When the police call, researchers require backup

Source: James Fryer

The only way that most researchers can defend confidentiality so wholeheartedly is with their university’s support. And therein lies the rub

Two recent cases have thrown a spotlight on the gulf that can occasionally separate ethics and the law. The release of seven Belfast Project interviews with Republican paramilitaries to the Police Service of Northern Ireland and the seizure of Bradley Garrett’s “place-hacker” research material (“Access denied”, Features, 5 June) highlight an acute problem. If the state can turn researchers into informers, why would anyone disclose to researchers information that could land them in jail, lose them their job or cause them some other serious harm?

What can the academy do to bridge that gulf?

In the UK, one solution would be a research shield law, protecting researchers from being forced to disclose confidential information or sources in court. This would emulate such laws in the US, the most expansive of which is the US Secretary of Health and Human Services “confidentiality certificate” system, which protects participants in health research.

Research shield laws evolved in the US as a result of dozens of instances of courts, congressional committees, litigants and grand juries issuing subpoenas seeking confidential research information. In the UK, however, as far as we know, no legal authority has ever issued a subpoena for that purpose, so legislators may have little appetite to create a shield law. This leaves researchers having to assert “research-participant privilege” using common law.

In Anglo-American common law, the principle that courts are entitled to everyone’s evidence is not absolute. Occasionally, courts weigh their need for evidence against another public interest, such as the sanctity of confidential relationships. A confidential communication protected against compelled disclosure is “privileged”. Courts have deemed lawyer-client confidentiality to be so important that it enjoys a “class privilege”, where the onus is on the party seeking access to a communication to show why the privilege should be set aside. Courts consider the value of other confidential relationships – such as priest-penitent, journalist-source, psychiatrist-patient – case by case. Here, the onus is on the party claiming the privilege to show why the court should recognise it.

Just such a case occurred in Canada this year. Police used a search warrant to seize the audio recording and transcript of an interview after learning that a person accused of murder had been interviewed by two University of Ottawa researchers years earlier for a project on sex work. The interview remained sealed until a court heard the researchers’ objections. The court recognised researcher-participant privilege and returned the interview without granting the prosecution access.

When not eligible for shield law protection, US researchers have turned to common law with great success. We have found only four cases where a legal authority ordered disclosure. Two involved grand juries. In both cases, the researchers chose to go to jail rather than to violate confidentiality. In the third case, the court used the researchers’ “limited confidentiality” warning to justify disclosure. The fourth case involved the Belfast Project interviews.

There are two main approaches to the legal limits on research confidentiality, which could be described as “law-first” and “ethics-first”.

All researchers seek to be both ethical and legal, but situations may arise where the two aims conflict. A law-first approach makes law the absolute limit of confidentiality: under such a principle, a researcher tells participants that they will comply with an order to disclose confidential research information. An ethics-first approach holds that it would be unethical for researchers to collect data knowing that they would become informants for legal authorities. Further, because it would so severely compromise reliability and validity, limiting confidentiality makes research involving sensitive and controversial social issues hardly worth doing.

Yet the importance of confidentiality is not lost on law-first researchers. While their conscience would have them comply with a disclosure order, they would use every legal means possible to avoid disclosure, taking their case to the highest court in the land if necessary. Ethics-first researchers would do the same – except in the last instant, when their conscience would lead them to defy a disclosure order. The only way that most researchers can defend confidentiality so wholeheartedly is with their university’s support. And therein lies the rub.

Take Garrett’s experience. When he was prosecuted, where was Royal Holloway, University of London, which approved his research? It ran for cover. What about the Belfast Project subpoenas? Initially, Boston College made a limp effort to challenge them, but then, incredibly, without the court compelling it, it handed over all the Republican interviews to the judge so that the content could be assessed. Where was the University of Ottawa when police seized the sex worker interview? It refused to fund the fight. If the Canadian Association of University Teachers had not provided funding to defend the researchers, what turned out to be a precedent-setting case might have been stillborn.

This is the most regrettable lesson of these experiences: the failure of university administrators to stand up for research confidentiality is an attack on academic freedom.