On the evening of March 2024, a 52-year-old woman entered a licensed dining establishment in Leduc, Alberta, was seated by staff, and approximately 40 minutes later was asked to leave the premises. What happened in between those 2 moments would become the central evidentiary question in a human rights complaint alleging discriminatory removal, and the documentation failures that surrounded that 40-minute window would expose the owner of the establishment to liability risk that better contemporaneous records could have substantially mitigated. The woman's complaint to the Alberta Human Rights Commission alleged that she was removed because of a protected ground, while the establishment's position was that removal followed from conduct that violated the premises policy adopted 18 months earlier. The gap between those 2 accounts would ordinarily be bridged by documentary evidence, but the records created on the night in question contained omissions severe enough to complicate the defense and trigger difficult questions under the establishment's $2 million in commercial general liability coverage. This lesson examines the specific evidentiary gaps that characterized the March 2024 incident and analyzes how those gaps translate into legal exposure when a respondent must answer a human rights complaint.
The foundational principle governing evidentiary gaps in discrimination defense is that the legal burden structure in human rights proceedings creates asymmetric consequences for missing documentation. When a complainant establishes a prima facie case of discrimination, the evidentiary burden shifts to the respondent to provide a credible, non-discriminatory explanation for the impugned conduct. The quality of that explanation depends almost entirely on the contemporaneous record, because tribunals and courts assessing credibility give substantial weight to documents created at or near the time of the incident and treat after-the-fact reconstructions with appropriate skepticism. A respondent who cannot point to specific, documented observations that supported the decision to remove a patron faces the problem of asking a tribunal to accept a narrative constructed months later, often in consultation with legal counsel, that cannot be corroborated by the very records the respondent's own policy required staff to create. The absence of required documentation does not merely leave a gap that the respondent can fill with oral testimony; it creates an adverse inference that the tribunal is entitled to draw, namely that the documentation was not created because the observations it would have recorded did not occur or did not support the action taken.
The Alberta Human Rights Act establishes that discrimination in the provision of goods, services, and facilities customarily available to the public is prohibited. A licensed dining establishment falls squarely within this category, and the removal of a patron engages this protection directly. The complainant in the March 2024 incident needed to establish 3 elements to make out a prima facie case: that she possessed a characteristic protected under the Act, that she experienced adverse treatment in the form of removal from the premises, and that the protected characteristic was a factor in that adverse treatment. The first 2 elements were not in dispute. The woman was 52 years old, and age is a protected ground. She was removed from the premises, which constitutes adverse treatment. The contested question was whether her age, or another protected ground she alleged was operative, was a factor in the decision to remove her. The establishment's defense rested on the assertion that removal followed from policy violations unconnected to any protected ground, but the evidentiary record available to support that assertion contained gaps that would prove consequential.
The first significant gap concerned the absence of a contemporaneous incident report documenting the specific conduct that allegedly violated the establishment's policy. The policy adopted 18 months earlier required staff to complete an incident form whenever a patron was asked to leave, recording the date, time, staff member involved, description of the observed conduct, the policy provision allegedly violated, and the patron's response. On the night of March 2024, no such form was completed before the establishment closed. The manager on duty testified in the subsequent proceedings that the evening was busy, that the removal occurred near closing time, and that she intended to complete the paperwork the following morning. When she did complete a report the next afternoon, approximately 18 hours after the incident, the document contained vague references to the patron being disruptive and making other guests uncomfortable, without specifying what the patron had said or done, which other guests were affected, or how the manager had determined that the conduct violated the policy rather than reflecting mere subjective dislike. The 18-hour delay and the lack of specificity would become central to the complainant's argument that the stated reason for removal was pretextual.
The second evidentiary gap involved the absence of corroborating witness statements from staff who were present during the alleged policy violation. The establishment employed 4 servers and 1 bartender on the evening in question, in addition to the manager who made the removal decision. None of these 5 employees were asked to provide written statements on the night of the incident or in the days immediately following. When the human rights complaint was filed 6 weeks later, counsel for the establishment began gathering statements, but by that point the employees' recollections had faded, their accounts contained inconsistencies with each other and with the manager's delayed report, and 1 of the servers had left employment and could not be located for 3 months. The statements eventually collected were dated between 8 and 14 weeks after the incident, and the complainant's counsel would argue that these reconstructed accounts were unreliable and potentially coordinated, a characterization that the absence of prompt, independent documentation made difficult to rebut.
The third gap concerned the establishment's surveillance system. The premises was equipped with 4 security cameras that recorded footage to a digital system with a 14-day overwrite cycle unless footage was affirmatively preserved. The policy adopted 18 months earlier included a provision requiring managers to preserve surveillance footage whenever an incident report was completed, precisely to create a visual record that could corroborate or contradict witness accounts. Because no incident report was completed on the night of the removal, no footage preservation was triggered. By the time the human rights complaint arrived 6 weeks later, the footage from March 2024 had been overwritten 4 times. The establishment could not produce any visual record of the events leading to the removal, the removal itself, or the complainant's demeanor and conduct in the 40 minutes between her arrival and departure. The complainant's position was that she had been calm, polite, and engaged in nothing more than ordinary conversation with her dining companion before being approached by the manager and asked to leave. The establishment's position was that she had been loud, had made comments that disturbed nearby tables, and had responded aggressively when initially asked to moderate her volume. Without footage, neither account could be independently verified, but the evidentiary consequences of that uncertainty fell disproportionately on the party with the burden to provide a credible non-discriminatory explanation.
The fourth gap involved the absence of any documented history of progressive intervention before removal. The establishment's conduct policy contemplated a graduated response to patron behavior: a verbal reminder, a documented warning, and removal only if the conduct persisted. The policy specifically provided that removal without prior warning was appropriate only in cases of threatening behavior, physical altercation, or intoxication affecting safety. The manager's delayed incident report did not allege any of these 3 circumstances. It alleged disruptive conduct of a type that, under the policy's own terms, should have been preceded by at least 1 documented warning. No warning was documented, and the manager's testimony that she had verbally asked the complainant to keep her voice down before deciding on removal was unsupported by any contemporaneous record and was denied by the complainant. The gap between what the policy required and what the documentation showed created an inference that the policy was not actually followed, which in turn supported the complainant's argument that the stated policy basis was pretextual and that the true reason for removal was discriminatory.
The cumulative effect of these 4 gaps was to leave the establishment in a defensive posture that depended almost entirely on the after-the-fact credibility of its witnesses rather than on a documentary record created when events were fresh and memories were reliable. Human rights tribunals in Alberta approach credibility determinations by examining internal consistency, consistency with other evidence, consistency with the inherent probabilities of the situation, and the demeanor of witnesses. The first 2 of these criteria were substantially undermined by the evidentiary gaps. The manager's account could not be checked against a contemporaneous document she herself had created, because that document did not exist until 18 hours later and contained insufficient detail to anchor her testimony. Her account could not be checked against prompt witness statements from her staff, because those statements did not exist until months later. Her account could not be checked against video footage, because that footage had been destroyed through the ordinary operation of a system that the policy contemplated preserving. Her account could not be checked against a documented warning, because no warning was documented. Each gap, considered individually, might have been explained as an understandable lapse in a busy service environment. Considered cumulatively, the gaps created a pattern that the complainant characterized as reflecting either a removal decision made on improper grounds that the establishment was now attempting to rationalize, or a standard of documentation so lax that the establishment could not credibly claim to enforce its policy in a consistent, non-discriminatory manner.
The documentation failures also affected the comparative evidence analysis that human rights tribunals undertake. When a respondent claims that adverse treatment was based on conduct rather than a protected ground, tribunals examine whether other individuals who engaged in similar conduct were treated similarly. If patrons outside the protected group were treated more leniently for equivalent conduct, this supports an inference of discrimination. The establishment's 18-month enforcement record, examined in the preceding lesson, showed that incident reports were completed for only a fraction of the removal decisions made during that period, and that the reports which did exist varied widely in specificity and completeness. This inconsistency made it difficult for the establishment to demonstrate that the complainant was treated the same as others who had engaged in similar conduct, because the documentary record of how others had been treated was itself fragmentary and unreliable. The complainant's counsel was able to argue that the evidentiary gaps were not random but rather correlated with the characteristics of the removed patrons, a serious allegation that a complete and consistent documentation practice would have made much easier to rebut.
The legal standard applicable to this comparative analysis requires the respondent to show not merely that it had a policy, but that the policy was applied consistently. Consistency cannot be established by assertion alone; it must be demonstrated through records showing that the same conduct triggered the same response regardless of who engaged in it. Where documentation is spotty, inconsistent, or missing, the respondent cannot meet this evidentiary requirement, and the tribunal is entitled to draw the inference that consistency was not achieved. The gaps in the March 2024 documentation thus implicated not only the specific incident but the establishment's entire enforcement posture over the preceding 18 months. The absence of a reliable contemporaneous record for the complainant's removal became a lens through which the tribunal could view the entire enforcement history as unreliable, amplifying the legal exposure beyond what a single incident, properly documented, would have generated.
The evidentiary gaps also created complications for the credibility of the manager who made the removal decision. In her testimony, she provided specific details about the complainant's conduct that were not contained in her delayed incident report. When asked why these details had not been recorded at the time, she explained that she had summarized rather than provided a complete account, and that she was now filling in the details from memory. This explanation created a credibility problem familiar to human rights practitioners: testimony that goes beyond contemporaneous documentation is treated as potentially reconstructed, self-serving, or influenced by knowledge of the complaint and the need to justify the decision. A witness who recorded specific observations at the time and testifies consistently with those observations is far more credible than a witness who recorded vague generalities and now supplies specifics that conveniently support the defense. The manager's credibility was further undermined by the absence of corroborating witnesses who could confirm her expanded account, since the other staff members' delayed statements did not contain the same level of detail and, in some cases, appeared inconsistent with particular aspects of her testimony.
The doctrine governing adverse inferences from missing evidence provided the complainant with a powerful argument. When a party who has control over evidence fails to produce it, and the evidence would be expected to favor that party if their account were true, the tribunal may infer that the evidence would not have supported their account. The establishment had control over whether an incident report was completed promptly, whether staff statements were gathered immediately, whether surveillance footage was preserved, and whether the policy's graduated response was documented. It failed in each respect. The complainant argued that these failures should give rise to an inference that the documentation, had it existed, would have shown that the complainant's conduct did not in fact violate the policy, that the removal was not in fact preceded by a warning, that the surveillance footage would not in fact have supported the manager's characterization of events, and that the staff would not in fact have corroborated the manager's account if asked while their memories were fresh and before they were aware of the legal proceedings. This argument was difficult to rebut precisely because the evidence that might have rebutted it did not exist.
The procedural posture of human rights complaints in Alberta compounded the significance of these evidentiary gaps. The Alberta Human Rights Commission has investigative powers, and the investigation stage often determines whether a complaint proceeds to a hearing. An investigator reviewing the complaint would have asked the establishment to produce its documentation, and the gaps in that documentation would have been apparent at an early stage. The absence of a prompt incident report, the absence of witness statements, the absence of preserved footage, and the absence of documented warnings created a situation where the investigator had limited material with which to assess the establishment's account, while the complainant's account, though also lacking independent documentation, was at least internally consistent and did not depend on documents that should have existed but did not. The evidentiary asymmetry was thus baked into the proceeding from its earliest stage, and the establishment's ability to resolve the complaint through investigation was compromised by its inability to produce a documentary record that might have persuaded the investigator that the removal was properly grounded in policy violations.
The relationship between evidentiary gaps and settlement dynamics deserves attention. A respondent who possesses strong documentation is in a position to take a complaint to hearing with confidence, knowing that the tribunal will have before it a contemporaneous record supporting the non-discriminatory explanation. A respondent whose documentation is weak or missing faces a different calculus. The risk of an adverse finding increases, and with it the risk of remedies including general damages for injury to dignity, feelings, and self-respect, which in Alberta human rights matters can reach significant amounts depending on the severity and duration of the impact. The establishment in the March 2024 incident would have been advised that the evidentiary gaps substantially increased the probability of an adverse finding and that settlement at an earlier stage, while requiring payment, might be economically preferable to proceeding with a weak record. This is a common consequence of documentation failures: they increase not only the probability of liability but the settlement value of claims, because a complainant's counsel recognizes that the respondent's ability to defend is compromised.
The evidentiary gaps also implicated the establishment's obligations to its insurer under the $2 million commercial general liability policy. General liability policies typically require the insured to cooperate fully with the insurer's investigation and defense, to preserve evidence that may be relevant to covered claims, and to notify the insurer promptly when circumstances suggest a claim may arise. The failure to preserve surveillance footage and to create prompt documentation could be characterized as a breach of the insured's duty to preserve evidence, particularly if the policy contained explicit evidence preservation provisions or if the insurer could demonstrate that the lost evidence would have been material to the coverage determination or defense. The insurer in this matter would have been concerned that the evidentiary gaps were not merely unhelpful to the defense but were created by the insured's own failures to follow its stated policy, and this concern would inform the coverage analysis examined in the following lesson.
The nature of the evidentiary gaps in the March 2024 incident illustrates a broader principle applicable to all premises liability and human rights defense matters: the time to create evidence is when the event occurs, not when the claim arrives. A human rights complaint may not be filed for weeks or months after the incident, and during that interval, memories fade, employees leave, footage is overwritten, and the opportunity to create a reliable contemporaneous record is permanently lost. The establishment's conduct policy recognized this principle by requiring prompt incident reports, immediate witness statements, and automatic footage preservation. The failure lay not in the policy's design but in its execution on the specific evening in question. A policy that exists on paper but is not followed when the moment arises provides no better protection than no policy at all, and in some respects provides worse protection, because it creates an expectation of documentation that, when unmet, supports an inference that the conduct justifying the adverse action did not occur as claimed.
The complainant's legal theory drew strength from each evidentiary gap in a manner that a complete record would have foreclosed. She alleged that the true reason for her removal was a protected characteristic, and that the stated policy basis was pretextual. A pretext argument succeeds when the respondent's explanation is shown to be unworthy of belief, and evidentiary gaps are precisely the conditions under which explanations become difficult to credit. If the establishment had produced a detailed incident report completed within minutes of the removal, signed by the manager, describing specific statements made by the complainant, specific tables whose occupants complained, and specific policy provisions violated, the pretext argument would have faced a substantial obstacle. The complainant would have needed to explain why the document should not be believed, and in the absence of contradictory evidence, tribunals are generally reluctant to disbelieve contemporaneous records created by disinterested parties. The gaps inverted this analysis. Instead of the complainant needing to overcome strong documentation, the establishment needed to overcome the absence of documentation, a far more difficult burden.
The implications for other operators of licensed establishments, hospitality venues, and premises open to the public are significant. The March 2024 incident demonstrates that documentation failures do not merely create inconvenience in subsequent proceedings; they fundamentally alter the legal landscape by shifting evidentiary burdens, enabling adverse inferences, undermining witness credibility, and increasing both the probability and the cost of adverse outcomes. An operator who invests in developing a comprehensive conduct policy but does not invest equally in training, monitoring, and enforcing compliance with documentation requirements has built a structure on an unstable foundation. The policy's protective function depends entirely on its consistent execution, and that execution must occur in real time, on busy evenings, when staff are tired and the temptation to defer paperwork to the next day is strongest. The cost of that deferral, as the March 2024 incident illustrates, can be exposure to a human rights complaint that a properly documented record might have resolved at the investigation stage or defeated at hearing.