Ethics and Confidentiality: Reflections and Lessons Learned Post-Parent and Bruckert v R and Magnotta

Abstract

In May 2012, a former research assistant contacted the Montréal police about an interview he had conducted with Luka Magnotta for the SSHRC-funded research project Sex Work and Intimacy: Escorts and their Clients four years previously. That call ultimately resulted in the Parent and Bruckert v R and Magnotta case. Now, a decade later, we are positioned to reflect on the collective lessons learned (and lost) from the case. In this paper, we provide a lay of the Canadian confidentiality landscape before teasing out ten lessons from Parent c R. To do so, we draw on personal archives, survey results from sixty researchers, twelve key informant interviews with qualitative sociolegal and criminology researchers, and documentary analysis of university research policies. The lessons, which range from the clichéd, to the practical, to the frustrating, have implications for the individual work of Canadian researchers and for the collective work of academic institutions.

Résumé

En mai 2012, un ancien assistant de recherche a contacté la police de Montréal au sujet d’une entrevue qu’il avait réalisée avec Luka Magnotta dans le cadre d’un projet de recherche financé par le Conseil de recherches en sciences humaines du Canada (CRSH), un projet intitulé Sex Work and Intimacy: Escorts and their Clients et qui avait été réalisé quatre ans auparavant. Cet appel a donné lieu à une affaire juridique, soit Parent et Bruckert c R et Magnotta. Dix ans plus tard, nous sommes maintenant en mesure de réfléchir aux leçons collectives tirées (et perdues) de cette affaire. Dans cette foulée, cet article donne un aperçu du paysage canadien de la confidentialité avant de tirer dix leçons de Parent c R. Pour ce faire, nous nous appuyons sur des archives personnelles, les résultats d’un sondage mené auprès de soixante chercheurs, douze entrevues avec des informateurs clés, des chercheurs en sociologie juridique et en criminologie qui utilisent des méthodes de recherches qualitatives, ainsi qu’une analyse documentaire des politiques de recherche universitaire. Ces leçons, qui vont des clichés aux frustrations en passant par les aspects pratiques, entraînent des répercussions sur le travail individuel des chercheurs canadiens et sur le travail collectif des institutions universitaires.

Introduction

On May 29, 2012, Canadians across the country awoke to the news of a gruesome murder in Montréal. Over the ensuing days, developments were salaciously reported in what came to be known as the “body parts case” (Blatchford and Banerjee 2013). We read that Luka Magnotta—quickly identified as the primary person of interest—was a “porn star” whose history of disturbing behaviours had been reported to the police, who had not acted upon the information (e.g., Hodge 2017, n.p.). The media breathlessly recounted that Magnotta fled the country, speculated wildly about possible disguises, and meticulously detailed the international manhunt (e.g., Chung 2012; Hamilton 2012; Warmington 2012).

This was the context in which, on May 31, 2012, four days before Magnotta was apprehended in a Berlin cybercafe, a research assistant “reached out to the Montréal police,” advising them that they had interviewed Luka Magnotta (under the pseudonym “Jimmy”) in 2007 (C. Bruckert, personal communication, 2012). Why a former research assistant hired by Colette Parent and Chris Bruckert to conduct interviews for a SSHRC-funded project entitled “Sex Work and Intimacy: Escorts and their Clients” proactively contacted law enforcement remains unclear. What is clear, however, is that in so doing, this individual breached the promise of research anonymity and confidentiality and set in motion a series of events that would culminate in a court case, a wide-ranging (and sometimes intense) debate within the academic community, a protracted battle between the Canadian Association of University Teachers (CAUT) and then University of Ottawa president Allan Rock, and an updated iteration of the Tri-Council Policy Statement: Ethical Conduct for Research Involving Humans (the policy governing all research ethics boards across Canada) (CIHR, NSERC, and SSHRC, 2018).

Now, a decade after that phone call to law enforcement, we are positioned to look back and reflect on some of the other impacts the case has had on researchers and universities—the collective lessons learned, and lessons lost—and consider possible impacts on how research projects are organized, on the day-to-day practices of researchers, and more broadly on the policies and actions of institutions. In so doing we endeavour to build on, and contribute to, the ongoing examination of research ethics in policy and practice by criminological, sociological, and sociolegal researchers studying complex social phenomena (van den Honaard 2001, 2002, 2014; Haggerty 2004; Palys and Lowman 2014). To this end, we start by laying out the Canadian research confidentiality context before using a chronological narrative structure to anchor the intertwined stories of Chris Bruckert (one of the researchers involved in the case), Canadian researchers more generally, and academic institutions. To tell these overlapping stories, we draw on personal archives, survey results and key informant interviews with qualitative sociolegal and criminology researchers across Canada, and documentary analysis of university research policies.

The Lay of the Land

Within this journal (and elsewhere), we find an ongoing conversation between sociolegal researchers, criminologists, and sociologists about the significance of confidentiality in research into contentious, deviant, or illegal phenomena (van den Honaard 2002; Palys and Lowman 2000, 2006, 2014; Palys and MacAlister 2016; Palys, Turk, and Lowman 2018). Subpoenas or seizure warrants which force—or endeavour to compel—researchers to release sensitive data risk undermining the trust researchers have painstakingly established, making participants hesitant to take part in research (Palys and Lowman 2000, 2014; Palys and Ivers 2018; Maillé 2018; Kolla and McClelland 2017). In short, the potential that the communications between members of marginalized or criminalized communities or groups and researchers might be shared with the police, the courts, or other state apparatuses can have a chilling effect on research, with deep epistemological consequences; ultimately, it may limit what comes to be known about the social world, threaten the integrity of the research produced, and restrict our collective ability to understand complex sociolegal phenomena. An inability to honour the promise of confidentiality can also imperil the safety and well-being of individual research participants. A breach of confidentiality can have devastating impacts, including ostracization from family and friends; loss of employment, social assistance, and housing; the apprehension of children by the state; harassment, bullying, and violence; and social exclusion and marginalization (Bostock 2002; Fitzgerald and Hamilton, 1996; Palys and Lowman 2014; van den Hoonaard 2002).

The importance of research confidentiality begs the question, how privileged (and therefore how protected) is researcher–participant communication? There are three principal kinds of legal privilege recognized in Canada. The first, constitutional privilege, refers to parliamentary privilege under the Constitution Act and the rights and freedoms guaranteed by the Canadian Charter of Rights and Freedoms (see Palys and Lowman 2000 for a fulsome discussion of statutory privilege). The second, class privilege, is a comprehensive protection that prohibits the disclosure of communications occurring within the context of very specific relationships (e.g., lawyer–client, doctor–patient, police–informant); the lawyer, doctor, or police officer cannot disclose the privileged information they have received and cannot be compelled to do so by court processes. Finally, there is case-by-case privilege. In Canada researcher–participant (and journalist–confidential source) communications fall under the latter category. In real terms, a researcher (like a lawyer, doctor, or police officer in the above-noted relationships) receives information that is privileged; that privilege belongs to the participant. It follows that the participant—but not the researcher—can waive that privilege should they so choose; however, the courts can also attempt to compel the researcher to disclose the information. Whether or not a researcher is required to share privileged researcher–participant communication is decided by the courts on a case-by-case basis.

Canadian experts on sociolegal research ethics, however, have cautioned that it is important not to overemphasize the disclosure risk (Palys and Lowman 2000, 2014). There have only been a handful of cases in Canada where research data was seized or subpoenaed, and, to date, no publicly known cases where researchers were required to breach the protection of confidentiality. ¹ At the time of Parent c R, there was only one known case of the court attempting to compel the disclosure of confidential research data. In 1994, the Vancouver Coroner issued a series of subpoenas related to an ongoing inquiry initiated in 1991, requiring Russel Ogden to testify about his master’s degree research (at Simon Fraser University [SFU]) on assisted dying for people with AIDS-defining illnesses (Unknown Female, Vancouver Coroner’s Court, 1991). Notwithstanding SFU’s refusal to provide legal representation, Ogden declined to answer the Coroner’s questions and, threatened with contempt of court charges, successfully mobilized arguments to align his work with the legal test laid out in the Wigmore criteria (Palys and Lowman 2000, 2014; Palys and MacAlister 2016). The Wigmore criteria outline the following four points, which must be proven for a relationship and its communications to be deemed privileged:

• • The communications must originate in a confidence that they will not be disclosed.

• • This element of confidentiality must be essential to the full and satisfactory maintenance of the relation between the parties.

• • The relation must be one which in the opinion of the community ought to be sedulously fostered.

• • The injury that would inure to the relation by the disclosure of the communications must be greater than the benefit gained for the correct disposal of litigation. (Wigmore 1961)

When the coroner accepted Ogden’s arguments, his became the first Canadian case to recognize researcher–participant privilege (Lowman and Palys 2000). ² Following the Ogden case, there was a surge of scholarship on research confidentiality and participant privilege. This body of literature highlights the need for researchers to take a proactive and robust ethical stance towards the protection of confidentiality and not to default to a broad (and potentially mistaken) understanding of what has come to be known as the “duty to report” (Palys, Turk, and Lowman 2018; Palys and Ivers 2018; Palys and Lowman 2006; Lowman and Palys 2000, 2014; Palys and MacAlister 2016). It also details the strategies that researchers can take to ensure they operationalize their ethical stance in research practices and protocols (Lowman and Palys 2000; Palys, Turk, and Lowman 2018; Palys and Lowman 2006, 2014; Palys and MacAlister 2016). Scholarship has also outlined how to proactively anticipate the court’s evidentiary requirements when a claim of privilege is invoked, to align research with the Wigmore test, and to help ensure future jurisprudence is as favourable and respectful as possible to research participants and academic freedom (Palys and MacAlister 2016). Furthermore, the literature has also examined the role of statute-based protection for researchers and assessed protections that have been in place for decades in the United States (Palys, Turk, and Lowman 2018; Palys and Lowman 2006).

Methodology

As well as drawing on documents related to the ethics breach navigated by Parent and Bruckert, this paper mobilizes findings from a larger ongoing research project entitled “The Tactics and Practices of Ethical Scholarly Inquiry,” which explores ethics practices with an eye to informing the protection of confidentiality in research. This mixed methods study is primarily grounded in qualitative inquiry, working with multiple (qualitative) sources (interviews, survey, documents, and reflexive engagement with personal archives) supplemented by quantitative survey data (Creswell 2010).

Between April and July 2020, we conducted twelve semi-structured one-on-one key informant interviews with Canadian scholars who undertake qualitative research with vulnerable populations. ³ These virtual interviews, lasting between one and two hours, focused on the researcher’s experiences, the challenges and obstacles they navigate, and the strategies they deploy to protect their research participants. Interviews were digitally recorded and transcribed verbatim. Given the rather small participant pool, the decision was made to provide limited demographic information (e.g., gender, research area) to protect the anonymity and confidentiality of our key informant interviewees. ⁴

In May 2020, we launched the “Confidentiality and Research: Tactics and practices for ethical scholarly inquiry” online survey. The survey, which ran until December of that year, was completed by sixty participants from across Canada. Respondents included graduate students (28% n = 17), university-based researchers who were early career (0–5 years post-PhD; 25% n = 15), mid-career (5–15 years; 26.6% n = 16), and established (15+ years; 13% n = 8), as well as community-based researchers not affiliated with a university (6.7% n = 4). All participants were, or had been, actively working on qualitative projects on or with criminalized or socially sanctioned communities. The survey was comprised of a series of Likert scales and open-ended questions on perceptions of confidentiality and the tactics and practices used to protect participant data. We also asked respondents to reflect on how they would react to various ethically challenging research scenarios (e.g., the disclosure of violence in the context of an interview).

For the interviews and open-ended survey results, we used NVivo qualitative data management software and employed an iterative approach to data analysis guided by the emergence of themes as they arose. After the emergent themes were coded, they were classified into main and residual categories and subjected to analysis (Braun and Clarke 2006). Quantitative survey data were analyzed using a variety of statistical methods, including correlations and multiple regression. Quantitative survey findings were examined in relation to a range of qualitative data sources to provide a richer analysis.

Finally, to fill out the picture in relation to institutional research ethics, we contacted the research ethics offices of thirty-five universities across Canada, asking them to supply us with their policies related to the protection of researchers in the event of a threatened breach due to subpoena or warrant. The sixteen responses that we received were subjected to documentary analysis, which entailed extracting information into tabular summaries and analyzing themes related to the overall scope of the policy, researcher and participant protections provided, and data protection procedures. These policy documents were triangulated with qualitative perspectives and interpretations from key informant interviews, reflexive engagement with archival records, and our survey results, to examine the complex interplay and disconnects between written policy and real-world practice.

Lessons Learned

We now turn to ten lessons learned (and sometimes simultaneously lost) by individual researchers and institutions of higher learning in Canada in the aftermath of Parent c R. The lessons, sometimes deeply personal and hard learnt, are organized in chronological order. As we detail the narrative of events, we weave together reflexive commentary with documentary sources and our survey and interview data; in so doing, we link the firsthand experiences of Bruckert to the experiences and insights of others in the field, and the practices and policies of institutions.

1. The Weakest Link Truism

On May 31, 2012, shortly after the former research assistant “reached out to the police,” they left a voicemail for Bruckert (who had hired, trained, and mentored them when they were a University of Ottawa student), indicating their confidence that together they could “help the police with their investigation” (C. Bruckert, personal communication, 2012). That voicemail sent Bruckert spinning (and leaving messages pleading for assistance on answering machines across the University of Ottawa campus). Her hope that the police would not be interested in an interview conducted almost five years previously—four years before the victim, Lin Jun, was even in the country—were soon dashed. Within the hour, Bruckert received a call from Detective Guy Gélinas of the SPVM (Service de police de la Ville de Montréal) who advised her that he would be securing the necessary warrant to seize the “Jimmy” interview. Bruckert’s rather frantic efforts to explain that she simply did not know whether she had done an interview with Magnotta was firmly rebuffed—apparently the research assistant had provided compelling enough information to convince the police of the interview’s existence and relevance.

Clearly, the first lesson is that confidentiality is only as solid as the weakest link. The research team’s careful protocols (e.g., rigorous anonymization, oral consent, honorariums signed with a pseudonym, the deletion of all potentially identifiable information such as email addresses) were for naught; all it took was one team member to remember the participant and his pseudonym and reach out to police despite their ethics training. The potential that one member of a research team could jeopardize the ability of all team members to honour their confidentiality commitments resonated with the Canadian researchers we spoke with. One of our key informants explained: “I’m really concerned about this [protecting research participant confidentiality], in part, because of Chris Bruckert’s experience. Researchers not really being able to have a lot of control over what people do after the process. As many times as you can have conversations about the importance of confidentiality, you don’t know what other people are going to do.”

This lesson is not specific to the Parent and Bruckert case, of course. One key-informant researcher described an experience that speaks to the same issue. During a community-based research project, a research assistant—who had been entrusted with rigorously de-identifying data—failed to adequately do so. When the narratives were brought back to the community for discussion, community members were able to identify participants. The researcher noted:

I’m doing less of the research myself. I’ve got enough funding and I’ve hired people. I have research assistants, researchers working for me. So now I’m actually not doing everything. In this case, for example, I’d asked the researcher to pull out the narratives. We read them over together. We identified them as good narratives. And then I asked someone else to de-identify them. I didn’t pay much attention to what they did. I think in retrospect, I would have caught some of them, but I can’t be sure because I wasn’t actually thinking hard enough about the people in the room. I don’t want to be righteous and say I would have caught that, and they didn’t. But I do know that because my hands weren’t on every piece of it, that’s an issue. When you’re not doing everything yourself, it’s different.

In short, there are potential risks when one’s program of research begins to expand, and aspects of data collection or analysis start to be delegated. In the context of a team, those organizing and/or conducting interviews are in many ways the individuals in the confidentiality chain who have the most information (e.g., they have met and may have corresponded with the participants). They may also know the pseudonyms. Indeed, there is another lesson to be learned here in relation to pseudonyms (and the potential risk when participants select their own pseudonym). Had Magnotta not chosen Jimmy (his sex work alias) it is unlikely that the research assistant would have been able to identify which transcript was Magnotta’s (see also Palys and McAlister 2016).

2. We Do Not Know What the Future Holds

Critical qualitative sociolegal studies researchers and criminologists (amongst others) routinely find themselves navigating stigmatic assumptions and sensationalized perceptions by Research Ethics Board (REB) officers who demand, for example, that researchers simplify their language (presumably prisoners and drug users cannot understand “big” words) and develop detailed safety protocols. One key informant—who was well situated, knew the participants, and whose project collected information on prisoners’ often-mundane day-to-day activities—found their REB holding up their project: “Anything that is like quote, unquote, crime, is a red flag within the university. They had a lot of questions, that it was like a high-risk project […]. There was a lot of resistance. I think in a weird way, they were scared for my safety. And I’m like, yeah, no.” Another researcher noted a similar problem:

I think it can be one of the kinds of broader cultural problems with ethics committees, is when they get a particular project that’s about criminal activity or sex, they go into these like “what if this happens?” and “what if that happens?” They go into all of these hypothetical scenarios. And then make the researcher respond in order to get their ethics approval, in a way that they don’t impose those kinds of hypothetical questions on people doing market research. But the same thing could happen.

As the researcher quoted above alludes to, we simply do not know what the future holds—for ourselves or our participants. One of the survey respondents spoke to this when they wrote: “Learning about Bruckert’s experiences reminded me of the real possibility of being asked to break confidentiality and the importance of having a plan in place.” However, in the context of stigma-driven REB decisions it is important to note that sex work was completely extraneous to the Parent c R case—Magnotta’s crime was unrelated to how he earned his living. In fact, this was noted by Justice Bourque, who presided over the case, when she rejected the Crown’s assertion that Magnotta’s past as a sex worker had relevance:

The Respondent submits that the Jimmy Interview contains information on Jimmy’s prostitution activities, which are related to the first-degree murder Magnotta allegedly committed because they both have a sexual dimension. […] [T]he Court cannot reasonably conclude that for a sex worker, both professional and personal sexual activities are the same. Also, this proposition may lead to an insinuation that a sex worker is the type of person more likely to commit murder. Thus, this argument cannot convince the Court.

(Parent c R, 2014, para. 156, 157).

Put another way, it is equally (un)likely that a nurse, a plumber, or an executive interviewed about their work would later come to the attention to law enforcement. It is also unlikely but nonetheless possible that our “banal” research may suddenly become hot. Here, the classic (albeit American) example is the experience of Mario Brajuha, whose ethnographic field notes were subpoenaed when the restaurant he was studying burned down under suspicious circumstances (Brajuha and Hallowell 1986). ⁵ In short, then, “the impossible is possible” cliché resonates at the same time as we must guard against assumptions of elevated risk based on the presumed deviance of specific research populations or sites.

3. Beware of REB Defaults

The May 31, 2012, phone call from Detective Gélinas of the SPVM pushed Bruckert to examine her unquestioning acceptance of REB defaults, including the preservation of the raw data following the completion of data collection. The University of Ottawa ethics application at the time indicated that, “generally, data should be conserved for a period of 5 to 10 years after time of publications” (University of Ottawa 2004, n.p.; emphasis added). The language has since changed to allow researchers to determine a timeframe, asking them to indicate “how long it [data] will be conserved” (University of Ottawa 2020, n.p.). In fact, while data preservation is reasonable for large quantitative studies where anonymized data can be aggregated (and therefore shared) it makes little sense for qualitative projects—after all, we can never share our interviews or field notes without breaching the very ethics protocols we are committing to in our REB applications. Nonetheless, Bruckert had unthinkingly indicated a five-year conservation period; in real terms, this meant the interview was still available to be seized.

We see the importance of perpetual vigilance to avoid inadvertently opening the door for a confidentiality breach in other REB defaults as well. For example, some university template consent forms specify that confidentiality will be preserved within the limits of the law, or that “all data will be kept confidential, unless release is required by law” (Carleton University 2020, n.p.). Reproducing this phrase—essentially promising only partial confidentiality—could seriously compromise a researcher’s ability to protect confidentiality should records be seized or subpoenaed (Palys and Lowman 2000). During our interviews, several key informants noted this concern. One researcher, then a doctoral student, challenged their university’s template consent form because of Parent and Bruckert’s experience and the fear of not meeting the Wigmore criteria. The researcher noted: “I wanted to remove the thing from their standard consent form saying that I would maintain confidentiality up until the limits of the law. It was on all their forms. I changed it to say I would maintain confidentiality up until the highest standards of ethics in my field. I removed the law component and we had to go back and forth a little bit. But ultimately, they agreed with me.”

The quotation above alerts us to the need for researchers to be proactive and persistent—something that can be daunting for a new hire or a graduate student. There is little doubt that, at the institutional level, research ethics offices and boards are principally concerned with liability and protecting the university (Onyemelukwe and Downie 2011; Haggerty 2004; van den Hoonaard 2001). Haggerty (2004) called the establishment of these prescribed codes “ethics creep,” an institutional process not particularly concerned with ethical issues arising through research but focused on the mitigation of risks (to the institution). That said, and contrary to the assumption of many of our colleagues and students, research ethics offices and boards are not necessarily peopled with autocratic and rigid gatekeepers. Rather, they can be comprised of individuals deeply committed to ethical research practices who are trying to navigate competing and sometimes conflicting interests.

Indeed, when we asked our survey respondents how accurate it was to state that their REB had been supportive in ensuring they could implement the protocols necessary to protect participant confidentiality, over half (53%) indicated “very accurate” or “accurate,” while only ten (16.6%) said “not accurate”; 70% of respondents stated that they have never had to make compromises with their REB that would undermine the confidentiality of their research. It would appear that, notwithstanding the general shift to actuarialism, REBs are often open to discussion and revisions if presented with a compelling ethics-first argument (McClelland has found liberally citing the TCPS in any argumentation or rationale to be fruitful). Ironically, it was the University of Ottawa Research Ethics Office who advised Bruckert that the five-year conservation period was not mandatory—albeit only after the police had seized the confidential research data. A more reflexive and proactive approach by Bruckert would have averted the case ever becoming a case.

4. When Sh*t Hits the Fan, There Is Little Practical Institutional Guidance on What to Do

One would assume, given the existence of the Interagency Advisory Panel on Research Ethics (the Tri-Council), offices of Research Ethics and REBs in every Canadian university, and the requirement that researchers articulate (usually in excruciating detail) how they will protect confidentiality prior to being authorized to undertake their research, that there would be a protocol in place when said confidentiality is threatened. Alas, that was not Bruckert’s experience. In fact, while the director of the Research Office and Chair of the Research Ethics Board were certainly engaged and eager to be of assistance, there was a deafening silence emanating from the Office of the Vice-President Research. Bruckert’s union, the Association of Professors of the University of Ottawa (APUO), was initially even less helpful—they advised her that she could “make a representation to the court” that she had promised confidentiality; however, if ordered, she would, “be legally required to abide by the Court’s order,” adding, “at the very least, your objection would be on the record” (C. Bruckert, personal correspondence, 2012; emphasis added). In short, no advice or bad advice.

In the absence of a protocol or even (ethical) guidance, Bruckert felt much like the proverbial headless chicken. Panicking that all her data might be seized by law enforcement and unaware of what actions to take, she made endless phone calls, sent myriad emails, conspired with a trusted colleague to impede access to her office, scoped the countryside for good hiding places to conceal data, consulted with sex working friends about data security, repeatedly overwrote her hard drive, and gave serious consideration to burning digital recorders that might hold ‘ghost’ interviews. ⁶ Fortunately for Bruckert, knowledgeable professors at other institutions stepped into the breach—notably, John Lowman and Ted Palys (both from SFU) and Russel Ogden (then at Kwantlin Polytechnic University). Indeed, the three scholars provided invaluable advice and assistance throughout the process, including submitting an affidavit (Lowman) and academic activism with the Tri-Council (Palys). They also contacted Jim Turk, then president of the Canadian Association of University Teachers (CAUT). It was Turk who took immediate action, including hiring Peter Jacobsen, LLB, to take on the case and represent Parent and Bruckert.

Again, Bruckert’s experience is not unique. One of our key informants described a research team error that resulted in a confidentiality breach. The researcher explained, however, that their REB provided no guidance on how to solve or resolve the issue:

The REB had no process to help me […]. When I went to the REB about it, I was like, this is the worst thing that could happen in research, to me, this is the worst thing to acknowledge. I breached confidentiality. They have this adverse event process. I had to report this adverse event. To me, the process, again, around dealing with that was bureaucratic […] it was transactional, not relational, and research is relational […]. I’m pushed towards a checkbox and how much I get questions about that checkbox, but don’t actually ever engage in real problem solving and the dilemmas. And I think that the ethics board feels like if we just cross all our T’s and dot all our I’s, we won’t have a problem of confidentiality.

5. Do Not Let the Raw Data Out of Your Control (or At Least Make Sure It Is Unusable)

On June 14, 2021, at the urging of Jacobsen, Parent and Bruckert sent the “Jimmy” recording and transcript to his law firm for safekeeping—they signed a retainer (with CAUT committing itself to covering the fees if the University of Ottawa refused to do so) that specifically stated, “We have been retained to provide advice and oppose and set aside any warrants that will require you to produce the documents” (Jacobsen 2012). Shockingly, just one week later, on June 21, 2021, police executed search warrant 500-26-071828-127 at the Toronto offices of Bersenas Jacobsen Chouest Thomson Blackburn LLP and the audio tape and transcript were seized (or, to be more specific, they were sealed and turned over to the courts; Québec Superior Court, 2012). The seizure of the documents reduced options for the researchers—no longer was it possible to refuse to hand over the documents or to destroy them and face the consequences—the only choice remaining was to fight to keep the documents sealed and out of the hands of law enforcement. The lesson here is clear and unequivocal: once the documents are out of your control, your ability to protect your participants’ confidentiality is critically compromised. The importance of retaining control of the data also highlights the imperative of ensuring data cannot be accessed through other means such as university or commercial cloud servers.

There is another implicit lesson here as well—the importance of ensuring there is nothing identifiable to seize. Mindful of this, researchers told us they make sure their data is thoroughly anonymized and that they promptly delete raw data. One researcher explained that they do not collect any names or personal information during field work and strip any identifying information from transcripts as soon as possible, noting: “In the extreme case that there was some kind of a mandate where transcripts or other aspects of the data would need to be turned over, then at least those individual identifiers would have been removed. And so, it would be very difficult at that point to actually figure out who those different individual speakers are in the first place.” Similarly, the research protocols of projects in which Bruckert is involved now ensure there is nothing for the police to seize; her REB protocols explicitly state that transcripts will be thoroughly anonymized immediately following transcription and that digital recordings will be destroyed once transcription accuracy has been verified.

6. Be Prepared to Justify Your Protocol (and Prove You Followed It)

Given the lack of understanding (and sometimes deeply embedded stigmatic assumptions) about qualitative research on sensitive subjects all too often exhibited by REBs, it is unsurprising that researchers perceive—often rightly so—that the rigid framework of the REB is inconsistent with the need to pivot in response to unanticipated (and unforeseeable) challenges that emerge in the field (Felices-Luna 2014). Such perceptions were widespread among the researchers with whom we spoke. For example, one key informant told us that they did not inform their REB when they adapted their research protocol in response to emerging community dynamics precisely because they feared a lack of understanding and support: “[I realized my protocol] is actually a bit different than what I proposed in my summary protocol. And I was like, I am not telling my university about this, because I feel like if I do, they’re going to make it impossible for me to actually function in the way the milieu functions. […] I felt that they were going to be like, pull the plug, now!”

Unfortunately, in the context of a court case, a researcher’s methodological approach is re-situated from a research context to a legal evidentiary one. In Parent c R, all research-related documents (e.g., REB application, consent forms, recruitment materials, training material for research assistants) were submitted to the courts to affirm not only that research confidentiality was taken seriously (e.g., that research assistants were trained in the importance of ethics and confidentiality and that this training was ongoing) but also that the research procedures had been meticulously followed. Had the court discovered the researchers had failed to follow their protocol, the case could have been significantly damaged—perhaps irreparably so.

Here the lesson is three-fold. First, researchers can use the REB process defensively to establish the evidentiary record demonstrating that research ethics and confidentiality were paramount. Second, researchers must, as Palys and Lowman (2014) have repeatedly argued, “walk your talk” given that, perhaps predictably, “courts expect those who claim confidentiality is essential to their work to behave in a manner consistent with that claim” (Palys and Lowman 2014, 205). The final lesson is directed at REBs—they must create a truly responsive and supportive environment and work collaboratively with researchers to accommodate the complex and dynamic nature of qualitative inquiry.

7. Heed Wigmore: Never Waver From the Promise of Absolute Confidentiality

The Parent c R court case—to set aside the above-noted search warrant—was heard at the Québec Superior Court in Montréal from April 3 to 4, 2013; Justice Sophie Bourque presided. Based in part on the precedent set by Ogden, Jacobsen successfully argued that the Wigmore test should be used. In and of itself, this is useful and affirms what John Lowman and Ted Palys have maintained for years, that researchers must develop their protocols with an eye to the protection of Wigmore by ensuring their projects stand up to the Wigmore test (Palys and Lowman 2000, 2014; see also Palys and MacAlister 2016). In other words, proactively engage a legal test that is conventionally deployed retroactively.

Truth be told, Parent and Bruckert did not consciously design their research project “Sex Work and Intimacy: Escorts and their Clients,” with an eye to Wigmore. Fortunately, however, like many researchers studying stigmatized and criminalized sectors, they had meticulously implemented tactics and strategies to protect participant confidentiality and guard against the possible seizure of the data set by police, not realizing that the Supreme Court of Canada in M (A) v Ryan, 1997, had ruled that “fishing expeditions are not appropriate where there is a compelling privacy interest at stake, even at the discovery stage” (at para 37). Parent and Bruckert had also promised absolute confidentiality—meaning they were able to meet the first of the four Wigmore criteria—that the communications must originate in a confidence that they will not be disclosed. In this regard, Justice Bourque accepted the research materials as evidence (including consent forms, recruitment materials, research assistant training modules), noting that “both the recruitment flyer and the recruitment letter sent to potential participants made explicit reference to the promise of confidentiality and anonymity attached to the research process” (Parent c R, 2014, para. 47) and that “there was no limitation to the promise of confidentiality given by the interviewers to the participants” (ibid., para. 53). ⁷ An important caveat to this promise is, of course, provincial/territorial child protection legislation (e.g., Quebec’s Youth Protection Act, para. 39), which enshrines everyone’s duty to report ongoing or imminent child harm to a designated person, organization, or government authority. ⁸

The lesson is simply that protocols, consent forms, and information documents must explicitly state that the project would not be possible without the protection of confidentiality (e.g., statements such as: “this research is contingent on confidentiality”; “all communications must originate in confidence”; and “I will do everything possible to maintain confidentiality” might prove to be helpful). It is also vital to ensure that there is no aspect of the informed consent process that can be construed as a waiver of privilege; quite simply, saying “within the limits of the law” anywhere in the project could jeopardize the ability to pass the Wigmore test.

Embedded in the imperative to proactively use Wigmore strategically and defensively is a helpful reminder about the fragility of the protections research participants are afforded. Given that researcher–participant communications are not guaranteed the protection of class privilege, the onus is on researchers to assert privilege on a case-by-case basis. As Justice Bourque wrote, “The burden of persuasion thus remains on the Petitioners [Parent and Bruckert] to show that the public interest in protecting the Jimmy–researchers relationship outweighs the public interest in investigating and prosecuting the alleged crime” (Parent c R, 2014, para. 148). Ironically, this lesson may have been undercut by the media attention afforded the case and headlines by CAUT asserting, “Court upholds researchers’ right to protect confidential information” (2014, n.p.). It is troubling that Bruckert is routinely thanked for ensuring that research confidentiality is protected. Sadly, Parent c R did no such thing. Absent guaranteed privilege for researcher–participant communications, such cases will continue to be decided on a case-by-case basis.

8. Even When You Win, You Lose

It was not until January 23, 2014, that Justice Bourque rendered her decision quashing the search warrant. There was reason to celebrate when the court ruled that, “the Confidential Interview is covered by the researcher–participant confidentiality privilege” (Parent c R, 2014, para. 212). In her decision, Justice Bourque noted that the “public interest in respecting the promise of confidentiality is high” (ibid., para. 210). Furthermore, she asserted that the social benefits of maintaining confidentially for the public, for the communities being researched, and for academics outweighed the potential benefits of disclosing information from the interview to the (then) ongoing criminal investigation.

Justice Bourque further noted that the interview—conducted more than four years prior to the offence—was decidedly not “the actus rea of the alleged crime” (ibid., para. 154), that the police had uncovered forensic and videographic evidence, and that “[t]he fact that more relevant contemporary material in regard to his [Magnotta’s] mental state does exist, that Magnotta has a reasonable expectation of privacy, and that Dr. Woodside’s ⁹ opinion that the 2007 material is likely to be of minimal assistance to a NCR [not criminally responsible] assessment militate in favor of recognizing a privilege” (ibid., para. 177). However, and notwithstanding the above, Justice Bourque deemed that, “the possibility remains that the information derived from the Confidential Interview could be of use to an NCR assessment” (ibid., para. 178). Accordingly, citing M (A) v Ryan, 1997, para. 39, Justice Bourque exercised her right to “examine the Confidential Interview transcripts kept under seal” (ibid., para. 183). In other words, to rule in favour of the petitioners’ right to protect confidentiality, confidentiality first had to be breached!

9. Do Not Assume Your Institution Will Have Your Back

University-based researchers might assume that their institutions would be invested in standing up for the protection of participant confidentiality—even if only motivated by the self-interested desire to safeguard the institution’s reputation (especially considering condemnation of SFU when they failed to support Russel Ogden; see Palys and Lowman 2014). That was certainly what Bruckert anticipated when she, along with Colette Parent, Jim Turk, and Christian Rouillard (then the APUO president), met with two University of Ottawa Vice-Presidents (Governance and Academics; surprisingly, the VP research was not in attendance) on June 20, 2012, to request that the University of Ottawa assume responsibility for legal fees. After all, the project was done in Parent and Bruckert’s capacity as professors, the SSHRC research funds were administered by the University of Ottawa, the project had received approval from the institution’s REB, and the ability to undertake research—essentially 40% of Parent and Bruckert’s workload—would be compromised if they lost community trust by failing to honour their promise of confidentiality. It took a full six months (until December 19, 2012) to receive a (negative) response from then University of Ottawa president Allan Rock:

The University of Ottawa recognizes its role pursuant to the Tri-Council Policy Statement on Research Involving Human Subjects in safeguarding entrusted information. However, the University does not consider that its role extends to the payment of legal costs if researchers decide to challenge the seizure of research records in the context of criminal proceedings. Consequently, the University will not assume responsibility for payment of the ongoing legal costs in this matter.

(A. Rock, personal correspondence, 2012) ¹⁰

The failure of the University of Ottawa to cover legal fees raises an important, and decidedly pragmatic, concern: what happens when CAUT is not willing or able to cover the significant legal fees incurred when researchers take a principled stance? Or what happens to student researchers where there is no CAUT or similar entity to act as a legal safety net? We do not know whether researchers will or have complied with the demands of law enforcement or the courts because of fiscal constraints, nor do we know whether researchers have quietly handed over documents after assessing the career/professional costs of challenging their (unsupportive) institution as “too high” and concluded that, on balance, their interests do not align with those of their participants. But we do know that the University of Ottawa’s response has engendered fear and cynicism amongst researchers. One key informant interviewee told us: “I don’t trust that the university would have participants’ confidentiality at their best interest.” Another explained their position: “Our universities are these monoliths that are interested in institutional risk protection. They’re not actually interested in the well-being of their graduate students or faculty doing research, as has been shown in several of these cases where they decided not to defend people or not to provide resources for faculty members who had data subpoenaed.”

The University of Ottawa’s response also begs the question: what does support look like? Notwithstanding Alan Rock’s assertion that the university “has expressed strong support for, and taken steps to defend, the confidentiality of the research records in the possession of Professors Bruckert and Parent” (A. Rock, personal correspondence, 2013), it is unclear what those steps were; not only was there a refusal to cover costs, but there was also nary a whiff of (any sort of) support from the administrative apparatus. Indeed, although many University of Ottawa faculty, REB members, and students (Feibel 2014), as well as scholars from other institutions, were vocal in their support, Bruckert was never contacted by the Office of the President or any of the five Vice-Presidents, or, for that matter, even her Dean at the time.

10. Some Universities Are Walking the Talk, but Clear (or Any) Policies Are Still Woefully Rare

In one of the two known cases ¹¹ where confidentiality was threatened post-Parent c R, the researcher’s experience of institutional support was dramatically different than Bruckert’s. In 2016, Dr. Greta Bauer, a tenured epidemiologist at Western University who had done research on transgender people’s health was asked to testify as an expert witness in the Québec Superior Court. The human rights charter challenge case claimed infringements to rights of integrity and security of the person, due to how identity documents—which did not conform to one’s identity—could lead to suicidal ideation for transgender people (Centre de lutte contre l’oppression des genres [Centre for Gender Advocacy] c Québec, 2016). The Attorney General of Québec, in response, and to verify the accuracy of her testimony, sought to access raw data from Dr. Bauer’s study, which was potentially identifiable. Not only did the university’s Associate Vice President for Research submit an affidavit on the damage that data disclosure would have on the study’s participants and to research more generally, but lawyers from the university worked collaboratively with those from the Center for Gender Advocacy. Ultimately, the judge ruled against the release of the data (A. McClelland, personal correspondence, 2018). The experience of Dr. Marie-Ève Maillé (then a doctoral student at the Université du Québec à Montréal [UQAM]) was less exemplary in terms of institutional support although she too successfully challenged a subpoena (initiated by a power corporation) requiring her to produce her doctoral research on rural Québec residents’ experiences of a wind power farm (Rivard c Éoliennes de l’Érable, 2017). Dr. Maillé accessed pro bono legal counsel through the Quebec organization Justice Pro Bono; UQAM filed for intervener status. Notably, after her experience, Maillé publicly called on researchers to “have a lawyer working for you who will defend your interests (and not those of your research institution)” (Couturier 2019, n.p.).

It would appear that, unlike the University of Ottawa (in Parent c R, 2014) and SFU (in Ogden’s case, Unknown Female, Vancouver Coroner’s Court, 1991), some universities are, at least to some extent, supporting their professors and graduate students when they challenge threats to research confidentiality. That said, individual institutions stepping up still leaves researchers unsure about the support they can anticipate should they find themselves in need of legal services. Indeed, when we asked our survey respondents whether their university (or community organization) had a policy to provide guidance in instances where research ethics come into conflict with the law (e.g., if a researcher’s data is subpoenaed), forty-two participants (70%) indicated that they did not know, while twelve (20%) indicated “no,” and only six (10%) indicated “yes.” Of those who answered “yes,” just two stated that the policy guaranteed the paying of legal fees if a researcher resists a legal order to disclose confidential data; three participants indicated they did not know, and one answered “no.”

This lack of clarity is shocking given that, as a result of academic activism on the part of Ted Palys and others (Palys and Ivers 2018), the most recent Tri-Council policy statement (2018) specifies that all universities should “establish policies, procedures or guidelines that explain how [they] will fulfill [their] responsibilities to support [their] researchers” in the context of a legally compelled breach (61). In 2018, Palys and Ivers found that most universities (87.7% of their sample) had not considered developing a policy, had started to but abandoned the endeavour, or were just beginning to develop such a policy. When we approached thirty-five university REBs with the same question in 2021, only sixteen got back to us; of these, three indicated the drafting of policies was “in progress” and only two (McGill University and the University of Waterloo) stated unequivocally “yes” to having a policy. ¹²

Arguably, this lacuna may speak to a lack of institutional urgency. Indeed, one university explained that their “Legal Counsel’s office does not believe that a policy is warranted, given the number of times that this has been an issue.” Researchers conducting qualitative research on criminalized and socially sanctioned issues may well beg to differ. As one of our key informants put it (when asked what kind of training is needed to support researchers): “It’s not fucking training sessions that people need. What people actually need is the ability to protect their legal rights. Like knowing that I would have access to good quality legal support if necessary. That would have my interests and not my university’s interests at heart. That would be the thing.” In short, the lesson learned is that while there has been some piecemeal progress, most researchers still cannot assume they will be supported should research confidentiality be threatened in the context of a legal process.

Conclusion

In this paper, we have drawn on personal archives, a survey, twelve interviews, and university policy documents to tease out ten lessons learned (and to highlight those that have not been learned) from Parent c R. These lessons range from the clichéd, to the practical, to the frustrating. At the micro level, there are a series of lessons that speak to practices individual researchers can take to mitigate the potential of data breaches—from the development stage to research ethics submissions to the day-to-day management of research project data. Here, we considered the importance of anticipating the unexpected; questioning REB default protocols; remembering that one weak link can unravel research protections; ensuring identifiable data is deleted as soon as possible; being mindful that research materials (e.g., ethics protocols) may be translated into evidence in a court; defensively mobilizing the Wigmore criteria as a protection mechanism; and—vitally—remembering that despite a favourable outcome in Parent c R, the current case-by-case examination of legal privilege means that research participants are still at risk. We also addressed a series of lessons related to navigating the more macro institutional level. Collectively, the lesson learned here is that—notwithstanding some universities establishing policies and stepping up—researchers must anticipate a lack of clear practical guidance on how to proceed in the context of a compelled data breach and an absence of support.

Lurking behind, but never addressed in this paper, was an elusive “dark figure.” The crime Magnotta was accused of—and ultimately convicted for—was, as the research assistant noted in the May 31, 2012, voicemail that so disconcerted Bruckert—“heinous.” As such, Magnotta is unquestionably a highly unsympathetic individual. Perhaps unsurprisingly, Bruckert was routinely confronted by individuals perplexed at her desire to “protect someone like that” and it is nearly impossible to find media articles from the time—even when they were supportive of the researchers’ claim—which do not also delve into the gruesome details of Magnotta’s crimes (e.g., Canadian Press 2013; Fine 2014; Ottawa Citizen 2014). Of course, who Magnotta is or what he did is extraneous to Parent c R; however, the focus on Magnotta certainly muddied the waters, and it is an open question whether public, collegial, and, perhaps most importantly, institutional support would have been more forthcoming if the case involved, for example, rural Quebec residents preoccupied by wind farms eroding their quality of life. Herein lies another lesson: when the case is salacious it is easy to lose the plot.

Perhaps, though, we can also take comfort that, despite the horrific nature of Magnotta’s crime, the courts upheld researcher–participant confidentiality. During our key informant interviews, researchers expressed a great deal of uncertainty and apprehension (and even fear of engaging in research on issues with criminal or social sanction) because of Parent and Bruckert’s experiences. Our goal with this paper, and our broader ongoing project, is to begin to defuse that apprehension and to start providing answers to address uncertainties. That said, absent guaranteed privilege, future researchers will—like Ogden, Parent, Bruckert, Bauer, and Maillé—find themselves in the courts fighting to protect their research subjects’ privilege rights. That fight will necessitate arduous work and (often expensive) legal expertise. It is long past time for all Canadian universities—self-congratulatory and rightly lauded for the innovative and important research undertaken in their institutions—to step up and truly walk their talk.