The Evolution of UK R&D Tax reliefs – Part 2

With UK R&D tax reliefs now over 20 years old, a retrospective look at their evolution is timely.

This second of three articles focuses on problems around the definition of R&D and the consequences of internal changes at HMRC.

Initial inconsistency over the definition of R&D

In its original, low enquiry approach, the Treasury was fairly na?ve as to customer behaviour.

It had failed to learn from the Individual Learner Account fiasco that if you put up a sign saying “free money”, and don’t carry out adequate checks, then this will quickly lead to abuse.

The absence of early enquiries encouraged the submission of poorer quality claims. Over time, as poor quality claims were accepted this problem grew. (An examination of the facts in the BE Studios case will show just how lax the HMRC approach could be in this early period, even when enquiries were underway).

Gradually the approach began to be tightened and this led to inconsistencies between different HMRC offices. One office was threatening legitimate claimants with penalties because it needed to boost its investigation statistics, while others (as in the BE Studios case) were accepting incompetent claims.

On top of this, the launch of the large company relief meant that much larger claims needed to be considered and so greater clarity and uniformity in the Inland Revenue (IR) approach became necessary.

To try resolving confusions and improve the IR performance, a new and better structured set of Guidelines defining R&D was produced in 2004, and also an internal manual (the CIRD manual) telling HMRC staff how to approach claims was written and published.

(At the time of this publication the IR treated all such published internal instruction manuals as though they were legally binding and so, as this was a new manual, covering a new and unexplored area, some reticence was shown in dealing with complex issues such as subcontracting and software. At the time it was intended that this guidance would be improved as experience grew, but HMRC signally failed to do this, for many years).

HMRC merger

In 2005 the Inland Revenue merged with Customs and Excise to form HMRC.

This led to many organisational issues. As well as dealing with the merger there were also?major staffing cuts in hand, which were probably premature while the new structure was bedding in.?

It was therefore an unfortunate time to also try a “root and branch” restructuring of IR functions such that (for example) the Corporation Tax incentives and reliefs section (responsible for R&D amongst other things) was merged with VAT and instructed to seek out non-existent synergies between the two fields of taxation.

This, along with many other incoherent changes led to a serious managerial loss of focus on bread and butter operational issues, and perhaps coincidentally, after a little over a year in post the HMRC chairman was suddenly and without warning removed from his post, and? moved sideways into the Treasury with a brief for “blue sky thinking”.

It has never been explained whether this was for operational or for other reasons, but this sudden change in leadership, on top of the ongoing re-organisations, did not enhance HMRC stability, and so operational performance slipped in the order of priorities.

That meant that the early problems in R&D compliance were allowed to grow.

New specialist R&D Offices

By 2006 it was clear that the inconsistencies in HMRC R&D approaches required significant changes to be made, and it was decided that, outside the Large Business Office (dealing with the very largest companies) all R&D claims would be dealt with in specialist R&D offices.

These would also advise the Large Business Office of best approaches to the R&D subject area to aim for uniformity.

The existence of these new offices enabled better training and monitoring, but was also subject to an unfortunate requirement that newly set up HMRC offices should have a standard ratio of qualified to less-qualified staff.

R&D relief is a very complicated area requiring an understanding of legislation, and of complex guidelines that function as legislation without being structured in that way.

At the same time, staff have to be able to consider evidence in order to form reliable judgements as to the technological state of play in specialist areas when they have had no relevant training or experience of these areas.?

That requires a fairly high calibre of staff which is comparable perhaps to an area like transfer pricing. Instead, under the new staffing model, smaller claims were worked by staff members with very limited experience of difficult areas requiring argument and reasoning.

So, for example, staff whose previous experience had been limited to making “own account” adjustments to sole trader’s accounts might need to decide if a new field of software might represent an advance in its field and whether the company relationships would meet the complicated contractual positions needed to qualify under the legislation.

Over time this expertise gap contributed to a decline in the quality of work relating to smaller SME claims.

In addition, there have been a number of further structural changes all intended to reduce HMRC head count as part of cost cutting exercises.

One of these was an operational policy that R&D claims below a certain size (that was still reasonably substantial) would not be subject to any careful scrutiny. That has been a significant contributing factor to the serious compliance failures that were recently? highlighted by the National Audit Office and which have led to HMRC throwing hundreds of inadequately trained staff into R&D review work to try and address the past neglect.

There is an important point here, which is that economising on staff as a cost saving measure without considering what they do and what the consequences might be can end up costing far more than the apparent savings.

It is arguable that saving millions of pounds in staff costs was achieved at the cost of hundreds of millions or even billions in inadequate claims.

(Although their figure is questionable, HMRC recently concluded that the proportion of SME claims that were unsatisfactory was of the order of 25%).

What is R&D – especially for software and engineering projects?

While the DTI had original responsibility for the R&D definition, its subsequent interpretation was left in the hands of IR/HMRC staff without any technology background.?

HMRC staff felt unable to automatically accept the assessment of R&D by company staff?claiming to be competent professionals but lacked the ability to adjudicate claims themselves.

This caused considerable difficulty in agreeing whether a project amounted in whole or in part to R&D. The difficulty was particularly acute in areas such as software and engineering where companies would often be trying to solve problems that were specific to their own framework or circumstances, and there was no generic technology information in the public domain that could be reliably applied to the particular circumstances.

The engineering problem remains unresolved to this day, but with the growing importance of software development to the UK economy, something had to be done to try to address the skill shortage at HMRC.

The first effort involved technologists from the claim advisory companies providing training to HMRC inspectors in about 2007 as to how software development took place and what characteristics it would show. This was generally recognised as helpful, but the situation then became more complicated following the financial crash in 2007/8.

Thereafter new capital adequacy rules for banks and insurance companies meant that these financial institutions all had to embark on major and system wide improvements to their software systems to enable risk to be properly assessed. These improvements were frequently the subject of very large R&D relief claims that HMRC felt unable to agree were all revenue in nature, or applicable to R&D.

The issue became stalled for a number of years and looked to be heading for litigation, but (perhaps unfortunately for the R&D field in general) some deals were reached with the various companies in private so that the courts never got the opportunity to clarify and explain how software claims should be approached.

Recognising their lack of software expertise, HMRC began to involve their own internal software staff in helping examine R&D claims, and while this represented some progress, it was not ideal, particularly for SME developments, as the HMRC staff generally had a “large platform operating” background with different development styles and issues to those of smaller companies, and they also frequently lacked a proper understanding of how broad the R&D definition in the Guidelines was, so at times made ludicrous suggestions such as that if a development was known to be feasible it could not be R&D.

The correctness of software claims remained a continuing issue and was partly resolved when a joint working party of HMRC technologists and external software experts put together some improved guidance in the CIRD manual in about 2018.

This was a welcome development and represented an unusually positive approach to discussion and cooperation in clarifying the meaning of R&D. Unfortunately it came rather too late in the day.

The HMRC pattern of behaviour before then had been to limit dialogue on technical questions and avoid resolving areas of difficulty such as the limits of subcontracting or what was meant by improving a company’s own knowledge rather than materially changing the underlying technology.?

With changes in its Head Office staff it also tried at times to make various unjustifiable? attempts to change its interpretation of qualifying R&D without any discussion, or foundation in the legislation.?

A particularly toxic example of this was the sudden claim that any sale of the product of R&D would invalidate an R&D claim. This caused major anger and paralysis of the claim system? around 2009 until HMRC had to acknowledge that their approach was not in accordance with the law.

Instead they finally did what they should have done at outset which was to raise their practical concerns with Ministers and secure appropriate and limited legislation changes restricting the costs that would qualify when a sale had taken place. Other examples could also be cited.

Over the years there has been a pattern of HMRC behaviour which does appear dysfunctional.

At an operational level the SME regime was allowed to deteriorate by non-enforcement of the rules, while occasionally there were attempts from the central technical team to change the law by fiat rather than by actual change to the legislation.

The foundation of any successful tax system has to be based on the correct application of the law, and for HMRC to have departed from this so often (most recently on questions of subsidy and subcontracting) has created a regrettable lack of trust in the reputable advisor community, which makes collaborative compliance harder to achieve.

The prolonged failure to police SME compliance has led to the current mess where large numbers of untrained staff have suddenly been instructed to take up enquiries into SME claims and to fight them and this has led to organisational chaos as the staff do not understand the subject area and the internal HMRC software staff have been unable to spare the resource need to reach valid judgements or even to discuss them.

Available now... click here for part 3 of my Evolution of UK R&D Tax reliefs series which looks at how reduced HMRC oversight allowed fraud to flourish, leading to significant losses and prompting a shift towards a single simplified scheme.

?

Note: this article series is based on exclusive interviews with a senior ex-HMRC inspector who was instrumental in the initial set-up and oversight of the R&D regime.

Interviews conducted by Rufus Meakin

Rufus Meakin helps companies prepare complex R&D Tax Credit claims where robust HMRC compliance is essential.

If you would like to discuss any aspect of your R&D Tax Credit claim then please feel free to call me on 0794 110 3285.