DO ETHICS BOARDS AND COMMITTEES POSE A THREAT TO CRITICAL RESEARCH?
In more than 35 years of academic research I've sought information from thousands of people, and done hundreds of interviews. During that time, ethical concerns, regarding both the substance of my research and my dealings with informants and respondents, have always been top-of-mind. The concerns I raise here are not with research ethics as such, but with bureaucracies that have gone awry in well-intentioned but misguided efforts to supervise research in politics and public policy.
In my experience - which antedates ethics bureaucracies by many years - two ethical concerns have stood out. One is my obligation to examine the way power is wielded, and look for ways of addressing shortcomings. For example, I've recently directed six case studies in three Canadian cities to look into how the federal government can fit national policies to the requirements of distinct communities. Some years ago, in studies of urban development in Edmonton and Winnipeg, I identified bad planning practices and looked for the administrative, political and socio-economic causes. All of this is main-line politics and policy research, typical of that being carried out by many of my colleagues.
Generally, I've found the causes of problems I identify in the course of my research to be systemic rather than personal, and I've never found it particularly useful to look for individuals to blame, but my findings do not always reflect well on individual power-holders. However, the primary ethical concern here is not their sensitivities. It is the public interest in good policy. It's a bedrock democratic principle that people in positions of power must be accountable for their actions.
Power-holders sometimes deal harshly with their critics, but that's a risk academic researchers must run, and our risk is limited by the institution of tenure - a guarantee that we may report our findings truthfully without fear of losing our academic appointments. The social purpose of tenure, therefore, is to ensure that the truth can be told despite the possibility of hostile reactions. Few things are more important to a democratic society than access to truthful accounts of how power is wielded. It follows that the ethical obligation of academics is to report truthfully and fearlessly. If we don't do that, there isn't a scintilla of justification for tenure.
My research, however, has not limited me to dealings with people in power. Over the years, I have acquired a great deal of valuable data from middle- or lower-level public servants, as well as hundreds of ordinary people. Here my primary ethical concern has been the protection of my informants from any possibility of reprisals. I've always tried my best to ensure that nothing in my publications identifies ordinary individuals who might be subject to reprisals.
Enter the ethics bureaucracies. In Canada, their job is to implement the Tri-Council Policy, so-called because the Canadian system of research ethics is based on a document jointly endorsed by the three funding agencies responsible for social science and humanities, natural science, and medical science research. There is a national supervisory apparatus, but the actual enforcement of the guidelines upon individual researchers is in the hands of departmental and university-wide ethics committees in each university. Individual researchers are required to file an explanation of their research project and tick off a set of assurances regarding their treatment of "research subjects". I'll restrict my comments to the Canadian system, though most, if not all, would apply to similar systems in other countries.
I'm in favour of the formulation of a set of ethics guidelines and a system of sanctions backing them. My concern is with the content of the Tri-Council guidelines, which actually militate against a social science researcher's obligation to look critically at the exercise of power, and, at the same time, deal inadequately with the potential plight of informants who might be subject to reprisals from their superiors. I'm also concerned about the way inappropriate scientific practises are introduced by ethics committees in individual universities, as they exercise their power to interpret the guidelines. I'll deal with each of these concerns separately.
THE RESEARCHER'S OBLIGATION TO ENGAGE IN CRITICAL SCRUTINY
When the Tri-Council guidelines were first introduced, researchers were not required to fill out an ethics application if their topic was limited to an examination of policy, how it works, and how it's implemented. Clearance became necessary only if the researcher initiated a line of enquiry that touched personal matters. I recently re-read the Tri-Council guidelines, and they now appear to require a researcher to file an application for investigation of policy, as well as enquiry into people's personal lives.
Many researchers concerned with politics and policy stay in regular touch with politicians and public servants and, in the process, ask them questions the answers to which may well be used in future publications. That's an essential part of the research process because regular contact with well-informed people makes it possible for researchers to stay abreast of events and identify important issues as they arise.
So when does a query become a research question and a conversation an interview that requires ethics review? The guidelines are little help in answering that question, but, if we take them literally, they would appear to have taken from university researchers a right that every ordinary citizen enjoys, namely that of picking up the phone and talking to a politician or public servant without applying for bureaucratic permission to do so. (See further comment on this point.)
In practice, of course, no one knows whether I'm talking to a politician, and no one cares. The problem is not that I will in fact be prevented from doing so. Rather it is that the existence of such severe strictures on ordinary enquiries sends a message to both researchers and public servants that the process of what was once free inquiry is now subject to bureaucratic control. Researchers are more likely to stick to "safe" topics, and public servants will feel emboldened to try to shut down criticism.
I have personally experienced a case in which a Manitoba public servant thought that the process of ethics review gave her the right to censor a manuscript I was in the process of preparing. I'm a senior professor with tenure and my response was a flat refusal, but were I less experienced, more junior, possibly still vying for tenure, who knows what my reaction might have been?
Ethics guidelines should not be written in such a way as to inhibit enquiry. Rather, they should underline a university researcher's right and obligation to engage in vigourous enquiry and factual, careful, but fearless critique. Rather than forcing researchers to apply for permission to do their work, it would make a great deal more sense to spell out clearly what they must and must not do, and put in place an enforcement mechanism with teeth. Hold us accountable, and make us pay for our mistakes if we fail, but don't stand over us like over-anxious parents and try to control our every move.
FAILURE TO PROTECT SUBORDINATES FROM REPRISALS
Top public servants and politicians are well accustomed to daily cut-and-thrust with newspaper reporters and political and bureaucratic rivals. It's obvious that they understand their rights and know how to protect themselves, yet ethics guidelines require researchers to treat both power-holders and subordinates as if they were vulnerable children. Ethics applications require the researcher to promise to begin all interviews with a reminder to the interviewee that she need not answer questions and may terminate the interview at any time.
In the best of circumstances, such procedures insult the intelligence of power-holders and make the researchers look like fools. But the most serious shortcoming of these regulations is that they make no distinction between superiors and subordinates. Just as researchers should be entirely free to speak to top public servants and politicians without restrictions, as any citizen is, so should they proceed under a strong mandate to protect subordinate informants from reprisals. It is not enough, and not even helpful, to go through the ritual of telling public servants what even the lowest-ranking of them knows perfectly well, namely that they cannot be compelled to answer questions.
Their problem is that they can't be certain which of their statements might prove damaging to them without knowing the details of how those statements will be contextualized in a future publication. The onus of protecting subordinates from reprisals belongs, not to the subordinate, but to the researcher. If an employee in a vulnerable position helps me to get at the truth, it ought to be my obligation, not his, to ensure that my publication doesn't expose him to reprisals. The Tri-Council guidelines appear to have been written in blissful ignorance of the distinction between superiors and subordinates in government offices.
I have, from time to time, served as an ethics bureaucrat myself. As a sometime member of the Politics Ethics Committee at my university, and a former member of a university-wide ethics committee, I've seen quite a few ethics applications, and those applications have made it clear to me that the strictures placed upon researchers are growing far beyond those in the Tri-Council guidelines, through self-censorship, as applicants impose limits upon themselves in an effort to forestall any possible future repercussions. In the process, concern with proper procedures of scientific investigation are taking a back seat. In particular, I have two concerns about items that I've seen repeatedly in ethics applications.
The first item is a promise to allow respondents to look at drafts of the research and to take their comments on board before doing a final draft. This is quite simply bad science. Scientific findings should be the truth, not a compromise document that offends no one. To be sure, it's reasonable to let people who have helped us with our research take a look at our findings, and correct any factual errors that may have slipped in. But it remains the researcher's responsibility to distinguish between verifiable correction of facts and "cleaning up the record" to make someone look better.
If we're genuinely concerned with scholarly ethics, ethics applications should contain no reference to anyone besides the researcher having a say in the final draft. It's university researchers who enjoy the protection of tenure, and it is they, and they alone, who should bear the responsibility that goes with it. Like Harry Truman, we ought to have a sign on our desk that says, "The buck stops here."
My second concern is an item that appears routinely in applications, though I've seen no reference to it in the Tri-Council guidelines: A promise to destroy data after five years. This is another instance of bad science. A fundamental of the scientific process is verifiability and falsifiability of data. Data that have been destroyed can neither be verified nor falsified.
Records that contain personal information which can be linked to individuals who are not in the public arena should either not be kept in the first place, even for a month, let alone five years, or be kept under lock and key. If records are properly kept, there can be no valid reason for destroying them. On the contrary, they must be kept, so that future researchers will be able to verify them independently of the original data-gatherers.
Nothing is more important to social science research than ethics, but the guidelines that govern Canadian politics and public policy research inadvertently do more to undermine research ethics than to support them.
The Northwestern University Law Review has devoted a special issue (101 ) to research ethics and institutional review.
For some other perspectives, go to Zachary M. Schrag's Institutional Review Blog.
I have put this topic up for discussion on the H-Urban e-mail list, under the subject line, "Does ethics review pose a threat to critical research?". For my post and responses to it, go to the H-Urban archive, and view the archives for September 2007.
For a quick overview of what's happening in the United States (a possible foretaste of what's in store for us) take a look at a recent New York Times article.