Good Tax: Status Issues

13 March 2018

Mark Groom provides an overview of the key employment status issues 

The title is intended to signpost the need for a deep dive review of the differing tax and national insurance (NI) treatment of employees, the fully self-employed, self-employed “workers”, gig-economy workers and those who engage through personal service companies. In a separate article, I look at the Tax and NIC facets in further detail. 

This whole area is rife with subjectivity, complexity, errors and in some cases sheer non-compliance at the boundaries; 90% non-compliance in the case of personal service companies and IR35 if HMRC’s statistics are correct. Without a holistic review, we’ll be in danger of applying sticking plasters on discrete problem areas as they arise, rather than achieving the optimum employment tax framework for the future. In this article I’ll scratch the surface of some of the issues arising.

Employment status

Accompanying the government’s response to Matthew Taylor’s excellent review was a welcome consultation on employment status. Broadly speaking, it asks how we can more clearly define the boundaries between statuses, what tests we should use and whether they should be legislated, if definitions for tax and employment rights should be aligned, and it includes a short section on platform based workers. It rightly notes that “the benefits of any legislative approach would have to be considered against any reduction in flexibility and the risk of making it easier to get around the legislation”.

Flexibility here refers to the ability of the courts to adapt to modern working practices.  However, doesn’t that also imply that a case heard today would not necessarily be decided in the same way as a case heard 10 years ago? That may not provide much comfort to the person who was first assessed 10 years ago, but whose appeal has taken that long to work its way to Tribunal! I would also ask, simply, if we think it is right that so many court cases are needed (and the time, cost and anxiety that goes with them) to determine the answer to employment status questions?

The questions posed in the consultation are a start, but other questions arise, including in relation to work elasticity and the overall economic and fiscal impact of any potential changes, which should be considered in a detailed tax review.

The three primary tests of employment

MacKenna J,’s summary of the essential conditions required for employment in Ready Mixed Concrete (South East) Ltd v Minister of Pensions and National Insurance [1968] 2 QB 497 will be familiar to most readers:

“(i) The servant agrees that, in consideration of a wage or other remuneration, he will provide his own work and skill in the performance of some service for his master. (ii) He agrees, expressly or impliedly, that in the performance of that service he will be subject to the other’s control in a sufficient degree to make that other master. (iii) The other provisions of the contract are consistent with its being a contract of service.”

The first condition combines “mutuality of obligation” with “personal service”. It also contextualises employment as a master-servant relationship, and leads naturally to the second condition, the servant submitting to the control of the master. The third condition is a check and balance which can negate an employment outcome if the remaining terms of the contract do not support that conclusion.

Mutuality – what does it mean and is it of any significance for tax purposes?

As an illustration of the confusion over the meaning of mutuality, look no further than Para 3.20 of the consultation document! This says that the first test in Ready Mixed Concrete is “An employer’s obligation to provide work, or to pay for work done, and the employee’s obligation to perform that work”. The italicised words are not quoted in the extract from the case noted above, but there are many other commentaries that interpret mutuality as including an obligation to provide work. 

Elias J explains in Stephenson v Delphi Diesel Systems Ltd [2003] ICR 471 that this is irrelevant:

“The significance of mutuality is that it determines whether there is a contract in existence at all. The significance of control is that it determines whether, if there is a contract in place, it can properly be classified as a contract of service, rather than some other kind of contract.

The issue whether the employed person is required to accept work if offered, or whether the employer is obliged to offer work as available is irrelevant to the question whether a contract exists at all during the period when the work is actually being performed…”

It will typically be the case that both employees and the self-employed enter into contracts to provide services for consideration (apart from volunteers), so mutuality doesn’t really help to differentiate between these two types of status for tax during periods actually worked. In contrast, a casual worker may need to establish a period of continuing employment to assert a claim for certain employment rights, which they may be able to do so if they can show ongoing mutual obligations subsisting in the gaps between tasks. 

However, even if a lack of any obligation on an engager to provide work and on a worker to do the work offered is not meaningful for tax purposes, it may still be indicative of self-employment. If nothing else, it may indicate a complete lack of control by the engager over what work the worker will do.

If we want to retain continuous periods of employment for employment rights, although this is not relevant for tax, it follows it will not be possible to truly align the employment definitions for these two purposes (as posed by Question 62 in the consultation document). However, there seems nothing wrong with that, provided there is clarity and certainty with the definitions for each purpose. Currently, mutuality seems to be at best, of very little practical consequence for tax purposes and at worst, extremely confused.

Personal service

Personal service (and its nemesis, substitution) has always been at the heart of employed / self-employed assessments. It is perhaps surprising that the only type of substitution that would support self-employment, was clarified only last year in Pimlico Plumbers Ltd and another v Smith [2017] EWCA Civ 51. This provides two guiding principles (1 and 2 below) and three examples of the second principle (3, 4 and 5 below). Anything less than this will not normally support self-employment:

  1. An unfettered right to substitute another person to do the work or perform the services is inconsistent with an undertaking to do so personally [i.e. supports self-employment].
  2. A conditional right to substitute another person may or may not be inconsistent with personal performance, depending upon the conditionality. It will depend on the precise contractual arrangements and, in particular, the extent to which the right of substitution is limited or occasional. 
  3. A right of substitution only when the contractor is unable to carry out the work will, subject to any exceptional facts, be consistent with personal performance. 
  4. A right of substitution limited only by the need to show that the substitute is as qualified as the contractor to do the work, whether or not that entails a particular procedure, will, subject to any exceptional facts, be inconsistent with personal performance. 
  5. A right to substitute only with the consent of another person who has an absolute and unqualified discretion to withhold consent will be consistent with personal performance. 

A point often missed is that substitution when a worker can’t work isn’t good enough; the worker must be able to send a substitute simply because they don’t want to work. Another important point is that it is acceptable for the engager to have a process to check that substitutes are suitably qualified. Provided this does not amount to an indirect right of veto for other reasons, this should not prevent the substitution from supporting the case for self-employment.

While this is very helpful in a world where personal service is retained in our status tests, it should be noted that this was done away with for tax purposes in April 2014 in the case of agency workers (see below). Also, there are many cases where it is simply unrealistic that a contractual right to substitution will ever be exercised; determining an individual’s true employment status in such cases is clearly much harder.   

Perhaps it is more expedient to ask, in what capacity is a person working when they are actually working: are they personally providing service (indicating employment), or are they delivering an outsourced service for which they take full responsibility (indicating self-employment)?

Control and the master servant relationship

Semantics like “master-servant” may conjure an image of some sort of Edwardian, Upstairs Downstairs relationship, but the term helps us identify the right sort of control i.e. that which may exist between the parties, which could indicate employment, as opposed to regulatory controls such as health and safety legislation, with which we must all comply and which should have no relevance to employment status.

In the 21st century, highly skilled/specialist roles, including the services provided by consultants present one of the biggest challenges to understanding and assessing the significance of control. However, Autoclenz v Belcher [2011] UKSC 41 reminds us that if a contractual right of control to a sufficient degree exists, then it doesn’t matter if that right is actually exercised.  Ultimately, if the engager wants to reserve the right to have the final say this will be an indicator of employment, even if de facto control is not necessary day to day.

We turn again to Ready Mixed Concrete where MacKenna J, identified control over what work is done, how it is done, and where and when it is done, as the four key components of control. When hiring a consultant, where and when are matters likely to be dictated by the particular work in question and will usually be a neutral indicator. Similarly, agreeing the scope of work that needs to be done and methodology to be used, will often be agreed between the parties in a grown up conversation, with very little (if any) de facto control while the work progresses. In these circumstances the control test can often reduce to, who has the ultimate right of control over such matters, should it need to be exercised.

Control over the manner in which work is done was chosen as the only test of control for the application of PAYE/NIC to agency workers under s44 ITEPA 2003 – see below. Learning points from that legislation should be reviewed when considering legislating employment status tests in due course. It may be that this is all that is needed in primary legislation, with other factors such as: the ability to profit from one’s own sound management; the provision of substantial equipment; financial risk; integration etc. included in secondary legislation, these comprising the third of MacKenna J’s tests in Ready Mixed Concrete.

Conclusions on employment status tests

The consultation on employment status is a first brave step forward, but are the questions asked bold enough?   For example, could we not also ask why, for tax and NIC purposes, there should be any boundaries between the employment statuses at all? It would be perfectly possible (and necessary for the purpose of establishing continuous periods of employment as indicated above) to retain the boundaries for employment law purposes, whatever is chosen for tax. Absent dividing lines for tax/NIC, there would be nothing left to deliberately try to “get around” (para 5.6 of the consultation) and the terror of incorrect employment status assessments and corresponding liabilities would be quelled.

Tax could still be collected in the most appropriate way for each type of status, so employees could still be subject to PAYE, but fully self-employed individuals could continue to pay under Self-Assessment; after all, if they are in business on their own account, they won’t know until the end of their accounting year whether they have made a profit or a loss, and so whether they have any tax to pay at all for that year. I return to the tax/NIC treatment of the employed and self-employed in the next article.

Worker status

Section 230(3)(b) Employment Rights Act 1996 seeks to provide certain rights and protections to self-employed people who are substantively and economically in a similar position to employees e.g. where they are to some degree subordinate to and dependent on their engagers. Hence Matthew Taylor’s suggestion to rename them “Dependent Contractors”. The legal definition of such a Limb (b) worker, is someone who undertakes to perform personally services for another party to a contract, unless that other party is a client of the worker’s business undertaking.  You have to turn to the case of Byrne Brothers (Formwork) Ltd v Baird & Others [2002] ICR 667 for confirmation of the policy intention noted above, which was indeed to protect those who may be “Dependent Contractors”.

Specifically, the main rights, entitlements and protections enjoyed by employees and Workers (but not the fully self-employed) relate to: protection for whistle-blowers; protection from unlawful deduction from wages; unlawful discrimination, working time, holiday pay, auto-enrolment pension contributions, and the National Minimum Wage/National Living Wage. The main rights and entitlements enjoyed only by employees (not workers or the fully self-employed) include SSP, Jobseekers Allowance, redundancy payments, statutory notice entitlement, flexible working rights, and protection under TUPE.

Commercially, a Worker is less likely to be exposed to financial risk. They will often be paid on an output or time basis so the profit or loss point faced by the fully self-employed is less relevant to the collection of taxes in the case of Workers. This suggests that some type of withholding may be appropriate, should there be any concerns over collection of tax. This could be administered through the platforms which are already sophisticated enough to be able to track workers and allocate work, which then forms the basis of payments to be made to them.

Question 42 of the consultation asks if we agree with the conclusion from Matthew Taylor’s review that the worker definition should place less emphasis on personal service (and by inference, substitution). If a Worker can send a substitute but nevertheless remains in a subordinate position in need of the rights and protections noted above, then perhaps substitution should not be allowed to stand in the way of Worker status. This was the approach taken with agency workers, using an amended application of personal service in April 2014 (see below). I see no reason why personal service in the Worker definition could not be amended similarly.

NIC presents more of a challenge. Para 6.19 notes Matthew Taylor’s suggestion to tax self-employed “Workers” like employees. In the article on tax/NIC below, I’ll explain why bringing Workers into PAYE/NIC may actually reduce the overall level of NIC paid by Workers, particularly in the gig-economy.

Agency workers

Agency Workers are yet another category of worker.   Agency workers are self-employed but they are subject to PAYE/NIC if (in short) they:

  1. personally provide services to another person under or in consequence of a contract between a client and a third person (the agency); unless
  2. it is shown that the manner in which they provide their services is not subject to (or to the right of) supervision, direction or control by any person (the “SDC” test).

This approach shines a light on any potential future employment status framework which might be enacted for employees/workers generally:

  • Note first that substitution has not been relevant to the PAYE/NIC treatment of agency workers since 2014.  Prior to that, an agency worker had to be under an obligation to personally provide service so, if they could send a substitute, the agency rules would not apply.  Since then, s44 applies whenever an agency worker personally provides service, even if he/she can send a substitute at other times. This will normally be the case for agency workers; however, exceptionally, their engager may be responsible for delivering an outsourced service to the client rather than supplying labour; in that case, s44 would be unlikely to apply and the relevant question should revert to whether or not the workers are employees of their engager.
  • Second, the control test presumes there is SDC (so that PAYE/NIC is the default tax treatment) unless the agency can show there is no SDC. Similarly, any new employment status test could deem employed status, unless it can be shown self-employed status applies.
  • Third, the SDC test does not consider all four components of control noted in Ready Mixed Concrete, but is only concerned with the manner in which (i.e. how) the services are provided.
  • Fourth, any overlay of the third set of tests noted in Ready Mixed Concreate (e.g. provision of own equipment) is abandoned in the case of agency workers.

Having said that, the SDC test is not without its own challenges, and it may be that this is too narrow a proxy for an employment status test more generally. However, it dispenses with mutuality and substitution which are often misunderstood, misapplied, and seem of limited relevance to achieving the policy objectives behind worker and agency worker status.  So, given its relative simplicity, it could provide a model for any future employment status framework, possibly also abandoning any overlay of the third set of tests noted in Ready Mixed Concreate.


One thing that doesn’t leap out in Matthew Taylor’s report, nor the government’s response to it, is that the rights, protections and entitlements intended to be provided to employees and Workers, largely go out the window if individuals engage through personal service companies (PSCs). This is on the basis that it’s not very helpful to suggest that the Worker’s entitlement can be self-delivered through his/her PSC.

So, going forward, if engagers find that contractors working for them are employees or Workers, and this was not intended or desired, they may simply insist on them engaging through PSCs to avoid all the Employment Law obligations they would otherwise be faced with. Workers may even find this attractive since, unless IR35 applies, they’ll also be able to access lower tax rates (see below).


This short paper cannot possibly surface all the issues that need to be thought about, but if it brings together things that people are already starting to discuss and crystallises any other thoughts, that will at least be something.