Royal Mencap Society v Tomlinson-Blake

It is very common in the care sector for workers to agree to “sleep in” overnight at premises where elderly, disabled or otherwise vulnerable people live, on the basis that they can be called on if assistance is required in the night but otherwise have no duties. The agreement may be either free-standing or an add-on to a contract of employment involving other duties and will typically be in return for a fixed amount, with an entitlement to further pay if the worker is in fact called on. Residential staff, both in the care sector and elsewhere, may also be required to be “on-call” overnight. The broad issue in both these appeals is whether the entirety of the period spent on the premises under such arrangements must be taken into account in calculating an employer’s obligations under the National Minimum Wage Regulations or only such time as is spent actually performing some specific activity.

There is a good deal of case-law relevant to that question, but its effect is not straightforward. Last year the Employment Appeal Tribunal (Simler P sitting alone) heard together three sleep-in cases with a view to giving, so far as possible, authoritative guidance. The cases were Focus Care Agency Ltd v Roberts, Frudd v The Partington Group Ltd and Royal Mencap Society v Tomlinson-Blake: I will refer to them collectively as Focus. Judgment was handed down on 21 April 2017 (reported at [2017] ICR 1186). Simler P reviewed the authorities and found in each case in favour of the claimant, though only in the Mencap case was her decision directly decisive of the national minimum wage issue. The employers in that case have appealed, with the permission of the EAT itself.

What is before us is the appeal in Mencap, together with Shannon v Rampersad, which is an appeal against a decision of HHJ Peter Clark (reported at [2015] IRLR 982) in another sleep-in case, though on rather untypical facts, which preceded the judgment in Focus. In this case it is the employee who is the Appellant.

In Mencap the employers are represented before us by Mr David Reade QC, leading Mr Niran de Silva. The Respondent employee is represented by Mr Sean Jones QC. Both also appeared in the EAT. In Shannon the Appellant is represented by Mr Caspar Glyn QC, leading Ms Chesca Lord, and the Respondents by Mr Mark Sutton QC. None of them appeared in the EAT.

Because of the importance of the case to employers in the care sector permission was given to Care England and the Local Government Association to intervene in the Mencap appeal. We had helpful written submissions on their behalf from, respectively, Mr Timothy Brennan QC and Ms Anne Redston; and both also made brief oral submissions. Both interveners also sought to adduce evidence addressing the policy aspects of the issue. We declined to consider the evidence filed by the LGA because the parties, and in particular the employees, had had no opportunity to answer it; but all or most of it was likely to have been inadmissible in any event, since the job of the Court is not to decide what the law ought to be in this area but what it is.

It will be convenient if I consider first the correct approach to sleep-in cases generally before turning to the particulars of the two appeals before us. The decided cases display considerable variations in their detailed facts, but the essentials of the situation which falls to be considered are that the worker is contractually obliged to spend the night at or near their workplace on the basis that they are expected to sleep for all or most of the period but may be woken if required to undertake some specific activity; and when I refer to “sleeping in” or cognate terms that is what I mean. In some cases the workplace where the worker sleeps in may also be their home; but in so far as that involves different considerations I will address them when we come to them.

Section 1 of the National Minimum Wage Act 1998 creates the basic right for a worker to be paid, in any “pay reference period”, the national minimum wage (“NMW”) in the form of an hourly rate of remuneration, to be prescribed from time to time by the Secretary of State. The pay reference period is also to be prescribed by the Secretary of State: for present purposes it is enough to say that it is typically a week or a month.

It is not always straightforward to identify at what hourly rate a worker has in fact been remunerated over a particular period, so as to know whether that rate is less than the prescribed minimum. Section 2 empowers the Secretary of State to make regulations providing for how to determine that question. I need refer specifically only to sub-section (3) (a), which provides that the regulations may make provision with respect to:

“circumstances in which, times at which, or the time for which, a person is to be treated as, or as not, working, and the extent to which a person is to be so treated”.

Section 5 (1) provides that before making the first regulations under (among other provisions) section 2 the Secretary of State shall refer the various matters listed at sub-section (2) to the Low Pay Commission. These include:

“what method or methods should be used for determining under section 2 above the hourly rate at which a person is to be regarded as remunerated for the purposes of this Act.”

Sub-section (3) provides:

“Where matters are referred to the Low Pay Commission under subsection (1) above, the Commission shall, after considering those matters, make a report to the Prime Minister and the Secretary of State which shall contain the Commission's recommendations about each of those matters.”

Sub-section (4) requires the Secretary of State, if he or she decides to make regulations which (among other things) “differ from the recommendations of the Commission”, to lay a report before each House of Parliament containing a statement of the reasons for that decision.

The First Report of the Low Pay Commission, addressing the matters identified at section 5 (2) of the Act, was published in June 1998. (Footnote: ¹) As appears from the preceding paragraph, it has a particularly important status in relation to the Regulations. The Secretary of State made no report to Parliament pursuant to section 5 (4) of the 1998 Act to the effect that the Government was not proposing to follow the Commission’s recommendations in making the 1999 Regulations: indeed the Government said in terms that it proposed to follow them. In those circumstances it was common ground before us that the recommendations in the Report are admissible as an aid to the construction of the Regulations: I should note that this was not the position in Focus – see paras. 44-46 below.

How the NMW should be calculated in the case of workers who are required to be at the place of work and available to work and, more particularly for our purposes, of workers who are required to “sleep in” at work was the subject of formal recommendations in the report. Recommendations 11 and 12 read:

“11.

The actual working time definition should define what constitutes working time for the purposes of the National Minimum Wage. The National Minimum Wage should also apply to all working time when a worker is required by the employer to be at the place of work and available for work, even if no work is available for certain periods. (para. 4.33)

12 For hours when workers are paid to sleep on the work premises, workers and employers should agree their allowance, as they do now. But workers should be entitled to the National Minimum Wage for all times when they are awake and required to be available for work. (para. 4.34)”

Paragraphs 4.33 and 4.34 of the report, as referred to in those recommendations, read as follows:

“4.33.

We recommend that the actual working time definition should define what constitutes working time for the purposes of the National Minimum Wage. The National Minimum Wage should also apply to all working time when a worker is required by the employer to be at the place of work and available for work, even if no work is available for certain periods. This definition has the advantage of covering agreed ‘downtime’ hours when workers are on-site but unable to work (e.g. because of machine breakdowns or lack of materials). It includes all agreed overtime hours, including call-out hours for emergencies, but it does not include standby or on-call periods away from the employer’s workplace or agreed rest periods.

4.34.

Certain workers, such as those who are required to be on-call and sleep on their employer’s premises (e.g. in residential homes or youth hostels), need special treatment. For hours when workers are paid to sleep on the premises, we recommend that workers and employers should agree their allowance, as they do now. But workers should be entitled to the National Minimum Wage for all times when they are awake and required to be available for work.”

Those recommendations, and the quoted passages explaining them, are of fundamental importance to this appeal. In particular, the Commission deals expressly with the case of workers who sleep in in “residential homes” but who are required to be “on-call”. In such cases, which are essentially the kinds of case with which we are concerned, it refers to what it evidently understood to be the existing practice (“as they do now”), namely that the worker would be paid an agreed “allowance”; but it recommends that the only time that would count for NMW purposes should be “when they are awake and required to be available for work”.

There have been regular reports from the Commission since its first report, together with numerous research papers and other publications. It appears, however, that there is nothing that casts light on the origins of the Amendment Regulations made in 2000.

The only other report to which we were referred is the Commission’s fourth, published in March 2003, which contains a section on pay in the social care sector. Paras. 3.55-3.56 read as follows:

“3.55

Having considered the overall economic impact of the minimum wage on the social care sector, we now turn to a more specific policy question of particular relevance to the sector, namely how the National Minimum Wage applies to ‘sleepovers’. ‘Sleepovers’ cover situations where, for example, someone works a day shift in a care home and then sleeps in the home overnight and is available to deal with emergencies but would not necessarily expect to be woken. Similar situations may arise with workers who care for an elderly or disabled person in his or her own home and sleep on the premises, or with wardens in sheltered housing who are available to deal with emergencies during the night.

3.56

In our earlier reports we concluded that for ‘sleepovers’, where the assumption is that the worker will not normally be woken, the National Minimum Wage should not apply (in practice an allowance is usually paid) but workers should be entitled to the National Minimum Wage for the times when they were awake and required to be available for work. We noted the difference between these ‘sleepovers’ and on-call or standby arrangements where a worker is required to be at the workplace outside of normal working hours with the expectation that he or she will be required to work, for which the National Minimum Wage is payable. The Government accepted and acted upon our recommendations, and we believe that these still reflect the right approach.”

Paras. 3.57 and 3.58 then go on to refer to “certain recent Employment Appeal Tribunal judgments [which] have held that the National Minimum Wage was payable in circumstances where the worker was able to sleep at times during the night” and say that there has been “concern that these judgments might imply that the National Minimum Wage was payable in the ‘sleepover’ cases which we had considered earlier”. The EAT decisions referred to appear to be those in the British Nursing Association and Scottbridge cases, which I discuss at paras. 49-62 below. The report says that it is clear that some uncertainty remains and continues:

“We believe it important to ensure that the position as set out in our earlier recommendations on ‘sleepovers’ is maintained and clarified. We therefore recommend that the Government should examine whether the present uncertainty over the treatment of ‘sleepovers’ can best be resolved through revised guidance, or whether a change to the Regulations is required.”

There is no evidence of the Government having given any further consideration to the issue, despite that recommendation. Certainly we were not referred to any published discussion or statement. (Footnote: ²)

The National Minimum Wage Regulations 1999 are the first regulations made under the 1998 Act and came into force on 1 April 1999. They were amended from time to time thereafter but with effect from 6 April 2015 they have been replaced by the National Minimum Wage Regulations 2015. These are significantly different in their structure from the 1999 Regulations, but it is stated in the Explanatory Memorandum that they are intended only as consolidating legislation, and they are thus not intended to change the effect of the Regulations as they stood previously. The 2015 Regulations are the relevant version in Mencap but the 1999 Regulations (as amended) remained in force at the period covered by the claim in Shannon. I shall accordingly need refer to both; but I would in fact have had to do so even if only Mencap was before us, because the authorities are almost all concerned with the 1999 Regulations.

The broad scheme of the Regulations is that in order to ascertain the hourly rate in fact paid in a pay reference period the total number of hours worked in the period is divided by the total remuneration paid in respect of it: see regulation 14 of the 1999 Regulations and regulation 7 of the 2015 Regulations. We are concerned with the first element in that calculation, i.e. the number of hours worked. That is the subject of Part 5 of the 2015 Regulations, which is headed “Hours Worked for the Purposes of the National Minimum Wage”. This is divided into five Chapters. Chapter 1 is essentially introductory. Regulation 17 reads:

“In regulation 7 (calculation to determine whether the national minimum wage has been paid), the hours of work in the pay reference period are the hours worked or treated as worked by the worker in the pay reference period as determined—

(a)

for salaried hours work, in accordance with Chapter 2;

(b)

for time work, in accordance with Chapter 3;

(c)

for output work, in accordance with Chapter 4;

(d)

for unmeasured work, in accordance with Chapter 5.”

The organisation of the 1999 Regulations was less elegant, but it used the same fourfold categorisation. The purpose of the categorisation is not, as such, to define what is meant by “work” but to recognise that different kinds of contract require different approaches to measuring the hours worked under them and to provide accordingly.

In Mencap the claim is that the worker was doing time work during the relevant period and in Shannon that he was doing salaried hours work. Those are accordingly the categories with which we are directly concerned. But I also need, because it is relevant to one of the authorities which we will have to consider, briefly to summarise the provisions about “unmeasured” work.

“Time work” is defined in regulation 3 of the 1999 Regulations as follows:

“In these Regulations ‘time work’ means –

(a)

work that is paid for under a worker's contract by reference to the time for which a worker works and is not salaried hours work;

(b)

work that is paid for under a worker's contract by reference to a measure of the output of the worker per hour or other period of time during the whole of which the worker is required to work, and is not salaried hours work; and

(c)

work that would fall within paragraph (b) but for the fact that the worker is paid by reference to the length of the period of time alone when his output does not exceed a particular level.”

There is no definition of “work”. The corresponding provision of the 2015 Regulations is regulation 30. I need not set it out because it is in materially identical terms.

Regulation 15 of the 1999 Regulations contains “provisions in relation to time work”, which address particular instances of what counts as such work. For our purposes the relevant provision is that concerning cases where a worker is required to be “available” for work. Such cases were initially covered by paragraph (1), which read:

“In addition to time when a worker is working, time work includes time when a worker is available at or near a place of work, other than his home, for the purpose of doing time work and is required to be available for such work except that, in relation to a worker who by arrangement sleeps at or near a place of work, time during the hours he is permitted to sleep shall only be treated as being time work when the worker is awake for the purpose of working.”

(The other paragraphs of regulation 15 concern such matters as time spent travelling or being trained.)

By regulation 6 of the National Minimum Wage Regulations 1999 (Amendment) Regulations 2000 (“the Amendment Regulations”), which came into effect on 1 October 2000, that paragraph was substituted, and a new paragraph (1A) added, as follows:

“(1)

Subject to paragraph (1A), time work includes time when a worker is available at or near a place of work for the purpose of doing time work and is required to be available for such work except where –

(a)

the worker's home is at or near the place of work; and

(b)

the time is time the worker is entitled to spend at home.

(1A) In relation to a worker who by arrangement sleeps at or near a place of work and is provided with suitable facilities for sleeping, time during the hours he is permitted to use those facilities for the purpose of sleeping shall only be treated as being time work when the worker is awake for the purpose of working.”

The corresponding provision of the 2015 Regulations is regulation 32 (part of Chapter 3), which reads:

“(1)

Time work includes hours when a worker is available, and required to be available, at or near a place of work for the purposes of working unless the worker is at home.

(2)

In paragraph (1), hours when a worker is ‘available’ only includes hours when the worker is awake for the purposes of working, even if a worker by arrangement sleeps at or near a place of work and the employer provides suitable facilities for sleeping.”

I will refer to paragraph 15 (1) and its successors, and their cognates in relation to salaried hours work, as the “availability provisions”.

The provisions of regulations 3 and 15 thus establish what counts as “time work”. Regulation 20 then provides straightforwardly that “[t]he time work worked by a worker in a pay reference period shall be the total number of hours of time work done by him in the pay reference period”.

“Salaried hours work” is defined in regulation 4 of the 1999 Regulations. Paragraph (1) reads:

“In these Regulations ‘salaried hours work’ means work–

(a)

that is done under a contract to do salaried hours work; and

(b)

that falls within paragraph (6) below.”

For present purposes I need not set out paragraph (6) or any of the other somewhat elaborate provisions of the definition. The definition is extensively re-cast, without any change in the meaning, by regulation 21 of the 2015 Regulations, paragraph (1) of which reads:

“‘Salaried hours work’ is work that is done under a worker’s contract and which meets the conditions in paragraphs (2) to (5) …”

Again, I need not set out paragraphs (2)-(5). What matters for our purposes is that the definition, like that of “time work”, refers to “work” but does not define it.

Regulation 21 of the 1999 Regulations prescribes how to determine the number of hours of salaried hours work done in a pay reference period in the paradigm case where (broadly speaking) the worker simply works his “basic” hours; regulation 22 provides for the case where those basic hours are exceeded. The corresponding provisions of the 2015 Regulations are regulations 24-28.

Regulation 16 contains “provisions in relation to salaried hours work” which are in the material respects identical to those of regulation 15, and it was similarly amended by the Amendment Regulations. I need not set out the original regulation 16 (1), but I should set out paragraphs (1) and (1A) as introduced by those Regulations:

“(1)

Subject to paragraph (1A), time when a worker is available at or near a place of work for the purpose of doing salaried hours work and is required to be available for such work shall be treated as being working hours for the purpose of and to the extent mentioned in regulation 22(3)(d) and (4)(b) except where–

(a)

the worker's home is at or near the place or work; and

(b)

the time is time the worker is entitled to spend at home.

(1A) In relation to a worker who by arrangement sleeps at or near a place of work and is provided with suitable facilities for sleeping, time during the hours he is permitted to use those facilities for the purpose of sleeping shall only be treated as being salaried hours work when the worker is awake for the purpose of working.”

The corresponding provisions of the 2015 Regulations are at regulation 27 (1) and (2), which read:

“(1)

The hours listed in sub-paragraphs (a) to (c) are treated as worked for the purposes of determining whether the worker works more than the basic hours in the calculation year … and, where the worker does, the number of hours of salaried hours work in that year … —

(a)

…

(b)

hours a worker is available at or near a place of work for the purposes of working, unless the worker is at home;

(c)

…

(2)

In paragraph (1)(b), hours when a worker is available only includes hours when the worker is awake for the purposes of working, even if a worker is required to sleep at or near a place of work and the employer provides suitable facilities for sleeping.”

“Unmeasured work” is defined in regulation 6 of the 1999 Regulations as

“any other work that is not time work, salaried hours work or output work including, in particular, work in respect of which there are no specified hours and the worker is required to work when needed or when work is available”.

Regulation 44 of the 2015 Regulations is in identical terms as to the primary definition but omits the words from “including” onwards.

The amount of unmeasured work done in a pay reference period is stated by regulation 27 to be “the total of the number of hours spent by him … in carrying out the contractual duties required of him under his contract to do such work”. However, there is provision in regulation 28 for the worker and the employer to make an agreement

“determining the average daily number of hours the worker is likely to spend in carrying out the duties required of him under his contract to do unmeasured work on days when he is available to carry out those duties for the full amount of time contemplated by the contract”.

Provided the average so agreed is “realistic”, it will apply for NMW purposes.

The provisions relating to unmeasured work contain no provision equivalent to the availability provisions as regards time work and salaried hours work, or therefore for workers sleeping in at or near their place of work.

I believe it will be helpful to begin by analysing the effect of the Regulations as regards sleep-in cases without reference to the case-law. For the reasons already given, I will need to refer to both the 1999 and the 2015 Regulations: I will use the formula “regulation X/Y” to refer to the (substantively) identical regulation in both versions. I start with the provisions relating to time work.

The starting-point is that it is clear that the draftsman of the original 1999 Regulations regarded the situations covered by regulation 15 (1) as time when the worker was not “working” within the meaning of regulation 3. That is quite explicit from the introductory words “in addition to time when a worker is working”. In the re-drafting effected by the Amendment Regulations those introductory words disappear; and they are likewise absent in the 2015 Regulations. But I see nothing to indicate that that is intended to remove the distinction which is explicit in the original version: on the contrary, it seems to me to remain inherent in the structure of the Regulations.

There are thus two separate kinds of “time work” – namely “work” as referred to in regulation 3/30 (to which I will refer as “actual work”), and, by virtue of regulation 15/32, availability for work (in the defined circumstances). How the dividing-line between actual work and availability for work applies in sleep-in cases is the essential question in these appeals.

Regulation 15 (1) and its successors have sometimes been described as deeming provisions. That may not be strictly accurate, since, as already noted, the formal effect of the provisions is to treat availability for work not as “work” but as a distinct category of “time work”; and it may be for that reason that the draftsman does not use the classic deeming terminology of “treated as” but instead says that time work “includes” time when the worker is required to be available. (Footnote: ³) However, in a loose sense the description is reasonable enough.

I turn to the substance of regulation 15 (1) and its successors. The provision which they make for availability cases comprises four elements – or, more accurately, two elements, each with a qualification.

First, the worker must be required to be at or near a place of work (“at work” for short). This element is constant through all three versions, and no issue arises about it for our purposes.

Second, the “at work” element is qualified, in the original paragraph (1), by the words “other than his home”: I call this “the at home exception”. The Amendment Regulations substitute more elaborate wording but I can see no difference of substance. That wording is in turn replaced by the more succinct language of paragraph 32 (1) of the 2015 Regulations; and in this case we know for sure that no change in meaning is intended (see para. 17 above). The purpose of the at home exception is evidently to cover the case where the worker’s home is at or near his or her place of work: there are of course plenty of residential jobs where that will be so. Its effect is that, where a worker is required to be available for work but is at home, the hours in question (awake or asleep) will not count for the purposes of the NMW. The thinking is understandable: the effect of a requirement to be available for work might reasonably be judged to be qualitatively different if the worker can be in his or her own home.

Third, the worker must be, and be required to be, available for the purpose of working (“available for work” for short). As we have seen from the report of the Commission, and will see in the authorities, that situation is sometimes referred to as being “on-call”. That phrase may be used in ordinary industrial parlance in a variety of circumstances, not all of which are necessarily covered by regulation 15 (1) or its successors; and although I will myself occasionally use it for convenience it is important where accurate analysis is required to stick to the language of the Regulations.

Fourth, the availability element is qualified, in the original paragraph (1), by the exception for workers who by arrangement sleep at work. In their case the only hours that count for NMW purposes are those when they are “awake for the purpose for working”: I call this “the sleep-in exception”. This provision is modified in the Amendment Regulations so that employers can only take advantage of the exception if they provide suitable sleeping facilities. The phrase“is permitted to sleep” is also changed to “by arrangement sleeps”. The amended language seems to me clearly to connote a situation where the worker is positively expected to sleep and thus to perform no substantive activities: the only obligation is to be available to work if called on. (Footnote: ⁴) The previous phrase – “is permitted to sleep” – is more ambiguous, because it could if taken literally apply to any kind of work, at any time of day or night, when there is no objection to an employee taking a nap between tasks. I doubt if it was ever the intention that it should have so broad a meaning, but the amended wording puts the position beyond doubt. This element is much more radically re-cast in regulation 32 (2) of the 2015 Regulations, but the essential language remains the same and, again, we know that no substantive change is intended. (Footnote: ⁵)

In summary, the effect of all three versions of the Regulations is that a worker who is, and is required to be, (a) available for the purpose of working (b) at or near his or her place of work is entitled to have the time in question counted as time work for NMW purposes unless

The position as regards salaried hours work is substantially identical, subject to one point. The default position under regulation 21 of the 1999 Regulations (regulation 22 of the 2015 Regulations) is that a worker is treated in any pay reference period as having done his or her “basic hours”, to the calculation of which the availability provisions are irrelevant. Those provisions only become relevant where the worker claims to have exceeded their basic hours: see regulation 22 (3) (d) (regulation 26 (1) (d)). Mr Glyn made a particular submission on that basis: see para. 101 below.

I turn to consider how those provisions apply in a sleep-in case. If we are concerned only with the terms of the sleep-in exception it is of course obvious that they are caught by it: they are sleeping, by arrangement, at their place of work. Logically, of course, that is not the first step: since the sleep-in exception only applies in availability cases and not in cases where the worker is actually working, it is strictly necessary to ask first into which of those two categories a worker who is sleeping in falls. However, that is for practical purposes an unnecessarily elaborate approach. The self-evident intention of the relevant provisions is to deal comprehensively with the position of sleep-in workers. The fact that their case is dealt with as part of the availability provisions necessarily means that the draftsman regarded them as being available for work rather than actually working. That is hardly surprising: it would not be a natural use of language, in a context which distinguishes between (actually) working and being available for work, to describe someone as “working” when they are positively expected to be asleep throughout all or most of the relevant period.

Although I would if necessary reach that conclusion on the basis of the Regulations alone, it is in my view strongly reinforced by the fact that it gives effect to the recommendations of the Low Pay Commission. It is convenient to address here how Simler P dealt with this aspect in her judgment in Focus. At para. 20 of her judgment (p. 1194 F-H) she said that she found the recommendations of no assistance partly because she did not regard the Regulations as ambiguous or obscure (and thus as falling within the scope of Pepper v Hart [1993] AC 593), and partly because they did not “solve the issues raised on this appeal”. I respectfully disagree with her on both points.

As regards Simler P’s first reason, before us Mr Reade disavowed any reliance on Pepper v Hart and relied instead on the principle enunciated by Lord Diplock in Fothergill v Monarch Airlines Ltd [1981] AC 251, at p. 281 B-D that:

“Where the Act has been preceded by a report of some official commission or committee that has been laid before Parliament and the legislation is introduced in consequence of that report, the report itself may be looked at by the court for the limited purpose of identifying the “mischief” that the Act was intended to remedy, and for such assistance as is derivable from this knowledge in giving the right purposive construction to the Act.”

I agree that that principle is apposite; but the present case is a peculiarly strong example of it because of the effect of section 5 (4) of the 1998 Act. In circumstances where the Secretary of State was bound either to follow the Commission’s recommendations in its first report or explain to Parliament why he had not done so (which he did not do), it would, frankly, be extraordinary if the Court were not permitted to have regard to those recommendations in order to help understand their effect.

As regards her second reason, I agree that the terms of the report do not by themselves solve the issues in this appeal. But an explicit recommendation that workers who are “required to be on-call and sleep on their employer’s premises (e.g. in residential homes …)” should not have the hours in question counted for NMW purposes plainly covers cases of the kind with which we are concerned here and we should approach the issues with a view to achieving that outcome so far as possible.

It follows that, on a straightforward reading of the Regulations (in the light of the first report of the Commission), workers sleeping in under arrangements of the kind identified above will only be entitled to have their sleep-in hours counted for NMW purposes where they are, and are required to be, awake for the purpose of performing some specific activity. However, the authorities do not straightforwardly support that analysis. In particular, there is authority at EAT level, which Simler P followed in Mencap, that sleepers-in in a care home are actually working throughout their shifts, within the meaning of regulation 3/30, rather than merely available, so that regulation 15/32 does not bite. I must accordingly turn to consider the case-law.

There are a large number of authorities that are either directly or tangentially relevant to the issue before us. I shall have to consider some in some detail, but others I can take much more shortly.

The problems that have arisen in the case-law start with the decision of this Court in British Nursing Association v Inland Revenue [2002] EWCA Civ 494, [2003] ICR 19. As will appear, I have no problem with either its reasoning or its outcome, but I do not believe that it is decisive in cases of the kind with which we are concerned in this appeal.

The claim was brought by the Inland Revenue (which had enforcement functions under the 1998 Act) against the British Nursing Association (“the employers”) in relation to the NMW rights of the staff of its emergency bank nurse booking service. The service was available on a 24-hour basis. The night shifts were worked by staff from their homes. The agreed statement of facts recited:

“When performing their shifts, the BNA staff are instructed that they must answer telephone calls within four rings or apologise for keeping the callers waiting. They must be available throughout the duration of the shift to answer telephone calls for the appellant (there is no provision for ‘protected’ periods of sleep during which no calls will be received). Subject to the foregoing, BNA staff are able to spend part of the shifts asleep or doing other activities, e g watching television. In practice, the busiest times are from 20.00 to 23.00/23.30 and from 05.30/06.00 to 08.00/09.00. Calls are infrequently received in the period 23.30–05.30.”

The issue in the proceedings was whether employees working the night shift were entitled to have the entirety of the shift counted for NMW purposes or only the time spent actually dealing with calls.

The case proceeded in the ET and the EAT on the basis that the situation was covered by regulation 15 (which at the material time was still in its original form). The ET held that the employees were engaged in time work in accordance with regulation 15 “at times when they are awake and awaiting calls at home” but not when they were asleep, and that decision was upheld by the EAT. This Court dismissed the employers’ appeal, but it made it plain that it considered that the case had proceeded on the wrong basis and that regulation 15 did not apply at all. I will not quote extensively from the ex tempore judgment of Buxton LJ (with whom Peter Gibson LJ and Neuberger J agreed) because its structure reflects the unsatisfactory way in which the issues arose. The essential points for our purposes can be summarised as follows.

The starting-point is that the ET had made a finding in the course of its reasoning, though it had not formed part of its formal decision, that the employees were “working” throughout the shift, whether or not they were actually dealing with calls and (it seems) even if they had gone to sleep: see paras. 26-27 of its reasons, quoted by Buxton LJ at para. 9 of his judgment (p. 21 G-H). Buxton LJ endorsed that finding. At para. 12 (p. 22 D-F) he said:

“I have to say that not only was it open to the employment tribunal and to the appeal tribunal to find that the workers were working throughout their shift, but also, as an issue of the ordinary use of the English language, it seems to me self-evident on these facts that they were indeed so working. No one would say that an employee sitting at the employer's premises during the day waiting for phone calls was only working, in the sense of only being entitled to be remunerated, during the periods when he or she was actually on the phone. Exactly the same consideration seems to me to apply if the employer chooses to operate the very same service during the night time, not by bringing the employees into his office (which would no doubt impose substantial overhead costs on the employer and lead to significant difficulties of recruitment), but by diverting calls from the central switchboard to employees sitting waiting at home.”

Buxton LJ held that it followed from the finding that the employees were working throughout the shift that regulation 15 had no application. He said at para. 14 (p. 23C):

“[R]egulation 15 only arises in a case where a worker is not in fact working, but is on call waiting to work.”

Again, at para. 15 he said (p. 23 E-F):

“Once the tribunal had found that the employees were in fact working throughout their shift, regulation 15 only applied thereafter to situations of a particular sort in addition to what can be properly characterised as work; and on the tribunal's findings the situation that they were addressing was not a deemed piece of time work but an actual piece of time work.”

The phrase “in addition to” is of course a reference to the opening words of regulation 15 (1), as it then stood.

Notwithstanding that conclusion, because of the way the case had been argued Buxton LJ was obliged to say something about regulation 15 (1), while protesting at the artificiality of having to do so and saying that he was deciding nothing “as to the proper application of regulation 15 (1) in a case such as the present”: see para. 16 (p. 23A). At para. 17 he summarises its effect as follows (p. 24 A-B):

“Regulation 15(1) relates to workers who are, in colloquial terms, ‘on call’. When a worker falls into that category, he has to be paid the minimum wage for his waiting hours, unless he is on call at home. I respectfully agree with the very clear analysis of the Employment Appeal Tribunal to that effect in paragraphs 28 to 30 of its judgment. However, if the worker is permitted to sleep when on call, the hours during which he is permitted to sleep and when he is not actually working do not count as the equivalent of time work.”

In the remainder of the paragraph, and at the beginning of para. 18, he returns to the point that regulation 15 should not have been treated as having any application in the present case. Although, since that had happened, the ET would have to undertake the task of distinguishing between periods of “actual work” and periods when the employees were “permitted to sleep”, he predicted real difficulties in making that assessment. He concludes at para. 19 (p. 24 F-G):

“Having said all that, the alternative that is apparently contended for by the BNA, that the employees are only working when they are actually dealing with phone calls, with all the periods spent waiting for calls excluded, would, in my view, effectively make a mockery of the whole system of the minimum wage.”

The British Nursing Association decision clearly establishes that regulation 15 (1) – now, regulation 32 – has no application in cases where workers are “actually” working: it covers only cases where they are “on call” (more accurately, where they are “available for the purposes of working”). That is entirely consistent with my analysis above.

The more important question, however, is what the case decides about what counts as “actual” working. In the passage from para. 12 of his judgment quoted at para. 52 above Buxton LJ says that it cannot sensibly be said that the booking staff working the ordinary day shifts are not actually working, i.e. within the meaning of the regulation 3, for the entirety of the shift simply because there may be slack periods during which there are no calls to answer. That must be right. Although for some purposes a worker might be said to be only “working” during the time that he or she is performing a specific task, that cannot be the sense in which the word is used in regulation 3. There must be many kinds of work where specific tasks only come up intermittently but where for the purpose of the Regulations a worker remains “actually” working even during periods where they have nothing to do.

Buxton LJ then extends that approach to workers who work the night shift at home. His essential reasoning is that while the lulls in work, or periods of inactivity, may have been longer at night there was no qualitative difference from the position during the day shift. I have no problem with that conclusion either, as long as it is appreciated that it is a decision about the facts of a particular case. The decision certainly establishes that the fact that a worker is entitled to go to sleep in the intervals between particular tasks is not necessarily inconsistent with them actually working, for the purposes of regulation 3, during the entirety of the period; but it does not establish that that will be so in a case of the kind with which we are concerned in these appeals, where the essence of the arrangement is that the worker is expected to sleep. Buxton LJ was not considering such a case.

I should note, finally, that Buxton LJ mentions at para. 20 of his judgment that counsel for the employer had referred the Court to the Working Time Regulations 1998 (“the WTR”) and case-law on the underlying Directive (in particular the decision of the ECJ in Sindicato de Medicos de Asistencia Publica (SIMAP) v Conselleria de Sanidad y Consumo de la Generalidad Valenciana [2001] ICR 1116). He deprecated such reference, partly on the basis of “the different objectives of the different pieces of legislation” (p. 24H). It is convenient to say at this point that I agree that reference to the WTR and the case-law under it is likely to be unhelpful. Although similar issues about what constitutes “work” may arise under both the WTR and the NMW Regulations, the legislation is different and derives from different sources: the WTR are made pursuant to an EU Directive while the NMW Regulations are wholly domestic.

In Scottbridge Construction Ltd v Wright [2003] IRLR 21 the respondent employee was employed as a nightwatchman at the premises of a construction company in Glasgow. The facts are sufficiently summarised at paras. 2-3 of the judgment of Lord Cullen (Lord President) in the Inner House of the Court of Session, as follows (pp. 21-22):

“2.

... [The employee] was required to attend [the] premises at 5 pm each evening and to remain there until 7 am the following morning. He was so employed by them between June and November 1999. In general his duties included maintaining security of the premises, admitting company employees to the premises when required, closing up afterwards, manning the telephones, and some small-scale cleaning tasks, including emptying the bins. These tasks were not onerous.

3.

In the evening between 5 and 6 pm the respondent could have been fully engaged while the day shift were completing their tasks, and it is possible that he might have been required to answer the telephone after that time. This was only an occasional requirement. He was not required to carry out any tasks on a regular basis from that time till around 11 pm when the night shift arrived to collect their vehicles and load them before proceeding to site. He was not involved in any tasks after midnight until 5 am, when the nightshift returned and unloaded. Very rarely was he required to open up the premises during this five-hour period. As nightwatchman he was permitted to sleep when he was not required to carry out work. Within the office premises there was a television set, cooking equipment and a mattress for sleeping, which was used by him for this purpose.”

The employee brought proceedings claiming that the entirety of the hours for which he was obliged to be on the premises should count for NMW purposes. His claim was dismissed by the ET, which proceeded on the basis that his case fell within the terms of regulation 15 (1) of the original Regulations, and held that he was only required to be “awake for the purpose of working” for four hours a night. The EAT allowed his appeal.

The Inner House upheld the decision of the EAT. The Lord President’s reasoning is succinct. He held that the facts were indistinguishable from those of the British Nursing Association case and that the employee should be treated as having been working, within the meaning of regulation 3, for the entirety of the period that he was required to be on the premises, with the result that regulation 15 was not engaged. Lord Cullen observed, at para. 11 (p. 22):

“The work which was paid for under his contract by reference to the time for which he worked was, for the purposes of reg. 3, his attendance as a nightwatchman for the whole of those hours. ... [T]he fact that the respondent had little or nothing to do during certain hours when he was permitted to sleep does not take away from the fact that he was throughout in attendance as a nightwatchman and required at any time to answer the telephone or to deal with alarms.”

The judgment in Scottbridge does not advance the argument on the issues of principle, since the Inner House simply adopted the analysis of this Court in the British Nursing Association case. However its decision that, on the application of those principles to the facts before it, the employee should be held to have been actually working for the entirety of the shift is arguably more problematic, and I will have to return to it later.

The specific issues in Walton v Independent Living Organisation Ltd [2003] EWCA Civ 199, [2003] ICR 688, were different from those with which we are concerned, but aspects of the Court’s analysis are significant. The main facts relevant for our purposes are stated in paras. 2-3 of the judgment of Aldous LJ (pp. 689-690) as follows:

“2.

The employer, the Independent Living Organisation Ltd, provides carers for people who because of their age or disability need assistance to maintain their independence and remain in their own homes rather than being placed in a residential home. Miss Walton was a carer employed by the employer. She was allocated to a Miss Jones who suffered from epilepsy and had fits on a regular basis. She was a relatively easy client needing a minimum of supervision. She could attend to her own needs, i.e. she could feed herself, tidy her room, attend to her personal hygiene. The employee was responsible for her washing, ironing, shopping, preparation of meals and medication. Miss Jones spent most of her time watching television and doing jigsaws. She usually retired to bed between 9.30 and 10 p m and slept through the night. Thus Miss Walton was very rarely disturbed during the night.

3.

Miss Walton worked three days on, four days off. … When not providing a service to Miss Jones she could please herself as to what she would do, but she was required to be on the premises in case Miss Jones required assistance which occasionally she did, sometimes during sleeping hours. …”

It is also important to record that the employee and the employer had together assessed the time she required for each of the tasks that she carried out for the client and agreed that they amounted, on average, to 6 hours and 50 minutes each day (see para. 5 of Aldous LJ’s judgment (p. 690 E-F)).

It was the employee’s case that she was doing time work throughout the entirety of the 72-hour period that she was required to be present at the client’s home and that her hours worked should be calculated for NMW purposes on that basis. Her counsel, Mr Robin Allen QC, quoted Milton – “they also serve who only stand and wait” (see para. 36 of Arden LJ’s judgment (p. 697C)). More prosaically, he submitted, as Aldous LJ records at para. 19 (p. 694B):

“… that both the employment tribunal and the Employment Appeal Tribunal had gone wrong in that they had confused ‘work’ in the sense of physically doing something with what was ‘work’ for the purposes of the legislation. Whilst Miss Walton was not at times physically working she was under contract to be mentally alert to the needs of Miss Jones. She was as much carrying out work during the time she was at Miss Jones's home as a night watchman would be when relaxing or a nurse when having a cup of tea off-ward.”

As Aldous LJ noted, that submission was partly based on the decisions in the British Nursing Association case and Scottbridge.

It was the employer’s case that the employee was doing unmeasured work and that the agreement that she worked 6 hours and 50 minutes per day constituted a realistic daily average agreement within the meaning of regulation 28 of the 1999 Regulations.

This Court found for the employer. Both Aldous LJ and Arden LJ delivered substantive judgments, with which Jacob J agreed. I take them in turn.

The first issue considered by Aldous LJ was whether the employee was doing time work. For that purpose he was prepared to assume that she was indeed “working” for the entirety of the 72 hours, within the meaning of regulation 3; but he held that even on that basis her work was not “paid for by reference to” the entirety of those hours and that the case accordingly fell outside the terms of regulation 3 (a): see paras. 25-32 of his judgment (pp. 695-7). However, there was a further issue as to whether the agreed average relied on by the employer under regulation 28 was “realistic”. Mr Allen argued that it was not realistic, because “the duties required of [the employee] under her contract to do unmeasured work”

“meant the same as ‘work legally required’ and in context should be interpreted in the same way as the British Nursing Associationcase”,

and that she was accordingly performing her duties for the entirety of the 72 hours. Aldous LJ rejected that argument, saying simply, at para. 34 (p. 697F):

“TheBritish Nursing Association and Scottbridge cases are not analogous. Miss Walton stayed at Miss Jones's home for three days. I do not believe that she ‘worked’ a continuous period of 72 hours.”

Thus at that stage in the argument he rejected the contention which he had at the previous stage been prepared to assume.

Arden LJ likewise rejected the argument that the employee was doing time work. As regards the regulation 28 point, she said (pp. 698-9):

“40.

There is, as I see it, a clear distinction in the 1999 Regulations between ‘working’ and being ‘available for work’: see, for example, regulations 3 and 15. However, that antithesis is not drawn in the context of unmeasured work.

41.

The tribunal's finding was that, when not performing her specified tasks, Miss Walton was not required to give Miss Jones her full attention …. In view of this finding, in my judgment it cannot be said that Miss Walton was continuously performing her contractual duties for 24 hours each day for the purpose of regulation 28. Not every worker who ‘only stands and waits’ carries out contractual duties: it is a question of fact.

42.

Had it been intended that regulation 28 should apply to hours for which a worker doing unmeasured work is required to be available for work, but not actually carrying out her contractual duties, it would have been necessary to have some equivalent of regulation 15 and, if recommendation 4.34 were being followed in the case of unmeasured work, to exclude time spent asleep.”

The reference at para. 42 to “recommendation 4.34” is of course to para. 4.34 of the first report of the Low Pay Commission, to which Arden LJ had referred earlier in her judgment, and which Aldous LJ had set out in full. The first sentence of para. 40 is consistent with my own analysis and with Buxton LJ’s reasoning in the British Nursing Association case.

Formally, the decision of the Court on the regulation 28 point is concerned only with whether the employee was “carrying out the duties required … under [her] contract” for the entirety of the 72 hours. But the case that she was doing so was advanced, as I understand it (though the summary of Mr Allen’s submissions is rather condensed), on the basis that that was the same question as whether she was “working” within the meaning of regulation 3; and that was certainly the language that Aldous LJ used in rejecting it. On that basis, Walton is authority for the proposition that a carer who is required to be on call overnight, albeit permitted to sleep, is not for that reason to be treated as working. Aldous LJ evidently regarded Mr Allen’s case to the contrary as so self-evidently wrong as not to require detailed rebuttal. It is true that the appeal was not concerned with sleeping time as such, but the artificiality of treating such time as working time might be thought to have weighed in particular in Aldous LJ’s thinking; and it clearly weighed with Arden LJ – see her point (at para. 42) that if Mr Allen’s submission were correct special provision for “time spent asleep” would have been necessary.

In MacCartney v Oversley House Management [2006] UKEAT 0500/05, [2006] ICR 510, the claimant was the manager of a home for the elderly, living in a flat on the premises. In addition to her duties during the day, she was on call overnight four days a week. This aspect of her duties was described in the judgment of HH Judge Richardson in the EAT as follows (p. 513 D-F):

“13.

… She was required to be on site or within a three minute radius – just far enough for her to take her dog for a walk. She could not socialise in the town, or visit her daughter or family. On the other hand, she could receive visitors, listen to music, eat, undertake other activities at home and, of course, sleep in her own bed.

14.

While she was on call, she had to answer emergency and non-emergency calls. She kept a record of calls answered between 6pm and 8am for the eight months from August 2003 to March 2004. It averaged out at 3-4 emergency calls each month, and 10-11 non-emergency calls each month. So, on average, the manager on duty would be called about every other day at some time between 6pm and 8am. These calls could no doubt vary from the relatively trivial to the full scale emergency. Apart from those calls, she was (as the Tribunal put it) not at the beck and call of the residents between 6pm and 8am.”

The principal issue on the appeal to the EAT was whether the periods during which the claimant was on call constituted “working time” for the purpose of the WTR. It held that it was. It relied principally on the decision of the decision of the ECJ in Landeshauptstadt Kiel v Jaeger [2003] IRLR 804, which it said established that “workers who are on call at a place where they are required by their employer to remain may be said to be ‘working’ when sleeping or resting”. However, there was also a claim under the NMW Regulations. The EAT dealt with this very briefly. On the question whether the time on call required to be counted for NMW purposes it said only, at para. 65 (p. 523 D-E):

“For the reasons we have given, Mrs MacCartney was at work throughout the period when she was providing on site cover, so Regulation 15(1) does not arise: see British Nursing Association v Inland Revenue at paragraph 18.”

The case is worth mention because it is the first in which the EAT considered the application of the NMW Regulations to hours where a worker was on call overnight in a residential home. But it carries limited weight because the issue appears to have been marginal to the main argument and the reasoning is tainted by the reference to authorities on the issue of working time: see para. 58 above. It also appears to have treated the British Nursing Association as authoritative as regards a factual situation which was plainly different.

Burrow Down Support Services Ltd v Rossiter [2008] UKEAT 592/07, [2008] ICR 1172, represents an important step in the development of the case-law. The facts are summarised in the judgment of Elias P in the EAT, as follows (p. 1173 B-E):

“1.

… The claimant ... commenced employment with the appellant company in November 2001. The hours when he had to attend work were from 10 pm to 8 am, two nights a week. His job was to ensure the security of the work premises, a care home for people with learning difficulties. He had to monitor health and safety and to be accessible should any emergencies occur. He was required to be awake for a quarter of an hour to effect a handover of duties, and to assist with the breakfasts of the residents between 7 am and 8 am. Otherwise he could be asleep, save when his duties required him to be awake, such as if he heard noises requiring investigation or if anything else untoward occurred. Indeed, his job title was 'night sleeper'.

2.

The tribunal found that the contract he entered into envisaged a wage which was described as '£20 per night (sleeping only)'. This was later raised to £24. This was the basic amount paid for being present during the night; a separate payment was made for the time during which the claimant was awake and working. ...”

The employee contended that the entirety of his shift counted as “time work” for the purpose of the Regulations (which were in their amended form). The ET upheld that claim, relying on the British Nursing Association and Scottbridge decisions. The EAT dismissed the employers’ appeal. Elias P’s reasoning was as follows.

At paras. 11-19 he reviewed the decisions in British Nursing Association and Scottbridge. At paras. 14-15, having analysed Buxton LJ’s judgment in the former, he says (p. 1175 C-E):

“14 As this analysis makes plain, the original regulation 15 is a deeming provision. It is treating as time work periods when an employee is in fact not working but only available for work.

15 The exception, which was relied upon by the employer, ensures that certain cases when the employee is available for work will not count as time work because it is taken out of the deeming provision. However, once it is determined that for the whole period of the shift the worker is actually working, he falls firmly under the scope of regulation 3 as a time worker. His status is not that of someone who is available for work but rather someone actually working. It follows that there is no scope for regulation 15 to operate. If that regulation is inapplicable then so is the exception. The claimant is relying on work actually done, not work deemed to be done by virtue of regulation 15.”

At paras. 21-23 he rejects an argument that the effect had been changed by the language of the new paragraphs (1) and (1A). Again, that accords with my own analysis.

Having established the legal framework, Elias P went on to endorse the ET’s conclusion on the facts. He says (pp. 1176-7):

“24.

It follows that, in our judgment, the analysis in the two earlier cases was plainly correct and is equally applicable here. The claimant was at work for the whole of the shift, essentially for the reasons given in the Scottbridge Construction case. Like the claimant in that case, even during the time when he was permitted to be asleep, he was still required to deal with anything untoward that might arise in the course of his shift. It was not a case where he was deemed to be at work although only available to work. So neither regulation 15 (1) nor regulation 15 (1A) were ever engaged. In our judgment, the tribunal was right to say that Scottbridge was indistinguishable and should be followed.

25.

We recognise that there is some artificiality in saying that someone is working when he is sleeping, but the justification for this, and the steps which the employer might take to ensure that he is getting value for the wage paid, were summarised as follows by Lord Johnston when hearing the Scottbridge Construction case in the appeal tribunal [2001] IRLR 589, para 9:

‘It is wholly inappropriate for the employer while requiring an employee to be present for a specific number of hours, to pay him only for a small proportion of those hours in respect of the amount of time that reflects what he is physically doing on the premises. The solution for an employer who wishes an employee to be present as a night watchman or the equivalent, is to provide him with alternative and additional work on the premises which enables him both to provide the employer with remunerated time and also the protection of someone on the premises for security reasons.’”

In my view Burrow Down was wrongly decided. The essential reasoning – that the worker was throughout his shift actually working within the meaning of regulation 3, so that regulation 15 had no application – is contrary to what I believe to be the clear meaning of the Regulations, reinforced by the terms of the Low Pay Commission recommendations: see paras. 43-44 above. It is based on the EAT’s understanding of the British Nursing Association decision and, more particularly, Scottbridge. I take those decisions in turn.

For the reasons given at para. 57 above, I do not think that the British Nursing Association decision requires the conclusion to which the EAT came in Burrow Down. Entirely reasonably, Buxton LJ believed that on the facts of that case the staff were indeed actually working throughout their shifts, notwithstanding that there were likely to be lulls in activity during which they could sleep. It does not follow that he would have taken the same view about a contract under which the worker was expected to sleep throughout the period. Indeed I am strongly inclined to think that his reaction to such a case would have been the same as Arden LJ’s in Walton.

The real problem is Scottbridge. Mr Jones submitted that, although the decision did not concern the care sector, the core facts were indistinguishable from those of the kinds of case with which we are concerned: although the employee had some duties at either end of his shift, for the central part of it he was indeed positively expected to sleep. He submitted that the EAT was right to regard itself as bound by the decision of the Inner House, and that we were similarly bound. I see the force of that submission, but I have come to the conclusion that Scottbridge cannot be regarded as determining the outcome in these cases. Starting, as I do, with the proposition that the Regulations, even read on their own but a fortiori when taken with the recommendations of the Commission, require the case of a sleeper-in in a residential home to be treated as a case of availability for work and not as one of actual work, it is necessary to consider narrowly whether there is in truth no distinction between their case and that of the nightwatchman in Scottbridge. I do not believe that they are necessarily identical. The essence of a “sleep-in” contract is that the worker, in the words of the Regulations, “by arrangement sleeps [at the workplace]” (and is given suitable facilities for sleeping – which would normally at least mean a proper bed in an area set aside for sleeping). I do not think that that would be a natural characterisation of what a nightwatchman does, even one who appears to have had so few duties as the employee in Scottbridge and who is given a mattress to sleep on in the office. (Footnote: ⁶) It is also material that the employee in that case had significant duties at either end of the shift, beyond mere hand-over: the period during which he could normally count on being able to sleep was only five hours. I quite accept that the distinctions are subtle, but they are in my view sufficient to justify a difference in outcome: it must be borne in mind that the decision which side of the line dividing “actual work” from “availability for work” a given case falls is factual in character, and in marginal cases different tribunals might well assess very similar facts differently.

I feel less discomfort in reaching this conclusion because the Inner House was not referred to the first report of the Low Pay Commission. If it had been, the question whether the case before it could be distinguished from that of sleepers-in in a care home would have had to be confronted. For the reasons already given I do not believe that it would have been open to the Inner House to find that sleepers-in were actually working; and it would then have had to decide, with a good deal fuller reasoning, whether the position of the employee in the case before it was distinguishable. Maybe it would have decided that it was, for the kinds of reason given in the previous paragraph, or maybe it would not; but I am only concerned to make the point that I am not differing from an explicit decision on the part of the House that no such distinction is possible. (Footnote: ⁷)

I should add that there was some discussion before us about whether this Court was in any event bound by Scottbridge. Normally a decision of the Inner House would only be of persuasive authority in this Court. But I cannot think that it would be right for us to depart from such a decision where it unequivocally decides the effect of a statute which applies throughout Great Britain.

I am also less reluctant to differ from the decision in Burrow Down given that the EAT likewise was not referred to the Commission’s report; nor was it referred to Walton. Indeed the employee appeared in person and the respondent was represented by a lay representative (albeit one of considerable experience). It is worth noting that Elias P was clearly concerned that there was “some artificiality” in the result (see para. 25 of his judgment). He went on to say that that was “justified” by the point (which he adopted from the judgment of Lord Johnston in the EAT in Scottbridge) that an employer can always ensure that it gets full value from a nightshift worker by finding specific tasks for them to do throughout the shift. With respect, I find that unconvincing. Even if it were always practicable to find worthwhile tasks to occupy a sleeper-in for eight hours (which I doubt), that does not address the basic artificiality of describing someone as “working” – still more, as actually working – during a shift when it is positively expected that they will spend substantially the whole time asleep.

Following Burrow Down there have been a number of decisions of the EAT concerning the NMW rights of workers required to sleep in, mostly but not wholly in the care sector. Some have followed Burrow Down, while others have sought to distinguish it. Since I believe Burrow Down was wrongly decided I see no advantage in reviewing them. For the record, however, the cases to which we were referred which followed Burrow Down, on the basis that the worker was actually working, are: Smith v Oxfordshire Learning Disability NHS Trust [2009] UKEAT 0176/09, [2009] ICR 1395 (Footnote: ⁸); Whittlestone v BJP Support Ltd [2013] UKEAT 128/13, [2014] ICR 275; Esparon (t/a Middle West Residential Care Home) v Slavikovska [2014] UKEAT 217/12, [2014] ICR 1037; Governing Body of Binfield Church of England Primary School v Roll [2016] UKEAT 0129/15, [2016] IRLR 670 (Footnote: ⁹); Focus itself; and Abbeyfield Wessex Society Ltd v Edwards [2017] UKEAT 0256/16. (Footnote: ¹⁰) The cases in which Burrow Down was distinguished, on the basis that the worker was only available for work and not actually working, are South Manchester Abbeyfield Society Ltd v Hopkins [2010] UKEAT 79/10, [2011] ICR 254; Wray v J.W. Lees & Co (Brewers) Ltd [2011] UKEAT 102/11, [2012] ICR 43; City of Edinburgh Council v Lauder [2012] UKEAT 0048/11; and Shannon itself. (Footnote: ¹¹)

As the tribunals in some of these authorities remarked, the basis on which cases have been held to fall on one side of the line or the other are hard to understand. In her judgment in Focus Simler P made a valiant attempt to reconcile them. She concluded that the decision whether in any given case a person who sleeps in at work is to be regarded as actually working during the entirety of the period or as only being available for work was one to which no bright line could be applied. The correct approach was multifactorial, and she discussed some of the factors that might be relevant, while acknowledging that there was no “magic key”. I mean no disrespect to her in saying that even in the light of her careful analysis the kinds of distinctions that are required seem to me elusive; and one advantage of a conclusion that Burrow Down was wrongly decided is that this difficult and intractable case-law can be simply put to one side.

I should, because it features a good deal in the authorities, mention one possible basis of distinction, which was floated in the South Manchester Abbeyfield case, namely between workers for whom their sleep-in hours are their only obligations to the employer in question (as in Burrow Down itself) and those where they are an add-on to a separate “core” contract. Simler P in Focus (see para. 28 of her judgment (p. 1197 C-E)) held that that could not be a material distinction; and I respectfully agree.

For the reasons which I have given I believe that sleepers-in, in the sense explained at para. 6 above, are to be characterised for the purpose of the Regulations as available for work, within the meaning of regulation 15 (1)/32, rather than actually working, within the meaning of regulation 3/30, and so fall within the terms of the sleep-in exception in regulation 15 (1A)/32 (2); and we are not bound by authority to come to any different conclusion. The result is that the only time that counts for NMW purposes is time when the worker is required to be awake for the purposes of working.

I have been able to reach that conclusion without placing any weight on the decision of this Court in Walton. Although I incline to think that it is at least implicit in the ratio of that case that time when a worker is expected to be asleep cannot be “work” for the purpose of regulation 3/30, that may be debatable; and in a case where any conflict of authority is so muffled it would be unsatisfactory to decide this appeal on the basis that we can escape from Scottbridge on Young v Bristol Aeroplane grounds. But it is an advantage of the conclusion which I favour that the contrary conclusion would have created a very awkward discrepancy between the treatment of time spent sleeping in for the purpose of time work and salaried hours work and the treatment of the employee’s overnight on-call time in a case involving unmeasured work, as expounded in Walton.

As I hope is clear, my reasoning does not undermine the point of principle first established in the British Nursing Association decision, namely that the at home and sleep-in exceptions only apply in cases where the case falls into the “available for work” rather than “actual work” category. I only say that since Burrow Down the line between those two categories has been wrongly drawn so as to put sleep-in cases on the actual work side of the line. It will still be necessary to draw the line in some other cases, in order to establish whether the at home exception applies. There are of course many cases in which workers who have accommodation at work – such as caretakers or residential managers – are required to be “on call” outside normal working hours but not at times when they are expected to be asleep. (Footnote: ¹²) The focus of the arguments before us was on the sleep-in exception, and I do not think it is appropriate for this Court to offer general guidance on other kinds of case. I would only make two points.

First, it was not fully explored before us whether the at home and sleep-in exceptions are mutually exclusive. I am inclined to think that they are, as Mr Glyn submitted: if the worker is at home, even if the home is at or near their place of work, it is rather awkward to describe them as sleeping there “by arrangement” or as being provided with suitable facilities for sleeping. However, so far as I can see the question ought not to matter in practice. For the reasons already given the worker will be available, rather than actually working, during the hours that they are expected to sleep (notwithstanding that they are on-call), so that either exception or both will be engaged as regards those hours.

Secondly, Mr Glyn pointed out that it was increasingly common for workers to work their ordinary daytime hours entirely or in part from home; and he submitted that we ought not to make any decision which allowed employers inappropriately to invoke the at home exception in such cases. I do not believe that my reasoning carries any such risk. If a worker is actually working, notwithstanding that the work is of a character that only generates specific tasks intermittently (and permits them to sleep in between), it makes no difference whether they are doing so at home or at work; and indeed that is clearly demonstrated by Buxton LJ’s approach to the facts of the British Nursing Association case.

The outcomes of the individual appeals are effectively determined by the decision on the issue of principle discussed above. I should nevertheless briefly set out the facts and explain how my decision applies.

The facts in Mencap were summarised by Simler P at paras. 47-50 of her judgment in Focus (pp. 1202-3) broadly as follows (though I have made some minor changes):

Ms Tomlinson-Blake’s claim was that she was entitled to have the totality of her hours spent sleeping in counted as time work for NMW purposes. The ET and the EAT, following Burrow Down, upheld that claim, on the basis that she was actually working for the whole period so that the sleep-in exception did not apply. It follows from my conclusions above that that was wrong. The claimant slept by arrangement at her place of work and was provided with suitable facilities for doing so. (I also note, though this is not legally significant, that she was paid an allowance of the kind referred to in the first report of the Low Pay Commission.) It follows that she is to be treated as being available for work during those hours and not actually working and that the sleep-in exception applies. The result is that only those hours during which she was required to be awake for the purpose of working count for NMW purposes.

I note that the ET emphasised that the claimant was obliged “to keep a ‘listening ear’” even while asleep and emphasised the need for her to exercise a professional judgment as to whether intervention was required. But I do not see how either point affects the analysis. Any sleeping-in worker has to have a “listening ear”, in the sense that they have to be prepared to be roused by the occurrence of something untoward: that is what they are there for. The phrase is in any event metaphorical: the fact is that she was expected to, and almost always did, get an uninterrupted night’s sleep. And it cannot make any difference what kind of decisions had to be made if and when she was roused.

I should for completeness mention a submission advanced by Mr Brennan to the effect that the correct analysis of Ms Tomlinson-Blake’s sleep-in shifts was that they constituted unmeasured work, with the one hour’s pay referred to at para. 92 (2) above constituting a “daily average agreement”. Clearly, as Walton shows, there are some cases involving care workers that can be analysed in this way, but I do not think it would be right to approach the present case afresh at the instance of an intervener and in circumstances where neither party wished to adopt Mr Brennan’s submission.

I can likewise broadly adopt Judge Peter Clark’s summary of the facts in Shannon (at paras. 4-9 of his judgment). Clifton House is a registered residential care home in Surrey. It provides for up to 16 elderly residents. There is a top-floor staff flat known as “the Studio”. Before the Respondents took over the home in May 2013 it was owned by a Mr Sparshott. In May 1993 Mr Sparshott offered the claimant, Mr Shannon, who was a family friend, employment as “on-call night care assistant” with accommodation in the Studio. He was required to be in the Studio from 10 p.m. until 7 a.m. He was able to sleep during those hours but was required to respond to any request for assistance by the night care worker on duty at the home. In return he was provided with free accommodation with all utilities provided free of charge, together with a payment of £50 per week, rising eventually to £90. It seems that originally he was entitled to take some holiday, but from 1996, following an incident when the flat was used without his consent while he was abroad and his equipment damaged, he slept there every night. In practice he was very rarely asked to assist the night care worker. He had day jobs as a driver. In contemplation of the Respondents’ acquisition of the home Mr Sparshott asked the Claimant to sign a contract of employment that included a wider range of duties than he had previously carried out; he was also required to provide cover for Abbey Lodge, a care home opposite Clifton House then owned by the Respondents though in the event he never did so.

Mr Shannon’s claim was that he was entitled to have the entirety of the hours between 10 p.m. and 7 p.m. counted as salaried hours work for NMW purposes for 365 days a year. The arrears that he claimed on that basis were calculated to amount to almost £240,000.

The case was argued in the ET and the EAT on the basis that the crucial question was whether during the periods in question the Claimant was actually working, or available for work, and the case-law on that question was expressly considered. The ET found that he was only available for work, so that regulation 16 (1) and (1A) applied. It dismissed the claim on the basis that the case fell within the at home exception. In the EAT Judge Peter Clark seems, though the analysis is not entirely clear, to have relied on the sleep-in exception as well or instead. However, as I have noted above, the difference does not really matter. Whichever applies, the crucial question is whether the Claimant was actually working or only available for work. In my view the ET’s conclusion about that was plainly right. Even if we were concerned only with the at home exception (and so could not rely on the point made at para. 43 above (Footnote: ¹³)), that is inescapably correct: it is impossible on any common-sense approach to describe the Claimant as actually working except when he was called on to assist the night care worker.

Mr Glyn submitted that the ET had been wrong to mention the point that the Claimant was not the primary person on call but was only required when the night care assistant at the home required help. I can see nothing in that. The point reinforced, though I am not sure reinforcement was needed, that on any ordinary meaning of language the Claimant was available for work rather than actually working. He also said that it was a regulatory requirement that the Respondents employ at the home an adequate number of staff, at night as well as by day. Even on the assumption that that meant that without their agreement with the Claimant they would have been in breach of their regulatory obligations I cannot see that it assists on the question of whether he was actually working or available for work (Footnote: ¹⁴).

Mr Glyn also emphasised the restriction on the Claimant’s liberty resulting from his being unable to leave the flat, for any purpose, between the agreed hours. That is true as far as it goes, though the hours are only those when he might be expected to be at home (and the factual findings, such as they are, suggest that for idiosyncratic reasons he wanted to stay in every night anyway). But I do not see how it advances the argument: any agreement to be available for work at a particular place necessarily involves a restriction on liberty, but the Regulations choose to treat that differently for NMW purposes where the place in question is the worker’s home.

Those points essentially rehearsed the arguments that had been deployed below. In his oral submissions before us Mr Glyn took a point unheralded in either his grounds of appeal or his skeleton argument, namely that the earlier debate was irrelevant because the hours between 10 p.m. and 7 a.m. were the Claimant’s “basic hours” within the meaning of regulation 21 and had thus to be counted in full: as noted at para. 42 above, the availability provisions only apply in salaried hours cases where the worker claims to have worked more than his basic hours. Mr Sutton protested, with justification, that he had had no notice of this point and was not in a position to deal with it. I do not think it would be right to allow it to be taken for the first time on appeal. My strong provisional view is that it is in any event a bad point because the contract could not be analysed as one in which the Claimant was paid an annual salary in respect of the hours in question; but it is not necessary to express a concluded view.

I would allow the appeal in Mencap and dismiss the claim. I would dismiss the appeal in Shannon.