Curtilage revisited (yet again)

One of the nightmares a lawyer has is that he or she has overlooked some relevant judicial authority – a judgment that has somehow escaped their notice, and which could potentially undermine some cleverly constructed argument that they have carefully devised. The same dilemma faces legal authors (including bloggers), and so in our seminar last month, I had a sudden frisson of doubt about the vexed issue of “curtilage”, the definition of which has repeatedly troubled the courts over the years.

What prompted my sudden anxiety was the citation by Alistair Mills, in his excellent survey of recent case law, of the judgment in Challenge Fencing Ltd v SSHCLG [2019] EWHC 553 Admin. This was a judgment that was handed down in March 2019, but which I had somehow missed until Alistair Mills drew attention to it in our seminar.

It turns out that this judgment does not change the accepted view of how “curtilage” is to be defined, but it did prompt me to revisit the issue of “curtilage” in light of the various judicial authorities on the subject that go back over the past 70 years or so. It is, I would suggest, a good idea to carry out this exercise occasionally, rather than repeating time and again the same old summary that one has been accustomed to trotting out whenever the subject arises.

One point that immediately became apparent to me from a re-reading of the cases is that very little attention seems to have been paid over the years to what I have always regarded as the locus classicus on this topic, the judgment in Sinclair Lockhart’s Trustees v Central Land Board (1950) 1 P.& C.R. 195. So far as I am aware, this judgment has never been over-ruled or distinguished, and so it remains good law.

The essential point that I derive from that case is that, in order to be within the curtilage of a building, the land in question must not only have a close spatial relationship with the building (a point to which the courts have repeatedly returned) but it must also share a functional relationship with the building. To quote the relevant passage in the judgment:

“The ground which is used for the comfortable enjoyment of a house or other building may be regarded in law as being within the curtilage of that house or building and thereby as an integral part of the same although it has not been marked off or enclosed in any way. It is enough that it serves the purpose of the house or building in some necessary or useful way.”

Although the Sinclair Lockhart judgment was not referred to in R (Egerton) v Taunton Deane BC [2008] EWHC 2752 (Admin), it is clear from the characteristically thorough and carefully reasoned judgment of Sullivan J (as he then was) that he certainly had in mind the need for a functional relationship between the alleged curtilage building and the principal building, as well as a spatial relationship. This was a strand of thought that ran through the entire judgment (which has to be read in its entirety to understand the stress laid on this point). It may be briefly illustrated by this passage:

“…..In 1984 Mill Barn and the Old Granary were not being used for purposes that were ancillary to the use of the farmhouse as a dwelling house; they were being used for the purposes of the general farming enterprise which was being carried on at Jews Farm. They, and the agricultural buildings to the west, were being used for agricultural purposes. They were not being used, for example, to garage the farmer's car, to store his domestic items, as a children's playroom, staff quarters etc cetera.”

In Egerton, Sullivan J held that, both in relation to their spatial relationship with the listed farmhouse (being separated from it by a wall) and in the lack of any functional (domestic) relationship with the farmhouse, the nearby barns were not within the curtilage of the farmhouse, and so they were not listed buildings. It is clear, therefore, that for one building to be within the curtilage of another building, their relationship must be both functional and spatial.

So far as the spatial relationship is concerned, a number of points have been established by the various judicial authorities over the years. In Dyer v Dorset County Council [1989] 1 Q.B. 346, the court referred to the definition of the word “curtilage” in the OED:

“a small court, yard, garth or piece of ground attached to a dwellinghouse, and forming one enclosure with it, or so regarded by the law; the area attached to and containing a dwellinghouse and its outbuildings”

The reference in the dictionary definition to the curtilage being “small” was held to be irrelevant by the Court of Appeal in Skerritts of Nottingham Ltd v SSETR (No.2) [2000] 2 P.L.R. 102; [2000] J.P.L. 1025, but Dyer is still authority for the proposition that (irrespective of its size) land forming the curtilage of a building must be ‘attached’ to a dwellinghouse (and not be separated from it by any intervening land) and must form one enclosure with it, a point affirmed by the High Court in Burford v SSCLG [2017] EWHC 1493 (Admin). (Practical examples of the application of this rule were provided by Collins v SSE [1989] P.L.R. 30 and McAlpine v SSE [1994] E.G.C.S. 189. )

One aspect of the decided cases with which I take issue is the repeated references to the judgment of the Court of Appeal in A-G (ex rel. Sutcliffe) v Calderdale BC (1982) 46 P. & C. R. 399, where it had been held that there were three factors to be considered when identifying the curtilage of a (listed) building. These were: (1) the physical ‘layout’ of the listed building and the structure, (2) their ownership, past and present, and (3) their use or function, past and present.

Stephenson LJ had said in that case:

“Where they are in common ownership and one is used in connection with the other, there is little difficulty in putting a structure near a building or even some distance from it into its curtilage. So when the terrace was built, and the mill was worked by those who occupied the cottages, and the millowner owned the cottages, it would have been hard, if not impossible, to decide that the cottages were outside the curtilage of the mill.”

It seems to me, however, that reliance should no longer be placed on that judgment. First, the buildings in that case may not (at the relevant time) have been in one and the same planning unit. The well-known rule in Burdle established that the unit of occupation is the appropriate planning unit to consider, until or unless a smaller unit is identified which is in separate use, both physically and functionally. That alone suggests that by the time that the mill building in Calderdale came to be listed, the adjoining terrace of millworkers cottages cannot have been within the same planning unit as the mill itself. Although the cottages were clearly built and were originally (and perhaps still) owned by the millowner, they were separately occupied, whether by millworkers or others. Even if they constituted ‘tied accommodation’, they clearly constituted separate units of occupation, and (even though they may have been attached, at least at one end of the terrace, to the mill building), they were in separate use, both physically and also functionally, as dwellings.

Even if they continued to be occupied by millworkers as tied accommodation, the terrace could not now be regarded as ‘ancillary’ to the mill building, bearing in mind the House of Lords decision in Westminster City Council v British Waterways Board [1985] A.C. 676; [1984] 3 All E.R. 737 that a use in one planning unit cannot be ancillary to an activity carried on in a different planning unit, even an adjoining one.

So far as physical attachment to another building is concerned, it is clear from section 1(5) of the Listed Buildings Act that any structure fixed to a listed building forms part of the listed building. However, the House of Lords in Debenhams plc v Westminster LBC [1987] A.C. 396 insisted that a structure fixed to a listed building would itself be listed only if it was subordinate or ancillary to the building that was actually listed. An obvious example (cited in Debenhams) is a terrace of houses; the listing of just one of the houses in the terrace clearly does not apply to the houses on either side if these are in separate ownership or occupation, even though they are structures that are ‘fixed to a listed building’. I would submit that this applies equally to the situation in the Calderdale case, and that the terrace of millworkers cottages cannot, for the reasons explained above, be regarded as having been included in the listing of the mill building (in the absence of their having been specifically included in the listing description).

For the reasons explained above, I would respectfully disagree with Lieven J’s formulation in Challenge Fencing of the tests that are to be applied in assessing whether or not a building or an area of land falls within the curtilage of a particular building (as summarised in paragraph 18 of her judgment), at least in so far as they depend, to a degree, on Stephenson LJ’s observations in the Calderdale case.

The definition of what constitutes “curtilage” is (and, in the absence of any general legislative definition, must remain) a matter for the courts. However, I entirely agree that the actual identification in a particular case of the extent of the curtilage of a building is a question of fact and degree, and so this must be a matter solely for the decision-maker, subject to normal principles of public law. This, however, depends on the correct application of the definition of “curtilage” (as established by the judicial authorities referred to above).

Reference was made in Challenge Fencing to the judgment in Sumption v Greenwich LBC [2007] EWHC 2776 (Admin). That case established that the identification of the curtilage must be determined in relation to the factual position at the relevant time (for example, as at the date of first listing in the case of a listed building, or as at the date of the application in the case of an LDC application); it also established that the extent of the curtilage can very easily be changed from time to time, subject of course to any addition to the curtilage being incorporated in the same planning unit as the building within whose curtilage it is claimed to have been included, and also to its use at that time being lawful in planning terms.

So I would formulate the relevant indicators as follows:

(1) The identification of the curtilage of the building in question is a matter of fact and degree, to be determined by the decision-maker.

(2) The determination is to be made by reference to the relevant point of time in the particular case (e.g. the date of first listing in the case of a listed building, or the date of the application in the case of an LDC application).

(3) In order to be within the curtilage of Building A, the relevant land and/or other building(s) must, at the relevant time, have been within one and the same planning unit as Building A.
(a) This question is to be determined in accordance with the rule in Burdle, i.e. What was the unit of occupation at the relevant time?
(b) At that time, were the alleged curtilage land and/or other buildings in the same use as Building A, without having been divided off from it, or were they in separate use, both physically and functionally?
(c) Past or present ownership (as distinct from occupation) is unlikely to be of assistance in relation to this question, which will depend on the actual occupation and use of the various parts at the relevant point in time in that particular case.
(d) The historic layout or use of the land and buildings is unlikely to be of any assistance in answering this question; it is actual occupation and use at the relevant time that will be the determining factors.

The following questions will only fall to be answered if it has been determined in answer to Question 3 that the alleged curtilage land or buildings were at the relevant time in one and the same planning unit as Building A.

(4) At that time, were the alleged curtilage land and/or building(s) used for the comfortable enjoyment of Building A? Did they serve the purpose of Building A in some necessary or useful way? (For example, if Building A is/was a dwelling, were the other building(s) and/or land also in use for domestic purposes in connection with the use and occupation of Building A as a dwelling?)

(5) Were the alleged curtilage land and/or building(s) attached (in spatial terms) to Building A, and did the land and/or building(s) form one enclosure with Building A? If not, were the relevant land and/or buildings divided from Building A by a wall, fence or other means of enclosure or were the relevant land and/or buildings separated from Building A by any intervening land (e.g. uncultivated ground, rough grass, pasture, etc.)?

(6) In a case where Building A was, or became at that time, a listed building, was any building within the alleged curtilage physically attached to Building A? If so, was it subordinate or ancillary to Building A? (The relative sizes of Building A and any attached building may be a factor in answering this question, as well as the relative functional relationship of the respective conjoined buildings.)

(7) Finally, whilst the claimed curtilage need not necessarily be ‘small’, is its size and alleged necessary or useful function in relation to Building A proportionate to the size and function of Building A?

It will be seen from these proposed questions that the three factors listed by Stephenson LJ in the Calderdale case really don’t come into the equation (at least, not in the form in which they were stated in that judgment).

There are, of course other judgments on this issue that have not been cited above, such as Methuen Campbell v Walters 1979 1 QB 525.and Lowe v FSS [2003] EWHC 537 (Admin), but they don’t really add anything to the authorities already cited.

© MARTIN H GOODALL