1) Preamble: Research Ethics at SFU *

2) Can We Defend Researcher/Participant Privilege? *

3) Going the Distance *

4) Anticipating Elements of a Wigmore Defence *

5. How Do Other Forms of Privilege Considered by the Supreme Court Compare to the Researcher-Participant Claim? *

6. Our Appeal of the URERC’s Negative Decision *

7) Academic Freedom, Research Ethics and Law *

8. What the Future holds *

Appendix A: A Legal Opinion Regarding Researcher-Participant Privilege by Paul Jones, Legal Counsel to the Canadian Association of University Teachers (CAUT) *

Appendix B: An Interchange on the Disciplinary Standards of Sociology Regarding Confidentiality and Its Protection *

For some five years now, Simon Fraser University has struggled with the problems that arise when research ethics and law potentially conflict over guarantees of confidentiality. The dilemma began in 1994 when Russel Ogden was subpoenaed by the Vancouver Coroner and asked to reveal the identity of some of his research participants. Although it is supposed to cherish academic freedom, the university chose not to appear in court to defend its research ethics application decisions and the integrity of the research enterprise. At the behest of the VP Research in 1994, the URERC introduced a policy of "limiting confidentiality," ostensibly as a way of providing prospective research participants with information about the legal limits to confidentiality. However, the Committee’s minutes reveal that one of the purposes of the policy of limited confidentiality was to "protect the university." From that point on, a conflict of interest has compromised the research ethics approval process at SFU.

In 1995, the URERC began forcing researchers to use a "limited confidentiality consent statement," which says that:

In December 1997, one of us (Lowman) submitted a research ethics application to the University Research Ethics Review Committee (URERC) to study the owners and managers of escort services, massage parlours and body rubs. In February 1998, the two of us submitted an application to study sex workers in off-street venues. When we submitted these applications, we informed the URERC that we could not use the limited confidentiality statement because we believed that it compromised our ability to protect confidentiality to the full extent permitted by law. As there is no statutory protection of researcher/participant privilege, we argued that the only way to protect confidential information is to design our research to take maximum advantage of the Wigmore test, which the Supreme Court of Canada has ruled (R. v. Gruenke, 1991) is the appropriate way to establish privilege on a case by case basis. Because it can be interpreted as a waiver of privilege (cf. A. (L.L.) v. B. (A), 1995, and M. v. Ryan, 1997), and may thereby expose research participants to harm, the limited confidentiality consent statement is unethical.

On October 8, 1998, after lengthy negotiations with the URERC, the Committee refused to accept our alternative consent statement. In December 1998, we appealed this decision. On April 1, 1999, we received a consent statement from the VP Academic, the University authority who heard the appeal, saying that we could proceed if we agreed to use his consent statement. For reasons that we elaborate below, we cannot use his statement either. We are now faced with the prospect of invoking Section IV.3 of the SFU Code of Faculty Ethics and Responsibilities (SFU Policy A30.01), which says that faculty members:

We believe the URERC and the VP Academic have infringed our academic freedom. We have acted diligently, trying to accommodate the URERC’s unauthorized substantive change to the SFU Research Ethics Policy, but without success. Recently the URERC twice approved a consent statement that satisfies our ethical and legal concerns. Consequently, we cannot understand why we should not invoke section IV.3 of SFU Policy A30.01 and proceed with our research using a consent statement we know the URERC has already approved for two other researchers. Also, we have initiated a grievance against the VP Academic using Article 12 (Resolution of Disagreements) of the "Agreement on the Framework for Collective Bargaining and Consultation Between Simon Fraser University and the Simon Fraser University Faculty Association" (the Framework Agreement).

This, our third submission to the SFU Research Ethics Revision Task Force, tells the story of our appeal using the appeal mechanism contained in the SFU Research Ethics policy (R20.01) and considers the implications of recent Supreme Court decisions for defending researcher-participant privilege. Also, it briefly examines the Statistics Act, which requires that Statistics Canada researchers take an oath that they will abide by a code guaranteeing unlimited anonymity and confidentiality.

Our position regarding the ethics and law of confidentiality has been outlined in our two previous submissions to the Task Force:

• The History of Limited Confidentiality1 looked back at the circumstances and approach that guided SFU in its treatment of Russel Ogden, and in its subsequent imposition of a policy of "limited confidentiality." The general conclusion of that document was that the policy was not properly enacted, did not meet its purported objectives of protecting research participants and, instead, seemed to be guided by university image management and liability considerations. It did this by engaging a caveat emptor approach to research ethics, i.e., where "being ethical" meant giving research subjects fair warning that the researcher and university would abandon the participant and divulge confidences if a court "required" the information. Although the university has apologized and made amends to Russel Ogden, the experience of our recent appeal under the SFU Research Ethics Policy (R20.01) reveals that the University is still prepared to use the ethics application process to manage its image and minimize liability.
• Informed Consent, Confidentiality and the Law2 articulated a strategy we believed would most effectively protect research confidentiality and research participants to the full extent permitted by law, as the Tri-Council Policy Statement3 requires. The strategy is based on designing research in anticipation of invoking the Wigmore criteria4 when and if researcher-participant confidentiality is challenged, in order to try to establish class privilege in common law for researcher/participant communications. A properly researched independent legal opinion would have helped this endeavour. However, neither the URERC nor the University provided one.5

A professional legal opinion would have been helpful for everyone involved in the ethics dilemma at SFU, and the controversy surrounding the limited confidentiality consent statement. In their review of the "Russel Ogden Decision" Blomley and Davis (1998)6 note that the university never sought an independent legal opinion regarding confidentiality and the law. Nor did the VP Research or the Ethics Committee obtain one in 1995 prior to imposing use of the "limited confidentiality consent statement" on researchers collecting self-report information about law violations. In 1997, we asked the university to obtain an independent legal opinion about the limited confidentiality consent statement because independent legal commentary posted on the Internet7 gave reason to believe that the statement the URERC was using actually exposed research participants to harm. Nevertheless, the committee continued to require use of the statement without the benefit of a formal legal opinion, and without offering a defence of its position in response to the specific legal concerns we raised (other than to say "Wigmore won’t work," although they were aware it had already "worked" for SFU researcher Russel Ogden).

To the extent that the URERC's refusal of our two ethics applications was grounded in points of law, the committee had an obligation of fairness to acquire the information needed to properly evaluate our claims. There was no appropriate legal expertise on the committee itself,8 and the SFU ethics policy empowers the committee to seek appropriate outside expertise when it does not exist among the committee membership. Eight months later, a budget was allocated; fifteen months later, the Committee has still not obtained a properly researched independent legal opinion.

More recently, in lieu of the University and the URERC obtaining an independent legal opinion about the issues preventing approval of our two ethics applications,9 the SFU Faculty Association sought and obtained a legal opinion (see Appendix A) from the Canadian Association of University Teachers (CAUT). Mr. Paul Jones, Legal Counsel to CAUT, provided the opinion. We thank CAUT and SFUFA for their assistance. Reading and considering the implications of Mr. Jones's opinion only adds to our dismay that the University and the URERC forged ahead with "limited confidentiality" without following established university policy and procedure, and in the absence of a properly researched legal opinion.

The CAUT legal opinion responded to the following questions:

1. What jurisprudence supports the use of the Wigmore test to establish researcher/ participant privilege?
2. Does the document entitled "Informed Consent by Subjects to Participate in a Research Project of Experiment" compromise a claim to privilege based on the Wigmore test?
3. How can this "Informed Consent" document be designed to maximize chances of successfully asserting the Wigmore criteria?
4. Are there alternatives to the Wigmore test that could be advanced to establish researcher/participant privilege?

In his opinion, Paul Jones identifies four leading Supreme Court cases relevant to considerations of privilege: R. v. Gruenke (1991); R. v. O'Connor (1995), A.(L.L.) v. B.(A.) (1995), and M. v. Ryan (1997). The Gruenke case, which involves a claim of priest-penitent privilege, is one we have considered at length before. The three post-Gruenke cases all deal with issues of psychiatrist-client privilege in the context of therapy for victims of sexual assault. Collectively, these cases provide valuable clues about how the courts might interpret future claims of privilege using the Wigmore test. In the ensuing discussion, we consider the four cases Jones cites, plus Smith v. Jones (1999)10 a Supreme Court judgement involving solicitor-client privilege that was published on March 25^th, 1999, a month after Paul Jones sent the Simon Fraser University Faculty Association his opinion.11

We formally entered the current debate in response to an article that Dr. Clayman, the VP-Research and then-Chair of the Ethics Committee, wrote for Simon Fraser News, entitled "The Law of the Land."12 This article displayed Dr. Clayman’s view that "The Law" is defined solely by statute. An independent observation made by Dr. Philip Hanson, a member of the URERC at the time we began our dialogue with the ethics committee in December 1997, nicely captured the difference between the URERC’s static conception of law and a more dynamic one:

From the law-as-statute perspective, given there is no statutory protection of the researcher-participant relationship (except for Statistics Canada researchers via Section 18 of the Statistics Act), a policy of limited confidentiality may make some sense. As Bruce Clayman argued, if the absence of statutory protection means that there is no defence to be made of confidentiality, then all you can do is "alert potential subjects about a risk to which they would expose themselves — under Canadian law — if they agree to participate in the research."15 However, there is more to the law of privilege than statute.

In our dealings with the URERC, we pointed out that the Ogden case was significant because it showed that lower level courts would follow the Supreme Court's exhortation in R. v. Gruenke (1991) that the Wigmore test is the appropriate measure for any claim of privilege on a case-by-case basis. The Ogden case also showed that the researcher-participant relationship could pass the Wigmore test. Against this view, the VP-Research and URERC asserted that the Ogden case set no precedent16 and that the Wigmore defence would never work,17 in which case there was nothing a researcher could do but warn subjects that courts might "require" them to give up confidential information.18 In lieu of a statutory protection of confidentiality, there is no legal defence, in which case a subpoena effectively becomes the "limit" of the law. Using this interpretation, protecting confidentiality to the "full extent permitted by law" means until a court order is made, at which point confidentiality is lost. Dr. Ogloff, for example, Chair of the Canadian Psychological Association Ethics Committee and a member of the URERC, stated:

In a more recent statement, Dr. Horvath reaffirmed this view:

Paul Jones's (1999) legal opinion is more optimistic:

Clearly then, the Wigmore test can be used to defend research confidentiality in court. Although no class privilege currently exists for researcher-participant communications (since this would require case law, of which there is none other than the Ogden case on this issue), the Supreme Court has not precluded its development.

A central premise of the URERC’s "limited confidentiality" approach — that there is no legal way to protect confidentiality against court-ordered disclosure — is thus incorrect. The Supreme Court of Canada has recognized the validity of the Wigmore test for establishing privilege and, as long as all four criteria are met, a claim to privilege can be made for any communication. The common law relating to privilege thus is the relevant law when it comes to protecting confidential research information from court-ordered disclosure. If the researcher and the university are to live up to their obligation to defend academic freedom and promote "the highest ethical standards" in research, they must defend confidentiality vigorously in court.

Nevertheless, Jones (1999) warns us that successfully invoking a claim of privilege will not be easy:

The difficulty of winning researcher/participant privilege aside, in the short term, protecting confidentiality "to the extent possible within law" — the minimum standard set by the Tri-Council Policy Statement (p. 3.2) — will mean using arguments based in common law to fight court-ordered disclosure of confidential research information all the way to the Supreme Court of Canada, if necessary. In the longer term, Universities should become engaged in a campaign to win the same statutory protection enjoyed by Statistics Canada researchers (see section 5.v below).

We believe that there is good reason for Simon Fraser University to take up these challenges. We have yet to see a Code of Ethics that does not recognize the importance of confidentiality to the research enterprise, and exhort researchers to engage in its protection. For example,

• Researchers must respect the rights of citizens to privacy, confidentiality and anonymity, and not to be studied. Researchers should make every effort to determine whether those providing information wish to remain anonymous or to receive recognition and then respect their wishes. (Canadian Sociology and Anthropology Association Code of Ethics, section 5.1)
• Sociologists have an obligation to ensure that confidential information is protected. They do so to ensure the integrity of research and the open communication with research participants and to protect sensitive information obtained in research, teaching, practice, and service. (American Sociological Association Code of Ethics, Section 11)
• Informants have a right to remain anonymous. This right should be respected both where it has been promised explicitly and where no clear understanding to the contrary has been reached. These strictures apply to the collection of data by means of cameras, tape recorders, and other data-gathering devices, as well as to data collected in face-to-face interviews or in participant observation. Those being studied should understand the capacities of such devices; they should be free to reject them if they wish; and if they accept them, the results obtained should be consonant with the informant's right to welfare, dignity and privacy. (American Anthropological Association Statements on Ethics, Section 1(c))
• [S]cholars also have a professional duty not to divulge the identity of confidential sources of information or data developed in the course of research, whether to governmental or non-governmental officials or bodies, even though in the present state of American law they run the risk of suffering an applicable penalty. (American Political Science Association Code of Ethics, Section 6.2)

Professionals in other fields also recognise this duty of academics. Judge Steinberg of the British Columbia Provincial Court, for example, in his written opinion in Russel Ogden v. SFU, chastized the university's meek "letter of support"21 and failure to appear alongside Ogden in Coroner's Court to defend the rights and interests of his research participants as a "hollow and timid" defence of academic freedom and research confidentiality, and reminded us that,

Much the same view was expressed by Blomley and Davis (1998) in their independent review of the university's decision-making regarding Russel Ogden:

More recently, the Tri-Council Policy Statement (1998) also reaffirms that confidentiality is worth going the distance for:

In sum, researchers are expected by the professional codes of ethics and policies that govern them to defend confidentiality to the fullest extent. Given these various exhortations, the CAUT legal opinion (Jones, 1999) helps to illustrate just what it means to defend confidentiality to the "fullest extent possible" by drawing our attention to the manner in which it was defended in M. v. Ryan (1997).

The Supreme Court’s reasoning in M. v. Ryan confirms that maintaining confidentiality to the "extent possible within law" involves resisting court-ordered disclosure all the way to the Supreme Court of Canada if necessary. The case involved a 17-year old girl (M.), who was indecently assaulted by her psychiatrist (Dr. Ryan). After the assault, the victim went to a second psychiatrist (Dr. Kathleen Parfitt) for treatment. Dr. Parfitt explicitly discussed the possibility that a court might, at some point, order disclosure of her therapy records. The girl explicitly stated that confidentiality was important to her, and hence that she did not want the records revealed at any point to anyone, especially a court. Dr. Parfitt stated that she would do "everything possible" to ensure no information was disclosed. The girl subsequently sued Dr. Ryan for damages. Dr. Parfitt's records (but not her personal notes) were subpoenaed, and the trial court did indeed order disclosure. In this case the principle at issue involved whether the defendant’s right to secure records potentially relevant to testing the plaintiff’s case against him outweighed the plaintiff’s expectation that communications with her psychiatrist would be kept in confidence.

Protecting confidential information "to the extent possible within the law" means doing exactly what the patient and Dr. Parfitt were prepared to do. By challenging court-ordered disclosure all the way to the Supreme Court, their actions offered the best evidence that their communications originated in a confidence that it would not be disclosed. Indeed, it was not until the case reached the Supreme Court, despite negative and mixed decisions in the Provincial trial and appeal courts, that Dr. Parfitt and M. could say they had done "everything possible" to protect confidentiality "to the extent possible within the law." As an institution whose policies require us to maintain "the highest ethical standards" (see SFU Policy R60.01, Ethics in Research and Misconduct in Research), the defence of research confidentiality requires no less. By overturning the Ogden decision and pledging to support researchers in court, Simon Fraser University appears to have fully accepted this principle.

Having decided that research confidentiality can and should be defended, one must then consider how best to do it. In this regard, we follow Michael Traynor’s advice (1996):

In developing this defence strategy, several principles need to be kept in mind from the beginning of one's research.

If one is going to argue a claim of privilege, then, as Paul Jones notes, "an expectation of confidentiality is absolutely critical to establish privilege" (p.3). This is precisely the point we asserted in our earlier writings25 on the Supreme Court's decision in R. v. Gruenke (1991), Mary Marshall's (1992) legal analysis of the implications of that decision for claiming privilege, and the US literature on court-ordered disclosure (see, for example, US v. Atlantic Sugar, 1980). And this is precisely why we could not use the SFU Ethics Committee’s limited confidentiality consent statement.

Jones's opinion points out that, "Gruenke is not the last word from the [Supreme] Court on the issue" (p.3). The Supreme Court's decisions in R. v. O'Connor (1995), A.(L.L.) v. B.(A.) (1995), and M. v. Ryan (1997) add important information for our understanding of what the Supreme Court of Canada takes to be evidence that there is a clear "expectation of confidentiality" among parties in a given relation.

The feature of R. v. Gruenke that most attracted our attention was the Supreme Court’s interpretation of the Wigmore requirement that for a communication to be privileged it must arise in the confidence that it will not be disclosed. In Gruenke the claim of privilege concerned communications between a pastor, a counsellor and the accused in a murder trial. Because it was not clear that these communications originated in a confidence that they would not be disclosed, the Supreme Court ruled that the communications were not privileged.

Reflecting on the Supreme Court’s ruling in this case, lawyer Mary Marshall (1991) advised:

Following this reasoning, we concluded that the SFU limited confidentiality statement warning that, "it is possible that … the researcher may be required to divulge information obtained in this research to a court or other legal body" effectively kills the possibility of a successful Wigmore defence.26 Consequently, we could not use this statement because it undermines our ability to protect confidential information to the "full extent permitted by law." Unmodified, the warning is unethical because it potentially exposes research participants to harm.

To resolve this dilemma, we suggested that warnings about the risk of court-ordered disclosure be accompanied by assurances that everything possible would be done to prevent disclosure. In order to comply with its own policies, we suggested this would entail the University assisting researchers to resist court-ordered disclosure all the way to the Supreme of Canada if necessary. And in order to maintain the undertaking of "strict confidentiality"27 the URERC requires researchers to sign before approving their ethics applications, this would mean maintaining confidentiality in the face of a charge of contempt of court.

As Jones observes, M. v. Ryan introduces new nuance to the law regarding privilege. The BC trial court behaved as we predicted on the basis of Gruenke alone, deciding against M. and Dr. Parfitt on the grounds that their discussions about the possibility of a court order to disclose implied recognition of a limit to confidentiality, and hence could not be considered completely "confidential." However, the BC Court of Appeal and the Supreme Court provided additional criteria for "a clear expectation of confidentiality." Writing for the majority at the Supreme Court, McLachlin J. stated:

We suggest the italicised phrases — "do everything possible to keep them confidential" and "over their objections" — describe the acid test of an "expectation of confidentiality" under the extant case law cited by Jones (1999). If the parties in the relation do not vigorously resist court-ordered disclosure, their actions de facto show that the expectation of confidentiality was missing. One must walk one's talk.

In addition to being an example of what it means to do "everything possible" to maintain confidentiality, Ryan implies that doing anything less than attempting to maintain confidentiality to the "extent possible within the law" may have other consequences too. For example, anything less than vigorous resistance to court ordered disclosure would likely undermine our ability to meet criteria 2 and 3 of the Wigmore test, which require that confidentiality be integral to the researcher-participant relation and that the community sees the relation as one that should be sedulously fostered. Acquiescing to a court order without a fight suggests that confidentiality is neither, thereby further undermining any prospective claim of privilege.

With respect to the design of research, the unequivocal "expectation of confidentiality" must be built into one's consent statement. Jones (1999) advises that a researcher must affirm that:

Given the demonstrated importance of these two elements in Ryan, we can now consider whether the URERC's limited confidentiality consent statement meets these criteria.

Jones's (1999) analysis of Ryan suggests that a warning about court-ordered disclosure per se does not necessarily constitute a waiver of privilege. The Supreme Court upheld the BC Court of Appeal court ruling that although Dr. Parfitt and the victim-complainant had explicitly considered the possibility of court-ordered disclosure, their communications nevertheless originated in a confidence that they would not be disclosed. The victim-complainant explicitly stated she did not want her communications with Dr. Parfitt disclosed to a court or anyone else and, at one point, Dr. Parfitt stopped taking her usual notes. If there were no notes, a court could not seize them. There was thus clear evidence that both Dr. Parfitt and M understood the importance of confidentiality to their communications — regardless of the apprehended threat of court-ordered disclosure — and acted in a manner that would maximize that confidentiality.

The most important lesson to be learned from this case is that the actions of M and Dr. Parfitt were consistent with their expectations and claims. Although they entertained the possibility of court-ordered disclosure, this did not represent a conscious waiver of privilege. Indeed, as Supreme Court Justice McLachlin put it, "Far from waiving privilege, the appellant has asserted it throughout the proceedings" (M. (A.) v. Ryan, p.6).

The SFU Ethics Committee has argued that subjects in studies of self-reported law violations must be forewarned of the possibility of court-ordered disclosure, even though, as far as we know, no Canadian law court has ever ordered such a disclosure. As we stated in the introduction to this submission, we cannot use this statement fearing that a court could construe it as a waiver of privilege. With the benefit of a properly researched legal opinion now at our disposal, let us look again at the URERC's limited confidentiality consent statement:

Does this statement unambiguously affirm that the researcher and university will go the distance to protect confidentiality? No it does not. From the perspective of providing sufficient information for informed consent and for satisfying the first criterion of the Wigmore test, there are at least three problems caused by the ambiguity of this statement:

a) Stating that one will protect confidentiality "to the full extent permitted by law" is ethically and legally acceptable only if "the full extent permitted by law" includes common law as well as statute, and challenging court ordered disclosure all the way to the Supreme Court of Canada, if necessary. However, the actions of the university when Ogden was subpoenaed, and the denials by URERC members that there is any way to legally defend confidentiality (for examples, see Section 2 above) indicate that, in their eyes, the receipt of a subpoena is effectively the limit of the law. Construed this way, a court might rule that the consent statement represents only a qualified assurance of confidentiality, and consequently that it does not pass the first criterion of the Wigmore test.

b) The statement does not clarify who will "require" the researcher to divulge the information — it could be the university as well as a court. Indeed, the phrase "may be required to divulge information … to a court" gives our research subjects the impression that the University expects the researcher will violate confidentiality if ordered by a court to do so. Our subsequent experience with the appeal of the URERC’s negative decision on two of our ethics applications reveals that this is precisely the University’s expectation. But is this posture a matter of research ethics, or is the ethics application procedure being used by the University to achieve some other end?

Research ethics policies are supposed to protect research subjects from harm and safeguard their rights. However, the URERC’s minutes reveal that limited confidentiality was designed in part to "protect the University."28

The experience of limited confidentiality in other jurisdictions reveals a similar logic. Talking about research practices in the USA, Traynor (1996) suggests that limited confidentiality consent statements do a better job of protecting universities and researchers from liability than they do of protecting research participants from harm:

This is exactly what happened in Atlantic Sugar, Ltd. v. The United States of America, described by Traynor (1996).29 In 1980 the US Customs Court entered an order granting the plaintiffs, Atlantic Sugar Ltd., access to confidential documents from the administrative records of an anti-dumping proceeding filed in court by the International Trade Commission. One of the records in question was Amstar Corporation’s response to an International Trade Commission survey questionnaire. Asserting its right as an intervenor, Amstar sought a limitation on disclosure until all persons whose confidential information was contained in the questionnaire had been given notice of the proceedings and a chance to comment. The court’s response in this case confirmed our fear that the SFU limited confidentiality consent statement could be read as a waiver of privilege:

This is an example of precisely why we could not use the SFU limited confidentiality consent statement — it is unethical because it exposes research participants to harm, and thus violates SFU’s Research Ethics Policy.

c) The limited confidentiality consent statement implies that researchers will always obey a court order when, in fact, they have an ethical option not to (see Appendix B for the American Sociological Association’s position in this regard). The Tri-Council Policy Statement recognises that:

The SFU limited confidentiality consent statement does not provide the information necessary for informed consent because it gives no inkling that a researcher has the right to choose to disobey a court order on ethical grounds. The limited confidentiality consent statement is overly constraining because it implies a doctrine that absolutely subjugates research ethics to law. This doctrine infringes academic freedom, a fundamental principle of the university which is guaranteed by SFU Policy A30.01 and the SFU Framework Agreement.

The concept of a waiver of privilege appears often in the Supreme Court’s decision-making regarding privilege, particularly in the Ryan case (see also A.(L.L.) v. B.(A.) (1995)). Recall that M’s and Parfitt's claims for privilege were rejected at trial because of not having met the first Wigmore criterion. The BC Court of Appeal disagreed with this reasoning, and asserted that mere discussion about disclosure orders did not in itself inevitably undermine the expectation of confidentiality required to meet Wigmore criterion 1. The Appeal court nonetheless also rejected M’s claims of privilege, but did so on the grounds that she had waived privilege by failing to assert it right from the start. The Supreme Court disagreed with the BC Court of Appeal on that score, stating that:

If a researcher tells a subject that a court might "require" disclosure of information, can it be claimed that the communication is really "confidential?" We are concerned that researchers and/or the University can treat this statement as a waiver of privilege so that they don’t have to go the distance and defend research participants against the harm potentially created by court-ordered disclosure. Indeed, this is precisely why we protested the way the URERC, through one of its representatives, used the limited confidentiality statement to "cover" researchers and the University. Neither the URERC nor the VP Research responded to our memo protesting their actions in this case. We suggest their actions could not pass muster under the Tri-Council Policy Statement.31

The Policy Statement is unequivocal that, "The consent of the participants shall not be conditional upon, or include any statement to the effect that, by consenting, subjects waive any legal rights" (p. 2.6). This suggests that the URERC's approach and limited confidentiality consent statement is unethical in at least two respects:

1. because researchers can interpret it as a waiver of privilege, it effectively makes subjects waive their legal rights; and
2. contrary to the principle of informed consent, it secures this waiver of privilege without explicitly informing subjects that is what they might be doing by accepting the URERC's statement.

The Supreme Court’s reasoning in Ryan does not mitigate the problems we identified with the SFU limited confidentiality consent statement. To begin with, it does not explain what the limit of the law is. It says that a court might "require" disclosure of research records. But it does not assert that the researcher and the university will "do everything possible to keep them confidential," hence our concern that a court, the university and/or a researcher could claim that the statement constitutes a waiver of privilege. If the researcher and the University were to treat the limited confidentiality statement as a waiver, then no doubt a court would, too. The chances are that the subject would not even know that the case was in court, and even if they did, they probably would not show up to assert privilege, as they would reveal their identity in the process.

Our refusal to use the SFU limited confidentiality consent statement without a proviso about the university's and our ethical responsibility to resist court ordered disclosure is consistent with the reasoning in Ryan.

One of the main implications of Ryan is that an undertaking of confidentiality must not only be said to have occurred, but also must be seen to have occurred. If we do not begin with an understanding of privilege, and do not assert it in court, we effectively waive claim to it. Assuming that the University and the researcher were to assert privilege all the way to the Supreme Court if necessary, just as the patient "M" did, the Court would presumably conclude that they had, indeed, done everything possible to maintain confidentiality.

But consider what would have happened had a Simon Fraser University researcher ended up in the same British Columbia court that heard M’s case using the SFU limited confidentiality statement and its caveat emptor approach to research ethics. It is no small irony that M was engaged in her court battle at precisely the same time that the URERC was forcing SFU researchers to use its limited confidentiality consent statement. Had an SFU researcher ended up in that same court using the limited confidentiality consent statement, it appears that the court would have ordered them to disclose the information on the grounds that court-ordered disclosure was anticipated.

Under the caveat emptor philosophy of research ethics, the statement that research material "will be kept confidential to the full extent permitted by law" apparently did not include asserting privilege all the way to the Supreme Court if necessary. In fact, it did not include asserting privilege at all. At this point, had the researcher followed the URERC’s philosophy, exemplified by Dr. Ogloff’s statement that there is no way to protect confidentiality in court, the confidential information would have been divulged.

Because the University and the URERC restricted their definition of "law" to statute, the limited confidentiality statement effectively is a waiver of privilege, and the essence of SFU’s caveat emptor approach to research ethics. "Informed consent" became a mechanism for limiting the University’s liability and was based on the University’s decision that it would go only so far to protect privilege, and apparently that was nowhere at all. Of course, this position, as exemplified by its treatment of Russel Ogden, abrogated the University’s responsibility to defend academic freedom and uphold the highest ethical standards (cf. Blomley and Davis, 1998).

However, if one recognizes that the law includes common law as well as statutes, then the URERC’s promise to maintain confidentiality "to the full extent permitted by law" might take on a different meaning according to the Supreme Court’s reasoning in Ryan. We can only speculate what would have happened if Dr. Parfitt had told her patient that she would "maintain confidentiality to the full extent permitted by law" as opposed to doing "everything possible" to maintain confidentiality. Because "M" asserted privilege all the way to the Supreme Court, her conduct likely would not have supported a finding of waiver, in which case the Court might well have concluded that the mention of the possibility of court ordered disclosure did not constitute a waiver. But without a corresponding effort to actively assert privilege in court, obviously any claim to privilege is doomed.

Perhaps SFU researchers should emulate Dr. Parfitt's pledge to "do everything possible" to maintain confidentiality. Indeed, a consent based on this wording would have several advantages over the SFU limited confidentiality consent statement, including the following:

1. It has met the criterion established by the Supreme Court;
2. Unlike the limited confidentiality consent statement, it cannot be confused as a waiver of privilege; and
3. It commits both the university and the researcher to defend confidentiality at least until there are no more courts of appeal to turn to.

With regard to (c), the notion of doing "everything possible" to maintain confidentiality includes a principled refusal by a researcher to obey even a Supreme Court order to disclose confidential research information. In other words, it recognizes that, "ethics and law may lead to different conclusions" (Tri-Council Policy Statement, p. i.8) and that as a matter of principle researchers should follow ethical rather than legal dictates.

The SFU Research Ethics Policy says that one of the purposes of ethics review is to consider the "cultural values and sense of propriety of the persons who are asked to participate in and/or be the subjects of research." In our first brief to President Blaney (November 17, 1997) regarding problems with the limited confidentiality consent statement, we suggested that it compromised criminological research because of the specific cultural value that prison guards, police officers and career criminals attach to "ratting" (informing on each other).

Notwithstanding what an informed consent statement might communicate from our point of view, an "expectation of confidentiality" is also defined by what research subjects believe a statement says. The primary reasons that courts order the disclosure of confidential information are for the correct disposal of litigation or a defendant’s right to a fair trial. However, when our subjects hear the limited confidentiality consent statement, they are concerned about a different eventuality: that pressure will be put on us to divulge their names so that they can be prosecuted. The URERC’s consent statement is like a large red flashing light above a sign saying, "don’t participate in this research."

Because of their lifestyle and cultural values, our research subjects believe the URERC's statement is not a guarantee of confidentiality of any kind, and they react accordingly ("Why would I talk to you when I know you’ll rat on me?"). Since researcher-participant privilege is partly the subject’s privilege, and because the participant’s understanding of a communication is relevant to the Court’s determination as to whether it originated in confidence, we must take our subject’s definition of the situation into account, as must the URERC, in designing a consent statement. Indeed, the Tri-Council Policy Statement requires researchers and universities to incorporate research participants into its ethics decision-making process.32

We suggest that, because of the specific sub-cultural values of offenders, prison guards and police, the drafting of ethics policies should be careful to follow the Tri-Council’s directive to "accommodate the needs of specialized research disciplines" (Policy Statement, p. i.2) such as criminology.

Another consideration raised by R. v. O'Connor (1995)33 and A.(L.L.) v. B.(A.) (1995)34 is that the adjudication of privilege is intertwined with a consideration of the Charter rights of persons involved in a particular court proceeding. As Justices La Forest, L’Heureux-Dubé and Gonthier of the Supreme Court explained, "Although the justification for privilege has traditionally been utilitarian, a view based on fundamental values, such as privacy, has recently begun to emerge" (A.(L.L.) v. B.(A.), p.4).

The non-utilitarian view of privilege is founded on abstract premises, and particularly on the view that privilege is vital to the protection of fundamental individual values. Confidentiality is thereby viewed as a privacy interest that in and of itself may justify attenuation of the truth-seeking process.

In the case of therapist-client privilege discussed in O'Connor, A.(L.L.) v. B.(A.), and Ryan privacy rights of a victim of sexual assault were pitted against the rights of an accused to receive a fair trial in which all relevant evidence is considered. There is clear division among the Supreme Court Justices on what should be given priority in this particular circumstance. Their lengthy debates with each other make fascinating reading for the many considerations involved. One area of consensus is that the Charter does not prioritize one right over any of the others (see Dagenais v. Canadian Broadcasting Corp. [1994] 3 S.C.R. 835). Any balancing of considerations must be done in the context of the facts of each case and the specific rights that are at odds.

How Charter considerations might play themselves out in the context of the researcher-participant relationship is a matter worth considering. Ironically, the situation in our research on prostitution, where research subjects may tell us about undetected criminal behaviour — which is why the URERC originally insisted that we use its limited confidentiality consent statement — may be protected by Charter considerations. As the various cases dealing with sexual assault make clear, sexual behaviour is a matter of privacy. Prostitution is a legal sexual activity. However, it is highly stigmatized. Does a prostitute have a right to disclose information to a researcher about prostitution, prostitution law, and so forth, with an expectation of privacy?

It is well known that researchers usually try to avoid learning the identities of their research participants, and when they do learn the identities of some of their subjects, as we are likely to in our proposed research on off-street prostitution, they remove all identifiers from their research records. This would be our standard practice in the case of interview transcripts and field notes.

Although we have asked, the URERC still has not provided a justification for its formulaic decision to force only those researchers who collect self-report information about law-breaking activity to use the limited confidentiality consent statement. When we examine cases in the US, it turns out that researchers engaged in this type of research are no more likely than others to be subpoenaed.36 The URERC’s discriminatory treatment of certain types of research has changed the shape of research in the School of Criminology, especially among graduate students who are working against the clock in their degree programs, because the limited confidentiality consent statement makes certain kinds of survey and field research impossible.

Added to all this is the dubious value of research information for a court. In the dissenting opinion in R. v. O'Connor, Justices La Forest, L'Heureux-Dubé, and Gonthier offered the following observations regarding therapists’ records:

We suggest much the same is true of researchers' records, mutatis mutandis, for reasons similar to those outlined in the dissenting opinion. There needs to be more discussion about the probative value of research records, and a legal opinion in this regard may be helpful for the deliberations of the SFU Research Ethics Policy Revision Task Force.

In Informed Consent, Confidentiality and the Law (Palys and Lowman, 1999), we differentiated among three different situations where ethical and legal conclusions might conflict. These include (1) the possibility of being ordered by a court to supply the identity of our research subjects; (2) being exposed in one's research to situations in which mandatory reporting laws are triggered; and (3) situations involving what we called "heinous discovery."

With respect to the prospect of court-ordered disclosure of the identity of a research subject we would be put in the extremely unusual situation where the law is effectively made up after the fact. Researchers have to make choices ahead of time in an occupation in which one’s role is to go where no one has gone before and, in field research, to carve out relationships in good faith in the interest of understanding more about the worlds of other people. The balancing act each researcher must consider at the outset is one that the Supreme Court gets to consider after the fact. We have no choice but to determine ahead of time what commitments to make. As the Chair of the American Sociological Association (ASA) Committee on Professional Ethics (COPE) recently clarified,

We know that it will be impossible to conduct meaningful research in field settings with people involved in prostitution if they think we will simply hand over their identities to a court. Ethics codes prohibit us from engaging in covert activities for police purposes (see e.g. Tri-Council Policy Statement p. 1.12) or acting as informants for authorities (Policy Statement p. 2.4). It would be unethical to engage in research and then turn it over to the authorities so that our subjects could be prosecuted. As to other reasons that a court might order the disclosure of the identities of our research participants, for reasons that we have explained elsewhere,38 we do not see why research on law breaking behaviour should have been singled out by the URERC, when the US experience shows that those engaged in criminological research are no more likely than any other researchers to have their pact of confidentiality challenged.39

Mandatory reporting laws are a second situation that warrants consideration. As the case law introduced by Jones (1999) does not deal with mandatory reporting, and because we consider these laws elsewhere,40 we will not discuss them here.

The third situation — heinous discovery — happens to have been the subject of a case decided recently by the Supreme Court, in Smith v. Jones,41 which involves the legal limits to solicitor-client privilege. The case began when a Vancouver lawyer42 met with his client, who was charged with a sexual assault. The lawyer referred the client (Jones) to a psychiatrist (Smith) for assessment. Solicitor-client privilege is a class privilege developed through the common law. It now enjoys implicit Charter protection. In this case, solicitor-client privilege extended to the psychiatrist-client interaction, since it was part of the preparation of a legal defence, and hence part of the accused's right to a fair trial. Once in the presence of the psychiatrist, the accused referred to the offence for which he had been charged as a "trial run" to see whether he had it in him to kidnap, sexually torture, and then kill a prostitute. Jones also revealed that he had made concrete and extensive preparations to serially murder street prostitutes working in Vancouver’s Downtown Eastside.

The psychiatrist believed that the accused was dangerous and would in all likelihood carry out his plan, and informed the lawyer that he wanted the judge to know this assessment at the sentencing hearing, since the accused had agreed to plead guilty. The lawyer did not reveal the information to the court, presumably believing it to be to the disadvantage of his client. The psychiatrist began litigation against the accused and his lawyer to obtain the court's permission to reveal to the sentencing judge the information he had gathered in confidence under the protection of solicitor-client privilege.

The case was heard at trial and on appeal in secret, with publication bans surrounding the case until the release of the decision by the Supreme Court. The names Smith and Jones are pseudonyms. Southam, Inc. acted as an intervenor at the Supreme Court, arguing the publication ban surrounding the whole affair should be lifted.

The decision of the Supreme Court in this instance is relevant because the Court’s claim that solicitor-client privilege is the, "…highest privilege recognized by the courts. By necessary implication, if a public safety exception applies to solicitor-client privilege, it applies to all classifications of privileges and duties of confidentiality" (p.15).

We do not understand the Supreme Court’s statement that solicitor-client privilege is the "highest" recognized by the courts. In fact, Section 18 of the Statistics Act would appear to give Statistics Canada researchers and research subjects significantly greater protection than solicitors and their clients (see Section 5.v below). Nevertheless, Smith v. Jones is a very important case for understanding judicial limitations on privilege.

In Smith v. Jones, the Supreme Court agreed unanimously that, in this circumstance, public safety considerations outweighed solicitor-client privilege. The 6-3 vote of the court reflected a disagreement among the Justices only as to the conditions that should attach to the disclosure. Given the importance of this case, we quote the reasoning at some length. Writing for the majority, Cory J. stated:

• Under the "clarity" factor, it is important, as a general rule, that a group or person must always be ascertainable. In some situations, great significance might be given to the clear identification of a particular individual or group of intended victims, even if the group of intended victims is large. At the same time, a general threat of death or violence directed to everyone in a city or community, or anyone with whom the person may come into contact, may be too vague to warrant setting aside the privilege. However, if the threatened harm to the members of the public was particularly compelling, extremely serious and imminent, it might well be appropriate to lift the privilege. All the surrounding circumstances will have to be taken into consideration in every case.
• The "seriousness" factor requires that the threat be such that the intended victim is in danger of being killed or of suffering serious bodily harm.
• With respect to the "imminence" factor, the nature of the threat must be such that it creates a sense of urgency. This sense of urgency may be applicable to some time in the future. Depending on the seriousness and clarity of the threat, it will not always be necessary to impose a particular time limit on the risk. It is sufficient if there is a clear and imminent threat of serious bodily harm to an identifiable group, and if this threat is made in such a manner that a sense of urgency is created. If after considering all appropriate factors it is determined that the threat to public safety outweighs the need to preserve solicitor-client privilege, then the privilege must be set aside. When it is, the disclosure should be limited so that it includes only the information necessary to protect public safety.43

All the Justices agreed that the clear, serious and immediate danger to the women warranted a violation of solicitor-client privilege, but they disagreed how far and in what form the information should be distributed. The majority decided that the file containing the psychiatrist's report should be unsealed, that portions "which do not fall within the public safety exception" should remain privileged, and that a ban on the publication of any other parts of it should be lifted. The three dissenting Justices viewed this as too broad a distribution, reminding the court that:

Writing for the minority, Major. J. further explained that the extent of distribution should be more constrained in order to minimize the long-term impact that widespread knowledge of this disclosure might have on the likelihood of future offenders revealing such information, thereby endangering other lives:

We had considered the possibility of heinous discovery in our earlier writings, particularly in what it implied one might insert in one's informed consent statement to subjects. In Informed Consent, Confidentiality and the Law, 44 we stated:

The decision in Smith v. Jones is also interesting insofar as it also mentions a variety of situations in which public safety considerations do not outweigh solicitor-client privilege. For example,

The court also advised that, in instances where public safety considerations warranted violation of a pledge of confidentiality, there are many ways a professional could handle the situation, not all of which involve reporting an individual to criminal justice authorities. Writing for the majority, Cory J. notes:

We would expect researchers in this situation to ensure that confidentiality would be preserved to the greatest extent possible, as long as the safety of the intended target was assured. In this regard, we suggest the decision to violate a confidence because of heinous discovery can be done in a more ethical and sensitive manner than the spectacle Smith v. Jones created, such that the confidence is minimally violated, i.e., only to the extent required to ensure the safety of the innocent third party. We also note that, if faced with this situation in our own research, the solution likely will have nothing to do with a court. The ethical dilemma would be ours to resolve at the time, on the basis of whatever course of action best maintained confidentiality, while ensuring also that an innocent third party is not killed.

In his legal opinion, Paul Jones (1999) explains "class" privilege as follows:

The importance of the duty to give evidence in furthering a court’s search for truth is beyond dispute. Canadian courts have been extremely cautious in restricting the power of an accused to call evidence in his or her defence, a reluctance founded in the fundamental tenet of our judicial system that an innocent person must not be convicted (R. v. Seaboyer, [1991] 2 S.C.R. 577 p. 611).

In certain rare circumstances, the identity of an informer has been disclosed, even though it may jeopardize the informer’s welfare. The informer’s common law class privilege has been set aside if the confidential material can demonstrate the innocence of an accused person,45 as has solicitor/client privilege in very rare circumstances.46

However, to be weighed against the interest of the correct disposal of litigation and the defendant’s right to a fair trial is the compelling policy interest in maintaining conditions in which research subjects can provide information without fear of that information being used against them. Like the duty to maintain an undertaking of confidentiality, this duty to provide information to the courts is not absolute.

While we recognise that the courts are reluctant to grant privilege on a case-by-case basis, and even more reluctant to grant it on a class basis, we think it is important to understand how the Canadian courts might conceptualize the policy issues and the public good when it comes to adjudicating claims to researcher/participant privilege.

Despite our drawing the URERC’s attention to the Supreme Court’s use of the Wigmore test to adjudicate ad hoc claims to privilege, the Committee has never bothered to engage in any kind of analysis of the rationale researchers might invoke to assert researcher/participant privilege. We believe that it was irresponsible to give up the fight before it even started.

In our second submission to the Research Ethics Policy Revision Task Force we tried to lay the foundations for a defence of confidential research information using the Wigmore test. When it comes to the fourth criterion, the adjudication of privilege is a question of public policy. We agree with Judge Steinberg in Ogden v. SFU: the "principles of academic freedom and privilege … are fundamental to the operation of any accredited University."

The Tri-Council considers the primary contribution of research subjects to be advancing the human condition (p. i.7). Given the voluntary nature of subject participation, and its long run collective benefit for all of us, we continue to believe that it would only be in the rarest of circumstances that a court would ever thwart the researcher-participant relationship, which is the bedrock from which that collectively beneficial knowledge arises.

There are many policy reasons to recognise researcher/participant privilege. Ironically, one is that the privileging of researcher/participant relations and protection against court-ordered disclosure is itself a desirable way to further the ends of the search for truth that lies at the heart of all court procedure. One of the main purposes of criminological research, for example, is to further our knowledge of crime and justice, and the discipline as a whole attempts to embrace a broad range of supportive and critical perspectives, thereby maintaining a disciplinary independence that is fundamental to maintaining credibility and integrity. Effectively transforming us into agents of the state, even if only on an ad hoc basis, is not only unethical according to the Tri-Council and most other disciplinary codes (e.g., CSAA, ASA, AAA, APSA), but also self-defeating in terms of the long-term interests of justice.

More and more, researchers are providing expert testimony in both criminal and civil courts. Both of us have provided expert testimony (Lowman on prostitution and Palys on pornography). In some cases the research that provides the basis of this testimony could not have been done if the research subjects thought that their identities would be divulged to a court. Lowman’s interviews with prostitutes, which formed the basis of his testimony in a civil suit, are a case in point. The quality of information we can bring before the courts in these roles is directly affected by the expectation, or not, of confidentiality. The rejection of privilege in these circumstances would involve the court willingly placing itself in an informational Dark Age, by severing the hand that helps maximize the authority and accuracy of its judgements.

At the time of writing, the BC Supreme Court is in the process of hearing what the Vancouver Sun describes as, "one of the most provocative and controversial legal cases in Canadian history."48 The case (R. v. Sharpe), which may well find its way to the Supreme Court of Canada, will determine whether or not the law against possession of "child pornography" is unconstitutional. We found the case interesting because the appellant, Crown and intervenors are all presenting research findings to establish opposing positions about the effects of "child pornography" on the viewer. Depending on which study one reads child pornography either reduces child abuse, increases it, or is unrelated to it! According to the Sun report, the Crown’s factum argues that it is difficult to obtain evidence establishing that child porn leads to child abuse because, "The sexual abuse and exploitation of children is a clandestine activity and those engaged in it are loathe to volunteer as test subjects." The Crown goes on to argue that it, "should not be necessary to wait until a causal link is definitely established by the social scientists before measures to protect children are taken." Regardless of the merits of this as a legal argument, the empirical debate cannot be resolved without reliable and valid research on the self-motivated (as opposed to experimental) viewer of child pornography. Given the clandestine nature of child porn viewing, it is difficult to imagine how one might conduct research on it without a guarantee of unlimited confidentiality.

We suspect that because research plays a key role in cases like this one, the Courts will have good reason to recognize the vital importance of privilege to the researcher/participant relationship and, in turn, to research on crime and justice, should a case end up in court. Because of its connection to the justice system itself, criminology and related disciplines may have one of the stronger cases to make for the recognition of researcher/participant privilege. This leaves us all the more concerned that the URERC has singled out criminological and related research for imposition of its limited confidentiality consent statement.

One of the main points we have made in our discussion of employing the Wigmore test to protect confidential research information is that the researcher/participant relationship is unique. Notwithstanding the many helpful cues that come from the Supreme Court cases cited by Jones (1999) we must avoid a wholesale transportation of the court’s reasoning into the research arena. We need to understand how the concepts used by the court would play themselves out in the research context. Given the differences among the researcher-participant relation and other claims to privilege — it shares some attributes, but is unique — we need more than an abstract analysis of tensions between research ethics and law. Consequently, in the ensuing discussion we try to imagine what kinds of situations might arise.

Most claims to privilege involve an information provider who has something tangible to gain by divulging the information. Charges against the informer are dropped, the patient is counselled, the penitent absolved, etc. Further, it is typically the patient or client who approaches the doctor, therapist, or lawyer for a professional service, and the penitent who approaches the priest.

Research participants are different in both those respects. Our research subjects divulge information in confidence about their own criminal activity (prostitution offences) and sexual activity to a person who has asked them to divulge the information, with the full knowledge they are offering us "data" that will at some point be compiled, analyzed and published. The researcher usually initiates the interaction and, in our experience, the respondent divulges the information only on the condition that they are not named. Since the interaction would not have happened if we had not initiated it, a tremendous ethical burden is placed on us to ensure no adverse effects befall the participant because of our entry into their lives.

It is further noteworthy that research participants typically receive nothing direct or tangible for their participation,49 other than, perhaps, a hope that their voice will be heard. The primary motive for our research subjects is to provide information for the purpose of learning and understanding, often with an eye to enhancing the social good.50 To the extent that research often provides valuable information about and for society, there are compelling policy reasons to protect the integrity of the academic research enterprise.

Universities have a very special responsibility to protect the identity of the persons who, for primarily altruistic reasons provide information that could seriously harm them if their identities are revealed. In these circumstances, we are ethically bound to do everything possible to protect confidentiality in order to maintain the integrity of the researcher-participant relationship. SFU’s caveat emptor approach to research ethics undermines this objective.

When we examine the judicial discourse about privilege, it mostly relates to the production of documents relating to someone whose identity is already known. In one case (R. v. Gruenke), the claim to privilege concerned a conversation. In all the cases dealt with by the Supreme Court of Canada, the Court knows the identity of the person who is the subject of the records or conversation.

The preponderance of discussion about privilege has centred on psychiatric and counselling records, and whether there should be "full" or "partial" disclosure. Challenges to privilege involve the court’s or a defendant’s interest in records, and occasionally conversations, relating to known subjects.

For our research subjects, the issue is not whether our research documents will be ordered disclosed. The purpose of our collecting the information in the first place is to publish it. In criminological research, we are most concerned about a different situation — one in which the information is public and a third party wants to ascertain to whom the record pertains, in order to prosecute them. By saying our information is "public," we mean to say that, as academics, it is not our mission or desire to hide information. Quite the contrary, academic tradition has it that the information gained in research is for the benefit of all. The data we produce, subject to the researcher's claim to priority of publication, are most beneficial when others can scrutinize it, criticize our interpretations and conclusions, and/or engage in secondary analyses.

Of course, our records are usually anonymized. According to academic principles, the records are not of interest because they were produced by a specific person, but because they were produced by someone in a particular role, or with a particular life experience, or who falls into a particular demographic category. Once they are anonymized, anybody who wants them is usually quite welcome to have them. Indeed, we have published reports in which anonymized verbatim transcripts of complete interviews with police officers, Crown attorneys, prostitutes, pimps, and agency workers are right there in the Appendices for anyone to see. Granting agencies, such as SSHRC, usually require that databases become publicly available after a researcher has exhausted his/her priority of publication.

The timing and means of anonymization in our field research is dictated by the types of threats we can envision for our research participants, either because of what they say about their own lives, or because of what they tell us about others. For example, when interviewing a prostitute or a pimp, Lowman anonymizes the interview as it is taking place, e.g., by stopping and rewinding the tape whenever a name is mentioned, deciding jointly on a pseudonym for the particular individual, and using the pseudonym at that point and thereafter. Immediately after transcription, the tapes are destroyed, so that the only record remaining is the anonymized transcript.

Sometimes anonymization is done at the point of transcription, so that no one can ever determine who "Respondent A" is. In a recent study Palys was involved in, in which First Nations people were interviewed about how the community is addressing problems of sexual abuse, this was the method used. The resulting publication does not even name the actual community, in respect of the community's wishes to maintain community confidentiality.51 Indeed, the Supreme Court does much the same thing when it needs to. Smith v. Jones is an example of case analysis in which all important facts and analysis are reported on the Internet for all the world to see, but one would never be able to ascertain from the record the identity of "Smith," "Jones" or the lawyer. Nor is it particularly relevant that we do.

Again, bearing in mind the nature of academic research and the anonymization of records, we would like to know exactly why the URERC is forcing us and not all SFU researchers to mention the risk of court-ordered disclosure of the identity of our research participants. We simply cannot see why the URERC thinks this risk is greater in our research than any other. We do not understand why the URERC has discriminated against criminology and other researchers who collect self-reported information about law violations.

In our review of US cases where attempts were made to obtain confidential research material, we noted that the courts rarely ordered disclosure of the names of research participants. In those very few cases when they did, it was either because confidentiality was not guaranteed, or there was an ambiguous relationship with a research subject (whose identity was already known), or a researcher gave up the fight.52 Generally, then, the US courts have followed a philosophy that the Supreme Court of Canada has referred to as the principle of "partial disclosure." In considering the doctor’s records in Ryan, the Supreme Court distinguished between:

As well as following a principle of partial disclosure, a court can also attach a strict protective order to information that it does order to be disclosed, as did the US Customs Court in Atlantic Sugar Ltd. v the United States.

Our purpose in mentioning these distinctions is to point out that they are probably not relevant to the predicament many researchers face, and are certainly not relevant to our prostitution research. Most of the discussion about court-ordered disclosure of confidential research information relates to the balancing a court is likely to engage in under the fourth criterion of the Wigmore test, i.e. between the "correct disposal of litigation" and the adverse effect on the relation for which privilege is being sought. In Canada, much of this discussion has taken place in relation to doctor/patient or counsellor/patient privilege, and three of the four cases mentioned in the CAUT legal opinion relate to sexual assault.

In each case the identity of the victim and offender is known, the struggle over disclosure relates to written records, and the person to whom the records relate appears in court to assert privilege. It should also be acknowledged that these cases are particularly troublesome when it comes to balancing the interests of the doctor/patient relationship against the correct disposal of litigation or the defendant’s right to a fair trial. In such cases, the person claiming privilege wants to protect information from their adversary; i.e., the person they have accused of an extremely serious crime (e.g., as in R. v. O'Connor) or are suing for damages (e.g., as in M. v. Ryan). Any claim to privilege by a complainant is further compromised if they have already turned the records over to the Crown, as is sometimes the case, because "… it would be difficult to argue that the complainant enjoys an expectation of privacy in records that are held by the Crown" (R. v. O’Connor, p.15).

The requests for medical records arise frequently and are important enough that the Criminal Code has been changed to clarify the ground rules for disclosure and to provide certain protections for confidential medical records.

But, as stated above, we are quite prepared to share our anonymized records. The main possibility that concerns our research participants is that the court might try to find out the identity of a respondent in order to prosecute them. Although the URERC has never explained its formulaic approach to the risk of court-ordered disclosure, apparently this is one reason why the URERC forces only those researchers collecting self-report information about infractions of law to use the limited confidentiality consent statement.53 When our prospective research participants read the URERC’s limited confidentiality statement, many of them are left with the impression that SFU expects researchers to divulge to the state the identities of research subjects who self-report criminal behaviour so that they can be thrown in jail. As a result, the statement represents the kiss of death for our research. And yet, as far as we understand rules of evidence and procedure — but again this is where we would benefit from a properly researched independent legal opinion — a court is unlikely to order disclosure for this purpose.54

A case like M. v. Ryan is instructive because it shows the importance of asserting privilege against court-ordered disclosure all the way to the Supreme Court, if necessary. However, in our research, the equivalent of "M" (i.e., one of our research subjects) would likely never come forward, because one possible scenario in court — like the situation Russel Ogden faced in Coroner's Court — would be over the question of whether a certain individual should be named. The ethical obligation on us to protect the identity of the participant, given that s/he is completely dependent on us to advance and protect his/her interests, is considerable.

A central feature of SFU's policy of limited confidentiality is the assertion that no form of unlimited confidentiality exists in Canada. This view is erroneous. Section 18 of the Statistics Act requires Statistics Canada researchers under oath to make unlimited guarantees of confidentiality and anonymity to their research subjects. Paul Jones (1999) makes reference to this Act in his discussion of existing statutory recognition of privilege, of which only two that are recognized in federal law — communications between spouses (recognized in Section 4.3 of the Canada Evidence Act, although the privilege is subject to qualification) and communications between Statistics Canada researchers and their research subjects. In this regard, Jones (1999) notes, "Section 18 of the Statistics Act provides a limited privilege to certain researcher/ participant relationships," its primary limitation being that it protects only Statistics Canada researchers and their research subjects.

The protection it offers, however, is almost "absolute" (the only limit is for the purpose of a prosecution under the Act), although presumably the Act is open to a Charter challenge. The section states,

We would note that, to the extent it clearly recognizes that unlimited confidentiality is necessary to conduct valid and reliable research, the state has reserved for itself the sole right to gather valid and reliable information.

Thus, when the Supreme Court in Smith v. Jones (1999) suggests that solicitor-client privilege is "the highest privilege recognized by the courts," it appears not to have taken into consideration the power of privilege bestowed by the Statistics Act. Indeed, we find it surprising that the Statistics Act is not mentioned in any of the Supreme Court jurisprudence on privilege.

As it happens, over the past four months, the two of us each have been approached by Statistics Canada to participate in two of their research projects (Palys, as a parent, for the "National Longitudinal Study of Children and Youth," and Lowman for the "Survey of Financial Security"). We have seen firsthand, therefore, as prospective research subjects, the emphasis that Statistics Canada places on ensuring the confidentiality of information, and, implicitly, the extent to which Statistics Canada believes the clarity of that guarantee is a prerequisite for the gathering of valid and reliable data. Under the heading "The law protects what you tell us," prospective subjects of the "Survey of Financial Security," for example, are told:

Why would the state reserve this right of absolute confidentiality for its own researchers if it did not consider confidentiality and anonymity to be vital to research? And on what basis would it deny similar protections for university researchers who, although we co-operate with government and respect state law, can best fulfil our roles as social critics by maintaining our independence from the state, and with an equivalent privilege? The very fact that the legislature recognised that Statistics Canada researchers needed this privilege could be of considerable value for academic researchers asserting privilege via the Wigmore test, and for the development of statutory protection of research done independently of the Federal government.

Research decisions, and the question of relevant law, become more complicated when the researcher occupies more than one role in the life of the research participant (e.g., the professional is a therapist who uses information gained in the therapeutic context for research purposes). It also becomes more complicated when the researcher has dual allegiances to people in the research setting (e.g., the researcher is paid by the managers of a corporation to gather data from workers). The ethics of such scenarios seem to have been a key consideration in the formulation of the Tri-Council Policy Statement, in large part because shifts in the funding environment have made universities more dependent on corporate largesse and the development of research partnerships with the private sector. Section 4 of the Policy Statement, entitled "Conflicts of Interest," begins:

One of the Policy Statement’s requirements is that researchers who hold multiple allegiances need to be clear about their different roles, declare potential conflicts of interest, and ensure they avoid agreements that would compromise the rights and interests of their research subjects.

When the professional occupies dual roles in the life of the subject, the Tri-Council suggests that, "researchers should separate their role as researcher from their roles as therapists, caregivers, teachers, advisors, consultants, supervisors, students or employers and the like" (Policy Statement p. 2.8). It is vital that they treat all their research subjects as research subjects with all the rights and protections that should accrue to them in this role. Although the Tri-Council discusses the ethical underpinnings of this approach, there are also differences among professional activities that have implications for the degree of legal protection that a person enjoys. Our analysis suggests that researcher-participant claims to privilege may well enjoy a more solid legal footing than the therapist-client, or physician-patient relationships. Furthermore, the professional in each of the latter dyads does his or her client/subject a disservice when their research is confused with their service role.

In this respect, we were concerned about Dr. Horvath’s (Chair of the URERC) repeated reference to research subjects as "clients" in our early meetings with the Committee (prior to his assuming the Chair), a term he continued to use despite our protest. A year later in an article offering his views about limited confidentiality, he continued to refer to research subjects as "clients:"56

From our reading of the Supreme Court cases identified by Jones (1999), we wonder whether the court will look differently at "pure" research in comparison to research activities infused with other motives (entrepreneurship, consulting fees, etc.) or when performed by persons wearing more than one professional hat.

In both the O’Connor and Ryan cases the Supreme Court seems to have been influenced by the fact that the sexual abuse victims in each case were not simply persons in therapy whose privacy was being invaded, but that they had other motives mixed into the situation. In O’Connor, for example, the Court noted that, given the victims’ therapy records had helped convince the Crown that laying charges was warranted, it would seem unfair to withhold these same records from the accused, thereby diminishing his ability to make full defence. Similarly, in Ryan, while acknowledging that victims of sexual assault have every right to pursue civil damages, the Court noted that the claim of "privileged" status for M’s therapy records had a self-serving element because granting it would have the additional effect of placing Dr. Ryan at a litigious disadvantage.

A different dynamic was evident in one US case. A sociology Doctoral student (Rik Scarce) at Washington State University at Pullman was engaged in research with animal rights activists. At one point, when he went on an extended family holiday, one of the activists house sat for Scarce. At about the same time, an act of vandalism occurred at a university facility, for which animal rights activists were deemed responsible. Scarce was subpoenaed by a Grand Jury and asked to provide information about one person who was considered a prime suspect in the case — none other than the individual who house sat for him. Scarce claimed researcher/participant privilege, and refused to testify about information gained in confidence as part of his research, but was unsuccessful both at the Grand Jury and on appeal, and ended up spending several months in jail for contempt of court.

The case is an unusual one, as Scarce appears to have worn several hats in the process of writing about animal rights activists. At the time he began his Doctoral research he had already published a "journalistic account" of animal rights activism. The credits in the book included the name of the animal rights activist who was subsequently the focus of the Grand Jury’s attention. In contrast to academic conventions, journalists often name their sources, whereas academics generally do not. Had the link between Scarce and the animal rights activist not already been established in a publication, one wonders if the Grand Jury would have shown the same interest in Scarce’s research. Also, at the heart of both the trial and appeal court’s rejection of Scarce’s claim to privilege is the appearance of a mixed motive — was Scarce claiming privilege because it was rightfully deserved? Or was he hiding behind a claim of privilege in order to protect a friend?57

The lesson we draw from these cases is that to the extent university researchers are engaged in something other than "pure" research, and are perceived to have motives that go beyond strict ethical concern regarding the rights and interests of their subjects, a claim of privilege may be diminished. Researchers engaged in multiple roles with subjects should do everything possible to follow the Tri-Council dictum of creating a clear separation of roles. For example, they should keep separate sets of records that are clearly demarcated as part of the research relationship. They should, if possible, distinguish when data are being gathered in relation to their service role as opposed to their research role. And they should inform research subjects about the existence of any conflict among their roles, and its possible impact on the legal protection of their research participants’ rights.

Related to this point, we note a disturbing tendency among disciplinary associations to amalgamate their codes of ethics across roles. For example, over the past decade the Canadian Psychological Association (CPA), the American Psychological Association (APA), and the American Sociological Association (ASA), have moved from separate ethics codes (e.g., one for psychologists as researchers, another for psychologists as service providers) to single over-arching codes to deal with the multiplicity of roles in which members find themselves. Although there may be similar ethical considerations in different realms — psychologists, for example, have ethical obligations to maintain confidentiality in both their researcher and service roles — the resulting mega-codes create a mix of considerations that are not always compatible, and which may dilute or complicate the legal protection of research subjects.

Once again, therefore, we suggest that all researchers should make as clear a distinction as possible between their "research" lives, and other life domains. In field research — as the Scarce case indicates — the legal danger may arise when the long-term rapport we develop with subjects crosses the role boundary to become also a friendship.

In "medical model" types of research, another danger may be in the long-term affiliations that some researchers develop with their contractors, particularly when the researcher has a direct or indirect commercial interest with that contractor.

Two of our research proposals have now been held up for seventeen and fourteen months respectively because, for ethical and legal reasons, we could not use the URERC’s limited confidentiality consent statement. We proposed several alternatives, but the Committee refused to accept any of them. In October 1998, the Committee sent us a letter telling us that it would not discuss the matter of the consent form any further.58 In December, we submitted our appeal of the Committee’s negative decision.

Before describing the appeal and its outcome, we think that it would be useful to compare the URERC’s limited confidentiality consent statement with the consent statement we proposed in April 1998, which both the URERC and the VP Academic refused to accept.

Here again is the SFU limited confidentiality consent statement:

We cannot use this statement because it:

• can be construed by the courts as a waiver of privilege by the research subject, and hence may place subjects at risk;
• can be considered by the researcher and/or the university as a waiver of privilege, and hence may place subjects at risk (the Tri-Council requires that participation in research shall not be contingent on any waiver of rights by the participants);
• does not unequivocally commit the researcher and the university to do "everything possible" to protect research confidentiality as case law requires, and thus falls short of the unambiguous "expectation of confidentiality" required to pass Criterion #1 of the Wigmore test, and hence places research participants at risk;
• falls short of informed consent because it does not tell subjects the extent to which the University and researcher will resist court-ordered disclosure;
• falls short of informed consent, because it puts research participants in the position of waiving their rights without explicitly informing them that is what they are doing;
• does not protect either research confidentiality or research participants, as R20.01 and now the Policy Statement require.59

In April 1998 we pointed out to the URERC that in order to satisfy the Wigmore criteria and provide the information necessary for informed consent, any warning about court-ordered disclosure should be accompanied by a statement saying that everything possible would be done to prevent it. Also, to facilitate informed consent, researchers should declare whether they intend to obey a court order to disclose confidential research information. In other words, the approach we asked the URERC to accept is the researcher’s equivalent of Dr. Parfitt and M’s approach described in M. v. Ryan, i.e. one in which we assure research participants that we will do "everything possible" to maintain confidentiality. The informed consent statement we presented to the URERC for approval in April 1998 was as follows:

We suggest this consent statement is consistent with existing case law, SFU policy, our disciplinary ethics code and, as it turn out, the Tri-Council Policy Statement because:

• It is designed to provide maximum protection for research subjects;
• It involves using common law (the Wigmore test) to do "everything possible within the law" to protect confidential research information;
• It clearly states the willingness of the researcher and the University to do everything possible to protect confidentiality, and cannot be misinterpreted as a waiver of privilege;
• It provides all the information needed for informed consent, including the University’s and researcher’s actions in response to subpoenas and court orders.

In justifying the lengthy delay of our research the URERC has confirmed that ONLY ONE aspect of our consent statement was problematic.

When it comes to the URERC’s delay processing the two applications, by the Chair’s own account, our negotiations with the Committee faltered in April 1998 when we included information in our informed consent statement about the "University's obligations" to support a faculty researcher in court.60 The URERC informed us they would accept our consent statement if we agreed to delete the following two sentences.

In correspondence with the URERC we quickly arrived at an understanding about the Committee’s responsibilities should a researcher be subpoenaed. Consequently, the only obstacle preventing approval of our applications related to our statement regarding the university’s prospective actions, i.e., that, "Because the University "supports the highest ethical standards" (R.60.01), it too will staunchly defend confidentiality in court." The URERC claimed that "this specific sentence went beyond the policy at SFU as we knew it" (Adam Horvath, Chair of the URERC, e-mail dated March 26, 1999).

Given that we were merely quoting university policy R60.01 (Integrity in Research and Misconduct in Research) in the first part of the sentence, the URERC’s concern is presumably about the second part of the sentence, which says that the University will defend research confidentiality in court.

We do not understand how the URERC could have problems with our statement about the University defending faculty researchers in court. The URERC must have known that the University is obliged to follow the Framework Agreement, which took effect in July 1995. Article 17 of the Agreement states:

If URERC members were unaware of Article 17 when we originally submitted our applications in December 1997 and February 1998, they most certainly should have taken it into consideration after an article by Dr. Clayman, the VP Research, appeared in the July 16, 1998 Simon Fraser News. In our article in that same issue, we had said:

In response to our question, Dr. Clayman stated, "The answer to their question about legal representation is contained in Article 17 of the framework agreement with the Faculty Association," which he then proceeded to quote.62

With Dr. Clayman having reminded SFU's research community of the university’s obligations to faculty members under the 1995 Framework Agreement, it was clear the university was obliged to "defend research confidentiality in court," as our informed consent statement articulated. The only outstanding policy issue after July 16, 1998 was whether the protection provided by Article 17 should be extended to graduate students. That question was answered in November 1998, when the VP-Research/Dean of Graduate Studies announced that it would.

Incredibly, despite our protest, the URERC persisted in trying to force us to use a consent statement that the University could interpret as a waiver of our right to legal representation.63 We do not think the URERC's apparent ignorance of Article 17 is an acceptable justification for infringing our academic freedom by blocking our research for well over a year.

We appealed the decision of the Ethics Committee using the appeal procedure outlined in the SFU ethics policy. Dr. David Gagan, the VP-Academic, heard the appeal.

There is an obvious institutional conflict of interest in Dr. Gagan’s involvement — the same one that led Dr. Clayman to resign eight months earlier from the Chair of the URERC. Dr. Gagan’s hearing of the appeal directly contravenes the Tri-Council Policy Statement’s requirement (section 4.C) that university administrators have nothing to do with the ethics evaluation process. It also violates the principles of natural justice and fairness, given that Dr. Gagan’s intervention involves the university determining whether the university violated its own policies and procedures in the imposition of limited confidentiality and the adjudication of our research ethics applications. Although the Tri-Council Policy Statement technically does not take effect until September 1999,64 we assume that the university follows the principles of procedural fairness and natural justice nonetheless.

Furthermore, the SFU Framework Agreement says that, "should conflicts arise between University policies and the express provisions of this Agreement, the latter shall prevail." The duty to fairness spelled out in Article 6.2 of the Framework Agreement thus supersedes the flawed appeal procedure contained in R20.01. When we first submitted an outline of our appeal to Dr. Gagan, we reminded him of the Tri-Council guidelines and suggested he could mitigate the institutional conflict of interest by appointing an independent panel to hear the appeal. He refused. We then submitted our appeal under protest of the conflict of interest and appearance of bias inherent in the appeal procedure.

Our documentation to Dr. Gagan included a point-form summary of documented policy and procedural violations by the VP-Research and the URERC, and our first two submissions to the Research Ethics Policy Revision Task Force — The History of Limited Confidentiality, and Informed Consent, Confidentiality and the Law. Dr. Gagan’s decision failed to address the many allegations of administrative violations of policy and procedure, and opted instead to articulate the criteria by which the URERC would generate a new informed consent statement for our research. Then he went beyond the jurisdiction of an appeal, and retained the right to censor any consent statement the URERC might propose.

Given that its members have previously insisted the URERC is independent from the administration65 we were curious how the Committee would respond to the Appeal Decision. Would it remind Dr. Gagan that his retaining the right to censor the Committee violates the Tri-Council Policy Statement’s requirements for administrative independence? Would it resist the more general violation of the principles of natural justice and procedural fairness involved in the Appeal Decision? Or would it ignore the principle of fairness guaranteed by the SFU Framework Agreement and comply with this unfair procedure?

In his Appeal Decision, Dr. Gagan invited the URERC to submit a new consent statement based on the following criteria:

• the researchers will use their best efforts to maintain the confidentiality of their data and its sources;
• the University is committed to the promotion and protection of the academic freedom of its researchers and will therefore use its best efforts to maintain the confidentiality of these data and their sources;
• the foregoing notwithstanding, the University can not and will not counsel its researchers to disobey the law, or support them in doing so;
• hence, there may be limits beyond which the confidentiality of personal information and its sources cannot be guaranteed.

On March 29, 1999, the URERC, "formulated a consent statement incorporating [Dr. Gagan’s] directives with respect to the consent document" which read as follows:

This statement informs subjects that researchers may make additional commitments to protect confidential research information thereby implying that they could resolve to disobey court-ordered disclosure. However, having reserved the right to censor the URERC's statement, Dr. Gagan did just that. Specifically, he removed all reference to the promises a researcher might make over and above the protection possible within law. Also, he expanded the section declaring the University’s doctrinaire subordination of research ethics to law, and gave us the choice between not doing our research and using the following consent statement:

Dr. Gagan provided us with no opportunity to comment on his statement, which we cannot use because:

With Dr. Gagan’s statement, the University has declared that it subscribes to a doctrine that absolutely subjugates research ethics to law.

We can envisage very few situations where an academic would disobey the law in the process of doing research. Nevertheless, as our own disciplinary code of ethics (see, for example, the American Sociological Association Code) and the Tri-Council Policy Statement (p. i.8) recognise, law and ethics may lead to different conclusions. Furthermore, while the Tri-Council Policy Statement is explicit that, "researchers must comply with all applicable legislative requirements" relating to the "competence" of a prospective subject to give informed consent (p. 2.9), and requires that tissue researchers, "must be aware of, and conform to, the specific requirements of applicable law" (p. 10.2), nowhere does it require that researchers obey a court order to disclose confidential research information.

Dr. Gagan’s Appeal Decision violates Article 1.2(c) of the Framework Agreement, which ensures "freedom from institutional censorship." By reserving the right to censor any consent statement the URERC might propose, Dr. Gagan went beyond the jurisdiction of an "appeal." On May 6, 1998 the URERC formally approved the clause in our consent statement saying that, "In the event that … we are ordered to reveal confidential information by a court or public body, we will make an ethical decision not to do so." This statement is consistent with our disciplinary ethics code, and Dr. Gagan knows that the URERC had already formally approved it. Also he knows that, since November 1998, the URERC already approved at least two ethics applications employing consent statements containing this clause (both projects involve our students).

To approve our ethics application, Dr. Gagan requires that we include a statement that,

Stopping us from doing our research if we do not mention the University’s doctrine infringes our academic freedom, and is thus a violation of Article 1.2 of the Framework Agreement, which affirms our "right to investigate … without reference to prescribed doctrine." Also, by censoring information about our intentions should a court order disclosure of confidential research information, Dr. Gagan is preventing us from providing information that is necessary for properly informed consent.

The purpose of codes of ethics is to protect research participants, whether maintaining that promise is comfortable or not. If SFU is to live up to the Tri-Council Policy Statement and its own policies regarding ethical standards, the integrity of the research enterprise and academic freedom, it has no choice but to defend confidential research material all the way to the Supreme Court of Canada, if necessary.

Of course, this still leaves the problem of what a researcher should do if the Supreme Court of Canada orders us to divulge the names of our research participants. Because a promise that may or may not be kept is no promise at all, in the interests of informed consent, we must inform our research subjects if we will obey a court order to reveal their names. It was presumably for this reason that on May 5, 1998, the URERC formally accepted a line in our proposed consent statement explaining to research participants that we would not obey a court order to disclose their names. However, Dr. Gagan has censored this component of the consent statement.

Once again it appears that the University is using the ethics approval process to minimize the University’s liability and protect its image. Dr. Gagan acted on the institutional conflict of interest when he censored previously approved elements of our informed consent statement, which the URERC has allowed other researchers to use. Given that this was a research ethics appeal, his sole concern should have been protection of the research subject. However, it appears to us that his "Law of the Land" consent statement has more to do with transmitting a particular image of the University and protecting it from liability than it does with research ethics.

In the process of struggling to overcome the URERC’s and University’s infringement of our academic freedom, we have examined a variety of social science ethics codes to ascertain the position they take on court-ordered disclosure of confidential research information. We have yet to discover a code that requires researchers to submit to court-ordered disclosure.

Most codes do not actually mention the problem of court ordered disclosure. One code that does is that of the American Sociological Association (ASA). In our November 17, 1997 brief71 to President Blaney outlining the problems with the URERC’s limited confidentiality consent statement, we cited this code as relevant to the kind of research we had proposed, because the 1990 version explicitly stated that:

When we examined the 1997 version of the code we realized that the wording of this section had changed somewhat, and so we wrote to ASA asking for clarification. Our query and the response of the current and two former Chairs of the ASA Committee on Professional Ethics (COPE) was published in the February 1999 issue of Footnotes, ASA’s monthly newsletter. Upon reading this exchange, Dr. Ellen Gee, Chair of the SFU Research Ethics Policy Revision Task Force (the "Ethics Task Force") contacted the current Chair of COPE to seek clarification on some of the issues raised in the initial query/response. Their interchange is worth quoting at some length,72 because it establishes that a pledge of unlimited confidentiality can be made ethically:

In keeping with the ASA code, we have considered the relevant law, we are prepared to offer unlimited guarantees of confidentiality to our research subjects, and we are prepared to live with the consequences of so doing.

Given the requirement of the SFU ethics policy and the Tri-Council Policy Statement to respect disciplinary standards we protest Dr. Gagan’s requirement that we can proceed with our research only if we agree to remove references to our ethical stance and mention the University’s "prescribed doctrine" subjugating ethics to law.

Dr. Gagan’s consent statement misleads the subject by saying that information "may be disclosed." This is not true: we do not intend to disclose our subject’s identities. While the statement implies that this information may not be disclosed, for fully informed consent we must explain when/how it will not be disclosed. The resulting consent statement does not comply with ASA’s ethical standards, which require one to make a guarantee and live with it, and fully inform subjects in advance of exactly what that guarantee is.

If Research Ethics Boards force researchers to warn prospective research subjects of the minuscule risk of a court order to disclose confidential research information, considerations of informed consent oblige us to also tell them what we intend to do should we receive such an order. In the process of trying to come to grips with this dilemma, the University has taken the position that it will not, under any circumstance or in any way, take a position that could be interpreted as supporting a researcher’s decision to defy a court order. We find the idea that the university would never "support" a researcher troubling, to say the least. This a priori subjugation of ethics to law flies in the face of a centuries old tradition of civil disobedience on grounds of conscience and principle. The tradition begins with Socrates and continued with the likes of Henry David Thoreau, Mahatma Ghandi, and Martin Luther King.

With a sense of considerable irony, we would note that some of the Supreme Court justices have acknowledged that some individuals may choose to follow ethical dictates rather than obey a Supreme Court order. They clearly recognise the powerful statement that such disobedience expresses. In their minority opinion in R. v. Gruenke (1991), in which the issue of priest-penitent privilege was addressed, Justices L’Heureux-Dubé and Gonthier stated:

We believe that the secular obligation of confidentiality in the research enterprise is every bit as important as the sacred obligation is to the priest. Indeed, Statistics Canada reflects this reverence of confidentiality by requiring its researchers to take an oath of secrecy. The Statistics Act specifically forbids the courts from requiring Statistics Canada researchers to give "oral testimony or to produce any return, document or record with respect to any information obtained in the course of administering this Act" (Section 18(2)). Ethically, as far as we are concerned, university researchers do not owe their subjects any less than this guarantee. This does place us in a more difficult situation — we have the same ethical obligations as Statistics Canada researchers, but do not enjoy the same statutory protection — and therefore conclude that, until similar statutory protection is available for university researchers, the only ethical way to fulfil these obligations would be to disobey a court order. In contrast, by subjugating ethics to law, the University has devalued the importance of confidentiality to the research enterprise, thereby compromising the legal defence of confidential research information and diminishing academic freedom.

The University’s justification of limited confidentiality is based on the premise that any refusal to obey a court order puts that researcher "above the law," thereby undermining the entire Rule of Law. We take a different view. The researcher who, for reasons of principle or conscience, defies a court order is not "above the law" because, as the Tri-Council recognises, "ethical principles cannot pre-empt the application of law" (Policy Statement, p. i.8). The researcher who defies a court order can expect to be sentenced to up to two years in prison under Canadian Criminal Code section 127.(1) ("Disobeying order of court"). Furthermore, the literature on the "obligation to disobey" recognises that the principled refusal to disobey a law does not represent a challenge to the entire legal system or democracy itself.74 Indeed, some classical political philosophers75 recognised that civil disobedience — the principled refusal to obey an unjust law — is a vital component of democracy. In this regard, it is important to distinguish revolutionary groups, whose aim is to bring down the entire legal system, from special interest groups who, through civil disobedience, desire to make larger society recognise the primacy in some area of social life of a particular ethical or moral principle over a particular law — or, in our case, the judicial application of common law relating to privilege. As Michael Walzer76 points out:

This same confusion apparently underlies Simon Fraser University’s a priori subjugation of research ethics to law. We believe this position potentially undermines the very foundation of the University: academic freedom. As academic researchers, we believe that ethical criteria, not law, must be the ultimate arbiter of research decisions. A university that does not take this position weakens its ability to protect academic freedom from the state.

Given the Tri-Council acknowledges that ethics and law may lead to different conclusions (Policy Statement, p. i.8), we do not understand how any academic, university administrator or not, would impose a priori the view that universities must always subvert principles of ethics to the rule of state law. If at some level the University does not retain its capacity to assert its independence from the state, it compromises its very raison d’être.

Instead of absolutely subjugating ethics to law, we suggest that the University should respond to acts of civil disobedience by its members on a case-by-case basis. It is relatively easy for us to envision a situation where taking the "ethical" rather than the "legal" option would be the only one our conscience would permit, and one we hope the university would support in principle. For example, what if the Supreme Court ruled that confidentiality was not essential to the researcher-participant relationship (Wigmore criterion 2), that the conduct of research was not particularly important to Canadian society, and that the researcher-participant relation did not need to be sedulously fostered (Wigmore criterion 3)? As an act of civil disobedience, we would disobey an order to divulge the identity of one of our research subjects, and accept the consequences. We would hope that the University would offer words of support in this situation.

In a more general sense, we believe that a promise to maintain confidentiality is fundamental to our research on prostitution if that research is to have any validity and integrity. It would be unethical for us to imperil the liberty of persons who volunteer to give us access to their lives for research purposes. Because the law regarding privilege operates on a case-by-case basis, it is effectively made up after the fact. In a sense, it thus contradicts one of the fundamental principles of natural justice, i.e., that rules be made known in advance. When we are doing research, we know full well that a promise that may or may not be kept is no promise at all, particularly for research subjects who are concerned that the information they give us may be used to prosecute them. On the basis of our belief that research satisfies criteria 2 and 3 of the Wigmore test, if we are to undertake it at all we have to make an undertaking of unlimited confidentiality and stick to it. The main exception to this would be cases where we discover prospective serious harm, such as intent to commit murder or the imprisonment of an innocent person, in which case we would be prepared to violate a guarantee of confidentiality because a higher ethic takes precedence.

In the fall of 1997, when we first asked Dr. Blaney to reconsider the university’s decision-making in the Ogden case, he delegated it to Dr. Clayman, the VP-Research, who had been a part of the university administration that made the decision to abandon Russel Ogden’s research participants in court. When Dr. Clayman was asked whether he and his colleagues made a proper decision, he said "yes." His decision to oblige the President’s request violated a principle of natural justice, i.e., that no one should sit as a judge in his/her own cause. Eventually, Dr. Blaney convened an independent panel (Drs. Blomley and Davis) to review the case. The panel reached quite the opposite conclusion from Dr. Clayman, judging the university’s behaviour to be unethical. They recommended that the university apologize to Ogden and rectify its wrongdoing.

In the fall of 1997 we alerted President Blaney to the URERC’s unauthorized substantive change to the SFU Research Ethics Policy and the various legal and ethical problems with the URERC’s limited confidentiality consent statement. However, rather than establishing an independent body to review our complaints, the President referred the matter back to the URERC for consideration. In so doing, he put the URERC in the position of acting as a judge in its own cause, thus again violating a core principle of procedural fairness and natural justice.

When we appealed the URERC’s October 1998 decision to deny approval of our research ethics applications, we pointed out the institutional conflict of interest in a university administrator hearing the appeal. Once the Tri-Council Policy Statement takes effect in September 1999, such practices will no longer be allowed. Because the SFU Framework Agreement bestows a duty of fairness on administrators, and because the Framework Agreement supersedes university policy, we remain amazed that the VP Academic proceeded to hear the appeal. We are now in the process of grieving his decision, and are faced with the prospect of invoking a section of the Code of Faculty Ethics and Responsibilities (SFU Policy A30.01) which authorizes faculty to ignore university policies that infringe their academic freedom.

We can only hope that the SFU Research Ethics Policy Revision Task Force will produce a new policy that will enable researchers to protect confidential research information to the extent possible within law and beyond, and that it will protect academic freedom in the process. If it succeeds in this respect, it will produce a policy that other Canadian universities would do well to emulate.

Because the courts have always taken a very conservative approach to recognizing privilege, researchers need to carefully design their research protocols with an eye to satisfying the Wigmore criteria. This is the core philosophy of the approach that we recommend be adopted in the new SFU research ethics policy77 and at all other Canadian universities until such time as legislators are convinced that researcher-participant communications are worthy of statutory protection.

In the conclusion of his legal opinion, Paul Jones notes that:

We encourage the Simon Fraser Faculty Association and Simon Fraser University to take up CAUT’s offer and would welcome an opportunity to assist in this endeavour. There are many reasons why the development of such a privilege is merited.

The defenders of limited confidentiality have insisted that it is not possible to give a guarantee of unlimited confidentiality under Canadian law. That is not correct. Section 18 of the Statistics Act is specifically designed to allow guarantees of unlimited anonymity. When asked to participate in research, prospective research subjects of Statistics Canada's "Survey of Financial Security" are told:

What better example can one find of the necessity of anonymity for valid and reliable research? Academic research requires no less. Such a protection would go far in easing the tensions between researchers and university administrators that have surfaced at SFU over the past five years. What presumably will stop once the Tri-Council Policy Statement comes into full effect is the (ab)use of ethics procedures by university administrators for purposes that go beyond ethics. Ethics policies are designed to protect the rights and interests of research participants. They are fundamental to the long-term maintenance of the integrity and viability of the research enterprise, and hence of the viability of the university itself.

February 15, 1999

David Bell

Executive Director

The Faculty Association of Simon Fraser University

8888 University Drive

Burnaby, B.C.

Dear Mr. Bell:

You have requested a legal opinion on researcher/participant privilege. The specific questions you have posed are as follows:

1. Are there alternatives to the Wigmore test that could be advanced to establish researcher/participant privilege?

All relevant evidence is admissible in a legal proceeding unless it is excluded by a specific exception. Privilege, the legal protection that shields communications within certain relationships from disclosure, is one such exception.

The common law recognizes privilege on a "class" and "case-by-case" basis. The most familiar example of a class privilege is the one extended to the solicitor-client relationship.

Communications within a relationship enjoying class privilege are presumed inadmissable as evidence in a legal process. In order for a class privilege to be recognized, compelling policy reasons must be demonstrated and the relationship at issue must be inextricably linked with the justice system.

In contrast case-by-case privilege is established through the application of the "Wigmore Test". This test, created by the American professor of law John Henry Wigmore, posits that a communication is privileged if the following conditions are met:

1. The communication must originate in a confidence that it will not be disclosed;

1. The injury that would inure to the relationship by the disclosure of the communication must be greater then the benefit thereby gained for the correct disposal of the litigation.

John Henry Wigmore, Evidence in Trials at

Common Law, vol. 8 (McNaughton rev.) (Boston:

Little, Brown and Company, 1961) at page 527

The Supreme Court of Canada has accepted the validity of the Wigmore principles. The Court enunciated the test in Slavutych v. Baker [1976] 1 S.C.R. 254 and formally adopted it in R. v. Fosty [1991] 3 S.C.R- 263 (sub non R. v- Gruenke).

However, the Supreme Court indeed all Courts, have adopted a conservative approach in applying the Wigmore principles. The accepted judicial view is that the search for truth is best accomplished through the admission of all relevant evidence, and that any departure from this rule is an extraordinary exception. In reference to such exceptions, Chief Justice Burger of the United States Supreme Court stated in United States v Nixon, 94 S.Ct. 3090 (1974) at 3108 that:

Wigmore himself, in Evidence in Trials at Common Law, vol. 8 (McNaughton rev.) (Boston: Little Brown and Company, 1961) stated at page 70:

The leading Canadian cases on privilege are R. v. Gruenke, P. v. O'Connor (1995) 44 C.R. (4^th) 1 (S.C.C.), A. (L.L.) V. B. (A.) (1995) 44 C.R. (4^th) 91 (S.C.C.) and A.M. v. Ryan [1997] 1 S.C.R 157.

In Gruenke, the Court dealt with an individual accused of murder who sought to protect from disclosure an admission of guilt she had made to a religious advisor. The accused argued that such "priest-penitent" communications were privileged.

The Court held that the policy reasons supporting a class privilege for religious communications must be as compelling as the reasons underlying the class privilege for solicitor-client communications, namely that the relationship and communications are essential to the effective operation of the legal system. The Court found that religious communications, notwithstanding their social importance, are not inextricably linked with the justice system and thus rejected the argument that they should enjoy class privilege.

While the Court held that there should be no aromatic presumption that religious communications are protected, it did find that it would be appropriate to apply the Wigmore criteria on a case-by-case basis to determine if a particular religious communication should be privileged. In her minority decision, Madame Justice L'Heureux-Dubé took a stronger position and concluded that a new class of privilege should be recognized for religious communications.

On the facts of the case, the Court found that there was no expectation of confidentiality connected with Ms Gruenke's confession to her pastor. As an expectation of confidentiality is absolutely critical to establish privilege, neither the majority nor minority judgements found that the communications at issue were protected, either on a class or case-by-case basis.

The Gruenke decision stands for the proposition that the Wigmore criteria can appropriately be utilized to determine on a case-by-case basis if privilege should be extended to communications within a particular relationship. However, Gruenke is not the last word from the Court on the issue. In the more recent companion decisions R. v. O'Connor and A. (L. L.) V. B. (A) the Court dealt with the issue of whether privilege should extend to sexual assault counseling communications, that is, whether a sexual assault victim's private therapeutic records should be protected from disclosure to the person accused of the assault.

Discussion of privilege in these cases was left largely to Madame Justice L'Heureux-Dubé in her dissent. In contrast to her approach to religious communications in Gruenke, she declined to extend a class privilege to sexual assault counseling communications. Moreover, she also retreated from the Gruenke position on case-by-case privilege. Mirroring the majority decision of Chief Justice Lamer and Sopinka J., she proposed that instead of the application of the Wigmore test, the necessity of disclosure could better be determined on a case-by-case balancing of the Charter values of privacy and the right to make full answer and defence. While L'Heureux-Dubé stated that an approach based on the Wigmore criteria was possible, she indicated at page 122 of A.(L.L) "I do not think that such exceptions to the general evidentiary rule of admissibility and disclosure should be encouraged".

Two years later, in A,M. v. Ryan, a case involving the production of sexual assault counseling records in the context of the disclosure process of a civil trial, the Court provided further clarification of its position on privilege. At paragraph 20 of Ryan, the Court restated the appropriateness of the Wigmore criteria to determine if a privilege exists. The Court further affirmed that privilege could be expanded to include new categories:

At paragraph 30 the Court continued its discussion from O'Connor and A. (L.L.) on the balancing of Charter rights, incorporating this balancing process into the fourth Wigmore criterion:

In addition to importing the notion of balancing Charter values, the Court also proposed a more flexible application of the Wigmore principles. At paragraph 3 the Court stated:

Where then does Canadian law stand on the issue of the application of the Wigmore principles?

An examination of the case law from Gruenke to Ryan indicates that the Supreme Court has accepted the Wigmore test as the means to determine if privilege should be extended on a case-case basis to a relationship. Moreover, these cases, Ryan in particular, emphasize that the categories of privilege are not closed and that the identification of new relationships that meet the Wigmore criteria is possible.

Thus, at a theoretical level, the establishment of researcher/participant privilege through the Wigmore test is conceivable. However, privilege jurisprudence is filled with admonitions about the caution with which any exceptions to the basic rule of admissibility must be applied. The Court's recent statements about "partial privilege" and balancing Charter rights further indicates the extent to which this is a complex and evolving area of law calling for a nuanced, rather than mechanistic, understanding. As such, it is not possible to predict with certainty the eventual recognition of researcher/participant privilege or the exact form it may take.

However, the decisions of the Honourable L.W. Campbell Coroner, in Inquest of Unknown Female (October 20, 1994 - 91-240-0838) and of the US Court of Appeals in Cusumano and Yoffe, in which researcher/participant privilege was recognized, are important signals. While neither decision would be binding on a Canadian Court their thoughtful consideration of the researcher/participant relationship carry considerable persuasive value. With respect to the Coroner's decision in particular, it is important to remember that advancements in the law, although eventually affirmed by the highest Courts, often commence at the most basic levels of the adjudicative process. Further, the Courts, in contrast to the situation even a decade ago increasingly utilize as evidence the very type of social science that would be protected and encouraged by the re-cognition of researcher/participant privilege. Such judicial familiarity would support arguments in favour of the expansion of privilege to capture the researcher/participant relationship.

My understanding is that the current version of the document entitled "Informed Consent by Subjects to Participate in a Research Project or Experiment" contains the following proviso:

Confidentiality is the foundation of the Wigmore criteria. In Gruenke, the Court stated that an expectation of confidentiality was absolutely critical to establish a claim for privilege. Indeed, it was the very lack of a promise of confidentiality in Gruenke that prevented the Court from privileging the religious communications at issue,

However, confidentiality is never absolute, even in the most established of class privileges. It is a fact that a party in any privileged relationship may be ordered to disclose information. What then is the impact if the parties, before they commence the exchange of confidential information, are aware that disclosure may be ordered? Does this negate any subsequent claim for privilege?

In Ryan, the patient informed her psychiatrist that all their communications must remain confidential. The doctor indicated to the patient the possibility of court-ordered disclosure, but assured the patient that everything possible would be done to ensure that their discussions would remain confidential.

This recognition by the parties that confidential communications could be ordered disclosed was held to be fatal to a claim for privilege at the lowest level of judicial consideration. The issue was appealed, eventually to the Supreme Court of Canada. In its decision, McLachlin J., writing for the majority, stated at page 173:

With this statement the Supreme Court clearly indicates that exposure to the knowledge communications may be subject to disclosure does not negate a subsequent claim for privilege. While I have concerns about the particular language of the "Informed Consent" document, it is my opinion that a warning of the potential of disclosure, if properly worded and situated in the context of meaningful commitments to confidentiality, will not automatically undermine the application of the Wigmore criteria.

In Ryan, the Supreme Court made the following observations about the doctor/patient communications at issue:

Notwithstanding the fact that both the patient and the doctor were aware that court ordered disclosure was a possibility, the Court still found that the expectation of confidentiality was high enough to support the application of the Wigmore criteria. In my view, the importation into the consent document of language that assured the research participant that:

would override the effect of the warning of potential disclosure and ensure the applicability of the Wigmore criteria.

In addition to common law privilege (class or case-by-case), certain forms of privilege have also been created by statute.

In respect to the criminal law, communications between spouses are granted privilege in s. 4('3) of the Canada Evidence Act. In civil proceedings provincial legislation generally provides for spousal communication privilege. Quebec and Newfoundland have enacted statutes recognizing religious communication privilege and Section 18 of the Statistics Act provides a limited privilege to certain researcher/participant relationships. The Criminal Code was amended, in response to the O'Connor case, to provide a partial shield for sexual assault counseling communications. Finally, in a variety of U.S. jurisdictions specific statutory privilege has been created to cover the researcher/participant relationship.

The creation in Canada of statutory privilege for the researcher/participant relationship represents an alternative to the Wigmore test as a means of ensuring the confidentiality of researcher/ participant communications. However, as the passage of such legislation is as much a political as legal issue, detailed discussion of the creation of such privilege is beyond the scope of this opinion. If the Faculty Association is interested in pursuing this question, CAUT would certainly be open to requests for support and advice.

I trust this information is of assistance. If you would like to discuss this matter further, please do not hesitate to contact me.

Yours truly,

Paul Jones

Legal Counsel

Canadian Association of University Teachers

We are writing to seek clarification of the confidentiality provisions of the 1997 ASA Code of Ethics (Section 11) in view of how changes from the 1989 version may affect our ability to offer guarantees of unlimited confidentiality to research participants, and defend them by resorting to the so-called "Wigmore criteria," a common law test that can be used to establish researcher-participant privilege on a case-by-case basis. These criteria were successfully invoked by Russel Ogden, the only researcher in Canada ever to be faced with a charge of contempt of court for refusing to divulge confidential information to a court (Lowman & Palys, 1998).

The Wigmore criteria require that:

1. The communications must originate in a confidence that they will not be disclosed;
2. This element of confidentiality must be essential to the full and satisfactory maintenance of the relation between the parties;
3. The relation must be one which in the opinion of the community ought to be sedulously fostered; and
4. The injury that would inure to the relation by the disclosure of the communications must be greater than the benefit thereby gained for the correct disposal of litigation. (Wigmore, 1905, p.3185)

Two changes in the 1997 Code may adversely affect use of these criteria:

1. The addition of admonitions to researchers to consider legal "limitations" to confidentiality in Section 11.02(a); and
2. The removal of a statement in the 1989 principles (section E5) that "Confidential information provided by research participants must be treated as such by sociologists, even when this information enjoys no legal protection or privilege and legal force is applied." The 1997 Code (Section 11.01(b)) replaces this with:

Any apparent acceptance of externally imposed a priori limitations to confidentiality may undermine the first criterion of the Wigmore test. We are thus concerned a court could interpret the recognition of "limitations" to confidentiality, combined with deletion of the phrase "and legal force is applied," as evidence the ASA had softened its commitment to research confidentiality by subjugating research ethics to law. The statement in the 1989 Principles about maintaining confidentiality even when "legal force is applied," was valuable precisely because it represented unambiguous evidence that confidentiality is crucial to the research enterprise, and worth fighting for (Wigmore criterion 2). Also, it clearly recognized a researcher's freedom to pledge unlimited confidentiality to research participants (criterion 1), which legal opinion in both Canada (Marshall, 1992) and the United States (Traynor, 1996) suggests are essential to the legal defense of confidentiality agreements.

We wonder whether the changes noted above arose because the architects of the new Code felt the "legal force" clause was redundant, or a shift in the ethical philosophy underlying it. Because these issues could be crucial for researchers who find themselves subpoenaed, we request clarification about how to interpret the newer Code of Ethics with regard to confidentiality and its limits.

Lowman, J., and Palys, T.S., The liability of ethics. CAUT Bulletin, 1998, June, p.28.

Marshall, M.A. When is a secret not a secret? http://www.cookdukecox/newsletters/issue6-1992/secret.htm, 1992.

Traynor, M. Countering the excessive subpoena for scholarly research. Law And Contemporary Problems, 1996, 59(3), 119-148.

The query from John Lowman and Ted Palys concerning confidentiality and its treatment in the new ASA Code of Ethics is an important one. One's ability to pledge confidentiality to research participants is a cornerstone of research involving human subjects. Further, a researcher's obligation to maintain the confidentiality of information lies at the heart of the ethical principles of science and research. Thus, the changes in the ASA Code of Ethics regarding the limitations to confidentiality were crafted with the utmost care and reflection. By clarifying the principle, the goal was to strengthen, not to erode it.

The changes in the Code were not intended to undermine a researcher's commitment to confidentiality or to subjugate research ethics to law. In the 1997 Code, there are now seven clauses in Section 11.01 specifically addressed to the significance and scope of Maintaining Confidentiality. As Section 11.01(b) states, "Confidential information provided by research participants, students, employees, clients, or others is treated as such by sociologists even if there is no legal protection or privilege to do so." Although the phrase "even when … legal force is applied" has been eliminated in the 1997 Code, this was not to imply that sociologists should divulge confidential information when pressure from the courts or legal sanction is brought to bear.

However, the new Code was written in recognition of the realities of ethical decision making specifically with respect to human protection. Researchers may be confronted with unanticipated circumstances that are clearly health- or life-threatening to research participants, students, employees, clients, or others. When such circumstances arise, it may require rethinking the guarantees of confidentiality, regardless of the law and/or application of legal force to divulge information.

For example, if a researcher discovers child abuse while gathering ethnographic data about family life, what should the researcher do? The underlying intent of the ethical principles applied to research with human subjects is to protect the subjects from harm or risk. In this situation, the researcher faces a choice. Report the child abuse with a potential breach of confidentiality promised to the parent or maintain the confidentiality of the information and risk future harm to the child. The recognition of such a narrow limit to confidentiality is expressed in Section 11.02(b) of the Code: "Sociologists may confront unanticipated circumstances where they become aware of information that is clearly health- or life-threatening to research participants. … In those cases, sociologists balance the importance of guarantees of confidentiality with other principles in the Code of Ethics, standards of conduct, and applicable law."

Further, consistent with the ethical standard enunciated in Section 11.02(b), the 1997 Code in Section 11.02(a) cautions sociologists before commencing research to be fully informed about all laws or rules which may limit or alter guarantees of confidentiality. In such situations where laws and/or rules exist (and hence can be anticipated), it is incumbent upon researchers to determine their ability to guarantee absolute confidentiality. Given this reality, the Code of Ethics also includes a standard whereby researchers are obliged to discuss confidentiality and its limits with research subjects. Section 11.03(a) states: "When sociologists establish a scientific or professional relationship with persons, they discuss (1) the relevant limitations on confidentiality, and (2) the foreseeable uses of the information generated through their professional work. These are critical sections of the 1997 Code, since it is now incumbent upon a researcher to be fully informed about all laws and rules that may limit or alter guarantees of confidentiality. Further, researchers may not promise absolute confidentiality if their ability to do so is limited or if they are unwilling to maintain confidentiality in these situations.

Although limitations to confidentiality under health- or life-threatening circumstances may be the exception, the Committee on Professional Ethics (COPE) recognized the importance of writing a Code that sociologists would not violate when a decision is made to divulge confidential information under such circumstances. Further, the 1997 Code places greater responsibility on sociologists to be aware of these limitations and to discuss them with research participants, students, employees, clients or others at the outset of their research.

If you have any further questions about this or other sections of the 1997 Code of Ethics please contact Joyce Iutcovich at iutcovic001@gannon.edu or (814) 453-4713.

Dr. Ellen Gee, Chair of the SFU Research Ethics Policy Revision Task Force — which is charged with developing a new ethics policy for SFU that is harmonious with the Tri-Council Policy Statement on ethics — read the Footnotes query and response, and sought further clarification. She e-mailed the current Chair of the ASA Committee on Professional Ethics (COPE), asking for clarification on several issues. COPE's response, which intermingles Dr. Gee's queries with their response, is shown below:

From: Ellen Gee <gee@sfu.ca

To: ethics-taskforce@sfu.ca

Cc: palys@sfu.ca; lowman@sfu.ca

Subject: Response to Ethics Query

Date: Tuesday, March 23, 1999 10:02 AM

...

Below is the response that I received back from the American Sociological Association.

...

Ellen

Return-Path: <iutcovic001@mail1.gannon.edu

X-Sender: iutcovic001@mail1.gannon.edu

Date: Tue, 23 Feb 1999 01:42:15 -0500

To: gee@sfu.ca

From: Joyce Iutcovich <iutcovic001@mail1.gannon.edu

Subject: Response to Ethics Query

Cc: levine@asanet.org

Dear Ellen:

The following is a response to your query about the 1997 ASA Code of Ethics with regard to limited/unlimited confidentiality.

In your e-mail of 3/6/99 you stated that your reading of the new ASA Code

1) With regard to (a)--In Canada (and the U.S., as far as I understand), there are no legal guarantees regarding academic privilege. Does this then mean that researchers are NEVER in a position to guarantee unlimited confidentiality?

2. With regard to (b)--this appears to mean that the ASA Code requires researchers to tell their subjects that any data obtained from them COULD be disclosed. Does this not impede research that criminologists (especially, but not exclusively) engage in, given that research subjects will either not want to participate or will hide information from the researcher? Also, does this not preclude researchers from using the Wigmore criteria (re criteria #1) in a court of law in any attempt to protect research-research subject confidentiality/academic privilege?

The essence of the 1997 Code is this: researchers are obligated to be fully informed, only make promises they are willing to back up, and be up front with the research subjects about what they are promising. It may be that a researcher is willing to promise confidentiality except in health- or life-threatening situations. If that is the case, then that is what can be promised.

I hope this answers your questions. If you want any further clarification or you have some additional questions, please contact me.

Sincerely,

Joyce Iutcovich

***************************************