The Controversial Nature of Fencing Easements: Express Grant and Covenants, Lecture notes of Law

Download The Controversial Nature of Fencing Easements: Express Grant and Covenants and more Lecture notes Law in PDF only on Docsity!

Fencing Easements, their true Character and how they arise Churston Golf Club Ltd v Haddock [2019] E.W.C.A. Civ 544 Introduction The recent Court of Appeal case of Churston Golf Club Ltd v Haddock concerned a covenant made between neighbouring freeholders for the upkeep of a boundary fence which had fallen into disrepair. It is of particular interest, however, because the court also had to consider the respondent’s argument that the covenant, contained in a conveyance, had at the same time created a fencing easement. It may be recalled that the fencing easement is a rare, and rather limited, exception to well-established rules that otherwise constrain the application of covenants and easements. Put simply, it is a right in fee simple to require a neighbouring freeholder to maintain a shared boundary fence.^1 The respondent in Churston claimed a fencing easement because the covenant itself was not enforceable on account of the long-standing decision in Austerberry v Oldham BC.^2 Austerberry makes it clear that the burden of a positive covenant cannot be attached to the land so as to be enforceable directly against successors in title of the original covenantor, and the covenant in question here had not been made by the appellant. A fencing easement, however, certainly would have attached to his freehold, also burdening the adjoining land, permitting Mr Haddock to enforce the fencing obligation against the golf club. The possibility that Austerberry might be avoided in this way should be of interest to conveyancers, particularly if the principle could be extended beyond fencing arrangements. There seems no obvious reason why it could not be extended to other boundary features. The difficulty of enforcing liability for the repair of a shared boundary feature directly against a covenantor’s successors was one of the key estate management problems that prompted the Law Commission to produce its 2011 reform proposals. 3 Methods conveyancers use to avoid the problem of Austerberry include the imposition of a restriction on the affected title, selling by way of a long leasehold, and even reliance upon the application of the equitable principle from Halsall v Brizell.^4_._ None of these “workarounds” is without its disadvantages.^5 The Law Commission’s proposals have yet to make it to the statute book, and the problem remains. The fencing easement argument failed in Churston because the court saw no reason to interpret what was a professionally drafted covenant as something else just because it related to a boundary fence.^6 This highlights a significant problem: it seems inevitable that any attempt to create a fencing easement in express terms would also create a covenant, and therefore fall foul of Austerberry in any event.^7 The law of covenants is well-established in comparison with the arguably shaky foundations on which fencing easements rest,^8 and it (^1) See the comments of Patten LJ in Churston Golf Club Ltd v Haddock [2019] EWCA Civ 544 at 10. (^2) (1885) 29 ChD 750. (^3) See the Law Commission’s 2011 report “Making Land Work: Easements, Covenants and Profits à Prendre” Law Com No 327 at para 6.28. (^4) [1957] Ch 169. For an example of the risks of relying solely on this principle see Wilkinson and others v Kerdene Ltd [2013] EWCA Civ 44_._ (^5) For a discussion of these issues, see the Law Commission’s 2011 report “Making Land Work: Easements, Covenants and Profits à Prendre” Law Com No 327, at paras 5.21 to 5.28. (^6) See Patten LJ’s reasoning in Churston in particular at 34_._ (^7) The point is made by Diplock LJ in Jones v Price [1965] 2 QB 618 at 634 and reiterated by Patten LJ in Churston at 22. (^8) Patten LJ makes the point in Churston at 11: “Fencing easements…have a long history but an uncertain legal basis”.

would be a bold decision for any court to ignore Austerberry in favour of what could be considered something of an archaic anomaly. It is significant, however, that the court did not need to rule upon the broader point of whether any form of wording could ever have succeeded, so made no judgment on it. It remains a tantalising possibility, therefore, that an ingenious conveyancer might devise an effective form of wording and so enable the fencing easement to reinvent itself as a useful tool in modern estate planning. As well as the possible implications for conveyancers, Churston is important because of the issues it raises concerning the fundamental nature of proprietary rights in land, and in particular what can and cannot effectively be granted. What distinguishes Churston from all previous reported cases involving fencing easements is the proposal that it is possible to create the interest by express grant in a deed. Fencing easements have never been recognised by the courts other than through long use.^9 Can the benefit of an obligation ever be asserted as a right that “lies in grant” when it is really a covenant? Is it justifiable for the law to make that assertion on the basis of a fictitious historic grant that could never have been made? If a fencing easement genuinely lies in grant, why does it appear to be impossible to grant in practice? Churston offers no definitive answers to these questions other than to confirm that an express grant is in theory possible; the court’s analysis of these issues is nonetheless of importance. The facts, initial decision and the first appeal The facts of the case are reasonably straightforward. Churston Golf Club owned its golf course freehold until 1972 when they sold it to the local authority, occupying it thereafter as the council’s tenants. In the 1972 conveyance the council covenanted to maintain in perpetuity certain boundary fences separating the golf club from neighbouring farmland. That adjoining land was part of a large trust, and the covenant was in favour of the trustees. The clause reads as follows: “The purchaser hereby covenants with the Trustees that that the Purchaser and all those deriving title under it will maintain and forever hereafter keep in good repair at its own expense substantial and sufficient stock-proof boundary walls fences or hedges along all such parts of the land hereby conveyed as are marked T inwards on the plan annexed hereto”.^10 Mr Haddock later acquired the farmland in question as a dairy farm, and, as the fences were no longer stockproof, he sued both the golf club and the council for losses he claimed his business had suffered as a consequence of his animals not being fenced in. He settled with the council, leaving the golf club as sole defendant in the action. The judge at first instance found the golf club liable for two reasons: the clause in the 1972 conveyance had created a fencing easement, but even if it had not, the burden of the covenant the clause contained had passed to the golf club pursuant to s79 of the Law of Property Act 1925. On hearing the first appeal in the Chancery Division, Birss J unsurprisingly rejected the lower court’s finding that s79 might have passed the burden of the positive covenant to the golf club. 11 It is well settled that, absent contrary wording, s79 makes covenantors liable for their (^9) It should be noted that Crow v Wood [1971] 1 QB 77 is an example of the technical grant of a fencing easement by conveyance pursuant to the effect of s62 of the Law of Property Act 1925, but the terms of the arrangement were long established and not expressly created by the conveyance itself. (^10) Churston at 3. (^11) Haddock v Churston Golf Club Ltd [2018] E.W.H.C. 347 (Ch) at 41.

successors’ breaches, if the covenant relates to the relevant land. Arguments that its effect is also to bind those successors to performance have been considered and rejected in a particularly strong line of cases^12 culminating in Rhone v Stephens^13 in which the House of Lords confirmed again that s79 is a word-saving provision only^14 and so cannot circumvent the rule from Austerberry v Oldham Corporation.^15 On the question of whether it was possible that the covenant could have created an easement, Birss J was equally forthright. It could, and it had. The preliminary point had been considered and settled that fencing easements could be created by grant.^16 The question of whether an express covenant might constitute a grant if that was its effect had also been answered in the affirmative by the House of Lords in Russell v Watts.^17 A second appeal, to the Court of Appeal, was allowed under the Civil Procedure Rules because the case involved “an important point of principle or practice”.^18 There were two points to be considered. The first was whether fencing easements could ever be created with express words. If they could, the question then arose whether the drafting had actually succeeded in the case at hand. This second question, however, was not a point of general importance, but the fact that it was “so obviously bound up”^19 with the preliminary issue meant that there was “a compelling reason for the Court of Appeal to hear it”.^20 The decision on the second Appeal The leading judgment in the second appeal is that of Patten LJ, with whom both Baker LJ and Nugee J concurred. Patten LJ begins his judgment with a conclusion. Fencing easements are, he states, “no more than instances where the law will impose on the owner of land an obligation to keep his land fenced for the benefit of the owners or users of the adjoining land.”^21 The rule from Austerberry v Oldham Borough Council^22 is clear that positive obligations cannot burden land. What Patten LJ describes is unquestionably a positive obligation, even if it is not a covenant made in a deed. His subsequent assertion that fencing easements are not easements at all, and that in the judgment he will refer to them as such for convenience only,^23 does little to assist. If not easements, this begs the question of what they are: exceptions to Austerberry , perhaps, or simply sui generis? The judgment proceeds with a useful and (^12) Notably Sefton v Tophams Ltd [1967] 1 A.C. 50 and Federated Homes Ltd v Mill Lodge Properties Ltd (1980) 1 W.L.R. 594. (^13) [1994] 2 A.C. 310. (^14) Rhone v Stephens [1994] 2 A.C. 310 per Lord Templeman at 322. (^15) (1885) 29 ChD 750. Austerberry firmly established the rule that the burden of positive covenants, however proprietary in nature, cannot burden land at law except between landlord and tenant. (^16) Crow v Wood [1971] 1Q.B. 77. (^17) (1885) 10 App. Cas. 590 per Lord Blackburn at 611. (^18) Civil Procedure Rules 1998, rule 57.2(a). (^19) Churston Golf Club Ltd v Haddock [2019] E.W.C.A. Civ. 544 per Patten LJ at 2. (^20) Civil Procedure Rules 1998, rule 57.2(b). (^21) Churston at 10. (^22) (1885) 29 Ch.D. 750. (^23) Churston at 11.

essential review of the three main Court of Appeal cases involving fencing easements: Jones v Price ,^24 Crow v Wood ,^25 and Egerton v Harding.^26 Patten LJ notes that in all three previous cases there had been consensus that fencing easements undoubtedly exist and can arise by common law prescription by virtue of immemorial use.^27 In Jones v Price , however, Winn LJ had thought that a lost modern grant could not possibly be assumed, as any such grant would inevitably have been a positive covenant and “could not prevail against the Austerberry decision”.^28 In the same case Diplock LJ had expressed the firm view that fencing easements do “lie in grant” and that the doctrine of a lost modern grant could therefore apply, but that if an express grant were possible he could not envisage its form.^29 The divergence in judicial opinion in Jones v Price is at the heart of the anomaly, and if Winn LJ was right, that would appear to rule out the possibility of express grant. Patten LJ, however, finds that the matter was resolved in Diplock LJ’s favour because of the subsequent decision of the Court of Appeal in Crow v Wood. In Crow v Wood the court had held that whatever the true origin of a fencing easement, s62 of the Law of Property Act 1925 could convey it and in that sense it could be granted; its validity could not then be questioned on account of an uncertain origin, owing to the curative effect of s62.^30 Having concluded his review of the authorities, Patten LJ leaves hanging the question of whether fencing easements could be created by express grant, and the point remains undecided. He nevertheless now turns to the second question. In finding that the wording used failed to create a fencing easement, he reasons that, even if express grants were possible, this wording could not amount to one. The clause was a covenant to maintain a fence, quite simply, was professionally drafted, and there were no grounds for re-interpreting it as something else. Accordingly it took effect as a positive covenant and, because of the Austerberry decision, would not bind the covenantor’s successors.^31 Critical analysis of the decision The Court of Appeal rejected the opportunity afforded to it by Churston to express a firm view on whether the express grant of a fencing easement is possible. The court was, however, prepared to confirm that they do lie in grant.^32 The grant by deed of something that lies in grant ought, presumably, to be straightforward. That it appears not to be is a contradiction that demands to be resolved. It is submitted that there are two interrelated causes of the paradox: firstly,“grant” is used in two separate senses, and, secondly, the interest itself is only partly “capable of grant” in the orthodox sense of that expression. These have to be unravelled. (^24) [1965] 2 Q.B. 618. (^25) [1971] 1 Q.B. 77. (^26) [1975] Q.B. 62. (^27) Churston at 11 - 19 in which Patten LJ cites Jones v Price [1965] 2 Q.B. 618 at 639 per Diplock LJ and at 636 per Willmer LJ, Crow v Wood [1971] 1 Q.B. 77 at 84 per Lord Denning and Egerton v Harding [1975] Q.B. 62 at 71 per Scarman LJ. (^28) Jones v Price [1965] 2 Q.B. 618 at 638. Winn LJ went on to say: “Only a duty independent of any contract or covenant and existing immemorially could possibly prevail” at 639. (^29) Jones v Price [1965] 2 Q.B. 618 at 634. (^30) Examples of the far-reaching effect of s62 of the Law of Property Act 1925 include Wright v Macadam [1949] 2 K.B. 744 and P&S Platt Ltd v Crouch [2004] 1 P. & C.R. 242. (^31) See Churston at 34. (^32) Churston at 36 approving Crow v Wood [1971] 1 QB 77.

Although Patten LJ treats himself as bound by Crow v Wood^33 on the point that fencing easements lie in grant, that decision must be viewed in its proper context. Crow v Wood decided that a fencing easement could be “granted” to a buyer in a conveyance pursuant to section 62 of the Law of Property Act 1925. S62 “grants” interests in a limited sense only. By conveying what already exists, in certain circumstances it can give the impression of granting proprietary interests afresh, but this is illusory. S62 would convey the benefit of a maintenance covenant of a suitably proprietary kind, if already attached to the land,^34 but it is submitted that this would not mean that covenants “lie in grant”. Crow v Wood is not, with all due respect to Patten LJ, authority for the general proposition that fencing easements lie in grant. The obiter comments of Diplock LJ in Jones v Price are of more general application than those of Denning LJ in Crow v Wood , but still leave the problem unresolved. In his judgment he is adamant that fencing easements “lie in grant”, but gives no explanation of what this really means. The assertion seems to be founded on a recognition that the benefit can become enforceable through long use, which is usually relevant only to genuine easements. It also looks probable that there is a conflation between long use and the concept of the lost modern grant. It is submitted that the appropriate sense of “grant” is that found in the judgment of Danckwerts J in Re Ellenborough Park :^35 “…capable of forming the subject matter of a grant”.^36 The original paradox therefore remains, because an obligation to fence a boundary is not something that is capable per se of forming the subject matter of a grant, being a promise to perform rather than an interest in land. The sense in which fencing easements do assume a more proprietary character is revealed by a comparison of the two distinct situations in which they arise. The first is where a common is surrounded by numerous farms, all of which put their livestock on it. Here, a custom is likely to arise that each farmer accepts responsibility for fencing their own boundary to the common. Whilst this is partly a matter of convenience, it being impractical for all involved to enter into formal arrangements with each other, it is also needed because no individual freeholder has the right to erect a fence on the common itself or indeed on anyone else’s land. Each farmer, however, has an interest in both protecting their land from damage by livestock and avoiding liability for the escape of their own animals. Where the custom is established, all the freeholders acquire both benefit and burden of a system of mutual fencing easements. The resemblance to a true easement comes not from the servient owner’s duty to fence, but from the defence it affords the dominant owner against cattle trespass^37 should his beasts find their way onto a neighbour’s farm, as had occurred in Egerton v Harding.^38 Conceptually, therefore, in this sense the right closely resembles other easements the court has recognised permitting private nuisances such as polluting water^39 or excessive noise.^40 The second situation involves two adjoining freeholds, as was the case in Jones v Price.^41 Here the position changes, because only the servient owner has an obligation to fence, so the dominant owner needs not only a defence against cattle trespass, but also the ability to claim (^33) [1971] 1 QB 77. (^34) Rogers v Hosegood [1900] 2 Ch. 388 established that the benefit of a covenant that relates to the land may be annexed to that land if the covenantee has a legal estate to which the benefit may attach. (^35) [1956] Ch. 131. (^36) Re Ellenborough Park [1956] Ch.131 at 140. (^37) Liability for cattle trespass arises under the principle of Rylands v Fletcher (1868) LR 3 HL 330. (^38) [1975] QB 62. (^39) See Baxendale v McMurray (1867) Ch App 790. (^40) See Ball v Ray (1873) 8 Ch App 467. (^41) [1965] 2 QB 618.

damages when the servient owner’s animals enter and cause damage to the dominant land. It can be seen that custom is not really a satisfactory explanation of such an arrangement, at least not in the same compelling way that it applies in the common land scenario. In this second situation there are only two parties involved and they could reach any arrangement they wished, or none. In the absence of custom as a possible origin of the right, a dominant owner must be able to establish either immemorial usage or at least the possibility of a lost modern grant.^42 Fencing easements seen as obligations to fence are in essence no more inherently proprietary than covenants that relate to land, the only obvious distinction being that a covenant is set out in a deed.^43 Viewed as a defence against trespass, they resemble quite closely a right to stray animals, but is this enough? It is settled law that a true easement requires “no more than sufferance” 44 on the part of the servient owner, and the existence of a duty to repair is fatal. It appears at first view that the “defence against trespass” right breaches this condition and so justifies Patten LJ’s finding that even in this form fencing easements are “not easements”.^45 It is, however, possible to treat the failure to fence not as the breach of an obligation, but as a condition precedent to the emergence of the interest itself. Indeed for so long as the servient owner maintains the fence, the easement-like defence never arises. If the fencing easement were used only as a defence against trespass, it might therefore be possible to dismiss its being referred to as a duty to fence as a mischaracterisation and thus resolve the difficulty. Fencing easements are in practice, however, not restricted to this use. In Lawrence v Jenkins^46 the claimant’s animals were injured while on the defendant’s land, the defendant being liable for failure to maintain a boundary fence. In this form the term fencing easement has to be seen as a misnomer; it is simply a positive covenant that relates to land. Fencing easements of this type can, it is submitted, never be granted in the same sense that an easement is granted, whatever wording is used to create them. Nevertheless, it remains theoretically possible under Churston, and the question remains whether the court could or should have upheld the respondent’s argument. What can be granted, and what cannot The existence of any liability in this appeal hinged on whether the clause in question constituted a fencing easement at law or was instead merely a covenant.^47 If the latter, the matter was easily disposed of, the law being clear that the golf club could not have acquired the burden of the covenant either at law^48 or indeed in equity.^49 Successive owners of the dominant land might, however, retain the benefit^50 but Mr Haddock’s precaution of taking an express assignment of the benefit of the clause^51 was redundant because the appellant was not the original covenantor.^52 (^42) See Churston at 22. (^43) See Ball v Ray (1873) 8 Ch. App. 467. (^44) Jones v Price [1965] 2 Q.B. 618 at 631 per Willmer LJ. (^45) Churston Golf Club Ltd v Haddock [2019] E.W.C.A. Civ. 544 at 10. (^46) (1872-73) L.R. 8 Q.B. 274. (^47) See Churston Golf Club Ltd v Haddock [2019] E.W.C.A. Civ. 544 at 27. (^48) Austerberry v Oldham Corporation (1885) 10 App. Cas. 590. (^49) Rhone v Stephens [1994] 2 A.C. 310. (^50) Spencer’s Case (1583) 5 Co. Rep. 16a.; Rogers v Hosegood [1900] 2 Ch. 388. (^51) Patten LJ explains this in Churston Golf Club Ltd v Haddock [2019] E.W.C.A. Civ. 544 at 5. (^52) Austerberry v Oldham Corporation (1885) 10 App. Cas. 590.

Patten LJ’s decision that the covenant could not also be a fencing easement seems to assume that different wording was both necessary, and achievable: “Had that been the draftsman's intention one would have expected in a carefully drafted conveyance of this kind to see the easement included in terms as an express grant of such a right by way of exception to the land conveyed to the Council. As it is, clause 2 is framed (like clause 3) in the language of a covenant and in my view should be treated as one.”^53 The flaw in this reasoning is that it is only possible to except something from a conveyance of the fee simple if it is one of the constituent parts of that estate. An exception is, by definition, something retained by the transferor. It might be corporeal or incorporeal. So, on a sale of part, the seller can keep back part of the right to walk across the land sold, in exactly the same way as he might retain the mines and minerals. With a fencing easement, if the “right” that is to be created consists of the ability to enforce an obligation to fence, it is an entirely new right, an artificial accretion to the fee simple, contractual in nature; clearly no such right inheres in the freehold to start with, and therefore it cannot be excepted from a transfer of that estate. The court cannot legitimately expect any draftsman, however skilled, to attempt to achieve the impossible. The wording in the Churston conveyance must be interpreted as a positive covenant; of that there can be no debate, but that conclusion is not the end of the analysis. Indeed to assert that this fact alone means it cannot take effect as a fencing easement is to go against the reasoning of the House of Lords in Russell v Watts^54 in which Lord Blackburn stated “a covenant may, if it is necessary in order to carry out the intention, operate as a grant”.^55 If the duty to fence does lie in grant, as Crow v Wood seems to have established, there therefore seems to be no obvious reason, Austerberry aside, why a covenant of the Churston kind could, in another case, also take effect as a fencing easement. Conclusion Fencing easements may be anomalous,^56 but still have to fit within the broad parameters of the law of property. It is submitted that fencing easements are sui generis, existing in two distinct forms. Firstly there is the simple duty to fence, and secondly a consequential, but separable, easement-like right which is the defence against trespass. The latter should not be affected by the rule in Austerberry , as it is not a covenant of any kind. Neither does it require any positive act on the part of the servient owner. In fact quite the reverse is true, as due maintenance of the boundary would nullify the right as opposed to supporting it. It is conceptually close to other permissive easements, with the difference that it must presumably be a defence only. Even though it is possible to conceptually to separate the defence against trespass from the duty to fence, there would still remain difficulties for the conveyancer to attempt to frame such a specific right without reference to the duty that underpins it. It would however be possible to include clarificatory wording that would assist in making the parties’ intentions (^53) Churston at 27 per Patten LJ. (^54) (1885) 10 App. Cas. 590. (^55) (1885) 10 App. Cas. 590 at 611 per Lord Blackburn. (^56) Diplock LJ so describes it in Jones v Price [1965] 2 QB 618 at 634.

clear. This should ensure that the mere existence of the covenant element should not nullify the right entirely. The right to sue one’s neighbour for failing to maintain the boundary fence does not obviously resemble a true easement. It is the benefit of a positive covenant that has been held to arise through long use. In this prescriptive form, its legitimacy has remained unchallenged, upheld, it seems, on the erroneous ground that it must be a form of genuine easement. In any event the ability to enforce the obligation against successive freeholders has survived Austerberry. The court in Churston (and in the cases that preceded it) has set much store by the notion that Austerberry does not apply to an obligation that arises without express creation. In Rhone v Stephens,^57 however, Lord Templeman confirmed that Austerberry is not a technical rule but a manifestation of a broader principle, explaining that the common law has no power “to enforce a personal obligation against a person who has not covenanted.”^58 Yet fencing easements do precisely this. It therefore follows that, because fencing easements that arise through custom or prescription already breach Austerberry , an express covenant constitutes no greater offence against the rule. Churston leaves open the possibility for inventive conveyancers to attempt to create fencing easements by express grant. Future decisions may yet rule this out, but until that happens there is nothing to be lost in the attempt. A repair covenant in standard terms could be coupled with a reference to the parties’ intention to create “a fencing easement in fee simple” in favour of the dominant land. This would prevent the court from treating the covenant element as the sole purpose of the wording. As explained above, it is not possible simply to except or reserve out of a freehold a duty to do something. It is possible that a conveyance takes place in circumstances similar to those of Crow v Wood , where the relevant duty already exists and can simply be conveyed pursuant to the effect of s62 of the LPA 1925. It would of course be a pre-requisite that there should be no blanket exclusion of s62 in the conveyance. If s62 is excluded, as is commonly the case, the buyer’s conveyancer would need to make sure express reference to the duty is made. Churston serves to emphasise the advantages the Law Commission’s 2011 reform proposals would bring to conveyancers seeking to append positive estate management covenants to land. Under these proposals it will be possible to directly burden land^59 with positive obligations that touch and concern the dominant owner’s land.^60 A fencing covenant should have no difficulty passing that test, as the duty to fence is clearly beneficial to successive owners of the adjoining land who are saved the cost themselves and receive the protection and privacy the fence affords. In the event that it proves possible to grant express fencing easements, it is difficult to predict whether the courts would limit their application to fences as opposed to other boundary features such as walls or ditches. All of the existing cases involve fences in rural locations. (^57) [1994] 2 A.C. 310. (^58) Rhone v Stephens [1994] 2 A.C. 310 at 321. (^59) To see how this would be achieved, see s1, ss 3(b)(i) ofthe draft Bill at Appendix A of the Law Commissions 2011 report “Making Land Work: Easements, Covenants and Profits à Prendre” Law Com No 327. (^60) Such obligations would, according to the Law Commission, be expected to pass the existing “touch and concern” tests in the cases, notably P&A Swift Investments v Combined English Stores Group plc (1989) AC 632; see Law Com No 327 at para 6.19.

The strongest justification for limiting the ambit to this would perhaps be the custom-based common land situations which dealt with a specific rural requirement that is not replicated in other settings. The contrary argument, that other boundary features might be included, is that the law should distinguish cases on the basis of legal principle only as opposed to historical origin. This seems to be the better argument. Churston is a decision limited to its particular facts, and it purports to establish no new principle. Importantly, however, it does not rule out the possibility of creating an express fencing easement. This is most likely to be needed in sales of part where obligations are allocated to either seller or buyer. If the buyer’s land is to be burdened, it is plainly impossible to except from a freehold transfer the benefit of a positive obligation that has not yet been created. If the seller’s land is to be servient, it is highly doubtful whether the benefit of a duty to repair can be granted in simple terms, even on the assumption that such a grant is meaningful at all. It seems inevitable that any express terms must involve a covenant, at least in their effect, whatever wording ingenious conveyancers might come up with. Churston is ultimately a case about a covenant, and there are those who will see it as little more than this. If fencing easements were to be rebranded as “proprietary fencing covenants” and created as such, some of the difficulties might recede. Austerberry is still the obvious problem, but it may not be the obstacle that it has been portrayed to be; fencing easements pre-date Austerberry and, it is submitted that they may be treated as entirely outside the remit of that decision, whether arising expressly or through long use. Finally, it should be noted that all of the difficulties of express creation will disappear if the Law Commission’s 2011 proposals are enacted. These reforms so far as they relate to covenants are, it is submitted, long overdue. Andrew Cash August 2019