The Proper Approach To Making Policy Decisions In Workplace Investigations

We recently reviewed an arbitration decision by Mark Hart. There’s enough to say about this decision that we have divided this into two editions of From Case to Practice. This Part 1 focuses on making policy decisions. Investigators are typically asked to make factual findings (“What happened?”) in their workplace investigations. However, most are also asked to decide whether, based on the facts they have found, some applicable policy was breached. This case provides helpful insight for those making these types of policy decisions.

Where we are referring throughout to “policy”, this can equally be a reference to any language against which factual findings are to be measured. This could include a collective agreement, like in this case, legislation (like the Ontario Occupational Health and Safety Act (“OHSA”)), a Code of Conduct, or any other document that governs the conduct of the parties.

The Case¹

Two employees working for a union (the Canadian Office and Professional Employees Union) alleged that the Executive Director (the “Respondent”) bullied and psychologically harassed them. The employees’ own union (The International Association of Machinists and Aerospace Workers, Local Lodge 1922) filed a grievance on their behalf in December of 2023. Between March and December of 2024, an external investigator conducted an investigation (which included 38 interviews) and issued reports in January and March of 2025 which substantiated a number of the allegations.

Unusually, the parties in this case agreed with the external investigator’s findings. This meant that the only thing Arbitrator Hart was left to decide was whether the facts as found breached the applicable Collective Agreement and/or OHSA. The Collective Agreement in this case had the type of anti-harassment language that is often contained in the policies investigators are asked to consider. In this Part 1, we review how the Arbitrator decided that the Collective Agreement had been breached².

How to decide when a policy has been breached in a workplace investigation?

1. Determine What Happened (Factual Findings)
   Establish the facts before evaluating compliance
   
   This case begins with a reminder that, before a decision can be made about whether a policy, a statute or a collective agreement has been breached, it must be first be decided whether the matters which are alleged actually occurred. These are frequently referred to as “factual findings” – and the first job of an investigator is to determine the facts: did this happen or not? As noted, in this case the investigator’s findings were adopted and so the Arbitrator was asked to consider whether the facts as found breached the Collective Agreement and OHSA.
2. Establish the Relevant Definitions and Elements
   Break down the policy or statutory language
   
   Arbitrator Hart next reviewed the specific Collective Agreement language to be considered. In the Collective Agreement, the definition of “bullying/psychological harassment” contained the following elements:
   1. that the behaviour at issue was vexatious;
   2. that the person who engaged in the behaviour at issue must have known, or ought reasonably to have known, that it was unwelcome; and
   3. one or more of the following:
      1. that the behaviour adversely affected the employee’s dignity, or psychological or physical integrity,
      2. that the behaviour took the form of repeated conduct which could reasonably be regarded as intending to intimidate, offend, degrade or humiliate, or
      3. that the behaviour resulted in a harmful work environment.
3. Match the Behaviour to the Definitions/Elements
   Apply the established facts to the applicable definition language
   
   Arbitrator Hart carefully considered the investigator’s findings (which included that the Respondent spread rumours about an affair between two people at work, referred to a board member as “crazy, “drunk all the time” and “mentally unstable,” and accused someone at work of stealing) and found that the Respondent’s behaviour was vexatious. “Vexatious” was not defined in the Collective Agreement and so Arbitrator Hart used a definition from a legal decision in which it was held to mean “words or actions…[that] cause annoyance, frustration and/or worry to an employee…”.The Arbitrator went on to consider whether the Respondent knew, or ought to have known whether their behaviour was unwelcome. Given the Respondent’s position as Executive Director and in light of the specific comments made, the Arbitrator again had no trouble concluding that the Respondent ought reasonably to have known that her behaviour was unwelcome.
   
   Finally, Arbitrator Hart considered whether the Respondent’s behaviour adversely affected the grievors’ dignity and psychological integrity. He reviewed evidence from the grievors about how the Respondent’s behaviour affected them. Because the grievors had provided this evidence, this was sufficient to satisfy all of the elements of the definition. However, the Arbitrator also concluded that the behaviour in question was repeated, and “could reasonably be regarded as intending to intimidate, offend, degrade or humiliate” and that this “resulted in a harmful work environment.” In doing so, Arbitrator Hart showed that there were multiple ways to view the Collective Agreement as having been breached.
4. Evaluate Any Necessary Context
   Consider power dynamics, patterns, and overarching themes
   
   Depending on the circumstances, context can be important when determining whether a policy has been breached. For example, in this case when considering one of the grievances in particular, the Arbitrator decided to consider the found behaviour as an overall pattern, as opposed to on a piecemeal basis, allegation by allegation. This is discussed further below.

Does the conduct in question need to be specifically referenced in a policy?

Many policies include lists of examples of specific types of behaviours that would generally be deemed to be breaches of the policy. Arbitrator Hart’s analysis provides two important insights about these lists.

First, it is not enough that a type of behaviour is listed as an example of a breach in a policy; it still has to satisfy the definition of the prohibited behaviour. In this case, “spreading malicious rumours or gossip” was listed as an example of a behaviour that could constitute “bullying/psychological harassment.” Arbitrator Hart did not simply find that the policy was breached because gossip was included in a list of examples. As noted above, he first reviewed the elements of the definition of “bullying/psychological harassment” and showed how the Respondent’s specific behaviour satisfied those elements. Then he noted that in addition that same behaviour was included in the examples.

Second, by approaching the analysis in this way, Arbitrator Hart noted that this would have allowed him to decide that the Collective Agreement had been breached even if the behaviour in question had not been listed as one of the examples. Lists of examples like this are generally not intended to be exhaustive.

Should separate findings be analyzed individually or as a whole?

If there is only one finding, investigators will obviously have to consider whether that is sufficient to breach a policy. In cases where there are multiple findings, it is possible that each type of behaviour might breach the policy on its own – as might be the case if the individual behaviours are all independently serious. However, if the multiple findings are all less serious behaviour or if they can collectively be tied to an overarching theme, then it may be that they breach the policy as a whole, instead of individually.

This was the case here where Arbitrator Hart found that the findings relating to one of the grievors (Grievor 2) were all “part and parcel of the same power dynamic in the workplace” and consequently an abuse of power in breach of the Collective Agreement. The same elements of the definition were considered, just as described above, except that in this case they were considered in relation to the pattern of behaviour instead of the individual findings.

What does this mean for the practice…

• Making factual findings and determining whether a policy has been breached are two distinct elements of an investigation.
  One of the pitfalls for investigators is when they blend factual findings and policy analysis in their investigations, and proceed throughout to be making one decision: whether a policy was breached. This can result in shortcuts and missed steps. This case serves as a solid reminder of the proper approach to take in a case where the investigator has two distinct tasks: making factual findings and determining whether a policy has been breached.
• When a policy lists certain behaviours as examples of a breach of policy, these are merely meant to be illustrative and not restrictive.
• Where there are multiple findings, investigators may need to assess them collectively to determine whether a policy has been breached. In certain cases, the breach may be due to a pattern of behaviour rather than one specific act.

In Part 2, we will examine how Arbitrator Hart applied these findings to the statutory threshold for workplace harassment under the OHSA.

Footnotes

1 International Association of Machinists and Aerospace Workers, Local Lodge 1922 v Canadian Office and Professional Employees Union, 2026 CanLII 50092 (ON LA).

2 We will examine Arbitrator Hart’s review of OHSA in Part 2.