IR35 explained – or maybe not…

Self-confessed ‘IR35 addict’ Dave Chaplin takes a tongue-incheek look at the bizarre world he’s inhabited for the past 20
years.

Do you know the subtle differences between the driver of a cement mixer, an exotic dancer, a housekeeper, a plumber, a car washer, Catholic school boarders, a hotel waiter, a football referee, Polish nationals or indeed a caravan park in South Wales? No, well unfortunately I do. Let me try and explain my world – IR35.


It all started 20 years ago

Just over 20 years ago the then Labour Government dreamt up the idea that everyone should be assessing their relationships with their clients, to establish if really they are only ‘employed for tax purposes’ – which means tax should be paid like an employee, but without actually being an employee. Makes perfect sense, right?

This came in under the banner of ‘IR35’ in April 2000, where contractors running limited companies had to ask themselves the status question, the result of which could see them losing 25% of their income, and they unsurprisingly concluded it didn’t apply to them. Who could have guessed?


Time to double up on crazy

HMRC tried to enforce this through the courts and found they could only win less than half of cases, and for a long period gave up even trying. Then, in 2016, HMRC thought it would be a good idea to get businesses to ask the status question instead and renamed the now-toxic brand of IR35 as ‘Off-payroll Working’. It was introduced in April 2017, all under the waving banner of “tax fairness”.

This time it was the Conservative Party, tasked by the Treasury, to introduce more madness – whereby they had to convince everyone it was a good idea to oil the wheels of industry by making firms try to solve the equivalent of the Rubik’s Cube blindfolded every time they wanted to hire someone urgently. What could go wrong? So they gave 60,000 businesses, hiring over half-a-million contractors, the perilous task of conducting status determinations.

But this status stuff is easy, right? One just has to examine the relationship between the worker and the firm they work for and determine if it’s one of ‘disguised employment’ or not. Piece of cake – let me explain how you munch your way through it.


Let’s start mixing the status ingredients

Firstly, familiarise yourself with about 50 years of case law on the subject, and start by learning about a cement mixer driver from 1968 (Ready Mix Concrete).

This case will tell you three areas that need to be examined to work out if someone is really an employee – but only for tax. Luckily for us, having gathered all the facts, we had a judge tell us (Hall v Lorimer 1994) that it would be useful to look at all of them.

He explained that some facts in some instances are useful, but not in others, but maybe, but not so much, but either way, it should all be looked at. Can you feel the mist clearing yet?

Thanks to the ruling on the cement mixer driver, one thing we do know for sure is that an employee must be a single person – ground-breaking, eh?

Because if more than one person can show up to do the work, then it’s not just one person, although this isn’t now always the case. Trouble had brewed on the “is it personal service?” matter for almost 50 years, due to charging “armies of lawyers” (Kalwak 2007) writing words in contracts akin to “anyone can do this” – which managed to circumvent the “is this a single person” problem.

But then a car washer sponged over those rules (Autoclenz 2011) and, in the process, kickstarted a contractual interpretation revolution – where we started to question whether what everyone wrote down was really what everyone meant, especially if one party held more bargaining power than the other – do they ever not?

Some caravan owners helped us on our journey a little (Arnold v Britton 2015) when we then learnt that actually what we wrote down, however imprudent to one party, might just be tough on the other. A couple of years later an insurance broker (Wood v Capita 2017) helped us work out that where there was ambiguity in the things we thought we had written down, that we could resolve this with the novel idea of looking at the facts, and what actually happened.

This approach, where we could semi ignore the written words, and consider the reliable pipeline of facts, was then burst open by a plumber (Pimlico Plumbers 2018), whose case finally provided some much-needed clarity on whether more than one person was really just actually one person. Head hurting yet?

Because you see, if the person tries to swap themselves out for someone else, it may not count, if they were unable to do the work, rather than unwilling to – well, of course, that’s obvious, right?

And assuming they could swap out if they were unable to do the work, it still might be an employment issue if the firm they are helping makes too many kinds of checks before letting them do the work.

And if they can turn them away, for whatever reason, then sorry, game over.

It could be perceived as employment, even though they could turn them away, but never actually do, because the right to do so is more important than what actually happens (Kalwak again). Phew – just the clarity we needed as we dance our way around these rules towards the altar of contractual (un)certainty.


Bring on the dancer

One such dancer, of the exotic kind (Quashie 2021), emerged from the darkness of gentlemen’s club Stringfellows to come to the rescue and cast some light – she finally helped us all understand that if someone didn’t pay someone else, then they weren’t in a relationship – of the contractual kind, of course.

The dancer conundrum was related to the second test we should consider, called mutuality of obligation, which was served up into the mix thanks to some temporary waiters in 1983 (O’Kelly v Trusthouse Forte). This is the idea that promises are made between the masters and their servants, and many agree that this involves some sort of guarantee where there is an ongoing promise to offer and do work or get paid regardless – you know, like what happens when you hire an actual employee.

But the one party who has disagreed for years on this is HMRC, who publicly claimed from their well-dug trench that it just meant the master paid the servant if they worked – which is normally called consideration in legal terms, not mutuality of obligation. I say publicly, because behind closed doors HMRC appears to roll the dice on this one.

Variations I’ve seen range from the ‘policy view’ that it’s just consideration, which often appears in IR35 Opinion letters, which then differs from arguments made in tribunal Statements of Case, which then differs to ones in their skeleton arguments. And sometimes all these variations appear within the same case – I kid you not.


Red faces and red cards

Many of us, not necessarily football fans, took a four-year interest in the referees of the game, who had been trying to blow the whistle arguing this topic. It finally reached the Court of Appeal this year, where we hoped to finally get clarity. HMRC threw a ‘1’ and went with their dodgy consideration argument, which was ironic considering that while waiting for the decision they also sent our defence team a skeleton argument having thrown a ‘2’. HMRC lost the legal argument at the Court of Appeal, where the judges didn’t give a final decision on the case, but they did tell us that the correct test was to look at “sufficient” mutuality of obligation – without telling us what that was, except giving us pointers to there being something called the “right kind”. Thanks for that.

And now, using some of those firm foundations – sigh – we can move onto the next test and start to look at whether the relationship contains enough control to make the single person a servant of their master. Because if they aren’t, then they might not be a pretend employee.


Who’s in charge here?

Controlling someone splits into four elements: how, what, where and when – why? I don’t know personally, you’ll have to ask the cement mixer guy. And we need to think about different factors and how strongly they might point towards or away from employment. Let’s start with ‘how’, which is easy. We learnt in 2021 from a very unpleasant case (Various Claimants v Catholic Child-Welfare Society and others) that if the worker is a professional, one can turn a blind eye to the ‘how’ aspect when dealing with ‘skilled people’ who are left to do as they please. I suggest you take this as gospel because after reading the case, you can’t unread it.

As for the other factors, we have a useful tip based on a housekeeper visit to the tribunals (White v Troutbeck 2013) where we learnt that establishing whether someone could control someone, even if they actually didn’t, but could if they wanted, was what we needed to focus on. Facts apparently aren’t relevant here.


Imagine all the contracts

Now, if you are thinking this is all sounding far too easy, it’s time to introduce you to the wisdom that was introduced when an IT contractor took on the revenue in 2004 (Usetech v Young). What the programmer inspired the judge to conclude was that we should only open one eye to all the existing paperwork and use the other one to look at anything else that comes to hand. Then close both eyes and form an imaginary contract in our minds, and then apply the law to that. Stop laughing – the concept of the ‘hypothetical contract’ is accepted law. I mean it, stop chuckling, it’s not funny.


One final artistic step

So, to recap all of the above, this is the process: words written down should mean what they say, and should not be interfered with really, and may just be tough on one person, except in instances where there is an imbalance of power, or some ambiguity, in which case maybe we should look at the facts, which might override the words.

And when we look at the facts, we need to consider things that may or may not have happened but could have happened, and we need to see if enough of it has or could have happened. Once we’ve done that, we close our eyes and dream about what an agreement might look like, and then apply some of those rules again, but not all of them.

There is a final step – having formed in our minds what might be in a new contract, we then have to get our brushes out and now paint a picture of the contract on some canvas, and stand back from it, and work out if the picture means the person is an employee. For we are all artists now.


Assured mutual self destruction


Much of these cases took place in the employment tribunals, and I’m a firm believer that false self-employment and denying people their rights should be eradicated. But the arguments developed in these cases are undermining the very fabric of the market economy, freedom and certainty of contract. This uncertainty is being leveraged by HMRC to collect more tax – but it’s driving companies and people offshore into friendlier more certain jurisdictions, leaving less money in the UK to hire the very workers that should normally get rights.


Rock and roll

Hopefully, this explanation written purposefully against the backdrop of absurdity (did you get that one Uber fans?) of IR35 explains why I led a campaign for four years to try and stop this nonsense. But alas, that battle was lost, and we are where we are. Ironically, for me, it’s party time in the IR35 world with firms seeking help with this stuff, and I’m referred to as “Mr IR35”.

So, the sun is now shining on me – rock and roll! But if you do meet me at a party, I’ll be the one rocking in the corner. Sweet IR35 dreams everyone.


• Dave Chaplin is CEO of IR35 compliance solution IR35 Shield and author of IR35 & Off-Payroll Explained