This time, reports have emerged that HMRC cannot prove that CEST - the tool the taxman built to make IR35 assessments following public sector reform - is accurate.
This has fuelled debate to the extent that arguments are being made for a full inquiry by the Public Accounts Committee into the build of HMRC’s IR35 tool.
Of course, many IR35 experts, including Qdos Contractor, have consistently voiced concerns about the reliability of the tool, while 81% of independent workers only recently stated they would be deterred from working with a client or through an agency if CEST was used.
That said, news of HMRC admitting to Contractor Calculator that it holds no evidence to show the tool is capable of making accurate IR35 decisions, is perhaps the first time we’ve heard it from HMRC themselves.
In response to a Contractor Calculator request, HMRC revealed ‘the CEST tool testing was done by workshop’ and ‘the only documented output of the workshops is the set of rules used by the tool.’
Furthermore, HMRC does ‘not have a record of how each question was answered as part of that testing, only the end determination’, signalling that it holds no detailed test data or evidence to show CEST’s accuracy.
Given HMRC has been adamant that CEST is reliable and accurate, this news will not sit well with contractors nor public sector engagers who have both been encouraged to trust the IR35 tool implicitly. Furthermore, HMRC has, in the past at least, shown what now looks to be blind confidence in the tool.
It was only in January of this year that Mel Stride, The Financial Secretary to the Treasury, backed the controversial technology, arguing that CEST “ has been tested for accuracy against known case law and settled cases. It provides an answer for 85 percent of uses, and HMRC will stand by the result given unless a compliance check finds the information provided isn’t accurate.”
This news will not have won HMRC much sympathy though, particularly since it emerged that they tested the tool against only seven IR35 tribunal cases. Is this enough of a foundation to build a technology which, should further IR35 reform materialise, could be used to set the status of the UK’s 2 million or so contractors? Arguably not.
Since its release just over a year ago, CEST has been used some 400,000 times. Granted, this figure might not represent the number of actual IR35 determinations it has delivered, but it does make you wonder how many incorrect answers it has given which were then used as the basis to set a contractor’s IR35 status.
After losing an IR35 case against a construction worker (MCDM Ltd) just the other week, this latest series of events reflects how things have gone from bad to worse for HMRC. Further analysis has shown that CEST - despite having not been built at the time of the working arrangement in question - was ‘unable to determine the tax status’. This only goes to strengthen the argument for a thorough examination into the logic of the tool.
If the shoe was on the other foot, there’s every reason to believe HMRC would have no hesitation in launching a full investigation. Having signalled its intent in recent months with regards to IR35, you get the feeling HMRC would leave no stone unturned in getting to the bottom of such a mistake.
Given the potential impact of incorrect IR35 decisions is colossal, for contractors, their public sector clients and perhaps in time, private sector engagers too, our sector is well within its rights to expect answers.
You only need to cast your mind back to last month - when it emerged a BBC worker attempted suicide due to the organisation’s handling of their tax status - to see the monumental repercussions of inaccurate IR35 decisions.
As part of the IR35 consultation, the Government has promised to review CEST. However, reports that HMRC supposedly built a tool of such significance without the expected level of detail does raise a pertinent question; is it time they faced a formal and thorough inquiry into the CEST project by the Public Accounts Committee?
Ask away! One of our team will get back to you!