The betrayal of research confidentiality in British sociology 1

Abstract

Research confidentiality in Britain is under attack. Indeed, in some quarters the ‘Law of the Land’ doctrine that absolutely subjugates research ethics to law is already a fait accompli. To illustrate the academic freedom issues at stake, the article discusses: (i) the Cambridge Psychology Research Ethics Committee’s ban of interview questions about a research participant’s involvement in criminal acts; (ii) the awarding of damages against Exeter University when it reneged on its agreement to uphold a doctoral student’s guarantee of ‘absolute confidentiality’ in his research on assisted suicide; and (iii) the controversy around the UK government’s attempt to obtain confidential records from the Belfast Project − an oral history of paramilitaries involved in the Troubles in Northern Ireland. The article urges British researchers to practice – or, at least, defend the academic freedom of their colleagues to practice – the ‘ethics-first’ doctrine of strict confidentiality that several North American disciplinary associations encourage.

When conducting human research, a researcher’s primary ethical duty is to protect research participants from harm. In certain kinds of inquiry researchers cannot help but learn the identity of their research participants, even if they remove names and other identifiers from interview transcripts and databases. When anonymity is not possible and research participants are asked to divulge information that, if attributed to them, could seriously harm them, researchers are ethically obliged to protect participant identities. ² This duty to maintain confidentiality in research on ‘sensitive topics’ – health, sexual behaviour, crime etc. – is enshrined in all human research ethics codes.

The duty to maintain confidentiality is particularly important in research that asks participants to divulge information about illegal behaviour. Their reward is partly altruistic: providing information for the benefit of social science and the understanding of society. Given that it usually has limited personal benefit, what person would divulge information about crimes they committed knowing that the researcher would turn it over to a prosecutor or police to be used against them? The same problem applies to all research on sensitive topics. Why would anyone provide information for research purposes if they thought it could end up being used to harm them? Conversely, why would any ethical researcher gather information that could harm research participants with the intention of disclosing it if pressured?

In view of its importance to research on sensitive topics, it is thus difficult to imagine what could have happened in the UK for British Sociological Association (BSA) President John D Brewer (2012) to assert that ‘Social researchers working on sensitive topics know that the confidentiality of interviewees cannot be guaranteed’. His comment was made in relation to the pledge of confidentiality that Boston College gave to participants in the Belfast Project − an oral history of paramilitaries involved in ‘The Troubles’ ³ in Northern Ireland. In its Donor Agreement with interviewees, Boston College promised that no interview material would be released until after the interviewee died or gave his/her permission. That Agreement is now being challenged by two sets of subpoenas that the US Attorney issued pursuant to a request by the UK Government under the UK−US Mutual Legal Assistance Treaty (MLAT), ostensibly to obtain material relevant to the investigation of a murder that occurred in 1972. The case and reaction to it raise important questions about the extent to which UK researchers can promise confidentiality given the threat, however remote, of court-ordered disclosure: should law set an absolute limit to research confidentiality? ⁴

Brewer’s ‘confidentiality cannot be guaranteed’ perspective represents but one side of the lengthy debate in the US and Canada about what kind of confidentiality guarantee a researcher should be permitted to make given the possibility that a court or other legal authority could use subpoena power and rules of evidence to compel disclosure of confidential research information. On one side are ‘law-first’ researchers who subjugate ethics to law on the ground that, according to their conscience, the ethical course of action would be to obey the law. Because most of the published advocates of this position (e.g. Clayman, 1997; Piron, 2006; Stone, 2002) assert that researchers must obey the law, we refer to this as the Law of the Land doctrine (Lowman and Palys, 2007). Against this view, advocates of an ‘ethics-first perspective’ hold that the ethical course of action would be to defy a court order given that it would be based on ex post facto arguments about the relative weight of competing interests, such as law enforcement vs science. This act of civil disobedience would entail the researcher accepting whatever consequences might follow, including imprisonment for contempt of court.

Beauchamp and Childress (2001; also see Israel, 2004) identify three main arguments supporting non-disclosure: (a) the potentially far-reaching consequences of disclosure on all research on ‘sensitive topics;’ (b) the rights of research participants to privacy and control over their personal information; and (c) fidelity-based arguments which hold that researchers have a fiduciary duty to prevent harm to research participants: a research participant cannot trust a confidentiality guarantee that may or may not be kept. Following Dingwall (2008), we would add a fourth: researchers must resist being co-opted as agents of the state lest the state become the sole producer of knowledge about crime and other sensitive topics. Speaking against the system of centralized ethics regulation that is threatening to stifle social science research in North America and the UK, Dingwall (2008: 10) noted:

When we give up doing participant observation with vulnerable or socially marginal groups because of … regulatory obstacles, then a society becomes less well-informed about the condition of those who it excludes and more susceptible to their explosions of discontent. How helpful is it when the only ethnographers of Islamic youth in the UK are undercover police or security service agents?

The three following examples from the UK suggest a culture of disclosure may be developing in the that will leave undercover police and security agents as the only enthnographers of criminal behavior:

• The Psychology Research Ethics Committee’s Handbook at Cambridge – one of the top-ranked universities in the world, housing one of the top-ranked criminology research institutes – advising that it cannot approve questions about a participant’s past criminal activity.

• An Exeter University doctoral dissertation on assisted suicide that could not be completed because of the way that Exeter first approved, but later withdrew, its promise to support the student’s guarantee of ‘absolute confidentiality.’

• The Belfast Project and Brewer’s argument that the researchers should have known, and should have told their participants that confidentiality ‘cannot be guaranteed.’

A brief description of the debate over research confidentiality in North America sets the stage for a discussion of the culture of disclosure that has taken hold in Britain. The paper aspires to defend an ethic of strict confidentiality against those who would impose the Law of the Land perspective on their peers in the UK.

The threat of court-ordered disclosure in North American research

In contrast to the UK, where we cannot find any examples of researchers receiving subpoenas and being ordered to disclose confidential research information, there are dozens of cases in the US. The FBI’s threat to subpoena the sex research records of the Kinsey Institute at Indiana University was the first such threat reported in the literature (Carroll and Knerr, 1973). By the mid-1970s Carroll and Knerr (1975) had identified more than 20 instances of law enforcement agencies, grand juries, congressional committees and prosecutors attempting to force researchers to yield information to facilitate investigation and prosecution of research participants. In two instances, researchers went to jail – Samuel Popkin for eight days (Carroll and Knerr, 1973) and Richard Scarce for 159 days (Scarce, 1994) – rather than violate their pledges of confidentiality.

Although the number of incidents involving prosecutions and criminal investigations declined in the US after the 1970s, other researchers found themselves being drawn into various forms of high stakes civil litigation, receiving subpoenas to disclose all manner of information, some of which was confidential. These cases had become so numerous that in 1996 a special issue of Law and Contemporary Problems was devoted to court-ordered disclosure of academic research in what its editors characterized as the clashing values of science and law (Cecil and Wetherington, 1996). There have been numerous other cases since then, one of the most notable involving Harvard Professor Michael Cusumano and MIT Professor David Yoffie, who had interviewed 40 Netscape employees about their ‘browser war’ with Microsoft. To facilitate its defence in an anti-trust case, Microsoft subpoenaed both researchers and asked for all their interview tapes, transcripts, data files, field notes and correspondence regarding the browser wars. Recognizing the importance of confidentiality to research and that disclosure would disrupt the free flow of information to the public, thus ‘denigrating a fundamental First Amendment value,’ ⁵ the court quashed the subpoena. For US researchers, the First Amendment ⁶ may well hold the best potential for defending research confidentiality.

In Canada, only one researcher – Russel Ogden – has ever received a subpoena that sought confidential research information. Ogden’s MA research at Simon Fraser University involved interviewing people who had been involved in assisting a suicide. Because of the potential consequences of a person disclosing their involvement in an assisted suicide – up to 14 years’ imprisonment – the SFU Research Ethics Review Committee approved Ogden’s undertaking to promise participants ‘absolute confidentiality.’ After his completed thesis drew wide media attention, the Vancouver coroner noticed that it referred to two persons who informed Ogden about the death of an ‘unknown female.’ The Coroner subpoenaed Ogden to appear at an inquest into that death. Rather than mounting a staunch defence of research confidentiality in coroner’s court, SFU effectively abandoned Ogden and his research participants. To prevent having to mount a defence of research confidentiality against any future threats of subpoena, the SFU research ethics committee imposed a regime of limited confidentiality, a liability management strategy that required researchers to warn prospective research participants that if the researcher received a subpoena seeking confidential research information, s/he would spill the beans (Lowman and Palys, 2000).

What SFU did not anticipate was Ogden successfully defending his pledge of research confidentiality using the Wigmore test (Palys and Lowman, 2000), which the Supreme Court of Canada has deemed to be the appropriate common law mechanism for adjudicating claims of privilege. ⁷ Although Ogden lost his subsequent law suit against SFU to recover his legal expenses, the judgment castigated SFU for its ‘hollow and timid’ defence of academic freedom. ⁸ An internal SFU review of the case (Blomley and Davis, 1998) reached the same conclusion, at which point the university issued an apology, reimbursed Ogden’s legal fees and introduced a policy that extended the same legal indemnification to graduate student research that it provides for faculty (Lowman and Palys, 2000). At the same time, two SFU researchers (us) refused to use the REB’s limited confidentiality consent statement, and brought a grievance against the administration for infringing their academic freedom to follow an ethics-first approach to research confidentiality instead of the university’s Law of the Land doctrine.

The researchers won the grievance and, after a prolonged struggle, the right of all SFU researchers to adhere to an ethics-first approach. They and other advocates of the ethics-first perspective then engaged in a struggle at the national level to prevent the Law of the Land doctrine being enshrined in Canada’s national research ethics policy, the Tri-Council Policy Statement: Ethical Conduct for Research Involving Humans (Canadian Institutes of Health Research, Natural Sciences and Engineering Research Council of Canada, and Social Sciences and Humanities Research Council of Canada, 2010; hereafter TCPS2). This second incarnation of the code says that ‘Researchers shall maintain their promise of confidentiality to participants within the extent permitted by ethical principles and/or law.’ This phrasing indicates that, although researchers should do their utmost to comply with law, ⁹ they can promise confidentiality to the extent permitted by ethical principles, which, as we shall see, includes maintaining confidentiality even when legal force is applied. TCPS2 also says that ‘Researchers shall avoid being put in a position of becoming informants for authorities or leaders of organizations.’ ¹⁰ This statement implies that researchers should not collect information and then hand it over to authorities for the purpose of prosecuting a research participant.

In view of the well-established history in the US of civil disobedience in support of research confidentiality, and given the issues at stake, imposition of the Law of the Land doctrine on UK researchers would constitute a fundamental infringement of their academic freedom to conduct research. Although the risk of court-ordered disclosure is minuscule – we know of no UK court ordering disclosure of confidential research information – formulaic a priori limitations of confidentiality could have a decidedly chilling effect on research on sensitive topics, particularly if it concerns government, military or corporate malfeasance, or any kind of criminal behaviour.

‘Guilty knowledge’: Losing the plot

One way of preventing researchers from being co-opted as informers is to prohibit research on what Brewer (2012) refers to as ‘guilty knowledge’, as apparently the Cambridge Psychology Research Ethics Committee (CPREC) has done, although it remains to be seen how often this policy was put into effect or has otherwise influenced research at Cambridge. The CPREC’s Handbook describes two concerns that stood out when it encountered a proposal to interview 14 to 18 year old youths about various activities, including their drug-taking habits:

The first area concerned the possibility of adverse publicity … if it became known by the press that a member of the University … were in possession of information concerning drug users and/or drug dealers, and the fact that in this situation the confidentiality of personal information about the participants could not be guaranteed …

The second area concerned the fact that … if the police became aware … that a member of the University … were in possession of information related to their case, then such a member … might be called to give evidence, and they might be held in contempt of Court if they refused …

The Committee … cannot give ethical approval to any project which involves questioning participants about their participation in, or actual knowledge of criminal acts … ¹¹

CPREC’s rationale ignores how other disciplines routinely conduct interview research with law-breakers using a strict confidentiality guarantee. For example, the Academy of Criminal Justice Sciences ethics code (s.19) holds that ‘Confidential information provided by research participants must be treated as such by criminologists, even when this information enjoys no legal protection or privilege and legal force is applied.’ The American Sociological Association code holds that researchers are ethically obliged to be aware of relevant law, make an ethical decision about the degree of confidentiality they are prepared to promise and then abide by that pledge. ASA has confirmed that this statement does not imply that researchers should divulge information under pressure from the courts (Iutcovich et al., 1999).

Ironically, in raising the spectre of how the press would treat research on law-breaking, the CPREC Handbook is oblivious to the ethics of journalism. Many reporters guarantee strict confidentiality for the same reasons that ethics-first researchers do. In cases on both sides of the Atlantic, journalists have refused to divulge sources, even if it means being jailed for contempt of court. Their publishers and editors usually support them. As the US First Circuit remarked:

[S]cholars … are information gatherers and disseminators. If their research materials were freely subject to subpoena, their sources likely would refuse to confide in them. As with reporters, a drying-up of sources would sharply curtail the information available to academic researchers and thus would restrict their output. Just as a journalist, stripped of sources, would write fewer, less incisive articles, an academician, stripped of sources, would be able to provide fewer, less cogent analyses. ¹²

The CPREC Handbook neglects to mention the value of research and the need to defend research participants in court. Instead, it prioritizes liability and image management in the process of creating a research prohibition that would end criminology and the sociology of deviance as we know it. Many field research classics would have been banned under this regime. Criminologists would have to give up interviewing anyone but convicts about the crimes for which they have been convicted; ‘successful’ criminals would be ruled out of bounds. What would criminology be if it studied only the people who get caught? Imagine applying this ‘ethic’ to research on white collar crime, workplace crime, corporate crime, environmental crime, male violence against women, drug trafficking, and the many other arenas of criminal activity where most offences are not reported, let alone a charge laid or a conviction registered.

Although a discussion of the effects of the broader regime of ethics regulation lies beyond the scope of this article, the CPREC approach illustrates precisely the sort of problem that animates opponents of the current centralized system of research ethics regulation in North America and the United Kingdom. An alternative approach is to allow interview research on illegal behaviour and leave it to a researcher’s conscience to decide how to resolve potential conflicts between research confidentiality and laws of disclosure. The two research projects described next suggest that if the Law of the Land approach holds sway and researchers studying illegal activity allow themselves to be turned into informers, the ‘dark figure’ of crime will become much darker.

Co-opting researchers as informants: Research on assisted suicide

One instance where the commitment of UK sociologists to research confidentiality was put to the test involved Exeter University sociology PhD student Russel Ogden’s (1994) research on assisted suicide and euthanasia among persons with HIV/AIDS. Ogden arrived at Exeter on the heels of his experience at SFU.

Exeter’s ethics review process

In the hope of avoiding the problems he experienced at SFU, Ogden warned the Exeter ethics administration about the risks his research posed to both research participants and the University. Given the subject of his research, would the University fight a subpoena should one arrive? In response the Sociology Department ethics sub-committee stated:

In the knowledge of the applicant’s previous experience of the need to actively uphold the commitments to confidentiality given to research subjects, the members of the Committee express their personal support for the researcher in continuing to meet such commitments. They recognise that entry into commitments of this kind is integral to the pursuit of truth through sociological research, and accept the obligation to support and sustain those who do so.

Believing that the second sentence of this statement meant that he had Exeter’s full support, Ogden started his research. With the respect he earned as a result of his highly publicized resistance to the Vancouver coroner’s subpoena, his snowball sampling procedure delivered in spades. However, as his research progressed, his relationship with his supervisor deteriorated. When the supervisor encouraged Ogden to restrict his inquiry to attitudes of persons with HIV and discouraged him from presenting information about the experience of persons who aided deaths, Ogden began to worry about Exeter’s commitment to support his confidentiality protocol.

In 1998 Ogden twice requested a formal statement outlining what support Exeter would give him if he received a subpoena. After the second query, the Department Head told Ogden that his research approval had not been rescinded, but offered no information about what support he could expect. When Ogden informed several research contacts that he was concerned about Exeter’s willingness to protect his research participants, two wrote to the Sociology Head requesting documentation from the ethics committee clarifying what kind of support they would receive should a subpoena arrive. They were sent a copy of what was supposed to be the Sociology ethics sub-committee report on Ogden’s proposal. But the document was not the same as the one in Ogden’s possession. The sentence, ‘They recognise that entry into commitments of this kind is integral to the pursuit of truth through sociological research, and accept the obligation to support and sustain those who do so,’ had disappeared. The document was dated five days later than the meeting at which Ogden’s research had been approved. The ethics subcommittee chair had excised a crucial part of the document, but neglected to inform Ogden. When Ogden asked for clarification, Exeter informed him that ethics approval ‘does not as a matter of law … oblige the University to provide a defence of the student’s work to third parties or to participants. Nor does the process … involve any obligation on the University to provide, or to pay for, any legal assistance or advance to the student in connection with that advice.’

With the rug pulled from under his confidentiality guarantee, how could Ogden use the approximately 100 interviews he conducted in Canada, the Netherlands, the US and the UK knowing that he had unwittingly misinformed his research participants about the support Exeter would provide?

The inquiries

In 1998 Exeter suspended Ogden on the grounds that he was not making progress on his research or paying fees. In 1999 an internal Committee of Inquiry found that Exeter had mishandled the ethics review and not provided Ogden with adequate supervision. Although the Inquiry attributed Ogden’s problems to the administration’s ‘serious incompetence’ and ‘misjudgement’ in situations where policies were ‘mishandled’ and procedures ‘deficient,’ Exeter offered no remedy for damages, and declined to honour the support the Sociology ethics sub-committee had promised. Exeter then rewrote its ethics policy and applied it to Ogden retroactively.

In light of Exeter’s refusal to pay damages, Ogden petitioned the ‘Visitor of Exeter University,’ Her Majesty the Queen. She appointed the Lord Chancellor to conduct an investigation. Exeter continued to reject Ogden’s damage claim, arguing that irregularities in the ethics review process had only potentially undermined his research, but not actually undermined it. Exeter argued that, under its new research ethics policy – the policy developed in response to the Committee of Inquiry’s recommendations – Ogden’s guarantee of absolute confidentiality was not acceptable. The revised policy requires that the researcher shall:

not give unrealistic guarantees of confidentiality and anonymity and be aware that legal challenge, moral duty and mishap may preclude the honouring of such a guarantee. It may be necessary to inform research participants of obligations under law, such as the possibility that the researcher will be required to give evidence or reveal documents, which may make it impossible for certain information to be kept confidential without breaking the law (Appendix 1, para. 12.2 (b)).

This prescription leaves open the question of whether violation of a court order is ethically permissible. The policy also spelled out the university’s obligations:

(b) The University is committed to offering access to professional legal advice to a researcher who, in good faith, encounters potential legal problems when planning their research or when conducting research approved under the Universities ethics procedures.

(c) The University is committed to covering reasonable legal costs associated with any challenge to research approved under the University’s ethics procedures, in respect of guarantees of confidentiality made in good faith by the researcher. This includes reasonable challenge to any subpoena or other legal process that seeks disclosure (Appendix 1, para. 12.3).

Although these changes indicated Exeter’s intention to defend confidentiality in court, they did nothing to address the more immediate problem of Ogden’s guarantee of absolute confidentiality that Exeter had approved. Nor did they clarify whether Exeter would permit a researcher to adhere to an ethics-first perspective.

The Visitor’s judgment

The Visitor’s conclusion regarding Exeter’s culpability was unequivocal:

Despite the fact that the findings of the Enquiry are stated in its rules to be ‘final’, these findings have not led to any positive result for Mr Ogden. I have found the reasons put forward by the University for its inaction to be unconvincing. Accordingly, I think that Mr Ogden is entitled to some remedy from the University. ¹³

The Visitor found wanting Exeter’s excuse for failing to act on the Inquiry’s recommendation to support Ogden’s confidentiality guarantee:

The University’s failure to do anything to implement this recommendation is unfortunate. … (I)t would have been perfectly possible for the University to have produced a written statement on the University’s letterhead containing the appropriate assurances which could have been given to Mr Ogden for him to send to all his data subjects without any breach of confidentiality. ¹⁴

The Visitor concluded that Exeter denied Ogden the opportunity to obtain a PhD and awarded him £62,000 in damages.

Exeter’s revision of its research ethics policy to preclude ‘unrealistic guarantees of confidentiality’ begs the question, what exactly is a ‘realistic guarantee,’ and from who’s standpoint? Although absolute confidentiality may be unrealistic – who can say if they will spill the beans in the event that they are tortured, or how can they absolutely ensure that data will not be stolen – is this any reason to create a priori limitations to confidentiality that preclude a researcher violating a court order? Is Exeter saying that the ethics-first perspective is not permissible in research conducted under its auspices? Is the ethics-first perspective permissible at any British university?

The Belfast Project ¹⁵

On 23 January 2012 Massachusetts Senator John Kerry, Chair of the US Senate Committee on Foreign Relations, wrote ¹⁶ to Secretary of State Hillary Clinton urging her ‘to work with the British authorities to … revoke their request’ under the UK-US MLAT to acquire confidential interviews provided for the Belfast Project. Kerry explained that the fallout from yielding up the interviews could derail the entire peace process that the 1998 Good Friday Agreement initiated. ¹⁷

The Project involved interviews with members of paramilitary organizations ¹⁸ on both sides of the Troubles, including the Provisional Irish Republican Army (IRA) and loyalist Ulster Volunteer Force (UVF). Its purpose was to:

gather and preserve for posterity recollections that would help historians and other academicians illuminate the intricacies of the Northern Ireland conflict in studies and books, and that would advance knowledge of the nature of societal violence in general, through a better understanding of the mindset of those who played a significant part in the events in Northern Ireland. ¹⁹

Boston College historian Thomas Hachey and Burns Librarian Robert O’Neill enthusiastically supported and oversaw the project, which the College funded, hiring Moloney as Director. ²⁰ Moloney, in turn, hired former Provisional IRA volunteer Anthony McIntyre to conduct interviews with Republicans, and Wilson McArthur, who was familiar with the UVF, to conduct interviews with loyalists.

When describing their paramilitary experiences some interviewees disclosed systematic illegal conduct, including their involvement in or knowledge of abductions and murders for which no one has ever been charged. If their identities became known, the IRA participants could be treated as informers and executed by dissident IRA members, as could McIntyre given his former role as a Provisional. The Belfast Project thus represents the quintessential example of research that could not have been conducted without a cast-iron guarantee of confidentiality. To that end, each research participant signed an ‘Agreement of Donation’ (hereafter, the Donor Agreement) with Boston College’s Burns Library, which stipulated that:

Access to the tapes and transcripts shall be restricted until after my death except in those cases where I have provided prior written approval for their use following consultation with the Burns Librarian, Boston College. Due to the sensitivity of content, the ultimate power of release shall rest with me. After my death the Burns Library of Boston College may exercise such power exclusively. ²¹

The first set of subpoenas arrived in May 2011 and sought the interview of one Republican who had died and another who was still living. A second set of subpoenas seeking information about a 1972 murder arrived in August that year. A rift soon developed between the researchers and Boston College authorities over how to proceed. The researchers have used every legal avenue available to them to challenge both sets of subpoenas and have said that they would defy any order to violate their strict confidentiality pledge. In contrast, Boston College rolled over after the first negative ruling on its application to have the first subpoenas quashed, at which point it began asserting that its allegiance to participants was only ‘to the extent that American law allows.’ The conflict between the College and researchers became irreconcilable when Boston College failed to appeal the initial negative ruling and then, in response to the second set of subpoenas, handed over the entire Republican archive so that the Judge hearing the case could decide which interviews were responsive to the MLAT request. ²²

At the time of writing the Belfast Project researchers are attempting to win standing to appeal the release of the first interview that Boston College turned over to the court. Boston College has since appealed the order to hand over interview material that is beyond the purview of the second set of subpoenas. It remains to be seen how these appeals will play out. One of the researchers is also seeking a judicial review in Northern Ireland of the MLAT request, but without Boston College’s support.

Inconsistency in the Belfast Project contracts

One lesson to be learned from both Ogden’s experience and the Belfast Project is the importance of ensuring that a university administration, the researcher(s) and research participants are all in agreement about the guarantee of confidentiality being made and that the contracts encapsulating it are clear and consistent. The conflict between Boston College and the researchers after the subpoenas arrived concerned what guarantee of confidentiality was provided. In the minds of the research participants and the researchers it was unlimited. Every statement that Boston College personnel made prior to the arrival of the first set of subpoenas substantiated this interpretation. Indeed, the College wore the cast iron guarantee as a badge of honour; without it there would be no Belfast Project archive. It was not until Boston College decided to hand over Republican interviews that it began claiming that the confidentiality guarantee was limited by law.

For Project Director Ed Moloney and researchers Wilson MacArthur and Anthony McIntyre the confidentiality guarantee had no limit, and nor did the Donor Agreement that Boston College made with interviewees. However, its contract with the Project Director contained a clause saying, ‘Each interviewee is to be given a contract guaranteeing to the extent American law allows the conditions of the interview and the conditions of its deposit at the Burns Library, including the terms of an embargo period if this becomes necessary, as outlined herein.’ ²³

Moloney understood the ‘extent of law’ clause to refer to copyright, not confidentiality. ²⁴ Nevertheless, he and the two researchers sought clarification from Boston College about potential legal threats to the Belfast Project archive once it was housed at Burns Library. ²⁵ McIntyre and McArthur have reported that a representative of Boston College when visiting Northern Ireland assured both of them independently that, once housed in the Burns Library, the interviews would be immune from legal threats. As McArthur told a Boston College student newspaper, ‘It could not have been more clear in all discussions with BC staff that this project would never have been possible without the absolute guarantee of confidentiality which predicated the whole thing.’ (Cote, 2012).

Given the personnel conducting the Belfast Project – McIntyre was a former paramilitary who had spent 18 years in Northern Ireland prisons including the Maze and lived in a world where a breach of confidentiality could mean a death sentence, working under the supervision of a journalist whose job depends on source confidentiality ²⁶ – the Boston College officials responsible for the research must have understood that it could not proceed without a cast-iron promise of confidentiality. As a result of feeling misled, the researchers have said that the entire archive should be destroyed or returned to them. ²⁷ At least two interviewees have asked for their interviews to be returned.

For its part, Boston College has defended its decision to release confidential interviews to the court by arguing that the law clause in Moloney’s contract meant that confidentiality was guaranteed only to the extent that law allows. If that was its intention, it is difficult to understand why that vital qualification did not appear in Boston College’s agreement with interviewees. If it had, none of them would have agreed to be interviewed, nor would the research team have conducted the interviews. Clearly, Boston College failed to provide interviewees with sufficient information for informed consent.

Academic reactions to the subpoenas: A study in contrasts

Kerry’s letter to Clinton explained that the UK request for Belfast Project interviews could not only derail the Northern Ireland peace process, but also would have ‘a profound impact on Boston College’ and on ‘confidentiality of other research projects of this nature.’ Several other letters from members of Congress and Senate ²⁸ to Clinton endorsing Kerry’s concerns were but part of a growing wave of criticism of the UK’s assault on research confidentiality. On 1 February 2012 the ASA Council added its weight to the fray, arguing that reliable guarantees of confidentiality are a ‘core component of … research based knowledge that is critical to an informed society and its well being:’

The release of the ‘Belfast Project’ interview data threatens the academic freedom to study difficult and controversial topics. It undercuts the willingness of potential participants in future research to share valuable information. In the short run, such intrusion in research seeking to understand past tragedies can harm the processes through which Northern Ireland now seeks political stability. And in the long run, we must weigh the potential damage to social science that can provide a firmer knowledge base for avoiding these types of conflicts in the future. ²⁹

In stark contrast, the BSA has not made any statement about the UK government’s threat to research confidentiality. BSA’s silence could be interpreted as endorsing the UK government’s position that investigating a murder that occurred in 1972 – an investigation that has been the subject of wide-ranging criticism ³⁰ – outweighs the importance of research confidentiality. At about the same time Kerry was writing to Clinton, BSA President John Brewer (2012) was arguing that:

Social researchers working on sensitive topics know that the confidentiality of interviewees cannot be guaranteed. This is why informed-consent forms always explain that confidentiality will be maintained only to the full extent provided under the law – and why respondents’ attention must be drawn to the risks.

On this basis, Brewer concluded, ‘This is not a tale of researchers betrayed by Boston College or hounded by callous lawyers … the problem is primarily that the journalists ³¹ felt themselves under no obligation to act responsibly … they are not the victims here.’ Brewer leaves the impression that research ethics must be absolutely subordinated to law.

Although Brewer did not make these comments in his capacity as BSA President, would BSA endorse it? We turn to disciplinary ethics codes at this juncture because the Economic and Social Research Council’s Framework for Research Ethics – with which researchers must comply in order to receive ESRC funding – states that “confidentiality must be respected” without specifying any exceptions or addressing potential conflicts between research ethics and laws of evidence. Section 38 of the BSA Code of Ethics implies that it would: ‘Research data given in confidence do not enjoy legal privilege, that is they may be liable to subpoena by a court and research participants should be informed of this.’ ³² Would the British Society of Criminology endorse this view? Section 4.iv of its Code of Ethics for Researchers in the Field of Criminology implies that it would:

Offers of confidentiality may sometimes be overridden by law: researchers should therefore consider the circumstances in which they might be required to divulge information to legal or other authorities, and make such circumstances clear to participants when seeking their informed consent; ³³

The Sociology and Criminology codes require this kind of warning even though the risk is entirely theoretical in the UK – even the subpoenas relating to the Belfast Project were issued in another country. The implication is that if subpoenaed, the researcher will hand the information over without a fight – caveat emptor. Worse, it is possible that a UK court could treat this a priori limitation of confidentiality as a waiver of privilege, as occurred in a US case in which participants had been warned that the information they were providing would be confidential “except as required by law. ³⁴ In this instance, the court held that, because research participants had been warned that a court might require the information, now that the court did require the information, it should be handed over, because this was precisely the limitation to confidentiality about which participants had been warned. Instead of buttressing a researchers’ ethical commitment to confidentiality, a priori limitations could sabotage a researcher’s ability to invoke privilege on the participants’ behalf. Warning the participant that a court might order disclosure devolves into a form of caveat emptor dressed up as ethics, allowing the researcher to roll over without a fight, sacrificing participant rights at the altar of informed consent.

If research participants must be informed about the risk of subpoena in the name of achieving informed consent, surely researchers and universities should also inform research participants what they propose to do if a subpoena arrives. Would the university indemnify the researcher’s attempt to contest it as Exeter now claims that it would? Given the importance of confidentiality to research, would the university appeal any negative decision to the full extent permitted by law, i.e. to the top court in the land? ³⁵ If attempts to contest the subpoena are not successful, would the researcher do what many journalists do in these circumstances, i.e. refuse to disclose sources?

If Brewer is correct that UK researchers always promise confidentiality only to the extent permitted by law, then they have sold out academic freedom before the battle to save it from attempts to turn researchers into informants was ever waged.

US confidentiality certificates: Reconciling research ethics and the Law of Disclosure

In the US in 1970 the first protection of research confidentiality beyond the national census was introduced. To enable research on drug abuse among soldiers returning from Vietnam, the Comprehensive Drug Abuse Prevention and Control Act 1970 authorized the Secretary of Health Education and Welfare to issue ‘certificates of confidentiality’ ³⁶ to researchers that prevented compelled production of research information on drug use (Madden and Lessin, 1983). In 1974 confidentiality certificate coverage was expanded to mental health research in general. ³⁷ As of 1988 it covers all research on sensitive topics that fits under the mandate of the National Institutes of Health.

Section 301(d) of the Public Health Service Act (42 U.S.C. 241(d)) authorizes the US Secretary of Health and Human Services (DHHS) ³⁸ to issue confidentiality certificates to researchers involved in any health research where confidentiality is deemed to be essential for producing valid and reliable information, regardless of whether DHHS funds it. Topics include, but are not limited to:

• Research on HIV, AIDS, and other STDs;

• Studies on sexual attitudes, preferences, or practices;

• Studies on the use of alcohol, drugs, or other addictive products;

• Studies that collect information on illegal conduct;

• Studies that gather information that if released could be damaging to a participant’s financial standing, employability, or reputation within the community;

• Research involving information that might lead to social stigmatization or discrimination if it were disclosed;

• Research on participants’ psychological well-being or mental health;

• Genetic studies, including those that collect and store biological samples for future use;

• Research on behavioural interventions and epidemiologic studies. ³⁹

Confidentiality certificates ‘allow the investigator and others who have access to research records to refuse to disclose identifying information on research participants in any civil, criminal, administrative, legislative, or other proceeding, whether at the federal, state, or local level.’ By protecting researchers and institutions from being compelled to disclose information that would identify research subjects, confidentiality certificates ‘help achieve the research objectives and promote participation in studies by helping assure confidentiality and privacy to participants.’ ⁴⁰ A battery of other Federal and state laws shield other kinds of government funded research, to the point that there is now a veritable ‘armamentarium’ of protections (Fanning, 2007). ⁴¹

To our knowledge Federal research confidentiality certificate legislation has been challenged only once. ⁴² The decision described confidentiality certificate protection as ‘absolute’ ⁴³ and dismissed the challenge. The US Supreme Court declined to hear an appeal. Had the Belfast Project or Ogden’s research been conducted with such protection, many of the issues the current article raises would be moot.

Protecting research confidentiality

The existence of confidentiality certificates is a reminder of just how important confidentiality is to research, and how researchers and universities ought to fight to achieve such protection rather than blithely accepting that ‘confidentiality of interviewees cannot be guaranteed.’

With its armamentarium of protections, the US government has recognized that a promise of confidentiality that may or may not be kept cannot be relied upon to produce reliable and valid data. In both the US and Canada, ethics-first researchers and the disciplinary associations that support them continue to exercise their academic freedom not to restrict their promise of confidentiality to the extent permitted by law unless they are protected by a confidentiality certificate or its equivalent. How else does one emphasize the importance of confidentiality to research other than by making a trustworthy pledge and standing by it? Walking one’s talk is particularly important in jurisdictions where claims of research confidentiality would be determined case by case according to common law.

In the US, research not protected by a confidentiality certificate or other shield law, and in Canada and the UK, where there are no shield laws for academic research, a court’s decision about whether to order disclosure would be based on an ex post facto balancing of competing interests. In contrast, researchers have to decide the importance of confidentiality at the outset, knowing that their research protocol will become part of the evidence if a court should be asked to weigh the importance of research confidentiality against some other social interest, such as criminal law enforcement. Although there are numerous examples in the US of legal challenges to research confidentiality, we have yet to see a situation where disclosure would have been ethically appropriate, which is why we do not anticipate complying with a court order. Nevertheless, if there is a reasonably foreseeable risk of a court ordering disclosure research participants should be warned. Also, to comply with the ethical requirement of informed consent, they should be informed about the researcher’s and university’s intentions should a court order a researcher to violate confidentiality.

Although Canada and the UK have not created a system of confidentiality certification – perhaps because there has not been a single example of court-ordered disclosure in either country – research ethics boards routinely require researchers to warn participants about the threat. By encouraging risk aversion, limited confidentiality already has had a chilling effect on research (Palys and Lowman, 2010).

The warning about a threat that has never materialized has created two problems for Canadian researchers studying sensitive topics. First, having mentioned the threat of court-ordered disclosure, researchers then find themselves in the position of having to declare whether they adhere to a Law of the Land or ethics-first approach. Second, TCPS2, Canada’s national ethics code, advises researchers that they ‘shall avoid being put in a position of becoming informants for authorities or leaders of organizations’ (Canadian Institutes of Health, Natural Sciences and Engineering Research Council, and Social Sciences and Humanities Research Council, 2010: 59). In light of this admonition, it would be unethical for researchers to collect information about crime knowing that they would turn it over to the authorities if ordered to do so. Criminologists in particular have to be mindful of the conundrum created by this clause in combination with other sections of the code. Given that the TCPS2 also says that ‘Researchers shall maintain their promise of confidentiality to participants within the extent permitted by ethical principles and/or law’ (2010: 58), it does not preclude the ethics-first approach in those rare cases where a court might order disclosure. However, if Brewer is correct, imposition of the Law of the Land doctrine in the UK is a fait accompli.

The implications of the Law of the Land doctrine for sociology and criminology are truly disturbing. Already, the Belfast Project subpoenas have had a chilling effect on research. The Guardian reported that as a result of the US justice department’s pursuit of parts of the Belfast Project archive, ‘police officers, soldiers and spies who fought the IRA have pulled out of a proposed parallel project potentially involving a London-based college’ (McDonald, 2012). How complete is our knowledge of society’s underdogs and those with the most power when the only people who get to be heard are government-approved spokespersons and administrative hacks? There could be a huge price to pay if the state is able to turn researchers into informants.

Although ethics-first researchers do their utmost to comply with law, the presence or absence of confidentiality certificate protection does not change their ethical obligation to protect research participants from harm. They are prepared to engage in civil disobedience to support the principle that confidentiality certificates are designed to protect. Such disobedience does not put researchers ‘above’ the law, as some commentators have argued (e.g. Stone, 2002); they remain subject to the law, which is why they would accept imprisonment as a consequence of their actions (Lowman and Palys, 2003). Universities and disciplinary associations should support their academic freedom to make that choice – or, at the very least, to not stand in their way.