Cases considered: Kolias v. Owners: Condomimium Plan 309 CDC, 2008 ABCA 379.
PFD Version: An exemplar of reasoning from precedent in a real property law context
This reserved judgment written by Mr. Justice Jean Côté of the Alberta Court of Appeal is strong on justifications for the decision reached and an excellent example of stare decisis and the doctrine of precedent at work. It reverses a decision of the Court of Queen’s Bench and discharges a restrictive covenant on the basis that the dominant tenement was not easily ascertainable in the deed creating the restrictive covenant, as required by the Supreme Court of Canada in Galbraith v. Madawaska Club, [1961] S.C.R. 639, 29 D.L.R. (2d) 153. The decision is also a good example of the priority given to certainty and predictability in property law.
I have been critical in recent posts about decisions made without reasons or without reference to statutory or case law authority: see Partition or sale of co-owned property? and Is there really any question about the test for part performance in Alberta? Justice Côté’s decision is a refreshing change and I want to use his judgment and this comment to make a few general points about legal reasoning. For example, there are many reasons why judicial justification is, at the least, a good idea in law. One of those is explained by Martin P. Goldberg, a Professor of Philosophy and Law at Duke University, in his book on Legal Reasoning (Broadview Press, 2001) in the following terms:
Most people who are at the losing end of a case are not very happy with the outcome. The stakes at issue might have been high: years behind bars in a criminal case, a great deal of money in a civil case. Even if the issue in a civil case seems trivial to an outsider, it must have been important enough to the parties for them to have pursued their dispute in a court of law. So the loser may complain about the result. But is he entitled to complain? One of the most important functions of the reasoned decision – a decision for which the judge or official articulates the (justifying) reasons – is to enable this question to be answered (at 6, emphasis in the original).
I was also critical of the Queen’s Bench decision in this case, noting that its reasoning and result were out of step with the weight of authority at common law and that there was no support in the Land Titles Act, R.S.A. 2000, c. L-4, for a change in the common law: see Ascertaining a dominant tenement with a right to a view. Justice Côté’s decision brings the result and reasons in this case into conformity with previous decisions in Alberta and other Canadian provinces.
The language of restrictive covenants – the subject matter of the case and this comment – might need some introduction. Technically, a “covenant” is a promise made in a contract under seal. In the restrictive covenant context, it is a promise made by a covenantor (who bears the burden of performing the promise) to a covenantee (who enjoys the benefit of the promise’s performance). The land burdened by the promise is called the servient tenement and the land benefitted by the promise is called the dominant tenement. Promises made in contracts are normally only enforceable by and binding upon the original parties to the contract. However, restrictive covenants over land can be enforced against subsequent owners of the servient tenement by subsequent owners of the dominant tenement if the covenants meet certain legal requirements. University of Alberta Law Professor Bruce Ziff describes the requirements for a valid restrictive covenant in his Principles of Property Law, 4th edition (Toronto: Thomson Carswell, 2006) at pages 380-384. I have broken them down and somewhat simplified them in the following list:
- The covenant must be negative in substance, that is, compliance with the promise is possible by doing nothing.
- The burden must have been intended to burden the servient tenement, and not just the covenantor.
- The servient tenement must be sufficiently described in the document creating the restrictive covenant.
- The covenant must have been intended to benefit the dominant tenement, and not just the covenantee.
- The dominant tenement must be sufficiently described in the document creating the restrictive covenant.
The facts and the procedural history
The restrictive covenant at issue in this case was contained in an agreement dated September 9, 1971 and made between Glenview Construction Ltd., then the owner of the Covenant House land in Calgary, and The A.D. Gelmon Development & Management. Lot 10 was owned at the time by Gustave Engbloom. On September 17, the day that the restrictive covenant was registered against Lot 10 at the Land Titles Office, Lot 10 was transferred to Glenview Construction Ltd., then to Gelmon Development & Management, and finally to The Gelmon Corporation. Title to Lot 10 later passed to Eva Gelmon, and then to the appellants, Ike and Lisa Kolias, in 1998. In the restrictive covenant, Lot 10 was divided into three areas and height restrictions were imposed on two of those three areas. No structure or hedge over six feet in height could be built or placed in the first of those areas. Nothing over fourteen feet in height (except chimneys or radio or T.V. antenna) was allowed in the second area. Although not specified in the restrictive covenant itself, its purpose was probably to protect the sight lines of the units in the six storey condominium on the adjacent lot, the six-storey condominium known as Covenant House. At least some of the condominium units in Covenant House have views of Heritage Park, the Glenmore Reservoir and the mountains to the west, across Lot 10.
The Koliases applied under section 48 of the Land Titles Act for a discharge of the restrictive covenant in 2006. Presumably they or their intended purchaser wanted to build one or more structures on Lot 10 that would interfere with the views currently enjoyed by some of the unit holders in Covenant House. They argued that the document creating the covenant did not identify the dominant tenement, the fifth requirement for formal validity listed above. In other words, it did not describe the parcel of land that benefited from the height restrictions that burdened Lot 10. As a result, they argued, the covenant could not be enforced against them as subsequent purchasers. It was a personal covenant only, binding upon Glenview Construction Ltd. and The A.D. Gelmon Development & Management, the original parties to the September 9, 1971 agreement.
The respondents, the owners of the condominium units in Covenant House, did not dispute that no specific term in the restrictive covenant identified the land to be benefited by the height restrictions on Lot 10. Nevertheless, they made three arguments in favour of the validity of the restrictive covenant. First, they argued that it was readily ascertainable from the restrictive covenant itself that the Covenant House land was the dominant tenement. Second, and in the alternative, they argued that extrinsic evidence should be admitted to identify the dominant tenement. That argument was the one that succeeded in the lower courts. Finally, and in the further alternative, the respondents asked that the deed containing the restrictive covenant be rectified.
In December 2006, Judith Hanebury, a Master of the Court of Queen’s Bench, refused to discharge the restrictive covenant. She had concluded that the dominant tenement could not be properly ascertained on the face of the document but she found that extrinsic evidence could be used to resolve the matter. In November 2007, Mr. Justice Alan D. Macleod denied the appeal: Kolias v. Owners Condominium Plan 309 CDC, 2007 ABQB 714. He held that the Court was allowed to look at the evidence of the surrounding circumstances at the time the restrictive covenant was created. He also held that, had one searched the state of affairs surrounding Lot 10 and the parcel of land on which Covenant House was erected at the time the September 9, 1971 agreement was signed, it was easily ascertainable that Glenview owned the Covenant House lands which immediately joined Lot 10 to the east and that they were the only lands adjoining Lot 10 that were owned by Glenview at the time. Justice Macleod therefore concluded (at para. 27) that “This is clearly not a case where the dominant tenement was not traceable such that the appellants could be left in the dark.” The Koliases appealed yet again to the Alberta Court of Appeal, making the same argument they had made in the lower courts, and this time they were successful.
As I have already indicated, the respondents did not dispute that no specific term in the restrictive covenant identified the dominant tenement. The leading Canadian case on what is required for a valid restrictive covenant is the Supreme Court of Canada decision in Galbraith v. Madawaska Club Ltd. The most often quoted passage from that case, found in the judgment of Judson J. at page 653, contained the following rule:
“[T]here is no evidence anywhere in the record to indicate whether the Club had any such land capable of being benefited. … This fails to meet what I think must be regarded as the minimum requirements that the deed itself must so define the land to be benefited as to make it easily ascertainable” (emphasis added).
The Supreme Court further held that a restrictive covenant that failed to meet this minimum requirement did not bind subsequent purchasers of the lands burdened by the restrictive covenant. The rationale for the rule, according to Ziff (at 382), is that a person who acquires the servient tenement has a right to know who may enforce the covenant, i.e., who holds the dominant tenement.
As Justice Côté notes (at para. 13), the Alberta Court of Appeal had already adopted the Galbraith v. Madawaska Club “easily ascertainable” test in Guaranty Tr. Co. of Can. v. Campbelltown Shopping Centre (1986), 72 A.R. 55, 44 Alta. L.R. (2d) 270 (C.A.). Why the reference to the “adoption” of the Supreme Court of Canada test, rather than to “following” it as a binding precedent? The respondents had argued that the Supreme Court of Canada’s statements about the required elements of restrictive covenants were merely obiter dicta. In other words, they argued that the passage from Judson J., quoted above, was not required in order to decide the question before the court and, therefore, the “easily ascertainable” test set out by Judson J. did not need to be followed.
In Galbraith v. Madawaska Club, the Supreme Court of Canada held that a covenant inserted into a deed by a golf club, which restricted who could occupy the servient tenement, did not run with the land. In deciding the question, Judson J. said (at para. 21):
The covenant in question here gives the Club the right to choose the persons who shall occupy the servient land, if the owner wishes to go outside the club membership. This has nothing to do with the use to which the land may be put, but relates only to the kind of person who may be given occupation. It is imposed by the vendor for its own benefit as a club. It does not touch or concern the land, as being imposed for the benefit of or to enhance the value of land retained by the club. It calls into being the exercise of an unfettered personal discretion by the club management and its plain purpose is to preserve the amenities of the club.
This passage from Galbraith v. Madawaska Club confirms that the ratio decidendi of the case related to the fourth requirement for formal validity listed above – the need for the covenant to benefit the dominant tenement – and not the fifth requirement. The “easily ascertainable” test does indeed appear to be obiter dicta, as the respondents argued. However, Justice Côté did not base his use of the “easily ascertainable” test on that point in Galbraith v. Madawaska Club being binding upon the Alberta Court of Appeal.
Instead, Justice Côté noted that, even if the “easily ascertainable” test set out by Judson J. is obiter, the test is still persuasive authority. The passage was part of “a considered discussion.” As persuasive authority, the test should be followed unless there is good reason not to. Justice Côté sets out two different types of reasons in determining that there was good reason to follow the “easily ascertainable” test from Galbraith v. Madawaska Club. First, he notes many Canadian cases that have followed and applied that test and that all of the leading Canadian property law texts treat the words of Judson J. as accurately reflecting Canadian law. Authority is therefore on the side of applying the “easily ascertainable” test. It has persuaded many already that it is correct. Second, he notes broader principles and policies in favour of not changing the law. Reliance is a major reason; in certain areas of the law such as real property conveyancing, it is assumed that people have come to rely upon the law and to structure their transactions based upon that law.
Of the cases that have followed or applied the easily ascertainable test from Galbraith v. Madawaska Club, Justice Côté cites Re Sekretov and Toronto, [1973] 2 O.R. 161, 166, 33 D.L.R. (3d) 257 (C.A.); Hi-Way Housing (Sask.) v. Mini-Mansion Constr. Co. [1980] 5 W.W.R. 367, 4 Sask. R. 415 (C.A.), as well as the previously referred to decision of the Alberta Court of Appeal in Guaranty Trust. Re Sekretov and Toronto is worth elaborating upon as the Ontario Court of Appeal was unequivocal in its endorsement of the “easily ascertainable” test when that court held (at pp. 261-262):
The law of Ontario and of other common law Provinces plainly require that the dominant land for the benefit of which a restrictive covenant is imposed in a deed from the covenantee to a purchaser of other lands of the covenantee must be ascertainable from the deed itself; otherwise, it is personal and collateral to the conveyance as being for the benefit of the covenantee alone and not enforceable against a successor in title to the purchaser. This was laid down by Judson J., in Galbraith v. Madawaska Club Ltd, supra, in the plainest terms, the minimum requirement being, as the learned jurist stated, that the deed itself must so define the land to be benefited as to make it easily ascertainable.
There are many more such cases. In the past two years, cases relying on the easily ascertainable test from Galbraith v. Madawaska Club include Save the Heritage Simpson Covenant Society v. Kelowna (City), 2008 BCSC 1084 (concerning two parcels of land sold to the City of Kelowna on condition the City would use the land for municipal purposes only and would not sell the property or use it for commercial or industrial purposes) and Mohawk Square Developments Ltd. v. Suncor Energy Inc. 62 R.P.R. (4th) 100 2007 CarswellOnt 5894 (concerning the enforcement of a restrictive covenant prohibiting the defendant from selling food, novelties, gifts and souvenirs from its store adjacent to the plaintiffs).
It is true that the requirement for specificity concerning the dominant tenement in the document creating the restrictive covenant has been relaxed in England, where so-called “implied annexation” is permitted in cases where there is no real doubt about the parties’ intentions. See Marten and others v. Flight Refuelling Ltd., and another, [1961] 2 All E.R. 696 (Ch. D.). However, as the Ontario Court of Appeal noted in Re Sekretov and City of Toronto (at 262), “The language used in the Madawaska Club case clearly excludes any notion of annexation to dominant land by implication.” The English approach is not the Canadian approach.
There is also some controversy in Canada about just how specific the identification of the dominant tenement must be. Some cases, such as Sawlor v. Naugle et al. (1990), 101 N.S.R. (2d) 160 at 166 (T.D.) require “either a metes and bounds description of the lands, a reference to a plan identifying the lands or some other specific reference by which the lands to be benefited may be readily identified or ascertained.” The British Columbia Court of Appeal in Kirk was “uncomfortable with the level of description that they seem to require of the instrument” (para. 32). However, the level of specificity of the description of the dominant tenement was not the issue in this case.
The relevant rule of law, therefore, was that “the deed itself must so define the land to be benefited as to make it easily ascertainable”: Galbraith v. Madawaska Club at 653.
The appellants’ first argument was that it was easily ascertainable from the September 9, 1971 agreement that the Covenant House land was the dominant tenement. Unfortunately, in so doing, they apparently put forward a multi-stage explanation that was not that easy to grasp. Even Justice Macleod, who ruled in their favour, described the dominant tenement as merely “traceable” (at para. 27), not identified. In addition, the condominium plan was registered at the Land Titles Office before the restrictive covenant was registered. As a result, there was, in law, no single parcel of land adjacent to Lot 10. Instead there was a vertical condominium development with a number of different parcels of land. They would not all benefit from the restrictive covenant on Lot 10.
The appellants’ second argument was that extrinsic evidence should be admitted to identify the dominant tenement. As already mentioned, Justice Macleod had admitted extrinsic evidence of surrounding circumstances to identify the dominant tenement, relying on Kirk v. Distacom Ventures (1996) 81 B.C.A.C. 5, 4 R.P.R. (3d) 240 (C.A.). Justice Côté, however, noted that Kirk involved a restrictive covenant which included an ambiguous description of the intended dominant tenement. The British Columbia Court of Appeal admitted the extrinsic evidence in order to resolve that ambiguity. Justice Macleod also relied upon Guaranty Trust when he admitted the extrinsic evidence. Justice Côté noted that Guaranty Trust, like Kirk, involved a restrictive covenant which did make some reference to the dominant tenement. Justice Côté therefore distinguished Kirk and Guaranty Trust on the basis that the September 9, 1971 agreement in this case did not contain any reference to the intended dominant tenement. There was no ambiguity; there was an omission. While extrinsic evidence might be allowed to resolve ambiguous references to dominant tenements, there was no ambiguity in this case. Therefore, extrinsic evidence of circumstances surrounding the September 9, 1971 agreement was not admissible.
In distinguishing Kirk and Guaranty Trust, Justice Côté was entirely correct, as a review of the Kirk decision itself would confirm. In that case, the British Columbia Court of Appeal considered the authorities, including Guaranty Trust, and came to the following conclusion about the use of extrinsic evidence to identify the dominant tenement (at para. 32):
Galbraith is clear authority for the proposition that extrinsic evidence cannot be adduced to identify dominant lands where an instrument does not purport to do so. However, where an instrument does contain a term purporting to identify dominant lands, extrinsic evidence may be used to explain that term or to identify its subject matter . . .
In other words, if there is no reference at all to the lands to be benefited in the deed, the use of extrinsic evidence of surrounding circumstances is precluded by the authorities.
At this point in his judgment, having dealt with the appellants’ second argument, Justice Côté does not move on to consider their third argument. Instead, in a section headed “Framing Workable Law For All,” he acknowledges (in para. 27) that “[i]t is very easy to be sympathetic with someone trying to enforce a restrictive covenant or other encumbrance against someone who later took title knowing of this encumbrance, probably fully understanding what it entailed.” He adds that ruling in favour of someone who knew about the encumbrance, as did the Koliases, might seem like relying on a technicality to defeat a just claim.
Nevertheless, Justice Côté concludes (at para. 26) that “[c]ertainty of conveyancing and respect for precedent must trump dislike of a claim jumper or top-leaser.” Analogizing the Koliases to top leasers and claim jumpers might seem a little harsh. A top leaser is someone who wants to sign a mineral lease that will take effect at the end of a current lease of those same minerals. A claim jumper is a person who overstakes a claim which was already staked by another prospector. The top leaser and claim jumper language drives home Justice Côté’s point that fairness or justice in this individual case have to give way to certainty and predictability in real property cases. Instead of focusing on the two parties to this case, Justice Côté asks us to consider (at para. 28) “what rules of conduct the law lays down for future transactions, and what conduct it wishes to encourage or discourage. . .”.
In terms of legal reasoning in general, and reasoning based on adherence to precedent in particular, Justice Côté’s shift in focus from the parties before the court to the persons who might in the future be affected by this decision of the court is a common argument and justification in law. Precedent is not only backward looking; it is also forward looking. However, a focus on future parties and consequences is not uncontroversial. The well known legal realist author and judge of the U.S. Court of Appeals for the 2nd Circuit, Jerome Frank, cautioned in his concurring decision in Aero Spark Plug Co. v. B.G. Co. 130 F.2d 290, at 294-99 (1942):
Perhaps the central theme in most discussions of the judicial process is the obligations of judges to consider the future consequences of their specific decision. Such discussions usually stress the “rule” or precedent aspect of decisions. Thus Dickinson, in 1927, paraphrasing (it may be unconsciously) Aristotle’s remarks made almost twenty-two hundred years earlier, writes “One danger in the administration of justice is that the necessities of the future and the interests of parties not before the court may be sacrificed in favour of present litigants”; he thinks it imperative that judges “should raise their minds above the immediate case before them and subordinate their feelings and impressions to a practice of intricate abstract reasoning . . . .
Although much can be said for that attitude – of considering a case primarily with reference to its significance in future cases – it is sometimes given too much weight. Excessive concentration of attention, by some upper court judges, on the formulation in their opinions of so-called legal rules, with an eye chiefly to the impact of those rules on hypothetical future cases not yet before the court, sometimes results in their allotting inadequate attention to the interests of the actual parties in the specific existing cases which it is the duty of the court to decide. Such judges never quite catch up with themselves; for, in cases which actually occur, they are deciding future cases which may never occur. . . .
What broader principles of or policies underlying real property and conveyancing law are in play in this case that would require the court to rule in favour of the Koliases? Justice Côté identifies two (at para. 29):
(a) there must be few and carefully-defined restrictions on free alienation of land; and
(b) conveyancing needs a high degree of certainty, and profits greatly from conveyancing which is quick and cheap.
The common law has long promoted the right to transfer property freely as an inherent attribute of ownership and therefore unacceptable restraints on alienation of property are discouraged. For a brief discussion of the policy ideas underlying this idea, see Ziff at 241-243. As for the latter policy, it is emphasized by our land titles system, a Torrens title registration system. That system was designed to be simple, efficient and inexpensive. As Justice Côté notes (at para. 31), it “features a simple title, easy to read, subject only to a few types of encumbrance, noted directly on the relevant title, which encumbrances are themselves easy to access and read.”
“Bright line” rules, clear and easy to apply, are typically a priority for real property law. In Galbraith v. Madawaska Club Ltd. the Supreme Court of Canada laid down such a rule. That case requires that “the deed itself must so define the land to be benefited as to make it easily ascertainable.” Easily ascertainable is what is required, and not ascertainable after hours of explanation or a complicated process of deductions and inferences. Identification in the deed itself, and not in extrinsic evidence, is what is required.
Given how Justice Côté dealt with the first two arguments, the appellants’ third argument could be dealt with in a cursory fashion, based as it was on extrinsic evidence and notions of fairness. The appellants sought rectification of the restrictive covenant. Rectification was explained by the Ontario Court of Appeal in the following terms in H.F. Clarke Ltd. v. Thermidaire Corporation (1973), 33 D.L.R. (3d) 13 at 20-21:
In order to succeed on a plea of rectification the Court must be satisfied that the parties were in complete agreement as to the terms of their contract but wrote them down incorrectly. It is not a question of the Court being asked to speculate about the parties’ intention but rather to make an inquiry to determine whether the written agreement properly records the intention of the parties as revealed in their prior agreement.
The respondents wanted the court to add a new clause to the September 9, 1971 agreement identifying the Covenant House lands as the dominant tenement. Justice Côté noted two problems with this request. First, without the inadmissible extrinsic evidence, how could the court tell which parcel of land the parties intended to be the dominant tenement? If the evidence was inadmissible to identify the land, how could it be used to rectify the contract that did not identify the land? Second, rectification is an equitable remedy and it would not be just to amend a contract made between Glenview Construction Ltd. and The A.D. Gelmon Development & Management so that the contract bound third party subsequent purchasers such as the Koliases. In addition, it would not be consistent with the principles behind the Torrens System that allow subsequent purchasers to rely on the register and the certificate of the title they are buying.
In its policy reasons, this case is a good example of the usual tension seen in cases dealing with property rights in general and the Torrens system of land registration in particular, the tension between certainty and predictability on the one hand and justice in the individual case on the other hand. The Torrens system of title registration always favours the former. It was that body of law known as equity that paid attention to justice in individual cases and concerned itself with whether the conscience of someone like the Koliases ought to be affected. Equity did not favour those, such as the Koliases who had notice of another’s claim to Lot 10 prior to acquiring their own claim to Lot 10. Neither did the lower courts. However, in the face of a Supreme Court of Canada precedent adopted in Alberta and elsewhere which provides a “bright line” test for distinguishing valid restrictive covenants from invalid ones, the Alberta Court of Appeal opted for the justice that comes with numerous parties being able to organize their transactions in reliance upon the rules. As Justice Hugh H. McLellan of the New Brunswick Court of Queen’s Bench succinctly put it in Lockmac Holdings Ltd. v. Earle (1997), 185 N.B.R. (2d) 360, 12 R.P.R. (3d) 142 at para. 13, in another restrictive covenant case applying Galbraith v. Madawaska Club:
Those previous decisions are called precedents. By following precedents and other legal authorities, courts decide similar cases the same. In that way the law remains equal, predictable and common.
