.png)
.png)
KEY POINTS
- Delays in processing VAT registration applications were caused by more stringent checks.
- Discrepancy in March 2008 and June 2008 processing figures caused by capping.
- Why did HMRC misreport the figures?
- HMRC have to act honestly if they expect taxpayers to do so.
Every month from August 2007 to March 2008, HMRC deliberately understated the extent of the VAT registration delays. When calculating and reporting the average waiting time for a VAT registration, HMRC capped the maximum waiting time of any application at 60 days, even where the application had in fact been outstanding for much longer.
The result has misled Parliament, the National Audit Office and taxpayers generally, and inevitably calls into question the accuracy of HMRC's other published figures on their VAT registration performance.
The problem
The background was the serious and increasing VAT registration delays in 2006-07, which by July 2007 had led to a backlog of over 60,000 registrations. HMRC VAT registration offices were advising of waiting times of up to 70 days after receipt before an application was even opened.
The problem had been steadily building up over the previous 12 months because HMRC had introduced much more stringent checks on VAT registration applications in 2006 to counter missing trader intra-community fraud, but had failed to allocate more staff to the task.
Faced with the mounting delays, the Institute of Chartered Accountants in England & Wales and other professional bodies became increasingly critical of HMRC's lack of resource allocation.
In August 2007, the ICAEW published TAXREP 53/07, Delays in registering businesses for VAT, detailing many cases of serious delay and making 12 recommendations for improvements.
Recommendation 2 was that HMRC should publish monthly registration performance figures.
Following constructive discussions at the Joint VAT Consultative Committee, the main platform for discussion between HMRC and representative trade and professional bodies, HMRC agreed to publish their monthly performance figures.
The ICAEW also recommended that the methodology should be agreed in advance with the professional bodies and the results scrutinised by the National Audit Office, but, unfortunately as it turns out, this was rejected.
In the next few months, HMRC devoted more resources to tackle the backlog and refined their approach. The result was a considerable success in bringing back the delays to an acceptable level. HMRC's figures showed that the average time to process a correctly completed, low risk VAT registration (some 70% of the total) had dropped from 42 days in August 2007 to 11 days in March 2008.
But in April to June 2008 HMRC's figures worsened dramatically: they showed that the average processing time for this category had more than doubled to 25 days in June 2008.
In response to questions why this had happened, HMRC explained that, when collating and reporting their performance statistics, they had previously capped the maximum individual delay at 60 days, but had ceased to do so from April onwards. As HMRC state in the Joint VAT Consultative Committee minutes for the 9 July 2008 meeting:
'The average time to process a low risk application is show(n) as having risen. While this might appear cause for concern, the apparent slippage is due to changes in the way we measure the age of a case.
'From 1 April we have been able to measure the full age of every case closed, previously this was capped at 60 days, which has led to the increase in the average time taken for April/June.'
NAO and Parliament also misled
HMRC's 2007-2008 accounts were published in mid-July 2008, together with the National Audit Office report on them (HMRC 2007-08 Accounts, Report by the Comptroller and Auditor General).
The NAO report devotes ten pages, some 20% of the total, to an analysis of the VAT registration delays, but it is based on HMRC's understated, capped, figures. The report makes no reference to the capping, of which the NAO appear to be unaware.
This is all the more strange since HMRC had published their (uncapped) April 2008 registration figures in May 2008, some two months before the NAO Report was published, and so would have been well aware at a senior level of the capping and that they had ended it.
Similarly, there was no mention of systematic capping in parliamentary written answers on the delays given by ministers in response to questions from MPs.
Why was it done?
The 'why' may never be known, but the effect is clear: the average delay was greatly understated since the average itself was skewed. We will never know the true figures, but HMRC's changeover to correct accounting in March/April 2008 more than doubled the average number of days.
Applied to August 2007, that would have meant an average of 84 days' wait for a VAT registration, which was certainly the experience of many taxpayers, and in line with the delays reported by VAT registration offices themselves. In addition, HMRC began to attack the backlog on a last-in-first-out basis, processing the most recent applications first. The 60-day cap masked this as well.
Reflecting the experience of many taxpayers, the NAO report paints a picture approaching melt-down in HMRC's VAT registration processes in the middle of last year. When HMRC began to publish their statistics, no one expected them to be 100% accurate.
It may well not have been possible for HMRC to carry out an accurate count, but no one would have objected to the use of best estimates, honestly made and based on statistically valid sampling techniques. Systematic undisclosed capping, which was bound to understate and mislead, was hardly the best answer.
HMRC have told Taxation that they had 'in no way tried to mislead anyone over VAT registration statistics' and that they had 'been consistently clear about the basis on which statistics on VAT registration are calculated, including that there have been limitations in the past'.
Why then, even assuming that for some extraordinary reason the 60-day cap was the only way in which HMRC could record their monthly performance figures, did they not disclose this systematic bias for almost a year?
Does it matter?
In one very superficial sense, it probably does not matter. Whatever the true figures, there can be no doubt that HMRC have now tackled the vast bulk of the delayed VAT registrations and that a more normal service has been resumed.
But HMRC are quick to lecture taxpayers on the need to be honest and transparent when dealing with them. By failing to act honestly and transparently themselves on such a major public issue, HMRC have damaged their own reputation and integrity.
They can hardly be surprised if in future representative bodies and taxpayers are more prepared to question and seek details for the basis of HMRC's figures, whether on service delivery or when used to support a change in tax policy. Calls for independent audit will become harder to ignore. None of this can be good for tax policy.
That is why it is in both the public and their own interest for HMRC to give a full, clear and credible explanation.
John Arnold is the director of European VAT Services for Ernst & Young in the Netherlands, and chairman of the ICAEW VAT and Duties Committee.
