The Risks of Assessing Ethical Risks
by Michael Rustin
University of East London
Sociological Research Online, 15 (4) 18
Received: 15 Nov 2010 Accepted: 20 Nov 2010 Published: 30 Nov 2010
Introduction1.1 We first need to differentiate, in considering the ESRC’s Framework for Research Ethics 2010, between issues of principle and issues of practice. In regard to the former, the issue is not whether research should be conducted according to ethical principles, but rather whether there should be some specified form of accountability for this, both in the submission of research proposals, and in the subsequent conduct of research. There are some researchers who hold that while research should self-evidently be conducted in an ethical manner, especially in regard to the potential risks of harm to research subjects, the inconvenience, costs and possible damage brought about by requirements of formal accountability outweigh any likely benefit from them. (Hammersley 2006). Such critics point out that the risks of research in the social sciences are much less than in other spheres, such as medical research, from which the pressures for ethical regulation have derived much of their impetus. FRE 2010 itself draws attention to this lower element of risk in social science compared with medical research, while nevertheless setting out extensive and mandatory protocols for its ethical regulation.
1.2 The idea that, while research should necessary be conducted on an ethical basis, nevertheless formal procedures of accountability for this are unnecessary, implies that trust can and should be placed in authorised researchers. We can think of this as trust in them as professional social scientists to conduct research in ethical ways, not least since professional identity has an inherently ethical dimension. And also that one can reasonably trust those empowered to assess research proposals to take ethical issues into account in the normal process of making their assessments, especially if their attention is specifically drawn to their obligation to do so as the ‘ethical’ dimension of an assessment protocol.
1.3 These are by no means unreasonable ideas, in themselves. Where all those submitting research proposals of any kind are required to take account of the ethical issues raised in their projected research, it seems reasonable to believe that competent assessors and supervisors can be trusted to make appropriate judgements about ethical risk, and give or withhold approval, or require revision of projects, accordingly. In so far as the ESRC under its 2005 FRE did allow some projects to avoid specific ethical scrutiny, following an explicit submission that this was unnecessary, this was the previous ESRC position at least for some research projects. Liz Stanley and Sue Wise’s objection to the obligatory reference of all ESRC research proposals to Research Ethics Committees is related to their view that competent ESRC assessors can themselves take sufficient account of ethical issues.
1.4 Where research projects are being undertaken as part of Masters or undergraduate courses in social sciences, as is now often the case, this ‘professional trust’ is now widely accorded to academic supervisors, and sometimes where research projects are required to have more formal approval, to assessment boards. It is assumed in such cases that a well-functioning system of academic teaching and supervision will take account of ethical risks to research subjects, and ensure that these are avoided. There are highly practical reasons for taking such a view, since the costs in time and resources of introducing formal ethical scrutiny for every piece of masters or undergraduate level research undertaken as part of degree programmes would be very substantial and harmful. It is worth drawing attention to this dimension both to indicate that trust in the competence of academics does continue to have an important place in the larger field of ‘research’, broadly defined, but also to indicate the dangers that lie along the road of insisting on formal ethical accountability for every form of investigation whatsoever.
1.5 The promulgation of the ESRC FRE 2005 document, and its revision in 2010, of course indicates that trust in the competence and ethical commitments of researchers and those who determine on scientific grounds whether projects are to be supported is no longer deemed to be a sufficient safeguard against risk. One should note that neither in the 2005 nor 2010 FRE is any empirical evidence given for this view, in the form of reports of actual harms suffered in consequence of research projects having been approved without formal ethical scrutiny. (Some reference is made in FRE 2010 to information being internally available about these matters, but no indication is given of what this was.)
1.6 In the absence of such evidence, one surmises that the reasons for the adoption of these Frameworks lie elsewhere than in catalogues, or even notorious single cases, of actual harms arising from lax ethical consideration. It seems likely that the perceived problem for the ESRC was a different one, namely the risk that it would be seen to be neglectful of ethical considerations that other public bodies (such as the NHS) had long been obliged to take seriously, and might be vulnerable to attack for this reason. The ESRC might have felt that social scientists, and thus itself, would be particularly vulnerable to criticism, should any harm be discovered to have resulted from ESRC-funded social science research. While the stated concern of FRE 2010 is of course to protect members of the public against ill-conducted research, another precautionary concern might well be to protect social scientists themselves from the damage which they and the ESRC could suffer, should researches ever be shown to have caused harm. It is possible that the choice of 2010 for the publication of this latest FRE was not altogether random, since the election year of 2010 was always likely to begin a new period in which ESRC might have again to defend its credibility with governments and publics.
1.7 Although the recent affair of the leaked climate change emails at the University of East Anglia was not directly of concern to the ESRC, it certainly does constitute a cautionary tale for any institution which supports research which is both controversial, and which has powerful enemies.
The Social Context of Regulation2.1 The pressures for the precautionary ethical regulation of research can best be understood in the context of much larger social trends than those concerning research alone. One is the expansion of audit and inspection procedures in almost all fields of life. This tendency has been especially strong in institutions subject to regulation by government and its subsidiary agencies. A trade-off has been made, in many parts of the public sector, between on the one hand the conferral of greater autonomy upon providing agencies (e.g. privatised utilities, schools, universities, NHS Trusts), and on the other hand the subjection of such agencies to statutory regulation, via entities such as Ofsted, Monitor, The Quality Assurance Agency for Higher Education, etc.. Where market or quasi-market competition is deemed to be unable to maintain or improve standards for consumers or service-users, regulation has been widely adopted as its preferred complement. Sometimes this is to restrict monopoly power (e.g. in the case of utilities), and sometimes it is because the consequences of failure by participants in competitive markets (e.g. for the patients or pupils of under-performing NHS Trusts or schools) are unacceptable for reasons of equity, even though competition itself has been chosen as the main engine for bringing about improvements in average standards.
2.2 Although this regulatory machinery has become most often discussed in connection with the governance of the public sector, as ‘the New Public Management’ (McLaughlin et al 2002, Hood 1991, Dunleavy and Hood 1994, Hood and Peters 2004, Newman and Clarke 2009) it is also a prominent element in the management of corporations operating in the market place. Large firms such as Toyota or Amazon maintain the consistency of their operations through continuous audit and inspection of performance, in which solicited feedback from customers is one of their sources of information. One of the anxieties moving this process forward has been the desire to find means of ensuring rising efficiency and increased value-for-money in public services which it is feared will otherwise constitute an increasing and non-productive burden for the economy. Although the ethical regulation of research may seem a particular function, not closely linked to broader spheres of performance management, its expansion does seem to be part of this much wider growth in the audit culture.
2.3 It seems that systems of audit and regulation, once established, are subject to an inherent logic of expansion, especially in the public sector where there seem to be few constraints of cost to inhibit them (Power 1994). One explanation has been found in the drive to homogenisation and standardisation of all procedures and practices which seems to be endemic in these system. It can be seen to be an aspect of the autopoeitic production, in Niklas Luhmann’s terms, of an institution’s sovereignty over the processes which take place within its boundaries (Luhmann 1993, Rustin 2004, Strathern 2000). Nothing can be left to chance, or to group or individual discretion, in a well-functioning modern organization, if rules of conduct and accountability for it can be specified. It calls for a synthesis of Max Weber’s ’iron cage’ with Foucault’s ideas of governmentality, to comprehend these processes. Inspection regimes invade the mentalities of the primary providers they regulate, as their demands are internalised, as well as responded to by calculation.
2.4 One justification of these systems of regulation is that they protect customers, consumers and clients of these services. The methods of the new public management are legitimised by an ultimate conception of consumer sovereignty, which is deemed to be endangered by the powers of monopolies, bureaucracies, or professionals. The paradox is that in a society increasingly regulated and subject to surveillance, the individual is upheld as the primary source of value. Injury to individuals, such as those arising from failures of public services, are defined as injustices, and populist rage can be mobilized against those institutions and professionals held responsible. This is the ‘risk society’, (Beck 1992, Luhmann 1993) in which the more vulnerable and fragile individuals feel, the more intrusive the measures of surveillance and regulation that can legitimately be adopted to protect them. The ethical regulation of research is a part of this system, set up to protect individuals from the possible harm that might ensue to them as a consequence of research activity. This public sensitivity to individual harms does not of course stand in the way of governmental actions in the sphere of for example employment or housing which may bring serious harm to many thousands of citizens.
2.5 Even given this context, it is puzzling why these systems of regulation should have become so expansive and invasive. Perhaps it is because while the direct and up-front costs of operating a system of regulation or quality assurance seem to be small, as a proportion of the total budget of a public enterprise, the indirect and ‘exported’ costs of these systems are so widely distributed across organisations and their members as to be largely invisible to decision-makers at higher levels. Whose responsibility is it to assess whether the burden of a regulatory requirement is excessive, or not? Suppose the start of a research project is delayed by as much as six months, by ethical approval procedures, who is counting, except for the unfortunate researchers? For all the emphasis in the modern public sector on ‘efficiency’ (as in the ‘efficiency savings’ which institutions in the NHS and in higher education are expected to achieve every year), one has rarely heard of ‘efficiency savings’ being sought in the sphere of quality assurance. Regulatory systems have now become widely ‘internalised’ within institutions, as mechanisms for their own governance. If institutional success is going to be assessed by public indicators of performance (for example through various kinds of ‘league tables’ ) then institutional managements will deem it in their own interests to quality assure and regulate everything in sight, even where, as sometimes happens, higher authorities have become less demanding in these respects.
2.6 It seems that public sector institutions find themselves in this respect in the worst of all worlds. Profit-seeking enterprises have to be most attentive to the economic outcomes of their activities, and therefore have reason to regard expenditures on quality control as one investment to be balanced against others in terms of the returns expected. But the plurality of goals set for public enterprises, however marketised they may be, means that the threat of failure takes many forms, such that no expenditure to avoid risk can seem excessive. We have gone from a situation when governments entrusted professional bureaucracies to deliver services authorised by legislation, and were concerned mainly with safeguards against corruption (Du Gay 2000, 2005), to one where public service-providers are regarded as innately self-interested, needing to be continually harassed by imposed performance targets and by threats of public exposure if they are to achieve what is asked of them. ‘Full marketisation’, such as is now being enacted for the university system, or for NHS Foundation Trusts, can then even come to seem a libertarian solution to the problems which ‘new public management’ has itself created, at least for those institutions who envisage their competitive position in the ensuing market to be a strong one.
2.7 The consequence of this development has been that in schools, universities, and NHS Trusts, and many other public institutions, there has been a huge expansion in recent decades in the resources which are required to be spent in meeting regulatory requirements. That is, in maintaining records, preparing for inspections, and operating regulatory committees. Not to mention the psychological concomitants of these procedures, in the distracting preoccupations and anxieties which they generate. (Rustin and Rustin 2009). All of these activities necessarily have an opportunity cost - that is to say that they have to be undertaken at the expense of time and effort devoted to other activities. The ‘other activities’ in question are mostly what used to be regarded as the primary tasks of these institutions, such as caring for patients, teaching students, undertaking research, or keeping up with the literature. It seems that there is a very widespread feeling among those who work in these settings that these regulatory burdens have become excessive. So strong was this feeling that in the recent General Election the parties which have now formed the Coalition pledged to restore trust in professional employees, and to reduce the regulatory burden upon them, though whether this will happen in practice remains to be seen.
2.8 One striking fact about this development is that although it has been largely sanctioned by the idea that practices should be shaped as far as possible by empirical measures of their effectiveness, systems of regulation have themselves remained largely immune from such measures. There has been little audit of the consequences of audit and inspection, though it would be possible to devise measures which compare the different ways (e.g, ‘light’ or ‘heavy’ touch) in which such regimes are conducted. Also notable is the failure to make use of systems of audit to develop new understandings of how the inspected systems work and could work, that is to say as means of research in their own right. This is because audit regimes are generally designed merely to measure compliance, by established norms, not to detect originality or to find hitherto unrecognised causes of success of failure.
Questions of Practice3.1 Liz Stanley and Sue Wise (2010) draw attention to three new features in FRE 2010, compared with its earlier 2005 version. The three new features to which they draw attention in FRE 2010 are indeed significant. These are (1) the requirement that all ESRC funded proposals need to receive ethical consideration by Research Ethics Committees. (2) the reconfiguration of the RECs set up following the 2005 FRE to include all research proposals accepted by the ESRC and other funding bodies and (3) the dependence of funding on REC review, with its purview extending throughout a project’s life.
3.2 But while these are significant extensions of regulation, and can readily been seen as aspects of ‘ethical creep’, I am somewhat less concerned about these specific changes than Stanley and Wise. One reason for this is that I understand the function and purpose of FRE 2010 as largely precautionary in its intention. The ESRC appears to have asked itself, was there any incompleteness in our original proposals that leaves us prospectively vulnerable to criticism? Thus, if ESRC holds ethical consideration to be a specialised function, needing to be distinct from scientific assessment, how can it be seen to exempt some proposals from this, at the stage of scientific evaluation? Or, if ESRC regards it as necessary for research to be undertaken in an ethically accountable way, must it not extend this accountability in principle to the actual conduct of research, rather than merely to its prior design and approval? Once can see an administrator’s drive for consistency and comprehensive specification in these revisions, as if there were a Platonic Ideal of ethical accountability to which practices should theoretically be brought into accord.
3.3 It may also have seemed anomalous that university and other institutional RECs might find themselves considering individual PhD research proposals, as a matter of course, while much more substantial externally- funded research programmes, like those supported by ESRC, could avoid such consideration. How could it be defensible for an institution to be giving ethical scrutiny to very minor pieces of student research, while omitting to give this to major projects? If there is going to be such a system, the argument may have been, it is logical that it should be comprehensive in its scope.
3.4 It can sensibly be argued that ethical issues can properly and responsibly be examined as one aspect of scientific, or indeed in the context of research training, supervisory, competence. In the social sciences no good empirical evidence has as far as I know come to light to suggest that this earlier assumption was unwarranted. However, the fact is that this argument has been lost, and it has now become the conventional wisdom that ethical issues must be dealt with as a specialist function, in response to the broader cultural changes referred to above. That being the case, the practical problem is one of finding a sensible balance between the risks of harm that might result from research activity, and the considerable and sometimes disproportionate costs of insuring against these risks through the procedures of ethical scrutiny. It should be noted that the much-maligned element of ‘trust’ is not removed whatever procedure is adopted, but merely removed to a different specialised institution.
Research Ethics Committees4.1 It is clear that according to the ESRC 2010 Framework, and in the real world of today, the crucial institutions and processes are those of the university and other RECs, and it to the operations of these that practical attention must be given. FRE 2010 seems to me to be more pragmatic in this respect than Stanley and Wise allow. In one part of the document at least, it does provide for RECs to be constituted on a Faculty or discipline basis, so far as their detailed consideration of projects is concerned, and proposes that second-tier RECS working at an institution-wide level should confine themselves to broad oversight, to a quality assurance function in effect, and should not engage in consideration of specific research proposals. The provision for review of ongoing projects suggests only ‘occasional’ investigations, against suggesting that the ESRC is anxious to cover every theoretical eventuality, without wanting to submerge research under a burden of continuous surveillance. (One could make a logical case for researchers having to report every year – every three months – on their own ethical conduct, but even the most anxious regulators can see the need for some proportionality).
4.2 Stanley and Wise (2010) are also concerned about the overlap between the ethical and the scientific evaluation of proposals, evident in the requirement for the full specification of a project as a basis for its evaluation. Once again, as with all these criteria, one can see the real danger of ‘mission creep’, as zealous RECs, given privileged material support for their activities, may choose to revisit the scientific evaluation which a project has already received. But it is also the case that without an adequate understanding of the scientific basis of a research proposal, it is impossible for RECs to sensibly assess the ethical issues they raise. Most of the problems I have observed in the consideration of research proposals by RECs, both within a university and an NHS context, have come about out through the inadequate understanding of the scientific purposes and methods of a project, or in health and social care research of its basis in clinical practices, by those seeking to assess its risks
4.3 These potential malfunctions are themselves now a serious hazard, and threaten effectively to reduce the resources available for research by diverting them into regulation. However I think these risks will now have to be addressed through the work of the RECs themselves, rather than in reformulating the ESRC Guidelines, which are in any case no more demanding than those which many RECs now routinely follow. (Indeed it is clear from FRE 2010 that its authors have been at pains to ensure their conformity with wider protocols for ethical research practice).
4.4 It would be helpful for RECs to be clearly instructed to avoid the hazards that lie in their path – e.g. exceeding their ‘ethical’ remit, making assessments without adequate specialist knowledge, causing needless delays to research, failing to seek feedback from their researcher clients in order to improve their own practices. But I doubt if there is now any way forward for researchers but to seek to make this REC system work constructively, thereby building into it the elements of trust and responsibility which are the essence of ethical practices of any kind.
ReferencesBECK,U. (1992) Risk Society. London: Sage.
DU GAY, P.(2000) In Praise of Bureaucracy. London: Sage.
DU GAY, P. (ed) (2005) The Values of Bureaucracy. Oxford: Oxford University Press:
DUNLEAVY, P. and Hood, C. (1994) ‘From old public administration to new public management.’ Public Money and Management, Vol. 14, No. 3 (9-16)
ESRC (2010) Framework for Research Ethics (FRE) <http://www.esrcsocietytoday.ac.uk/ESRCInfoCentre/Images/Framework%20for%20Research%20Ethics%202010_tcm6-35811.pdf>
HAMMERSLEY, M. (2006) ‘Are ethics committees ethical?’ Qualitative Researcher, Issue 2, (4-8).
HOOD, C. and Peters, G. (2004) ‘The Middle Aging of New Public Management: Into the Age of Paradox?’ Journal of Public Administration and Theory, 14 (267-282).
LUHMANN, N. ( 1993) Risk: a Sociological Theory. London: Sage
MCLAUGHLIN, K. ,Osborne, S.P., Ferlie, E., (eds) (2002 The New Public Management: Current Trends and Future Prospects. London: Routledge.
NEWMAN, J.E. and Clarke, J. (2009) Publics, Politics and Power: Remaking the Public in Public Services.
POWER , M. (1994) The Audit Explosion. London: Demos.
RUSTIN, M.J. (2004) ‘Rethinking Audit and Inspection’, Soundings 26, pp
RUSTIN M.E. and M.J. (2010) ‘States of Narcissism’, inE. McGinley and A. Varchkever, A., Mourning, Depression and Narcissism throughout the Life Cycle. London: Karnac.
STANLEY, L., and WISE, S. (2010) The ESRC’s 2010 Framework for Research Ethics: Fit for Research Purpose? Sociological Research Online <http://www.socresonline.org.uk/15/4/12.html>. STRATHERN,. M. (ed) (2000) Audit Cultures: Anthropological Studies in Accountability, Ethics and the Academy. London: Routledge.