Enforcing a Termination Clause - Can You Sever a Partially-Illegal Termination Framework?

This pound of flesh which I demand of him

is dearly bought; 'tis mine and I will have it.

If you deny me, fie upon your law:

There is no force in the decrees of Venice.

-Merchant of Venice, Act IV, Scene 1.

Employers, concerned about having to provide potentially prohibitive 'pay in lieu of reasonable notice' when dismissing employees, often require employees to sign contracts limiting their entitlements on termination.

There are plenty of ways that employees can challenge these clauses following termination: A new employment contract for an existing relationship might be said to lack 'fresh consideration'; an old employment contract for an evolved relationship might be said to have lost its 'substratum'; a clause may be void for a conflict with the applicable employment standards legislation; etc.

In the case of Waksdale v. Swegon North America Inc., Ontario's Court of Appeal recently dramatically expanded the category of contracts void for conflicts with employment standards legislation, in a way that applies mostly to Ontario, but has implications elsewhere.

The Issue

Many employment contracts have multiple clauses governing termination by the employer.  At its most basic, written termination clauses usually have at least two facets, governing 'not for cause' and 'for cause' dismissals - along these general lines:

If we dismiss you without cause, we'll give you x notice or pay in lieu.

If we dismiss you for misconduct, you get nothing.

There can be others.  Termination by employee (resignation).  Termination by employer after termination notice by employee.  Termination by employer in certain other contingencies, like sale of business, etc.  But those are the main two - cause, and not-for-cause.

The challenge is that employment standards legislation has strict standards regarding minimum periods of notice or pay in lieu. It's illegal to reserve a right to dismiss while paying nothing, in circumstances where the statute says you have to provide notice, and that illegality voids the provision.  These statutory regimes do provide circumstances where employee misconduct can relieve you of the obligation to provide notice...but if you've drafted a contract where the 'you get nothing' termination category expands even narrowly outside the statutory exemption, then that's a problem.

This is particularly so in Ontario.  In many jurisdictions (including Alberta), the statutory exemption for termination notice uses the term "just cause" - this is essentially an invocation of the common law concepts that typically apply in a wrongful dismissal action, and which many lawyers instinctively use to draft contracts.  So in Alberta, a contract that says "We can dismiss you for just cause and pay you nothing" is not problematic.

In Ontario, on the other hand, the statutory exemption is more specifically defined:

An employee who has been guilty of wilful misconduct, disobedience or wilful neglect of duty that is not trivial and has not been condoned by the employer.

This is not quite aligned with the common law conception of just cause, which, among other issues, includes non-wilful misconduct in certain cases.  So, while an employee guilty of gross incompetence is entitled to statutory termination notice, a contract authorizing termination without notice for "just cause" would purport to contract out of that obligation.  Which, again, is void for illegality.

None of this is complicated or controversial - Ontario law figured out the significance of that difference a long time ago.  But where the question gets tougher is when we start to ask about whether or not the illegality of that clause impacts the enforceability of any other clause in the contract...and namely, the not-for-cause termination clause.

The challenging question here is this:  Where the 'for-cause' termination language is illegal, what impact does that have on 'not-for-cause' termination clauses?

And the answer, according to the Ontario Court of Appeal, is that the impact is to void the whole thing.

The Decisions

This is a surprisingly clean appeal:  The employer acknowledged that the 'for cause' termination language is illegal; the employee acknowledged that the language of the 'not-for-cause' termination language, in isolation, was compliant with the statute.  The question was simple: Under these circumstances, can the employer rely upon that not-for-cause language?

And the structure of this contract was about as favourable to the employer's circumstances as possible, which is what makes this decision so definitive.  The two different provisions were separate from each other, and there was a severability clause saying that the voiding of one clause didn't affect the rest of the contract.  (I've seen speculation for years that, depending on how its drafted, a problematic 'for cause' dismissal provision might bring down an otherwise-valid 'not-for-cause' provision...but this drafting is pretty much the high watermark for severability.)

Justice Morgan was the motions judge.  (Justice Morgan is a great judge, by the way. I may have critiqued some of his substantive results from time to time, but he's authored some fantastic judgments.)  He originally ruled in favour of the employer, finding that the illegality of the 'just cause' language didn't impact the enforceability of the other clause.

While acknowledging the public policy principles supporting holding employers to a high standard in ensuring that their contracts comply with the statute, and also endorsing the logic that formulaic deficiency in a clause cannot be remedied after the fact, Justice Morgan distinguished this case on the basis that the clause the employer sought to rely upon was not the same one that the employee attacked as void.

The Defendant does not seek to invoke or to in some way remediate the invalid Termination with Cause clause. Rather, it invokes, and seeks to enforce, the Termination of Employment for Notice clause on the very terms on which it was agreed. That clause was valid when the Agreement was entered into and remains valid upon the Plaintiff’s termination.

The Court of Appeal disagreed.

An employment agreement must be interpreted as a whole and not on a piecemeal basis. The correct analytical approach is to determine whether the termination provisions in an employment agreement read as a whole violate the ESA. Recognizing the power imbalance between employees and employers, as well as the remedial protections offered by the ESA, courts should focus on whether the employer has, in restricting an employee’s common law rights on termination, violated the employee’s ESA rights.

(Note: I've seen some question on the Twitterverse about the extent of that first sentence, and whether it extends to saying that a deficiency in some other part of the contract, like vacation entitlements or restrictive covenants, will void the termination language.  I don't take the court as necessarily saying that, nor do I think their reasoning particularly supports such an interpretation.)

Analysis

It's a tough question, fundamentally.  On a theoretical level, it boils down to whether or not the severability clause can coherently operate in such a way as to excise one termination clause without materially changing the meaning of the other terms of the contract (or undermining the public policy goals driving the illegality in the first place...there's some controversy about the test for severability).  And I think the Court of Appeal got this one right, for two central reasons:

1. Severing unenforceable termination language from enforceable termination language can easily create awkward results where the nature and effect of the enforceable language is dramatically altered by the omission; and
2. There's a broader trend in employment law jurisprudence against relieving against an employer's contractual overreach by allowing them to enforce the contract to the maximum degree enforceable.  The Supreme Court of Canada has been consistent on this type of issue for decades, from Machtinger to Shafron.

All due respect, and then some, to Justice Morgan, but I would suggest that he failed to fully appreciate the consequences of the plaintiff's argument "that the impugned provision of an employment agreement must be analyzed as at the moment of contracting, not at the time of breach."  He agreed with the principle, but not with the application on the facts.

The problem is that, viewed from the moment of contracting, excising the 'for cause' termination language, on its own, has some awkward and clearly unintended consequences.

Termination provisions, taken together and properly drafted, should create a coherent and comprehensive set of "If-then" statements.  "If a, then b.  In any other case, then c."  The intention of these different items is that the parties expect the various consequences to be ascribed to their respective causes, and nothing else.  To excise the first category leaves you with two interpretive options: Either you say that 'c' follows from 'a', despite the fact that the parties clearly did not intend that, or you break the comprehensive scheme and leave the contract silent as to one contingency, but not the other.

To be a little less abstract about it, the question in scenarios like this would be simple:  If the misconduct-driven termination language is void, and the 'not-for-cause' termination language is intact, what actually happens in the event of a misconduct-driven termination?

If we theorize a contract that, on its face, only specifies a not-for-cause mechanism, without a for-cause mechanism, challenging questions arise.  To some extent, it may depend on the framing, but there's certainly an argument - particularly in the context of an employment contract, typically drafted by the employer - that there's a resulting ambiguity as to whether or not the decision to provide only one mechanism for termination limits an employer to using only that mechanism.  That, after all, is why employers put 'for cause' language into the contract in the first place.  The ambiguity would tend to be interpreted to the advantage of the employee.

That result of excising 'for cause' language, however, has results obviously not intended by the parties.  Especially if the not-for-cause language is in any way generous, giving someone a golden parachute for his own misconduct seems obviously problematic.

If you were to ignore the ambiguity, however, and suggest that you're left with a scenario where the written terms of the contract govern termination not-for-cause, but implied terms govern termination for cause, that runs into a more cerebral problem, and particularly the problem the Court of Appeal alluded to when saying that employment contracts aren't to be interpreted on a piecemeal basis.

Fundamentally, the fact that there happens to be a common law default that more-or-less corresponds with the content of the void clause...is essentially coincidental, and irrelevant to the question of severability.

Under the common law default, an employer is entitled to terminate summarily for cause, or on reasonable notice without cause.  That, too, is a comprehensive structure.  But when you implement an express written termination clause, you are creating a different framework of obligations and entitlements, no matter how similar (or not) it may look to the implied default language.

In a case where the different contingencies don't cleanly line up with the common law categories, you'd end up superimposing a common law cause/not-for-cause analysis on one contingency within the agreement, while continuing to displace it for the other.

Consider the scenario of a contract that defines 'cause', with a list of different types of misconduct that the employer regards as being sufficiently serious to warrant summary dismissal.  This language is often used to bolster a just cause claim, saying that it was really clear to all parties from the outset just how serious this kind of conduct was...but it is not difficult to imagine a scenario where a court might interpret such language as overstepping, creating a right to summarily dismiss in circumstances where the applicable tests don't objectively call for it.

So suppose this hypothetical contract has language that limits not-for-cause entitlements to substantially less than common law reasonable notice, but tries to create a right to summarily dismiss an employee on the first instance of dishonesty, no matter how trivial.  If you voided this language, but left the not-for-cause language intact, you're looking at a question of what happens when an employee engages in dishonesty short of common law just cause, and gets dismissed as a result of that dishonesty.

Do we just treat that as falling within the intact entitlement-limiting language of the contract?  That's a problem for two reasons - theoretical, and practical.  The theoretical problem is that you're effectively extending the contractual not-for-cause category wider than it was intended by the parties, essentially rewriting the contract between the parties.  The practical problem is that it creates no disincentive for employers to draft contracts with overreaching 'for cause' language, if the courts are effectively just going to read it down to its maximum effect anyways.  This allows employers to establish a "banquet of Damocles" type environment - to invoke an image once used by the late Justice Echlin - where employees may be convinced that they're subject to summary dismissal for the slightest misstep.

This is reminiscent of the Supreme Court's reluctance in Shafron to 'read down' restrictive covenants, because doing so "provides no inducement to an employer to ensure the reasonableness of the covenant and inappropriately increases the risk that the employee will be forced to abide by an unreasonable covenant", which in turn invoked an 'incentive management' approach pioneered in Machtinger.  This is essentially the #2 reason I noted above, and goes to the principles set out in the larger body of employment law:  Recognizing the natural imbalance of power between employers and employees, and that contractual language is most often set by the employer, the courts don't want to create scenarios where employers can insert unreasonably onerous obligations in the contract on the assumption that the 'worst case scenario' is a court reading it down to the most significant obligation they could have contracted for.

So then, is the alternative to say that an employee dismissed for reasons within the voided category, but that don't constitute just cause, gets common law notice?

The law has tended away from that kind of result in the past, in scenarios that are similar but not the same:  Employees have argued, unsuccessfully, that an employer who alleges cause unsuccessfully should not be able to rely on contractual language limiting their obligations.  That would be a punitive framework, and moreover would ignore the fact that, absent cause, the contract delineates the employee's entitlements.  However, this is different, because it means that the factual underpinnings of the termination actually put the circumstances into a gap in the framework. The not-for-cause language was not expected or intended by the parties to be applied in circumstances where the employee engaged in dishonesty, so in a case where an employee is actually dismissed for provable dishonesty (that falls short of just cause at common law), the implication would seem to be that you can't apply the contractual 'not-for-cause' language to it.  This leads to plainly absurd results, with the dishonest employee having greater entitlements than the honest employee.

No matter which way you cut it, by leaving one contingency intact, and excising the other, you are fundamentally changing the operation, function, and impact of the intact contingency within the whole context of the employment contract.

This is over and above the usual mischief of severability: If you and I negotiate a contract, we are presumed to have concluded that the terms were acceptably fair to each of us.  Severing a provision will often have the impact of changing the entire balance of rights and obligations within the agreement, in a way that makes it something that one of us wouldn't have agreed to, on the whole.  Courts don't like to rewrite contracts.

That mischief is less obvious when we're talking about two unrelated terms:  If we agree to a mandatory arbitration clause, and to a fixed amount of notice of termination, does the unenforceability of the former mean that we shouldn't be held to the latter?  Sure, you can imagine a scenario where an employee might have said, "Okay, you can have your mandatory arbitration clause, but I want a concession on this other clause for it", but that seems relatively remote and unlikely.  And if there's an express severance clause, there's really no problem with separating them out.  But when you're talking about severing half of a termination framework from the other half of a termination framework, you're really talking about separating the two faces of a coin.

Other Viewpoints

Ontario lawyer Sean Bawden recently argued that Waksdale significantly diverges from the first principles of contract law, in that we're no longer trying to give effect to the intention of the parties. It's a worthwhile read, for a different perspective on this issue.

And with the greatest respect to Sean, I disagree with his interpretation of what's happening here, for a few reasons.

Firstly, his argument that the first principles of contract law require us to strive to maximally give effect to the intention of the parties...well, it echoes of the employer's position in Machtinger, and it fails for the same reasons: Employment law is and should be, in most respects, a subdiscipline of contract law, but it is expressly modified by statute.

All employment standards legislation across the country has a mechanism that expressly voids contractual provisions that attempt to circumvent the minimum standards therein. In Machtinger, the employer had implemented contracts with clearly deficient not-for-cause termination language, which entitled the employee to less than the minimum standard...but the employer nonetheless paid the minimum standard, and argued that the clear contractual intention to displace common law notice entitlements remained, so the better remedy was to read the minimum standard into the contract instead. The Supreme Court of Canada rejected this approach: If the courts just read down the impact of the language but tried to preserve the 'intention', then it would tacitly encourage employers to implement contracts offering less than the employment standards minimums.

Dealing with employees who generally don't understand their legal rights, this likely promotes exploitation of workers in a manner inconsistent with the language and intention of employment standards legislation.

Similarly, Waksdale is as much about the scope and application of the statutory voiding mechanism as it is about 'severance' more generally: Not just whether we can fix the contract by excising certain language and leaving the rest intact, but whether the statutory voiding mechanism operates narrowly enough to permit such severance at all. This is why the 'incentive management' approach - which is probably the most compelling policy argument underlying the Waksdale decision - makes sense; not because courts are trying to police the formation of contracts, but because they're trying to interpret legislation that expressly does exactly that.

Moreover, when we're talking about severability in general, we are, by definition, in an area where the courts have found some basis to not uphold the terms of the contract as entered into. By their very nature, severability questions necessarily entail divergence from the general law of contract.

It may (or may not) be that we're seeing an application of the doctrine of severability that's inconsistent with how it is applied in other contexts. If that's the case, I'm not sure it's as big a problem as Sean suggests.

Employment Law Versus Contract Law

As surely as employment law has always been a niche area of contract law, there have always been tensions between general principles of contracts and special treatments in employment contexts.

In general terms, trying to streamline employment law with the broader law of contract tends to be a step forward in terms of the consistency and comprehensibility of the law. However, that's not always just about changing the special treatments in employment law; sometimes, it's about changing the general principles of contracts.

At a truly fundamental level, the general law of contracts historically presupposed voluntariness of contracts and general equality of bargaining power. There were rare and exceptional treatments for things like duress or unconscionability, but it's an area of law deeply premised on the ability of either party to say "No".

In the employment context, these presuppositions have always been dubious, because of the core role of the employment contract in the lives and livelihoods of the employees, and the very substantial resulting imbalance in bargaining power. Often, where there's an unexplained and unusual treatment of employment contracts, what we're really doing is recognizing a gap or a problem in the broader canon of contract law.

Wallace damages are a good example of this: Historically, contract law generally eschewed non-economic damages, and in particular damages for mental distress, dating back to a 1909 decision of the English House of Lords, Addis. In 1997, Wallace created an exception to that doctrine, in recognition of the special kind of relationship that exists between employers and employees, permitting employees to seek a 'bump up' for bad faith employer conduct. From its inception, it was well-criticized (by, among others, Justice McLachlin) as arbitrary and doctrinally unsound...but it filled a gap.

In 2006, in Fidler v. Sun Life, the McLachlin court finally did away with Addis, allowing a simpler, older, and even more fundamental doctrine of damages (Hadley v. Baxendale) to govern mental distress damages; they're still not available in every case, but there's a non-arbitrary test for determining whether they're appropriate. Wallace damages were eliminated shortly thereafter in Keays, requiring employees to meet the consistent principled standard set in Fidler.

And so evolves contract law into the 21st century, as the current Supreme Court seems set on reinventing contract law to relieve against strict application of contractual terms, particularly as against those of reduced bargaining power. Many of the changes we've seen within the last year will need some time to mature, but concepts of contractual good faith obligations and recognition of bargaining power differentials have been realities for employment lawyers for decades, and I would suggest that we're now seeing contract principles start to 'catch up'. It's simply not unusual for employment law to exist at the cutting edge of contract law - and for that cutting edge to be rough.

So if Waksdale creates a break with broader principles of severability, then I would suggest that it's entirely possible that the problem lies in those broader principles (which, frankly, are not all that well established).

Implications

This will be a big deal in Ontario.  Unless this case (or a similar one) gets appealed to the Supreme Court of Canada, or a subsequent decision by the ONCA overturns the precedent, this case represents the law in Ontario, firmly binding upon lower courts.  And given how their ES termination exemptions are structured, a lot of contracts will get caught here.

Outside, it will be of less import.  Whether or not Alberta follows the precedent at all...I'd probably wager against it.  (Interestingly, the ONCA case was heard in writing only, apparently with no opportunity for oral submissions.  That may impact its persuasiveness outside of Ontario.)  But because our employment standards exemption uses the language of 'just cause', the typical framing of contracts allowing for termination without notice in the event of 'just cause' is probably still going to be okay.

But the question-mark arises when employers proceed to delineate cause.  As noted above, it is not unusual for employers to insert such language in a way that highlights that certain types of misconduct are considered very serious by the employer.  Zero-tolerance policies are theoretically problematic, because just cause is an inherently contextual question, but they sometimes 'wag the dog' into a finding of just cause by showing that the parties were aware of the seriousness of the conduct from the moment they entered into the contract.  So, historically, an employer might say in the contract "You can be summarily dismissed for cause if you whistle a C-sharp while in the workplace", knowing full well that the first instance of the C-sharp would never be upheld as cause, but on the communicative theory that, following multiple instances with progressive discipline, a judge would be more likely to accept that the employee understood its seriousness.

However, if Waksdale is followed in Alberta, then you might see the impact of such a clause become much more adverse to the employer, with courts not only disregarding the 'for cause' termination language altogether, but also voiding any not-for-cause termination language present.

Employers may have to find new ways of impressing the seriousness of certain types of conduct upon their employees.  If you have a bona fide zero tolerance policy (say, for safety violations), then make sure it's reflected in an actual written policy, and make sure employees have reviewed it and have ongoing access to it.  It's probably also prudent to post reminders around the workplace about such an important policy.

The alternative would be to change the structure of what you're defining in the employment agreement:  If you want to keep the old model of 'showing the serious stuff' in the termination provisions of the employment agreement, then maybe you do it without saying that these are automatically just cause:  Rather, you structure language in a way that defines just cause expressly with reference to the Code, but still highlighting that certain types of misconduct are very serious in determining whether or not just cause exists.  (Employers can contact me for help with their contracts. I have such contractual language, but hey, I don't give everything away for free.)

Other Food For Thought: Simpler Contract Drafting

Some Ontario lawyers use a structure along these lines:

If we dismiss you without cause, you get x.If we dismiss you for cause, you get only the minimum to which you are entitled by statute, if anything.

There's an appeal to this language, as it integrates the range of statutory treatments for the for-cause language, including the notice exemption for wilful misconduct:  If the 'cause' satisfies the statutory exemption test, the employee gets nothing; otherwise, the employee gets his or her statutory minimums only.

This language is arguably less useful when you're only giving the statutory minimums on a without-cause termination anyways.  In theory, you might be able to get away with one simple termination clause that guarantees the statutory minimums, if applicable - it builds a lot of content into that 'if' statement, including exemptions for the first three months of employment, misconduct, frustration of contract, refusal of reasonable alternative work, etc.  (I argued, in a parenthetical remark on probationary periods last year, that probationary language creates an unnecessarily high hurdle for employers to clear, as opposed to simply using ES minimum language for employees with less than three months of service.  In theory, this language would fix that.)  Elegant in its simplicity, really.

In practice, however, that may be too much content to build into those words, at least without explanatory or clarifying language, to ensure that its integration of the exemption categories is understood by employees:  Firstly, the inability of an employee to understand the impact of a term tends to push courts toward relieving against that impact. This will be particularly true in jurisdictions like Ontario where the exemption categories aren't even in the ESA itself, but rather buried in a regulation.  Secondly, such language loses the moral impact of having express contractual penalties for misconduct.


*****

Dennis Buchanan is a lawyer practicing labour and employment law and civil litigation in Edmonton, Alberta.

This post does not contain legal advice, but only general legal information.  It does not create a solicitor-client relationship with any readers.  If you have a legal issue or potential issue, please consult a lawyer.