.

HMRC considered that there were a number of lessons to be learnt from this experience concerning: the sensitivity and prioritisation of customers in the delivery of public services; the need for more thorough assurance about contingency planning; a question about the speed of escalation, so that people know how quickly issues get to decision makers who can resolve them; a question about whether third parties can fully understand the subtleties of delivering a public service; and whether an incentive contract was the right kind of mechanism in this case.53

Subsequently HMRC announced that, as of 13 November 2016, it had cleared all of the 181,000 incomplete cases that it took over from Concentrix and had completed around 28,500 of the roughly 32,500 requests for review of Concentrix decisions (known as mandatory reconsiderations). HMRC also noted that as a result of the contract ending, around 250 Concentrix staff had transferred to HMRC.54 We expect to return to this subject in the New Year when the results of an investigation into this matter being undertaken by the National Audit Office are available. 46 Public Accounts Committee Oral evidence: Quality of Service to Personal Taxpayers, HC 78 of Session 2016–17, 13 June 2016, Qq116, 117, Correspondence from the Permanent Secretary of HM Revenue and Customs relating to customer service and Concentrix, dated 26 July 2016, published on 8 September 2016 47 Qq 3–4, 7 48 Update on HM Revenue and Customs’ contract with Concentrix: Written statement, Financial Secretary to the Treasury, HCWS251, 14 November 2016 49 Q6 and Correspondence from Jon Thompson to Frank Field MP regarding Concentrix, 24 October 2016, published on Work and Pensions Select Committee website 28 October 50 Q 5 51 Qq 6, 18, 21–22, 24 52 Q 6 53 Q 39 54 Update on HM Revenue and Customs’ contract with Concentrix: Written statement, Financial Secretary to the Treasury, HCWS251, 14 November 2016 HM Revenue & Customs performance in 2015–16 13 3 Tax reliefs

Tax reliefs are integral to the tax system and an important tool for public policy. There are more than 1,100 tax reliefs and several different types.55 We asked HMRC (the Department) what assurance it could provide that it was monitoring tax reliefs to ensure they were delivering the intended policy objectives.56 The Department told us that it had two main responsibilities in relation to tax reliefs: managing their administration and monitoring and reporting the costs involved.57 HMRC noted that it would be for Ministers to decide whether “a wholesale review of tax reliefs” was required and said that it was committed to publishing cost information in cases where it has the relevant data.58

HMRC’s organisational structure is based around customer groups (such as small businesses, large businesses, wealthy taxpayers and organised criminals) and specific behaviour types (such as tax avoidance, tax evasion and the hidden economy), rather than specific tax reliefs.59 Nevertheless, each area of tax has a product owner who has the expertise in that tax to give policy advice on what needs to change. Responsibility for administering tax reliefs, including monitoring the cost of the reliefs, lies with the relevant product owner.60 In 2015, we made a number of recommendations to HMRC to improve how it monitors and reports on tax reliefs.61 In September 2015, HMRC, in response to our recommendations, issued good practice principles for governing and administering tax reliefs.62 The Department told us that it was too early to evaluate the impact of these guidelines on formal product plans as the product owners have not had time to adjust their plans in response.63

In 2015, we also recommended that HMRC should identify which reliefs it considers require monitoring and evaluation, and that it should publish this information to enable Parliament to decide which reliefs may require further scrutiny or legislative change.64 The Department recognised it has to be able to explain why it has decided to obtain data for some tax reliefs but not others, to address any perception that the information HMRC currently holds is based on arbitrary criteria.65

HMRC told us that as a matter of routine it monitored the costs associated with tax reliefs wherever it could. But it emphasised that it had to be conscious of the trade-off between getting the data required to monitor costs and making sure undue burdens were not imposed on businesses. In some cases HMRC may decide it is so important to have the data that it justifies the burden on the economy to collect them. The Department told us that, in addition to its routine monitoring of the costs associated with tax reliefs, it carried out periodic monitoring of certain tax reliefs, particularly if ministers were interested in reforming a particular area.66 55 C&AG Report, para 3.1 56 Qq 78, 98–99 57 Q 78 58 Q 102 59 Q 78 60 Qq 79-80; C&AG Report, paras 3.7-3.9 61 Committee of Public Accounts, The effective management of tax reliefs, HC 892, 26 March 2015 62 Q 81; C&AG Report, paras 3.10-3.13 63 Q 81; C&AG Report, para 3.14 64 Committee of Public Accounts, HM Revenue & Customs performance in 2014–15, HC 393, 4 November 2015 65 Q 103 66 Qq 83–84 14 HM Revenue & Customs performance in 2015–16

We questioned whether there was sufficient information available to policy makers and analysts to enable them to assess the effectiveness of tax reliefs, citing the fact that it was only because the National Audit Office highlighted that entrepreneurs’ relief was costing the taxpayer three times more than intended that HMRC and HM Treasury reviewed the relief and changed how it worked. HMRC told us that every December it publishes the costs of about 180 tax reliefs, including entrepreneurs’ relief, out of a total of 400 reliefs that it recognises.67 HMRC noted that, by international standards, 180 was “a very high number” to be able to report costs on and said that it was part of its new guidelines for managing tax reliefs that variances in the costs must be monitored and the reasons for those variances understood.68

HMRC acknowledged that its published data on tax reliefs were scattered over a number of different publications, which had an adverse impact on the accessibility of the information.69 The Department agreed to consider collating and publishing all its data on tax reliefs in its annual report.70 67 Qq 81–82, 85-86; Office of Tax Simplification, Full list of tax reliefs. HMRC does not recognise all of the 1,156 reliefs identified by the Office of Tax Simplification (OTS) as being, in practice, a relief. For example, OTS classifies the exemption given to air crew from air passenger duty as a tax relief whereas, in HMRC’s view, the exemption is just part of the definition of the tax base for air passenger duty 68 Q 85, 87 69 Q 88; C&AG Report, para 23 70 Qq 88–91 HM Revenue & Customs performance in 2015–16 15