In an important decision released this afternoon, the Court of Appeal held that the “your work” exclusion in a CGL policy (often referred to as a “business risk” exclusion) did not apply where the complained-of work had been performed by the insured’s subcontractor. In the course of its reasons, the C.A. made some interesting observations about CGL policies whose effects are likely to be felt beyond this case. The court made it clear that coverage is to be decided in accordance with the wording of the policy, not on the basis of “general principles” of insurance law.
In Bridgewood Building Corporation v. Lombard General Insurance Company of Canada, Lombard insured a general contractor. Lombard had denied indemnity to the contractor under a CGL policy, for a claim relating to the cost of repairing a number of new homes which had been built by the insured. The homes contained structural defects which related to defective concrete. The concrete had been supplied by a subcontractor.
Lombard took the position that “there is no coverage under a CGL policy for costs incurred by the insured for the repair or replacement of its own defective work or product.”
At issue was exclusion (j) in the CGL. The actual wording of exclusion (j) (which was key to the C.A.’s decison) excludes coverage for property damage to that part of the insured’s work that is defective. However, there is an exception to the exclusion (on which the court fastened), that reads as follows:
This exclusion does not apply if the damaged work or the work out of which the damage arises was performed on your behalf by a subcontractor.
Lombard seems to have placed primary reliance, not on exclusion (j) itself, but on the more general proposition that “a CGL policy, by its very nature, is intended only to cover an insured’s tortious liability to third parties, not including the cost of repairing or replacing the insured’s own defective work or product”. In fact, the reasons of the Superior Court judge from whose decision this appeal was taken, indicate that Lombard did not rely on any “work” or “product” exclusions at all.
In the Court of Appeal, Lombard argued that CGL policies do not cover the cost of restoring defective work and that an exception to an exclusoin cannot restore coverage. Thus, it contended, the subcontractor exception in exclusion (j) should not be read to trigger coverage. To do so, said Lombard, would mean that the exception had converted the CGL into a Performance Bond, which it was never intended to be. If coverage were found, Lombard submitted, “the liability policy effectively becomes ‘enormously expanded’, opening the floodgates to claims that would include building code infractions, substitution of sinks, repair of chipped tiles, drywall repair and essentially all of the [Insureds’] work.
The Court of Appeal rejected Lombard’s argument. It said that “the key to coverage lies not in the general nature of the policy itself but in whether the alleged acts fall within its provisions and coverage is not excluded”. So, it is not correct to proceed from the premise, that CGL policies are not meant to be Performance Bonds. These sorts of general principles are mere “interpretative aids”, said the court, that “cannot decide any issues by themselves”.
(Readers might recall last year’s decision of Ferrier J., in A.R.G. Construction Corporation v. Allstate Insurance Company of Canada. Although not referred to in the Court of Appeal’s decision, it is an example of one of many cases that have accepted the basic proposition on which Lombard was relying in Bridgewood. Justice Ferrier reviewed the law and summarized it this way:
It is common ground between the parties that CGL policies are not performance bonds. They are not intended to cover the costs of repairing or replacing the insured’s defective work or product. They are intended, rather to cover tort liability for injury to other persons or damage to their property. This general intent is expressed in the language used in the various policies to define both the coverage and the exclusions.
Justice Ferrier relied on another decision of the Court of Appeal, in Celestica Inc. v. ACE INA Insurance, where the Court of Appeal said expressed the view, that defective design or manufacture would not trigger coverage, because there were good policy reasons for refusing to find that such a cause could constitute an “accident”, as required by the policy. Although this issue was not addressed in Bridgewood, it is hard to square the two approaches.)
The correct analytical method, according to the Court of Appeal, is to look to the policy wording itself. This is usually a three-stage approach, in which the insuring agreement extends coverage, exclusions withdraw coverage and exceptions to the exclusions restore some coverage taken away by the exclusion.
Applying that method to the CGL wording, the court first looked at the insuring agreement. It said
We will pay those sums that the insured becomes legally obligated to pay as damages because of “bodily injury” or “property damage” to which this insurance applies.
Next, it looked at exclusion (j), which said that the insurance did not apply to that particular part of the insured’s work that is defective or actively malfunctions.
Finally, it considered the exception to the exclusion. As quoted above, it provided that exclusion (j) did not apply if the damaged work or the work out of which the damage arises was performed on behalf of the Insured by a subcontractor.
The Court of Appeal considered that, “on a plain reading, clause (j) would seem to indicate that coverage will be provided if the ‘damaged work or the work out of which the damage arises’ is performed on behalf of the insured by a subcontractor.” Lombard was forced to argue that the exception was really redundant, since there was no coverage for the cost of making good defective work anyway, whether the work was done by the Insured or by a subcontractor. The Court of Appeal did not agree. After discussing some U.S. authorities, it concluded that there was a reason for the subcontractor exception. It quoted with approval a decision of the Minnesota Supreme Court:
A general contractor has minimal control over the work of its subcontractors by definition, and the fact that the general contractor receives coverage will not relieve the subcontractor of ultimate liability for unworkmanlike or defective work. In such a case, an insurer will have subrogation rights against the subcontractor who performed the defective work. Presumably, the Business Risk Doctrine will preclude the subcontractor from recovering from its own insurer. Thus, the goal of the Business Risk Doctrine would still be achieved because ultimate responsibility for poor workmanship would lie with the one who performed it.
This case makes it clear, if it was not already, that coverage will be decided on the basis of what the policy actually says, not on general principles of insurance law. It will be important for the insurers of general contractors, to see to it that their insureds obtain evidence that their subcontractors have adequate liability insurance.