Delving into the National Audit Office report on HMRC’s service to taxpayers

Jon Thompson, the new Chief Executive of HMRC, has been set three key strategic objectives in HMRC’s departmental plan which has recently been updated:

  • To maximise tax revenues due and bear down on avoidance and evasion
  • To transform tax and payments to customers
  • To design and deliver a professional, efficient and engaged organisation

Within the documents some new customer service targets have also been set, presumably to be delivered primarily by Ruth Owen who now has the title of Director General of Customer Service (rather than Director General of Personal Tax). They are:

  • that telephone calls are to be answered on average within six minutes and that 85% of calls are to be answered
  • 80% of letters are to be answered in 15 days and 95% to 40 days
  • tax credit queries are to be answered from within the UK in 22 days and from international claimants within 92 days

So how do these new targets square with the National Audit Office report published on 25 May, and why were they published the day before the report not prior to the tax year as had been expected?

The report was extremely critical of HMRC’s service to taxpayers during 2014/15, stating that the taxpayers service ‘collapsed’, because the leadership team at HMRC overestimated the staff savings that could be made as a result of a move to automated telephony.  As call waiting times shot up, the leadership team decided to move 5,600 staff from dealing with PAYE queries to answering the telephone within call centres.

This led to a double whammy of problems, staff were not equipped with sufficient training to move onto call centre operations and the backlog of personal tax queries rose to 4.6 million by March 2015. According to HMRC, there was a significant improvement in performance in the autumn of 2015, when 2,400 new staff were recruited to call centres. They claimed that this meant personal tax staff were able to return to their core duties and managed to reduce the queries on hand to around 3 million by December 2015.

But let’s delve a little deeper into the National Audit Office (NAO) report…

  • Why did HMRC reduce the cost of its personal tax operations by £64 million more than requested (£237 million versus £193 million)? And why was it that personal tax bore all the cost and headcount reductions, whereas other parts of HMRC saw increased numbers? Perhaps it is because an assumption was made that RTI would reduce PAYE queries considerably,  HMRC say they automated processes to reduce the cost of managing PAYE from £144m to £31m, whereas in fact discrepancies shot up and not just because fewer people were working on them
  • Isn’t it odd that HMRC managed to maintain its service levels with reduced numbers until 2013/14, but as we reach the first year end under RTI, service began to deteriorate rapidly as calls increased by 22% (12m more calls). A perfect storm occurred: the volume of calls from employers and taxpayers increased just as the ability to get through using the new automated telephony system decreased
  • HMRC have told the NAO that of the 4.6 million tax record discrepancies in March 2015 3.2 million were ‘high priority’ cases. It’s not clear on what basis a discrepancy is classed as high or low priority, if there is an error in the data this is unacceptable and can infect other systems within government. One of the footnotes within the report indicates that tax code discrepancies are classed as low value items as they carry a ‘low risk of affecting someone’s tax’ – one would question why that is the case. There has been no discussion with stakeholders around HMRC’s criteria for suppressing ‘low value items’, or ‘ministerial priorities’.  Even these ‘ministerial priorities’ rose to a shocking 1.2 million from 20,000 by April 2015. HMRC believes that by the end of this tax year there will be just 20,000 discrepancies; but is that 20,000 individual records or 20,000 PAYE schemes, as one PAYE scheme can contain 150,000 records?
  • Given the numbers of PAYE schemes in dispute with HMRC, it seems highly unlikely that there are only 3.2 million high priority discrepancies. Equally, many PAYE schemes have not reported discrepancies (often because they’re advised by HMRC helpline staff to ignore them) and as payroll agents cannot view their clients’ tax accounts none of these PAYE schemes have reported discrepancies for them to be investigated. On this front the report notes HMRC’s failure to introduce a self-service facility for agents, which we now know will rely on the payroll software industry developing APIs rather than on HMRC providing access
  • Being able to answer the telephone more quickly is not the same as being able to answer queries correctly. It would be interesting to know if any research has been done, or is planned, that would indicate how many customers weren’t happy with the advice given and had to telephone again. In my experience call centre staff are becoming less knowledgeable about the tax system, particularly since the introduction of RTI
  • We’re told in the report that the 2,400 new call centre staff are simply ‘transitional’ and numbers are intended to go down again as the digital strategy is implemented. So how well have they been trained if they are temporary and how quickly are they likely to be removed?
  • HMRC’s measurements of customer satisfaction has remained stable since 2008 and yet these results do not appear to correlate with the research conducted by the NAO for this report, nor from the quoted research from the voluntary sector organisations that support taxpayers such as Citizens’ Advice, TaxAid and Tax Help for Older People. The NAO was surprised that HMRC does not ask taxpayers if they are satisfied with its service
  • For the first time the NAO calculated the cost to the taxpayer of waiting speak to HMRC. At a conservative £17 an hour (HMRC’s hourly rate attributed to taxpayers) the NAO calculated the cost to customers was £4 pounds for every £1 HMRC had saved.  If the costs to employers and agents had been factored in, which are much higher than £17 per hour, the cost burden would have been even more disconcerting.  The report notes with surprise the HMRC has no reliable data on the tax burden incurred by personal taxpayers or employers in respect of PAYE.  The quoted figures of a £63 million reduction in costs for taxpayers between April 2011 and March 2015 and a £300 million saving from introduction of RTI have been questioned by both the NAO and HMRC’s own the Admin Burdens Advisory Board.
  • HMRC contend that there is no evidence that ‘recent spells of poor service have affected tax revenues’.  One wonders how this can be?  if you can’t get through to discuss a disputed charge (or are told not to bother in the first place) when you know that the figures shown as tax receipts are incorrect, surely this is affecting recorded tax revenues?  Tax receipts have fallen from 15.9% in 2010/11 to 14.9% in 2014/15, which was not at all what the Office of Budget Responsibility expected to happen
  • Figures 5 and 10 within the report shows an interesting correlation between an increase in calls and implementation of major PAYE system changes.  There was, according to the NAO, a surge in calls relating to the new PAYE system in 2010/11 (this was the introduction of the National Insurance and PAYE service) and at the start of 2012/13 (when the RTI pilot began).  HMRC have not published any figures for calls being answered in 2010/11 some reason. The same is true of postal targets, these were not sustainable once RTI was live in 2014/15

In its summary the NAO states that HMRC’s plans to introduce digital services are very challenging in the light of their experience to date and that they must learn to plan sufficient capacity rather than expecting an immediate change in the behaviour of taxpayers in embracing digital services. They are urged to ‘build some realism into their assumptions’,  I would also add that agreeing those assumptions with stakeholders on the ground would make them more valuable and credible.

Given there now appears to be universal agreement from within HMRC (their Admin Burdens Advisory Board (ABAB)) and externally from the NAO, that predicted cost savings from RTI were actually translated into cost burdens, what will the conclusions be of the long-awaited Post-Implementation Review of RTI that we hope will see the light of day this summer?

HMRC have already told ABAB that they will take ‘a more rounded view of costs and benefits and do more to check and test what their model says against experience on the ground’, let’s hope so.