SUPREME COURT OF CANADA

Date: 20131213 Docket: 34472

Between:

IBM Canada Limited

Appellant

and

Richard Waterman

Respondent

Coram: McLachlin C.J. and LeBel, Fish, Abella, Rothstein, Cromwell, Moldaver, Karakatsanis and Wagner JJ.

Reasons for Judgment: (paras. 1 to 99) Dissenting Reasons: (paras. 100 to 155)

Cromwell J. (LeBel, Fish, Abella, Moldaver, Karakatsanis and Wagner JJ. concurring) Rothstein J. (McLachlin C.J. concurring)

IBM Canada Limited v. Waterman, 2013 SCC 70, [2013] 3 S.C.R. 985

IBM Canada Limited Appellant

v.

Richard Waterman Respondent

Indexed as: IBM Canada Limited v. Waterman

2013 SCC 70

File No.: 34472.

2012: December 14; 2013: December 13.

Present: McLachlin C.J. and LeBel, Fish, Abella, Rothstein, Cromwell, Moldaver, Karakatsanis and Wagner JJ.

on appeal from the court of appeal for british columbia

Employment law — Wrongful dismissal — Damages — Compensating advantage — Dismissed employee drawing pension benefits upon dismissal — Trial judge establishing appropriate notice period at 20 months without deduction for pension benefits — Whether pension benefits constitute compensating advantage — Whether pension benefits should be deducted from damages for wrongful dismissal.

IBM dismissed W without cause on two months’ notice. W was 65 years old, had 42 years of service, and had a vested interest in IBM’s defined benefit pension plan. Under the plan, IBM contributed a percentage of W’s salary to the plan on his behalf. Upon termination, W was entitled to a full pension, and his termination had no impact on the amount of his pension benefits.

W sued to enforce his contractual right to reasonable notice. The trial judge set the appropriate period of notice at 20 month and declined to deduct the pension benefits paid to W during the notice period in calculating his damages. The Court of Appeal dismissed the appeal.

Held (McLachlin C.J. and Rothstein J. dissenting): The appeal should be dismissed.

Per LeBel, Fish, Abella, Cromwell, Moldaver, Karakatsanis and Wagner JJ.: The rule that damages are measured by the plaintiff’s actual loss does not cover all cases. The law has long recognized that applying the general rule of damages — the compensation principle — strictly and inflexibly sometimes leads to unsatisfactory results. Employee pension payments, including payments from a defined benefit plan, should generally not reduce the damages otherwise payable for wrongful dismissal. Pension benefits are a form of deferred compensation for the employee’s service and constitute a type of retirement savings. They are not intended to be an indemnity for wage loss due to unemployment.

A compensating advantage arises if a source other than the damages payable by the defendant ameliorates the loss suffered by the plaintiff as a result of the defendant’s breach of a legal duty. However, not all benefits received by a plaintiff raise a compensating advantages problem. A problem only arises with a compensating advantage when the advantage is one that (a) would not have accrued to the plaintiff but for the breach, or (b) was intended to indemnify the plaintiff for the sort of loss resulting from the breach.

The question is whether the compensation principle should be strictly applied and the compensating advantage should be deducted. Considerations other than the extent of the plaintiff’s actual loss shape the way the compensation principle is applied. The deductibility of compensating advantages also depends on justice, reasonableness and public policy.

Benefits received by a plaintiff through private insurance are generally not deductible from damages awards. While there is no single marker to sort which benefits fall within the private insurance exception, the more closely the benefit is, in nature and purpose, an indemnity against the type of loss caused by the defendant’s breach, the stronger the case for deduction . Whether the plaintiff has contributed to the benefit also remains a relevant consideration, although the basis for this is debatable . In general, a benefit will not be deducted if it is not an indemnity for the loss caused by the breach and the plaintiff has contributed in order to obtain entitlement to it . Finally, there is room in the analysis of the deduction issue for broader policy considerations such as the desirability of equal treatment of those in similar situations, the possibility of providing incentives for socially desirable conduct, and the need for clear rules that are easy to apply . While this exception is called the private insurance exception, it has been applied by analogy to a variety of payments that do not originate in a contract of insurance.

Although the courts have not relied on any broad “single contract” rule, where a cause of action and a benefit arise under the contract of employment, the terms of a contract and the dealings between the parties will inform the analysis.

A compensating advantage issue arises in this case: W received his full pension benefits and the salary he would have earned had he worked during the period of reasonable notice; had IBM given him working notice, he would have received only his salary during that period . However, the private insurance exception applies to benefits such as pension payments to which an employee has contributed and which were not intended to be an indemnity for the type of loss suffered as a result of the defendant’s breach . As such, the compensation principle should not be applied strictly in this case.

In this case, the factors clearly support not deducting the retirement pension benefits from wrongful dismissal damages . W’s contract of employment is silent on this issue, but it does not have any general bar against receiving full pension entitlement and employment income . W’s retirement pension is not an indemnity for wage loss, but rather a form of retirement savings . While IBM made all of the contributions to fund the plan, W earned his entitlement to benefits through his years of service, as the plan’s primary purpose is to provide periodic pension payments to eligible employees after retirement in respect of their service as employees . Thus, this case falls into the category of cases in which the insurance exception has always been applied — the benefit is not an indemnity and W contributed to the benefit.

Although Sylvester v. British Columbia , 1997 353 (SCC) , [1997] 2 S.C.R. 315, is distinguishable on the facts, the factors it sets out support the conclusion that W’s benefits should not be deducted from his wrongful dismissal damages . The pension benefits were clearly not an indemnity benefit for loss of salary due to inability to work, and W’s interest in the pension bears many of the hallmarks of a property right . Looking at the contract as a whole, it is not a fair implication that the parties agreed that pension entitlements should be deducted from wrongful dismissal damages.

Finally, the broader policy concerns in this case support not deducting the pension benefits . The law should not provide an economic incentive to dismiss pensionable employees rather than other employees . The other policy concerns raised by Justice Rothstein or present in Sylvester either do not arise here or are highly speculative.

Per McLachlin C.J. and Rothstein J. (dissenting): This case requires an assessment of W’s loss under the terms of a single contract which gave rise to both a right to reasonable notice and a right to pension benefits. The private insurance exception has no application to such a case . Where a court is called upon to assess loss under a single contract, the plaintiff’s entitlement turns on the ordinary governing principle that he should be put in the position he would have been in had the contract been performed . In this case, that means that the pension benefits W received must be deducted in calculating his damages for wrongful dismissal; not deducting would give W more than he bargained for and would charge IBM more than it agreed to pay.

The governing principle for damages upon breach of contract is that the non‑breaching party should be provided with the financial equivalent of performance . Employer‑provided benefits are integral components of the employment contract, so deductibility turns on the terms of the employment contract and the intention of the parties . Under the terms of W’s employment contract, he would have been eligible to receive pension benefits only upon being terminated or retiring . Therefore, as in Sylvester , W’s contractual right to wrongful dismissal damages and his contractual right to his pension are based on opposite assumptions about his availability to work. Damages cannot be paid on the assumption that he could have earned both.

This conclusion is necessitated by the fact that the pension plan at issue here is a defined benefit plan . Unlike a defined contribution plan, a defined benefit plan guarantees the employee fixed predetermined payments upon retirement for life . Deducting the benefits would provide the wrongfully terminated employee with exactly what he would have received had the employment contract been performed: an amount equal to his salary during the reasonable notice period and thereafter defined benefits for the rest of his life.

This is materially different from a defined contribution plan, which provides an employee with a finite total amount or lump sum of retirement benefits. Deducting benefits that a wrongfully terminated employee receives from a defined contribution plan would leave the employee in a worse position that he would have been in had his employment contract not been breached.

In this case, W’s wrongful dismissal had no impact on his pension entitlement, and he could not have received both his salary and his pension benefits had he continued to work for IBM through the reasonable notice period . Whether the benefit is non‑indemnity or contributory does not answer the question of whether the plaintiff will be provided with the financial equivalent of performance or will receive excess recovery under the governing principle of contract damages.

Furthermore, the private insurance exception is not applicable to cases that involve a single contract that is the source of both the plaintiff’s cause of action and his right to a particular benefit. In such circumstances, there is no justification for resorting to the private insurance exception because the plaintiff’s entitlement to the benefits is established based on the terms of his contract . If the plaintiff is entitled to the benefits under his contract, he will receive the benefits based on the ordinary governing principle that he should be placed in the position he would have been in had the contract been performed. There will be no need to reach the collateral benefit exception. A straightforward reading of Sylvester demonstrates that it is a fully applicable authority supporting the proposition that, under a single contract of employment, barring contractual provisions to the contrary, an individual cannot receive salary as if he is working and pension benefits as if he is retired . These are opposite, incompatible assumptions. Thus, applying Sylvester to this case, salary and pension income are not payable at the same time.

Cases Cited

By Cromwell J.

Distinguished: Sylvester v. British Columbia , 1997 353 (SCC) , [1997] 2 S.C.R. 315; Ratych v. Bloomer , 1990 97 (SCC) , [1990] 1 S.C.R. 940; discussed : Cunningham v. Wheeler , 1994 120 (SCC) , [1994] 1 S.C.R. 359; referred to: Phillips v. Western Company of North America , 953 F.2d 923 (1992); United States v. Price , 288 F.2d 448 (1961); Sloas v. CSX Transportation, Inc. , 616 F.3d 380 (2010); Parry v. Cleaver , [1970] A.C. 1 ; Attorney General v. Blake , [2001] 1 A.C. 268 ; Bank of America Canada v. Mutual Trust Co. , 2002 SCC 43 , [2002] 2 S.C.R. 601; Redpath v. Belfast and County Down Railway (1947), N.I. 167; Jack Cewe Ltd. v. Jorgenson , 1980 177 (SCC) , [1980] 1 S.C.R. 812; Canadian Pacific Ltd. v. Gill , 1973 2 (SCC) , [1973] S.C.R. 654; Grand Trunk Railway v. Beckett (1887), 16 S.C.R. 713 ; Quebec Workmen’s Compensation Commission v. Lachance , 1971 184 (SCC) , [1973] S.C.R. 428; Guy v. Trizec Equities Ltd. , 1979 232 (SCC) , [1979] 2 S.C.R. 756; Chandler v. Ball Packaging Products Canada Ltd. (1992), 2 C.C.P.B. 101, aff’d (1993), 2 C.C.P.B. 99; Emery v. Royal Oak Mines Inc. (1995), 1995 7074 (ON SC) , 24 O.R. (3d) 302; Canadian Human Rights Commission v. Canada (Attorney General) , 2003 FCA 86 , 301 N.R. 321; Bradburn v. Great Western Railway Co. (1874), L.R. 10 Ex. 1 ; National Insurance Co. of New Zealand Ltd. v. Espagne (1961), 105 C.L.R. 569 ; Graham v. Baker (1961), 106 C.L.R. 340 ; Smoker v. London Fire and Civil Defence Authority , [1991] 2 A.C. 502 ; Hopkins v. Norcross plc , [1993] 1 All E.R. 565 ; Knapton v. ECC Card Clothing Ltd. , [2006] I.C.R. 1084; Gilbert v. Attorney‑General , [2010] NZCA 421, 8 N.Z.E.L.R. 72.

By Rothstein J. (dissenting)

Girling v. Crown Cork & Seal Canada Inc. (1995), 1995 954 (BC CA) , 9 B.C.L.R. (3d) 1; Sylvester v. British Columbia , 1997 353 (SCC) , [1997] 2 S.C.R. 315; Cunningham v. Wheeler , 1994 120 (SCC) , [1994] 1 S.C.R. 359; Chandler v. Ball Packaging Products Canada Ltd. (1992), 2 C.C.P.B. 101; Parry v. Cleaver , [1970] A.C. 1 ; Guy v. Trizec Equities Ltd. , 1979 232 (SCC) , [1979] 2 S.C.R. 756; Canadian Pacific Ltd. v. Gill , 1973 2 (SCC) , [1973] S.C.R. 654; Jack Cewe Ltd. v. Jorgenson , 1980 177 (SCC) , [1980] 1 S.C.R. 812; United States v. Price , 288 F.2d 448 (1961); Phillips v. Western Company of North America , 953 F.2d 923 (1992); Bank of America Canada v. Mutual Trust Co. , 2002 SCC 43 , [2002] 2 S.C.R. 601.

Statutes and Regulations Cited

Canadian Forces Superannuation Act , R.S.C. 1985, c. C‑17.

Employment Insurance Act , S.C. 1996, c. 23, s. 45.

Authors Cited

Burrows, Andrew. Remedies for Torts and Breach of Contract , 3rd ed. Oxford: Oxford University Press, 2004.

Cassels, Jamie, and Elizabeth Adjin‑Tettey. Remedies: The Law of Damages , 2nd ed. Toronto: Irwin Law, 2008.

Fleming, John G. “The Collateral Source Rule and Contract Damages” (1983), 71 Cal. L. Rev. 56.

Kaplan, Ari, and Mitch Frazer. Pension Law , 2nd ed. Toronto: Irwin Law, 2013.

Marks, John. “Symmetrical Use of Universal Damages Principles — Such as the Principles Underlying the Doctrine of Proximate Cause — to Distinguish Breach‑Induced Benefits That Offset Liability From Those That Do Not” (2009), 55 Wayne L. Rev. 1387.

McCamus, John D. The Law of Contracts , 2nd ed. Toronto: Irwin Law, 2012.

Ogus, A. I. The Law of Damages . London: Butterworths, 1973.

Perillo, Joseph M. “The Collateral Source Rule in Contract Cases” (2009), 46 San Diego L. Rev. 705.

Sproat, John R. Wrongful Dismissal Handbook , 6th ed. Toronto: Carswell, 2012.

Swan, Angela, and Jakub Adamski. Canadian Contract Law , 3rd ed. Markham, Ont.: LexisNexis, 2012.

Waddams, S. M. The Law of Damages , 5th ed. Toronto: Canada Law Book, 2012.

APPEAL from a judgment of the British Columbia Court of Appeal (Finch C.J. and Prowse and Levine JJ.A.), 2011 BCCA 337 , 20 B.C.L.R. (5th) 241, 308 B.C.A.C. 304, 521 W.A.C. 304, 336 D.L.R. (4th) 481, [2011] 10 W.W.R. 425, 91 C.C.P.B. 60, 92 C.C.E.L. (3d) 289, [2011] B.C.J. No. 1453 (QL), 2011 CarswellBC 2023, affirming a decision of Goepel J., 2010 BCSC 376 , 2010 CLLC ¶210‑021, [2010] B.C.J. No. 510 (QL), 2010 CarswellBC 679 . Appeal dismissed, McLachlin C.J. and Rothstein J. dissenting.

D. Geoffrey Cowper , Q.C. , and Lorene A. Novakowski , for the appellant.

Christopher J. Watson and Matthew G. Siren , for the respondent.

The judgment of LeBel, Fish, Abella, Cromwell, Moldaver, Karakatsanis and Wagner JJ. was delivered by

Cromwell J. —

I. Introduction

[ 1 ] When IBM Canada Ltd. wrongfully dismissed its long-time employee, Richard Waterman, he had to start drawing his pension. The question before the Court is whether his receipt of those pension benefits reduces the damages otherwise payable by IBM for wrongful dismissal. The British Columbia courts decided not to deduct the pension benefits and IBM appeals.

[ 2 ] The question looks straightforward enough at first glance. The general rule is that contract damages should place the plaintiff in the economic position that he or she would have been in had the defendant performed the contract. IBM’s obligation was to give Mr. Waterman reasonable notice of dismissal or pay in lieu of it. Had it given him reasonable working notice, he would have received only his regular salary and benefits during the period of notice. As it is, he in effect has received both his regular salary and his pension for that period. It therefore seems clear, under the general rule of contract damages, that the pension benefits should be deducted. Otherwise, Mr. Waterman is in a better economic position than he would have been in had there been no breach of contract.

[ 3 ] On closer study, however, the question raised on appeal is not as simple as that. The case in fact raises one of the most difficult topics in the law of damages, namely when a “collateral benefit” or a “compensating advantage” received by a plaintiff should reduce the damages otherwise payable by a defendant. The law has long recognized that applying the general rule of damages strictly and inflexibly sometimes leads to unsatisfactory results. The question is how to identify the situations in which that is the case.

[ 4 ] In my view, employee pension payments, including payments from a defined benefit plan as in this case, are a type of benefit that should generally not reduce the damages otherwise payable for wrongful dismissal. Both the nature of the benefit and the intention of the parties support this conclusion. Pension benefits are a form of deferred compensation for the employee’s service and constitute a type of retirement savings. They are not intended to be an indemnity for wage loss due to unemployment. The parties could not have intended that the employee’s retirement savings would be used to subsidize his or her wrongful dismissal. There is no decision of this Court in which a non-indemnity benefit to which the plaintiff has contributed, such as the pension benefits in issue here, has ever been deducted from a damages award.

[ 5 ] I would dismiss IBM’s appeal and affirm the result arrived at by the British Columbia courts.

II. Overview of Facts and Proceedings

[ 6 ] When IBM dismissed Mr. Waterman without cause on March 23, 2009, he was 65 years old and had 42 years of service. He was a long-standing member of IBM’s defined benefit pension plan, which I will refer to simply as “the plan”. IBM contributed a percentage of his salary to the plan on his behalf and the plan guaranteed specific benefits, which became vested over time, upon retirement.

[ 7 ] At the time of the termination, there was no longer a mandatory retirement policy in place for IBM employees. However, Mr. Waterman was entitled to a full pension under the plan and his termination had no impact on the amount of his pension benefits. IBM told Mr. Waterman that on termination, he would be treated as a retiree and that he must begin receiving monthly pension payments as of that date.

[ 8 ] An employee like Mr. Waterman, who is entitled to retire with his full pension but has not reached the age of 71, cannot receive both pension and employment income from IBM at the same time. That changes at age 71, when he or she must start drawing benefits and may continue working and earning employment income from IBM. We have not been referred to any provision in the plan that would prevent a retiree, regardless of age, from receiving benefits under the plan and employment income from a different employer.

[ 9 ] Mr. Waterman sued for wrongful dismissal and the matter proceeded to summary trial in the Supreme Court of British Columbia. The trial judge, Goepel J., found that the appropriate period of notice was 20 months. IBM’s position was (and is) that Mr. Waterman’s pension benefits (approximately $2,124 per month starting June 1, 2009) should be deducted from the salary and benefits otherwise payable during this period. The trial judge rejected this position: 2010 BCSC 376 , 2010 CLLC ¶210-021.

[ 10 ] IBM’s appeal from this decision was dismissed by the British Columbia Court of Appeal. Writing for the court, Prowse J.A. relied on this Court’s judgment in Sylvester v. British Columbia , 1997 353 (SCC) , [1997] 2 S.C.R. 315 . However, she concluded that the distinctions between the benefits and the intentions of the parties in the two cases led to a different conclusion in this case: 2011 BCCA 337 , 20 B.C.L.R. (5th) 241.

III. Positions of the Parties

[ 11 ] On its appeal to this Court, IBM makes two main points. It submits, first, that the result reached by the British Columbia courts is at odds with the compensatory goal of damages for wrongful dismissal. IBM points out that even if it had given Mr. Waterman adequate working notice of his termination, he would not have received both his employment income and his pension benefits during the notice period. By awarding him damages for the full notice period without deduction of the pension benefits received during that period, the British Columbia courts have placed him in a better economic position than he would have been in had IBM performed the contract. Second, IBM maintains that the Court in Sylvester held that these sorts of benefits are part of an integrated employment relationship and, unless deducted, the employee collecting them would receive greater compensation than would an employee lawfully dismissed with working notice.

[ 12 ] Mr. Waterman urges us to reject IBM’s position. He submits that the pension is the property of the employee that is earned through work and consists of a benefit that is part of the employee’s remuneration package. The pension is like a “nest egg”, RRSP or savings account, which IBM could not take advantage of to offset the damages awarded. Mr. Waterman could have transferred the value of his pension to another vehicle if he had left employment with IBM before reaching the age of 65 and his retirement savings would consequently have been out of reach. As for the intention of the parties, there is no provision in the pension plan expressly prohibiting concurrent reception of salary and pension benefits. It was therefore up to the courts to determine the parties’ intention, which the Court of Appeal correctly did in its decision.

IV. Analysis

[ 13 ] In my respectful view, both of IBM’s main arguments must be rejected. The general principle of compensation is not a full answer to the issue. The question is whether this case falls within an exception to it and in my view it does. The Court’s decision in Sylvester is distinguishable and, in fact, its reasoning supports the conclusion that the pension benefits should not be deducted.

[ 14 ] There are three key matters that need to be considered in order to answer the question posed by the appeal. I will set them out here with a summary of my conclusions.

A. Why is there a “collateral benefit” problem in this case?

[ 15 ] A collateral benefit is a gain or advantage that flows to the plaintiff and is connected to the defendant’s breach. This connection may exist either because there is a “but for” causal link between the breach and the receipt of the benefit or because the benefit was intended to provide the plaintiff with an indemnity for the type of loss caused by the breach. The problem raised by collateral benefits is the question of whether they should be deducted from the damages otherwise payable by the defendant on account of the breach. This case raises a collateral benefit problem because there is a “but for” causal link between the IBM’s breach of contract and Mr. Waterman’s receipt of the benefit. He would not have received the pension benefits and full salary in lieu of working notice “but for” the dismissal.

B. Is the compensation principle the answer to the problem?

[ 16 ] The principle that the defendant should compensate the plaintiff only for his or her actual loss is not, on its own, an answer to the problem. There are exceptions to the strict application of this principle, the most important of which is the exception for private insurance and other benefits which, for this purpose, are considered analogous to private insurance. That exception applies not only to insurance benefits in the strict sense, but also to other benefits such as pension payments to which an employee has contributed and which were not intended to be an indemnity for the type of loss suffered as a result of the defendant’s breach.

C. Does the Court’s decision in Sylvester support IBM’s position that the pension benefits must be deducted?

[ 17 ] In my view, it does not. Sylvester is distinguishable. The reasoning in Sylvester in fact supports the conclusion that Mr. Waterman’s pension benefits should not be deducted from the wrongful dismissal damages otherwise payable by IBM.

[ 18 ] My more detailed analysis follows.

A. Why Is There a Collateral Benefit Problem in This Case?

[ 19 ] It will be helpful to start by explaining what a collateral benefit problem is and why we have one here.

(1) What Is a Collateral Benefit Problem?

[ 20 ] In general terms, there is a collateral benefit when a source other than the damages payable by the defendant ameliorates the loss suffered by the plaintiffs as a result of the defendant’s breach of legal duty: J. Cassels and E. Adjin-Tettey, Remedies: The Law of Damages (2nd ed. 2008), at p. 416. For example, if an employee is wrongfully dismissed, but receives employment insurance benefits, those benefits are a collateral benefit. The problem is whether they should be deducted from the damages the defendant will pay for wrongful dismissal.

[ 21 ] If we simply apply the compensation principle — that the plaintiff should recover his or her actual economic loss but not more — the answer is straightforward. If we do not deduct the collateral benefit, the plaintiff will be in a better position than he or she would have been in had the employment contract been performed. To apply the compensation principle, we should consider not only the plaintiff’s losses but also any gains that flow from the defendant’s breach. The collateral benefit problem asks whether we should apply the compensation principle and deduct or depart from it and not deduct.

[ 22 ] There is considerable overlap between the collateral benefit problem and the questions of mitigation. The main distinction is this: mitigation is concerned with whether the plaintiff acted reasonably after the defendant’s breach in order to reduce losses. The collateral benefit question, in contrast, is concerned with whether some compensating advantage that was in fact received by the plaintiff, most often as a result of arrangements made before the breach, should be taken into account in assessing the plaintiff’s damages: see A. I. Ogus, The Law of Damages (1973), at pp. 87-88.

(2) When Does a Collateral Benefit Problem Arise?

[ 23 ] Not all benefits received by a plaintiff raise a collateral benefit problem. Before there is any question of deduction, the receipt of the benefit must constitute some form of excess recovery for the plaintiff’s loss and it must be sufficiently connected to the defendant’s breach of legal duty.

[ 24 ] For example, there is no excess recovery if the party supplying the benefit is subrogated to — that is, steps into the place of — the plaintiff and recovers the value of the benefit. In those circumstances, the defendant pays the damages he or she has caused, the party who supplied the benefit is reimbursed out of the damages and the plaintiff retains compensation only to the extent that he or she has actually suffered a loss: see, e.g., Cunningham v. Wheeler , 1994 120 (SCC) , [1994] 1 S.C.R. 359, at pp. 386-88, per McLachlin J., as she then was, dissenting in part. (The employment insurance example that I mentioned earlier is now resolved in this way by statute: see below, at para. 44.)

[ 25 ] Even if there is some form of excess recovery, however, there is only a collateral benefit problem if the benefit is sufficiently connected to the defendant’s breach. This requirement of sufficient connection serves a purpose with respect to collateral benefits that is analogous to that served by rules of causation and remoteness with respect to damages. Just as plaintiffs cannot recover all losses, no matter how loosely related to the defendant’s breach or how far beyond the parties’ reasonable contemplation, so too the defendant does not get credit for all benefits accruing to the plaintiff, no matter how loosely connected to the defendant’s wrongful conduct.

[ 26 ] Before turning to the nature of the required link, I note that scholars have objected to the term “collateral benefit” because it assumes the answer to the question. The word “collateral” suggests that the benefit should not be taken into account. But of course the legal problem is whether or not the benefit should be deducted. Scholars have suggested that the term “compensating advantages” is a better one and that is the term I will use in my reasons: see, e.g., Ogus, at pp. 93-94; A. Burrows, Remedies for Torts and Breach of Contract (3rd ed. 2004), at p. 156; S. M. Waddams, The Law of Damages (5th ed. 2012), at para. 15.700.

[ 27 ] Another problem with the terms “collateral benefit” or “collateral source” is that they suggest that the test for whether a benefit is deductible is whether it is “collateral”, that is, independent of the relation between the plaintiff and the defendant. Some of the American jurisprudence, for example, has recognized that this “independence” test is an oversimplification which does not explain the treatment of benefit in the cases: see, e.g., Phillips v. Western Company of North America , 953 F.2d 923 (5th Cir. 1992), at pp. 931-33. Moreover, it can lead to fruitless semantic debates about whether a benefit is or is not “collateral” or “independent” rather than furthering principled analysis. As one court put it, that a benefit “comes from the defendant tortfeasor does not itself preclude the possibility that it is from a collateral source. The plaintiff may receive benefits from the defendant himself which, because of their nature, are not considered double compensation”: United States v. Price , 288 F.2d 448 (4th Cir. 1961), at p. 450; Sloas v. CSX Transportation, Inc. , 616 F.3d 380 (4th Cir. 2010), at p. 389. As we shall see, several factors other than the source of the benefit may be considered in order to determine whether it should be deducted.

[ 28 ] Returning to the issue of connection between the benefit and the breach, the question is what sort of link is required before the issue about deduction arises. The cases suggest two answers. The advantage must either be one that (a) would not have accrued to the plaintiff “but for” the defendant’s breach or (b) was intended to indemnify the plaintiff for the sort of loss resulting from it. If neither of these conditions is present, there is no issue about deduction. If either of these conditions is present, there is.

[ 29 ] In relation to the “but for” connection between the breach and the advantage, consider this example. A plaintiff who has been injured by a defendant’s negligence buys a lottery ticket, as is his usual practice, and wins a large sum of money. No one would argue that the amount of the winnings should be deducted from the damages payable by the defendant. There is no “but for” causal connection between the defendant’s negligence and the plaintiff’s purchase of the winning ticket: see Burrows, at p. 156.

[ 30 ] Even if there is no “but for” causal link between a benefit and the breach, there may still be a problem about whether a benefit should be deducted. This will occur where the benefit and the breach are connected in the sense that the benefit is intended to indemnify the type of loss caused by the breach — Sylvester is an example. Mr. Sylvester was unable to work and receiving disability payments under his employment contract when he was wrongfully dismissed. There was clearly no causal link between the employer’s failure to give reasonable notice of termination (or payment in lieu of notice) and the receipt of the disability benefits. Nonetheless, the Court found that there was a compensating advantages problem. As Major J. pointed out, the disability benefits were intended to be a substitute for Mr. Sylvester’s regular salary: para. 14. In other words, the benefit was intended to be an indemnity for the loss of the regular salary, precisely the sort of loss that resulted from the defendant’s breach of the employment contract.

[ 31 ] The existence of these sorts of links between the breach and the benefit identifies whether there is a compensating advantage problem. But the existence of such a link is not a reliable marker of whether a particular benefit should be deducted. Relying on strict principles of causation, for example, often conceals unarticulated policy concerns: see, e.g., Parry v. Cleaver , [1970] A.C. 1 (H.L.), at pp. 34-35 , per Lord Pearce; Ogus, at pp. 225-26; Ratych v. Bloomer , 1990 97 (SCC) , [1990] 1 S.C.R. 940, at pp. 965-66. Similarly, the indemnity factor is not a reliable marker of which benefits are or are not deductible. This is clear, for example, from the Court’s decision in Cunningham . In issue were disability benefits provided for under collective agreements. They were clearly intended to provide an indemnity for wage loss arising from an inability to work. Nonetheless, the Court held that the benefits should not be deducted.

[ 32 ] To sum up, a potential compensating advantage problem exists if the plaintiff receives a benefit that would result in compensation of the plaintiff beyond his or her actual loss and either (a) the plaintiff would not have received the benefit but for the defendant’s breach, or (b) the benefit is intended to be an indemnity for the sort of loss resulting from the defendant’s breach. These factors identify a potential problem with a compensating advantage, but do not decide how it should be resolved.

(3) Why Is There a Problem About Deduction in This Case?

[ 33 ] A compensating advantage issue arises in this case. First, there is an element of excess compensation. Mr. Waterman has received his full pension benefits and, in addition, the salary he would have earned had he worked during the period of reasonable notice (less an allowance for his earnings from other employment). Had IBM not breached the contract of employment and instead given him working notice, he would have received only his salary during that period and not his pension. Second, there is a “but for” causal relationship between IBM’s breach of contract and Mr. Waterman’s receipt of the pension benefits. One could say that it was the pension plan rather than IBM’s breach of contract that gave rise to the benefit, but it is artificial to suggest that there is no “but for” causal link between IBM’s breach of contract and Mr. Waterman’s receipt of his pension benefits: “but for” the breach, there would have been no termination and, “but for” the termination, Mr. Waterman would not have started to collect his pension. Given that there was double recovery and that the benefit would not have arisen but for IBM’s breach, we must decide whether the benefit should or should not be deducted from damages otherwise payable by IBM.

B. Is the Compensation Principle the Answer to the Problem?

[ 34 ] IBM’s first main point is that the compensation principle requires the pension benefits to be deducted. Mr. Waterman is better off as a result of the damage award than he would have been if IBM had given reasonable working notice. It follows, in IBM’s submission, that the pension benefits must be deducted so that the damage award places Mr. Waterman in the economic position he would have been in had IBM given him reasonable working notice. This is essentially the position adopted by my colleague Rothstein J.

[ 35 ] While I agree that the damage award is a departure from the compensation principle, this in itself is not an answer to the problem posed by the appeal. As I will explain, the compensation principle cannot be, and is not, applied strictly or inflexibly in a manner that is divorced from other considerations. The question is whether the compensation principle should be strictly applied in this case. In my view, it should not. To explain why, it is helpful to look first at why the compensation principle is not applied strictly, or at all, in various situations.

(1) When Does the Compensation Principle Not Apply Strictly?

[ 36 ] Considerations other than the extent of the plaintiff’s actual loss shape the way the compensation principle is applied and there are well-established exceptions to it. For example, the rule that contract damages compensate only the plaintiff’s actual loss is not the only rule that applies to assessing contract damages. As a leading English case put it, “Damages are measured by the plaintiff’s loss, not the defendant’s gain. But the common law, pragmatic as ever, has long recognised that there are many commonplace situations where a strict application of this principle would not do justice between the parties. Then compensation for the wrong done to the plaintiff is measured by a different yardstick” : Attorney General v. Blake , [2001] 1 A.C. 268 (H.L.), at p. 278 . In some cases, for example, an award of damages in contract may be based on the advantage gained by the defendant as a result of the breach rather than the loss suffered by the plaintiff: see, e.g., Bank of America Canada v. Mutual Trust Co. , 2002 SCC 43 , [2002] 2 S.C.R. 601, at para. 25 . The rule that damages are measured by the plaintiff’s actual loss, while the general rule, does not cover all cases. In addition, through the doctrines of remoteness and mitigation, the compensation principle gives way to considerations of reasonableness in relation to whether the plaintiff’s expectations of the contract and his or her conduct in response to the breach of it were reasonable.

[ 37 ] Finally, there are well-recognized exceptions in which benefits flowing to plaintiffs are not taken into account even though the result is that they are better off, economically speaking, after the breach than they would have been had there been no breach. These exceptions are ultimately based on factors other than strict compensatory considerations. As Lord Reid put it in Parry , “[t]he common law has treated [the deductibility of compensating advantages] as one depending on justice, reasonableness and public policy”: p. 13. Or, as McLachlin J. wrote, this issue raises a question of “basic policy”: Ratych , at p. 959.

(2) What Factors Help to Identify When Compensating Advantages Are Not Deducted?

[ 38 ] What are some of these considerations of justice, reasonableness and policy? An answer may be found by looking at the two well-established situations in which compensating advantages are not deducted: charitable gifts and private insurance.

(a) Charitable Gifts

[ 39 ] The first is the less controversial. The rule is that charitable gifts made to the plaintiff are generally not deductible from the plaintiff’s damages even though they were made as a result of and in response to the injury or loss caused by the defendant’s wrong: see, e.g., Waddams, at paras. 3.1550-3.1560; Cassels and Adjin-Tettey, at pp. 420-21. Two concerns explain the exception: first, that if these charitable gifts were deducted, “the springs of private charity would be found to be largely, if not entirely, dried up” and, second, that it rarely makes practical sense to spend the time and effort required to take these sorts of gifts into account: Redpath v. Belfast and County Down Railway (1947), N.I. 167 (K.B.), at p. 170. See also Ogus, at p. 223; Waddams, at para. 3.1550; Cassels and Adjin-Tettey, at pp. 420-21; Cunningham , at p. 370.

[ 40 ] These explanations of the exception suggest we may take into account the broader incentives created by deducting or not deducting a benefit as well as pragmatic considerations relating to whether the applicable rule is clear, coherent and easy to apply: Cunningham , at p. 388, per McLachlin J.

(b) Private Insurance

[ 41 ] A second and more controversial exception relates to payments from the plaintiff’s private insurance. The core of the exception is well established: benefits received by a plaintiff through private insurance are not deductible from damage awards. However, both the precise scope and the rationale of the exception have been the subject of judicial and scholarly debate. Its practical importance is limited given the widespread use of subrogation, which avoids the compensating advantage issue altogether. While the exception more typically arises in tort cases, it has also been applied in contract actions, including actions for wrongful dismissal: Jack Cewe Ltd. v. Jorgenson , 1980 177 (SCC) , [1980] 1 S.C.R. 812. The approach in both areas of law is the same in principle, although the terms of the contract and the dealings between the parties will inform the analysis in contract cases.

[ 42 ] One area of controversy relates to the sorts of benefits which fall within the private insurance exception. Does it apply to both indemnity and non-indemnity insurance? Does it extend to disability benefits, employment insurance or pensions payable on retirement? The Court has held that the answer to all of these questions is yes, but not, as we shall see, without well-reasoned dissent. In short, the so-called private insurance exception has been applied by analogy to a variety of payments that do not originate in a contract of insurance.

[ 43 ] In Canadian Pacific Ltd. v. Gill , 1973 2 (SCC) , [1973] S.C.R. 654, the Court applied the insurance exception to prevent deduction of the present value of Canada Pension Plan benefits available to surviving dependents from the damages awarded in a fatal injuries claim. Spence J., for the Court, held that the payments were “so much of the same nature as contracts of insurance that they also should be excluded from consideration when assessing damages under the provisions of that statute”: p. 670; see also Grand Trunk Railway v. Beckett (1887), 16 S.C.R. 713, at p. 714, and Quebec Workmen’s Compensation Commission v. Lachance , 1971 184 (SCC) , [1973] S.C.R. 428, at pp. 433-34.

[ 44 ] In Guy v. Trizec Equities Ltd. , 1979 232 (SCC) , [1979] 2 S.C.R. 756, Mr. Guy’s injury led to his retirement and receipt of pension benefits. They were not deducted from damages for loss of earnings. Ritchie J., for the Court, viewed pensions, whether contributory or non-contributory, as flowing from the employee’s work and part of what the employer was prepared to pay for the employee’s services. He agreed with Lord Reid’s conclusion, in Parry , as quoted by Spence J., in Gill , that “[t]he fact that they flow from past work equates them to rights which flow from an insurance privately effected by [the employee]”: Guy , at p. 763. Similarly, in Jack Cewe , the Court did not deduct a dismissed employee’s unemployment insurance benefits from his wrongful dismissal damages. The benefits, wrote Pigeon J., for the Court, were a consequence of the contract of employment making them similar to contributory pension benefits: p. 818. (The collateral benefit issue that arose in Jack Cewe is now addressed by s. 45 of the Employment Insurance Act , S.C. 1996, c. 23, which states that a claimant who receives benefits and is subsequently awarded damages for the same period, “shall pay to the Receiver General as repayment of an overpayment of benefits an amount equal to the benefits that would not have been paid if the earnings had been paid or payable at the time the benefits were paid”.)

[ 45 ] In Ratych , the Court found that sick leave benefits should be deducted from damages otherwise payable for loss of earning by the party whose negligence was responsible for the injuries. For the majority, McLachlin J. wrote that it may well be appropriate not to deduct benefits where the employee can show a contribution equivalent to payment of an insurance premium. In other words, benefits may not be deductible when they come about because the plaintiff has prudently obtained and paid for insurance. However, that was not the case in Ratych , making it a different situation than one in which the benefits flow from the employer/employee relationship: pp. 973-74. In Cunningham , disability insurance benefits payable under the terms of collective agreements were held not to be deductible because there was evidence that the plaintiffs had paid for these disability plans through reduced wages. The Court’s earlier decision in Ratych was distinguished on this basis.

[ 46 ] Finally, in Sylvester , non-contributory disability benefits received during the notice period were deducted from wrongful dismissal damages otherwise payable. The benefits were intended to be an indemnity for lost wages while the plaintiff was unable to work, the plaintiff had not contributed to acquire the benefit, and policy considerations favoured deduction.

[ 47 ] The two cases in which the private insurance exception was not applied ( Ratych and Sylvester ) involved benefits that were intended to be an indemnity for the type of loss that resulted from the defendant’s breach and to which the plaintiff had not contributed. Retirement pension benefits, which are not an indemnity for loss of wages resulting from inability to work and to which the employee contributes directly or indirectly, have been held by this Court and others to fall within the private insurance exception: Guy ; Gill ; Chandler v. Ball Packaging Products Canada Ltd . (1992), 2 C.C.P.B. 101 (Ont. Ct. J. (Gen. Div.)), aff’d (1993), 2 C.C.P.B. 99 (Ont. Ct. J. (Div. Ct.)); Emery v. Royal Oak Mines Inc. (1995), 1995 7074 (ON SC) , 24 O.R. (3d) 302 (Gen. Div.); Parry .

[ 48 ] IBM relies on Canadian Human Rights Commission v. Canada (Attorney General) , 2003 FCA 86 , 301 N.R. 321, but, in my view, this reliance is misplaced. The human rights complainant in that case, Master Corporal (retired) Carter, complained that his release from the Canadian Forces by virtue of his age constituted discrimination; in other words, his claim was not that his employer had failed to give him reasonable notice of termination, but that it could not lawfully terminate him. Following his release from service, a proper legislative basis for compulsory retirement was put in place, thus ending the discrimination. The question was whether the compensation awarded by the Human Rights Tribunal for lost wages during the period of discrimination should be reduced by the amount of pension benefits received during that period. The Federal Court of Appeal held that they should. However, it specifically declined to decide the case on the basis of the private insurance exception: para. 20. Instead, it reasoned that Master Corporal Carter should be treated as a member of the regular force during the period of discrimination. But, by virtue of the applicable provisions of the Canadian Forces Superannuation Act , R.S.C. 1985, c. C-17, a person may either be a member of the regular armed forces contributing to the superannuation account or a person who has ceased to be a member and entitled to benefits, but not both at the same time. On that basis, his claim for both pension benefits and his full salary was inconsistent with the nature of his claim and the governing legislation. This reasoning cannot apply to this case, however. The private insurance exception applies to wrongful dismissal actions: Jack Cewe . In addition, the contractual provisions here, unlike the statute that governed Master Corporal Carter’s case, do not have any general bar against receiving full pension entitlement and employment income.

[ 49 ] A second area of controversy concerns the basis of the private insurance exception. It has been explained on various grounds, which may be grouped under three main headings. One is concerned with the strength of the causal connection between receipt of the benefit and the defendant’s breach, a second relates to the nature of the benefit, and a third concerns a variety of policy considerations that may be served by either deducting or not deducting the benefit.

[ 50 ] Before turning to those issues, however, I must address a contention advanced by my colleague Rothstein J. He maintains that application of the collateral benefit or private insurance exception is not appropriate where the plaintiff’s cause of action and his right to a particular benefit arise from the same contract. I respectfully do not accept that there is or should be any such categorical “single contract” rule in relation to compensating advantages. This proposition is not consistent with this Court’s jurisprudence.

[ 51 ] In Jack Cewe , unemployment insurance benefits were not deducted from wrongful dismissal damages. The Court held that the benefits were the “consequence of the contract of employment”, making them similar to contributory pension benefits: p. 818. Thus, although the Court considered that the benefits and the claim for damages arose as a consequence of the same contract, the benefits were not deducted from the wrongful dismissal damages. Thus, my colleague’s proposition is contradicted by a leading authority from this Court on the deduction of benefits from wrongful dismissal damages.

[ 52 ] The Sylvester case, from this Court, does not lay down any such broad “single contract” rule. If that had been the Court’s view, it would have provided a much simpler solution to the issue in Sylvester than the one it unanimously adopted. Of course, in Sylvester , the sick leave benefits and the claim for wrongful dismissal damages both arose from the contract of employment, but the Court did not rely on, or even mention, the broad “single contract” rule advanced by my colleague. On the contrary, Major J., writing for the Court, was careful not to articulate any broad “single contract” rule in relation to compensating advantages. He stated that

[t]here may be cases where an employee will seek benefits in addition to damages for wrongful dismissal on the basis that the disability benefits are akin to benefits from a private insurance plan for which the employee has provided consideration. This is not the case here. . . . The issue whether disability benefits should be deducted from damages for wrongful dismissal where the employee has contributed to the disability benefits plan was not before the Court. [Emphasis added; para. 22.]

Of course, whether the employee contributes to the benefits or not, they equally arise under the employment contract. The fact that the Court explicitly left this point open is inconsistent with the Court intending to adopt the broad “single contract” rule espoused by Rothstein J. Sylvester teaches that, where a cause of action and a benefit arise under the contract of employment, we must look first to that contract to determine the issue of whether an employment benefit should be deducted from wrongful dismissal damages. As in Sylvester , Mr. Waterman’s contract of employment is silent on this issue, so we must attempt to discern the parties’ intentions in light of the express terms of the contract of employment.

[ 53 ] I return to the three areas of controversy in relation to the basis of the private insurance exception .

(i) Strength of Connection to the Defendant’s Breach

[ 54 ] The strength-of-connection factor has often been referred to in the cases. The argument is that private insurance benefits (and benefits considered analogous to them) should not be deducted because they result from the plaintiff’s contract of insurance, not from the defendant’s wrongful act. This was part of the reasoning in Bradburn v. Great Western Railway Co. (1874), L.R. 10 Ex. 1 , but at the distance of 140 years, this analysis seems artificial. Moreover, scholars have pointed out that decisions about legal as opposed to factual causation often simply disguise the true policy reasons underlying the decisions: see, e.g., Ogus, at p. 94; Burrows, at p. 162. In the leading English case on the private insurance exception, Parry , Lord Pearce commented that strict principles of causation do not provide a “satisfactory line of demarcation” between benefits that are and are not deductible: p. 34. While, as discussed, considering the connection between the breach and the benefit helps to identify that there is an issue about whether the benefit should be deducted, principles of causation do not provide reliable markers of whether a benefit should be deducted or not.

(ii) The Nature and Purpose of the Benefit

[ 55 ] The nature and purpose of the benefit, on the other hand, is often a better explanation of why private insurance benefits should or should not be deducted. Two factors relating to the nature of the benefit have been particularly important: whether the benefit is an indemnity for the loss caused by the defendant’s breach and whether the plaintiff has directly or indirectly paid for the benefit.

[ 56 ] I will not attempt to lay down general principles that will cover all possible types of benefits. However, as we shall see, a review of this Court’s jurisprudence supports the following general propositions (subject, of course, to statutory or contractual provisions to the contrary).

• Benefits have not been deducted if (a) they are not intended to be an indemnity for the sort of loss caused by the breach and (b) the plaintiff has contributed to the entitlement to the benefit: Gill ; Guy.

• Benefits have not been deducted where the plaintiff has contributed to an indemnity benefit: Jack Cewe ; Cunningham .

• Benefits have been deducted when they are intended to be an indemnity for the sort of loss caused by the breach but the plaintiff has not contributed in order to obtain entitlement to the benefit: Sylvester ; Ratych .

[ 57 ] The pension benefit in this case was not intended to be an indemnity for lost wages and Mr. Waterman contributed to the acquisition of his pension through his years of service. This, no doubt, is why it has never been argued that the benefits should be deducted under the principle of mitigation. The pension benefit, therefore, is the type of benefit which should not be deducted. The reasoning leading me to this conclusion follows.

[ 58 ] I begin my review with the decision of the House of Lords in Parry , which is the foundation of much of the Canadian jurisprudence. Lord Reid ultimately based his conclusion that the benefit (a pension) should not be deducted based on its “intrinsic nature”: “A pension is intrinsically of a different kind from wages. . . . [W]ages are a reward for contemporaneous work, but . . . a pension is the fruit, through insurance, of all the money which was set aside in the past in respect of his past work. They are different in kind”: p. 16. Lord Pearce also considered the nature and purpose of the benefit when he asked: “Is there anything else in the nature of these pension rights derived from work which puts them into a different class from pension rights derived from private insurance? Their ‘character’ is the same”: p. 37. Lord Wilberforce also focused on the nature of the pension benefit, noting that it did not prevent the injured officer from taking other paid employment, whether it be for a wage that was less, equal to or more than his police officer’s salary: p. 42.

[ 59 ] The nature and purpose of the benefit was central to the minority’s reasoning in Cunningham . While the majority was concerned with authority, fairness and deterrence, the minority refocused the analysis on the nature of the benefit, distinguishing between “indemnity” and “non-indemnity” insurance. The former should be deductible, while the latter should not:

This distinction is critical to a discussion of collateral benefits. If the insurance money is not paid to indemnify the plaintiff for a pecuniary loss, but simply as a matter of contract on a contingency, then the plaintiff has not been compensated for any loss. He may claim his entire loss from the negligent defendant without violating the rule against double recovery. [pp. 371-72]

[ 60 ] Importantly, the minority judges accepted that the dominant tide of the jurisprudence in the common law world is that non-indemnity pension benefits should not be deducted: Cunningham , at p. 376. Although they mostly do not rely on the private insurance exception, Commonwealth decisions conclude that pension benefits should not be deducted from a damages award because pension benefits are not meant to compensate the plaintiff for the injury or breach of contract or to act as wage replacement. See for example: National Insurance Co. of New Zealand Ltd. v. Espagne (1961), 105 C.L.R. 569 ; Graham v. Baker (1961), 106 C.L.R. 340 ; Parry ; Smoker v. London Fire and Civil Defence Authority , [1991] 2 A.C. 502 (H.L.) . In Hopkins v. Norcross plc , [1993] 1 All E.R. 565 (Q.B.) , the High Court applied this reasoning to the deductibility of pension benefits in a wrongful dismissal suit. The reasoning is also consistent with the decision of the U.K. Employment Appeal Tribunal in Knapton v. ECC Card Clothing Ltd. , [2006] I.C.R. 1084. The non-deductibility of pension benefits was affirmed by the New Zealand Court of Appeal in Gilbert v. Attorney-General , [2010] NZCA 421, 8 N.Z.E.L.R. 72. This is consistent with the approach in Guy , discussed earlier, which concerned pension benefits that were clearly not intended to be an indemnity for loss of earnings due to an inability to work. They were held not to be deductible from damages for loss of earnings payable by those responsible for the plaintiff’s inability to work.

[ 61 ] The nature of the benefit was also an important factor in the Court’s decision to deduct employer-funded disability payments from wrongful dismissal damages in Sylvester . The Court’s analysis looked first to the nature and purpose of the benefit and, in particular, to the question of whether the benefit is in the nature of an indemnity for the sort of loss caused by the defendant’s breach of contract. The fact that the benefit was intended to be an indemnity for wage loss was one of the reasons for the Court’s conclusion that the benefit should be deducted.

[ 62 ] Reliance on the distinction between indemnity and non-indemnity benefits is sound in principle. As McLachlin J. pointed out in her dissenting reasons in Cunningham , if the benefit “is not paid to indemnify the plaintiff for a pecuniary loss, but simply as a matter of contract on a contingency”, the benefit cannot be seen as having compensated the plaintiff for that pecuniary loss: pp. 371-72. If that is the case, the arguments in favour of deducting the benefit are weaker in the sense that IBM is asking to deduct apples from oranges.

[ 63 ] The fact that Mr. Waterman’s pension comes from a defined benefit plan does not change its nature as a non-indemnity benefit.

[ 64 ] The Court in Sylvester also considered another factor — that the plaintiff had not contributed to obtain the benefit by paying for it directly or indirectly — in support of its conclusion that the benefit should be deducted from the damages. This factor has often been mentioned and relied on in the cases.

[ 65 ] For example, the Court first applied Parry in the 1973 case of Gill , and reaffirmed it in Guy . In both cases, the Court emphasized that the plaintiff had directly or indirectly paid for the benefit in question. As Ritchie J., writing for the Court, put it in Guy :

. . . this contributory pension is derived from the appellant’s contract with his employer and . . . the payments made pursuant to it are akin to payments under an insurance policy. This view is in accord with the judgment of the House of Lords in Parry v. Cleaver , which was expressly approved in this Court in the reasons for judgment of Mr. Justice Spence in Canadian Pacific Ltd. v. Gill . . . . [p. 762]

[ 66 ] This line of reasoning was repeated in Jack Cewe , which held that contributory unemployment insurance benefits were not deductible from wrongful dismissal damages. This factor was also an important one in Cunningham . As Cory J. put it, on behalf of the majority: “The application of the insurance exception to benefits received under a contract of employment should not be limited to cases where the plaintiff is a member of a union and bargains collectively. Benefits received under the employment contracts of non-unionized employees will also be non-deductible if proof is provided of payment in some manner by the employee for the benefits ”: p. 408 (emphasis added). The majority found that there was evidence of such payment and held that the benefit should not be deducted.

[ 67 ] While the cases from this Court have referred to whether the plaintiff has directly or indirectly contributed to the benefit, there are strong arguments against giving this consideration much weight as an explanation of why particular benefits should or should not be deducted. As McLachlin J. pointed out in her partial dissent in Cunningham , reliance on this factor may be seen as inconsistent with legal principle and logic. With respect to legal principle, the defendant takes the plaintiff as he or she is and the plaintiff is compensated for his or her actual loss and no more. As a matter of logic, it does not seem right to say that deducting the benefits deprives the plaintiff of the contributions made to gain entitlement to those benefits — whether deducted from damages or not, the plaintiff receives the benefits: Cunningham , at pp. 381-83; for a critique of reliance on this factor, see also Ogus, at pp. 226-27.

[ 68 ] The pension benefits in issue in this case are not an indemnity for loss of wages and, as we shall see, pension benefits earned through years of service are invariably found to be contributory. The fact that the pension plan here is a defined benefit plan does not detract from that conclusion. As a result, the problem highlighted in the difference between the majority and the dissent in Cunningham , i.e. how to treat indemnity benefits to which the plaintiff contributed, does not arise in this case.

[ 69 ] I conclude from this review that whether the benefit is in the nature of an indemnity for the loss caused by the defendant’s breach and whether the plaintiff has directly or indirectly paid for the benefit have been important explanations of why particular benefits fall, or do not fall, within the private insurance exception. The Court has been sharply and closely divided on the issue of the deduction for an indemnity benefit to which the plaintiff has contributed. However, there is no decision of the Court of which I am aware that has required deduction of a non-indemnity benefit to which the plaintiff has contributed, like the pension benefits in this case.

(iii) Broader Policy Considerations

[ 70 ] Three main policy considerations have often been advanced to explain why a benefit should or should not be deducted: punishment, deterrence, and the provision of incentives for socially responsible behaviour.

[ 71 ] The private insurance exception has often been justified on the basis that deducting the benefit from the damages reduces their punitive and deterrent value. However, the notion that the exception was intended to have a punitive and deterrent value has been widely, and, in my view, soundly, criticized. Authors agree that punitive and deterrent value ought not to be relied on to explain why a benefit is or is not deducted: see J. G. Fleming, “The Collateral Source Rule and Contract Damages” (1983), 71 Cal. L. Rev. 56, at pp. 58-59; J. Marks, “Symmetrical Use of Universal Damages Principles — Such as the Principles Underlying the Doctrine of Proximate Cause — to Distinguish Breach-Induced Benefits That Offset Liability From Those That Do Not” (2009), 55 Wayne L. Rev. 1387, at p. 1420; J. M. Perillo, “The Collateral Source Rule in Contract Cases” (2009), 46 San Diego L. Rev. 705, at p. 716; Ogus, at p. 225; Burrows, at pp. 162-63. This view is supported by both the High Court of Australia and the House of Lords: see National Insurance Co. , at p. 571, per Dixon C.J., and Parry , at p. 33. In Parry , Lord Pearce put it this way at p. 33: “The word ‘punitive’ gives no help. It is simply a word used when a court thinks it unfair that a defendant should be saddled with liability for a particular item.” I would add that it is hard to defend punishment and deterrence as rationales against the incisive critique advanced by McLachlin J. in her dissenting reasons in Cunningham , at pp. 383-84. I conclude that it is unsound to rely on a punitive or deterrent justification for the private insurance exception, particularly in breach of contract cases where fault is not an operating concept.

[ 72 ] This is not to say, however, that the approach to damages does or should ignore the underlying purposes of the substantive obligations the breach of which they seek to remedy. If, for example, an important purpose of the law of contracts is to protect the reasonable expectations of the parties to a contract, it is appropriate to consider how well the award of damages furthers that purpose in a particular case: see, e.g., A. Swan and J. Adamski, Canadian Contract Law (3rd ed. 2012), at §1.27. This consideration may be taken into account along with the other principles of damages law in order to ensure that there is a good “remedial fit” between the breach of obligation and the remedy.

[ 73 ] The private insurance exception has also been justified by the incentives it may provide. For example, deducting benefits that plaintiffs have provided for themselves might discourage plaintiffs from acting prudently in obtaining that sort of protection. This, however, has been a controversial explanation. The majority relied on it in Cunningham , but it was trenchantly criticized by the dissent and a similar critique has been made by scholars: see, e.g., Ogus, at pp. 226-27.

[ 74 ] In my view, we should be cautious about relying too heavily on the incentives that may result from deducting or not deducting. There will sometimes be little basis in fact for supposing that either deducting or not deducting certain benefits will have any impact on people’s behaviour. For example, do we think it likely that deducting insurance benefits will discourage people from buying insurance? The coverage is not limited to situations in which there will be legal recourse against a defendant. Even when legal recourse is available, it will likely require a longer and more expensive process, as compared to making an insurance claim. Nor is it likely that people will be less ready to buy insurance if they are not doubly compensated in cases in which fault can be established. It seems to me that we should generally rely on these broader policy concerns only when they are directly related to the particular benefit in issue and when there is some reasonable basis in fact or experience to suppose that deducting or not deducting will actually serve the policy objective.

[ 75 ] Sylvester provides an example of grounding policy considerations in the facts of the case. The result in that case was supported by the fact that deducting the disability benefits from wrongful dismissal damages ensured that all affected employees would receive equal damages: if the benefits were not deducted, a dismissed employee collecting disability benefits would receive more compensation than would the employee who is dismissed while working (para. 21). In the same paragraph, the Court considered the incentives created by the deduction or non-deduction of the disability benefits: failing to deduct the disability benefits could be an undesirable deterrent to employers establishing disability benefit plans. These concerns are directly related to the benefits in question and have a reasonable basis in fact.

[ 76 ] From this review of the authorities, I reach these conclusions:

(a) There is no single marker to sort which benefits fall within the private insurance exception.

(b) One widely accepted factor relates to the nature and purpose of the benefit. The more closely the benefit is, in nature and purpose, an indemnity against the type of loss caused by the defendant’s breach, the stronger the case for deduction. The converse is also true.

(c) Whether the plaintiff has contributed to the benefit remains a relevant consideration, although the basis for this is debatable.

(d) In general, a benefit will not be deducted if it is not an indemnity for the loss caused by the breach and the plaintiff has contributed in order to obtain entitlement to it.

(e) There is room in the analysis of the deduction issue for broader policy considerations such as the desirability of equal treatment of those in similar situations, the possibility of providing incentives for socially desirable conduct, and the need for clear rules that are easy to apply.

(3) Application to This Case

[ 77 ] Where would these factors lead us in this case? In my view, they clearly support not deducting the retirement pension benefits from wrongful dismissal damages. The retirement pension is not an indemnity for wage loss, but rather a form of retirement savings. While the employer made all of the contributions to fund the plan, Mr. Waterman earned his entitlement to benefits through his years of service. As the plan states, its primary purpose is “to provide periodic pension payments to eligible employees . . . after retirement . . . in respect of their service as employees”: art. 1.01, A.R., at p. 117. Thus, it seems to me that this case falls into the category of cases in which the insurance exception has always been applied: the benefit is not an indemnity and the employee contributed to the benefit. This result is consistent with the dominant view in the case law and among legal scholars: Guy ; Gill ; Chandler ; Emery ; Parry ; Ogus, at p. 223.

[ 78 ] To conclude, the compensation principle should not be applied strictly in this case because the pension benefits fall within the private insurance exception and should not be deducted from the wrongful dismissal damages.

C. Does the Court’s Decision in Sylvester Support IBM’s Position That the Pension Benefits Must Be Deducted?

[ 79 ] I turn to IBM’s second main argument, that the Court’s decision in Sylvester supports its position that the pension benefits must be deducted here. In my view Sylvester does not support that result.

[ 80 ] The issue in Sylvester was whether damages for wrongful dismissal should be reduced by the amount of disability benefits paid during the notice period from an employer-funded plan. The Court’s analysis addressed three factors: the nature of the benefit, the intentions of the parties as reflected in the employment contract, and some broader policy considerations. When these factors are considered in light of the facts of this case, they lead to the opposite conclusion than they did in Sylvester .

[ 81 ] The Court in Sylvester began by looking at the nature of the benefit. Was it intended to be a substitute (i.e. an indemnity) for wages payable during the period of reasonable notice? For two reasons, the Court determined that they were. First, the disability benefits were a wage replacement benefit. It was clear from the terms of the plans that the benefits were intended to continue the employee’s earnings in the event the employee was unable to work due to illness or injury. Second, the disability benefits would be reduced by other income received by the employee, including other disability income, wage continuation plan benefits, pension benefits, workers’ compensation benefits and salary from other employment: para. 14. They were therefore not freestanding entitlements — they were linked to and defined by the extent of actual income loss. (As I have already noted, the Court was also careful not to opine on whether the result would be the same if the employee had contributed money or money’s worth in order to obtain the benefit. The Court specifically left open the question of whether “disability benefits should be deducted from damages for wrongful dismissal where the employee has contributed to the disability benefits plan”: para. 22.)

[ 82 ] The benefit in issue in this case is of an entirely different nature. Unlike the disability benefits in Sylvester , the pension benefit is clearly not an indemnity benefit for loss of salary due to inability to work. The purpose of the pension benefits, as expressed in the plan documents, “is to provide periodic pension payments to eligible employees . . . after retirement and until death in respect of their service as employees”: art. 1.01, A.R., at p. 117. The pension plan is, in essence, a retirement savings vehicle to which an employee earns an absolute entitlement over time. Benefits are determined by years of service and salary level. An employee who leaves employment after 10 or more years of service receives either a deferred pension or a transfer of the lump sum commuted value of the pension entitlement to a locked-in retirement vehicle. Pensionable earnings are credited at 100 percent of salary while on approved unpaid leave or short-term disability. Moreover, unlike the disability payments in Sylvester , pension payments or entitlements are not in general reduced by other income or benefits received by the recipient. Mr. Waterman could have retired, drawn his full pension, and drawn a full salary from another employer. Pension benefits are clearly not intended to provide an indemnity for loss of income.

[ 83 ] There is an even more fundamental difference. As Prowse J.A. points out in her reasons in the Court of Appeal, pension benefits like those in issue here bear many of the hallmarks of a property right. They, as she put it, are regarded as belonging to the employee:

. . . although the payments under the [Defined Benefit Pension] Plan are made wholly by IBM, they are made “on behalf of” the employee. This is also reflected in IBM’s [Defined Contribution] Plan, where employer contributions are attributed to a fund in the name of the employee. In both instances, the pension benefits are regarded as belonging to the employee . They have the right to designate beneficiaries of the benefit; they can elect to transfer their pension account to another locked-in RRSP or to another employer after 10 years of service upon leaving IBM; there is a provision for a lump-sum pay-out on retirement in the case of “small pensions” (of lesser magnitude than that enjoyed by Mr. Waterman (Article 10.08)); and, in many jurisdictions, their pension rights are divisible between spouses on marriage breakdown. [Emphasis added; para. 60.]

[ 84 ] This view is supported by basic principles of pension law. Mr. Waterman’s pension was vested. As A. Kaplan and M. Frazer explain in Pension Law (2nd ed. 2013), at p. 203:

Vesting is the “foundation stone” of employee protections upon which pension regulation is based . . . . An employee who is vested has an enforceable statutory right to the accrued value of his or her pension benefit earned to date, even if the employee terminates employment and plan membership prior to retirement age. It is the vesting of pension benefits that shift our perception of pensions from purely contractual entitlements to quasi-proprietary interests.

[ 85 ] Pension benefits have consistently been viewed as an entitlement earned by the employee. As Lord Reid put it in Parry , at p. 16: “The products of the sums paid into the pension fund are in fact delayed remuneration for [the employee’s] current work. That is why pensions are regarded as earned income.” The pension is therefore a form of retirement savings earned over the years of employment to which the employee acquires specific and enforceable rights. This is no less the case because the pension benefits were not reduced by the wrongful dismissal; had they been, there would be no collateral benefit problem and no question of deduction. It is useful to ask this question: In light of the contract of employment, would the parties have intended to use an employee’s vested pension entitlements to subsidize his or her wrongful dismissal? In my view, the answer must be no. As Joseph M. Perillo writes:

Suppose an employer fires an employee without justification, breaching a contract of employment, and the employee turns to his or her savings account for living expenses. No one would argue that the employee’s recovery against the employer should be diminished by the employee’s withdrawals from savings. The savings account is a collateral source. To the extent that another collateral source resembles a savings account, the plaintiff should be able to recover damages without a deduction for the amount received from the collateral source. [Emphasis added; p. 706.]

[ 86 ] My colleague Rothstein J. does not accept that the different nature of the benefits in issue here and in Sylvester is a relevant distinction between the two cases. However, Major J., writing for a unanimous Court in Sylvester , clearly thought it was. His first reason for deciding that the benefits ought to be deducted was that “the disability benefits were intended to be a substitute for the respondent’s regular salary”: para. 14. In other words, it was a key aspect of the Court’s reasoning in Sylvester that the benefit in issue was intended to be an indemnity for wage loss. I find it impossible to dismiss the first reason the Court in Sylvester gave for its decision as irrelevant.

[ 87 ] The Court in Sylvester then turned to the contract of employment. The goal was to see if it shed any light on the parties’ intentions with respect to the receipt of both damages for wrongful dismissal and disability benefits. Contrary to the view of my colleague Rothstein J., the relevant question was not what Mr. Sylvester was entitled to under his contract in the event that his employer had not breached it. The question was whether the contract expressly or impliedly provided for him to receive both disability benefits and damages for wrongful dismissal: para. 13. Although the employment contract in Sylvester (as in this case) did not expressly address that question, it did so by implication. The receipt of both disability benefits and wages was not possible in any circumstances under the contract of employment. Moreover, other income of any nature had to be deducted from the amount of the disability payments. This suggested that the parties did not intend Mr. Sylvester to receive both disability benefits and damages representing lost wages during the notice period. As Major J. put it:

The respondent’s contractual right to damages for wrongful dismissal and his contractual right to disability benefits are based on opposite assumptions about his ability to work and it is incompatible with the employment contract for the respondent to receive both amounts . The damages are based on the premise that he would have worked during the notice period. The disability payments are only payable because he could not work. It makes no sense to pay damages based on the assumption that he would have worked in addition to disability benefits which arose solely because he could not work. This suggests that the parties did not intend the respondent to receive both damages and disability benefits. [Emphasis added; para. 17.]

[ 88 ] As I read Sylvester , this analysis does not suggest that we should focus narrowly on the precise provisions of the employment contract, unless of course they deal expressly with the issue of whether pension benefits should be deducted from wrongful dismissal damages. In the absence of such an explicit provision — and, as in Sylvester , there is no explicit provision in this case — we must look at the contract in an attempt to determine what the parties intended with respect to the receipt of both wrongful dismissal damages and pension benefits.

[ 89 ] When we examine the employment contract in this case, the picture is much less clear than it was in Sylvester . It is true that because Mr. Waterman was between the ages of 65 and 71 at the time of his dismissal and qualified for his full pension, he could not in fact receive both employment income from IBM and pension benefits. However, looking at the contract as a whole, it is not a fair implication that the parties agreed that pension entitlements should be deducted from wrongful dismissal damages.

[ 90 ] First, an employee who is dismissed before his date of retirement would receive, without deduction, wrongful dismissal damages and all of his or her entitlements under the plan (for example, a deferred pension or its commuted value transferred to a locked-in savings vehicle). No one has suggested that these amounts would in any way affect wrongful dismissal damages. In fact, the value of any pension entitlements lost during the notice period would be a compensable loss in a wrongful dismissal action: see, e.g., J. R. Sproat, Wrongful Dismissal Handbook (6th ed. 2012), at pp. 6-51 to 6-52.6. Second, a retired employee would receive, in full, both his pension benefits and any employment income earned from another employer. There is nothing before us to suggest that a retired IBM employee could not obtain employment with another employer and keep both his or her pension income and the new employment income. Third, once an employee reaches age 71, he or she could receive in full both employment income from IBM and pension benefits: plan description, at p. 2 (A.R., at p. 103); plan art. 9.02 (A.R., at p. 132). In Sylvester , not only was it impossible in all circumstances to receive salary and disability benefits, it was clear that the amount of disability benefits would be reduced by any other income, whatever its source, received by the employee: para. 14. Unlike Sylvester , it cannot be said here that the rights to damages for wrongful dismissal and to pension benefits are based on opposite or incompatible assumptions. This conclusion is also consistent with the understanding of vested pension entitlements as being akin to property rights which accrue over time for the employee’s benefit.

[ 91 ] I conclude that, unlike the situation in Sylvester , Mr. Waterman’s receipt of pension benefits and wrongful dismissal damages is not based on opposite assumptions about his ability to work and it is not incompatible with the employment contract that he could receive both pension benefits and employment income.

[ 92 ] Finally, the Court in Sylvester turned to the broader policy concerns, notably that dismissed employees should be treated alike and that the incentives should encourage rather than discourage employers from setting up disability plans. As Major J. put it, at para. 21:

If disability benefits are paid in addition to damages for wrongful dismissal, the employee collecting disability benefits receives more compensation than the employee who is dismissed while working. Deducting disability benefits ensures that all affected employees receive equal damages . . . . If disability benefits are not deductible, employers who set up disability benefits plans will be required to pay more to employees upon termination than employers who do not set up plans. This deterrent to establishing disability benefits plans is not desirable. [Emphasis added.]

[ 93 ] These factors are also relevant here, although, in this case, they support not deducting rather than deducting the benefits. Unlike in Sylvester , non-deduction in this case promotes equal treatment of employees. If deduction is permitted, an employee who is eligible to receive his or her pension but has not reached 71 years of age can, by means of wrongful dismissal, be forced to retire and draw on his or her pension benefits. By contrast, an employee who is not entitled to his or her pension receives either a deferred pension or the commuted value of it plus full damages for wrongful dismissal and an employee over the age of 71 receives both pension and employment income. Deducting the benefits only in the case of employees in Mr. Waterman’s situation would constitute unequal treatment of pensionable employees. Moreover, deductibility seems to me to provide an incentive for employers to dismiss pensionable employees rather than other employees because it will be cheaper to do so. This is not an incentive the law should provide. While this is a broader policy consideration, it is directly related to the benefit in question and has a reasonable basis in fact.

[ 94 ] My colleague Rothstein J. is of the view that there is no such incentive because “with respect to the cost of dismissing pensionable and non-pensionable employees, there is a difference only in form, not substance”: para. 134. Respectfully, I cannot agree. The suggestion implicit in this is that there is a dollar for dollar correlation between the amount of the pension benefits that IBM claims should be deducted and the amount IBM contributed over time in order to fund those benefits such that it is not cheaper to dismiss a pensionable employee than one who is not eligible to collect a full pension. This proposition, however, is based on a considerable oversimplification of how pension benefits are funded and, in my respectful view, is not accurate.

[ 95 ] My colleague Rothstein J. suggests that failure to deduct earned pension benefits from wrongful dismissal damages may disadvantage other employees in the future because it may “incentivize” employers to require an employee to work through the duration of the reasonable notice period to the potential disadvantage of employees. However, the risk of such an incentive seems to me to be highly speculative. There are pluses and minuses for both the employer and employee of giving (and receiving) working notice. From the employer’s perspective, it may not be advantageous to have the employee remain on the employer’s premises during the period of working notice. In addition, the employer loses the benefit of the employee’s efforts to mitigate damages by finding alternate employment, a benefit that is often unpredictable at the time of termination. The employer is always able to negotiate before firing an employee rather than firing without first negotiating. In light of these considerations, among others, it seems to me to be highly speculative to say that refusal to deduct pension benefits will encourage employers to give working notice rather than offer severance.

[ 96 ] Finally, there is no parallel, from a policy analysis perspective, between this case and Sylvester. The Court in Sylvester was concerned that failure to deduct the non-contributory wage replacement benefits in issue there might make employers reluctant to fund wage replacement benefits. This concern does not arise here, given that the pension benefit is not intended to be an indemnity for wage loss and that the employees contribute to the cost of the pension benefits. Moreover, any employer who has this concern (and it must be said that the scarcity of reported cases on the point suggest that it arises very uncommonly) can address it by adding appropriate language to the pension plan text.

[ 97 ] To conclude: in this case, the pension benefits are markedly different in nature than the disability benefits in issue in Sylvester , the intention of the parties in relation to the issue of deduction is much more uncertain in this case than in Sylvester and the broader policy considerations point in the opposite direction. Unlike the disability benefits in Sylvester , the pension benefits are not an indemnity for loss of earnings, they are not reduced by other benefits or income received and the employee over time receives a legal entitlement to the commuted value of the benefits. Unlike the situation respecting disability benefits in Sylvester , there is no general bar against an employee receiving both pension income and employment income, and receipt of the benefits and income is not based on opposite or incompatible assumptions. Pension benefits are not reduced by other income. Not deducting the pension benefits serves the goal of equal treatment of employees and provides better incentives for just treatment of all employees.

[ 98 ] I conclude, therefore, that Sylvester does not support IBM’s position in this case, and that it, in fact, supports the conclusion that the pension benefits should not be deducted from the wrongful dismissal damages.

V. Disposition

[ 99 ] I would dismiss the appeal with costs throughout.

The reasons of McLachlin C.J. and Rothstein J. were delivered by

Rothstein J. (dissenting) —

I. Introduction

[ 100 ] Richard Waterman brought this suit alleging that his employer, IBM Canada Ltd., breached his employment contract by failing to provide him with reasonable notice of his termination. The trial judge found, and it is now undisputed, that Mr. Waterman was entitled to 18 months more notice than he was given, and that he is accordingly entitled to the salary he would have earned if he continued to work during that period. During the 18-month period, IBM paid Mr. Waterman monthly pension benefits under the assumption that he was retired. The sole issue in this case is whether the pension benefits that IBM paid to Mr. Waterman during the 18-month notice period must be deducted in calculating the appropriate damages award.

[ 101 ] I agree with the majority that a straightforward application of the governing principle of contract damages — that the non-breaching party be placed in the position he would have been in had the contract been performed — leads to the conclusion that deduction is required (see para. 2). The parties agree that, had Mr. Waterman been given reasonable notice and worked through the reasonable notice period, he would have received his salary, but not his pension, until the notice period elapsed. Deducting the pension benefits IBM paid him during the reasonable notice period thus puts him in the position he would have been in had the contract been performed and failure to deduct gives him a windfall.

[ 102 ] However, the majority accepts Mr. Waterman’s argument that he should be allowed a windfall because his pension benefits are subject to the “private insurance” exception. I would reject that argument. This case requires the Court to assess Mr. Waterman’s loss under the terms of a single contract which gave rise to both Mr. Waterman’s right to reasonable notice and his right to pension benefits. The private insurance exception has no application to such a case. Where the Court is called upon to assess loss under a single contract, the plaintiff’s entitlement turns on the ordinary governing principle that he should be put in the position he would have been in had the contract been performed.

[ 103 ] It is important to note that not all pension plans are alike. Mr. Waterman’s pension plan is a defined benefit plan, under which IBM undertook to provide Mr. Waterman with pension benefits from the time of his retirement until the time of his death, based on a predetermined formula. That is to say that, from Mr. Waterman’s perspective, upon retirement, he would receive his defined benefits from an unlimited fund for the rest of his life. For this reason, Mr. Waterman’s receipt of pension benefits during the reasonable notice period did not affect his future entitlement to pension benefits and deducting the benefits does not have the effect of taking anything away from Mr. Waterman. Rather, not deducting has the effect of giving Mr. Waterman more than he bargained for and charging IBM more than it agreed to pay.

II. Factual Background

[ 104 ] Mr. Waterman was an employee of IBM for approximately 42 years. At the time he was terminated, he was 65 years old.

[ 105 ] As an employee of IBM, Mr. Waterman became a member of the company’s defined benefit pension plan. Under the terms of the plan, IBM was required to make contributions to the pension plan on behalf of its employees and, upon an employee’s eligibility to receive benefits, IBM would provide the employee with monthly benefits according to a predetermined formula until the employee’s death. An employee became eligible to receive his monthly benefits upon retiring after reaching the age of 65. An employee whose employment was terminated prior to the age of 65 could receive his pension benefits upon turning 65 or could elect to transfer the actuarial equivalent of his accrued pension to a new employer. An employee also became eligible to receive his benefits upon reaching the age of 71, independent of whether he had been terminated or retired, which, according to the parties, was necessary for the plan to comply with income tax regulations. At the time Mr. Waterman was terminated by IBM, the monthly payment he would receive upon becoming eligible had already been determined for several years.

[ 106 ] IBM terminated Mr. Waterman in March 2009. It provided him with two months’ working notice, after which it would consider him retired and begin paying him his pension benefits. The trial judge found, and it is now undisputed, that IBM was required to give Mr. Waterman an additional 18 months of notice.

[ 107 ] The termination letter also offered Mr. Waterman a separation payment in exchange for a general release from liability. As explained later, the separation offer would have provided Mr. Waterman with more than he would have earned had he been given the full 20-month notice period and worked through the notice period. Mr. Waterman declined IBM’s separation offer. He continued to work for IBM during the two-month notice period that he was given, and thereafter began collecting monthly pension benefits from IBM. On June 11, 2009, Mr. Waterman initiated this action to enforce his contractual right to be provided with reasonable notice of his termination.

[ 108 ] In September 2009, Mr. Waterman obtained alternative employment as a part-time insurance salesman.

III. Procedural History

A. Supreme Court of British Columbia, 2010 BCSC 376 , 2010 CLLC ¶210-021

[ 109 ] After a summary trial, Goepel J. found that IBM breached Mr. Waterman’s employment contract by failing to provide him with reasonable notice. Goepel J. held that IBM was required to provide Mr. Waterman with an additional 18 months of notice beyond the two months that had been provided. As a result, Mr. Waterman was entitled to the salary he would have earned and benefits he would have accrued if he had continued to work for IBM during that time.

[ 110 ] Goepel J. did not deduct the pension benefits that IBM paid to Mr. Waterman during the notice period in calculating his damages. Goepel J. expressed the view that he was bound by the Court of Appeal for British Columbia’s decision in Girling v. Crown Cork & Seal Canada Inc. (1995), 1995 954 (BC CA) , 9 B.C.L.R. (3d) 1, in which it was held that pension benefits should not be deducted from wrongful dismissal damages. He acknowledged the possibility that Girling was no longer an accurate statement of the law in light of this Court’s decision in Sylvester v. British Columbia , 1997 353 (SCC) , [1997] 2 S.C.R. 315, but found it incumbent upon him to follow Girling for reasons of judicial comity.

[ 111 ] Based on this reasoning, Goepel J. awarded Mr. Waterman $93,305 in damages, which reflected the salary and benefits he would have earned if he had worked through the additional 18 months of notice, less the income earned from his new employment during that period.

B. Court of Appeal for British Columbia, 2011 BCCA 337 , 20 B.C.L.R. (5th) 241

[ 112 ] Writing for a unanimous panel, Prowse J.A. dismissed IBM’s appeal.

[ 113 ] Prowse J.A. observed that the approach the Court of Appeal had previously taken in Girling was rejected by this Court’s decision in Sylvester . In particular, Sylvester rejected the Court of Appeal for British Columbia’s approach of treating agreements for employee benefits as contracts distinct from the employment contract. According to Prowse J.A., under Sylvester , Mr. Waterman’s entitlement to both salary and payment of his pension benefits during the notice period turned on the construction of the contractual arrangement between the parties.

[ 114 ] After reviewing the terms of Mr. Waterman’s employment contract and IBM’s defined benefit plan, Prowse J.A. found that there was no express provision addressing Mr. Waterman’s rights in the event of wrongful dismissal. Prowse J.A. turned to consider what the parties would have intended had they put their minds to that circumstance. She concluded that, although there was no evidence regarding the parties’ intention, had they considered the issue, they would not have intended for Mr. Waterman’s pension benefits to be deducted from wrongful dismissal damages.

[ 115 ] Prowse J.A. also concluded, at para. 62, that “the pension benefits in issue are also properly characterized as a form of non-deductible, non-indemnity insurance”, as described by McLachlin J., as she then was, in Cunningham v. Wheeler , 1994 120 (SCC) , [1994] 1 S.C.R. 359.

IV. Issue

[ 116 ] The only issue before this Court is whether the pension benefits IBM paid to Mr. Waterman during the reasonable notice period should have been deducted in calculating his damages.

V. Analysis

[ 117 ] My analysis proceeds in two stages. First, I consider whether it is necessary to deduct the pension benefits Mr. Waterman received during the reasonable notice period in order to put him in the position he would have been in had the contract been performed — i.e. had he been given reasonable notice and worked through the end of the reasonable notice period. Second, I consider whether there is a basis for applying the private insurance exception, which allows a plaintiff to receive excess compensation in certain circumstances. I conclude that, to put Mr. Waterman in the position he would have been in had the contract been performed, the pension benefits he received must be deducted. The private insurance exception is not applicable to this case.

A. Contract Damages for Wrongful Dismissal

[ 118 ] The governing principle for damages upon breach of contract is that the non-breaching party should be provided with the financial equivalent of performance (J. D. McCamus, The Law of Contracts (2nd ed. 2012), at p. 871). With respect to wrongful dismissal, damages should “represent the salary the employee would have earned had the employee worked during the notice period, less any amounts credited to mitigation” ( Sylvester , at para. 1 ). I agree with the majority that applying this rule leads to the deduction of the pension benefits in this case.

[ 119 ] In Sylvester , this Court considered whether disability benefits received by a wrongfully terminated employee during his reasonable notice period should be deducted from damages for wrongful dismissal. Major J., writing for a unanimous Court, held that deduction was required. He explained that employer-provided benefits should not be considered as “distinct from the employment contract, but rather as integral components of it” (para. 13). As such, “[t]he question of deductibility . . . turn[ed] on the terms of the employment contract and the intention of the parties” (para. 12).

[ 120 ] Major J. went on to explain that damages for wrongful dismissal were “ based on the premise that the employee would have worked during the notice period” (para. 15). The employee’s “contractual right to damages for wrongful dismissal and his contractual right to disability benefits [were] based on opposite assumptions about his ability to work and it [was] incompatible with the employment contract for the respondent to receive both amounts” (para. 17). Based on this analysis, Major J. concluded: “It makes no sense to pay damages based on the assumption that [the plaintiff] would have worked in addition to disability benefits which arose solely because he could not work” (para. 17).

[ 121 ] It follows from a straightforward application of Sylvester that deduction is required in this case. In particular, Mr. Waterman’s wrongful dismissal damages must be “ based on the premise that the employee would have worked during the notice period” ( Sylvester , at para. 15 ). Under the terms of Mr. Waterman’s employment contract, he would have been eligible to receive pension benefits only upon being terminated or retiring. Therefore, as in Sylvester , Mr. Waterman’s contractual right to wrongful dismissal damages and his contractual right to his pension are based on “opposite assumptions” about his availability to work (para. 17). It thus “makes no sense” to pay damages on the assumption that he could have earned both ( ibid. ).

[ 122 ] This conclusion is necessitated by the nature of the pension plan at issue in this case — a defined benefit plan. This plan is materially different from a defined contribution plan, and the distinction between these two types of pension plans is at the heart of my disagreement with the majority.

[ 123 ] A defined contribution plan “operates in much the same way as group registered retirement savings plans”, in that it provides an employee with a finite total amount or lump sum of retirement benefits (A. Kaplan and M. Frazer, Pension Law (2nd ed. 2013), at p. 89). It would be inappropriate to deduct pension benefits that a wrongfully terminated employee receives from a defined contribution plan because deduction would leave the employee in a worse position that he would have been in had his employment contract not been breached.

[ 124 ] In particular, in the case of a defined contribution plan, if the employee’s employment contract is performed (i.e. he is given reasonable working notice of his termination and he continues to work through the notice period), he would expect to receive his salary through the notice period and the full lump sum he would have accrued in his savings account or defined contribution plan by the end of the reasonable notice period, including whatever additions should have been made to the plan during that notice period. If, instead, the employee is wrongfully dismissed and draws benefits from his finite lump sum during the reasonable notice period, deducting the pension benefits would leave the employee with an amount equal to his salary through the notice period and the lump sum, less the amount he had withdrawn during the notice period. He would thus be awarded less than he was entitled to under his employment contract.

[ 125 ] Throughout Mr. Waterman’s argument before this Court, he has made submissions that his pension operates like a savings account. He is not alone in this respect. The Court of Appeal, at para. 48, quoted with approval language from Kent J. in Chandler v. Ball Packaging Products Canada Ltd. (1992), 2 C.C.P.B. 101 (Ont. Ct. J. (Gen. Div.)), that pension payments should be viewed “as akin to a registered Retirement Savings Plan” (para. 4).

[ 126 ] The majority too accepts the analogy. It holds that “[p]ension benefits . . . constitute a type of retirement savings” (para. 4; see also para. 85). It quotes, with emphasis, the following language from J. M. Perillo, “The Collateral Source Rule in Contract Cases” (2009), 46 San Diego L. Rev. 705, at p. 706:

Suppose an employer fires an employee without justification, breaching a contract of employment, and the employee turns to his or her savings account for living expenses. No one would argue that the employee’s recovery against the employer should be diminished by the employee’s withdrawals from savings. The savings account is a collateral source. To the extent that another collateral source resembles a savings account, the plaintiff should be able to recover damages without a deduction for the amount received from the collateral source. [Emphasis added by Cromwell J.; para. 85.]

These references may wrongly suggest that Mr. Waterman’s pension benefits came from a finite account and thus came at a cost to him. If Mr. Waterman had needed to draw from his own savings due to his wrongful dismissal, the amount he withdrew would have to be reflected in the damages award in order to put him in the position he would have been in had the contract been performed. However, analogizing Mr. Waterman’s pension to a savings account misconceives the nature of the defined benefit pension plan at issue in this case.

[ 127 ] Unlike a defined contribution plan, the defined benefit plan at issue in this case is fundamentally different from a savings account. The defined benefit plan did not provide Mr. Waterman with a finite lump sum that was partially depleted by the pension funds he received during his reasonable notice period. Rather, the plan guaranteed him fixed predetermined payments upon retirement for as long as he would live. For that reason, deducting Mr. Waterman’s pension benefits in this case does not have the effect of “taking away” benefits that he would have been entitled to had IBM not breached the contract. On the contrary, deducting provides him with exactly what he would have received had the employment contract been performed: an amount equal to his salary during the reasonable notice period and thereafter defined benefits for the rest of his life.

[ 128 ] The different outcome in the cases of defined benefit and defined contribution plans turns on a straightforward application of the governing principle of contract damages — that the non-breaching party should be placed in the same position he would have been in had the contract been performed. It has nothing to do with the collateral benefit, compensating advantages or private insurance exception. As my colleague correctly observes, those exceptions are relevant only where the plaintiff experiences “excess recovery” (para. 23). However, as the analysis above demonstrates, there is no excess recovery when pension benefits received from a defined contribution plan are not deducted or where benefits received from a defined benefit plan are deducted. In each case, the result is to put the employee in the position he would have been in had the contract been performed.

[ 129 ] At the time Mr. Waterman was wrongfully dismissed, the amount of pension benefits he was to receive upon retirement had already been determined for some time and could not have gone up if he had continued to work for IBM. If Mr. Waterman’s pension benefits could have increased during the notice period, his wrongful dismissal damages would have compensated for this loss. However, in this case, Mr. Waterman’s wrongful dismissal had no impact on his pension entitlement and, as the parties agree, there is no need to make adjustments to his damage awards based on pension entitlements that would have accrued had he worked through the reasonable notice period.

[ 130 ] The majority states that the fact that Mr. Waterman’s pension comes from a defined benefit plan does not change its nature as a contributory, non-indemnity benefit (paras. 63 and 68). However, the nature of the benefit as non-indemnity or contributory does not answer the question of whether the plaintiff will be provided with the financial equivalence of performance or will receive excess recovery. With respect, the majority reasons conflate the analysis of contract damages for wrongful dismissal with what considerations should apply with respect to the private insurance exception to contract damages. Under the governing principle of contract damages, the fact that the pension plan at issue is a defined benefit plan leads to the conclusion that the benefits must be deducted from Mr. Waterman’s wrongful dismissal damages.

[ 131 ] As an aside, not distinguishing between defined benefit and defined contribution plans may also be why the majority’s policy concern about making pensionable employees cheaper to dismiss is incorrect. The majority suggests that deducting the benefits IBM paid to Mr. Waterman during the reasonable notice period would “provide an incentive for employers to dismiss pensionable employees rather than other employees because it will be cheaper to do so”. The majority states that “[t]his is not an incentive the law should provide” (para. 93).

[ 132 ] This incentive argument is based on a false premise: that deducting pension benefits from reasonable notice damages would make it cheaper to dismiss a pensionable employee than a non-pensionable employee. That is not the case. The pension benefits that Mr. Waterman received during the notice period did not come out of thin air. With a defined benefit pension plan, the employer is solely responsible for providing the employee with the guaranteed defined benefits. In the event the payment of the defined benefits results in an actuarial deficit in the pension fund, the employer will be required to top up the fund to meet its statutory obligation to keep it fully funded. Alternatively, if the fund is operating at an actuarial surplus despite payment of the benefits, the contribution holiday that the employer may otherwise be able to take — i.e. the break from its regular contributions to the pension fund — would be reduced. In this way, withdrawal of the benefits from the pension fund, like any other payment, affects the employer’s bottom line.

[ 133 ] The majority alleges that this analysis is an oversimplification and is inaccurate (para. 94). This assertion seems to misunderstand the impact of IBM having paid pension benefits to Mr. Waterman. The analysis has nothing to do with funding the benefits over time. Rather, the analysis is simply how the pension benefits paid by IBM impacted IBM’s obligation to ensure the actuarial solvency of the pension fund, such that it would be necessary to either top up the pension fund or to refrain from taking a contribution holiday to the extent of those pension benefit payments.

[ 134 ] It follows that, with respect to the cost of dismissing pensionable and non-pensionable employees, there is a difference only in form, not substance. That is to say, in the case of an employee who is not eligible to receive his defined pension benefits, the employer compensates a dismissed employee by paying him damages equal to the salary he would have earned during the reasonable notice period. In the case of an employee who is eligible to receive his defined pension benefits, the employer pays: (1) pension benefits from the employer’s pension fund, which it is responsible for maintaining, and (2) damages equal to the salary the employee would have earned during the notice period less what it has already paid from the pension fund. In both cases, the cost to the employer is the same: an amount equal to the salary the employee would have earned had he worked through the reasonable notice period. There is thus no incentive to terminate pensionable employees.

[ 135 ] The majority emphasizes Mr. Waterman’s “specific and enforceable rights” in relation to his pension (para. 85). It is not disputed that Mr. Waterman’s pension benefits are vested and that this gives him specific and enforceable rights. However, his specific and enforceable rights remain subject to the provisions of the plan text which govern the conditions under which benefits will be paid. As a result, even though Mr. Waterman’s pension plan had vested, he could not have demanded to receive both his salary and his pension benefits had he continued to work for IBM through the reasonable notice period.

B. Applicability of the Private Insurance Exception

[ 136 ] The majority agrees that putting Mr. Waterman in the position he would have been in had the contract been performed would lead to the conclusion that the pension benefits must be deducted (para. 2). According to the majority, however, the pension benefits that IBM paid to Mr. Waterman under his employment contract on the assumption that he was retired may be treated as a “private insurance” and, thus, need not be deducted under the private insurance exception. I disagree with that conclusion. In my view, the private insurance exception has no applicability to this case.

[ 137 ] This case involves the interpretation of a single employment contract that gives rise to Mr. Waterman’s right to wrongful dismissal damages and his right to pension benefits. This Court has determined that employer-provided benefits “should not be considered contracts which are distinct from the employment contract, but rather as integral components of it” ( Sylvester , at para. 13 ). The majority is correct that the words “‘single contract’ rule” do not literally appear in Sylvester , but the reasoning in Sylvester can lead to no other conclusion (para. 52).

[ 138 ] As I will explain, in the context of a single contract, the collateral benefit or private insurance exception has no application. The reason is straightforward: where the plaintiff’s cause of action and his right to a particular benefit arise from the same contract and the plaintiff is indeed entitled to the benefits — i.e. he has “insured” himself in a manner that requires the defendant to pay the benefits — then the plaintiff will receive the benefits based on the ordinary governing principle that he should be placed in the position he would have been in had the contract been performed. There will be no need to reach the collateral benefit exception.

[ 139 ] Said another way, given that Mr. Waterman’s pension flows from the same contract under which the court must assess his loss, the need to reach the private insurance exception is itself a concession that Mr. Waterman’s pension was not “private insurance” that covered the breach in the first place. If he had “insured” the breach, he would get the benefits under the governing principle that he should be provided with what he would have expected to receive under the terms of the contract.

[ 140 ] For this reason, the majority’s approach to this case contains an inherent inconsistency: the majority concludes that Mr. Waterman had “private insurance” that allows him to keep his pension benefits in addition to his salary. To the extent Mr. Waterman had such “private insurance”, it must have come from his employment contract. However, if Mr. Waterman’s employment contract indeed allowed him to have pension benefits in addition to his salary, there would be no need to reach any exception: he would get the benefits based on what he would have expected under the terms of the contract.

[ 141 ] In addition to this troubling inconsistency, applying the private insurance exception to this case would not be consistent with the justification for the exception. The rationale for the private insurance exception is that it would be “unjust and unreasonable to hold that the money which he prudently spent on premiums and the benefit from it should enure to the benefit of the tortfeasor” ( Parry v. Cleaver , [1970] A.C. 1 (H.L.) , per Lord Reid, at p. 14). Accepting the assumption that Mr. Waterman’s work for IBM over the years is analogous to paying premiums to obtain his pension plan, it remains that the contractual terms of the pension or “insurance” he paid for allowed him to receive salary or pension benefits, but not both at the same time. In other words, this is not a case where deduction would lead to some benefit that the plaintiff paid for enuring to the benefit of the defendant. Quite to the contrary, as explained above, deducting is necessary to provide the plaintiff with the pension or “insurance” he paid for. Not deducting has the effect of the plaintiff receiving more than he expected to receive under the terms of his contract and requiring the defendant to pay more than it agreed to pay.

[ 142 ] This distinguishes the case before the court from all other cases in which the private insurance exception has been applied. Each of Guy v. Trizec Equities Ltd. , 1979 232 (SCC) , [1979] 2 S.C.R. 756, Canadian Pacific Ltd. v. Gill , 1973 2 (SCC) , [1973] S.C.R. 654, and Cunningham involved a plaintiff who was personally injured by the defendant and, upon being sued, the defendant sought to pay less than what it owed under ordinary principles of compensatory damages based on a distinct contractual or statutory benefit that the plaintiff received from a third party.

[ 143 ] Similarly, Jack Cewe Ltd. v. Jorgenson , 1980 177 (SCC) , [1980] 1 S.C.R. 812, involved a wrongful dismissal suit, in which the employer sought to have its damages reduced based on the employee’s distinct statutory entitlement to unemployment benefits. Contrary to the majority’s assertion that the benefits were derived from the employment contract, the source of the benefits was a third party — the government. As with Guy , Gill , and Cunningham , Jack Cewe was not a case in which the plaintiff’s cause of action and the benefit he received came from a single contract whose terms did not allow the plaintiff to receive both salary and benefits at the same time.

[ 144 ] Considered in terms of the justification for the private insurance exception, in each of Guy , Gill , Cunningham and Jack Cewe , the Court was faced with two choices: (1) not deduct the benefits and thus require the defendant to pay the amount equal to the plaintiff’s loss determined by ordinary principles of tort or (in the case of Jack Cewe ) contract damages, even though the plaintiff would receive more than his actual loss as a result of the benefits he received; or (2) allow the defendant to pay nothing or some amount less than the plaintiff’s loss, such that the plaintiff does not get the benefit resulting from the premiums he paid to the third party. The Court decided, consistent with the rationale for the private insurance exception, that the plaintiff — not the defendant –— should receive the benefits associated with the premiums he paid.

[ 145 ] The choice in this case is very different. The options are to (1) not deduct, requiring the defendant to pay more than it agreed to pay the plaintiff under the terms of the employment contract and awarding the plaintiff more than he bargained for; or (2) deduct, requiring the defendant to pay an amount equal to the plaintiff’s loss (i.e. the amount required to put the plaintiff in the position he would have been in had the contract been performed) and awarding the plaintiff an amount equal to his loss. In neither case do benefits that the plaintiff actually paid for enure to the defendant. The issue is whether the defendant should be required to pay twice, such that the plaintiff receives more than he bargained for under his contract, or pay once, such that the plaintiff receives exactly what he bargained for under his contract. In my view, the latter is appropriate.

[ 146 ] The fact that this case involves a single contract also distinguishes the cases the majority cites, such as United States v. Price , 288 F.2d 448 (4th Cir. 1961), and Phillips v. Western Company of North America , 953 F.2d 923 (5th Cir. 1992), in which employees sued their employers in tort for personal injuries caused by the employer. In each of those cases, the employer sought to pay less than the plaintiff’s loss from the injury, according to ordinary tort principles, based on benefits that flowed to the plaintiff from his employment contract. In neither case did the facts before the court establish that it would be inconsistent with the terms of the employment contract for the plaintiff to receive both tort damages and his employment benefits. That is in contrast to this case, where, as described above, Mr. Waterman’s contract provided that he could receive salary or pension benefits, but not both.

[ 147 ] Further, the choice before the courts in Price and Phillips was whether to (1) require the defendant to honour both of its legal duties (the legal duty to take reasonable care under tort and the legal duty to pay the plaintiff the amount promised under his employment contract) such that the plaintiff would receive compensation for his loss and the benefits he was entitled to under his employment contract; or (2) allow the defendant to offset the damages for breaching its duty of care using the benefits that it had separately promised the plaintiff in his employment contract, such that those benefits would enure to the defendant. Again, there is no parallel here. This case involves a single legal duty to honour the terms of an employment contract. The terms of that contract provided that Mr. Waterman would receive only his salary during his reasonable notice period.

[ 148 ] In sum, I would reject the idea that the private insurance exception is applicable to cases that involve a single contract that is the source of both the plaintiff’s cause of action and other benefits. In such circumstances, there is no justification for resorting to the private insurance exception because the plaintiff’s entitlement to the benefits is established based on the terms of his contract.

C. The Majority’s Treatment of Sylvester

[ 149 ] The majority has devoted an extensive portion of its reasons in attempting to distinguish this case from Sylvester and at the same time attempting to rely on Sylvester (paras. 80-98 ). As I read the majority’s reasons with respect to Sylvester , they say that, under the ordinary principles of contract damages, Sylvester would support the proposition that Mr. Waterman is entitled to both his salary and his pension benefits at the same time (paras. 88-91). Indeed, if Sylvester was authority for such a result, it is difficult to understand the majority’s resort to the private insurance exception. With respect, the majority’s analysis of Sylvester is strained. In my respectful opinion, a straightforward reading of Sylvester demonstrates that it is a fully applicable authority supporting the proposition that under a single contract of employment, barring contract terminology to the contrary, an individual cannot receive salary as if he is working and pension benefits as if he is retired. These are opposite, incompatible assumptions. Thus, salary and pension income are not payable at the same time.

D. Efficient Breach

[ 150 ] My colleague appropriately cautions against speculation about “policy” and the future impact of deduction rules. I would not resolve this case based on policy or speculation. In my view, the case should be resolved based on the terms of the parties’ contract.

[ 151 ] Only in response to the majority’s concerns about policy, I point out that while the majority’s conclusion would operate to Mr. Waterman’s benefit in this case, it would do so at the cost of other employees in the future. It is often advantageous for both employers and employees to agree to an amount for reasonable notice, rather than having the employee work through the notice period. For instance, in the case of an employer who must terminate an employee, it may be advantageous for the employer to offer the employee at least the amount he would have earned throughout the notice period in order to end the employment relationship immediately. In those circumstances, it will generally also be economically favourable for the employee to accept the offer because he will receive the full salary he would have earned if he worked through the notice period without having to work through the period. He would then be free to earn additional income from alternate employment.

[ 152 ] In fact, the record reveals that this was precisely the case here: IBM offered Mr. Waterman a separation agreement that would have provided him with even more than he would have earned if he had worked through the reasonable notice period. If IBM had provided Mr. Waterman with the additional 18 months of notice to which he was entitled and Mr. Waterman had worked through the entire notice period, he would have earned approximately $112,000 in salary and accrued benefits. Under the separation offer Mr. Waterman turned down, he would have received an $80,000 separation payment, plus an additional $38,000 in pension payments during the 18-month period. He thus would have received approximately $118,000. In addition, he would have been free to obtain income from alternative employment.

[ 153 ] This is an example of efficient breach. This Court has previously described efficient breach and cautioned courts from discouraging such a breach:

Efficient breach is what economists describe as a Pareto optimal outcome where one party may be better off but no one is worse off, or expressed differently, nobody loses. Efficient breach should not be discouraged by the courts. This lack of disapproval emphasizes that a court will usually award money damages for breach of contract equal to the value of the bargain to the plaintiff.

( Bank of America Canada v. Mutual Trust Co. , 2002 SCC 43 , [2002] 2 S.C.R. 601, at para. 31 )

[ 154 ] The majority’s approach discourages efficient breach in the context of an employer with a defined benefit pension plan who wishes to terminate an employee. This is because, all things equal, the majority approach incentivizes the employer to require the employee to work through the notice period (and avoid paying out the pension benefits) instead of offering the employee a separation package that would be economically superior for the employee. While there are always a number of competing factors that govern whether an employer makes a separation offer and what that offer contains, the majority’s approach encourages, at least to some extent, giving working notice rather than severance.

VI. Conclusion

[ 155 ] The pension benefits IBM paid to Mr. Waterman during the reasonable notice period should be deducted in assessing Mr. Waterman’s damages for wrongful dismissal. I would allow the appeal with costs throughout.

Appeal dismissed with costs throughout, McLachlin C.J. and Rothstein J. dissenting.

Solicitors for the appellant: Fasken Martineau DuMoulin, Vancouver.

Solicitors for the respondent: MacKenzie Fujisawa, Vancouver.