England and Wales Court of Appeal (Civil Division) Decisions

IN THE SUPREME COURT OF JUDICATURE
COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM THE HIGH COURT OF JUSTICE
CHANCERY DIVISION
(Ferris J)

		Royal Courts of Justice Strand, London, WC2A 2LL Friday 27th July 2001

MR JUSTICE CARNWATH:

Introduction

1. The taxpayers in these three appeals were employed during the years of assessment on one or other of two similar "jack-up rigs" (mobile oil-drilling rigs) operating in the North Sea, named respectively "Santa Fe Magellan" and "Santa Fe Monarch". The short issue is whether their earnings were "emoluments from employment as a seafarer", for the purpose of para 3(2A) of Schedule 12 of the Income and Corporation Taxes Act 1988. If they were, then they could take advantage of more generous provisions in relation to exemption from income tax for work performed abroad. The detail of these somewhat complicated provisions is not important for present purposes. In very broad summary, their effect is that, in the normal case, interruptions of the periods of employment abroad must not exceed 62 days or one sixth of the total number of days in the period; in the case of a "seafarer" the period of interruption is increased to 183 days or one half of the period.
2. The expression "employment as a seafarer" is defined as:

"employment consisting of the performance of duties on a ship (or of such duties and others incidental to them)".

3. The word "ship" is not defined in the 1988 Act. However, before the Commissioners and the judge, reliance was placed on the definition in the Merchant Shipping Act 1894:

"ship includes every description of vessel used in navigation not propelled by oars."

4. The 1894 Act definition, which replaced an identical definition in the 1854 Act, was still current in 1988, when the relevant provisions were introduced. Modifications made for certain purposes in the interim (e.g. by Merchant Shipping Act 1921 s 1) do not materially affect the argument. The definition has since been replaced by the Merchant Shipping Act 1995, s 313(3).
5. We are concerned with three separate decisions of General Commissioners, two from Norwich and the other from Great Yarmouth, all decided against the Revenue. The two Norwich decisions related to the same taxpayer, Mr Perks, who worked as a senior drilling foreman on one rig in 1991-2 and the other in 1992-3. The Great Yarmouth decision related to two taxpayers, Mr Newrick and Mr Grainger, who were "rig workers" on one or other of the rigs in 1992-3. The facts and reasoning in that case, which was the latest of the three, were stated somewhat more fully than in the others. However, it is accepted before us that no distinction should be drawn between the facts of the different cases.
6. We were also referred to another case involving a similar "jack-up rig", the Glomar Adriatic 9, which was decided by the Special Commissioners in December 1999. In that case, the Commissioners arrived at the same ultimate conclusion, but the evidence and arguments were not identical to the present cases. That decision is itself subject to appeal, but no arrangements have been made to include it with the present group of appeals.

The Commissioners' findings

7. In all three cases the Commissioners held that the rigs were "ships", and that, accordingly, the taxpayer's earnings were from employment as "seafarers", within the relevant definition.
8. The primary facts as found by the Norwich Commissioners were identical in each case (save for the name of the rig), as follows:

"(a) That the appellant was employed on the jack-up oil rig Santa Fe Magellan during the tax year in question as a senior drilling foreman.

(b) That the jack-up rig Santa Fe Magellan had a floating hull and retractable legs. From the profile drawing and photograph of the Santa Fe Magellan on station the Commissioners were satisfied that when the jack-up was stationary for the purposes of drilling the legs would be down so that the feet would stand upon the sea bed and the hull would be jacked up so that it was clear of the water. From the photograph of the Santa Fe Magellan under tow the Commissioners were satisfied that the legs could be retracted upwards to enable the floating hull to be towed from place to place.

(c) That the jack-up rig Santa Fe Magellan has no propulsive thrusters or engines of its own.

(d) The Santa Fe Magellan was moved by being towed by at least two tugs. A tow master controlled the towing operation, particularly as to the speed and direction of the tugs, giving instructions to the tugs from the bridge of the jack-up rig.

(e) An alternative method of transporting the jack-up rig is for it to be carried "piggy-back" style upon a cargo vessel."

9. They noted the submissions of the Inspector, which included the following:

(b) As there was no definition of "ship" in ICTA 1988, and as yet no case law on the subject, Revenue relies on the Merchant Shipping Act 1894.

(c) In deciding whether a jack-up is a "ship" the test to be applied is as follows:

i Is the jack-up capable of navigation?

ii Is the jack-up actually used in navigation?

iii Is the jack-up capable of free and ordered movement upon the sea?

iv It is contended that a jack-up fails each of these tests

10. It appears that the taxpayer did not dispute the tests, but contended that on the facts the jack-up rig satisfied them. He submitted also that the official documentation required for a jack-up rig confirmed its status as a ship.
11. The Commissioners recorded that they had been referred to certain cases (to which I will come below) and concluded as follows:

"The Commissioners decided that the jack-up rig Santa Fe Magellan was a ship for the purposes of the Taxes Acts as it was capable of navigation, the motive power being the engines of the towing tugs with all aspects of navigation being controlled from on board the rig".

12. In the second Norwich case, the Commissioners' decision was expressed in the same words, but they added:

"The Commissioners listened to arguments from both sides regarding the definition of a ship but the most compelling evidence was the definition contained in IR CA 2451 and CT 1060 defining reliefs to be given in certain circumstances."

13. The two documents there referred to were extracts from guidance notes prepared by the Revenue for inspectors, which were listed by the Commissioners as two of the documents "admitted in evidence". I will return to the content of those guidance notes a little later.
14. In the Great Yarmouth case, the Commissioners had somewhat fuller evidence and made further findings of fact. As finding (i), they accepted as proved the facts in a statement of written submissions by the taxpayer. They also found:

"(ii) That the jack-up rigs Santa Fe Monarch and Santa Fe Magellan were sister rigs with no rudder or motive power identical save for minor variations in specification;

(iii) that the said rigs were built in Singapore and transported to Rotterdam by sea to be used in the North Sea;

(iv) that the said rigs were towed from Rotterdam across the North Sea to the oil fields in the British Sector;

(v) that from time to time the said rigs were moved by being towed to different positions in the North Sea and then jacked up for drilling purposes;

(vi) that in all towing operations at sea the navigation of both the rig and the towing vessel was under the command of a towing master on the rig itself on which was installed a satellite navigation system known as Global Positioning System as well as a compass and depth sounder;

(vii) that the said rigs were (a) registered as ships with the Panama Merchant Marine Registry (b) flagged (c) marked with a load line (d) had a registered tonnage (e) had a radio call sign (f) carried lights as required by international regulations (g) complied with safety at sea regulations (h) carried certified coxswains (i) were fitted with the required radio equipment and global marine distress and safety equipment operated by a certified radio operator (j) were required to comply with the international convention for the prevention of pollution and (k) were subject to annual surveys as ships;

(viii) that both the said rigs were treated as ships by (a) port authorities (b) customs authorities (c) the Department of Transport and (d) the Health and Safety Executive."

15. That statement of written submissions referred by way of introduction to some previous decisions of the Revenue and Commissioners in relation to different types of drilling vessel. It contained the following additional information about the method of navigation:

"Jack-up vessels are capable of being used in navigation, with a few exceptions they are towed around by an Anchor Handler/Tug/Supply vessel on a towing bridle or moved on long voyages by transport vessels. When towed, a GPS (Global Positioning system) is set up on the Jack-up Vessel. The position of the vessel is shown on a digital display and the movement of the vessel on a VDU screen. When under tow, all control of the operation and navigation is under the command of the Towing Master. When transported on the deck of a transporter vessel the navigation equipment on the transporter vessel is used to give a position. All navigation is done by electronic means."

16. The statement also set out a list of the various regulatory controls applicable to ships, which were summarised in findings (vii) and (viii) above.
17. The Inspector had submitted a bundle of documents, one of which was an extract from Halsbury's Laws (4th Ed. Vol 43 (1) para 102), relating to the meaning of "ship" for the purpose of the Merchant Shipping Acts. I shall return to that extract. The contentions of the Inspector, in summary, were that the rigs were not ships because a ship must be "capable of free and ordered movement under its own power"; and further that any movement of the rigs was "incidental to their real purpose which was oil exploration". The Commissioners expressed their conclusions as follows:

(i) We accepted as an accurate summary of the law on the definition of a ship the said paragraph in Halsbury's Laws of England

(ii) On the facts we found that both the Santa Fe Monarch and the Santa Fe Magellan were vessels used in navigation at sea and were therefore ships within the meaning of the Merchant Shipping Acts and for the purposes of the Income and Corporation Taxes Act 1988

(iii) We rejected the contentions of Mr. Ash on behalf of the Inspector of Taxes that it was necessary to the definition of a ship that a vessel should be able to navigate under its own power and that the said vessels were not constructed for navigation.

The Judge's decision

18. The appeal from the Commissioners was on a point of law only. The Judge explained his approach as follows:

"I approach these appeals on the basis that both sets of the findings are applicable to both the structures with which I am concerned. To the extent that these findings represent findings of primary fact they are binding upon me. The meaning of the term 'ship' is, however, a matter of law on which I am entitled to give effect to my own view, as is the question of whether the facts found bring the case within that meaning." ([2000] STC p483C).

(I shall return to that passage, because the appellant's primary contention before us is that that approach is wrong in law.)

19. The Judge commented that, in addition to the facts recorded in the decisions, he had been shown various agreed documents and had also found particularly helpful the illustration of the rigs and their specifications and profiles. He then noted, on the one hand, the absence of any definition of the term "ship" in the 1988 Act itself, and, on the other, the definitions in successive Merchant Shipping Acts. He reminded himself of the dangers in having resort to the Merchant Shipping Act definition, not least because it was not expressed as an exhaustive definition. He observed:

"The correct approach must be, in my view, to examine the authorities for indications of the characteristics which, in the particular context, were considered to be relevant to the decision whether a particular structure is or is not a ship and then to consider whether those characteristics are relevant in the context of para 3(2A) of Schedule 2 and, if they are, whether they are possessed by the vessels with I am concerned."

20. He then reviewed in detail the authorities (to which I shall return), and referred to the conclusions of the Commissioners. Three particular aspects of their reasoning attracted his criticism. The first was that the Norwich Commissioners had in part relied on the fact that the motive power was the engines of the tugs used for towing. The Judge commented:

"This suggests that the Commissioners have looked not just to the rigs themselves, which is what they ought to have considered, but to the composite item consisting of the rig and the tugs which are moving it"

21. However, he accepted that, according to the authorities, absence of a self-contained means of propulsion was not a crucial factor. In any event, I think he may have been reading too much into the Commissioners' observation. As I understand it, they were simply noting, as was a fact, that the motive power came from the tugs, without thereby implying that they regarded the tugs as part of the subject matter of their consideration.
22. The other criticisms related to the use made of the Revenue's guidance, and of information relating to compliance with regulatory requirements. I shall return to these points when considering the challenges made to the Judge's decision.
23. The Judge concluded his review of the authorities by observing that it was impossible to give a general definition of the term "ship":

"Even the Merchant Shipping Act does not purport to do this, for its definition is not exhaustive. Insofar as any one characteristic may be said to be required for something to be a ship it is, I think, that of being capable of being used, and in fact used, for the purpose of navigation. Beyond this the exercise must, I think, be to examine the particular features of the vessel in question and to reach a conclusion whether such a vessel is properly and reasonably described as a 'ship'".

24. He then set out the features which he regarded as relevant as follows:

"(1) Each rig has a hull which makes it capable of floating and moving upon water.

(2) The rigs do not have their own means of propulsion. However this, as the barge cases show, is a matter of comparatively minor significance.

(3) The rigs do not have rudders or other means of controlling their direction of travel. For this they are dependent, as I understand it, upon the giving of instructions to independent towing vessels by the tow-master who is on board while they are being moved.

(4) Although the rigs have quite sophisticated positioning equipment which can be, and presumably is, used in connection with the movement of the rigs from place to place, the main purpose of this equipment is to enable the rigs to be positioned accurately in the place where they are required to drill into the sea bed.

(5) It cannot realistically be said that the function of the rigs is 'navigation' in the sense of conveying persons or cargo from place to place by water. No doubt personnel, who may be called a crew, are present on board when the rigs are moved from place to place on the sea and stores and equipment are carried. But these are incidental to the real function of the rigs.

(6) That function is to provide a firm base at particular points on the earth's surface to enable the drilling part of the rig to drill to great depths under the bed of the sea. Once in position the rigs perform in a similar manner to that in which a drilling rig on land would perform. No part of them is floating. They are supported by their retractable legs which are positioned on the ocean floor. Their hulls are raised some considerable number of feet above the surface of the water and their entire weight is carried by the sea bed rather than by the water. They differ from land based drilling rigs only in that they are moved from place to place on water and, while drilling, they stand in and are surrounded by water."

25. He observed:

"While some of these features are possessed by vessels or constructions which may fairly be described as 'ships' it seems to me that most of them are not and that the preponderance lies on the side of those features which are not characteristic of ships. I attach particular significance to those which I have mentioned under (5) and (6), taken in conjunction with the limited manoeuvrability of the rigs."

26. He referred to Polpen Shipping Co Ltd v Commercial Union Assurance Co Ltd [1943] 1 KB 161, where Atkinson J had said that a ship is a –

"…hollow structure intended to be used in navigation, i.e. intended to do its real work on the seas or other waters, and capable of free and ordered movement thereon from one place to another…"

27. He concluded:

"Correspondingly, the work of a jack-up drilling rig is to be positioned at a point on the surface of the earth, there to perform the static function of drilling into the earth's crust. Its ability to float and to be moved from place to place by means of outside assistance (not even under its own power and steerage as in the case of flying boat) is merely incidental to its static, non-floating, work. Overall it does not appear to me that the rigs with which I am concerned have sufficient characteristics of ships to lead to the conclusion that each of them is a ship for the purposes of para 3(2A) of Sch 12 to the 1988 Act. I hold that they are not ships for that purpose."

28. Finally, he dealt with a point raised by the Revenue, based on statements in Hansard. That point is no longer relied on, and I need say nothing more about it.

Fact or Law?

29. The appellants' primary ground for challenging the Judge's conclusion is that he was wrong to proceed on the basis that the question was one of law, on which he was entitled to substitute his own view of the facts for that of the Commissioners. Mr Davey submits that the conclusion that a particular structure is a "ship" is an inference of fact drawn from the primary facts. The judge could only interfere with the Commissioners' conclusion if it was one which could not reasonably have been reached on the basis of the primary facts as found (see Edwards v Bairstow [1956] AC 14, 36 per Lord Radcliffe).
30. The application of that principle may vary depending on the nature of the word or phrase being interpreted. In Edwards v Bairstow itself, the expression was "adventure in the nature of trade". Such an expression may need some judicial interpretation, to identify "the characteristics which distinguish such an adventure" as a matter of law (ibid p31, per Viscount Simonds). Other words are used in an ordinary sense which needs no judicial interpretation. Thus, in Cozens v Brutus [1973] AC 854, the expression "insulting….behaviour" (in the Public Order Act 1936 s5) was a pure jury question. As Lord Reid said (at p861):

"The meaning of an ordinary word in the English language is not a question of law. The proper construction of a statute is a question of law. If the context shows that a word is used in an unusual sense the court will determine in other words what that unusual sense is. But here there is in my opinion no question of the word 'insulting' being used in any unusual sense. It appears to me… to be intended to have its ordinary meaning. It is for the tribunal which decides the case to consider, not as law, but as fact, whether in the whole circumstances the words of the statute do or do not as a matter of ordinary usage of the English language cover or apply to the facts which have been proved…"

31. A useful summary of the effect of these cases is in Ransom v Higgs [1974] 1 WLR, 1594, 1618C-E, per Lord Simon of Glaisdale:

"The meaning of a word or phrase in an Act of Parliament is a question of law not fact; even though the law may then declare that the word or phrase has no statutory meaning beyond its common acceptance and that it is a question of fact whether the circumstances fall within such meaning (Cozens v Brutus). But many words and phrases in English have many shades of meaning and are capable of embracing a great diversity of circumstance. So the interpretation of the language of an Act of Parliament often involves declaring that certain conduct must as a matter of law fall within the statutory language (as was the actual decision in Edwards v Bairstow); that other conduct must as a matter of law fall outside the statutory language; but that whether yet a third category of conduct falls within the statutory language or outside it depends on the evaluation of such conduct by the tribunal of fact. This last question is often appropriately described as one of 'fact and degree'."

32. That case, like Edwards v Bairstow, concerned the meaning (in a taxing statute) of the word "trade". A similar summary, in relation to the word "plant" in the Taxing Acts, is found in IRC v Scottish and Newcastle Breweries [1982] 1WLR 322, 327, per Lord Lowry,

"(1) It is a question of law what meaning is to be given to the word "plant", and it is for the courts to interpret its meaning, having regard to the context in which it occurs. (2) The law does not supply a definition of plant or prescribe a detailed or exhaustive set of rules for application to any particular set of circumstances, and there are cases which, on the facts found, are capable of decision either way. (3) A decision in such a case is a decision on a question of fact and degree and cannot be upset as being erroneous in point of law unless the Commissioners show by some reason they give or statement they make in the case stated that they have misunderstood or mis-applied the law in some relevant particular. (4) The Commissioners err in point of law when they make a finding which there is no evidence to support. (5) The Commissioners may also err by reaching a conclusion which is inconsistent with the facts which they have found."

33. Applying those principles to the present case, the word "ship" is as ordinary an English word as one could imagine. I will come in a moment to consider the relevance of the analogies with the definitions in the Merchant Shipping Acts and other legislation. However, whether the proper view of the law is that it is to be treated as an ordinary English word simpliciter, or is to be given some more refined or expanded meaning, the application of that meaning to the facts of the particular case is a question of fact, not law. The decision is in the province of the Commissioners, not of the courts. That is so, whether one is speaking of the findings of primary fact or of the inferences to be drawn from those facts (see Furniss v Dawson [1984] 1AC 474, 527-8 per Lord Brightman).
34. On this question accordingly I think Mr Davey is right, with respect, to criticise the learned judge's approach. Once the meaning of the term "ship" had been established, the question whether the facts found brought the case within that meaning was not a question of law (as he identified it - see p348C), but a question of fact, subject only to the conclusion falling within the limits of reasonableness as defined by Edwards v Bairstow.
35. It is fair to add that the relevant caselaw on this question does not seem to have been examined in much detail before the judge. Indeed, he may have been misled by one of the cases relied on before him, Addison v Denholm Ship Management (UK) Ltd [1997] ICR 770 (a decision of the Employment Appeal Tribunal in Scotland). The issue in that case was whether mobile offshore accommodation-vessels, known as "flotels", were "ships" for the purpose of the Transfer of Undertakings (Protection of Employment) Regulations 1981. I shall return later to the substantive decision in that case. In the present context, the point was taken that this was a question of fact, excluded from the jurisdiction of the appeal tribunal. Having quoted the well-known judgment of Sir John Donaldson MR in O'Kelly v Trust House Forte plc [1983] ICR 728, 761, the tribunal said this:

"…it is our opinion that, while the determination of whether a set of facts applies to a particular object is essentially for the tribunal of fact, and where the issue is determined upon the selection of certain facts which may either be competing or diverse again is solely within the province of the tribunal of fact, the application of facts found proved or agreed to achieve a definition on any statutory basis is, in our opinion, a question of law. The question, therefore of status or categorisation of a flotel is one upon which this appeal tribunal has jurisdiction to consider as a matter of law, against a given state of facts agreed or found proved by the Industrial Tribunal." (p 781 F-G).

36. I do not, with respect, find that passage easy to interpret. However, insofar as it appears to be saying that the evaluation of the facts found against the statutory definition is a question of law, it is in my view inconsistent with the principles outlined above.
37. On the first issue, therefore, the appellant is correct. The judge's decision cannot be upheld on the basis on which he reached it. It remains to consider whether, nonetheless, the Commissioners' decisions were erroneous in law, as the Revenue contends. Before us, Mr Brennan put the Revenue's case in two ways: (1) that the Commissioners were wrong in law to base their conclusions on cases relating to the definitions in the Merchant Shipping Acts, and other contexts, when there is no such definition in the Taxes Act; (2) that, even using that definition, they failed to apply the correct test emerging from the cases (which he refers to as the "real work" test), or failed to reach a conclusion reasonably open to them on that test.
38. The former submission comes as something of a surprise, not only because it was common ground before the Commissioners that the Merchant Shipping Act definition, and the cases relating to it, were of relevance; but also because, as will be seen, the Revenue's own internal guidance specifically adopts that definition. It will be convenient, therefore, to start by dealing with second issue, accepting for the time being the basis on which the matter was argued before the Commissioners; before going on to consider whether it is open to the Revenue to adopt a different line on appeal and, if so, what are the consequences.

The "real work" test

39. Mr Brennan (in the words of his skeleton argument) submits that the proper approach is to consider "what is the real purpose and use of the structure, taken by itself. What is it designed to do? What is its real work?" Like Ferris J, he derives that test from the judgment of Atkinson J in Polpen Shipping Company Ltd v Commercial Union Assurance [1943] 1KB 161, 167. The facts of that case were very different. The question was whether a flying boat was a "ship or vessel" within the meaning of an insurance policy. Having explained (in the passage noted by Ferris J – see above) that a "ship or vessel" must be "intended to do its real work on the seas…", he said:

"A flying boat's real work is to fly. It is constructed for that purpose, and its ability to float and navigate short distances is merely incidental to that work." (p 167)

40. Adopting that formulation, Mr Brennan submits that the "real work" of the jack-up rig is its stationary drilling function. Its mobility is purely incidental to that function.
41. As I have said, the judge had reviewed in detail a series of judicial decisions on the term "ship", mainly in the context of the Merchant Shipping Acts. It is unnecessary to repeat that review. Their general effect is fairly summarised in the extract from Halsbury's Laws (4^th Ed vol 43), which was adopted by the Great Yarmouth Commissioners. That reads as follows:

"102. Meaning of 'ship'. Unless the context otherwise requires, 'ship' includes every description of vessel used in navigation. Whether a vessel comes within that meaning of a ship depends on the facts of each case; the statutory definition is intended to enlarge the meaning of 'ship'. To be a ship a vessel must be used in navigable waters, either inland or at sea, and, although she must be constructed for navigation, it is not necessary to the definition that she should be able to navigate under her own power. The presence of a rudder and the manning of the vessel with a crew are important as showing that a vessel is a ship, but the absence of either does not mean that a vessel is not a ship. The purpose for which a vessel has been and is being used is also material when considering whether she is used in navigation."

42. The first sentence is a direct reference to the definition in the Merchant Shipping Act 1995 section 313(1). The other references include many of the cases referred to before us, including the decision of the House of Lords in Wells v The Owners of the Gas Float Whitton No 2 [1897] AC 337.
43. As that passage shows, the key to the interpretation of that definition in the cases has been the reference to "use in navigation". The test has enabled the courts to include within the scope of the definition structures of very specialised kinds, for example:

(1) A "Hopper-barge", used in connection with dredging (The Mac (1882) 7PD 126: The Mudlark [1911] PD 116):

"She is used for dredging purposes and has a double bottom. She has no means of propulsion, but her owners employ a tug to take her to and from her destination when carrying dredged materials to sea for the purpose of discharging them." (ibid at p 117)

(2) A "Flotel" (in the Addison case, above):

"They comprise a platform attached by legs or columns to pontoons which enable them to float on water. The platform carries accommodation for several hundred workers; it has offices, workshops and storage areas. It is equipped with large cranes and helicopter landing facilities, and is capable of supplying power to the installation. A flotel is taken to the location of the installation, possibly under its own power, but usually under tow." ([1977] ICR at p774C).

(3) A "backhoe dredger" (the "Von Rocks" [1998] 2 Lloyd's Rep 198, a decision of the Supreme Court of Ireland):

"a type of maritime dredger called a backhoe dredger which was primarily used in harbours, channels or estuaries to deepen the waters at such locations. When not in operation it was a floating platform comprising ten individual pontoons bolted together. When in use it was held in position on the sea bed by three spud legs which were capable of being hydraulically lowered and raised. The backhoe dredger had no bow, no stern, no anchors, no rudder or any means of steering and no keel or skeg. It had no means of self propulsion mechanical or otherwise and it had no wheelhouse." (headnote at p198).

This case was not cited to the learned judge. Although it is not binding on this court, its correctness was not questioned before us.

44. These examples seem to me to falsify Mr Brennan's main submission. In each of these cases it could have been said that the "real work" of the vessel or structure was something other than navigation, and that its mobility was purely incidental to its main function. However, that was not the approach of the courts. Thus in the Irish case, Keene J said this:

"The Von Rocks undoubtedly lacks some of the characteristics one would normally associate with a 'ship'. It is not self propelled, it normally is not manned by a crew and it has no form of rudder or other steering mechanism. But it is a structure designed and constructed for the purpose of carrying out specific activities on the water, is capable of movement across water and in fact spends significant periods of time moving across the seas from one contracting site to another." (p208)

45. By contrast, the cases which have been held to fall on the other side of the line have been ones in which "navigation", in the sense of "moving across the seas", was minimal or non-existent. Thus, the following have been held not to be "ships":

(1) A "gas float" moored in tidal waters to give light to vessels (Wells v The Owners of the Gas Float Whitton No 2 [1897] AC 337):

"It is used for purposes connected with navigation in the same sense as a lighthouse, or as a buoy, whether used as a beacon or for mooring a ship; but it appears to me to be wholly unfit for the purpose of being navigated as a vessel, and that it never was used, or intended to be used, for any such purpose." (p348 per Lord Watson).

(2) A "pontoon crane" (see Merchants Marine Insurance Co Ltd v North of England Protecting and Indemnity Association (1926) 25 Ll.L.Rep 446; (1926) 26 Ll.L.Rep 201):

"A structure upon which a crane is fixed, and permanently fixed. …..It is undoubtedly capable of being moved, but it is obviously so unseaworthy that it can only be moved short distances, or comparatively short distances, and only when the weather is exactly favourable…. I think there are about five or six times when it has been moved since 1914…" (p202 per Bankes LJ).

(3) A flying boat (the Polpen Shipping case - see above). The flying boat -

"did not travel on the surface of the water, except when taking off or alighting." ([1943] 1KB at 162).

46. The Merchants Marine case is of particular interest, because, notwithstanding the very specialised nature of the structure and the limited movement involved, it was not regarded as a clear-cut case. The judge (Roche J) accepted that there might be floating cranes which would be "ships or vessels" within the definition (25 Ll.L.Rep p447). In the Court of Appeal, Scrutton LJ would say no more than that he was "not convinced that the learned judge below was wrong", and he considered it impossible to lay down a satisfactory definition ((1926) 26 Ll.L.Rep p203).
47. Those examples show that, so long as "navigation" is a significant part of the function of the structure in question, the mere fact that it is incidental to some more specialised function, such as dredging or the provision of accommodation, does not take it outside the definition. There may be an issue of degree as to the significance of the navigation on the facts of a particular case, but that, as the observations of Scrutton LJ show, is a question for the fact-finding tribunal. Those examples also show that "navigation" does not necessarily connote anything more than "movement across water"; the function of "conveying persons and cargo from place to place" (in the judge's words) is not an essential characteristic.
48. In the present cases, it was clear, as the Commissioners found, that the rigs were capable of, and used for, navigation, in the sense described in the cases. There does not seem to have been much discussion of the extent of actual movement. There was some evidence of the voyage details. For example the Santa Fe Magellan, having arrived in Rotterdam by transport from Singapore in October 1992, seems to have been moved four times between Rotterdam and locations in the North Sea between November 1992 and April 1994. However, Mr Brennan did not suggest that the relative infrequency of movement was in itself a reason for challenging the Commissioners' decisions. He realistically accepted that in most cases the categorisation of a structure, as a ship or not, should be governed by its design and capability, rather than its actual use at any time.
49. Accordingly, in my view, he has failed to show any error of law in the decision of the Commissioners. For the same reasons, the judge was wrong, in my view, to interfere with their decision. It is fair to add, however, that, if he had been referred to the Irish case of The Von Rocks, which on the facts is very similar to the present cases, he might well have reached a different conclusion.
50. Finally, I should refer to the two specific criticisms made by the learned judge. The first related to the use made by the second Norwich panel of the Revenue guidance. Somewhat surprisingly they seem to have been referred to guidance on the meaning of "ship" for the purposes of capital allowances (CA 2451) and capital gains (CG 1060), but not to the similar guidance, included in the papers before us, relating to the foreign earnings deduction (SC 4530). However, the effect of the guidance is essentially the same in each. Inspectors are invited to approach the matter, in the absence of a definition of "ship" in the relevant Act, on the basis that the word has the same meaning as in the Merchant Shipping Act 1894. Specifically it is suggested that oil rigs and platforms are not generally regarded as ships "since they do not normally move about and are not 'used in navigation'". By contrast, "semi-submersible oil rigs and similar vessels in the oil and gas industry may be ships, depending on the use made of them."
51. The judge fairly criticised the Norwich panel for referring to these notes as "evidence". Equally, they would have been wrong to treat these statements as in any way definitive of the law. However, there was no reason why they should not have been received by them, in effect as part of the submissions for the Inspector, with which they appear to have been wholly consistent. Thus the judge's criticism does not in my view undermine the Commissioners' decision in that case, and in any event the reference was not repeated in the other two cases.
52. The other point of criticism made by the judge relates to the material summarised in the findings of the Great Yarmouth panel, relating to the compliance with various regulatory requirements for ships, derived from the appellant's written statement. This contained a comprehensive summary of the regulatory requirements, which have to be satisfied for a "ship", and which are also applied to the jack-up rigs. In my view the Commissioners were entitled to take account of these as part of the background, but there is no indication that they regarded them as in any way definitive. As Mr Davey submits, they are if anything of negative significance, in that their absence might have been seen as a factor undermining the suggestion that these jack-up rigs are treated as ships for practical purposes.
53. Finally I should note a criticism which Mr Davey makes of the judge's conclusion relating to what he called the "sophisticated positioning equipment". This I take to be the "global positioning equipment" referred to in the evidence. The judge suggested that, while no doubt it could be used in connection with movement from place to place, its main purpose was to enable the legs to be positioned accurately for drilling. Although Mr Davey says there was no evidence to support this view, it seems a not unreasonable inference. Be that as it may, it was not one of the features picked out as having special significance by the judge, and it does not seem to me to add to the case either way.
Relevance of Merchant Shipping Act and other definitions.
54. Finally I come to Mr Brennan's submission that the Commissioners erred in law by having regard to cases decided under the definitions in the Merchant Shipping Acts and other related contexts.
55. The first issue is whether this contention is open to him, given that it is directly contrary to the way the case was presented for the Revenue before the Commissioners. It is clearly unsatisfactory, as well as potentially unfair to the taxpayers and the Commissioners, for the Revenue to attempt to change the basis of its case after the hearing. On the other hand, as Mr Brennan rightly submits, it is the Revenue's duty to administer the tax system in accordance with the law. If it becomes apparent to them, during the progress of a case, that a view taken of the law at an earlier stage is wrong, then in principle, he says - subject to proper procedural safeguards and issues of costs - they cannot be prevented from proceeding on the correct basis.
56. That seems to me correct as far as it goes. (For a recent example of the application of that approach, see F and I Services Ltd v Commissioners of Customs and Excise [2001] AllER (D) 289 para 73). However, it raises a difficulty where the change comes after the conclusion of the fact-finding stage of the proceedings. The rules allow service of a respondent's notice, asking the Appeal Court to uphold the Order on different grounds (CPR 52.5 (2)(b)). However, where the appeal is on law only, it is well-established that the parties will not be permitted to raise new points, unless all the necessary findings of fact are included in the case; nor, in normal circumstances, will the Court fill the gap, by use of its power to remit to the Commissioners (see the cases summarised in Yuill v Wilson [1980] 1 WLR 910, 920-1, per Lord Edmund-Davies). This rule applies to the Revenue, no less than to the taxpayer.
57. Mr Brennan submits that the primary facts here are not in dispute. However, except where there is only one conclusion which could properly be reached on the primary facts, the taxpayers are in my view entitled to have the decision of the Commissioners, on the application of those primary facts to the law. In this case, the Commissioners were never asked to reach such a conclusion on the basis of disregarding the Merchant Shipping Act cases. I do not think it would be right for the Court of Appeal now to undertake that task for them.
58. In any event, Mr Brennan has not put before us any material, which satisfies me that the Commissioners would have been likely to reach a different decision on the basis which now he proposes. If one simply looks at the word "ship", as it appears in the Taxes Act, one gets no guidance, except possibly that the reference to "seafarers" directs attention to seagoing vessels, rather than those confined to inland waters. Other guidance would have to be found in dictionaries. The primary definition of "ship" in the Oxford English Dictionary is "a large seagoing vessel". Comparing that with the definition in the Merchant Shipping Acts, one finds the word "vessel" repeated in both. The main points of distinction are that the dictionary definition requires not merely a vessel, but one which is both "large" and "seagoing". Both characteristics were undoubtedly fulfilled by the rigs in this case.

Conclusion

59. In my view the Commissioners reached a conclusion which was open to them on the facts found. The learned judge was wrong to treat the issue as one of law, on which he could substitute his own views for those of the Commissioners. The Commissioners did not err by failing to apply the "real work" test, as advocated by Mr Brennan. In view of the way the case was presented to the Commissioners, it is not open to the Revenue now to challenge their decision on the basis that they wrongly referred to definitions outside the Taxing Acts. Even if it were, it has not been shown that a different result would have been achieved. For all these reasons, I would allow these appeals and restore the decisions of the Commissioners.

LORD JUSTICE LONGMORE:

61. I agree.
62. This case shows that "watertight definitions do not exist even for ships", see H Meijer, The Nationality of Ships (1967) p. 15. On the facts, found by the Commissioners, the oil rigs Santa Fe Magellan and Santa Fe Monarch were used in navigation but navigation was ancillary to their real function of drilling for oil. The Commissioners, applying cases in which the definition of a ship in the Merchant Shipping Act 1894 had been considered, decided that the rigs were used in navigation and were thus ships. Ferris J decided that they were not ships, since navigation was only incidental to their real work and their natural function.
63. Normally one would have considerable reservations about construing a word, which is undefined in the relevant statute, by reference to its definition in another, apparently unconnected, statute. Such reservations would usually increase when one discovers that the other statute defines the relevant word by what it includes rather than by what it is. In this case the Merchant Shipping Act definition is:

"'Ship' includes every description of vessel used in navigation".

64. But ships and shipping have been governed by the Merchant Shipping Acts for so long (1854 to date) and the same definition has, in all essentials, been repeated so often that it would be unsatisfactory for the court (and indeed the Commissioners) to proceed on the basis that the meaning of the word "ship" in a taxing statute should be decided in the abstract without reference to the definition in the Merchant Shipping Acts. Moreover the Revenue's own guidance note expressly assumes that the Merchant Shipping Act definition is to apply.
65. In these circumstances I consider that the Commissioners made no error of law in asking themselves the question whether the oil rigs in question were used in navigation and answering that question in the affirmative. It is true that some of the authorities regard the function and purpose of the structure as important but the critical question is whether the structure is used in navigation. It was open to the Commissioners to find that they were and that must be the end of the matter.
66. It is not part of the function of this court to provide a definition of a ship, watertight or otherwise. It is, however, part of our function to encourage consistency of approach in fact-finding tribunals. Drilling ships and drilling barges must be ships. Semi-submersible oil rigs in which drilling operations are carried out while the rig is in a floating condition, submersible oil rigs in which drilling is carried out when the rig is resting on the sea bed, and jack-up drilling rigs which, when drilling, have legs resting on the sea bed (and are thus not subject to the heaving motion of the sea, in the same way as semi-submersible oil rigs and drilling ships) are all different forms of structure; it could be said that since the jack-up rigs cannot perform their main function without their legs being on the sea bed, they should be singled out and should not be regarded as ships. It would, however, be unsatisfactory if some forms of oil rigs were ships and others were not. One approach should be that all three forms of oil rig should either be ships or not ships. Mr Michael Summerskill observed in his Oil Rigs: Law of Insurance (1979) page 85 that there would be a certain logic in such an approach. I agree and, in any event, I do not consider that the Commissioners' findings or conclusions can be successfully assailed.
67. I would, therefore, allow this appeal.

LORD JUSTICE ROBERT WALKER:

69. I agree with both judgments and I add a few remarks of my own only because we are differing from the judge in a case of some general interest. The court was told that a very large amount of tax is at stake in respect of years of assessment before the relevant provisions were repealed by the Finance Act 1998.
70. The essential issue is whether the Norwich and Great Yarmouth General Commissioners made an error of law in their understanding of the word "ship" which appeared, undefined, in the statutory definition of "employment as a seafarer" in paragraph 3(2A) of Schedule 12 to the Income and Corporation Taxes Act 1988. The correct approach to that issue has been described by Lord Simon of Glaisdale in his speech in Ransom v Higgs [1974] 1WLR 1594, 1618, to which my lord, Carnwath J has referred.
71. The legislative purpose of paragraph 3(2A) was, in general terms, to accord more generous treatment to foreign earnings from employment as a seafarer, as compared with other foreign earnings. But it is impossible to be more precise about the social or economic reason why more generous treatment was expedient for that class of taxpayers. Mr Brennan QC correctly did not renew in this court the application to refer to material from Hansard which the judge had rightly rejected as not complying with Pepper v Hart [1993] AC 593.
72. On any view the legislative purpose cannot have been to accord the favourable treatment only to employees skilled in seamanship, since (as the official notes on foreign earnings deduction remark at SE 4530)

"Employees who are seafarers for this purpose will therefore include not only sailors but anyone whose work is carried out on ships such as cooks, entertainers and couriers on luxury liners."

73. There is no obvious reason why engineers and others doing skilled and demanding work on a jack-up oil rig should not qualify for the same treatment.
ORDER: appeal allowed with costs here and below and determinations of the three sets of General Commissioners restored. (Order does not form part of approved Judgment)