Corps de l’article


The principle of proportionality is designed to limit abuse of power and infringement of human rights and freedoms by governments and other public officials to the minimum necessary in the circumstances. As a philosophical notion, proportionality may be traced back to the ancient Golden Rule of “that which is hateful to you, do not do to your fellow.”[1] As a legal principle, it originated in the nineteenth century in Prussian administrative law, in which it imposed constraints on police powers that infringed an individual’s liberty or property.[2] Throughout the years, the principle of proportionality expanded and migrated to other European countries,[3] where it is now a central and binding public law principle,[4] and to other jurisdictions, including Canada, New Zealand, Australia, South Africa, Hong Kong, India, and countries in South America.[5] Furthermore, it has become part of many constitutional and international documents.[6] It is also relevant in other contexts, such as international law (e.g., the doctrine of just war, the laws of self-defence, and international human rights law)[7] and criminal law (e.g., punishment should be proportional to the offence).[8]

The principle of proportionality was first recognized in Canadian constitutional law in R. v. Oakes,[9] in which the Supreme Court of Canada interpreted section 1 of the Canadian Charter of Rights and Freedoms,[10] which allows the government to limit constitutional rights and freedoms to a reasonable extent,[11] as entailing a proportionality test. Similar to other jurisdictions,[12] the Court established a three-stage proportionality test that examines the relationship between the measure adopted by the government to achieve a legitimate objective and the legitimate objective itself. First, the measure adopted by the government must be rationally connected to the justifiable objective it aims to achieve. Second, the government must select the measure that is the least harmful to, or minimally impairing of, the right or freedom in question, but similarly achieves the objective. Third, there must be proportionality stricto senso between the harms caused by the measure and the benefits of achieving the important objective—“[t]he more severe the deleterious effects of a measure, the more important the objective must be.”[13]

In a neo-liberal capitalist era, employers often exert as much control over an individual’s life as governments do. Should the application of the principle of proportionality extend to the private sphere and impose limitations on employers’ actions? The question is not about constitutional cases; the constitutional analysis undoubtedly involves a proportionality analysis in labour and employment contexts, as in any other context. The question here is rather about non-constitutional cases, involving private sector employers: Can (and should) we demand that such employers conform to the requirements of proportionality when making decisions affecting employees? Can (and should) we place similar constraints on labour unions making decisions that affect employers and the public at large? A number of scholars have recently explored this possibility in other jurisdictions and advocated the use of proportionality in some labour and employment contexts.[14] The three-stage test appears to offer a useful structure for discretionary decision making, ensuring that decisions are both rational and considerate, and preventing abuse of power by both employers and unions.[15]

Geoffrey England has examined the impact of the Charter on employment contract law, including the application of proportionality in “just cause” cases.[16] But a complete account of the role that proportionality plays or should play in Canadian employment and labour law has not yet been offered. In this article, we wish to advance two main arguments: First, a survey of employment and labour decisions by courts and other adjudicators in Canada reveals that the principle of proportionality is already being used in certain contexts. Sometimes the application is explicit, even if incomplete (i.e., does not closely follow all three stages of the Oakes proportionality test). But more often, the application is implicit. That is, courts and other adjudicators analyze different situations using tests akin to the Oakes proportionality test without an explicit reference to proportionality. Second, we argue that this trend is normatively justified and that a more explicit and structured use of the proportionality test should be advanced in various employment and labour spheres.

The article proceeds as follows: Part I exposes the contexts in which proportionality is currently used in Canadian employment and labour law decisions. We argue at a descriptive level that proportionality already plays a major role—although often not explicitly—in Canadian labour and employment law. Part II turns to the normative level and explores the justifications for extending the application of proportionality to the private sphere, and more specifically to the employment relationship. First, we explain why a higher standard of behaviour is required in employment relationships as opposed to other contracts. Second, we defend the use of proportionality in these contexts, stressing its legal and analytical merits. Third, we demonstrate that the application of proportionality fits within contemporary legal doctrine and advances legal coherence. We therefore advocate a more explicit use and structured application of the three-stage proportionality test in the contexts mentioned above. Part III proposes additional applications of proportionality in the labour context, showing how this principle may provide a more balanced approach to the resolution of contemporary labour relations conflicts in Canada, limiting the use of excessive power by both employers and trade unions.

I. Proportionality in Canadian Employment and Labour Contexts

A. Introduction

The principle of proportionality is used both explicitly and implicitly in various employment and labour law decisions. In some cases, the Charter, including section 1 and the principle of proportionality, is directly relevant in an employment setting. At times, a governmental action or piece of legislation infringes the rights and freedoms of employees, trade unions, or employers guaranteed under the Charter.[17] Setting aside these constitutional cases, there are also scenarios in which a private dispute arising between an employer and an employee, or an employer and a trade union, is analyzed within a proportionality framework. Sometimes the court or the relevant adjudicator will make concrete reference to proportionality, but may not follow all three stages within the Oakes proportionality test. Occasionally, the legal analysis will not make explicit reference to proportionality, but will significantly resemble the three-stage test. In most cases the burden of proof is dictated by the legislation or common law, but in other cases it is an open question how to devise the legal rule in this respect. This Part will canvass several representative employment and labour law decisions to demonstrate this argument.

B. Explicit Use

1. Disciplinary Procedure and Just Cause

The most obvious example of an explicit use of proportionality in the employment sphere is found in “just cause” cases. In response to its recognition of both the imbalance of bargaining power between employees and employers and the importance of work to the lives of individuals,[18] the Supreme Court of Canada developed in McKinley v. B.C. Tel the notion of proportionality in disciplinary procedures.[19] The Court held that employee misconduct, in and of itself, does not necessarily warrant just cause for summary dismissal. The principle of proportionality helps to assess whether, in the context and circumstances, an employee’s misconduct was so serious[20] that it should give rise to just cause for dismissal. That is, employers claiming just cause for dismissal are required to show that the sanction imposed upon an employee was proportional to his or her misconduct. Only if the misconduct was very serious (for example, “theft, misappropriation or serious fraud”) would an employer have a just cause to summarily dismiss the employee without an advance notice or pay in lieu of that notice.[21] In other cases involving less serious misconduct, an employer should use progressive discipline (i.e., “lesser sanctions for less serious types of misconduct”).[22] Only when the misconduct or poor performance repeats itself or continues despite discipline and clear warnings would it amount to just cause for summary dismissal.[23]

The test for establishing just cause, developed by the Supreme Court, was named a “proportionality” test,[24] perhaps building on the well-established test for disciplinary action in labour arbitration jurisprudence.[25] The test for just cause includes two stages: “(1) whether the evidence establishe[s] the employee’s [misconduct] on a balance of probabilities; and (2) if so, whether the nature and degree of the [misconduct] warranted dismissal.”[26] While no reference was made to section 1 of the Charter or to the Oakes test, a closer inspection of this test reveals some similarity to the Oakes proportionality test. One might argue that when employers make a decision to either discipline or dismiss an employee, the decision infringes the employee’s right or interest to have job security or at least to receive advance notice. Assuming that the objective of either disciplining or dismissing an employee is to ensure that the workplace is composed of the most competent and cooperative workers, employers are required to show that the measure chosen to achieve this objective was proportional.

The test developed in McKinley resembles the first two stages of the Oakes proportionality test, although a more structured analysis could have been beneficial. First, the McKinley test requires a proof of incompetence or misconduct. This is necessary, as disciplining or dismissing without notice employees who were engaged in misconduct or incompetence appears to be rationally related to the aforementioned objective because it deters—or, in the case of dismissals, conclusively prevents—future misconduct or incompetence from the same employee. By contrast, where an employee’s action was just an error in judgment, trivial or unintentional, discipline or dismissal without notice does not seem to advance the objective. Certainly, if the employee is wrongly accused or if the accusations are not substantiated—if there is no proof of the alleged misconduct—the disciplinary measure will not advance the stated objective, and therefore no rational relationship between measure and objective exists.

Second, the McKinley test examines whether a less severe response is possible while still achieving the aforementioned objective. Summary dismissal is a severe punishment. A less severe response, such as a warning, is usually sufficient to achieve the objective when the misconduct is not very serious.[27] However, when the employee’s actions are serious, intentional, or numerous, the employer may argue that there is no less intrusive way to achieve its legitimate business objective other than to dismiss the employee without notice.[28]

The McKinley test was further developed in subsequent cases and now contains elements of all three stages of the Oakes proportionality test, including a requirement to balance the benefits gained against harms caused by the chosen sanction. In Dowling v. Ontario,[29] for example, the Ontario Court of Appeal held that the test requires a consideration of the particular circumstances of both the employee and the employer:

In relation to the employee, one would consider factors such as age, employment history, seniority, role and responsibilities. In relation to the employer, one would consider such things as the type of business or activity in which the employer is engaged, any relevant employer policies or practices, the employee’s position within the organization, and the degree of trust reposed in the employee.[30]

In the context of balancing harms against benefits, this contextual evidence is needed to assess the severity of the harm to the employee versus the importance of the objective to the employer in the specific circumstances—as in the third stage of the Oakes proportionality test.

The same analysis applies not only in dismissal cases, but also in disciplinary cases. In Haddock v. Thrifty Foods,[31] for example, an employee had been working for sixteen years for the same chain of grocery stores. His last position was as a seafood department manager. He had been a good employee for most of the period but, in the later years, had some personal problems that led to alcohol abuse. In response to changes in his workplace behaviour, he was warned twice in 2002 and 2003 and then, about a year later, he was demoted to a non-managerial position with a 16–20 per cent decrease in income. The Supreme Court of British Columbia held that demotion was not the proper response to his poor performance but rather amounted to constructive dismissal. The court also held that a further warning was needed before the employer could terminate without notice, due to the time that had passed since the previous warnings and also the employee’s efforts to rehabilitate himself and improve his performance during that period. The additional warning requirement echoes the second stage of the Oakes proportionality test, in that the employer should have chosen a less severe measure (i.e., a warning) to achieve its legitimate objective of having the most competent body of employees in the organization.

In unionized settings, use of the principle of proportionality in discipline and dismissal cases is even more established. Collective agreements generally require employers to establish just cause prior to the imposition of any form of discipline (i.e., oral and written warning, suspension, discharge, etc.).[32] Furthermore, legislation provides arbitrators with the power to substitute their authority for that of the employer and to reduce the penalty imposed by an employer to one that is “just and reasonable” in the circumstances.[33] Interestingly, when attempting to give concrete meaning to these vague concepts, arbitrators appear to use the proportionality test.

Arbitrators consider two main questions in just cause cases. First, they consider whether the conduct in question amounts to just cause for the imposition of some form of discipline. As noted, this part resembles the rational connection stage of the Oakes test: dismissing or disciplining only those employees who misbehaved or performed poorly is rationally connected to the employer’s objective of having the most competent body of employees. Second, arbitrators consider whether the method of discipline selected by the employer is appropriate in the circumstances. Various mitigating factors have been identified as potentially justifying the substitution of a lesser penalty in the place of discharge, including: whether the employee was confused or mistaken as to whether an act was permitted, whether the act was impulsive (i.e., non-premeditated), whether the harm to the employer was trivial, whether the employee sincerely acknowledged the misconduct, the past record of the employee, the length of service, and whether the penalty imposes severe hardship upon the employee given his or her age and personal circumstances.[34] This part combines both the second and third stages of the Oakes proportionality test. It requires the employer to choose the least intrusive punishment while still achieving its objective. It also balances between the benefits of achieving the employer’s objective and the harms imposed upon the employee.

2. Privacy in the Workplace

Another explicit use of the proportionality principle can be demonstrated in invasion of privacy cases—for example, when an employer requires his or her employees to pass a drug or alcohol test, or uses surveillance cameras, or monitors emails and computer use. The clearest examples are those that fall within the jurisdiction of the federal Personal Information and Protection of Electronic Documents Act (PIPEDA).[35] Section 5(3) of the PIPEDA stipulates that “[a]n organization may collect, use or disclose personal information only for purposes that a reasonable person would consider are appropriate in the circumstances.”[36] Section 5(3) of the PIPEDA was interpreted by the Privacy Commissioner, as well as arbitrators and federal courts, as including a proportionality test.

The Privacy Commissioner of Canada has set out a fourfold test for determining when personal information may be collected for purposes a reasonable person would find appropriate in the circumstances.[37] The Commissioner held that, when examining section 5(3), one has to consider the appropriateness of the organization’s purpose for collecting personal information, as well as the circumstances surrounding that purpose. Once the purpose is identified, in order to determine whether the collection, use, or disclosure was reasonable in the circumstances, one has to consider the following questions: “Is the measure demonstrably necessary to meet a specific need? Is it likely to be effective in meeting that need? Is the loss of privacy proportional to the benefit gained? Is there a less privacy-invasive way of achieving the same end?”[38]

This test, which has been upheld by the Federal Court[39] and is followed in many arbitration awards,[40] is very similar to the Oakes proportionality test. The first inquiry corresponds to the minimal impairment stage of the Oakes test because it examines whether the measure is necessary to meet the objectivethat is, whether there are less intrusive ways of achieving the same objective. The second inquiry is akin to the first stage of the Oakes proportionality test because it examines whether the measure chosen for the collection of information is effective in achieving the objective—that is, whether it is rationally connected to it. The third inquiry resembles the third stage of the Oakes proportionality test because it weighs the proportional benefits of collecting information against the harm to the employee’s privacy. Finally, the fourth inquiry, which asks whether the employer explored other less privacy-invasive ways of achieving the objective, is also similar to the minimal impairment stage of the Oakes test.

In Eastmond v. Canadian Pacific Railway,[41] for example, the video recording surveillance cameras installed in the work yard were held to be justified because the employer successfully demonstrated that it had used the least intrusive means available to accomplish a reasonable purpose. The Federal Court used the term “proportional”, although it did not refer specifically to the Oakes proportionality test. In reaching its decision, the court considered the above-mentioned questions.[42] The court found that the purpose of collecting information through video cameras was appropriate in the circumstances. The employer had successfully established a legitimate aim—taking preventative action motivated by numerous past incidents. The court mentioned the importance of these cameras for deterrence of theft and vandalism as well as for the increased security of individuals and goods. Furthermore, the court found the loss of privacy to be minimal. Collection of information was neither surreptitious (there were warning signs) nor continuous. It was not limited to employees only, but captured every person who walked in the yard. It did not measure work performance. The recorded images were kept under lock and key and were accessed only when an incident was reported. Otherwise, they were destroyed. Moreover, the employer explored other alternatives (such as fencing and security guards), but they were too expensive or unfeasible. Finally, the court found the loss of privacy proportional to the benefit gained from the collection of information.[43]

It is worth noting that these tests had been used prior to PIPEDA by arbitrators adjudicating privacy cases and balancing the interests of the parties involved. Indeed, in the CAW Canada case,[44] which deals with drug and alcohol testing in a unionized and federally regulated workplace prior to PIPEDA, proportionality is not mentioned explicitly, yet the arbitrator applied tests akin to the Oakes proportionality test and engaged in an analysis that required balancing “the interests of the employees in the privacy and integrity of their person with the legitimate business and safety concerns of the employer.”[45] In examining whether drug and alcohol testing violated the collective agreement, the arbitrator asked a series of questions. First, is there evidence of a drug or alcohol problem, or both, in the workplace and, therefore, a need for management’s policy (i.e., the “test of justification or adequate cause”)?[46] This parallels the first stage of the Oakes proportionality test. Second, has the employer considered the available alternatives and might the problem in the workplace be combatted in a less invasive way (i.e., the “test of reasonableness”)?[47] This is similar to the second stage of the Oakes proportionality test. The arbitrator held that the employer demonstrated the need for its policy because of the safety-sensitive nature of the national railway operations[48] and provided sufficient evidence to reasonably justify its substance abuse policy, including drug and alcohol testing.[49] Furthermore, the employer had explored other less intrusive alternatives to deal with the substance abuse problem.[50] However, the arbitrator took issue with some policy rules that did not meet these tests. For example, one rule stipulated that “a positive drug test is, of itself, grounds for discipline or discharge.”[51] As such, the rule did not distinguish between “a positive drug test, standing alone, and impairment while on duty.”[52] This rule was unreasonable because it made no “reference to any clearly demonstrated legitimate employer interest,”[53] and because there were less intrusive ways of achieving the goal of combating drug and alcohol use among employees in non-safety-sensitive jobs.[54] By contrast, it may be reasonable when employees in risk-sensitive positions are concerned.[55] The arbitrator held that for risk-sensitive employees, who “work in locations spread across Canada, often without supervision or with only partial supervision,”[56] the benefits of the rules for fitness assessment, discipline matters, and monitoring substance abuse in the workplace outweigh the cost of infringing the privacy rights of individuals, “whose expectations must conform to the risk-sensitive concerns of the industry in which they seek to hold employment.”[57] This last part clearly reflects the third stage of the Oakes proportionality test.

Note that the PIPEDA also applies to provincially regulated organizations and businesses that collect information in the course of “commercial activities”. However, it does not extend to all employee personal information that is collected and used by provincially regulated organizations and businesses, because this generally does not amount to “commercial activity”. Some provinces have passed specific legislation on privacy which covers employment, while others have not yet done so. Consequently, different jurisdictions and adjudicators use a variety of tests when it comes to employers’ potential invasion of the privacy of employees. However, there is an increasing recognition of the employee’s right to privacy in the workplace, and the principle of proportionality, which is well established in PIPEDA cases, has gradually penetrated into non-PIPEDA cases. As we will argue later, this evolving area of law would benefit from a more explicit, structured use of the proportionality test.

Alberta, for example, has adopted comparable legislation—the Personal Information Protection Act—in 2003.[58] Section 11 stipulates: “(1) An organization may collect personal information only for purposes that are reasonable; (2) Where an organization collects personal information, it may do so only to the extent that is reasonable for meeting the purposes for which the information is collected.”[59] In Parkland Regional Library,[60] an employer installed keystroke logging software to monitor the computer usage of one of its employees without his knowledge, following concerns about low productivity and suspicions of his inappropriate use for personal purposes. When the employee found out about the software, he filed a complaint with the Privacy Commissioner of Alberta. The commissioner held that the collection of personal information did not comply with the legislation. There was no legitimate reason for monitoring the employee, as there was no sufficient evidence to support the employer’s suspicions.[61] This can be seen as a lack of rational connection, although the commissioner did not refer to the first stage of the Oakes proportionality test. Moreover, the chosen method of collection was not necessary for managing the employee. While it provided a broad range of information about the employee, other computer-based methods might have assessed productivity more specifically.[62] That is, the chosen software was not the least intrusive way of collecting this information. The employer could have, for example, simply asked the employee to explain his apparently low productivity or used performance measures and reviews, which are widely accepted management tools.[63] Again, the commissioner was using the second stage of the Oakes proportionality test, minimal impairment, without referring to it explicitly.

Similarly, a Nova Scotia arbitrator held that the Regional Municipality of Halifax, which recorded and stored for one year all incoming calls at a call centre, violated provincial privacy legislation (i.e., the Municipal Government Act, which protects privacy of information collected or used by the municipality)[64] and the collective agreement with call centre employees which included a duty to act reasonably.[65] Although the municipality’s actions were for a legitimate business purpose (i.e., quality control, training, and dispute resolution), it was unnecessary in the circumstances, disproportionate to the invasion of the employees’ inherent privacy rights, and therefore unreasonable. Quality deficiencies had already improved through coaching and supervision. The arbitrator concluded that the invasion of privacy was “significantly out of proportion to any benefit, potential or actual, gained or to be gained, by the employer.”[66] Note that the arbitrator referred specifically to proportionality when balancing between the benefits of collecting information and the harms of invading the employees’ privacy: “Proportionality is a tool to assist in the assessment of whether justification has been made out. It calibrates the intrusion to the interest protected. The operating principle is that the more serious the intrusion, the heavier the burden will be, and vice versa.”[67]

In provinces where no such legislation exists, there is a distinction between unionized and non-unionized workplaces. In a unionized environment, employers are required to exercise their managerial rights and discretion reasonably.[68] In privacy cases, this reasonableness standard has evolved into a “balancing of interests” test: weighing the employer’s interest in running its business effectively and safely against the privacy interests of employees. Arbitrators often assess the reasonableness of the employer’s action or policy, the nature of the employer’s interests in advancing this action or policy, whether there are less intrusive means available to address these interests, and the impact of the employer’s action or policy on the employees.[69]

This reasonableness test embodies the first and second stages of the Oakes proportionality test, but some elements of the third stage may be identified too. One might argue, for example, that surveillance cameras placed in workplace washrooms are reasonably needed to prevent thefts in a workplace. That is, washroom cameras are the most effective way to prevent thefts, and there is no less intrusive way of achieving this purpose. However, most people will agree that this measure is still unreasonable, due to the severity of privacy infringement, which cannot be offset by the benefits of preventing thefts. That is, the reasonableness test sometimes entails a balancing act which is the third stage of the Oakes proportionality test.[70] As will be argued later, it would be useful to break down the reasonableness test, which is a vague standard overall, into the three more concrete stages of the Oakes proportionality test.[71]

Finally, in non-unionized workplaces, employers are generally allowed to collect and use information about their employees in the absence of specific legislation or common law rules. This has led courts to seek creative ways to remedy the situation of employees whose privacy was brutally invaded by their employers in some cases,[72] and to acknowledge employees’ reasonable expectation of privacy even where workplace policies allowing search and surveillance were in place.[73] Recently, a tort of invasion of privacy (called “intrusion upon seclusion”) was established in Ontario and may be used against employers who invade the privacy of their employees.[74] However, this is merely a partial solution because it covers only extreme cases of intentional action, thus allowing most employers to continue collecting and using information about their employees. Subjecting employers’ actions to the principle of proportionality, even in the absence of specific privacy legislation, may be an appropriate solution.

We have seen that proportionality is used explicitly in at least two labour and employment law contexts: just cause and privacy. However, while the term “proportionality” has been invoked by courts in these two contexts, this has not been consistent. In some cases the term has not been mentioned. Moreover, the three stages of the proportionality test, although they can often be found between the lines, are not usually applied separately and systematically.

C. Implicit Use

1. Introduction

In this Part, we describe restrictive covenants, workplace discrimination, picketing, and unfair labour practice cases, in which courts have developed legal tests that are very similar to the proportionality test yet lack any direct reference to proportionality.[75] The legal tests developed in some of these contexts are well-established and structured. One might then ask why using proportionality in an explicit manner will be beneficial in these contexts. Our answer is twofold: First, once it is demonstrated that the tests used in these contexts are, in fact, very similar to proportionality, our argument is that it could prove beneficial to start using all three stages of the test, a practice which, in some cases, would add additional relevant considerations into the analysis. Second, even if no change is made to the jurisprudence on this particular topic and the same tests prevail without referring to proportionality, by showing that courts are de facto using the proportionality tests, the argument we wish to advance is that proportionality tests are generally useful in various contexts of labour and employment law. They are already used in some contexts, and can be used in other contexts. In other words, we are taking a broad look at several contexts in labour and employment law, within which tests have developed in ways that appear unrelated to one another, and we show that, in fact, the tests are very similar in all of those contexts, and also very similar to the Oakes proportionality test. This observation is, in our opinion, useful as a general jurisprudential point: demonstrating the importance and prevalence of proportionality as a general principle of law, including in private law, even when it is not mentioned explicitly. This also supports the first argument that proportionality tests should be used in some additional contexts.

2. Restrictive Covenants

A prominent example of an implicit use of proportionality in the employment sphere is found in restrictive covenants cases. Generally, non-competition clauses in an employment contract are viewed as a restraint of trade and are presumed to be unenforceable, unless the employer shows that the non-competition clause is necessary to protect the employer’s legitimate proprietary or business interests, that the non-competition clause covers a reasonable length of time and geographic area, and that a non-solicitation clause would not suffice to protect the employer’s legitimate interests in the circumstances.[76] As the Supreme Court of Canada held:

A covenant in restraint of trade is enforceable only if it is reasonable between the parties and with reference to the public interest. ... [C]ompeting demands must be weighed. There is an important public interest in discouraging restraints on trade, and maintaining free and open competition unencumbered by the fetters of restrictive covenants. On the other hand, the courts have been disinclined to restrict the right to contract, particularly when that right has been exercised by knowledgeable persons of equal bargaining power. In assessing the opposing interests the word one finds repeated throughout the cases is the word “reasonable.” The test of reasonableness can be applied, however, only in the peculiar circumstances of the particular case.[77]

Reasonableness is a relatively vague legal concept. Proportionality, on the other hand—which could be seen as a concretization of reasonableness—provides more guidance through the three different stages. And, in fact, the test used by judges to assess restrictive covenants appears to be in line with proportionality. While proportionality is not mentioned explicitly in the prevalent legal analysis, two stages of the Oakes proportionality test can be clearly identified. The employer is required to identify a legitimate objective (i.e., a proprietary or business interest) and to explain why a non-competition clause is necessary to protect this objective. These requirements resemble the test of rational connection. Furthermore, the employer has to draft a reasonable restrictive covenant in terms of length of time and geographic area, and must use a non-solicitation clause (i.e., rather than a non-competition clause) when it is effective in fulfilling the legitimate objective. These requirements are very similar to the test of minimal impairment. When there are less intrusive ways of achieving a goal, the employer must choose these measures. As the Manitoba Court of Appeal held:

The onus of proving that a covenant is reasonable as between the parties falls upon the party relying on it, i.e., the plaintiffs in this case. The presumption is rebuttable by evidence showing that the covenant is reasonable in that it goes no further than is necessary to protect the legitimate rights of an employer, and does not unduly restrain the employee.[78]

3. Discrimination

Another test that turns out to be very similar to proportionality is the bona fide occupational requirement (BFOR) defence in workplace discrimination cases. In Meiorin,[79] the Supreme Court of Canada developed a three-stage test to determine whether an employer may use the BFOR defence after an employee or a job applicant has shown a prima facie case of discrimination. Proportionality was mentioned briefly when the Court explained why direct discrimination and adverse effect discrimination should both be subject to the same analysis, despite some semantic differences across provinces:

In both cases, whether the operative words are ‘reasonable alternative’ or ‘proportionality’ or ‘accommodation’, the inquiry is essentially the same: the employer must show that it could not have done anything else reasonable or practical to avoid the negative impact on the individual.[80]

By contrast, the test itself, which includes three limbs, does not refer to proportionality, but clearly includes elements of the Oakes proportionality test. To determine whether a prima facie discriminatory standard is a BFOR, an employer has to justify the impugned standard by establishing on a balance of probabilities that

the employer adopted the standard for a purpose rationally connected to the performance of the job; that the employer adopted the particular standard in an honest and good faith belief that it was necessary to the fulfilment of that legitimate work-related purpose; and that the standard is reasonably necessary to the accomplishment of that legitimate work-related purpose. To show that the standard is reasonably necessary, it must be demonstrated that it is impossible to accommodate individual employees sharing the characteristics of the claimant without imposing undue hardship upon the employer.[81]

The first limb of this test explicitly adopts a rational connection test (the first stage of the Oakes proportionality test). The employer’s justifiable objective is to assign jobs to the most competent employees, and for this purpose the employer develops workplace standards (i.e., the measures). The second limb, requiring honesty and good faith, can be understood as an additional check on the legitimacy of the employer’s purpose. If the employer acts in bad faith—in an attempt to achieve illegitimate goals—then it arguably fails the rational connection test. The third limb of the Meiorin test includes elements of all three stages of the Oakes proportionality test. It examines, first, whether the workplace standard is reasonably necessary to the accomplishment of the employer’s purpose. This requires a rational connection between the workplace standard and the employer’s purpose. Second, it also requires the employer to show necessity—that is, consideration of alternative and less intrusive ways of achieving the employer’s goals. Alternatives may include, for example, “various ways in which individual capabilities may be accommodated.”[82] As the Court explains, “there may be different ways to perform the job while still accomplishing the employer’s legitimate work-related purpose. ... The skills, capabilities and potential contributions of the individual claimant and others like him or her must be respected as much as possible.”[83] Finally, the third limb involves an act of balancing the interests and rights of the employer, and those of the employee and other workers, as part of the duty to accommodate and the undue hardship analysis. As the Court elucidates, this act of balancing includes factors such as “the financial cost of the possible method of accommodation, the relative interchangeability of the workforce and facilities, and the prospect of substantial interference with the rights of other employees.”[84]

The Court then lists a number of supporting questions that again reflect a very similar analysis to the Oakes proportionality test.[85] These questions—“Has the employer investigated alternative approaches that do not have a discriminatory effect, such as individual testing against a more individually sensitive standard?”; “If alternative standards were investigated and found to be capable of fulfilling the employer’s purpose, why were they not implemented?”; “Is there a way to do the job that is less discriminatory while still accomplishing the employer’s legitimate purpose?”—are all akin to the second stage of the Oakes proportionality test. The question, “Is it necessary to have all employees meet the single standard for the employer to accomplish its legitimate purpose or could standards reflective of group or individual differences and capabilities be established?” reflects the first stage of the Oakes proportionality test (or, alternatively, could be understood as referring to minimal impairment). The question, “Is the standard properly designed to ensure that the desired qualification is met without placing an undue burden on those to whom the standard applies?” resembles the third stage of the Oakes proportionality test.

Subsequent cases have followed the same line of analysis. In Entrop v. Imperial Oil,[86] for example, Imperial Oil adopted an employee alcohol and drug testing policy that included an automatic termination of employment sanction for positive tests. The issue in court was whether this policy was discriminatory on the basis of disability, which includes substance abuse. The legitimate purpose of minimizing risk of impairment due to substance abuse and ensuring a safe workplace, especially in safety-sensitive positions, was clearly identified. The Ontario Court of Appeal held that while drug testing was, in general, rationally connected to work performance, it could not measure present impairment of ability to perform work safely, only past drug use. Accordingly, the court held that this testing could not be justified as reasonably necessary to accomplish Imperial Oil’s legitimate goal, which in this case appears to suggest that the policy failed the rational connection test.[87] It was also held that the sanction for a positive test was too severe—“more stringent than needed for a safe workplace and not sufficiently sensitive to individual capabilities”—which appears to suggest that the policy also failed the minimal impairment test.[88]

4. Picketing

Another context we would like to discuss regarding implicit use of proportionality involves cases concerning picketing. Proportionality may be relevant to picketing in two different contexts. The first context is constitutional and examines whether picketing should be permitted or restricted by legislation or common law rules. The second context focuses on the relationship between the union and the employer and assesses whether the use of picketing is appropriate, which involves the application of the legislation or common law rule in the specific circumstances. The relevance of proportionality in the first, constitutional context is clear. Picketing is a form of expression and, as such, is protected under section 2(b) of the Charter. Imposing limitations on picketing may therefore be justifiable only in accordance with section 1 and the Oakes test. Our focus in this article is on the second context and its less obvious relevance to proportionality. When determining, in specific circumstances, whether to issue an injunction or not, courts examine a union’s action and require certain standards to be met (as sometimes imposed by an injunction order), in line with the principle of proportionality.

In the case of Pepsi-Cola,[89] the Supreme Court of Canada dealt with both contexts. First, it examined the constitutionality of banning or limiting secondary picketing at common law and held that secondary picketing should be allowed unless it involves a wrongful action (e.g., a crime or a tort) because picketing engages freedom of expression, which is constitutionally protected under section 2(b) of the Charter.[90] The Court held that the Charter is relevant despite the private nature of the matter because:

The Charter constitutionally enshrines essential values and principles widely recognized in Canada, and more generally, within Western democracies. Charter rights, based on a long process of historical and political development, constitute a fundamental element of the Canadian legal order upon the patriation of the Constitution. The Charter must thus be viewed as one of the guiding instruments in the development of Canadian law.[91]

Furthermore, since freedom of expression is not unlimited, being subject to reasonable limitation under section 1, the Court subjected the common law to section 1 values in the following: “Limitations are permitted, but only to the extent that this is shown to be reasonable and demonstrably necessary in a free and democratic society.”[92]

Next, the Court moved on to assess the dispute between the private parties—Pepsi-Cola and the trade union—and asked whether the use of secondary picketing was appropriate. The Court held that the protest outside the homes of Pepsi-Cola’s management personnel was tortious, and upheld the associated injunction order, but allowed the peaceful picketing outside retail outlets selling Pepsi-Cola products.[93] Although this “wrongful action approach” does not explicitly require a balancing act,[94] the Court recognized that courts and legislatures might have to provide supplementary guidelines:

Doubtless issues will arise around the elaboration of the relevant torts and the tailoring of remedies to focus narrowly on the illegal activity at issue. Doubtless too, circumstances will present themselves where it will become difficult to separate the expressive from the tortious activity. In dealing with these issues, the courts may be expected to develop the common law sensitively, with a view to maintaining an appropriate balance between the need to preserve third-party interests and prevent labour strife from spreading unduly, and the need to respect the Charter rights of picketers.[95]

In our view, maintaining an appropriate balance requires a proportionality analysis, which may be developed at common law. The Court recognized that although picketing may cause economic harm to employers and third parties, it is usually allowed because such economic harm is “anticipated by our labour relations system as a necessary cost of resolving industrial conflict.”[96] However, the “most problematic” picketing, “whose value is clearly outweighed by the harm done to the neutral third party” or which causes irreparable harm to the employer, will either not be allowed or be subject to some restrictions.[97] Thus, these statements have opened the door for a proportionality analysis in subsequent cases.[98]

In some jurisdictions, injunction orders in labour disputes are regulated by statute. In Ontario, for example, an employer or a third party has to show that there are activities taking place that cause danger of damage to property, or danger of injury to people, or obstruction of or interference with lawful entry or exit from the property.[99] While the statute does not explicitly refer to proportionality, post-Pepsi-Cola cases have engaged in an analysis that closely resembles the Oakes proportionality test. As will be argued later, a more explicit use could better guide the reasoning by providing detailed structure, and thus, provide much more useful precedent for future cases.

In Unilux Boiler Corp. v. Fraser,[100] for example, when employees committed tortious acts and criminal misconduct in the course of their picketing, Unilux sought an injunction that would, among other things, limit the number of picketers. The Ontario Superior Court of Justice examined this request and held that it was “an unmerited infringement on the Union’s ability to provide support for the remaining strikers and to exert pressure on their employers.”[101] In other words, the union’s actions met the first stage of the Oakes proportionality test: the picketing was rationally related to the aim of exerting pressure upon the employer. However, the court was willing to issue an order restraining the union from preventing entrance or exit for any time longer than five minutes.[102] That is, the union’s actions failed to meet the second stage of the Oakes proportionality test. Apparently, the court felt that a five-minute delay at the entrance was sufficient to achieve the goals of picketing, including conveying information about the dispute and exerting social and economic pressure on the employer, and accordingly, a longer delay could not be justified. Alternatively, perhaps the court felt that the union’s actions did not meet the third stage of the Oakes proportionality test, because the costs to the employer outweighed the benefits of picketing when people were delayed at the entrance for more than five minutes. It is not clear from the judgment which of the two tests was applied; the court only emphasized the fact that some of the union’s actions were unlawful.[103] But some form of a proportionality test is obviously required if injunctions are issued only against delays at the entrance that are longer than five minutes. Why five and not more or less? Either a minimal impairment test or the third stage of the proportionality test is necessary to justify such a conclusion.

In Ogden Entertainment,[104] striking workers at the Corel Centre in the City of Kanata, where NHL games were played, had set up large picket lines on nights with scheduled hockey games. They impeded the access of passenger vehicles, public transit, commercial vehicles, team buses, and more. Traffic jams resulted, causing traffic on the highway to back up for many miles. The picketers did not distribute leaflets or try to communicate with the occupants of any vehicles. The Ontario Court of Justice held that the picketing amounted to a criminal offence and nuisance, and stressed that the only thing the picketers achieved was the obstruction of vehicles.[105] It also stated that there might be a need for special rules to apply in cases that involve large numbers of people who are not party to the labour dispute.[106] The court issued an injunction restraining the picketers from interfering, blocking, or delaying any person or vehicle from entering or exiting the Centre.[107]

This is another example of how courts resort de facto to a proportionality analysis, and again, a more structured analysis in line with the three-stage Oakes proportionality test could have been beneficial. The court maintained that the picketers did not convey information to the public and achieved nothing other than the obstruction of vehicles. This appears to mean that their actions failed the first stage of the proportionality test: no rational connection to a legitimate goal. However, a better articulation of the union’s goal—an articulation that recognizes the need to exert economic pressure on the employer—may have led to a different conclusion. Arguably, the analysis would have benefited from a discussion of the second stage: did the union have other alternatives that were less harmful to the employer but also achieved its legitimate goals? The court did hold that the picketers committed the tort of nuisance, which is clearly harmful to the employer, but it did not consider whether other, less harmful ways to achieve its legitimate goals were actually available to the union. Moreover, as part of the “balance of convenience” examination that courts employ to consider petitions for injunctions, the court weighed the employees’ interest in obstructing traffic against the employer’s right to enjoy lawful entrance to and exit from its premises by its tenants, other employees, and members of the public.[108] Not surprisingly, it concluded in favour of the employer.[109] An explicit resort to the third stage of the proportionality test could have led to a better articulation of the rights and interests involved. Employees obviously do not have a right to obstruct traffic per se, but they have a right to exert pressure on the employer—or at least a legitimate interest in doing so—as a way to secure better work conditions. The court should have considered not only the damage to the employer and to the public but also the importance of the actions for the picketers themselves.

In Ledcor,[110] the workplace was under substantial renovations and, as a result of picketing at the entrance that included delays of vehicles, construction had to be shut down. The Ontario Superior Court of Justice allowed the picketing, but to ensure that construction workers were let in, the court limited the maximum number of picketers to twenty. The picketers were further prohibited from obstructing or blocking entrances to or exits from the site. This result resembles the minimal impairment test.[111] The court, in effect, concluded that there were less intrusive ways to achieve the objective and since the union had not chosen them, the court had to impose some limitations.

In Industrial Hardwood,[112] the strikers set up a picket line and blocked the entrance of vans carrying replacement workers. The delay was up to an hour and included harassment of replacement workers and damage to the vans. The Ontario Superior Court of Justice issued an order that prohibited picketers from preventing vehicular access to the workplace and also prohibited all picketing at the plant, except for the purpose of communicating information to those wishing to receive it, and only for a maximum of five minutes. The order also limited the number of picketers to four at each plant entrance. The Court of Appeal upheld the order except for the limitation on the number of picketers.[113]

This analysis again echoes the three-stage Oakes proportionality test. The Court of Appeal stated the union’s legitimate objective as follows:

[Picketing] provides striking workers with the collective opportunity to seek to persuade others of the rightness of their cause. It allows them to express through collective action their solidarity in pursuit of that cause. And it provides an important outlet for collective energy in what is often a charged atmosphere.[114]

Based on this starting point, the first stage of the Oakes proportionality test was not met: blocking the entrance and harassing replacement workers is not effective in achieving this objective. Providing replacement workers with information, on the other hand, as the order suggests, is rationally related to the objective. The second stage of the Oakes proportionality test was not met: there were less intrusive means to achieve the objective. As the Court of Appeal stated, “the delays were in the range of half an hour, considerably longer than reasonably necessary for the picketers to effectively communicate their position to the occupants of the vans.”[115] The order therefore limited the delay to five minutes. The third stage of the Oakes proportionality test was not met: the harms caused by the picketing outweighed its benefits. The case did not involve any property damage or personal injury, but the court considered the degree and duration of obstruction to entry or exit to be very substantial.[116] To be sure, the results may have been different had the court considered the goals of exerting pressure on the employer or preventing strikebreaking as well. Even if the three stages of the proportionality test are applied explicitly and separately, there is always room for discretion (and disagreement) on how to apply them. But such an analysis can help in pointing attention to all relevant considerations and in creating a more structured decision-making process.[117]

5. Unfair Labour Practice

A final example of an implicit use of proportionality revolves around unfair labour practice cases. Although the requirement of intentional interference does not appear in certain (often central) provisions regulating employers’ interference with trade unions, labour relations boards and courts have insisted on finding that employers were intentionally involved in unfair labour practice before holding them liable.[118] A more balanced approach might be to prohibit any sort of anti-union action by employers subject to the principle of proportionality. That is, employers often exercise their managerial prerogative to advance various actions in the workplace. These actions might interfere with trade unions. When an employer shows that: a) behind its action, there was a legitimate objective that is rationally connected to the action; b) the action was the least intrusive one to achieve the legitimate objective; and c) the harm to employees’ rights and interests is not disproportionate to the benefits of achieving that objective, the action would be allowed. In contrast, anti-union actions that are not proportional would be considered unfair, and thus illegal, regardless of the employer’s intention.

A few cases have followed this proportionality analysis, though implicitly. The leading example is CBC v. Canada (Labour Relations Board),[119] in which the union filed an unfair labour practice complaint, claiming that CBC had interfered with its actions when it forced the union president, Goldhawk, to choose between his job as host of a radio program and his role as union president. This move by CBC followed publication of an article written by Goldhawk which the network thought was in violation of its journalistic policy. The complaint was upheld by the Canada Labour Relations Board and the Supreme Court of Canada. While CBC’s actions were not intentionally anti-union, it was found liable because the unfair labour practice provision in the federal code, as well as in other Canadian jurisdictions, is not limited to intentional actions but broadly prohibits any interference.[120] As Brian Langille and Patrick Macklem describe it, it was clear that CBC had interfered with employee representation, yet

[t]he issue, just as it is in human rights and constitutional analysis, is of possible justification. This requires ... a balancing or proportionality analysis. And this is what the board and the Court did, with the result that the CBC was not able to justify its decision by reference to a compelling business justification of its action.[121]

In The Society of Energy Professionals v. Hydro One,[122] the Board explained which factors come into play when assessing an employer’s conduct to determine whether it constitutes an unfair labour practice. The board employed elements of proportionality, and especially elements of the third stage of the Oakes proportionality test, which measures the severity of the interference for the trade union against the benefits of achieving the legitimate business goal, when it looked for “more than incidental interference with the trade union” and examined whether there was an “imbalance of interests in favour of the protected activity,” “whether the conduct threaten[ed] the formation or very existence of a trade union,” and “whether the employer conduct [was] classic business activity, such as a bona fide exercise of a managerial prerogative, such as a layoff, or subcontracting decision.”[123] An explicit resort to all three stages of the proportionality test could be more beneficial. The test would ask whether the action that may interfere with the trade union would achieve the legitimate business goal, thus testing rational connection. It would also ask whether the action is necessary, or whether there are other, less intrusive ways to achieve the legitimate business goal.

II. Justifications for Applying Proportionality in Labour and Employment Law

A. Introduction

In the previous Part, we have shown that proportionality tests are already an important feature of Canadian labour and employment law. In the current Part, we turn from the descriptive to the normative. To justify the growing practice of resort to proportionality tests—and to suggest that this practice should be expanded and become more explicit and structured—we proceed in three steps. First, in Subsection B, below, we argue that it is justified to place a high standard of behaviour on employers vis-à-vis their employees—higher than the standard demanded in other contracts. It is similarly justified to demand a higher standard from unions when they are exercising powers that have the potential to harm employers and the public at large. Then, in Subsection C, we argue that the proportionality tests are an appropriate choice to guide such a higher standard, as they are more concrete than other (vague) standards, provide clear guidance, and generally refrain from intervening in the choice of goals, thus offering balanced solutions.[124] In this context, it will be shown that proportionality is already used by other legal systems and has proven useful to solving labour law questions. Finally, in Subsection D, we discuss the doctrinal issues. We argue that applying these proportionality tests is within the discretion of the courts in the development of the common law and, in some cases, when interpreting legislation. We also argue that applying these tests will have the added advantage of improved coherence within the legal system.

B. A Higher Standard of Behaviour is Normatively Justified

A market economy is based, to a large extent, on self-interest. People are allowed to act to advance their own interests. Indeed, they are expected to do so, and contract laws assume that a meeting of (self-)interests will lead to an agreement that is beneficial to both parties, and indirectly, to society at large. The law, therefore, generally supports such agreements without requiring individual actors to consider either the interests of others with whom they contract or other societal interests. There are exceptions, as we shall see shortly, but this is the default rule.

The government, on the other hand, is expected to uphold a higher standard. Government officials making a decision obviously think first and foremost about the government’s interests, and so they should. But they also have to consider the implications for others; if a decision harms someone, officials have to take this into account. The leading benchmark used in recent years to examine governmental decisions is proportionality. In Canada, as in many other countries, society expects the government to act in accordance with the standard of proportionality, meaning that the decision has to pass the three stages of the test mentioned above. Why does the law demand a higher standard of behaviour from the government, as compared with the standard required in dealings between private actors? One answer could be the fact that the government acts as our “long arm”, in that government officials represent us and make decisions on our behalf. It is only natural that we demand that they do so with a degree of respect for our interests—that they, at the very least, take them into consideration.

There is, however, another justification that is just as valid. Governments have power, and power should be used responsibly. In various contexts, the law is designed to prevent the abuse of power by those who hold it; public law can be seen as an example of this general legal rule. But private actors hold power as well. Corporations have significant powers, which the law limits in various ways—for example, with competition or anti-trust laws, or consumer laws. In these regulated areas, private actors can no longer act freely to promote their self-interest. Rather, the law creates limitations to ensure that the interests of other parties are considered—that is, that harms to others are minimized.[125] Employment standards are, in effect, another example of this general rule. They are based on the understanding that employment relationships are characterized by a power imbalance. Employers sometimes abuse their superior powers, and employment laws are designed to prevent that—for example, by setting a minimum wage. If the interests of employees are sufficiently considered, the wage cannot fall below a certain minimum.

In collective relations, both parties have powers, and the same ideas apply. Here it is not only the employer, but also the union that is expected to use its power responsibly. The law recognizes the right to strike (or picket, etc.) and protects striking employees; in effect, unions have been given a legal power to take collective industrial actions. At the same time, it is justified to demand that decisions concerning such actions, which create significant harms to both employers and the public at large, measure up to a high standard of behaviour, to ensure that the power is not abused.

C. Proportionality is an Appropriate Choice of a Higher Standard of Behaviour

It should be fairly easy to accept the argument in the previous section. After all, labour and employment laws already create various limitations on the exercise of power by employers and by unions. And, without getting into the details of these laws, their basic existence is uncontroversial. The question is, however: why should we add an additional limitation (proportionality) to the ones that are already detailed in legislation? Here, we offer two separate answers.

The first is that adopting the principle of proportionality does not necessarily create additional limitations. As we have shown in the previous Part, in many cases the current law—whether in the form of legislation or common law—already places limits on the use of power by employers or unions. Thus, for example, employers are legally entitled to dismiss employees, but this power has various limitations. Unions are legally entitled to organize picketing, but again, this power is not without limitations. Judges are left with broad room for discretion when applying these laws; thus, the three proportionality tests can be a useful aid. In other words, in many cases, proportionality would simply structure the analysis. While employee rights have been granted considerable weight in the case law, applying proportionality would strike an appropriate balance between the rights of employees and rights and interests of employers. As England argues, the use of proportionality is essential in ensuring that employee rights are not “advanced at the expense of unduly impairing employers’ economic efficiency, from which everyone ultimately benefits.”[126]

The second answer refers to situations in which proportionality would indeed create new limitations on employers or unions. We argue that this too is justified, at least in some contexts, and can be achieved by judicial development of the common law. We discuss the doctrinal viability of this proposal in the next Subsection. Here, we wish to justify the choice of instrument: Why proportionality and not some other standard?

The employment relationship is dynamic. Demands from an employee change over time. Mutual expectations evolve, and so do workplace norms and rules. New managers and co-workers replace old ones. Power can be used, and abused, in different and unexpected ways. Some of these ways are addressed by specific regulations, but regulations can never cover the entire range of possibilities. It is therefore useful to leave some degree of discretion for courts to prevent the abuse of power in unforeseen situations. This is accomplished through various open-ended standards. Indeed, legislatures have established, and adjudicators have also developed, a de facto requirement of fairness in some employment contexts.[127] And, as we have seen, employers are sometimes required by common law to measure up to a reasonableness standard.[128] Canadian courts have recognized implied contractual duties to treat employees with civility, decency, respect, and dignity,[129] and to exercise discretion reasonably, or at least honestly and in good faith, when discretion may adversely affect employees’ interests.[130] In other legal systems, a requirement of good faith in employment relations is increasingly gaining ground.[131]

The advantage of these open-ended standards is their ability to address new problems in an ever-changing landscape. There is, obviously, a price in terms of indeterminacy and vagueness.[132] To enable workers to know their rights and employers to know their obligations, we need concrete rules. To some extent, courts can develop such rules over the years by implementing the open-ended concepts, but such rules are always incomplete. We believe that the principle of proportionality offers a balance: it is open-ended and yet includes relatively concrete rules—the three-part proportionality test. Admittedly it does not offer clear-cut solutions for any given case. Yet the three-stage structure offers a principled way to analyze the problem and promises a degree of determinacy and predictability higher than what can be found in open-ended standards.[133]

Proportionality also offers a balance in terms of respecting the rights and interests of both parties. The default rule is that the employer is free to make any managerial decision, so the principle of proportionality does not generally intervene in business judgments and choices. The exception is that society insists on a degree of respect for the rights and interests of employees. Employers are not expected to completely internalize the costs of their decisions on employees. They are, however, expected to refrain from choosing means that do not advance their own goals, means that harm the employees more than necessary to achieve these goals, and means that infringe the rights of employees in a way that inflicts harms disproportionate to the expected gains. In short, the proportionality test ensures that the harms to employees are minimized, while also minimizing any intervention in business decisions.

This does not mean that every decision by every employer and every union must be subject to a proportionality analysis. Some decisions are entirely prohibited, and should remain so—for example, dismissing an employee because of union activities. Other decisions are entirely within the employer’s discretion, and should remain so—for example, choosing the managers. Our focus here is on decisions that fall somewhere in between—that is, allowed in principle, but subject to limitations. We argue that the proportionality test is a useful and appropriate way to articulate such limitations and to structure their analysis. We believe that the use of this test is warranted and justified in the various contexts discussed in the previous Part. We also believe that the same test could be useful and justified in other labour and employment contexts. We give two examples in the next Part; additional contexts could be considered in future research.

There are two possible critiques of proportionality that should be considered here. First, a relatively open-ended standard could be difficult to enforce. One could argue that such a standard would be relevant as a matter of practice only for high-level employees, those with access to legal advice and resources. This could lead to more disparity and higher inequality between workers. However, if employers change their decision-making process to consider the impact on employees—as required by proportionality tests—lower-income employees can be expected to benefit as well. Moreover, we do not propose to replace other (more concrete) standards, only to add another layer. There is no reason to believe that a requirement to avoid unnecessary or excessive harms to employees would detract in any way from the rights of other employees.

A second possible critique is that applying the principle of proportionality, especially the third part of the test, requires adjudicators to engage in an act of balancing and weighing various considerations. This might be problematic, especially in a private sector context in which the decision makers (in the current context, employers or trade unions) are in the best position to engage in such an analysis; decision makers’ discretion should not be replaced by the adjudicators’ discretion. Intervening in managerial decisions infringes the autonomy of employers and could be detrimental to efficiency. Intervening in union decisions could be similarly detrimental to unions’ autonomy and ability to achieve their goals.

Our response is twofold. First, as the examples in the previous Part show, courts are already required to engage in balancing when applying the law. We simply suggest replacing existing standards, such as reasonableness, with the more structured tests of proportionality. Second, proportionality analysis involves very little intervention in the choice of goals, except for very extreme situations in which certain goals will be deemed illegitimate. Employers and unions will thus continue to have very broad discretion in choosing their goals. The requirement to choose means that are rationally related to that goal, and that will minimize the negative impact on others as much as possible, is hardly a cause for concern. Rationality and minimal respect for others are not ingredients in a recipe for inefficiency—quite the contrary.[134] The situation is a bit different with regard to the third branch of the proportionality test, requiring employers and unions to internalize, to some extent, the costs to others of their decisions. To limit the harms of this demand, we suggest that the level of scrutiny vary depending on the type of decision that is in question. For example, when fundamental rights are at stake (such as equality, privacy, or freedom of association), stricter scrutiny is more appropriate compared with the protection of other interests (such as one’s job, as in just cause cases). In the latter cases, the third stage of the proportionality test could be relaxed, allowing intervention only in extreme cases of disproportionality.

The many advantages of a proportionality test delineated above probably explain the ever-growing reliance on proportionality in the labour and employment laws of other countries. Most notably, proportionality is heavily used as a labour and employment law standard in Germany (verhältnismäßigkeitsprinzip). Interestingly, the strongest example is found in cases on the legality of strikes.[135] The principle was established in 1971, when the Federal Labour Court held that, due to their negative impact on participants as well as third parties and the general public, strikes and lockouts have to comply with the principle of proportionality.[136] To meet this requirement, industrial action must be suitable and necessary to achieve legal aims; must be proportional to those aims; must be used only after all other negotiations have failed (i.e., the last resort or ultima ratio principle); must not exceed what is necessary to achieve the aim; and also, both parties must contribute to restoring peace as extensively and as soon as possible after the industrial action is over.[137]

In the European Union, the principle of proportionality applies in various private spheres, including discrimination law. In October 2000, the EU adopted the Directive establishing a general framework for equal treatment in employment and occupation for all people, irrespective of a range of factors.[138] While direct and indirect forms of discrimination are prohibited,[139] article 2(2)(b) provides that indirect discrimination can be justified if it serves a legitimate aim and the means of achieving this aim are objectively necessary and proportionate.

Furthermore, the European Court of Justice (ECJ) subjects trade unions to the principle of proportionality. In the controversial Laval case,[140] it was held that the right to take collective action (i.e., the right to strike) was a fundamental one, but was subject to certain restrictions. Since it might have infringed the right to provide services, which is one of the fundamental freedoms guaranteed by the Treaty on the Functioning of the European Union, it had to be exercised proportionately.[141] This case and other similar cases[142] were criticized for their outcomes, which prioritize economic interests over social interests. Brian Bercusson, for example, argues against the use of proportionality in the context of strikes because strikes are linked to a collective bargaining process, and it is difficult to apply proportionality to unions’ demands, which change and evolve through a process of negotiation. Also, applying proportionality may negatively affect the impartiality of the state in economic conflicts.[143] However, while the outcome of these cases was controversial, it does not mean that the application of proportionality should be eliminated altogether. Several commentators have proposed different ways of applying the proportionality test in this context.[144] Moreover, the justification for using this standard as a limitation of strikes becomes stronger when employers also have to conform to the same standard.

In the UK, the more structured principle of proportionality has replaced or, some have suggested, should replace the standard of reasonableness in various employment contexts.[145] The principle of proportionality is well-established in discrimination law.[146] This development was influenced by the jurisprudence of the ECJ applying EU directives concerning equal treatment.[147] These directives have led to the amendment of existing measures and to the adoption of new measures prohibiting employment discrimination on various grounds.[148] They have also required the application of a proportionality test as part of the defence in indirect cases of discrimination.[149] That is, a neutral-on-its-face provision, criterion, or practice can be justified if the employer shows that it was a proportionate means of achieving a legitimate aim.[150] While most cases deal with employers who discriminated against their employees, trade unions are also subject to the same analysis.[151] However, English courts often apply a test integrating proportionality and reasonableness; a test that requires an objective balance between the discriminatory effects of the measure and the reasonable needs of the discriminator, but avoids subjecting employers to the stricter ECJ standard, which demands that indirect discrimination be necessary to meet a real need of the business.[152]

Another application of a proportionality test is possible in cases of unfair dismissal under the Human Rights Act 1998, under which courts evaluate the justification for dismissing an employee relative to the infringement of his or her rights guaranteed under the Act.[153] The Act incorporates the European Convention on Human Rights protections into UK law.[154] Although the Act focuses on the public sphere, applicable in the current context to public sector employees, a claim can be invoked against private employers and trade unions by various indirect means prescribed by the Act.[155] Baker argues that the Act provides a great opportunity to enhance the application of proportionality in British discrimination law cases.[156] Similarly, it has been argued that the application of proportionality in workplace privacy cases may reconcile employee privacy with employers’ interests.[157]

In France, there is a general rule grounded in the Labour Code that prohibits any infringement of workers’ rights that is not in line with the principle of proportionality.[158] In Israel, labour courts have been using proportionality tests de facto for many years, and more recently they have started referring to the principle explicitly.[159] It is reasonable to assume that this trend will expand into many more countries in the near future.[160]

D. Applying the Proportionality Test is Doctrinally Possible and Will Improve Coherence

So far, we have argued that it is justified to apply the proportionality tests in the labour and employment sphere and to private sector employers as well. But some might question whether this is possible as a matter of doctrine, given the current jurisprudence of the Supreme Court of Canada concerning the inapplicability of the Charter in private relations. The proportionality test was developed in Oakes as an aid for the application of the Charter’s section 1. It is therefore not surprising that any mention of proportionality tends to evoke the idea that the Charter is being applied. Nonetheless, our argument does not rely on any change in constitutional jurisprudence. It squares perfectly well with the current jurisprudence, because we are not advocating the direct application of the Charter in relations between individuals. We simply use the same legal tool—proportionality—as an aid in another context. For this reason, the use of proportionality is not necessarily limited to situations in which fundamental rights have been infringed. While most of the examples considered in this article implicate fundamental rights, the same kind of analysis that we argue is useful for deciding labour and employment law cases can also be useful to analyzing the impacts on other interests deemed to be justified of protection.

There are two separate doctrinal routes in which we have argued that proportionality is used (and should be used): interpretation of legislative provisions and the development of common law rules. In both cases, proportionality tests can be infused into current doctrines, and to some extent, already have been infused. It does not mean that employees have constitutional rights vis-à-vis the employer. It simply means that limitations on employers and unions are placed into the same structure of analysis—the three-stage proportionality test—that is used in constitutional law.

The common law route is perhaps more controversial. Brian Langille has argued against any kind of “balancing” when applying the common law; he maintains that people should be able to exercise their freedoms without limitations, even when such freedoms negatively and substantially affect others and their interests, unless one has a legal right that limits another’s freedom.[161] Such a right, he argues, arises from legislation, and can also arise from the common law. At the same time, he assumes that judges cannot develop the common law to create new rights. It is here that we respectfully disagree. Take restrictive covenants, for example. Judges developed rules to decide cases involving such covenants—and to place limits on the freedom of employers to use such covenants—by relying specifically on the concept of reasonableness. It would be odd to suggest that judges are not allowed to further develop these rules, in replacing the vague reasonableness test with the three-stage proportionality test.

As the Supreme Court held in Pepsi-Cola, the law should be developed in line with the values enshrined in the Charter.[162] The principle of proportionality has been a central part of Canadian jurisprudence, used as an aid to implement Charter values.[163] Applying the same tests in the labour and employment sphere has the added advantage of increasing coherence within the legal system. As David Beatty argues, “[e]xempting judge-made rules that regulate how people interact personally and privately in civil society from having to conform to the principle of proportionality is worse than incoherence.”[164] It is inconsistent with the hierarchical relationship between supreme and subordinate laws. Furthermore, while some argue that applying proportionality in private law might threaten individual autonomy and freedom, the principle of proportionality is, in fact, sensitive to these values, and without it, the threat to privacy and autonomy is greater.[165]

We have argued in Part II that applying the three proportionality tests in the labour and employment context is normatively justified. This analysis followed the detailed exposition of the various ways in which these tests are already being applied in Canadian labour and employment law, in Part I. Our conclusion is, therefore, that the development exposed in Part I is justified. However, as the examples have shown, the use of proportionality tests has so far been incomplete and inconsistent. We believe that an explicit reference to the three-stage proportionality test and a separate application of each stage will be highly beneficial. First, our recommendation will ensure that all the right questions are asked and that the examination is structured and principled. Second, this will make it easier for employers, employees, and unions to anticipate the results of litigation and understand the requirements demanded of them. Finally, an explicit resort to proportionality will add coherence to labour and employment law. As we have seen, courts invoke many different tests that are, in fact, very similar, and can be replaced with the same proportionality test.

III. Additional Applications

As described in Part I, there are many employment and labour contexts in which the principle of proportionality applies either explicitly or implicitly. This raises the question of whether there are additional labour and employment law contexts in which proportionality could be relevant and its implementation beneficial. We believe that the answer is affirmative, and offer one example in this Part: setting limits for strikes and lockouts. We hope that additional contexts will be explored in future research.

Canadian law stipulates some requirements prior to commencing a lawful strike or lockout. These are mostly technical requirements, such as a conciliation process, a “no board” report, and a strike vote.[166] Currently, there are almost no substantive restrictions on strikes or lockouts.[167] For example, a union may commence a lawful strike once it meets the technical statutory requirements even if the strike is unreasonable in the circumstances, or too destructive to the employer’s business or to third parties. One might argue that a strike should not be limited once it meets the technical statutory requirements, because the right to strike is fundamental and arguably protected by the Charter.[168] However, there are two main justifications for the imposition of some limitations.

First, even the most fundamental and constitutional rights are not absolute. There is always a need to balance competing rights and interests. Second, when a strike is believed to be unreasonable or destructive, it is currently limited by provincial and federal governments through back-to-work legislation[169] or specific legislation denying the right to strike in some workplaces.[170] These drastic actions against unions, which have become increasingly popular, often result in major infringements of freedom of association.[171] A more balanced approach could involve subjecting the act of strike or lockout to the principle of proportionality. Importantly, such a system is already in existence in several other jurisdictions.[172] Instead of broadening the scope of what is considered to be “essential services” and consequently eliminating the right to strike altogether ex ante, the principle of proportionality would ensure that unions may go on strike, but use strike actions appropriately in a way that balances the interests of all parties. The Labour Relations Board, when asked to issue a back-to-work order, would have the authority to determine, on a case-by-case basis, whether the strike was proportional or not, taking into account all relevant factors and circumstances.[173] Obviously, such a proceeding would have to be swift, but there is no reason to think that this would not be possible. The default is that a strike is allowed, yet the Labour Relations Board may decide to issue an interim order until the final decision, if this seems justified in the circumstances.

For example, instead of preventing all Toronto Transit Commission (TTC) workers, regardless of their job position, from striking at all times (no matter if it is at rush hour or not),[174] TTC workers would generally be allowed to strike. If the strike were too destructive, the TTC would be able to file a complaint with the Ontario Labour Relations Board, which would examine the particular circumstances of the case and determine whether the union used the strike weapon proportionately. Assuming the justifiable objective of the union is to reach a collective agreement, the board would examine, first, whether a strike is effective in achieving this purpose. It would then determine whether there are other means of achieving this objective. It might be, for example, that the conciliation failed because the union did not cooperate and rushed into a lawful strike position. It might be that the union’s decision to strike during rush hours for more than a day was too intrusive. In such a case, the board would be able to limit the nature and scope of the work stoppage. Finally, the board would consider whether the damages of the strike outweigh its benefits.

The same proportionality analysis should apply to lockouts. When a trade union commences a partial strike, an employer may impose a lockout for a legitimate purpose of protecting the business against inefficient operation. However, sometimes the use of a lockout might be disproportionate—for example, when an employer uses a lockout to force the government into passing back-to-work legislation. In June 2011, the Canadian Union of Postal Workers went on rotating strikes for two weeks. In response, Canada Post decided to impose a full lockout and blamed the strikers for losses of $100 million.[175] This lockout created pressure on the government, pressure that ended with the enactment of back-to-work legislation. One might argue that a more balanced approach would be to subject the right to lockout to the principle of proportionality. The employer would then have to show that it had used lockout for a legitimate purpose and that the lockout was rationally connected to this purpose, that the lockout was the least intrusive way of obtaining this purpose, and that the benefits of a lockout outweighed its damages.[176]

We realize that these proposals assume a degree of faith in the judicial system, including courts and labour boards. Judges and adjudicators will be given a greater role and broad discretion to assess the means chosen by labour unions and by employers in light of their legitimate goals. Admittedly, in other countries where proportionality is used in this context, there are independent labour court systems, sensitive to the unique features of employment relations. We believe, however, that because of the important role played by proportionality in Canadian constitutional law, Canadian judges and adjudicators are well positioned to perform this kind of analysis, based on the default rule that the right to strike should be respected, and limitations must be justified.


Proportionality is seen by enthusiastic proponents as the ultimate rule of law.[177] One does not have to go very far to appreciate the usefulness of this principle as a legal tool. A demand that those holding power will use it carefully and responsibly finds a concrete legal expression in the three-stage proportionality test. An expectation that those infringing the rights of others will not do so gratuitously also materializes in the proportionality test. Ever since the seminal Oakes judgment, proportionality has become an important pillar of Canadian law. One cannot think about constitutional law or discuss it—in Canada as in many other countries—without referring to proportionality. We have argued that the same principle plays an important role in Canadian labour and employment law as well, a role not sufficiently acknowledged thus far. We further argued that proportionality should play an even greater and more explicit role.

It is crucial to understand that by referring to proportionality, we do not settle for an abstract, vague concept. We rather refer to the three separate stages of the proportionality test developed in the Oakes judgment, following other legal systems. These stages allow one to consider the means chosen to achieve a given goal. There is minimal intervention in the choice of goals: in practice, a goal simply has to be legitimate. This is appropriate for constitutional law, under which the elected branches of government should be given as much freedom as possible to pursue the goals of their choice, and it is similarly appropriate for labour and employment law, under which employers should be allowed to set their own managerial goals with as little intervention as possible. The focus of the proportionality test is on the goal-means connection: first, a rational relation must exist between them; second, the impairment of rights (or, more generally, the harm to others) should be as minimal as possible; and finally, the harm caused by the action through the chosen means should not be disproportionate to the benefits gained by it.

We began the article by canvassing the different contexts in which proportionality is already used in federal and provincial labour and employment laws. In some cases, courts have inferred from legislation a requirement that employers act according to this test (or some of its parts). In other cases, courts have developed similar tests as part of the common law when considering problems without legislative solutions. In a few instances, we found explicit reference to proportionality; in other cases, we found that seemingly unrelated tests used by courts and adjudicators, in fact, closely resemble the three-stage proportionality test. In all of these contexts, we believe that a more explicit application of all three stages of the proportionality test will prove useful to the analysis and the decision-making process.

At least in some cases, demanding that employers and unions comply with the three parts of the proportionality requirement means that a higher standard of behaviour is imposed. Is this justified? We have argued that it is. We further showed that one does not need to apply the Charter to private relations in order to accept this conclusion. We then explored the possibilities for further development and proposed an additional context in which proportionality tests can be used and potentially offer better solutions than current laws. Generally speaking, we believe that the incorporation of proportionality into labour and employment law could be an important and useful development. Our findings are also of great relevance to the more general discussion about the applicability of the principle of proportionality beyond the boundaries of public law.