US social scientists have long complained about the impact of the Common Rule, the main federal regulation governing the ethical review of biomedical and behavioral research by Institutional Review Boards (IRBs). This was first enacted in 1991 and last revised in 2005. In 2011, the relevant federal agencies announced a review, leading to the publication of revised regulations on the very last day of the Obama presidency. An international policy community has closely followed these US debates because of their influence on the frameworks for ethical regulation established elsewhere.
A 2014 report of the National Research Council of the National Academies (NRC) and by the draft regulations issued for consultation in 2016 raised hopes of sensible reform. The final drafting, however, seems to have been distracted by a major controversy over access to biospecimens. There are also signs of haste to enact regulations before the change of administration. Biomedical agendas have once again crowded out proper consideration of social science concerns.
The NRC set out a coherent approach that appropriately identified virtually all social science research as minimal risk. It should be ‘excused’ from ethical regulation on the basis that participants were well able to judge the risk and make their own decisions. A small number of experimental or intervention studies might require IRB review but everything else should just be registered. Specific consent should not be required for most studies – it could be inferred from willingness to fill in a survey or continue with an interview or focus group. Observations in public spaces, including social media, would also be ‘excused,’ as would most re-use of administrative data sets. ‘Vulnerable groups’ would no longer be listed but assumed normally to be capable of judging their own best interests. This approach was largely adopted by the draft regulations, which substituted the term ‘excluded’ for ‘excused’. IRBs would lose their jurisdiction over most social science research, unless it fell within narrow criteria or they could justify calling it in from the registration documents. There were still uncertainties about the status of participant observation or ethnography, but the approach was broadly welcomed by the community.
The final text, however, seems to have been driven by a concern to patch a few holes in the 2005 revision which, for the social sciences, did little to address the flaws of the original. It does not reflect the NRC’s attempts to introduce a more principled and proportionate approach. Some elements, like inferred consent have been taken over, but then applied inconsistently: the provisions on vulnerable groups are particularly muddled. There is silence on whether social media should considered part of the public sphere, presumably because this would have required a contestable decision to be made when the authors were already dealing with the fight over biospecimens.
At other places, the influence of particular lobbies seems to have been felt at the expense of consistency. The ‘excluded’ category has disappeared and the activities to be covered by it have been divided between ‘not research’ and ‘exempt.’ A life history interview by an oral historian, a journalist or a political scientist leading to the publication of individually identifiable material is not research. A biographical interview by a sociologist or anthropologist that will be anonymized and contribute to a generalized understanding of a phenomenon is research and will be regulated. Interviews with people involved in a lawsuit by a law professor are not research. Interviews with the same people by a social science professor are research – it is not clear what happens if the social science professor is on the faculty of a law school. Participant observation and ethnography are definitely research – but can be regulated more lightly if they are carried out in educational settings.
The removal of the ‘excluded’ category draws most social science research back into the IRB net. ‘Exempt’ is something of a weasel word: it does not mean ‘exempt’ in common sense terms but ‘exempt from full review.’ It must still be scrutinized by IRB staff and members, who will determine whether review is expedited or conducted in full. Paternalism and distrust remain the order of the day. In the original intentions of the revision, a roll-back of social science work was supposed to free regulatory staff and resources for an extension of their responsibilities in the biomedical field: the overall costs of the system would not increase. Predictably, the regulators have resisted any loss of responsibilities so that their empires could expand rather than remain static.
This incoherence suggests a rushed job to publish regulations before the arrival of the Trump administration. The work is, though, vulnerable to Republican plans to review all regulations issued in the last six months of the Obama presidency. If this document is not to be set in stone for a generation, social scientists may need to press the case in terms of the new administration’s deregulatory impulses.
Ethical regulation of the social sciences should be pulled out of the Common Rule and given distinctive attention. The NRC attempted to approach the topic in a more systematic fashion, but has been betrayed by the final text. Regulation needs to be based on the actual risks associated with particular methods, regardless of who is using them, rather than lobbying by disciplines – oral historians obviously have more friends in DC than anthropologists. In most social science research, the risks are everyday in nature and participants are better able to assess them than a distant IRB. Regulation must be proportionate to other risks – creating areas of ignorance, encroaching on First Amendment rights or academic freedom, chilling innovation, and denying participants’ rights to self-determination.
It is time to tear the house down and rebuild on stronger foundations instead of trying to paper over the obvious cracks.