Davies & Anor, R (on the application of) v HM Revenue & Customs

In 1936 the Income Tax Codification Committee complained that the absence of guidance as to whether or not a taxpayer was resident in the United Kingdom was intolerable. In 1973, the Revenue issued the first version of IR20, a booklet designed to provide what it described as general guidance in relation to the residence and ordinary residence of individuals. The appellant, Robert Gaines-Cooper, claims that, on a proper interpretation and application of IR20 (1999 version) he has been neither resident nor ordinarily resident since 1976. The appellants Robert Davies and Michael James assert that they too, in the tax year 2001-2, were neither resident nor ordinarily resident, were the Revenue properly to interpret and apply IR20.

The appellants contend that the Revenue (now HMRC) has failed to interpret IR20 correctly and has unlawfully refused to apply it. On the contrary, after the taxpayers had left the United Kingdom, in 2005 it implemented and applied an unannounced change of policy so as to deprive these appellants of the status of non-residence, to which, by virtue of the terms in which IR20 was expressed, they were entitled. In 2007, Mr Glyn Davies, a tax partner in Price Waterhouse Coopers LLP described this approach as a betrayal of trust.

This application for judicial review (as this appeal has become), therefore, raises the question of the status and interpretation of IR20 (1999 version). Although that booklet has now been replaced, that question is also of wider importance: it goes to the heart of the relationship between the Revenue and the taxpayer. It is not about whether as a matter of fact, on the application of authority, the appellants are resident and ordinarily resident. It concerns the public law obligations imposed on the Revenue, once it has chosen to give guidance to taxpayers as to how it will approach questions relating to their residence for the purposes of taxation.

It is notorious that the principles to be applied are to be found, not in the few statutory provisions (sections 334-336 ICTA 1988, now sections 829-832 ITA 2007), which do not purport to provide a statutory code but in case law, mainly from the late 19^th and early 20^th Century. As the Codification Committee recognised, only study of that jurisprudence would enable intelligent prediction of the outcome of an assertion as to residence or non-residence. All the more important, then, that guidance should be given on which taxpayers could rely.

On 28 November, 2006, the Revenue determined that the appellants, Davies and James, were ordinary residents in the UK and thereby triggered their right of appeal to the Special Commissioners. But the status of those appellants as not resident and not ordinarily resident within the UK has not yet been determined by that statutory process of appeal. The appellants contend that the Revenue has no right to insist upon any such determination. They assert that they left for full-time employment in Belgium in March 2001 and because they started that employment in the tax year 2001/2, and it lasted for at least a full year, they are entitled to be treated as non-resident and not ordinarily resident in the UK for the tax year 2001/2, in accordance with § 2.2 IR20, irrespective of any decision the Tribunal might reach. Alternatively, they are entitled to such treatment, by virtue of §§ 2.7-2.9 because they left the UK indefinitely, for a settled purpose.

The Revenue’s refusal to accept that status stems, they say, from its failure correctly to interpret IR20, from its refusal to apply IR20, properly construed, or, at the very least, from an attempt to apply a different and unannounced change to the interpretation and application of IR20 which it had previously adopted. The taxpayers sought permission to move by way of judicial review.

There then arose a dispute with the Revenue as to whether the hearing of the application for judicial review should take place before or after an appeal. It is of note that the Revenue contended that an application for judicial review would not be pre-empted by a hearing before the Special Commissioners, at that time charged with determining appeals. The Court of Appeal took the view that there was at least a real risk that a decision by the Special Commissioners would rule out a claim based on legitimate expectation and set aside the order of Burnton J staying the application ([2008] EWCA Civ 933 [18 and 19], [24]).

These preliminary but time-consuming and costly skirmishes bore little fruit. Wilkie J refused permission [2008] EWHC 1218 (Admin). The appellants appealed. On 10 July, 2009, with a fair appreciation of the direction in which the wind was gusting, Miss Simler QC accepted that the appeal should be allowed, and we decided to hear the application for judicial review once the Revenue had had the opportunity to serve evidence.

The appellant Gaines-Cooper’s contention that he had not been resident or ordinarily resident in the United Kingdom in the years 1993/4 to 2003/4 was rejected by the Special Commissioners in 2006. In contrast to the other appellants, Mr Gaines-Cooper says he was compelled to appeal to the Special Commissioners. But, contends Mr Gaines-Cooper, following the line adopted by the Revenue in the other appeal, such a determination does not preclude reliance on the legitimate expectation that IR20 would be applied, and that, on its proper construction it gave a binding assurance to him that he would be treated as not resident and not ordinarily resident since 1976 despite the conclusion of the Special Commissioners.

One month after a hearing lasting two days, Lloyd Jones J decided that the appellant’s claim was unarguable and refused permission [2008] EWHC 2608 (Admin). We allowed an appeal from that refusal at the same time as we allowed the other appeal and this hearing has proceeded as an application for judicial review by all the appellants. Neither Lloyd Jones nor Wilkie JJ had the advantage of the oral argument, lasting nearly five days, if one allows for the argument on the appeal in July 2009, by which we were helpfully guided through about 8 lever-arch files, and skeleton and supplementary skeleton arguments, still less the opportunity of taking an overall view of both applications, which raise the same legal, but not the same factual, issues.

I venture to suggest that if they had had that benefit, they might not have been led by the Revenue to the view that permission should be refused. Lloyd Jones J was faced with an order by another judge that the application should follow the permission hearing, but since, if permission had been given, the Revenue would have wished to serve evidence, that order would have had no effect; the order did not justify the Revenue’s approach to the permission application. The Revenue ought to have appreciated, much earlier, that these applications did raise important issues as to the effect of the guidance in IR20, and that their interpretation and application of the guidance had aroused serious concerns in the minds of serious practitioners, which the public interest required should either be allayed or confirmed, whatever the Revenue’s optimism as to the outcome. The procedural wrangles and opposition to permission have merely caused an all too costly delay to full consideration of the correct conclusion as to IR20. The sifting procedure in CPR Pt 54 was designed to protect public bodies against weak and vexatious claims (R v Secretary of Trade and Industry ex p Eastaway [2002] 1 WLR 2222 at 2227H). It was not designed for lengthy inter partes hearings; it was to enable judges to decide whether a case might be arguable on a quick perusal of the material available (per Lord Diplock in R v IRC ex p National Federation of Self-Employed and Small Businesses Ltd [1982] AC 617 644A). Nor, I suspect, was the process intended to afford an opportunity to a public body, such as the Revenue, to resist full consideration of matters of great importance not just to the taxpayer but to the Revenue itself.

The importance of the extent to which thousands of taxpayers may rely upon guidance, of great significance as to how they will manage their lives, cannot be doubted. It goes to the heart of the relationship between the Revenue and taxpayer. It is trite to recall that it is for the Revenue to determine the best way of facilitating collection of the tax it is under a statutory obligation to collect. But it should not be forgotten that the Revenue itself has long acknowledged that the best way is by encouraging co-operation between the Revenue and the public. Mr Beighton, a member of the Board of Inland Revenue, said as much twenty years ago (cited by Bingham LJ in R v IRC ex p MFK Underwriting Agents Ltd [1990] 1 WLR 1545 at 1568H). Co-operation requires fair dealing by the Revenue, and frank and open dealing by the public. Of course the Revenue may refuse to give guidance and re-create a situation in which the taxpayers and their advisers are left to trawl through the authorities to find a case analogous to their own, or, if they are fortunate, a statement of principle applicable to their circumstances. But since 1973, in a field fraught with borderline cases relating to an enormous variety of circumstances, the Revenue has chosen to confer what presumably it regarded as a benefit on taxpayers who wished to know whether they were likely to be treated as resident or not.

In R v IRC ex p MFK Underwriting Agents Ltd the taxpayer ferreted through correspondence and notes of meetings in attempting to identify a binding assurance from the Revenue as to the tax treatment of corporate bonds. No such binding assurance could be discovered. But it should be recalled that counsel for the Revenue attempted to persuade the court that the Revenue’s managerial discretion was not of such width as to allow the Revenue to give any binding assurance which might conflict with the taxpayer’s true liability in law. The Revenue was not permitted to indicate in advance that it would not collect tax which, on a proper application of the law, might prove to be due (the argument is at 1566 G-1567C). Bingham LJ and Judge J rejected this submission; the Revenue’s discretion was not limited to a decision as to the best and most economical method of collection:-

“No doubt a statement formally published by the Inland Revenue to the world might safely be regarded as binding, subject to its terms, in any case falling clearly within them” (per Bingham LJ 1569C).

MFK was concerned with answers the Revenue gave informally to different taxpayers and their advisers as to how, generally, they would treat transactions which had not yet taken place. But the Court recognised that the Revenue could be bound, through the medium of judicial review, to honour statements made to the public as to how it would treat a taxpayer in particular, defined circumstances (Judge J at 1572F-H and 1574H to 1575B).

As I have recalled, issues of residence may give rise to complex questions of fact. The principles depend on case law not on statutory definition. The Court of Appeal has recently adopted Dr Brice’s summary of relevant factors in Shepherd v HMRC and Lewison J’s identification of relevant features in Grace v HMRC [2009] EWCA Civ 1082 [6] and [7]. Many of the questions which must be asked to determine questions of residence, such as whether the purpose for which a person has adopted an abode is ‘settled’ [6(vi)] or whether there is a sufficient degree of permanence and continuity [6(iv)] lend themselves to no certain conclusion. They require value judgments, which may express a wide range of views, all of which are within the area of reasonable conclusion, even when they conflict.

IR20 reminds taxpayers and advisers of this uncertainty, and warns that the Revenue’s determination of residence will often depend on its assessment of the facts. The Preface reads :

“The notes in this booklet reflect the law in practice at October 1999. They are not binding in law and do not affect rights of appeal about your own tax.

You should bear in mind that the booklet offers general guidance on how the rules apply, but whether the guidance is appropriate in a particular case will depend on all the facts of that case. If you have any difficulty in applying the rules in your own case you should consult an Inland Revenue Tax Office ... Some practices explained in this booklet are concessions made by the Inland Revenue. A concession will not be given in any case where an attempt is made to use it for tax avoidance.”

Paragraph 1.1 provides:

“The terms ‘residence’ and ‘ordinary residence’ are not defined in the Taxes Acts. The guidelines to their meaning in this Chapter and Chapters 2 (resident status of those leaving the UK) and 3 (those coming to the UK) are largely based on rulings of the Courts. This booklet sets out the main factors that are taken into account, but we can only make a decision on your residence status on the facts in your particular case.

As mentioned in paragraph 1.4, even if you are resident (or ordinarily resident) in the UK under these rules, the terms of a double taxation agreement with another country might affect your final tax position if, for example, you are resident in both that country and the UK.”

But the undoubted need to recognise how dependent questions of residence are upon the facts of individual cases does not and should not detract from the utility and effect of the guidance the Revenue chose to promulgate. IR20 contains statements of normal practice, as acknowledged by the Revenue in its Brief 01/07, following reports of the Gaines-Cooper decision. It sets out a limited number of specific situations in which a taxpayer will be treated as non-resident. If a taxpayer falls within the situation described by the Revenue, the Revenue has given an assurance that it will treat the taxpayer in accordance with the terms of the guidance. If a taxpayer comes, for example, within the terms of 2.2, the Revenue has given an assurance that it will treat that taxpayer as not resident and not ordinarily resident. It will not be permitted to resile from that assurance, unless and until it announces that it proposes, for the future, to alter the circumstances in which it will accept non-resident status.

These propositions do not in any way conflict with the principles identified in cases such as R(Wilkinson)v IRC 77 TC 78 or Al Fayed v the Advocate General for Scotland 77 TC 273. In Wilkinson the Revenue’s statutory power of management under section 1 of the Taxes Management Act 1970 could not be invoked to compel it to introduce a widower’s bereavement allowance, when Parliament had chosen not to do so. In Al Fayed the Revenue had entered into an agreement outwith its powers and could not be compelled to apply it for the future. But the guidance, in IR20, was plainly within the Revenue’s powers of providing statements of practice and identifying how it proposed to deal with the residential status of taxpayers in particular circumstances. By doing so it gave some comfort (the extent of which is in dispute) to taxpayers on which they were entitled to rely. It assisted co-operation and, thereby, the collection of tax.

The Revenue does not dispute these general propositions. It accepts, to use its own words, that the taxpayer has a legitimate expectation that it will apply the guidance in IR20 to the facts of his particular case and if satisfied that the facts fall within one of the circumstances in Section 2 indicating a certain residence treatment, will treat him accordingly. Their evidence reiterates that assurance.

I stress the Revenue’s assurance and its origin in well-established principles of public law in relation to the Revenue’s dealings with the taxpayer, because, from time to time in the meandering course of these proceedings, it has appeared that the appellants, their advisers and the taxation profession have doubted whether IR20 can be relied upon at all; they harbour the suspicion that the Revenue regards itself as free to apply or to disapply its guidance, as it chooses. They describe IR20 as misleading and a trap.

This fear is, in part, attributable to the Revenue’s own stress, earlier in these proceedings, upon the limits of its managerial discretion, contending that it cannot lawfully treat taxpayers as non-resident irrespective of their status as a matter of law or irrespective of facts found by the Special Commissioners on appeal. They rely, for example, on Wilkinson [45]-[46] (see first skeleton argument 22.5.09 § 24-25). The Revenue’s description of IR20 as “no more than guidance” fuels that fear.

The Revenue’s emphasis appears to have influenced the approach of the judges who refused these appellants permission. In the Gaines-Cooper application, Lloyd Jones J described IR20 as “a statement of the governing legal rules and guidance as to how those legal rules apply” [21]. They provide “guidance which, if applied by the taxpayer to the facts of his individual case, will show what the approach of the Revenue will be… (they) cannot be construed as a binding promise that HMRC will ignore an individual’s residence status in law or will treat him as a non-resident by ignoring findings of fact that he is resident and ordinarily resident in the United Kingdom” [22]-[23]. It must be recalled that the judge was dealing with an argument that IR20 contained what Mr Milne QC, for Mr Gaines-Cooper, called “bright line tests” which the Revenue was bound to apply, irrespective of the law, in the context of facts already found in an unappealed determination of the Special Commissioners.

I shall turn later to Mr Milne’s description of IR20. But it seems to me that Lloyd Jones J’s rejection of the argument, even if correct, led him to diminish the important effect of IR20. Similarly, Wilkie J stressed the cautionary words in the Preface and 1.1 [39], in concluding that it was not designed to be a blueprint [49]. Whether the judges’ conclusions as to the interpretation and applicability of IR20 are correct or not, I wish to emphasise that it does provide guidance as to particular circumstances in which the Revenue gives a binding and lawful assurance that it will treat a taxpayer, whose case falls within the circumstances described, as not resident and not ordinarily resident. So to conclude is to do no more than echo the assurance the Revenue itself gives within section 2.2.

2.2 provides, under the heading “Working Abroad”:

“If you leave the United Kingdom to work full-time abroad under a contract of employment, you are treated as not resident and not ordinarily resident if you meet all the following conditions:

Your absence from the UK and your employment abroad both last for at least a whole tax year;

During your absence any visits you make to the UK:

Total less than 183 days in any tax year, and

Average less than 91 days per tax year ... "

2.3

provides:

“If you meet all the conditions in paragraph 2.2, you are treated as not resident and not ordinarily resident in the UK from the day after you leave the UK to the day before you return to the UK at the end of your employment abroad ...

If there is a break in full-time employment, or some other change in your circumstances during the period you are overseas, we would have to review the position to decide whether you still meet the conditions in paragraph 2.2. If at the end of one employment you returned temporarily to the UK, planning to go abroad again after a very short stay in this country, we may review your residence status in the light of all your circumstances of your employment abroad and your return to the UK.”

The Revenue chose to use this language “you are treated as not resident and not ordinarily resident if you meet all the following conditions”. The Revenue has used similar language in relation to the guidance on leaving permanently or indefinitely (2.7-2.9, cited [36]), for example at 2.8, “If you have left the UK permanently or for at least three years, you will be treated as not resident and not ordinarily resident”. This is the language of assurance, which the Revenue, in the interests of fair dealing, has bound itself to honour.

The real difficulty, exposed by these appeals, lies not in imposing well-established public law obligations on the Revenue, but in the interpretation of the assurances the Revenue has given and their application to the facts relating to these particular taxpayers. That there is difficulty is not surprising. Just as the application of the jurisprudence relating to residence may lead to a wide range of value judgements, so does the guidance itself. In 2.2, what is “full-time employment”? To what extent do remaining duties to be performed within the UK affect any conclusion that the work is full-time abroad? IR20 at 2.5 demonstrates that the issue may not be straightforward. Equally, 2.7 – 2.9 (cited [37]) require judgements to be made as to whether a person has gone abroad “permanently” and whether the purpose for which a person has gone abroad is “settled”.

The very terms of the paragraphs in IR20 on which the appellants rely demonstrate that they do not contain any bright-line test of certain application. They require consideration of particular facts and that a judgement be made on those facts. That seems to me to be the importance of the opening words in the preface and 1.1 (cited [16]). They underline the problem inherent in both the jurisprudence and the guidance, that a view will have to be taken on the facts, even though they are undisputed, and that there will be many cases in which opposite but equally reasonable views may be taken.

There are bright line tests contained within IR20, for example that a person will always be resident, if present in the UK for 183 days (1.2), a rule which merely echoes s.336(2) ICTA 1988. The day-count rules, to which I must refer later, have no statutory basis, but, if fulfilled, will protect a taxpayer against loss of non-resident status or the acquisition of resident status. But otherwise, unless the facts of the taxpayer’s case are beyond all reasonable dispute, the very terms of IR20 provide no certainty of the outcome of any claim to resident or non-resident status. The uncertainty is compounded by the possibility that one who is resident abroad may retain his resident status in the UK (1.4).

It is in the context of those principles, which afford only limited comfort to the taxpayer, that I turn to the questions of interpretation to which these appeals give rise.

Mr. Goldberg QC submits that the Revenue has misconstrued 2.2 (cited [24]), and, in particular:

“Your absence from the UK and your employment abroad both last for at least a whole tax year;

He contends that, properly construed, the reference at the first bullet point to “a whole tax year” is a reference to at least a whole tax year but not necessarily the whole of the tax year in which the Revenue seek to charge the appellants to capital gains as residents in the tax year 2001/2. He contends that the Revenue has erroneously transposed the definite for the indefinite article. Given the assurance contained in 2.2, he submits, Davies and James are entitled to be treated as not resident and not ordinarily resident in 2001/2002, since it is clear that their full-time employment has run for longer than a full tax year and continues to run (PWC letter 20 December 2004).

The importance of this submission stems from the facts which remain in dispute between the Revenue and these appellants. They left the UK in March 2001, but the Revenue do not accept that they did so “to work full-time abroad”, because they were not in full-time employment in April 2001. Accordingly, they cannot be treated as non-resident in the tax year 2001/2 since the extra-statutory concession A11, which permits a tax year to be split when a taxpayer leaves part way through the year, has no application to capital gains tax (see 1.5 IR20 (cited [33]]). The appellants accept that the Revenue is entitled to take the view that the appellants were not in full-time employment in April 2001 (Mr Glyn Davies’s 2^nd statement § 85) but attempt to finesse the point with their argument that the proper interpretation of 2.2 renders that view irrelevant.

In my view, properly construed, 2.2 does not entitle a person to non-resident status, for capital gains tax purposes, unless he leaves to work full-time either before or by the start of a tax year, in the instant case by 6 April 2001. To come within 2.2, a taxpayer must leave for and remain in full-time employment throughout the relevant tax year. Full-time employment throughout any subsequent tax years does not affect the date when a taxpayer first attained non-resident status; that date is determined by reference to the date the taxpayer left to work full-time abroad.

IR20 must be construed as a whole, by reference to all its provisions and so far as possible so that they do not contradict each other. It makes no sense to permit a taxpayer to claim non-resident status under 2.2 notwithstanding that the full-time employment started only part way through his tax year in the light of 1.5, headed “Leaving or coming to the UK part way through a tax year,”:

“Strictly, you are taxed as a UK resident for the whole of a tax year if you are resident here for any part of it. But if you leave or come to the UK part way through a tax year, the year may, by concession...be split. Where this applies, your tax liabilities on income which are affected by tax residence will be calculated on the basis of the period of your actual residence here during the year...this has the same effect as splitting the tax year into resident and non-resident periods.”

Mr Goldberg QC contended that this point of interpretation was overlooked by Mr West, the Inspector dealing with his clients’ case. Mr West accepts that due to an oversight he did not deal with it, in response to Glyn Davies’s letter of 20 December 2004. If it mattered, Mr Davies accepts that the point only occurred to him as an afterthought, after a meeting on 16 December 2004 (2^nd statement). It does not matter: the question of interpretation does not turn on this tu quoque exchange.

It might have mattered if the correct construction was unclear. If that were the view of the court, the Revenue contends that its own interpretation falls at least within the range of reasonable construction (R v MMC exp South Yorkshire Transport Ltd [1993] 1 WLR 23, 32C-33A). That the point only occurred to Mr Davies as an afterthought demonstrates, so Miss Simler QC argued, that the appellants’ construction was not the only reasonable construction. But it does not matter, since in my view the correct interpretation is plain. The Revenue has consistently taken the view that the appellants did not work full-time in Belgium in April 2001 and that that is fatal to their claim that they come within the terms of 2.2 (see the Revenue letters dated 23 February 2006, 20 March 2006, culminating in the decision letter dated 29 November 2006):-

“Based on the facts of the matter it has not been established that your clients’ employment was full time employment throughout the tax year to satisfy a claim to be non-resident under paragraph 2.2 of IR2O.”

That decision was based on a proper interpretation of 2.2. But it does not dispose of their claim to entitlement under 2.7-2.9, nor that of Mr Gaines-Cooper whose claim is restricted to entitlement under those paragraphs, nor of their assertions of an unannounced change of practice.

Paragraphs 2.7 to 2.9 are contained under a single heading, “Leaving the UK permanently or indefinitely”. They provide, in so far as is relevant, as follows:

The argument relating to these provisions lies at the heart of these appeals. The appellants contend that the Revenue has not only misconstrued these provisions but has adopted and applied that misconstruction only after these appellants left the United Kingdom. Before they left, the Revenue had construed these paragraphs in a way which would have led to their being treated as non-resident. The subsequent application of a misconstruction of these paragraphs amounts to an unannounced change of policy which infringes the appellants’ legitimate expectation that the Revenue’s earlier approach would be applied to their cases. Thus, these provisions give rise to two distinct issues: the correct construction of 2.7-2.9 and whether the Revenue did, in fact, alter its construction of those paragraphs and unfairly apply that altered construction to these appellants.

The dispute is focussed on whether, under these paragraphs, it is necessary for a taxpayer to demonstrate he has severed his ties to the extent that his previous social and family ties in the United Kingdom are no longer retained. That issue, contend the appellants, does not arise on a proper construction of these paragraphs, and even if it does, was not a construction applied by the Revenue until these cases provoked an unannounced change of policy.

The taxpayers contend that the question of construction turns on the meaning of the word “leave”. That word must be construed consistently throughout Section 2. It connotes no more than a departure; it does not require social and family ties within the UK to be cut.

I do not accept that anything turns on the meaning of the word “leave”. No light is shone merely by substituting another verb, such as “depart” or suggesting that it means no more than climbing aboard a Eurostar train to Brussels or whatever airline flies to the Seychelles. The argument gets the taxpayers nowhere because the word takes its meaning from the context in which it appears. In Section 2 wherever the word “leave” or the expression “gone abroad” appear, they are subject to a further qualification which the taxpayer must satisfy if he is entitled to the status of non-residence.

In 2.2, it is not enough that the taxpayer has left the UK, he must have left to work full-time. Absence is not sufficient, it must be absence whilst engaged on full-time employment for at least a whole tax year. No more is, however, required. The absence need be neither permanent nor indefinite. Accordingly, there was no dispute between the parties but that there is no requirement, under 2.2, for a taxpayer to demonstrate that he has severed family and social ties within the UK. The Revenue accepts that the taxpayer need do no more than establish that he has left the UK for full-time employment abroad, and that the employment has continued throughout the relevant tax year. There is no need to be concerned with any persisting social or family ties in the UK, unless those ties themselves cast doubt on whether the employment is genuinely full-time.

Once it is accepted that there is no need, save in the limited circumstances I have described, to consider whether social and family ties have been retained under 2.2, it follows, contend the appellants, that there is no such need under 2.7-2-9. I disagree. The adverbs “permanently or indefinitely” make, as a matter of construction, all the difference. The extent to which a taxpayer retains social and family ties within the United Kingdom must have a significant and often dispositive impact on the question whether a taxpayer has left permanently or indefinitely (for at least three years). It makes no sense to construe “leave” when qualified by the adverbs permanently or indefinitely as referring to the process of going abroad. They clearly require consideration of the quality of the absence and contrast with 2.1 which provides, under the rubric “Short Absences” :

“You are resident and ordinarily resident in the UK if you usually live in this country and only go abroad for short periods - for example, on holiday or on business trips”.

The description, in IR20, of the evidence required to establish permanent or indefinite absence, of which examples are given in 2.8, demonstrates the need to cut existing ties with the UK:

“This evidence might be, for example, that you have taken steps to acquire accommodation abroad to live in as a permanent home, and if you continue to have property in the UK for your use, the reason is consistent with your stated aim of living abroad permanently or for three years or more.” (my emphasis).

That example of evidence which may be required is an example which would demonstrate the severance of social and family ties hitherto maintained within the UK.

That such severance is necessary to establish permanent or indefinite absence finds further confirmation in the reminder, in 1.4, headed “Residence in both the UK and another country”:

“It is possible to be resident (or ordinarily resident) in both the UK and some other country...at the same time. If you are resident...in another country, this does not mean you cannot also be resident...in the UK. Where, however, you are resident both in the UK and a country with which the UK has a double taxation agreement, there may be special provisions in the agreement for treating you as resident of only one of the countries for the purposes of the agreement.”

The very concept of dual residence despite departure abroad reveals the adhesive quality of the previously held status of resident within the UK. Permanent or indefinite absence abroad connotes a severance of that which previously bound the taxpayer to the UK.

Accordingly, I reject the appellants’ contention that persisting social and family ties are irrelevant to 2.7-2.9, or that it is unnecessary, under IR20, for a taxpayer to demonstrate that he has cut those ties. That 2.2 imposes no such requirement provides no warrant for construing different provisions relating to permanent or indefinite absence in the same way. A taxpayer may establish that he is not resident and not ordinarily resident in two distinct ways, either by establishing that he left for full-time employment which he has maintained for at least a year, or by leaving permanently or indefinitely. The criteria for achieving that status are not the same and are not expressed in the same way.

Mr Goldberg QC sought to focus on the express terms of 2.9. It makes no reference to leaving permanently or indefinitely. Nor does it refer to any evidence which might be regarded as evidence of the severing of ties. It is sufficient to establish a settled purpose, and absence for a whole tax year. Even without a settled purpose, it contemplates that it may be sufficient if the taxpayer can show that his absence actually covers three years from departure.

2.9 must be read in the context of 2.7-2.9, all of which come under the same rubric “Leaving the UK permanently or indefinitely”. It is predicated on the taxpayer not being able to provide the evidence required under 2.8. It would be absurd if a taxpayer could acquire non-resident status on the basis of his claim that he has left permanently or indefinitely, without establishing that he has severed social and family ties in the UK, merely because he cannot provide the evidence he has done so and, thus falls within 2.9. 2.9 is designed to assist taxpayers who lack evidence, not to impose a less strict test than 2.8. Accordingly, any settled purpose, a term undefined in case law as well as 2.9, must be consistent with a distinct break, sufficient to cut pre-existing ties. Absence for three years must equally be consistent with such a break.

I note that the joint supplementary written submissions on the law of residence on behalf of all the appellants, dated 12 October 2009, explicitly accepted that 2.9 does not set out a “specific scenario”; it applies to a taxpayer who contends that he has left indefinitely or permanently but does not have the evidence to support that contention (footnote 6). Mr. Milne QC confirmed that interpretation of the rôle of 2.9 in his acceptance of the Revenue’s proposition that 2.9 flows from 2.7-2.8; 2.9 is designed to assist a taxpayer who does not have the evidence available to establish an intention to settle permanently or indefinitely abroad (supplementary skeleton argument for Gaines-Cooper, 19.10.2009, Ms McLean-Tooke, § 43).

Thus far, I have based my views that the concepts of permanent or indefinite absence connote a distinct break from previous ties within the UK on the terms of IR20. I am confirmed in that view by the objective of IR20 stated in the opening words of the Preface, that it is designed to reflect the law. It would, therefore, be surprising if IR20 had the effect of contradicting established jurisprudence. The adhesive quality of residence is reflected in the reference in s.334 ICTA to “occasional residence abroad”. The notion of a distinct break from previously held ties provides a clear test as to whether previously held residence, for example in the UK, has ceased permanently or indefinitely. It distinguishes exclusive residence abroad from dual residence, a concept recognised in Grace [6viii]. In Levene v CIR [1928] 13 TC 486, the taxpayer was held to be resident and ordinarily resident in the UK, despite the absence of any fixed home in the UK. It is of note that the Special Commissioners, whose conclusions were upheld all the way to the House of Lords, referred specifically to “his past and present habits of life” and “ties with this country” (490). Rowlatt J explained “ordinary residence” as a reference to “the ordinary course of a man’s life” (494).

The notion of a distinct break appears in IRC v Combe [1932] 17 TC 405, in which a full time apprenticeship was served in New York. Lord Sands (411) attached importance to the distinct break in residence in the UK (referred to in Grace at [6] (xiii) and see also the application of that concept in Reed v Clark [1985] STC 323 at 346h). Whilst IR20 is designed to guide and simplify, I cannot accept that it provides a warrant for ignoring so obvious a factor for determining whether a taxpayer hitherto resident and ordinarily resident in the UK has ceased to be so and has left permanently or indefinitely. IR20 itself, at 1.4, requires a value judgement to be made as to whether a taxpayer, claiming to come within 2.7-2.9, has ceased to be resident in the UK. There can be no sensible reason why one of the most telling features of such a cessation, a distinct break from family and social ties in this country, should be ignored. It would not create clarity or simplicity; it would merely remove from consideration an obvious test of permanent or indefinite absence abroad.

Under 2.2 the Revenue look no further than the question whether the taxpayer has left for full-time employment and has continued in that employment for at least a whole tax year; it does not consider persisting social and family ties. Thus, as we shall see, long distance commuters, who work full-time abroad, but return at week-ends, providing their presence remains less than the number of days specified, are treated as not resident and not ordinarily resident. This distinction between the factors relevant to 2.2 and those relevant to 2.7-9 is supported in s.335, headed “Residence of persons working abroad”. When a taxpayer works full time outside the UK, the question whether he is resident falls to be decided without regard to any place of abode maintained within the UK (s.335(1) ICTA). The statute’s exclusion of that factor demonstrates the significance of a distinct break, when non-resident status is claimed for reasons other than full-time employment abroad.

I conclude that the correct interpretation of Section 2, in the context of IR20 read as a whole, creates a clear distinction between 2.2 and 2.7-2.9. If a taxpayer claims to have left permanently or indefinitely, and thus, to have ceased to be resident and ordinarily resident in the UK, he must demonstrate a distinct break from former social and family ties within the UK. Under 2.2 he need establish no more than that he has left to work full-time abroad and has continued for at least the whole of the tax year in respect of which he asserts, for capital gains tax purposes, non-resident status. In my view that is not only a reasonable interpretation which the Revenue was entitled to adopt but is the correct construction of IR20.

There is one more discrete question of interpretation; it relates to the question of the number of return visits during absence (2.2) or since leaving (2.8). The appellants contended that the only relevant issue which arose under 2.7-2.9 was the number of visits they had made to the UK (further written submissions, 1.4). That seems to me an impossible construction. The number of return visits does not establish non-residence; for that the taxpayer must look to the substance of 2.2 or 2.7-9. The number of return visits are important only to establish whether non-resident status, once acquired, has been lost. Conversely, Section 3 explicitly provides that if a non-resident makes regular visits exceeding the number of days specified, a taxpayer will be treated as resident, and, thus, by reference to the number of regular visits as acquiring resident status in the UK (see 3.3 and 3.4). The only reference to the length of absence from the UK in Section 2, on the basis of which non-resident status may be acquired, are those references to absence for the whole of a tax year under 2.2 and 2.9 and absence for at least three years in 2.8 and 2.9. As 2.7 makes clear, if the specified number of return visits is exceeded, non-resident status will be lost, even if the taxpayer has gone abroad permanently. But the converse does not follow as a matter of logic or language: if the number of return visits is not exceeded, it by no means follows that the taxpayer has left or gone abroad permanently or indefinitely. If a taxpayer limits his return visits, that will not lead to acquisition of non-resident status. He must leave to work full-time, or permanently or indefinitely, as IR20 tells him.

My views as to the correct interpretation of IR20 by no means conclude these appeals. The taxpayers assert that the Revenue has changed its construction in an unannounced change of policy in 2004/5 and retrospectively applied that construction to the cases of these appellants, after they had left. Fairness dictates that it should not be permitted to do so.

The appellants contend that it was not until 2004/5 that the Revenue sought to require taxpayers to demonstrate that they had made a distinct break from ties in the UK in order to establish that they had left permanently or indefinitely. Such a change in approach, first applied to these taxpayers after they had left, breached their legitimate expectation that the Revenue would continue, as a matter of practice, to apply IR20 in the manner in which it had been applied consistently in the past, until it publicly announced that it proposed to change its practice for the future.

For the reasons I have given earlier, I accept that the specific assurances that taxpayers will be treated as non-resident in the circumstances identified in IR20 amount to statements of practice, or policy, which the Revenue was obliged as a public law duty to consider and apply, until they promulgated a change of practice for the future. R v IRC ex pUnilever 68 TC 205 concerned an unannounced change of practice which led to refusal of the taxpayer’s claim to set-off of losses on the grounds it was made outside the time limit prescribed by statute. The taxpayer had previously been permitted, for many years, to set off its losses, notwithstanding that its claim was consistently out of time. There is every reason why the obligation of fair dealing in relation to a particular taxpayer, identified in Unilever, should be imposed on the Revenue in the case of settled practices applied to taxpayers at large (see Bingham LJ in MFK 1569C,cited at [13]).

There is, however, an important evidential difference, which has a significant impact on these appeals. In order to establish the existence of a previous settled practice as to how the Revenue was interpreting and applying IR20 these appellants are driven to relying on their advisers’ experience of other cases and published papers on the topic. In Unilever the taxpayer had direct experience of the previous practice and the unannounced change.

Three witnesses on behalf of the Revenue, Susan McLean-Tooke, policy adviser in the Offshore Personal Tax Team of HMRC’s Charity Assets and Residence (CAR) between June 2001 and March 2009, Stephen Symonds, a technical specialist in residence issues between 2003 and 2007, and David West, the Inspector who dealt with the appellants Davies and James, deny that any change of practice has occurred (see e.g. § 9, McLean-Tooke, § 19, Symonds, § 77 West). Since these witnesses were in post at the time it is alleged the practice was changed, there is nothing in the appellants’ contention that they came too late to acquire knowledge of that change.

These witnesses contradict the formidable array of respected specialist advisers in the field of residence, Mr Glyn Davies, chartered accountant and tax adviser since 1981 with PWC and its predecessors, Mr Warburton, Senior Tax Partner of Grant Thornton, David Hilton-Gee, formerly a senior manager in PWC’s tax department, and Peter Vaines, a barrister and member of the Chartered Institute of Taxation, who has advised Gaines-Cooper. All speak of the Revenue’s unannounced change of approach (e.g. Warburton § 14-15, Vaines § 27-29). Mr Davies encapsulates this evidence:

“The point of IR20 is that it is supposed to prevent the sort of thing that has happened in this case from happening: if it does not do that it is not giving general guidance, which is what HMRC say they are doing in the preface, but setting a trap and misleading those who rely on it.

Secondly, although I have been in practice for 25 years and have advised very many taxpayers who have left the UK relying on IR20, I have never known HMRC raise the sort of issues they have raised in this case before now, nor have I seen them adopt so many different arguments, one after another, none of which have any merit.

Lastly, the point which most concerns me is that HMRC now seem to be putting emphasis on the words “leave the UK”, “go abroad” and “gone abroad” which appear, respectively, in the opening sentences of paragraphs 2.2, 2.7 and 2.9 of IR20…I believe that the emphasis on the phrases to which I have referred is novel and represents a most significant change of practice by HMRC which undermines - and I assume is intended to undermine - the whole purpose of IR20. It does not seem to me to put it too highly to say that HMRC’s conduct involves a breach of trust” (1^st statement 39-41).

Inevitably, this conflict of evidence has led the appellants to rely upon a substantial quantity of documents, letters and notes of meetings dealing with a variety of different issues and taxpayers. They can do little else, absent an application to cross-examine (which would probably have been refused in judicial review proceedings which tend to resolve contradictory evidence in favour of a defendant (R (oao Al-Sweady & Others) v SOS for Defence [2009] EWHC 2387)).

This creates difficulties in resurrecting the context and circumstances in which statements were made, not least because, on occasion, they relate to cases in which none of the witnesses have been involved. The Revenue is, of course, under an obligation to the court to do its best to assist by way of disclosure and by adducing full and accurate evidence. But its ability is limited by the fact that there is no central store for files, and weeding takes place after six years. Any picture is bound to be partial and incomplete. There is no reason to believe that the Revenue has not done its best in relation to disclosure.

I should also make clear, at the outset, what I mean by a practice from which the Revenue should not be permitted to resile without prior warning. There is no public law obligation of fairness which prevents the Revenue from increasing, without warning, the intensity of inquiry or scrutiny of claims to be non-resident. That the Revenue, in the past, may have adopted an attitude of laissez-faire, is not and ought not to bea guarantee that it will continue to do so, even without any prior warning of a change in attitude. Indeed, the absence of warning may be a powerful tool to deploy to ensure that taxpayers provide frank disclosure. Particularly, in these days of self-assessment, they do not know if their claims will be targeted for particular scrutiny. The essence of the system of self-assessment is that self-assessments will not all come under enquiry and some will be chosen for investigation at random.

The Revenue’s witnesses accept that claims to be non-resident have become subject to increased scrutiny. Ms McLean-Tooke ascribes this increase to a growth in the numbers of those claiming non-residence and, new or different interpretations advanced by those claimants. She attributes that growth to three factors: the abolition, in 1993, of the rule that any person who visited the UK and had available accommodation would be treated as resident here, the introduction of self-assessment, and thus reliance on the taxpayer’s own assessment as to whether he was resident here or not, and the abolition of “Foreign Earnings Deduction” (her statement § 10-13). Mr Glyn Davies denies that self-assessment led to an increase in inquiry or refusal of claims to be non-resident. It is unnecessary to resolve the dispute. The important point is, I suggest, not whether the Revenue has tightened up its “policing” of IR20 but rather whether it has changed its approach to and application of IR20.

A particular difficulty which has bedevilled this case has been a failure to distinguish between statements and comments which relate to taxpayers leaving for full-time employment abroad within 2.2, and 2.7-2.9. The appellants contend that hitherto the Revenue has made no distinction between those paragraphs in that it has never, before 2004/5, sought to insist upon a distinct break. But the Revenue contends that the need to look for a distinct break has always distinguished 2.7-2.9 from 2.2; under 2.2 a full-time worker abroad has not needed and still does not need to demonstrate any severance of social or family ties. Many of the statements on which the appellants rely plainly refer only to cases of leaving for full-time employment. Indeed, the original trigger for concern that the Revenue had changed its practice stemmed from a question in relation to “mobile workers”, an issue relating to full-time employment and 2.2.

In 1999, the issue of acquisition of the status of non-residence by reason of full-time employment arose in relation to mobile workers. Such workers are, typically, lorry drivers, living in the UK who drive through this country on their way abroad. They spend substantial time travelling abroad and claimed non-resident status on that ground. It became apparent that some inspectors accepted their claims to non-resident status purely on the basis that their day count did not exceed the number prescribed in IR20.

The Revenue took the view that that was erroneous and made public announcements to that effect. It considered that such workers had not, within the meaning of 2.2, left the UK to work full-time abroad. The Revenue made it clear that “a day count” was insufficient to acquire non-residence status. In FICO Newsletter 6, in November or December, 1999, the Revenue explained that their concerns related only to those

“whose job it is to travel abroad (for example, ferry or airline crew)…the individual will continue to be resident in the UK….The fact that they spend days partly in and partly outside the UK does not affect their residence position (as they do not count as days of “arrival” or “departure” in the context of 1.2 of IR20.” (See to similar effect the standard letter dated 20 February 2000, which spoke of a temporary concession to ameliorate the effect of the erroneous advice previously given to mobile workers.)

The Revenue’s approach to mobile workers was reiterated in a Tax Bulletin Article in February 2001 (on the Centre for Non-Residents’ Website) and published in the Tax Bulletin in April 2001. That Article made it clear that it was concerned with:

“mobile workers who usually live in the UK and have not genuinely left this country. Different considerations apply to those who have left the UK to live abroad permanently. Paragraphs 2.7-2.9 of booklet IR20 explain the circumstance in which such individuals may be treated as not resident and not ordinarily resident.” (paragraph 8).

But the Bulletin did refer to 2.2 ( at paragraph 5) :

“The treatment under paragraph 2.2 is aimed at individuals who leave the UK for a complete tax year to live and work on assignments abroad…In the case of individuals living in the UK but making regular short trips abroad, it is questionable whether they have genuinely left the UK in a residence sense, or can be said to be working full time abroad; and they could not satisfythe condition that their absence and the employment abroad lasted a whole tax year. They have not in our view made the clear break with the UK that the practice in paragraph 2.2 requires.” (my emphasis in italics)

These public pronouncements as to mobile workers caused concern amongst the tax advisory profession, particularly in relation to the Revenue’s approach to long distance commuters: those who work full-time abroad during the week but return at week-ends. The reference to “a clear break”, translated by some in the profession as “clean break” triggered particular concern. But those advisers received explicit assurances that the Revenue’s approach to mobile workers would not be applied to such commuters; they would continue to be regarded as coming within 2.2. (See Revenue letter 25 October, 1999, the record of Mr Hilton-Gee’s conversation 8 May 2001, and the response to KPMG’s concern expressed on 12 June 2001, in a letter dated 23 October 2001.)

None of that correspondence seems to me to assist the taxpayers in establishing any change of practice in relation to requiring a distinct break in relation to 2.7-2.9. On the contrary, the assurances given in relation to long-distance commuters, who did not and could not claim to have severed family or social ties in the UK, demonstrate that in relation to 2.2, then as now, no such distinct break with pre-existing social and family ties was required. The reference in the Tax Bulletin to a clear break was no more than a reference to the inference to be drawn from full-time employment abroad for at least a year, namely that that of itself established a clear break sufficient to establish non-resident status. The documents on which the taxpayers rely reveal to my mind a consistent approach by the Revenue to 2.2, which has been maintained in this hearing, namely that social and family ties did not need to be severed.

But it did not follow that a taxpayer could establish non-residence under 2.7-2.9 without such a distinct cut. None of the assurances to which I have referred suggested to the contrary. There is no basis for eliding the assurances given by the Revenue in relation to 2.2 with its approach under 2.7-2.9. Those who were concerned in relation to long-distance commuters and others employed full-time abroad were reassured that their clients did not have to demonstrate a distinct break from social and family ties in the UK. In response to the published statements concerning mobile workers, they did not ask for nor did they receive any assurance relating to the factors which the Revenue would regard as relevant to a claim under 2.7-2.9. Indeed, that difference in approach appears to have been recognised, at least by one established tax adviser at the time.

On 22 June 2001, the Revenue and representatives of the accountants Arthur Andersen met, at the accountants’ request, to discuss their concerns as to the Tax Bulletin. They received the firm reassurance to which I have referred that the Bulletin was solely concerned with mobile workers, who had not “genuinely left” the UK whereas those who had clearly left for employment abroad and had accommodation abroad would be treated as non-resident. The meeting records the Revenue’s view that:

“It was certainly not the intention that individuals going to work abroad could never qualify as not resident within the terms of paragraph 2.2 unless they severed every link with the UK.”

The meeting goes on to record Arthur Andersen’s acknowledgement of the change due to self-assessment, which had caused residence rulings, and exchanges between the profession and the Revenue to cease. This, they recognised, had had the effect that practitioners saw but few cases where the Revenue challenged an individual’s self-assessment as non-resident. The meeting records that the Revenue had set up CNR, later CAR, to concentrate expertise in one body and that it had concluded that the Revenue should publish more of their thinking and there should be more contact between the Revenue, representative bodies and practitioners.

There then follows an important paragraph expressing Arthur Andersen’s comments on the residence rules:

“They recognised the problems of deciding whether someone had ‘left’ the UK, but apart from that they found paragraph 2.2…straightforward. If an individual had full-time employment abroad, it was not necessary to look at the wider factors in paragraph 2.7 about personal circumstances such as accommodation, family life etc.” (my emphasis)

That passage demonstrates unequivocal recognition of the distinction the Revenue and IR20 draw between 2.2 and 2.7-2.9. Even Mr Glyn Davies has, on occasion, accepted the relevance of links retained with the UK:

“…what IR20 does (according to the understanding I have always had as a practitioner) is to set out certain factors which will be taken into account. Some of these factors relate to the quality of the links which the taxpayer has with another country (e.g. full-time employment for at least a whole tax year, settled purpose)…. And some of the factors relate to the extent of the links retained by the taxpayer with the UK (e.g. the number of days spent here, retaining a property in the UK). It follows from that that HMRC have set out their view of the quality of the links with another country and the extent of the remaining links with the UK which should together be taken into account in determining whether someone has ceased to be resident.” ( my emphasis.) (2^nd statement §15)

I do not believe the Revenue would quarrel with any of that understanding; they may argue only with the following sentence :

“The quality of the links with the other country are relevant insofar as they help to determine the extent to which the taxpayer has removed himself from the UK”.

They do help, but are not determinative, since the extent of ties retained is also relevant.

The record of the meeting of 22^nd June 2001 closes with the Revenue’s acceptance that changes to the residence rules were a matter for Ministers. That recognition has significance in relation to the provision of what was described in the meeting as a draft list of indicators. Such a list was produced (see letter from Revenue Policy International to Law Society July 2001). The list of indicators related to 2.2 and included as a factor suggesting “non-resident treatment potentially met”: “‘settled domestic life’ appears to be abroad” and, as an indicator pointing to the opposite treatment, “‘settled domestic life’ appears to be in the UK”.

That proposal, and it is important to stress it was only a proposal for consultation, led to concern and opposition to the suggestion that under 2.2 settled domestic life was relevant (see e.g. a letter from Unilever 3 August 2001). After all, the proposal was contrary to the assurances previously given in relation to long-distance commuters, returning at week-ends.

The most important response came from the largest body of taxpayers’ advisers, the Tax Faculty of the Institute of Chartered Accountants. It is plain that their concern and opposition to the proposed indicators was focussed on those who claimed non-resident status because of full-time employment abroad. After all, it was the Revenue’s proposal in relation to 2.2 which had triggered that concern, as the opening of the Institute’s comments makes clear. The Institute was concerned at the reference to what it described as a “clear” break at paragraph 5 of the Bulletin, and referred to long-distance commuters (§10). It commented that a “UK lifestyle was irrelevant”, but made it clear that its comments related to those in full-time employment abroad (1^st bullet point § 11). The specific comments on the indicators and particularly “settled domestic life” all relate to the application of the proposals to those in full-time employment abroad (see in particular 29-31).

The Institute protested that the changes proposed would undermine the credibility of IR20. It is of particular significance to these appeals that those proposals were never adopted. The fact that the Revenue appreciated that its proposals should be announced, promulgated for consultation and were, in the absence of approval by Ministers, never adopted is important: it is a powerful demonstration that it is most unlikely that the Revenue would have adopted a clandestine change of interpretation and application of IR20. When the Revenue suggested a change, the history shows quite the opposite.

Moreover, there is no justification for extrapolating from the Revenue’s proposals and from the opposition to them relating to 2.2, any conclusion in relation to the factors relevant to leaving permanently or indefinitely under 2.7-2.9. The Revenue had made no proposals as to changes under those paragraphs; nor had the Institute commented upon them.

Accordingly, I reject any suggestion that the discussions stemming from the Revenue’s approach to mobile workers affords any assistance whatever to the appellants’ case as to an unannounced change of interpretation and application of IR20. On the contrary, with the exception to which I shall now come, the documents show that the Revenue’s approach to IR20 did not change and was consistent with the interpretation for which it had contended and which I have accepted.

The one exception in the documents disclosed by the Revenue is contained in a letter dated 7 July 1999 from an Inspector (Residence), Brian Wilks, to the international tax manager, Mr Sawyer, of Wilfred T Fry Ltd. This is a letter which explicitly deals with the Revenue’s approach to 2.9. Since the appellants place understandable reliance on this letter I should quote from it:

“...I’m writing to confirm the way we approach the residence status of individuals who leave the UK for purported permanent residence but who cannot produce the sort of evidence mentioned in paragraph 2.9 of IR20.

Subject only to the caveat that the following guidance is general and particular cases will always need to be decided on their own specific facts, I can say that provided such an individual

•

lives outside the UK for 3 years or more from the date of departure, and

•

after departure has not visited the UK for as much as 183 complete days in any one tax year or 91 or more days a year on average

then we will, after the three years has elapsed, accept the claim to have become not resident and not ordinarily resident.

Specifically, circumstances such as

•

The spouse and/or children having continued to live in the UK

•

A residence having been maintained here

•

Duties having continued to be performed in the UK

will not prejudice the claim to non-residence”.

This letter, without doubt, supports the appellants’ case that the Revenue in 1999 was not insisting on severance of ties with the UK in cases under 2.7-2.9. Ms McLean-Tooke says that the letter was not as clearly expressed as it should have been (§ 16). That will not do. It was clear, and she should have acknowledged, that it contradicts the Revenue’s case.

It seems to me, however, that Mr Wilks’s letter does not bear the weight which the appellants seek to place on it. In subsequent correspondence in relation to the FICO newsletter and mobile workers he draws a clear distinction between such workers, who continue to “live” in the UK and those going abroad to work who fall within 2.2. He reiterates that there has been no change to the practices at 2.2 and concludes:

“that (in relation to) those who are not (or who are no longer living) in the UK the other tests in IR20 will continue to apply” (letter to Mr Sawyer 8 March 2000).

Those observations show that Mr Wilks did regard “living” as relevant to the claims of those who had not gone abroad to work; it contradicts his previous letter.

Additionally, that letter must be read in the context of the later documentation, all of which shows that a distinction was drawn in relation to those who claimed non-residence by reason of full-time employment abroad and other workers who “lived” in the UK. Mobile workers could only acquire non-resident status by ceasing to live in the UK (see the reference to lorry drivers living in Sweden in paragraph 5 of the Tax Bulletin). The underlying feature of that distinction must be that apart from those in full-time employment abroad, others seeking non-resident status must cease to live in the UK.

I conclude that the documentation does not establish that the Revenue, prior to consideration of the cases of these appellants, took the same approach to 2.2 as it took to 2.7-2.9. It does not establish that the Revenue declined to seek a distinct break from social and family ties in the cases of those who sought to come within 2.7-2.9.

It is in that context that the exchanges between the appellants, Davies and James and the inspector, Mr West, must be considered. On 20 December 2004, in a letter recording discussions at a previous meeting, PWC acknowledged that the Revenue did not accept that their clients were in full-time employment in Belgium from 1^st April 2001, took the point that that did not matter, and claimed non-residence on the alternative bases of absence for three years or more or going abroad for a settled purpose in accordance with 2.7-2.9. In his response of 17 January 2005, Mr West rejected the claim on that basis because the taxpayers had maintained “family, business, financial and other links with the UK” which “strongly suggest no permanent or indefinite move”. On 10 March 2005 Mr Glyn Davies protested that that approach amounted to a misconstruction of 2.7-2.9. He wrote:

“The basis for this dismissal appears to be the heading ‘Leaving the UK permanently or indefinitely’, but as I have explained above, I feel that the paragraphs must explain the heading rather than being limited by the heading. If what you say is correct, then there is no meaning to paragraph 2.9, and probably little meaning to paragraph 2.8. I do not think that the words ‘leaving the UK’ make the difference, because the words ‘leave the UK’ are also used at the beginning of paragraph 2.2.”

It is at that point that it seems to me that Mr West’s responses become unclear. On the same day, he replied by e-mail, repeating his request for information as to when full-time employment in Belgium started. He continued :

“A more fundamental question might be ‘When did your clients leave the UK?’ It is this issue of ‘leaving the UK’ or ‘going abroad’ that causes the most controversy in that we are talking here of leaving the UK in the residence/ordinary residence sense. To illustrate this point:

IR20 Para 2.2

If you leave the UK to work full time abroad…you are treated as NR/NOR if…

An obvious test here is whether or not the employment is full time. And to a certain extent we have become bogged down at this test. But the first test is whether or not your clients have ‘left the UK’. This is the fundamental R/OR test and again, on the basis of the evidence thus far, it is not at all clear that your clients have passed this test. Equally I am not suggesting that they have ‘failed’ any test.”

He said he would discuss the matter with Mr Steve Symonds and let Mr Davies have a substantive response.

Mr Glyn Davies’s response of 11 March 2005 asserted that his clients had in fact left in March 2001 “in order to establish a completely different life style and new activity in Belgium”. Mr West replied on 14 March accepting that Mr Glyn Davies had identified the key issue. But he rejected the assertion of a completely different life style, referring to the retention of residential property, some business links, and links to Swansea RUFC and the Area Health Authority.

On 14 March 2005, Mr West responded more fully to Mr Davies’s letter of 10 March, acknowledging that the Revenue is bound to follow the practices outlined in IR20 and that if the taxpayers’ circumstances place them within 2.2 or 2.7 etc the Revenue would accept the non-resident claim. But he rejected Mr Davies’s interpretation of those paragraphs and referred again to the strong links retained within the UK. He continues :

“If your clients have not ‘left’ the UK, they cannot be said to have left indefinitely. In my view detailed discussion on...2.7-2.9 simply confuses matters and obscures the fundamental issues.

The key question for me is whether your clients can reasonably claim to have left the UK to work full time abroad (IR202.2) without strictly having ‘left’ the UK for R/OR purposes. I will ask Steve Symonds, as technical lead here, for his views. Your general comments about areas of guidance possibly having very little meaning might be apposite here”.

I have emphasised the final sentence because Mr Goldberg contended that that amounted to an astonishing suggestion. Certainly it involved some acceptance of Mr Glyn Davies’s earlier disparaging comments on the Revenue’s interpretation of 2.7-2.9, although it was not consistent with his earlier acceptance that the Revenue was bound to follow the practices outlined in IR20.

The exchanges meandered on. But on 28 February 2006 Mr West did make clear that he was rejecting the claim both under 2.2 on the grounds that employment was not full-time from March 2001 and under 2.7-2.9. He repeated those distinct grounds for rejecting the claim on 20 March 2006. He reverted to his position that the taxpayers had retained all their ties and connections in the UK. His position culminated in the decision letter of 29 November 2006 dismissing the claim to non-residence both on the grounds that employment was not full-time and that the taxpayers had not left permanently.

Mr Goldberg’s criticisms of lack of clarity are justified. But such criticism does not carry his clients home. Detailed scrutiny and analysis of lengthy exchanges between taxpayer and inspector will often reveal ebb and flow and cross-purposes and misunderstanding. But this painstaking process adopted by Mr Goldberg cannot establish a breach of the Revenue’s public law obligations of fairness unless it demonstrates a breach of the legitimate expectations founded on the Revenue’s previous interpretation and application of the terms of IR20.

I accept that Mr West appeared in the e-mail of 10 March 2005 and the letter dated 14 March 2005, during the course of this lengthy process, to have failed properly to distinguish between the factors relevant to 2.2 and 2.7-2.9. He may have suggested that in both cases the taxpayers had to demonstrate a distinct break from social and family ties. But whether or not that is a fair reading, by the time of the decision in 2006, no such confusion is apparent. By his letters of 20 February 2006, and 20 March 2006, Mr West made it sufficiently clear that the rejection of the claim under 2.2 was because there was no full-time employment from the claimed date of leaving, whereas, in relation to the alternative case, under 2.7-9, he did not accept that the taxpayers had cut their ties with the UK. That that distinction was made and that the factors relevant to 2.2 were not confused with those relevant to 2.7-2.9 is further confirmed in Mr Steve Symonds’ internal note dated 10 November 2006 which gives the reasons for seeking approval of a determination of ordinary residence for 2001/2 (see p.2). It demonstrates that the Revenue did not reject the claims to non-residence on the grounds of full-time employment abroad because of a failure to cut social and family ties.

The determination was not reached as a result of any misinterpretation of 2.2. Nor does the correspondence read as a whole show that the Revenue rejected the claim to having gone abroad permanently or indefinitely as a result of a changed approach to 2.7-2.9.

Those conclusions are sufficient to dispose of the appeals of Davies and James. This court does not have to and should not decide whether the claims to full-time employment or permanent or indefinite residence are correct, as a matter of fact. It will be for them to establish before the First Tier Tribunal that the view the Revenue has taken of their factual circumstances is incorrect. In their case, the Revenue did not, in my view, renege upon the assurances given by IR20 or misinterpret its terms or alter its interpretation or application of IR20 to defeat the taxpayers’ claim to non-resident status.

As I have already recalled, Mr. Gaines-Cooper lost his appeal to the Special Commissioners (Decision dated 31 October 2006), in respect of the years of assessment 1993/4-2003/4 and has not appealed against that decision. Thus he can only succeed in his application for judicial review, if he can make good his contention that the Revenue has, in 2.7-2.9 IR20 given an assurance, which, if honoured and applied to the facts of his case, would result in his being treated as not resident and not ordinarily resident or that it made an unannounced change to its interpretation and application of those provisions, which it cannot apply retrospectively to his case.

Mr Gaines-Cooper contends that, in 1976, he left the UK permanently or for at least three years, and for the settled purpose of establishing himself as an international businessman. He was absent for at least a whole tax year and his return visits never exceeded the number of days specified in IR20. Further, he informed the Revenue when he left in 1976, but, apart from questions in 1980 as to his return visits no further queries were raised. In particular, it was not suggested that he had failed to establish a distinct break from the UK in the years in question.

The insuperable difficulty, as I see it, for Mr Gaines-Cooper lies in the findings of fact made by the Special Commissioners. They made a number of findings of fact which were relevant to the issue before them as to whether he had abandoned his domicile of origin. Even though those facts were relevant to that issue, they remain findings which are relevant to the question whether Mr Gaines-Cooper ever severed his social and family ties with the UK.

Of course, if the Revenue is not entitled to insist that the taxpayer demonstrate a distinct break, those facts are irrelevant. But for the reasons I have given, the interpretation of IR20, which the Revenue applied, was correct and did not result in Mr Gaines-Cooper attaining non-resident status. He needed to establish a distinct break from social and family ties and the Revenue asserted, and maintains its assertion that he did not make that break either in 1976, when he claims to have left permanently, or thereafter. Nor, for the reasons I have already given, has he succeeded in establishing an unannounced change of practice in its interpretation and application of the provisions on which he relied.

The Special Commissioners found that England remained the “centre of gravity of his life and interests” [142] and that his chief residence was in the UK [142] and [145]. He had, as they put it “so many ties” in Berkshire and Oxfordshire [146]. From the start of the years in issue, 1993/4, he had a settled abode in Henley-on-Thames, where he “dwelt permanently”, and he spent more time in the UK than the Seychelles or anywhere else [166].

Accordingly, there were ample findings of fact to justify the Special Commissioners’ conclusion that he was resident and ordinarily resident in the UK, cited by Lloyd Jones J at [33] to [34]. There is no need to rely on his counsel Mr Flesch QC’s apparent acceptance of that conclusion up to 1993/4, because until that time, available accommodation in the UK was fatal to any contrary finding.

There was discussion before this court, as there had been before the Special Commissioners as to whether, if the taxpayer had left permanently earlier, he had lost his non-resident status by virtue of section 3 of IR20 or s.336 ICTA. Since the Commissioners had found he had not lost his resident status in the UK (they said s.334 and not s.336 applied [170]), there is no need to consider whether he is correct in contending that he never lost his non-resident status or to consider his argument that what he called “return visits” did not exceed the prescribed limit. That controversy elicited the somewhat exhausting evidence of 150 days of flying, in one year.

My conclusion that the taxpayer would not have achieved non-resident status, even if he were in the same position as Davies and James, and the facts of his case had not been assessed by the Special Commissioners, leaves otiose the interesting debate as to the legal effect of their determination on Mr Gaines-Cooper’s claim that his legitimate expectation has been breached. That issue would only have arisen if IR20 ought to have been applied, as he contends, without regard to the question whether he had broken his social and family ties with the UK.

But I ought to record that there was a dispute between Mr Gaines-Cooper and the Revenue as to whether he had been compelled to take his case to the Special Commissioners before launching his claim to judicial review. The Revenue contended that he could have adopted the same stance as his co-appellants and insisted on the judicial review application first. I need not and shall not resolve that issue.

I ought also to record my view that the debated conflict between the view of this court in Davies v James [2008] EWCA Civ 933 and of Blake J in R (oao Lower Mill Estate Limited and Conservation Builders) v HMRC EWHC 2409 is probably more theoretical than real. As I have already recalled, Hughes LJ thought that a determination of the Special Commissioners might pre-empt the judicial review claim [19] and Keene LJ thought there was a real risk that such a claim would, as he put it, be “ruled out” [24].

Blake J described his perplexity, since:

“the whole doctrine of legitimate expectation is of benefit, and only real value, where, on a true understanding of the facts and the law, the taxpayer is, or may well be, liable to tax. Despite that, however, it would be oppressive or unjust...to require him…to pay the tax because the conduct of the tax authorities, in the exercise of their management powers, has legitimately created the belief that tax would not be payable for a particular period, for a particular reason”. [22].

He later qualified his application of his understanding of legitimate expectation to the cases of Davies and James, on the basis that there may have been no express representation as to how the Revenue would treat their particular cases by way of what he described as a “genuine pre-assessment” [28].

As I have endeavoured to explain, the very terms of IR20 themselves often do not contain any binding assurance as to how the Revenue will treat a particular claim. The tests of full-time employment and permanent or indefinite absence will often require value judgements, which may respectably give rise to differences between taxpayers and the Revenue. If the Revenue, on the facts of a particular case do make it clear that it is accepting a claim to non-resident status, it, generally, will not be permitted to resile from that assurance, merely because the statutory fact-finding body might have reached a different factual conclusion. But though the three situations described in Section 2 may be described as specific, they themselves give considerable scope for different conclusions to be reached and may, often, give little comfort to a taxpayer as to the attitude the Revenue will take to his particular claim to non-resident status.

It is in that sense that the facts will prove all important. When this court, in 2008, feared for Davies and James’ prospects, if the judicial review proceedings were to follow the Special Commissioners’ determination, I think the real danger lay in the risk of adverse findings of fact, which would prevent the taxpayer in the judicial review proceedings asserting to the contrary. The danger did not lie in any conflict between the Revenue’s duty to collect tax and the exercise of its management power. In a field, such as residence, where so much depends on facts and their evaluation, it is difficult to see how such a conflict could arise. If the Revenue chooses, on its evaluation of the facts, to treat a taxpayer as non-resident, applying IR20, without any appeal, then that does not conflict with its statutory obligations.

These appeals fail, in my view, not because of the jurisprudence in relation to residence, not because IR20 can be disregarded by the Revenue but because on a proper interpretation of that statement of practice, the taxpayers fell outwith the circumstances which would have gained them non-resident status. The Revenue has not been shown to have altered its interpretation and application of IR20 to these appellants’ cases. Accordingly, I would refuse their applications for judicial review.

I agree that the appeal should be dismissed for the reasons given in the judgment of Lord Justice Moses.

Throughout the hearing of this appeal I have had and continue to feel some sympathy for the appellants and I have entertained more doubt about the result of the appeal than my Lords betray.

I am not troubled by the Revenue’s construction of paragraph 2.2 of IR 20. The first condition for its application is that “absence from the UK and … employment abroad lasts for at least a whole tax year”. It is, therefore, perfectly reasonable for the Revenue to construe that condition to mean that the tax payers should be in actual employment at the beginning of the tax year and that an intention, albeit it firmly held at the beginning of the tax year, to take up employment during the year is not good enough if the employment only commence later in the year. Mr Goldberg Q.C. virtually concedes that by having resorted to the submission that “a whole tax year” does not mean “the whole tax year” and can mean a followingtax year. That argument simply does not run. The plain and ordinary meaning of those words is that the tax payer is to be treated as non-resident and not ordinarily resident if in the year in which he claims to have that status he not only has been absent from the United Kingdom but has also been in employment abroad for at least the whole of the year in which he claims exemption.

My difficulty has been with clauses 2.7 to 2.9. As the heading suggests, one can leave the United Kingdom permanently or one can leave the United Kingdom indefinitely. The concepts are different. The appellants fall into the latter category. Leaving indefinitely anticipates or at least does not exclude a return at some time in the future. It does not on the face of it require a complete cutting of ties with the United Kingdom. Nor does the concept of ordinary residence as it has been developed by the courts, the most recent authoritative exposition of which is contained in the speech of Lord Scarman in Reg. v Barnett L.B.C., Ex p. Shah [1983] 2 A.C. 309 applying the tax cases of Levene v Inland Revenue Commissioners [1928] A.C. 217 and Inland Revenue Commissioners v Lysaght [1928] A.C. 234. In Levene Viscount Cave L.C. said at p. 225:

“I think that [ordinary residence] connotes residence in a place with some degree of continuity and apart from accidental or temporary absences.”

Lord Warrington of Clyffe said at p. 232:

“′Ordinary residence′ also seems to me to have no such technical or special meaning [for the purposes of the Income Tax Act]. In particular it is in my opinion impossible to restrict its connotation to its duration. A member of this House may well be said to be ordinarily resident in London during the Parliamentary session and in the country during the recess. If it has any definite meaning I should say it means according to the way in which a man’s life is usually ordered.”

In Lysaght Viscount Sumner said at p. 243:

“I should think the converse to ′ordinarily′ is ′extraordinarily′ and that part of the regular order of a man’s life, adopted voluntarily and for settled purposes is not ′extraordinary′.”

In Shah Lord Scarman said at p. 342:

“I note that in the nineteenth century bankruptcy case In Re: Norris (1888) 4 T.L.R. 452 it was accepted that one person could be ordinarily resident in two countries at the same time. This is, I have no doubt, a significant feature of the word’s ordinary meaning for it is an important factor distinguishing ordinary residence from domicile.”

Thus he concluded at p. 343:

“Unless, therefore, it can be shown that the statutory framework or the legal context in which the words are used requires a different meaning, I unhesitatingly subscribe to the view that “ordinarily residence” refers to a man’s abode in a particular place or country which he has adopted voluntarily and for settled purposes as part of the regular order of his life for the time being, whether of short or of long duration.”

He added at p. 344:

“And there must be a degree of settled purpose. The purpose may be one; or there may be several. It may be specific or general. All that the law requires is that there is a settled purpose. This is not to say that the “propositus” intends to stay where he is indefinitely; indeed his purpose, while settled, may be for a limited period. Education, business or profession, employment, health, family or merely love of the place spring to mind as common reasons for the choice of a regular abode. And there may well be many others. All that is necessary is that the purpose of living where one does has a sufficient degree of continuity to be properly described as settled.”

I recite from those cases because the guidance in I.R. 20 rightly claims to follow the law. The language of paragraph 2.9 (“You have gone abroad for a settled purpose (this would include a fixed object or intention in which you are going to be engaged for an extended period of time), you will be treated as non-resident and not ordinarily resident …” is consistent with and an application of the principles of ordinary residence espoused in those cases. The appellants would seem to me to have an unarguably strong case for claiming to be ordinarily resident abroad.

That, however, is not the end of it. The appellants also have to show that they are not resident in the United Kingdom. Here Viscount Cave L.C.’s words in Levne at p. 233 present a problem. He said:

“But a man may reside in more than one place. Just as a man may have two homes – one in London and one in the country – so he may have a home abroad and a home in the United Kingdom and in that case is held to reside in both places and to be chargeable with tax in this country.”

Thus it seems to me the Revenue were in this regard fully entitled to look for a clear break – or a clean break - with this country. Even if the appellants have established an ordinary residence abroad, the fact that they return to sleep in their own beds in their own homes on a fairly regular basis and retain a string of connections with the life they led in this country, then the Revenue are entitled to question their claim to be non-resident here.

In this admittedly round-about way, I have reached the conclusion that the Revenue are entitled to construe paragraphs 2.7 to 2.9, even 2.9 standing alone, in such a way as to require the tax payer to show that he has severed social, domestic and family ties in the United Kingdom.

As for the suggestion that the Revenue has changed its construction in unannounced changes of policy, I agree with my Lords. I readily understand the appellants’ suspicions. The surprise of a large body of professional opinion at the recent turn of events causes me concern that there has been a change in policy. I am, however, persuaded that the change which has been perceived by the profession in the assessment of taxpayers’ claims to be treated as not resident and not ordinarily resident in the United Kingdom is the effect of a closer and more rigorous scrutiny and policing of the growing number of claims, which it is permissible for the Revenue to conduct and is not a root and branch change in policy.

Not without considerable hesitation I too would dismiss the appeals of Mr Davies and Mr James.

In the case of Mr Gaines-Cooper, I agree that the findings of fact are against him and that his appeal must also be dismissed.