(with whom McLachlin C.J. and Abella, Moldaver, Karakatsanis, Gascon, Côté and Brown JJ joined)
The outcome of these appeals hinges on the interpretation of an exclusion clause in a common form of all-risk property insurance, variably referred to as “builders’ risk”, “contractors’ risk”, “all risks”, “multi-risk” or “course of construction” insurance. This type of insurance covers physical damage on a construction site. It is usually issued to the owner of the property under construction and the general contractor, providing coverage for them as well as for all contractors and subcontractors working on the project. The exclusion clause at the heart of these appeals is a standard form clause that denies coverage for the “cost of making good faulty workmanship” but, as an exception to that exclusion, nonetheless covers “physical damage” that “results” from the faulty workmanship.
In the present case, a contractor was hired to clean the windows of a building under construction. In the course of the cleaning, the contractor scratched the building’s windows, which ultimately needed to be replaced. The windows’ replacement cost was claimed by the building’s owner and the general contractor in charge of the project under a builders’ risk policy issued in favour of the owner and all contractors involved in the construction, but the insurers denied coverage on the basis of the “cost of making good faulty workmanship” exclusion. The issue before the courts was thus to determine, where windows of a construction project are damaged from post-installation cleaning by a contractor responsible for only their cleaning, if the cost of the windows’ replacement was excluded from coverage under the faulty workmanship exclusion.
After determining that the work performed by the contractor amounted to faulty workmanship, the trial judge applied the contra proferentem rule against the insurers and concluded that the faulty workmanship exclusion did not exclude from coverage the damage that the contractor had caused to the building’s windows. Applying a correctness standard of review to the interpretation of the insurance policy, the Court of Appeal of Alberta overturned the trial judge’s decision and declared that the damage to the building’s windows was excluded from coverage, as the damage was physically or systematically connected to the very work the contractor had performed.
In my opinion, the appropriate standard of review in this case is correctness. Where, like here, the appeal involves the interpretation of a standard form contract, the interpretation at issue is of precedential value, and there is no meaningful factual matrix that is specific to the particular parties to assist the interpretation process, this interpretation is better characterized as a question of law subject to correctness review.
Regarding the appropriate interpretation of the faulty workmanship exclusion in all builders’ risk policies, I am of the view that the exclusion clause serves to exclude from coverage only the cost of redoing the faulty work. This interpretation is dictated by the general rules of contractual interpretation. It best represents the parties’ reasonable expectations, as informed by the purpose of builders’ risk policies, aligns with commercial reality, and is consistent with the jurisprudence on the matter. In this case, the cost of redoing the faulty work is that of recleaning the windows. Therefore, I would allow the appeals and hold that the windows’ replacement cost is covered under the insurance policy.
Station Lands Ltd. (“Station Lands”) is the owner of the recently built EPCOR Tower (“Tower”), an office building in Edmonton. Ledcor Construction Limited (“Ledcor”) was the general contractor for the Tower’s construction.
During construction, the Tower’s installed windows were dirtied with paint specks, dirt and concrete splatter. To clean these windows prior to the completion of construction, Station Lands hired Bristol Cleaning (“Bristol”). The service contract between Station Lands and Bristol stipulated that Station Lands would provide all-risk property insurance for the project, which Station Lands did in the form of a builders’ risk policy (the “Policy”). The scope of Bristol’s work under the service contract was to “[p]rovide all necessary equipment, manpower, [and] materials required to complete a construction clean” of the Tower’s exterior windows.
Unfortunately, Bristol used improper tools and methods in carrying out its cleaning work, scratching the Tower’s windows, which consequently had to be replaced. Station Lands estimated the replacement cost of the windows to be $2.5 million. Both Station Lands and Ledcor claimed this replacement cost against the Policy through their insurers at the time, the respondents Commonwealth Insurance Company, GCAN Insurance Company, and American Home Assurance Company (together, the “Insurers”).The Insurers denied the claim on the basis of clause 4(A)(b) of the Policy (the “Exclusion Clause”), which is an exclusion for faulty workmanship.
The relevant coverage provisions of the Policy provide that all risks of direct physical loss or damage to the property undergoing construction are insured, subject to certain outlined exclusions:
The Exclusion Clause excludes from coverage the “cost of making good faulty workmanship”, but provides an exception for “resulting damage”:
Station Lands and Ledcor (together, the “Insureds”) submitted their statement of claim before the Court of Queen’s Bench of Alberta, seeking enforcement of the Policy and coverage for the replacement cost of the damaged windows.
III. Decisions Below
A. Court of Queen’s Bench of Alberta, 2013 ABQB 585,  I.L.R. ¶ I-5495
The trial judge concluded that the cleaning work Bristol had carried out constituted “workmanship” and that it had been faulty. He declared, however, that the Exclusion Clause did not exclude from coverage the damage that Bristol’s faulty workmanship had caused to the Tower’s windows. In coming to this determination, he found the Exclusion Clause ambiguous and the interpretations of “making good” advanced by the Insureds and Insurers equally plausible. He therefore applied the rule of contra proferentem against the Insurers. The Insureds had argued that the “cost of making good” encompassed only the cost of redoing the cleaning work, whereas the Insurers had argued that it encompassed both the cost of redoing the cleaning work and the damage to the windows, as they were the very thing on which Bristol had performed the faulty workmanship.
B. Court of Appeal of Alberta, 2015 ABCA 121, 599 A.R. 363
On appeal, the Court of Appeal reversed the trial judge’s decision and declared that the damage to the Tower’s windows was excluded from coverage. Applying a correctness standard of review to the interpretation of the Policy, the court held the trial judge had improperly applied the rule of contra proferentem because the Exclusion Clause was not ambiguous.
The Court of Appeal proceeded from the premise that because the base coverage under the Policy was for “physical loss or damage”, as provided by clause 2, the Exclusion Clause had to exclude physical damage of some kind, or else it would be redundant. For the court, then, the key was to determine the dividing line between the physical damage that was excluded as the “cost of making good faulty workmanship” and the physical damage that was covered as “resulting damage”. To establish this dividing line, the court devised a new test of physical or systemic connectedness, based on three primary considerations, outlined at para. 50 of its reasons:
the “extent or degree to which the damage was to a portion of the project actually being worked on at the time, or was collateral damage to other areas”;
the “nature of the work being done, how the damage related to the way that work is normally done, and the extent to which the damage is a natural or foreseeable consequence of the work”; and
“[w]hether the damage was within the purview of normal risks of poor workmanship, or whether it was unexpected and fortuitous.”
In applying this newly formulated test, the Court of Appeal concluded that the damage to the windows was physical loss excluded as the “cost of making good faulty workmanship”, because it was not accidental or fortuitous but was directly caused by the scraping and wiping motions involved in Bristol’s cleaning work. According to the court, Bristol intentionally applied these motions to the windows, a core part of the work to be done, and the damage was not only foreseeable but highly likely.
IV. Issues on Appeal
The Exclusion Clause in the standard form builders’ risk insurance policy at issue in these appeals raises two questions that this Court must answer.
First, what standard of appellate review applies to a trial judge’s interpretation of a standard form insurance contract?
Second, what is the proper interpretation to be given to the faulty workmanship exclusion clause and the “resulting damage” exception to that exclusion contained in builders’ risk insurance policies?
A. The Standard of Review Is Correctness
In my view, the trial judge’s interpretation of the Policy should be reviewed for correctness.
These appeals present an opportunity to clarify how Sattva Capital Corp. v Creston Moly Corp., 2014 SCC 53,  2 S.C.R. 633, applies to the interpretation of standard form contracts, sometimes called contracts of adhesion.
In Sattva, Rothstein J. held that “[c]ontractual interpretation involves issues of mixed fact and law as it is an exercise in which the principles of contractual interpretation are applied to the words of the written contract, considered in light of the factual matrix” (para. 50). As a result, the palpable and overriding error standard of review applies to a trial court’s interpretation of a contract: Heritage Capital Corp. v Equitable Trust Co., 2016 SCC 19,  1 S.C.R. 306, at paras. 21-24. However, Rothstein J. acknowledged that the correctness standard of review still applies to the “rare” extricable questions of law that arise in the interpretation process, such as “the application of an incorrect principle, the failure to consider a required element of a legal test, or the failure to consider a relevant factor”: Sattva, at paras. 53 and 55, quoting King v Operating Engineers Training Institute of Manitoba Inc., 2011 MBCA 80, 270 Man. R. (2d) 63, at para. 21. This is consistent with the jurisprudence on the standard of review for questions of mixed fact and law: Housen v Nikolaisen, 2002 SCC 33,  2 S.C.R. 235, at para. 36. However, in this case, the Court of Appeal did not purport to identify an extricable question of law that arose in the interpretation process. Rather, it concluded that the interpretation of the contract itself should be reviewed for correctness, despite Sattva’s holding that contractual interpretation is a question of mixed fact and law and is owed deference on appeal: paras. 18-19.
Appellate courts have disagreed on whether this Court’s holding in Sattva on the standard of review of contractual interpretation applies to standard form contracts. Many appellate courts have held that Sattva does not apply, and have conducted correctness review:
Vallieres v Vozniak, 2014 ABCA 290, 5 Alta. L.R. (6th) 28, at paras. 11-13;
Portage LaPrairie Mutual Insurance Co. v Sabean, 2015 NSCA 53, 386 D.L.R. (4th) 449, at para. 13;
Precision Plating Ltd. v Axa Pacific Insurance Co., 2015 BCCA 277, 387 D.L.R. (4th) 281, at paras. 28-30;
Stewart Estate v 1088294 Alberta Ltd., 2015 ABCA 357, 25 Alta. L.R. (6th) 1, at para. 273, per McDonald J.A.;
MacDonald v Chicago Title Insurance Co. of Canada, 2015 ONCA 842, 127 O.R. (3d) 663, at paras. 40-41;
Monk v Farmers’ Mutual Insurance Co., 2015 ONCA 911, 128 O.R. (3d) 710, at paras. 22-24;
Daverne v John Switzer Fuels Ltd., 2015 ONCA 919, 128 O.R. (3d) 188, at paras. 12-14;
True Construction Ltd. v Kamloops (City), 2016 BCCA 173, at para. 34 (CanLII); and
Sankar v Bell Mobility Inc., 2016 ONCA 242, at para. 26 (CanLII).
In other cases, however, courts of appeal have applied Sattva and have deferred to trial courts’ interpretations of standard form contracts:
Kassburg v Sun Life Assurance Co. of Canada, 2014 ONCA 922, 124 O.R. (3d) 171, at para. 33;
Anderson v Bell Mobility Inc., 2015 NWTCA 3, 593 A.R. 79, at paras. 9 and 33-35;
Van Camp v Chrome Horse Motorcycle Inc., 2015 ABCA 83, 599 A.R. 201;
Industrial Alliance Insurance and Financial Services Inc. v Brine, 2015 NSCA 104, 392 D.L.R. (4th) 575, at paras. 40-41;
Ontario Society for the Prevention of Cruelty to Animals v Sovereign General Insurance Co., 2015 ONCA 702, 127 O.R. (3d) 581, at paras. 34-36;
Acciona Infrastructure Canada Inc. v Allianz Global Risks US Insurance Co., 2015 BCCA 347, 77 B.C.L.R. (5th) 223, at para. 35; and
GCAN Insurance Co. v Univar Canada Ltd., 2016 QCCA 500, at para. 40 (CanLII).
See also Stewart Estate, at para. 63, per Rowbotham J.A. (dissenting on this point).
I would recognize an exception to this Court’s holding in Sattva that contractual interpretation is a question of mixed fact and law subject to deferential review on appeal. In my view, where an appeal involves the interpretation of a standard form contract, the interpretation at issue is of precedential value, and there is no meaningful factual matrix that is specific to the parties to assist the interpretation process, this interpretation is better characterized as a question of law subject to correctness review.
The statements made in Sattva on the standard of review of contractual interpretation must be considered in their full context. That case concerned a complex commercial agreement between two sophisticated parties – not a standard form contract. Professor John D. McCamus has described standard form contracts as follows:
.... the document put forward will typically constitute a standard printed form that the party proffering the document invariably uses when entering transactions of this kind. The form will often be offered on a “take it or leave it” basis. In the typical case, the other party, then, will have no choice but either to agree to the terms of the standard form or to decline to enter the transaction altogether. Standard form agreements are a pervasive and indispensable feature of modern commercial life. It is simply not feasible to negotiate, in any meaningful sense, the terms of many of the transactions entered into in the course of daily life.
(The Law of Contracts (2nd ed. 2012), at p. 185)
Sattva did not consider the unique issues that standard form contracts raise.
Moreover, the Court in Sattva gave two reasons for concluding that contractual interpretation is a question of mixed fact and law subject to deferential review on appeal. As a general matter, those reasons are less compelling in the context of standard form contracts.
(1) Factual Matrix
The first reason is that the surrounding circumstances of the contract, or the factual matrix in which it was formed, are important considerations in contractual interpretation: Sattva, at para. 46. Rothstein J. stated that determining the intention of the parties is a “fact-specific goal” that requires a trial court to “read the contract as a whole, giving the words used their ordinary and grammatical meaning, consistent with the surrounding circumstances known to the parties at the time of formation of the contract”: paras. 47 and 49.
While a proper understanding of the factual matrix is crucial to the interpretation of many contracts, it is often less relevant for standard form contracts, because “the parties do not negotiate terms and the contract is put to the receiving party as a take-it-or-leave-it proposition”: MacDonald, at para. 33. Standard form contracts are particularly common in the insurance industry, as Professor Barbara Billingsley observed in General Principles of Canadian Insurance Law (2nd ed. 2014), at p. 56:
As part of its business considerations and in advance of meeting with any particular client, an insurance company decides the terms and conditions under which it is willing to provide insurance coverage for certain common types of risk. This means that, in most situations, an insurance company does not negotiate the detailed terms of insurance coverage with individual customers. Instead, before entering into any insurance agreements, an insurer typically drafts a series of pre-fabricated contracts outlining the terms upon which particular kinds of coverage will be provided. These contracts are known as “standard form policies”. The insurer then provides the appropriate standard form policy to clients purchasing insurance coverage.
Parties to an insurance contract may negotiate over matters like the cost of premiums, but the actual conditions of the insurance coverage are generally determined by the standard form contract: Billingsley, at p. 58.
My colleague Justice Cromwell accepts that, for standard form contracts, there are usually no relevant surrounding circumstances relating to negotiation (para. 106). However, he observes that other elements of the surrounding circumstances – such as the purpose of the contract, the nature of the relationship it creates, and the market or industry in which it operates – have a role in the interpretation process.
I agree that factors such as the purpose of the contract, the nature of the relationship it creates, and the market or industry in which it operates should be considered when interpreting a standard form contract. However, those considerations are generally not “inherently fact specific: Sattva, at para. 55. Rather, they will usually be the same for everyone who may be a party to a particular standard form contract. This underscores the need for standard form contracts to be interpreted consistently, a point to which I will return below.
In sum, for standard form contracts, the surrounding circumstances generally play less of a role in the interpretation process, and where they are relevant, they tend not to be specific to the particular parties. Accordingly, the first reason given in Sattva for concluding that contractual interpretation is a question of mixed fact and law – the importance of the factual matrix – carries less weight in cases involving standard form contracts.
(2) The Definitions of “Question of Law” and “Question of Mixed Fact and Law”
In Sattva, this Court gave a second reason for concluding that contractual interpretation is a question of mixed fact and law: contractual interpretation does not fit within the definition of a pure question of law. Questions of law are “about what the correct legal test is”: para. 49, quoting Canada (Director of Investigation and Research) v Southam Inc.,  1 S.C.R. 748, at para. 35. For instance, the content of a particular legal principle of contractual interpretation is a question of law. However, in interpreting contracts, courts apply the legal principles of contractual interpretation to determine the parties’ objective intentions: Sattva, at para. 49. Therefore, according to Sattva, contractual interpretation is a question of mixed fact and law, which is defined as “applying a legal standard” (the legal principles of contractual interpretation) “to a set of facts” (the words of the contract and the factual matrix): para. 49, quoting Housen, at para. 26.
In my view, however, while contractual interpretation is generally a question of mixed fact and law, in situations involving standard form contracts, it is more appropriately classified as a question of law in most circumstances.
The law of standard of review – including the distinction between questions of law and those of mixed fact and law – seeks to achieve an appropriate division of labour between trial and appellate courts in accordance with their respective roles. The main function of trial courts is to resolve the particular disputes before them: Housen, at para. 9. Appellate courts, however, “operate at a higher level of legal generality”: Association des parents ayants droit de Yellowknife v Northwest Territories (Attorney General), 2015 NWTCA 2, 593 A.R. 180, at para. 23. They ensure that “the same legal rules are applied in similar situations”, as the rule of law demands: Housen, at para. 9. Appellate courts also have a law-making function, which requires them to “delineate and refine legal rules”: ibid.
These particular functions of appellate courts – ensuring consistency in the law and reforming the law – justify reviewing pure questions of law on the standard of correctness. By contrast, appellate courts defer to findings of fact in part because they can discharge their mandate without second-guessing trial courts’ factual determinations: Housen, at paras. 11-14. For questions of mixed fact and law, the correctness standard applies to extricable errors of law (such as the application of an incorrect principle) because, again, a review on the standard of correctness is necessary to allow appellate courts to fulfill their role. However, where it is “difficult to extricate the legal questions from the factual”, appellate courts defer on questions of mixed fact and law: Housen, at para. 36; see also paras. 33-35.
In many cases, appellate courts need not review for correctness the contractual interpretation itself in order to perform their functions – namely, ensuring the consistent application of the law and reforming the law. That is because, in general, the interpretation of a contract has no impact beyond the parties to a dispute. As Rothstein J. commented in Sattva, at para. 52:
.... this Court in Housen found that deference to fact-finders promoted the goals of limiting the number, length, and cost of appeals, and of promoting the autonomy and integrity of trial proceedings (paras. 16-17). These principles also weigh in favour of deference to first instance decision-makers on points of contractual interpretation. The legal obligations arising from a contract are, in most cases, limited to the interest of the particular parties. Given that our legal system leaves broad scope to tribunals of first instance to resolve issues of limited application, this supports treating contractual interpretation as a question of mixed fact and law.
For the interpretation of many contracts, precedents interpreting similar contractual language may be of some persuasive value. However, it is the intentions of the particular parties, as reflected in the particular contractual wording at issue and informed by the surrounding circumstances of the contract, that predominate, and “[i]f that intention differs from precedent, the intention will govern and the precedent will not be followed”: G. R. Hall, Canadian Contractual Interpretation Law (3rd ed. 2016), at pp. 129-30; see also Tenneco Canada Inc. v British Columbia Hydro and Power Authority, 1999 BCCA 415, 126 B.C.A.C. 9, at para. 43.
These teachings, however, do not necessarily apply in cases involving standard form contracts, where a review on the standard of correctness may be necessary for appellate courts to fulfill their functions. Standard form contracts are “highly specialized contracts that are sold widely to customers without negotiation of terms”: MacDonald, at para. 37. In some cases, a single company, such as a bank or a telephone service provider, may use its own standard form contract with all of its customers: Monk, at para. 23. In others, a standard form agreement may be common throughout an entire industry: Precision Plating, at para. 28. Either way, the interpretation of the standard form contract could affect many people, because “precedent is more likely to be controlling” in the interpretation of such contracts: Hall, at p. 131. It would be undesirable for courts to interpret identical or very similar standard form provisions inconsistently, without good reason. The mandate of appellate courts – “ensuring the consistency of the law” (Sattva, at para. 51) – is advanced by permitting appellate courts to review the interpretation of standard form contracts for correctness.
Indeed, consistency is particularly important in the interpretation of standard form insurance contracts. In Co-operators Life Insurance Co. v Gibbens, 2009 SCC 59,  3 S.C.R. 605, at para. 27, Binnie J. recognized that “‘courts will normally be reluctant to depart from [authoritative] judicial precedent interpreting the policy in a particular way’ .... where the issue arises subsequently in a similar context, and where the policies are similarly framed”, because both insurance companies and customers benefit from “[c]ertainty and predictability”. And where an insurance policy is ambiguous, courts “strive to ensure that similar insurance policies are construed consistently”: Progressive Homes Ltd. v Lombard General Insurance Co. of Canada, 2010 SCC 33,  2 S.C.R. 245, at para. 23.
The definition of questions of law – “questions about what the correct legal test is” (Southam, at para. 35) – does not preclude classifying some questions of contractual interpretation as questions of law. There is no bright-line distinction between questions of law and those of mixed fact and law. Rather, “the degree of generality (or ‘precedential value’)” is the key difference between the two types of questions: Sattva, at para. 51. As Iacobucci J. stated in Southam, at para. 37:
If a court were to decide that driving at a certain speed on a certain road under certain conditions was negligent, its decision would not have any great value as a precedent. In short, as the level of generality of the challenged proposition approaches utter particularity, the matter approaches pure application, and hence draws nigh to being an unqualified question of mixed law and fact .... Of course, it is not easy to say precisely where the line should be drawn; though in most cases it should be sufficiently clear whether the dispute is over a general proposition that might qualify as a principle of law or over a very particular set of circumstances that is not apt to be of much interest to judges and lawyers in the future.
Contractual interpretation is often the “pure application” of contractual interpretation principles to a unique set of circumstances. In such cases, the interpretation is not “of much interest to judges and lawyers in the future” because of its “utter particularity”. These questions of contractual interpretation are appropriately classified as questions of mixed fact and law, as the Court explained in Sattva .
However, the interpretation of a standard form contract could very well be of “interest to judges and lawyers in the future”. In other words, the interpretation itself has precedential value. The interpretation of a standard form contract can therefore fit under the definition of a “pure question of law”, i.e., “questions about what the correct legal test is”: Sattva, at para. 49; Southam, at para. 35. Establishing the proper interpretation of a standard form contract amounts to establishing the “correct legal test”, as the interpretation may be applied in future cases involving identical or similarly worded provisions.
My colleague Cromwell J. suggests that the interpretation of a standard form contract will not be of much precedential value because “its application in other cases will ultimately be decided on a case-by-case basis in light of the particular circumstances of the particular case” (para. 120). I respectfully disagree. Settling on a consistent interpretation of a standard form provision is useful. Of course, the result of applying the interpretation in future cases will depend on the facts of those cases. The facts are for the trial judge to find, and those findings will be owed deference.
For instance, in this case, the Court of Appeal interpreted the Exclusion Clause as excluding damages physically or systemically connected to the faulty work. For the reasons I will give below, I am of the view that the Exclusion Clause excludes only the cost of redoing the faulty work. These are two different interpretations of the same standard form language. Selecting one interpretation over the other as correct will give parties certainty and predictability. This is true even though what constitutes the cost of redoing the faulty work will depend on the facts of future cases.
(3) Conclusion on Standard of Review
Sattva should not be read as holding that contractual interpretation is always a question of mixed fact and law, and always owed deference on appeal. I would recognize an exception to Sattva’s holding on the standard of review of contractual interpretation. Where, like here, the appeal involves the interpretation of a standard form contract, the interpretation at issue is of precedential value, and there is no meaningful factual matrix specific to the particular parties to assist the interpretation process, this interpretation is better characterized as a question of law subject to correctness review.
These criteria are met in the present case, so the standard of review applicable to the trial judge’s interpretation of the Policy is correctness. The trial judge’s underlying factual findings remain subject to deferential review, as mentioned above.
Depending on the circumstances, however, the interpretation of a standard form contract may be a question of mixed fact and law, subject to deferential review on appeal. For instance, deference will be warranted if the factual matrix of a standard form contract that is specific to the particular parties assists in the interpretation. Deference will also be warranted if the parties negotiated and modified what was initially a standard form contract, because the interpretation will likely be of little or no precedential value. There may be other cases where deferential review remains appropriate. As Iacobucci J. recognized in Southam, the line between questions of law and those of mixed fact and law is not always easily drawn. Appellate courts should consider whether “the dispute is over a general proposition” or “a very particular set of circumstances that is not apt to be of much interest to judges and lawyers in the future” (para. 37).
B. The Exclusion Clause
(1) Rules Governing the Interpretation of the Policy
The parties agree that the governing principles of interpretation applicable to insurance policies are those summarized by Rothstein J. in Progressive Homes. The primary interpretive principle is that where the language of the insurance policy is unambiguous, effect should be given to that clear language, reading the contract as a whole: para. 22, citing Non‑Marine Underwriters, Lloyd’s of London v Scalera, 2000 SCC 24,  1 S.C.R. 551, at para. 71.
Where, however, the policy’s language is ambiguous, general rules of contract construction must be employed to resolve that ambiguity. These rules include that the interpretation should be consistent with the reasonable expectations of the parties, as long as that interpretation is supported by the language of the policy; it should not give rise to results that are unrealistic or that the parties would not have contemplated in the commercial atmosphere in which the insurance policy was contracted, and it should be consistent with the interpretations of similar insurance policies. See
Progressive Homes, at para. 23, citing Scalera, at para. 71;
Gibbens, at paras. 26-27; and
Consolidated‑Bathurst Export Ltd. v Mutual Boiler and Machinery Insurance Co.,  1 S.C.R. 888, at pp. 900-902.
Only if ambiguity still remains after the above principles are applied can the contra proferentem rule be employed to construe the policy against the insurer:
Progressive Homes, at para. 24, citing Scalera, at para. 70;
Gibbens, at para. 25; and
Consolidated-Bathurst, at pp. 899-901.
Progressive Homes provides that a corollary of this rule is that coverage provisions in insurance policies are interpreted broadly, and exclusion clauses narrowly.
It is also important to bear in mind this Court’s guidance in Progressive Homes on the “generally advisable” order in which to interpret insurance policies (para. 28). Although that case involved commercial general liability policies and not builders’ risk policies, the two types of policies share a similar alternating structure: they set out the type of coverage followed by specific exclusions, with some exclusions containing exceptions. As such, the insured has the onus of first establishing that the damage or loss claimed falls within the initial grant of coverage. The parties in these appeals have conceded that this particular onus has been met: trial judge’s reasons, at para. 9. The onus then shifts to the insurer to establish that one of the exclusions to coverage applies. If the insurer is successful at this stage, the onus then shifts back to the insured to prove that an exception to the exclusion applies: see Progressive Homes, at paras. 26-29 and 51. Contrary to the Court of Appeal’s statement at para. 26 of its reasons that the exclusion and exception in this case must be interpreted “symbiotically”, I see no reason to depart from the generally accepted order of interpretation in analyzing the Policy and the Exclusion Clause.
(2) The Court of Appeal’s Approach to the Exclusion Clause
Before engaging in the interpretation of the Exclusion Clause, I believe it necessary to properly set out the Court of Appeal’s reasoning and explain why its new physical or systemic connectedness test was unnecessary.
At paras. 29 and 48 of its reasons, the Court of Appeal explained that because the base coverage under clause 2 of the Policy is for “physical loss or damage”, it follows that the Exclusion Clause needs to exclude from coverage some physical loss. In the Court of Appeal’s opinion, a different reading of the Exclusion Clause would risk rendering it redundant. Under this view, the “cost of making good faulty workmanship” cannot be limited to the cost of redoing the faulty work. Rather, that exclusion must be construed more broadly to also exclude from coverage some type of physical loss or damage.
As mentioned above, the Court of Appeal’s acceptance of this initial premise led it to search for a dividing line between physical damage that is part of the “cost of making good” and therefore excluded from coverage, and physical damage that is “resulting damage” and therefore covered as an exception to the exclusion. In its quest to establish this dividing line, the court fashioned a new test of “degree of physical or systemic connectedness”, which it said was “the key to determining the boundary between ‘making good faulty workmanship’ and ‘resulting damage’”: para. 50.
In my respectful view, the premise from which the Court of Appeal proceeded is flawed. The “faulty workmanship” exclusion need not encompass physical damage. Although “[e]xclusions should .... be read in light of the initial grant of coverage” (Progressive Homes, at para. 27; see also M. G. Lichty and M. B. Snowden, Annotated Commercial General Liability Policy (loose-leaf), at p. 1-10), this Court has stressed that “perfect mutual exclusivity [between exclusions and the initial grant of coverage] in an insurance contract is not required”: Progressive Homes, at para. 40.
Bearing the above-mentioned principle in mind, the Policy in this case contains exclusions that do not pertain to “physical loss or damage” otherwise covered under clause 2. For instance, clause 4(A)(a) of the Policy excludes from coverage “[a]ny loss of use or occupancy or consequential loss of any nature howsoever caused including penalties for non-completion of or delay in completion of contract or non-compliance with contract conditions”. This exclusion deals with a form of pure economic loss stemming from contractual breach, not physical loss or damage. Additionally, clause 28 of the “standard conditions” section excludes “costs, fines, penalties or expenses” imposed by governments under environmental legislation. This also does not relate to the Policy’s base coverage for physical loss or damage.
As such, perfect mutual exclusivity is neither provided for under the Policy nor should it be required when interpreting the Exclusion Clause. The Court of Appeal consequently erred by approaching its analysis of the Exclusion Clause from a premise that was not supported by the text of the Exclusion Clause or the Policy as a whole. Adopting this premise led the Court of Appeal down an improper analytical path toward establishing a new and unnecessary test. Indeed, as I will explain below, the general rules of contractual interpretation provide the answer to whether the damage to the Tower’s windows is covered under the Policy.
(3) Interpretation of the Exclusion Clause and the Policy
(a) The Language of the Exclusion Clause Is Ambiguous
The Insureds argue that the plain language of the Exclusion Clause, read in the context of the Policy as a whole, is unambiguous. They say it leads to the conclusion that only the cost of redoing the faulty work – in this case, cleaning the windows – is excluded from coverage. The consequences of the faulty work – here, the damage to the windows, necessitating their replacement – are covered as “resulting damage”.
The Insurers similarly argue that the Exclusion Clause is unambiguous, yet they arrive at a different conclusion as to its meaning. They say that which is excluded is not only the cost of redoing the faulty work, but also the cost of repairing that part of the insured property or project that is the subject of the faulty work. That which is covered as “resulting damage” is consequential damage to some other part of the insured property or project. They point to the case law in support, contending that the courts have consistently interpreted the language of the Exclusion Clause to bear this meaning. Accordingly, in this case, the Insurers say the Policy excludes both the cost of recleaning the windows and the cost of replacing the windows, the subject of the faulty work.
I am of the view that the language of the Exclusion Clause slightly favours the interpretation advanced by the Insureds, but is nonetheless ambiguous. The word “damage” figures only in the exception to the Exclusion Clause; it is not included in the language setting out the exclusion itself, i.e., the “cost of making good faulty workmanship”. As such, “making good faulty workmanship” can, on its plain, ordinary and popular meaning, be construed as redoing the faulty work, and “resulting damage” can be seen as including damages resulting from such faulty work.
That said, the language of the Exclusion Clause does not clearly point to one interpretation of “cost of making good faulty workmanship” and “resulting damage” over the other. The Policy does not define these terms. The general coverage provisions, clauses 1 and 2, do not resolve the ambiguity, and neither do the other provisions in the Policy.
Therefore, we must look to the general principles of contract interpretation. As I will detail below, the application of these principles points to one interpretation that is consistent with the reasonable expectations of the parties and commercial reality: the faulty workmanship exclusion serves to exclude from coverage only the cost of redoing the faulty work, as the resulting damage exception covers costs or damages apart from the cost of redoing the faulty work. As such, excluded under the Policy is the cost of recleaning the windows, but the damage to the windows and therefore the cost of their replacement is covered. This is consistent with previous interpretations of similar clauses in the jurisprudence. Indeed, as I explain below, I disagree with the Insurers’ contention that the case law consistently supports their interpretation of the Exclusion Clause.
In light of this determination, it is not necessary to turn to the contra proferentem rule to answer the second issue raised in these appeals.
(b) Reasonable Expectations of the Parties
Parties’ reasonable expectations with respect to the meaning of a contractual provision can often be gleaned from the circumstances surrounding the contract’s formation: Sattva, at paras. 46-47. However, as discussed above, there is no factual matrix here that would assist in ascertaining the parties’ understanding of and intent regarding the Exclusion Clause. The Policy is a standard form contract. And, as the Court of Appeal noted at para. 15 of its reasons, there is no evidence that the parties gave any thought to the cleaning of the windows, the relationship of faulty workmanship to resulting damage, or anything else that would help in determining their reasonable expectations.
Therefore, in my view, the purpose behind builders’ risk policies is crucial in determining the parties’ reasonable expectations as to the meaning of the Exclusion Clause. In a nutshell, the purpose of these polices is to provide broad coverage for construction projects, which are singularly susceptible to accidents and errors. This broad coverage – in exchange for relatively high premiums – provides certainty, stability, and peace of mind. It ensures construction projects do not grind to a halt because of disputes and potential litigation about liability for replacement or repair amongst the various contractors involved. In my view, the purpose of broad coverage in the construction context is furthered by an interpretation of the Exclusion Clause that excludes from coverage only the cost of redoing the faulty work itself – in this case, the cost of recleaning the windows.
“The raison d’être of insurance is coverage”: D. Boivin, Insurance Law (2nd ed. 2015), at p. 288. The purpose of builders’ risk policies in particular is to offer broad coverage, which benefits both insureds and insurers:
Urbanization and industrialization in the past 100 years have made the concept of an insurance policy covering all conceivable risks advantageous to both insureds and their insurers. The insured benefits from the extensive nature and scope of the coverage, and insurers benefit from the economies of managing and marketing a policy which, in terms of its scope, has certainty. For these reasons, the “all risk policy,” which creates a special type of coverage extending to many risks not customarily covered under other types of insurance policies, is attractive to both the insurance industry and consumers.
(E. A. Dolden, “All Risk and Builders’ Risk Policies: Emerging Trends” (1990-91), 2 C.I.L.R. 341, at pp. 341-42)
This Court stated in Commonwealth Construction Co. v Imperial Oil Ltd.,  1 S.C.R. 317, that the purpose of builders’ risk policies is to provide certainty and stability by granting coverage that reduces the need for private law litigation. The Court also recognized the complexity of industrial life and large-scale construction projects that involve many different individual contractors [p. 328]:
As already noted, the multi-peril policy under consideration is called .... a course of construction insurance. In England, it is usually called a “Contractors’ all risks insurance” and in the United States, it is referred to as “Builders’ risk policy”. Whatever its label, its function is to provide to the owner the promise that the contractors will have the funds to rebuild in case of loss and to the contractors the protection against the crippling cost of starting afresh in such an event, the whole without resort to litigation in case of negligence by anyone connected with the construction, a risk accepted by the insurers at the outset. This purpose recognizes the importance of keeping to a minimum the difficulties that are bound to be created by the large number of participants in a major construction project, the complexity of which needs no demonstration. It also recognizes the realities of industrial life.
Although such policies are said to insure against all risks, this description is not entirely accurate. As a general rule, insurance offers protection only for fortuitous contingent risk: Progressive Homes, at para. 45. Moreover, builders’ risk policies contain various exclusions, meaning indemnity is precluded in many circumstances of fortuitous loss: Dolden, at pp. 342-44.
Despite these qualifiers, builders’ risk construction policies are the norm, if not a requirement, on construction sites in Canada. In purchasing these policies, “contractors believe indemnity will be available in the event of an accident or damage on the construction site arising as a result of a party’s carelessness or negligent acts”, which are the most common source of loss on construction sites: Dolden, at pp. 345-46. And, in selling these policies, insurers
are prepared to insure risks relating to problems caused by faulty .... workmanship, but they are not prepared to insure the quality of .... the workmanship in a construction project per se. The argument is that the contractor is responsible for doing [its] job right and the insurance company is not there to provide compensation for inadequate performance by a contractor of the very work the contractor agreed to perform.
(Canadian College of Construction Lawyers, report of the Insurance & Surety Committee, “‘Covered for What?’: Faulty Materials and Workmanship Coverage Under Canadian Construction Insurance Policies” (2007), 1 J.C.C.C.L. 101, at p. 104)
Consequently, an interpretation of the Exclusion Clause that precludes from coverage any and all damage resulting from a contractor’s faulty workmanship merely because the damage results to that part of the project on which the contractor was working would, in my view, undermine the purpose behind builders’ risk policies. It would essentially deprive insureds of the coverage for which they contracted.
In my opinion, therefore, the Insureds’ position on the meaning of the Exclusion Clause better reflects and promotes the purpose of builders’ risk policies. In the words of this Court in Commonwealth Construction, it keeps “to a minimum the difficulties .... created by the large number of participants in a major construction project” and “recognizes the realities of industrial life” (p. 328). Their position finds additional support in some of this Court’s other comments in that case, at pp. 323-24, where it was emphasized that these policies exist to account for the fact that work of different contractors overlaps in a complex construction site and “there is ever present the possibility of damage by one tradesman to the property of another and to the construction as a whole”.
Further support for the Insureds’ position can be found in commentary contending that all-risk coverage under builders’ risk policies was intended to be broad, and the faulty workmanship exclusion narrow. For instance, Maurice G. Audet has discussed the original intent of the exclusion, reviewing the case law as well as annotated insurance policies and manuals: “Part II – Insurance” (2002), 12 C.L.R. (3d) 100; and “All Risks – a promise made or a promise broken?” (1983), 50:10 Canadian Underwriter 34, at pp. 40-42 and 93-96. He concludes that the faulty workmanship, materials and design exclusion was meant to be narrow, to exclude only the cost of replacing the fault or defect but to provide coverage for damage caused by it.
Other authors have remarked that the trend in the common law jurisprudence interpreting builders’ risk policies has been to widen the scope of the above-mentioned exclusion or narrow the ambit of the exception to the exclusion. See e.g. Dolden, at pp. 350 and 358; R. B. Reynolds and S. C. Vogel, A Guide to Canadian Construction Insurance Law (2013), at pp. 140 and 150; and P.-S. Poitras, “L’assurance et l’industrie de la constructionˮ, in Service de la formation permanente du Barreau du Québec, vol. 147, Développements récents en droit des assurances (2001), 181, at p. 195. I would not go so far as to question the jurisprudence. Consistency of interpretation is important, and these judicial interpretations have undoubtedly shaped parties’ reasonable expectations with respect to builders’ risk policies and their exclusion clauses. I simply note that the interpretation of the Exclusion Clause advanced by the Insureds in these appeals best reflects the original intent of such exclusion clauses, as compared to the interpretation advanced by the Insurers.
It should be mentioned that the service contract between Station Lands and Bristol has no bearing on the reasonable expectations of the parties to the Policy with respect to the meaning of the Exclusion Clause and whether the damage to the windows would be covered. The Insurers and Ledcor were not parties to that service contract, and it was entered into on June 16, 2011, almost three years after the Policy’s effective date of June 27, 2008. At most, the service contract could shed light on Station Lands’ understanding of the Policy and the Exclusion Clause, as it was a party to both. Still, the service contract was itself based on a slightly modified standard form contract published by the Canadian Construction Association.
Despite the service contract’s irrelevance to the parties’ reasonable expectations, at various points in its reasons the Court of Appeal seemed to use it to bolster its interpretation of the Exclusion Clause. For instance, at para. 35, the court determined it was artificial to draw the dividing line between the “cost of making good faulty workmanship” and “resulting damage” as falling between Bristol’s work and the work of other contractors, in part because under the service contract Bristol was responsible for repairing damage it did to the work of other contractors. Further, at para. 49, the court highlighted that an interpretation of “making good faulty workmanship” that included redoing the work and fixing the damage directly caused by the work was consistent with the service contract, because, again, the contract required Bristol to repair damage it did to the work of other contractors.
Even if the service contract were relevant to the reasonable expectations of the parties to the Policy, there are two other reasons why the Court of Appeal’s reliance on it – to the extent that there was such reliance – was problematic. First, a contractor’s or subcontractor’s stipulated responsibility under its work contract to repair or pay for certain damage does not necessarily preclude coverage under a builders’ risk policy, as recognized by this Court in Commonwealth Construction , at p. 330. For instance, insurance policies often have deductible amounts. In fact, clause 5 of the Policy provides that the Insurers’ liability is limited to the amount by which the loss or damage exceeds the deductible amount, and clause GC 11.1.6 of the service contract provides that Bristol shall be responsible for deductible amounts under the various insurance policies except where such amounts may be excluded from its responsibility by other terms of the contract, including those adverted to by the Court of Appeal. The stipulation in the service contract could thus serve to confirm responsibility for that deductible amount, even where loss or damage is covered under the Policy. In other words, the contract stipulation does not necessarily suggest the parties expected that Bristol would ultimately bear the entire cost of damages it caused to the work of other contractors.
Second, even if the stipulation did indicate such an expectation, the Court of Appeal’s new physical and systemic connectedness test does not reflect it. Under the court’s new test, and using its language, certain unforeseeable, collateral damage to areas on which Bristol was not working would likely be covered under the resulting damage exception in the Exclusion Clause. Yet Bristol would also be responsible for this damage under the service contract, which makes no such distinction with respect to the foreseeability or remoteness of the damage caused. In effect, there would be dual responsibility for payment, under both the Policy and the service contract, even though, as discussed above, the Court of Appeal stated it would be artificial to draw the dividing line where such dual responsibility would result.
(c) No Unrealistic Results
In discussing the interpretation of insurance policies in Consolidated-Bathurst, at pp. 901-2, Estey J. stressed the need to avoid interpretations that would bring about unrealistic results or results that the parties would not have contemplated in the commercial atmosphere in which they sold or purchased the policy. The interpretation should respect the intentions of the parties and “their objective in entering into the commercial transaction in the first place”, as well as “promot[e] a sensible commercial result” (p. 901). See also Guarantee Co. of North America v Gordon Capital Corp. ,  3 S.C.R. 423, at para. 62, where this Court restated the importance of commercial reality, albeit in a different context.Interpreting the Exclusion Clause to preclude from coverage only the cost of redoing the faulty work aligns with commercial reality and leads to realistic and sensible results, given both the purpose underlying builders’ risk policies and their spreading of risk on construction projects.
As already discussed above, the interpretation advanced by the Insureds in these appeals best fulfills the broad coverage objective underlying builders’ risk policies. These policies are commonplace on construction projects, where multiple contractors work side by side and where damage to their work or the project as a whole commonly arises from faults or defects in workmanship, materials or design. In this commercial reality, a broad scope of coverage creates certainty and economies for both insureds and insurers. In my opinion, it is commercially sensible in this context for only the cost of redoing a contractor’s faulty work to be excluded under the faulty workmanship exclusion. Such an interpretation strikes the right balance between the two undesirable extremes described by Estey J. in Consolidated-Bathurst, at pp. 901-2: “.... the courts should be loath to support a construction which would either enable the insurer to pocket the premium without risk or the insured to achieve a recovery which could neither be sensibly sought nor anticipated at the time of the contract”. Under the Policy, the Insurers did not undertake to cover the “cost of making good faulty workmanship”, but they did promise to cover “physical damage [that] results” from that “faulty workmanship”. It can hardly be said that recovery for the damages to the Tower’s windows in the circumstances of this case could not have been sensibly sought or anticipated when the Policy was purchased.
Furthermore, such an interpretation does not, in my view, transform the insurance policy into a construction warranty. It does not inappropriately spread risk, nor would it allow or encourage contractors to perform their work improperly or negligently. Importantly, Bristol is precluded from receiving initial payment for its faulty work and then receiving further additional payment to repair or replace its faulty work. SeeC. Brown, Insurance Law in Canada (loose-leaf), at p. 20-31; and Privest Properties Ltd. v Foundation Co. of Canada Ltd. (1991),57 B.C.L.R. (2d) 88 (S.C.).The cost of redoing faulty or improper work is excluded from coverage. The cost can be sizeable; in the instant appeals, for example, Bristol’s contract price for cleaning the windows was $45,000.
The Insurers argue that accepting the Insureds’ position would tether the application of the resulting damage exception to how the work is divided among various contractors on a project, a result which they say would not make commercial sense. This argument echoes the Court of Appeal’s concerns at para. 40 of its reasons:
This approach might create an incentive to artificially divide up the work as finely as possible, as then the maximum amount of damage would be covered by insurance. On the other hand, it would be dangerous for the owner to hire a single contractor to do all the work, as then nothing would be covered.
With respect, I do not find this persuasive. It is premised on a theoretical concern that does not reflect the commercial reality of construction sites on the ground. In my view, it is unreasonable to expect that the owner of a property or the general contractor on a construction site will divide up work exclusively on the basis of potential coverage under their insurance policy. Many other considerations, such as costs, subcontractor expertise and the risk of delay, will likely be more relevant in deciding how to allocate work.
I also note that interpreting the Exclusion Clause as precluding from coverage only the cost of redoing the faulty work breaks no new ground in the world of insurance, as it mirrors the approach courts have adopted when construing similar exclusions to comprehensive general liability insurance policies. These policies cover the risk that the insured’s work might cause bodily injury or property damage. However, they generally contain a “work product” or “business risk” exception, which excludes from coverage the cost of redoing the insured’s work: “Covered for What?”, at p. 122.
(d) Ensuring Consistent Interpretation
The purpose of builders’ risk policies and the need to prevent unrealistic results point to an interpretation of the Exclusion Clause that would exclude from coverage only the cost of redoing the cleaning work. Such an interpretation of the Exclusion Clause is also consistent with case law. Though the Court of Appeal stated, at para. 46 of its reasons, that “numerous cases .... hold that the exclusion is not limited to the cost of re-doing the faulty work, but also extends to the cost of repairing the thing actually being worked on”, with respect, I am of the view that many of these faulty workmanship and faulty design decisions can be read as limiting the faulty workmanship exclusion to only the cost of redoing the faulty work. As these cases are highly fact-specific, the results that courts reach will be largely dictated by the particular circumstances of each case. More specifically, whether certain damage falls within the resulting damage exception to the faulty workmanship exclusion will greatly depend on the scope of the contractual obligation pursuant to which the faulty workmanship was carried out.
In the appeals before us, Bristol’s obligation under its service contract with Station Lands was limited to cleaning the Tower’s windows after they had been properly installed. Redoing Bristol’s faulty work did not require Bristol to install windows in good condition. As such, the cost of the windows’ replacement represents “resulting damage” and is covered under the Policy. Conversely, if Bristol had been responsible for the windows’ installation, and the windows had been damaged in the course of the installation process, the damage done to the windows in such circumstances would not have constituted “resulting damage”. Indeed, redoing the faulty work would have required installing windows in good condition, as per Bristol’s (hypothetical) contractual obligation.
I will now review some of the faulty workmanship cases cited by the Court of Appeal to illustrate how an interpretation that limits the scope of the faulty workmanship exclusion to the cost of redoing the faulty work is consistent with the jurisprudence.
In Sayers & Associates Ltd. v Insurance Corp. of Ireland Ltd. (1981), 126 D.L.R. (3d) 681 (Ont. C.A.), an electrical subcontractor was to install two bus ducts in connection with the construction of an office building in Toronto. Because of the subcontractor’s failure to take adequate protective measures, which constituted faulty workmanship, rain water that had come into contact with concrete penetrated the ducts, which in turn caused a malfunction. The subcontractor argued that the damage to the equipment by water was “damage resulting from .... faulty workmanship” so as to come within the exception to the exclusion. The Ontario Court of Appeal disagreed, writing as follows, at pp. 684-85:
In the present case the “fault” that underlay the “faulty workmanship” was the failure of the appellant to take protective measures; but by the terms of its contract its “work” was to install the electrical equipment and to keep it dry and clean until the contract was completed. It would be taking too narrow a view of the case to isolate one part of the work from the total contractual obligation. The damage to the equipment was the product of the failure to take protective measures, and so that fault rendered the appellant’s performance of its contractual obligations “faulty workmanship”. The damage to the ducts and the switching gear was not, therefore, “damage resulting from such faulty .... workmanship ....”, so as to come within the exception to the exclusion.
The court’s statement is clear: since the subcontractor was contractually required to install the electrical equipment and keep it dry, and its failure to take adequate protective measures resulted in it failing to comply with said contractual obligations, the damage to the equipment could not be considered resulting damage.
In Ontario Hydro v Royal Insurance,  O.J. No. 215 (QL) (H.C.J.), a contractor was responsible for designing a boiler system, acquiring the material and supervising the commissioning of the boiler. After installation, as part of the testing of the boiler system, an acid wash of the superheater and reheater was carried out. But the acid wash was done improperly, ruining the reheater and resulting in extensive cracking of the tubing in the boilers. The contractor argued that the faulty workmanship exclusion served to exclude from coverage only the cost of the redoing the wash, and not the cost of replacing the tubing, which it said constituted “resultant damage”. The court disagreed, holding that the “cost of making good the improper workmanship is the cost of replacing the tubing which was the object of [the] procedure”: para. 37. In my view, the court reached this conclusion because replacing the tubing was necessary for the contractor to fulfill its contractual obligation to design the boiler system, to acquire the material and to supervise the commissioning of the boiler. Thus, the cost of redoing the work encompassed the cost of replacing the tubing.
In Bird Construction Co. v United States Fire Insurance Co. (1985), 24 D.L.R. (4th) 104 (Sask. C.A.), the collapse of a truss, caused by the subcontractor’s faulty workmanship in failing to properly erect it, resulted in the subcontractor failing to comply with its contractual obligation to fabricate and erect a truss in good condition. The cost of repairing the truss formed part of the cost of redoing the work, and thus did not fall within the resulting damage exception. The same reading can also be made of Greene v Canadian General Insurance Co. (1995), 133 Nfld. & P.E.I.R. 151 (Nfld. C.A.).
Though this interpretation of the resulting damage exception to the faulty workmanship exclusion may, at first glance, seem to run contrary to the interpretation generally given to it by courts in faulty design cases, it is actually consistent with those cases.
It is true that, in faulty design cases, courts generally interpret the resulting damage exception as encompassing damage done to something other than the property which is faultily designed. Such language may thus appear to be more closely in line with the physical or systemic connectedness test established by the Court of Appeal, as exclusion from coverage may appear to depend on whether the damage has been done to the very thing being worked on or to something else. Only in the latter case would the loss qualify as “resulting damage”. For instance, in British Columbia v Royal Insurance Co. of Canada (1991), 7 B.C.A.C. 172, the British Columbia Court of Appeal wrote that “[d]amage for faulty or improper design encompasses all the damage to the very thing that was designed faultily or improperly. Resultant damage is damage to some part of the insured property other than the part of the property that was faultily designed”: para. 11; see also Algonquin Power (Long Sault) Partnership v Chubb Insurance Co. of Canada (2003), 50 C.C.L.I. (3d) 107 (Ont. S.C.J.), at para. 204.
These decisions, however, are not inconsistent with holding that the faulty workmanship exclusion precludes from coverage only the cost of redoing the faulty work. Indeed, in faulty design cases, a contractor is obligated to design a given item, with the design being integral to the whole of that item. Thus, the cost of repairing the damages caused to that item will be included within the cost of redoing the faulty work, and the resulting damage exception will necessarily apply to damages caused to items other than the item being designed. As held in Simcoe & Erie General Insurance Co. v Royal Insurance Co. of Canada (1982), 36 A.R. 553 (Q.B.), at para. 34:
.... the total contractual obligation of [the engineer] was to design and supervise the construction of a bridge required by the [city]. The damage to the structure that [the engineer] first designed was the product of its failure to properly design a bridge, which in turn prevented it from properly performing its contractual obligations. It follows therefore that the contract was not performed until a stable bridge was constructed.
In any event, I disagree with the Insurers’ contention that the case law systematically supports one interpretation of the faulty workmanship exclusion. Though the jurisprudence addressing the resultant damage exception has generally interpreted it narrowly (S. C. Vogel, “Recent Developments in Construction Insurance Law”, in Review of Construction Law: Recent Developments (2012), 169, at p. 184), courts have not always been consistent in construing the exception, and parties cannot therefore adequately predict what sort of damage will or will not be caught by the exclusion: L. Ricchetti and T. J. Murphy, Construction Law in Canada (2010); “Covered for What?”, at p. 106; and Poitras, at p. 195. Even the Court of Appeal acknowledged the inconsistency in the jurisprudence at para. 47 of its reasons, citing as an example Foundation Co. of Canada v Aetna Casualty Co, of Canada,  I.L.R. ¶ 1-757 (Alta. S.C.).
Additionally, in Commercial union cie d’assurance du Canada v Pentagon Construction Canada Inc.,  R.J.Q. 1399 (C.A.), the Quebec Court of Appeal broadened the resultant damage exception and, thus, narrowed the faulty workmanship exclusion, commenting that damage to the thing that the faulty contractor is responsible for building is covered. Though these comments were made in obiter, as the court had already concluded that the workmanship at issue was not faulty, they help demonstrate that the case law on the interpretation of the faulty workmanship exclusion and resulting damage exception is not unanimous.
(e) Conclusion on the Interpretation of the Exclusion Clause
As outlined above, the language of the Exclusion Clause, read in light of the Policy as a whole, does not provide a clear answer to the question raised before us. That said, the parties’ reasonable expectations, informed largely by the purpose of builders’ risk policies, point to the faulty workmanship exclusion serving to exclude from coverage only the cost of redoing the faulty work. This interpretation aligns with commercial realities and is consistent with prior jurisprudence. In the circumstances of this case, the cost of redoing the faulty work is the cost of recleaning the windows – both parties agree that the recleaning falls under the Policy’s “cost of making good faulty workmanship” exclusion. The Insureds, however, have met their onus of demonstrating that the cost of replacing the damaged windows is covered under the “resulting damage” exception to that exclusion.
In any event, even if I were to determine that the general rules of contractual interpretation do not clarify the ambiguous Exclusion Clause, I would reach the same conclusion on the basis of the contra proferentem rule.
I would allow the appeals with costs throughout.
(delivered a concurring judgment)
I agree with the disposition of these appeals proposed by my colleague, Justice Wagner, in his carefully crafted and comprehensive reasons. However, I respectfully do not agree with him on two points: the applicable standard of appellate review and whether the contractual clause that we must interpret is ambiguous. As I will explain, our very recent decision in Sattva Capital Corp. v Creston Moly Corp., 2014 SCC 53,  2 S.C.R. 633, decides that the standard of review here is palpable and overriding error, not correctness, and in my opinion the trial judge did not err in finding that the clause is ambiguous. Like my colleague, I would therefore allow the appeals with costs.
The merits of the appeals turn on a straightforward question. When window cleaners destroy the windows they are supposed to clean, does the cost of replacing the windows fall within the expression the “cost of making good faulty workmanship”, in which case it is excluded from coverage provided by the insurance policy we must interpret in this case, or does that cost fall within the expression “physical damage .... resulting” from the faulty workmanship, in which case it is covered by the policy?
II. The Standard of Review
A. Sattva Brought Appellate Review in Contract Cases Within the Court’s General Framework for Appellate Review in Civil Cases
(1) The Standard of Review in Civil Appeals Turns on the Nature of the Question Under Review
The standard of review aspect of the Court’s judgment in Sattva must be understood in the context of the Court’s broader jurisprudence on standard of review in civil appeals. At least since Housen v Nikolaisen, 2002 SCC 33,  2 S.C.R. 235, the standard of appellate review has turned on the nature of the question being reviewed. Questions of law are reviewed for correctness and questions of fact for palpable and overriding error. Housen also holds that applying a legal standard to the facts is a mixed question of law and fact and is generally reviewable on appeal for palpable and overriding error: paras. 26-37. So, in a negligence case such as Housen, that is the standard that generally governs appellate review of the trial court’s application of the legal standard of negligence to the evidence.
I say “generally” because Housen recognized that this would not always be so. In some cases, the trial court’s application of a legal standard to the facts will attract correctness review on appeal. This will be the case when the basis for a finding under review can be traced to a pure legal error, such as a wrong characterization of the legal standard or the failure to consider a required element of the applicable standard. In cases of this sort, the reviewing court can “extricate” a purely legal question from the trial court’s analysis and having done so, apply to that purely legal question the correctness standard of appellate review: paras. 31-33. These sorts of cases are fairly rare, however. As Housen cautioned, it is often difficult to extricate the legal questions from the factual and therefore appellate courts should not be quick to find extricable legal errors in the trial court’s application of a legal standard to the facts: para. 36.
(2) Sattva Brought Contract Appeals Within this Framework
I review these basic points of Housen because, as I see it, the Court’s decision in Sattva brought appellate review in contract cases within this general standard of review framework. To put it in Housen’s terms, applying the text of a contract to a particular fact situation involves applying the legal standard set by the contract to the facts of the situation at hand. This interpretative process, therefore, generally gives rise to a mixed question of law and fact and should be reviewable on appeal for palpable and overriding error.
Sattva explained that this was an appropriate development for two related reasons. First, contractual interpretation is not simply a question of ascribing an abstract legal meaning to the words, but rather of understanding those words in their full context. Second, this process of interpretation should generally be considered to be the application of a legal standard to the facts; in other words, contractual interpretation is generally a mixed question of law and fact which, under the Court’s standard of review jurisprudence, is generally reviewed for palpable and overriding error. Both of these related reasons, as we shall see, apply to interpreting all types of contracts.
Consider the first reason. Contract interpretation cannot be understood as a process of determining the “legal” and immutable meaning of the text. “[W]ords alone do not have an immutable or absolute meaning” and therefore contractual interpretation does not often turn on ascribing immutable legal meanings to the contractual words: Sattva, at para. 47. Rather, the meaning of words often turns on context, such as the purpose of the agreement and the nature of the relationship between the parties: para. 48. Taking those sorts of contextual considerations into account – sometimes called the surrounding circumstances or the factual matrix – requires the court to understand the text of the agreement in light of them, not simply to ascribe purely legal meanings to the words taken in isolation. Thus, just as the Court in Housen cautioned against too readily finding that applying a legal standard to the facts gives rise to a purely legal question, the Court in Sattva cautioned that interpretation does not often give rise to a pure question of law. Interpretation is rarely a matter of ascribing some immutable legal meaning to the text considered apart from the surrounding circumstances.
A number of appellate courts and my colleague Wagner J. are of the view that this first rationale underlying Sattva does not apply to cases interpreting standard form contracts:
Vallieres v Vozniak, 2014 ABCA 290, 5 Alta. L.R. (6th) 28, at paras. 11-13;
Precision Plating Ltd. v Axa Pacific Insurance Co., 2015 BCCA 277, 387 D.L.R. (4th) 281, at paras. 28-30;
Stewart Estate v 1088294 Alberta Ltd., 2015 ABCA 357, 25 Alta. L.R. (6th) 1, at para. 273, per McDonald J.A.;
MacDonald v Chicago Title Insurance Co. of Canada, 2015 ONCA 842, 127 O.R. (3d) 663, at paras. 40-41;
Monk v Farmers’ Mutual Insurance Co., 2015 ONCA 911, 128 O.R. (3d) 710, at paras. 22-24;
True Construction Ltd. v Kamloops (City), 2016 BCCA 173, at para. 34 (CanLII); and
Sankar v Bell Mobility Inc., 2016 ONCA 242, at para. 26 (CanLII). I respectfully disagree.
I accept, of course, that standard form contracts generally do not have relevant surrounding circumstances relating to their negotiation because there was in no real sense any negotiation of their terms. However, standard form contracts, like all contracts, have many other surrounding circumstances: they have a purpose, they create a relationship of a particular nature and they frequently operate within a particular market or industry. These factors are all part of the context – of the surrounding circumstances – that must be taken into account in interpreting the text of the contract. As Lord Wilberforce put it in a passage cited with approval in Sattva, “In a commercial contract it is certainly right that the court should know the commercial purpose of the contract and this in turn presupposes knowledge of the genesis of the transaction, the background, the context, the market in which the parties are operating”: Reardon Smith Line Ltd. v Hansen-Tangen,  3 All E.R. 570 (H.L.), at p. 574, quoted in Sattva, at para. 47. This point is further developed in a short passage from Investors Compensation Scheme Ltd. v West Bromwich Building Society,  1 All E.R. 98 (H.L.), also quoted by the Court in Sattva, at para. 48 [p. 115]:
The meaning which a document .... would convey to a reasonable man is not the same thing as the meaning of its words. The meaning of words is a matter of dictionaries and grammars; the meaning of the document is what the parties using those words against the relevant background would reasonably have been understood to mean.
All contracts, whether standard form or not, have important contextual elements – elements of their surrounding circumstances – that are generally considered in applying the contractual language to a specific set of occurrences.
Unlike my colleague, I do not read this aspect of Sattva as holding that contractual interpretation is not generally a pure question of law simply because it involves assessing a “factual matrix” relating to negotiation. Rather, as I have discussed, Sattva sees contractual interpretation as not being a pure question of law because it involves understanding the words used in light of a number of contextual factors beyond negotiation, including the purpose of the agreement, the nature of the relationship, the market in which the parties are operating, and so forth. While the words have a consistent meaning, how they apply to the myriad of situations that may arise will most often turn on these sorts of contextual factors. My colleague’s interpretative analysis of the standard form contract before us in this case shows that this is so. That analysis relies on the nature of the particular work alleged to be faulty; the nature and cause of the particular damage in issue; the purpose of the contract; the market in which it operates (i.e. the construction industry); the parties’ reasonable expectations; and commercial reality.
The importance of taking these contextual matters into account is the first reason the Court relied on in Sattva to explain why contractual interpretation is generally not a pure question of law and applies to standard form contracts as it does to others. While negotiating history will generally not be relevant to such contracts, many other contextual matters are.
Turning to the second related reason given in Sattva, it too applies to the interpretation of standard form contracts. That second reason was that “the historical approach to contractual interpretation does not fit well with the definition of a pure question of law identified in Housen and [Canada (Director of Investigation and Research) v Southam Inc.,  1 S.C.R. 748]”: para. 49. Rather, contractual interpretation should be understood as generally giving rise to mixed questions of law and fact. As Rothstein J. wrote for the Court, “Contractual interpretation involves issues of mixed fact and law as it is an exercise in which the principles of contractual interpretation are applied to the words of the written contract, considered in light of the factual matrix”: para. 50. In short, Sattva brought appellate review of contractual interpretation into the general framework for appellate review in civil cases set out in the Court’s standard of review jurisprudence.
I see no reason to think that the interpretation of certain types of contracts should be excluded from these general principles that apply to appellate review in all civil cases. A number of appellate courts have reached the same conclusion:
Industrial Alliance Insurance and Financial Services Inc. v Brine, 2015 NSCA 104, 392 D.L.R. (4th) 575, at paras. 40-41;
Ontario Society for the Prevention of Cruelty to Animals v Sovereign General Insurance Co., 2015 ONCA 702, 127 O.R. (3d) 581, at paras. 34-36;
Acciona Infrastructure Canada Inc. v Allianz Global Risks US Insurance Co., 2015 BCCA 347, 77 B.C.L.R. (5th) 223, at paras. 34-35;
GCAN Insurance Co. v Univar Canada Ltd., 2016 QCCA 500, at paras. 37-42 (CanLII).
It is important to remember that Housen did not hold that all applications of a legal standard to the facts should be reviewed for palpable and overriding error. As I have discussed, Housen recognized that sometimes the analysis will turn on an extricable pure question of law. Sattva adopted this holding. Rothstein J. in Sattva acknowledged, echoing Housen, that it may sometimes be possible to identify an extricable question of law such as the application of incorrect principles, the failure to consider a required element of a legal test, or the failure to consider a relevant factor. I therefore agree with Wagner J. that “ Sattva should not be read as holding that contractual interpretation is always a question of mixed fact and law”: para. 46. Sattva was explicit on this point: para. 53.
However, again echoing Housen, Rothstein J. warned that courts should be cautious in identifying extricable questions of law in disputes over contractual interpretation; he noted that “the circumstances in which a question of law can be extricated from the interpretation process will be rare”: Sattva, at para. 55.
B. The Proposed Exception Does Not Conform to the General Principles of Appellate Review in Civil Cases
My colleague proposes an “exception” to Sattva’s holding: if an appeal involves the interpretation of a standard form contract, the interpretation itself is of precedential value and there is no meaningful factual matrix specific to the parties to assist the interpretation process, then the interpretation is a question of law and subject to correctness review (para. 46). I do not support the creation of this “exception”.
As I have outlined, the general principles of appellate review in civil cases turn on characterizing the nature of the question being reviewed as one of fact, law or mixed law and fact. The distinction between questions of pure law and questions of mixed law and fact turns on where the question is located along a “spectrum of particularity”: Housen, at para. 28. Questions of law are concerned with general legal propositions: Housen, at para. 28, citing Southam, at para. 37. As stated in Housen and repeated in Sattva, examples include applying an incorrect principle, failing to consider a required element of a legal test, or the failure to consider a relevant factor: Sattva, at para. 53.
As I see it, the three elements of the proposed exception do not assist in deciding whether the question is sufficiently general in nature so as to attract correctness review. Whether or not a contract is a standard form does not, as I see it, tell us anything about the degree of generality of the particular interpretative principle in issue in a particular case. The absence of a “factual matrix” is not of much assistance either. All contracts have a context which is important for their interpretation. As I mentioned earlier, aspects of the transaction such as its purpose and the market or industry in which it operates are important for interpreting all contracts, and so is the nature of the allegedly faulty work and the damage allegedly resulting from it. The absence of facts about negotiations does not mean that there are no contextual matters that inform the interpretative process and therefore tend to make it a mixed question of law and fact.
The third element of the proposed exception – whether the interpretation has precedential value – seems to me to simply ask the critical question, which is concerned with the level of generality of a legal principle, in a different and unhelpful way. Questions of law are reviewed on appeal for correctness because the decisions on such questions have precedential value: these sorts of decisions ensure uniformity among similar cases and serve the law-making function of appellate courts (Housen, at paras. 8-9). The more general the principle, the more the precedential value. To ask the question in terms of precedential value rather than the generality of the legal principle in issue seems to me to simply pose the key question in a different way and in one that simply sends the analysis back to the question of degree of generality.
As my colleague’s interpretive analysis shows, there are important contextual elements – surrounding circumstances – that inform how the text should be applied to the facts. This is not a case where there are no such contextual factors to consider. Focusing on the question of the generality of the legal principle in issue, I do not see a good case for correctness review on that basis either.
The question this case raises, boiled down to its essentials, is this: Is the cost of replacing a window that was scratched by a window cleaner while cleaning it the “cost of making good faulty workmanship” (which is excluded from insurance coverage) or the cost of repairing “physical damage [that] results” from faulty workmanship (which is covered)? The answer proposed by Wagner J. is that it is the cost of repairing the physical damage, because the exclusion applies only to the cost of redoing the faulty work, in this case, recleaning the windows: para. 5. The legal principle is that “making good faulty workmanship” means “the cost of redoing the faulty work”. However, this principle does not seem to me to operate at a very high level of generality.
Applying the principle turns on two considerations: the scope of the “faulty work” and the nature of “redoing” it. We could say that the window cleaners’ faulty work did not require them to install windows in good condition: para. 81. But this seems to me to be the assertion of the conclusion rather than a reason for it. Presumably, the window cleaners’ work was to clean the windows without destroying them; if their faulty work destroyed the windows, why should we say that “redoing” their work does not involve replacing the windows? In short, I am not convinced the principle that the exclusion only relates to “the cost of redoing the faulty work” can operate at a very high level of generality. Rather, its application in other cases will ultimately be decided on a case-by-case basis in light of the particular circumstances of the particular case.
The extensive jurisprudence cited to us tends to confirm the view that it is difficult to define the scope of the exclusion in general terms. The line basically has to be drawn on a case-by-case basis. For example, in Greene v Canadian General Insurance Co. (1995), 133 Nfld. & P.E.I.R. 151 (Nfld. C.A.), a contractor was hired by the appellants to construct the framework of a house. Shortly after the contractor completed its contract, the framework was damaged by high winds and, as a result, the house had to be substantially demolished and rebuilt: para. 2. It was uncontested that the loss was caused by the failure of the contractor to install temporary bracing.
The trial judge held that the loss was caused by faulty or improper workmanship and was an excluded peril under clause 9(a) of the appellants’ insurance policy with the respondent company. He rejected the appellants’ contention that the exclusion clause should be limited to the cost of remedying the improper installation of permanent or temporary bracing: Greene, at para. 3.
The Court of Appeal confirmed the trial decision, holding that “the defective or inadequate bracing was to stabilize the house during construction and the resulting accident caused the destruction of that house. I see no error in the decision of the trial judge that the loss suffered by the appellants was the cost of making good faulty or improper workmanship, not ‘resultant damage’, nor in his analysis of the applicable case law”: Greene, at para. 16. As I see it, the analysis of this case could easily have gone a different way. The court could have held, as we do in this case, that because the contractor was only responsible for constructing the framework of the house, the exclusion should only be the cost of redoing the framework of the house, and not the cost of fixing damages to the whole house.
Other examples include Bird Construction Co. v United States Fire Insurance Co. (1985), 24 D.L.R. (4th) 104 (Sask. C.A.), where a subcontractor was hired to build and install roof trusses and partially damaged the trusses due to faulty erection procedures. Although the faulty workmanship occurred during the erection procedures, the cost of replacing the trusses themselves ended up also not being covered because the court considered it would be making good faulty workmanship. In Ontario Hydro v Royal Insurance,  O.J. No. 215 (QL) (H.C.J.), the contractual obligation of the contractor responsible for the damage was to install a power-generating boiler. As part of the installation, the plaintiff requested that the contractor perform an acid wash on the boiler. The acid wash ultimately damaged the tubing and the court held that the cost of making good improper workmanship was not only the cost of rewashing the tubing, but also to replace it: para. 37.
I conclude that there are important surrounding circumstances that inform the interpretation of standard form contracts and that the legal principle is not of much precedential value. In short, the issue here involves applying a legal standard to a set of facts and did not give rise to any extricable question of law.
C. The Merits of the Appeals
I agree in substance with the trial judge’s analysis and conclusion: 2013 ABQB 585,  I.L.R. ¶ I-5494. Clackson J. made no legal error because he properly described and applied the Court’s decision in Progressive Homes Ltd. v Lombard General Insurance Co. of Canada, 2010 SCC 33,  2 S.C.R. 245.
The trial judge considered the competing interpretations of the relevant policy provisions. The insureds’ position was that excluding the cost of “making good” faulty cleaning simply excludes the cost of redoing the cleaning properly. The insurers’ position was that “making good” faulty cleaning extends to the damage done by the faulty cleaning. The trial judge found that both of these interpretations were reasonable as the policy did not clearly suggest one alternative over the other. The judge then took into consideration the general rules of contract construction: he looked at the context, the language of the contract, as well as the nature and purpose of an all risks policy, which helped him to determine the reasonable expectations of the parties. It remains that, according to him, these rules did not produce a clear result. He therefore applied the contra proferentem principle, interpreted the clause against the insurer and held that the exclusion did not apply. I see no reviewable error in this analysis.
I would allow the appeals, set aside the order of the Court of Appeal (2015 ABCA 121, 599 A.R. 363) and restore the order of the trial judge with costs to the appellants throughout.
 Although builders’ risk policies can provide coverage on either an all-risk or named-peril basis, only the former type of policy is at issue in this appeal. It is also the more common type of policy. Therefore, when I refer to builders’ risk policies in these reasons, I specifically mean builders’ risk policies that provide coverage on an all-risk basis.
 Between the date of the Policy and the date of judgment at trial, these respondents became the remaining respondents Northbridge Indemnity Insurance Company, Royal & Sun Alliance Insurance Company of Canada, and Chartis Insurance Company of Canada, respectively.
Eugene Meehan, Q.C., and Stacey Boothman (instructed by Supreme Advocacy, Ottawa; Stacey Boothman, Vancouver), for the appellant Ledcor Construction Limited.
Dennis L. Picco, Q.C., and Marie‑France Major (instructed by Dentons Canada, Edmonton; Supreme Advocacy, Ottawa), for the appellant Station Lands Ltd.
Gregory J. Tucker, Q.C., and Scott H. Stephens (instructed by Owen Bird Law Corporation, Vancouver), for the respondents.
all rights reserved