Martin Laker v The Commissioners for HMRC

It was common ground that Dr Laker was not in pensionable employment in February 2013. Accordingly, he claimed that HMRC cannot apply the definition to one year ending 31 August 2013 as that would not be a year of pensionable employment due to the February 2013 break.

Equally, Dr Laker submitted you cannot construct “the member’s last year” from two non-consecutive periods because a year means 12 consecutive months. Whilst the period of one year can start on any date (and continue for 12 months), a period (or, more accurately, periods) that are disjunctive are not a year. Dr Laker submitted that this is no more than the ordinary meaning of the word “year” which is an ordinary word, and its meaning depends on how it is ordinarily used.

In HMRC v Dolphin Drilling Ltd [2024] EWCA Civ 1 Nugee, Jackson and Newey LJJ said:

Mr Firth submitted that HMRC’s approach of treating non-consecutive periods as amounting to a year contradicts how the word “year” is ordinarily used. It was not consistent with any understood usage of that word. For instance, on HMRC’s approach, Dr Laker’s last year of pensionable employment did not include February at all and included the whole of August twice.

Adopting HMRC’s approach, according to Mr Firth results in any aggregation of a sufficient amount of time as a year. For instance, HMRC would say that every 7 June for the last 365 years, when aggregated, is a year. Anybody with common sense would know that it is not. In other words, once the requirement for a year to be a single, continuous period of time is removed, there is no way to limit the consequences of that step. One cannot, on any principled basis, according to Mr Firth, limit it to a maximum number of non-consecutive periods (if 2 periods are permissible, 3 must also be permissible and so on up to 365 or 366 non-consecutive periods or further if one includes parts of a day).

Similarly, if one can have regard to non-consecutive periods one month apart, why not two months apart, a year apart and so on? The result according to Mr Firth is a concept that is patently not a “year”.

Second, Mr Firth submitted that a period of 12 consecutive months is supported by the meanings of similar/related words such as “week” and “month”. A week is, self-evidently, 7 consecutive days. An aggregation of 7 non-consecutive days is plainly not a week. For instance, 13 January, 10 February, 20 June, and 1 to 4 December is a total of seven days, but they do not, together, amount to a week. Similarly, a “month” refers to the period from a date in one month to the immediately preceding date in the next month (e.g. 15 January to 14 February). An aggregation of 28, 29, 30 or 31 non-consecutive days is not, on any view, a month. The next step up according to Mr Firth, in terms of identifying time periods, would be a year. The same also applies to larger periods of time, such as decades, centuries and millennia. For instance, a decade patently refers (and only refers) to 10 consecutive years.

Third, depending on what measure HMRC decide to use (months or days), the ‘year’ varies. Based on 12 months, Mr Firth claimed that HMRC’s year is 368 days long in the present case (as a result of excluding February (28 days) and including August (31 days) twice, a net increase of 3 days). In other circumstances, HMRC’s “12 month” year could be as short as 336 days (twelve Februarys) or as long as 372 days (for 31-day months).

Even HMRC’s attempt to define a year in terms of a specified number of days (irrespective of whether they are consecutive) cannot justify choosing 365 days rather than 366 (sometimes 366 days is a year) according to Mr Firth.

Mr Firth then claimed the ordinary usage and meaning of “year” was confirmed by the dictionary meaning. The Oxford English Dictionary definition is:

What all these definitions have in common Mr Firth claimed is that a year is a single period of time, namely, 12 consecutive months and this is what Parliament must be regarded as having intended to communicate by picking this ordinary word. In R (oao) v Secretary of State for the Home Department [2022] UKSC 3 Lord Hodge said:

In R (oao Mawbey) [2019] EWCA Civ 1016 Lindblom J said:

Mr Firth claimed that the “last year of pensionable employment” must be a period that includes every month of the calendar and equally cannot include the same month more than once.

Finally Mr Firth pointed out that the definition of “pensionable service” has a provision that specifically deals with (and aggregates) disjunctive periods of pensionable service (and deals with leap years):

There is no such aggregation provision for “pensionable employment”. Pensionable service and pensionable employment are distinct concepts. This provision demonstrates that where Parliament intended to aggregate disjunctive periods, it did so explicitly. The absence of such a provision in relation to pensionable employment demonstrated that there was no intention to aggregate disjunctive periods (and no prescribed mechanism for doing so). Dr Laker’s last year of pensionable employment was, therefore, the year ended 31 January 2013, and, accordingly, his pensionable pay for calculating benefits in 2012/13 and 2013/14 was the same. The result is that the increase did not exceed the threshold and fixed protection was not lost.

The meaning of “final year’s pensionable pay” that is relevant to Dr Laker is “pensionable pay in respect of the member’s last year of pensionable employment, ending on the date the member ceases to be in such employment”.

Ms Belgrano claimed that the pensionable pay in respect of Dr Laker’s last year of pensionable employment, ending on the date he ceased to be in such employment refers to his pensionable pay in respect of the last 365 days of pensionable employment, ending on 31 August 2013 (i.e. the period running from 4 August 2012 to 31 August 2013, since the 28 days in February 2013 did not constitute pensionable employment as Dr Laker opted out of the Pension Scheme for the duration of February 2013).

The reference to “…member’s last year of pensionable employment…” (C1(6)) is referring to pensionable pay in respect of Dr Laker’s last year “of” pensionable employment.

It was therefore necessary according to Ms Belgrano to identify the “last year” i.e. the last 365-day period “of pensionable employment”. She pointed out that Regulation C1(6) itself uses the period of 365 days and, in the context of the Regulations “relevant daily proportion” means 1/365th of the amount that would apply in respect of one year. (Regulation A2).

Regulation C1(6) makes it clear that Dr Laker’s last year of pensionable employment is that “ending on the date the [he] ceases to be in such employment”. The date on which Dr Laker ceased to be in pensionable employment for the purpose of Regulation C1(6) was therefore 31 August 2013 (not, as Mr Firth argued, 31 January 2013).

Given the terms and context of Regulation E1, we should adopt a ‘wait and see’ approach to determine the “last year” of pensionable employment “ending on the date the member ceases to be in such employment”. The relevant date is therefore 31 August 2013 because, by virtue of Dr Laker’s second opt-out (the opt-out in August 2013), he was treated as having left pensionable employment on the date the opt-out notice took effect, which was 1 September 2013 (Regulation B4(1)) . This gives a sensible meaning according to Ms Belgrano to the words “last year” and “ending on the date the member ceases to be in such employment”.

Ms Belgrano claimed there were two fundamental problems with Dr Laker’s case. The first is that he seeks to give meaning to the word “year”, viewed in isolation and therefore fails to engage with the meaning of final year’s pensionable pay that is actually provided by Regulation C1(6), and also fails to engage with Regulation L4, which applies to him. The second fundamental problem with his case is that his “year” does not end (as it must, given the terms of Regulation C1(6) and Regulation L4) “on the date the member ceases to be in such employment” because it ends on 31 January 2013 even though Dr Laker was subsequently in pensionable employment from 1 March 2013 to 31 August 2013 and the “pre-break” and “post-break” periods of pensionable employment are not to be treated separately under Regulation L4, with the effect that Dr Laker ceases to be in pensionable employment on 31 August 2013.

Ms Belgrano submitted that the fundamental problem that affects the entirety of the Dr Laker’s case is that he seeks to construe the word “year” in isolation but that is not what Regulation C1(6) says. Thus, his reliance on the “ordinary meaning” of the word “year”, “week”, “month”, “decade” etc. as well as his reliance on a dictionary meaning of the word “year” in isolation adopts the wrong approach to the construction of Regulation C1(6). It is necessary to construe the phrase that is actually used in that Regulation, in the context of the Regulations as a whole including, importantly, Regulation .

Regulation L4 applied to Dr Laker because he left pensionable employment (in February 2013) without becoming entitled to a pension under any of regulations E1 to E5 and later returned to pensionable employment (in March 2013) before becoming entitled to a pension under the scheme (see Regulation L4(1)). Ms Belgrano advised that Regulation L4(2) applied to Dr Laker because he left pensionable employment with a preserved pension under Regulation L1 and returned to pensionable employment within 12 months after leaving. The effect of that Regulation applying is,

It should be noted that in Dr Laker’s case, “pensionable service”, which is “service which counts both for the purpose of ascertaining entitlement to benefits” under the Regulations and “for the purpose of calculating them”, means “any period of pensionable employment in respect of which the member contributes” to the scheme (Regulation C2(1)(a)).

Thus, on the face of it, the effect of Regulation L4(2) is that both the pre-break and post-break periods of pensionable employment count towards his pensionable service, and that when calculating his entitlement to benefits under the scheme, his pensionable service is treated as continuous. Putting it another way, Ms Belgrano argued the pre-break and post-break periods of pensionable employment are not treated separately when calculating his benefits. Regulation L4(8) makes express provision for the opposite situation i.e. where the pre-break period of pensionable employment and the post-break period of pensionable employment have to be treated separately:

Dr Laker’s position, that disjunctive periods of pensionable employment cannot be treated together for the purpose of calculating final year’s pensionable pay, is wrong according to Ms Belgrano. When Regulation L4(2) applies and L4(8) does not apply, the final year’s pensionable pay should be calculated by reference to the whole period of pensionable employment taken together (and not by reference to the pre-break and post-break periods of pensionable employment, viewed separately as would be the case under Regulation L4(8)).

Dr Laker’s construction of Regulation C1(6), which has the effect of ignoring pensionable pay in respect of pensionable employment, is also inconsistent with the scheme of the Pension Regulations as a whole: for example, and without limitation, Regulation N1(5) provides that:

Both parties accepted that Dr Laker was covered for death in service benefits prior to February 2013 and after March 2013. Regulation F1 provides:

The “last year of pensionable employment” is that which ends “on the date the member ceases to be in such employment” (Regulation C1(6)). HMRC’s construction gives that a sensible meaning since HMRC work backwards from 31 August 2013. By contrast, Dr Laker’s approach has the effect of treating the date he ceased to be in pensionable employment as being 31 January 2013 even though his pensionable employment resumed on 1 March 2013 and continued until 31 August 2013. Mr Firth seeks to justify that result on the basis that the legislation does not refer to the date on which the Appellant “permanently” ceased to be in pensionable employment. Regulation C1(6), in asking for the “date the member ceases to be in [pensionable] employment” is clearly looking for the date (singular) on which pensionable employment ceased for the last time.

Ms Belgrano argued that the word “year” should not be construed in isolation but in the context of the Regulations as a whole including importantly Regulation L4. What Regulation C1(6) says must be identified is the last year “of” pensionable employment. Dr Laker’s approach equates the language of C1(6) with a search for the last 12 consecutive months comprised solely of pensionable employment which is not what the Regulation says and there is no basis for that approach. The Regulations do not refer to “12 consecutive months” consisting “solely” of pensionable employment, as Dr Laker’s case requires.

Mr Firth argued that periods of pensionable service are expressly aggregated but there is no similar provision aggregating periods of pensionable employment. Ms Belgrano argued that on the face of it, the effect of Regulation L4(2) is that both the pre-break and post-break periods of pensionable employment count towards the Dr Laker’s pensionable service, and that when calculating his entitlement to benefits under the scheme, his pensionable service is treated as continuous. Put another way according to Ms Belgrano, the pre-break and post-break periods of pensionable employment are not treated separately when calculating his benefits.

Regulation L4(8) makes express provision for the opposite situation i.e. where the pre-break period of pensionable employment and the post-break period of pensionable employment have to be treated separately:

Dr Laker’s position, that disjunctive periods of pensionable employment cannot be treated together for the purpose of calculating final year’s pensionable pay, is wrong according to Ms Belgrano. When Regulation L4(2) applies and L4(8) does not apply, the final year’s pensionable pay should be calculated by reference to the whole period of pensionable employment taken together (and not by reference to the pre-break and post-break periods of pensionable employment, viewed separately as would be the case under Regulation L4(8)).

Mr Firth’s construction of Regulation C1(6), which has the effect of ignoring pensionable pay in respect of pensionable employment, is also inconsistent with the scheme of the Pension Regulations as a whole. For example, and without limitation,

HMRC understands Dr Laker accepts that during the period of “pensionable employment” that he seeks to ignore for the purposes calculating the “final year’s pensionable pay” he was covered for death in service benefits.

Regulation F1 provides as follows:

Ms Belgrano argued that the “last year of pensionable employment” is that which ends “on the date the member ceases to be in such employment” (Regulation C1(6)). Her construction gives that a sensible meaning since HMRC work backwards from 31 August 2013.

By contrast, Dr Laker’s approach has the effect of treating the date Dr Laker ceased to be in pensionable employment as being 31 January 2013 even though his pensionable employment resumed on 1 March 2013 and continued until 31 August 2013. He seeks to justify that result on the basis that the legislation does not refer to the date on which he “permanently” ceased to be in pensionable employment. Regulation C1(6), in asking for the “date the member ceases to be in [pensionable] employment” is clearly looking for a single date on which pensionable employment ceased for the last time.

Ms Belgrano submitted that Regulation L4(8) is consistent with HMRC’s case that “final year’s pensionable pay” should be calculated by counting backwards from 31 August 2013 and not by counting backwards from 31 January 2013 because, in the absence of the deeming in Regulation L4(8), the 31 January 2013 date is not treated as the date on which the member “ceases to be” in pensionable employment for the purposes of calculating the final year’s pensionable pay.

Ms Belgrano further submitted that, in Dr Laker’s case, the deeming in Regulation L4(8) does not apply (since the effect of Regulation L4(2) is that the pre-break and post-break periods of pensionable service are treated as continuous) a single calculation counting backwards from 31 August 2013 can be carried out (rather than needing to carry out two calculations, one as if the period up to 31 January 2013 had been the only period of pensionable employment and the other as if the period from 1 March 2013 to 31 August 2013 had also been the only period of pensionable employment). Mr Firth’s approach of working back from 31 January 2013 also creates an inconsistency between Dr Laker’s period of “pensionable service” which, it is understood he accepts , should be calculated up to 31 August 2013 and “last year of pensionable employment, ending on the date the member ceases to be in such employment” which Mr Firth says should be calculated only up to 31 January 2013. Given that in Dr Laker’s case pensionable service (as defined in Regulation C2(1)) is comprised of “any period of pensionable employment…”, it is inconsistent for the period of pensionable service to end on a different day to the period of pensionable employment.

Dr Laker’s case is also inconsistent with the approach of the BMA and the NHSPA who both considered that the final year’s pensionable pay was judged by reference to the year of pensionable employment ending on 31 August 2013.

Dr Laker’s construction also has the strange effect that if he had not been in NHS employment prior to 1 March 2013, Regulation C1(6)(b) would apply and his “final year’s pensionable pay” would be calculated by reference to his employment ending on 31 August 2013.

Finally, Ms Belgrano submitted that Dr Laker’s construction of “final year’s pensionable pay” also appears to be inconsistent with the language used in his contract of employment and with the publicly available NHS/BMA guidance on the interpretation and operation of the NHS Pension Scheme:

HMRC note that the Lifetime Allowance Charge has been abolished with effect from 6 April 2023: Dr Laker’s figures are understood to be out of date given the changes in the legislation. On the basis of HMRC’s current understanding of the facts the only lifetime allowance charge which will remain if Dr Laker loses his FP 2012 is the charge when he crystallised his NHS pension which was £13,970 and has been paid, although FP 2012 would also affect the quantum of the Lump Sum Allowance and his Lump Sum and Death Benefit Allowance.

At first blush Mr Firth’s arguments that the word “year” must mean 12 consecutive months seem compelling but on further consideration of the examples given by Ms Belgrano this Tribunal prefers the definition adopted by HMRC as being a total period of 12 months working back from 31 August 2013.The Tribunal notes that Dr Laker was paid the same monthly salary regardless of whether the month had 28, 30 or 31 days. It is a coincidence that Dr Laker opted out for the month of February 2013.

In paragraph 53 above we referred to Mr Firth’s argument that a week is always seven consecutive days. This is incorrect. The Tribunal knows of several contracts of employment where the employee’s annual holidays are stated to be, for example, five weeks. Nowhere in the contract does it state that the employee may only take seven consecutive days at a time. The employee is entitled to take individual days and/or consecutive days as holidays so long as the overall number of days does not exceed five weeks. The Tribunal applies a similar approach to the word “year” so that it means a combination of periods totalling a year.

Reading the legislation in context (and Dr Laker’s contract of employment and the NHS Pension Scheme Guide) leads the Tribunal to the conclusion that the word “year” in regulation C1(6) – last year of pensionable employment – must mean a combination of periods totalling 12 months.

Although not relevant to the current appeal there could be occasions where adopting Dr Laker’s definition of “year” the pension calculation would be less than fair if the member received a pay rise during the post-break period which period could be up to eleven months before the date the member ceases to be in employment.

Accordingly, using HMRC’s calculations, Dr Laker was slightly over the threshold and they were correct to revoke his FP 2012 certificate.

This document contains full findings of fact and reasons for the decision. Any party dissatisfied with this decision has a right to apply for permission to appeal against it pursuant to Rule 39 of the Tribunal Procedure (First-tier Tribunal) (Tax Chamber) Rules 2009. The application must be received by this Tribunal not later than 56 days after this decision is sent to that party. The parties are referred to “Guidance to accompany a Decision from the First-tier Tribunal (Tax Chamber)” which accompanies and forms part of this decision notice.