Quality of service to personal taxpayers and replacing the Aspire contract Contents

1Customer service

1.On the basis of two reports by the Comptroller and Auditor General, we took evidence from HM Revenue and Customs (HMRC) on the quality of service for personal taxpayers and replacing the Aspire contract.1

2.HM Revenue and Customs (HMRC) is responsible for collecting tax revenue and interacts with million of taxpayers (online, by telephone or by post) each year. Most people have little need for direct contact with HMRC, with 86% of taxpayers paying their income tax automatically via Pay-As-You-Earn (PAYE). The remainder self-assess their own tax obligations due to being self-employed or having other sources of income.2 HMRC’s customer charter outlines taxpayers’ right to expect a ‘helpful, efficient and effective service’, designed to help customers pay the right amount of tax at the right time.3 HMRC measures its performance through targets for the time it takes to respond to customers’ telephone calls and letters.

3.Under the 2010 Spending Review, HMRC agreed with HM Treasury to take £193 million (24%) from the cost of its personal tax operations, with these cost reductions being heavily concentrated in the last two years. Over the period, HMRC reduced its personal tax headcount from 26,000 in 2010–11 to 15,000 in 2014–15 (a reduction of 42%) and reduced gross expenditure by £257 million (32%).4 Under the 2015 Spending Review, HMRC needs to reduce costs further by 2019–20.5 It plans to do this by developing its digital services and increasing taxpayers’ take-up of these.

Performance

4.Over the last five years, HMRC has changed its customer service targets to reflect changes in its performance and resources. In 2011 HMRC set a target to answer at least 90% of calls to its taxes helpline by March 2013, but subsequently revised its target for the 2012–13 year to 75%. In December 2012 it told the previous Committee it was deploying 1,000 extra temporary staff until March 2014 to achieve its target of answering 90% of calls. However in 2013–14 HMRC recognised that it would have significantly less resources and it would not be able to meet its target to handle 80% of calls. In 2014–15 and 2015–16, over a quarter of calls to HMRC were left unanswered, with 71% and 72% of calls handled in each year respectively. In one week in October 2015, the average time taken for HMRC to answer the phone peaked at 34 minutes for taxes calls and 47 minutes for self-assessment enquiries.6, 7 HMRC acknowledged that its performance in 2014–15 and the early part of 2015–16 was ‘unacceptable’, having apologised to customers for the impact that changes to customer services had had on the customer experience during this time.8

5.HMRC recovered its performance in the second half of 2015–16, after customer service deteriorated to such a level that HMRC decided to recruit 2,400 additional staff to stabilise services.9 The average time taken to answer calls improved to around five minutes by January 2016.10 HMRC told us that this improvement had been maintained, with customer service performance now in a ‘fundamentally different place’ when compared to the time period covered in the National Audit Office report.11 HMRC stated that in April and May 2016, it had handled 90% of calls in less than six minutes12 and that it planned to reduce waiting times further, aiming to answer 90% of calls in less than five minutes within the next 12 months, to meet a long-aspired standard.13

6.The decline in HMRC’s customer services occurred at the same time as it reduced its headcount in personal tax services. It reduced expenditure on personal tax by £257 million over the 2010 Spending Review period, exceeding its cost reduction target of £193 million by one third.14 Of the 10,800 staff released over the 2010–11 to 2014–15 Spending Review period, over half (5,600) were lost in the last year when there was a ‘huge spike’in customer waiting times.15 HMRC told us that the drop in performance was a result of the cumulative impact of bringing in new internal systems at the end of 2014–15, a cloud-based telephony system and digital mailing system. It said that ‘teething problems’ experienced when these systems were introduced, and the learning curve for staff using these systems, went ‘further, longer and deeper’ than anticpated.16

7.HMRC achieved almost one quarter (2,500) of its headcount reduction in personal tax through redundancies.17 We asked HMRC why it had released 2,500 trained staff only to spend time and money recruiting and training 2,400 new staff at short notice in 2015.18 HMRC told us that the majority of the staff released had been made redundant prior to 2014–15, as they were located in offices which HMRC planned to close and their contracts did not require them to work evenings and weekends.19 Under its workforce strategy, HMRC is moving away from a distributed call-centre network, consolidating staff in fewer sites, as it believes that this will deliver ‘better expertise and productivity’.20

8.HMRC’s staff reductions were based on an expectation that taxpayers’ demand for contact via telephone and mail would reduce as customers moved to online services.21 However, taxpayers’ use of the telephone to contact HMRC has not declined. We questioned HMRC’s understanding of its customers, as it plans to further digitise its services and reduce the number of staff in personal tax by a further 34% by 2020–21.22 HMRC considered that where its digital services were more mature (for example, self assessment and tax credit renewals) there was evidence of customers shifting to digital channels. It stated, for example, that 90% of taxpayers now do their self-assessment online and that, in the last two years, this had halved the number of telephone calls around the January deadline.23 However, HMRC admitted that getting customers to use online methods of contact (for example, a digital tax account) was the key risk it faced in personal tax services. It also recognised the need to manage relationships with the digitally excluded.24

9.HMRC maintained that it had learned lessons from what had gone wrong and that it would not expose taxpayers to the risk of service levels falling as it seeks to reduce costs over the 2015 Spending Review period. The planned cost reductions would be phased more smoothly the over the next five-year period, with reductions of no more than 9% in one year. In contrast, the previous five year Spending Review period had required the majority of savings in the last two years, with over 20% in one year. HMRC also told us that it had built greater contingency into its plans, so that it could respond to unexpected demand and had put in place a number of monitoring systems to enable it to predict and act more quickly.25

10.We questioned HMRC about the balance between its own expenditure on customer service and the costs incurred by its customers. In order to deliver savings in its personal tax services, HMRC reduced the cost of its telephony services. These savings were more than offset by the increased burden on taxpayers, as the time taxpayers spent waiting for HMRC to answer the phone increased.26 According to the National Audit Office, each £1 that HMRC cut from its telephone transaction costs between 2012–13 and 2015–16 created an additional £4 cost for taxpayers.27

11.HMRC told us that it intended to carry out further research on the costs associated with specific improvements in customer services and waiting times. It cited the ‘Erlang’ queuing model as ‘industry standard practice’ and told us that it intended to investigate the use of this mathematical model further, with a view to using it as the basis for its methodology to predict costs from 2017. HMRC agreed that it was important to consider the cost to the customer in making decisions about how much to spend on customer services and agreed that it needed to take this further into account. However, it stressed that these costs must be balanced against its own administration costs; there is a point at which reducing waiting times would lead to wasteful administration costs in terms of staff waiting for the phone to ring.28 HMRC recognised that it was not yet at this stage, acknowledging that in 2015–16 it had answered less than three quarters of its calls. However, it stressed that in April and May 2016, it answered 90% of calls in under 6 minutes. HMRC expected its key indicator: the time it takes to answer the telephone, to come down fairly dramatically by March 2017. It planned to reduce the time to below five minutes and to keep answering at least 90% of calls.29

12.Unlike some other government services, for example Jobseekers Allowance, HMRC’s helpline is not a Freephone number.30 Calls to HMRC’s 0300 number cost taxpayers up to 10 pence per minute from landlines and, where they are not covered by free minutes in call packages, up to 40 pence per minute from a mobile phone.31 Callers incur these direct costs whilst on hold to HMRC. With customers spending over 4 million hours on hold in 2015–16, these call charges are substantial and, particularly for low-income families, can represent a large amount of money.32

Customer service and collecting tax

13.In November 2015, we recommended that HMRC should identify what impact its poor level of service was having on tax revenues.33 In response, HMRC said it would work to try to identify evidence of the link between service and compliance and report back to us on this in 2016.34 In response to further questions on this matter, HMRC witnesses told us that their ‘gut feeling’ was that the quality of customer service must have some impact on tax revenues. However, while academic research to date has found examples that might illustrate a relationship, in common with the Organisation for Economic Co-operation and Development and United States’ Internal Revenue Service, HMRC had been unable to demonstrate a quantitative link.35

14.HMRC had performed some ‘exploratory analysis’ in this area, which suggested that a one percentage point improvement in customer satisfaction might lead to a increased income tax revenue of £43 million per year. However, HMRC considered this to be a ‘highly speculative estimate’, based on an extrapolation of ‘extremely weak’ evidence—notably proxies for customer experience that HMRC did not feel were good enough.36 It agreed that it needed to do further analysis to bottom-out the relationship between customer satisfaction and the tax gap.37 It acknowledged that this information would help support any decisions about how much to spend on customer services, and to determine the right balance between this expenditure and optimising tax revenue. HMRC would share its work in this area with the National Audit Office, or do it with them, and return to the Committee on this matter.

15.HMRC informed us that taxpayers’ attitudes to compliance were impacted by both HMRC’s actions to tackle those who try to avoid tax, as well how it assists those that want to pay their fair share.38 While there are diverse factors which may also affect an individual’s willingness to comply, for example their values and risk appetite, we heard that customers’ ability to get in touch with HMRC where they have queries, as well as their perception of the service HMRC provides, must also play a part in whether HMRC collects the right amount tax at the right time.39 HMRC is currently trialling an Australian initiative with customers, of thanking them for their contribution in their annual tax summaries, and expects to complete testing within a few weeks to see what customers think of this.40


1 C&AG’s Report, The quality of service for personal taxpayers, Session 2016–17, HC 717, 25 May 2016; National Audit Office, Memorandum for the House of Commons Public Accounts Committee, Replacing the Aspire contract, June 2016

2 C&AG’s Report, para 1.1

3 C&AG’s Report, para 1.2

4 HC 78, Q 65; C&AG’s Report, para 5

5 HC 78, Q 80

6 C&AG’s Report, para 4

7 HC 78, Qq 65, 83

9 HC, 78, Qq 65–66

11 HC 78, 59

12 HC 78, Q 60

13 HC 78, Qq 139–143

14 HC 78, Q 65

15 HC 78, 65

16 HC 78, Qq 66–67

17 HC 78, 74

18 HC 78, Qq 70–75

19 HC 79, Qq 75–79

20 HC 78, Qq 75–78

21 HC 78, Qq 62,66

22 HC 78, Qq 62, 80

23 HC 78, Q 62

24 HC 78, 61

25 HC 79, Q 80

26 HC 78, Q 81; C&AG’s Report, para 2.16

27 HC 78, Q 87; C&AG’s Report, paras 2.17–18

28 HC 78, Qq 87–95

29 HC 78, Qq 91–94

30 HC 78, Q 99

31 HC 78, 96

32 HC 78, Q 81

33 Committee of Public Accounts, HM Revenue & Customs performance in 2014–15, Sixth Report of Session 2015–16, HC 393, November 2015, Conclusions and recommendations, para 6

35 HC 78, Qq 101–103, 107; C&AG’s Report, paras 3.7–3.8

36 HC 78, Q 107

37 HC 78, Qq 107–109

38 HC 78, Q 103

40 HC 78, Qq 99–100




© Parliamentary copyright 2015

25 July 2016