Researching the Powerful: A Call for the Reconstruction of Research Ethics

Introduction

We are seeking to resolve core ethical contradictions that arise when the research subject is the corporation and/or the individuals who control and manage the corporation. ¹ Yet our call to reconstruct a framework research of ethics also applies to other powerful institutions of government and ‘civil society’ and to researching elites more generally. The starting point for our argument is shaped by our experience as researchers who are primarily interested in studying corporate crime, that is, the illegal and harmful behaviour of corporations. We have been trained as social scientists, and this training is grounded in particular professional standards, rules, and codes. Yet when we have applied those rules and codes in our own research, we find them to be inadequate and not fit for purpose. At times, this has led us to pursue research strategies that break the established ethical rules (Alvesalo and Virta, 2003; Whyte, 2007).

This is not a new dilemma for social scientists and is familiar to many who share our research interests, or to those who seek to ‘study up’ rather than ‘study down’. Indeed, questions surrounding the applicability of ethical rules and ethical dilemmas have previously been the focus of numerous debates and deliberations within our profession. Those debates have been connected to prolonged struggles over the re-framing of ethical codes (Galliher, 1980; Haggerty, 2004) or focused on the appropriateness of a range of practices such as covert research (e.g. Bulmer, 1982; Erikson, 1967; Hammersley, 2014). This article draws on those debates, focusing on how ethical principles and the concepts and the guidelines that codify them might be reframed in the context of ‘studying up’.

We begin this article by summarising the challenges facing researchers who focus on powerful institutions and elites before embarking on a brief and selective history of the study of the powerful in key social science texts. This article then analyses the dominant ethical principles adopted in professional codes of conduct. We identify a twin obsession with professional (the social scientist) and institutional (the university) autonomy in those codes that hamper the development of a research ethics that is genuinely based on the public or common interest. In this argument, we do not advocate the rejection of ethical principles for research per se; neither do we seek to undermine the importance of codes of ethics designed to protect vulnerable research subjects. Rather, we argue for a reconstruction of social science research ethics. Recently, in this journal, Emmerich (2016) has argued for a renewal of professional autonomy in the upholding of ethical standards in the social sciences. While we agree with many of the reasons for his argument, we will argue that the solution is not to be found in practices that are internal to the profession. Rather, it is to be found in our external relationships and practices outside the academe.

Research methods and researching the powerful

Before we develop our analysis of the shortcomings of research ethics, we briefly explore some of the methodological dilemmas that arise when researching powerful elites. Scholars across different disciplines, such as anthropology, sociology, human geography, and business studies, have consistently drawn attention to a number of practical difficulties associated with researching the powerful. It has been argued that one of the key features and effects of power is the ability to operate beyond public scrutiny and thus accountability (Tombs and Whyte, 2003: 4). Research that stands outside of dominant research agenda, challenges power, or merely provides support for dissenting voices can create difficulties for social scientists (see, for example, Bell, 1978; Cormode and Hughes, 1999; Friedrichs, 2010: 27–29; Mungham and Thomas, 1981; Snider, 2000). A number of barriers can be identified throughout the research process: from the design stage, through the process of gaining access and gathering data, to the dissemination of findings (see, for example, Friedrichs, 2010: 36; Smith, 2006; Undheim, 2003).

First, it may be difficult to even contemplate some powerful organisations and individuals as legitimate research subjects at the design stage, since funders of research often set research agendas that may limit or prescribe the directions of research activity. Research not prescribed by funders or research subjects may also be publicly undermined by those whose interests are at stake (see, for example, Tomiv and Trumper, 2012: 241). This has a number of subsidiary consequences, such as the demand for policy-relevant impact. It has thus been argued that the paucity of research on the powerful presents awkward questions for the claims about the ‘value-neutrality’ of university funding and state-funded research councils (Whyte, 2000: 269). Large corporations, in particular, have proven themselves adept at influencing such research agendas (Michaels, 2008; Proctor, 2008). The control of research agendas and the limiting of doing what one sees as important is, of course, a common practice even where elites are not under the gaze of research.

Second, difficulties of gaining access to powerful corporations or institutions can prevent the research from taking place. Although the ‘powerful’ tend to be publicly visible and seemingly easy to contact (e.g. via information published on websites and social media online), they have remained secretive, placing high value on privacy and exclusion (see Bell, 1978; Ozga, 2011). Gathering data from interviews, observations, and other forms of ethnographic techniques using human subjects or the gathering of quantitative forms of data is, therefore, a difficult process (see, for example, Cunningham-Sabot, 1999) and requires careful, constant negotiation (Lancaster, 2017: 99). Methods of generating data often rely on the co-operation of the corporation and access can be denied without argumentation since they enjoy almost complete rights of ownership to information about their activities. Moreover, they are not generally obliged to disclose data in the same way public bodies are. Hence, the legal constitution of corporations is designed in ways that allow it to avoid public disclosure (Tombs and Whyte, 2003: 33).

Third, if research in these areas does yield evidence of harm or impropriety on the part of a corporation or other powerful actors, it can be difficult for critical researchers to disseminate their research to the public. They may face opposition from the organisations or the actors they have researched, who may have little difficulty in raising the resources—be they financial or in the shape of alternative ‘expert evidence’—to discredit, censor, or otherwise challenge the research findings (Michaels, 2008; Proctor, 2008). Suppression may happen in the field of the ‘journal industrial complex’ (Arrigo, 1999: 1) whereby critical scholarship, in particular, is denied recognition, and a variety of unconscious and conscious processes are used to exclude certain forms of work from prestigious publication (Whyte, 2000).

In summary, powerful institutions, private corporations, and public organisations occupy a social position that enables them to repel the advances of researchers, strategically opt out of participation in research and manipulate the research process in a range of ways. It is this ability to influence both the viability and the shape of a research project that renders researchers—rather than their research subjects—relatively powerless. Thus, the power relationship that studying up involves implies a set of dynamics that does not resemble the textbook approach to research methods in which the researcher is generally in a position of relative power. Those barriers to the development of research agendas, access, and dissemination force us to change the way we think about and approach the research process. In the next section of this article, we show how this is based partly on a historical tendency in social science that has encouraged a downward rather than upward research gaze.

The methodological lacunae in ‘studying up’

Some of the most cited and debated texts in sociology have identified the powerful as their object of study. This point is made commonly in relation to Marx, Weber, and Durkheim but consider also some of the ground-breaking texts in sociology at the turn of the century such as Ross’ (1907) remarkable commentary on the social organisation of the elite, Sin and Society. Thorstein Veblen’s (1994 [1889]) work also drew on detailed empirical analyses of elite behaviour. There is a strong tradition of foregrounding the study of elites in sociology—not least American sociology—that gathers pace throughout the 20th century, illustrated most clearly in the work of Wright Mills (1999 [1956]).

Perhaps, the most famous call to ‘study up’ was developed in the widely cited Becker–Gouldner debate, in which Alvin Gouldner (1968) raised key political and ethical questions about the trajectory of the researchers’ gaze, arguing for a shift of attention from the ‘underdog’ to the ‘overdog’. Yet, as is indicated in much of the work that came on the heels of this debate, a simple redirection of the gaze may not be so simple. After all, despite the continual presence of analyses that deal with the social science of ‘elites’, the empirical methods used in social science have almost exclusively been developed to face downwards rather than upwards.

While various movements in social science have from the late 20th century onwards been in a position to develop a range of segmented empirical approaches to studying down (studies of working and the labour process, studies of the changing family structure, and studies of ‘deviance’), social scientists interested in the powerful were, and are, still exploring a very general and broad object of analysis that does not have obvious sub-disciplinary loci. Some disciplines related to the social science tradition have developed methodological systems that do not sideline the study of elites as a peripheral object of study. In political science, the study of power is a central problem for the discipline. There exist well-developed techniques for understanding how political decision-making processes are established (‘process tracing’) as well as an established (if small) literature on elite interviewing. There is also a large body of work on researching elites in human geography (e.g. Desmond, 2004) and business studies (e.g. Welch et al., 2002). In anthropology, the study of elites has also been present (Hertz and Imber, 1995). Much of this work in anthropology has, in the context of the discipline’s tradition, been concentrated on kinship and tribal elites in traditional societies (Marcus, 1983). Yet as early as the 1940s, anthropology had not only developed a political focus which was largely focused on new political systems adopted by traditional societies but also opened the door for studies of Western elites (e.g. Weatherford, 1981). Post-colonial anthropology has retained a focus on bureaucratic governmental and economic elites (Cohen, 1981; Ferguson, 1994; Luhrmann, 1996), and a new generation of anthropologists have turned their attention to institutions of power at a global level (e.g. Galtung, 1986; Wedel, 2001). However, within this discipline, researchers have been slow and inconsistent to heed Nader’s famous call in 1972 for anthropologists to ‘study up’ (Aguiar, 2012; Nader, 1972); the discipline has generally continued to focus on ‘giving voice’ to marginalised others (Ho, 2012: 36).

Although there are some disciplinary exceptions, the study of the powerful has not been mainstreamed across the social sciences. A key consequence of this is that we do not have developed tools to guide us in empirical studies of powerful elites and institutions. Thus, we only have a scattered number of studies that use the ‘powerful’ as the subjects of primary data collection (as respondents to questionnaires, as the subjects of in-depth interviews and focus groups, or as the subjects of ethnographic studies). As the brief review above indicates, we are not arguing that a methodology for researching the powerful is wholly absent. Yet, it is barely developed in a practical sense. And this lack of a developed methodology has consequences for the primary concern of this article: research ethics. The following section of this article, therefore, explores how the lack of a developed methodology forces us to change the way we think about and approach research ethics.

Ethical rules for scrutinising power

Where there have been debates and discussions about the problem of researching the powerful, the relationship between research ethics and methods tends to be left ‘between the lines’. For example, Galliher (1980) raised the possibility that equal treatment of certain subordinated minorities calls for unequal treatment of various superordinate groups and that it may be necessary to violate the rights of the powerful in order to protect and advance the powerless. In a similar vein, Spencer (1982) argued for a principled form of rule-breaking in researching the elite since often deception was necessary to obtain results (see also Lilleker, 2003).

The ethical paradox that this raises can be summarised in the following logical progression: the ethical rules established for conducting social research seek to address the imbalance of power in the research process; those same rules protect all research subjects, regardless of their social position or status; the rules, therefore, protect relatively powerful individuals who are under scrutiny by researchers. In order to deepen our analysis of the problem that we face here, our argument is that we have to move beyond an understanding of the immediate research relationship—for it is only by moving beyond the research relationship that we can see power relationships in their totality. A research respondent in a research project focusing on poor housing conditions, for example, may be vulnerable in a research ethics sense. But this respondent was not made vulnerable by the research setting in the first place. Vulnerability is structured in all kinds of other, prior, ways.

If ethical principles and standards are really supposed to protect the relatively powerless, then they must address power relationships that exist outside a rarefied and synthetic research environment. The question for us, then, is why would a research project that challenges the validity of power relationships in the field of its study not also challenge the validity of power relationship as they are constructed in a research sense. It is for this reason that we argue that the analysis of this ethical paradox must go deeper into critiquing the structure of power relations in a much more social sense, in one that extends beyond the immediate research relationship. It is this approach that we take in the discussion that follows, as we revisit some key principles of research ethics.

Every undergraduate sociology student is tutored in three key ethical principles. These principles are the bedrock of the ethical codes that social scientists use to guide their professional conduct.

First, the harm principle, where steps must be taken to minimise ‘harm’ (normally referring to psychological or physical harm to participants). The second principle is that researchers should guard against the invasion of privacy. Anonymity and confidentiality guarantees are given either unconditionally, or sometimes on request, as a prerequisite of research. The third principle is the prevention of deception, whereby the researcher must avoid deceiving the research participant. In order to minimise deception, researchers apply the related principle of informed consent where research subjects are given as much knowledge about the research as is necessary to allow them to make an informed decision about their participation.

As we have indicated already, those principles are based on a particular concept of power in the research process and assume relative powerlessness on the part of the research subject. Yet, as we have seen, elite research subjects generally have the social or organisational capital to influence the research process. This structurally privileged position makes the ethical principles outlined above look very different. In order to develop a more detailed analysis of those three key ethical principles, let us briefly look at what they look like in the context of the broader social capital that elite research subjects, and in particular, corporate actors, are able to draw on.

First, we might well ask what the harm principle means when research subjects are in a strong position to generate or ameliorate social harms. By this, we mean, not simply those that generate particular harms in particular situations but research subjects who belong to larger organisations that have some control over the harms that we experience. How, for example, might the harm principle apply to research interested in understanding the way that corporate executives engage in denial about the environmentally harmful effects of particular production processes? A researcher engaged in such a study might legitimately regard the significant of generating knowledge about the legitimisation of environmental harm as good reason for suspending the ‘harm principle’. In this context, the harm principle moves beyond from being merely a question of power in the immediate research relationship, to a question of social power. The harm principle, bounded as it is by the narrow confines of the research process, is thus wholly dissected from the environmental harm, which may be the subject of the research. In short, a key ethical question that is raised when understanding the immediate research relationship in a broader social context is: what happens when the subject of research is the harm perpetrated by the powerful actors that the harm principle purports to protect?

Second, if we understand the issue of privacy within the broader social context of the immediate research relationship, a rather different issue arises. Very often members of elites who are research subjects enjoy prominent positions in public life as elected representatives, corporate executives, or functionaries in government (see, for example, Aguiar, 2012; Alvesalo and Virta, 2003; Snider, 2003: 62). They may have a public profile for precisely the same reason that they are research subjects. This may mean that their identity cannot always be preserved. But a trickier issue is whether—in the public interest—it should be preserved. Why should a corporate executive who works for a polluting company or a representative of a corrupt government department be regarded as conducting ‘private’ activities and, therefore, have their privacy protected? Our current definition of what constitutes a public or private domain is shaped by a legal formalism that constructs government as ‘public’ and corporations as ‘private’ actors. Corporations are able to draw on this formal distinction to define what they do as ‘private’ or ‘commercially confidential’ and so on—and, therefore, beyond the reach of some of the mechanisms of accountability that public bodies are exposed to. But should the protections afforded by this legal formalism extend to research ethics?

Third, what does the prevention of deception mean when we move beyond the immediate context of the research process? As we have seen, the powerful institutions and individuals that we research are very often able to control, distort, or shape the research process and its outcomes in a process that is itself very often deceptive. A valid question in this context, then, is whether a researcher, for example, interviewing corporate executives could be anything other than deceptive (pretence to gain access, pretence in interviews, etc.). A slightly different question relates to how in this context, research could be understood as a corrective to the systematic organisational deception that studies of elites seek to understand and expose? If we return to our example of the researcher seeking to understand how corporate executive legitimise environmental pollution, the research itself seeks to understand how elites systematically engage in deception. In such a context, does this mean we can legitimately develop a new set of research practices that are themselves deceptive? What might this mean for an ethical approach to research?

By placing the research relationship in a broader context of power relationships, we are not simply asking questions about the research process. That is, we are not segregating the process of research from a broader problem of power. Researchers are socialised to look at the world in a particular way: one in which their profession is given validity and one in which their enterprise makes sense. If ethical rules seek to establish the professional validity of the social scientist, the purpose of codes of ethics is to formalise this process. Yet if we place the codes of conduct that establish this ‘validity’ or make sense of the enterprise of the social researcher in this broader context of power, the principles that underpin the discipline begin to appear irrelevant. And this becomes a problem for the social scientist researching elites. For, if the principles underpinning the discipline are indeed irrelevant for the purpose of researching the powerful, what does that say about the ability of the profession to adapt to this purpose?

Our argument is that codes of ethics produced by different professional organisations representing social scientists have to be re-aligned to take account of the issues raised in this article. ² The following section of this article uses an analysis of a select group of those codes of conduct to deepen our understanding of the key claims around which the professional validity of the social scientist is formalised.

Codes of ethics and researching elites

A full list of the codes of ethics in this section we refer to can be found at the end of this article. Those codes of ethics have been selected because they are the codes we have used as a reference point to guide our own practice. All of the codes we reference are produced by the key social science professional organisations. All of those codes in some form set out the grounds for suspending the normal ethical rules of research when they are applied to researching the powerful. We have selected the parts of those codes that directly address dilemmas of researching elites and powerful groups for analysis and those parts that indicate circumstances under which the normal ethics rules of research might be suspended. The section that follows is, therefore, based on a selective overview of those codes, rather than a systemised content analysis.

Conclusion

This article has reflected on the ethical paradoxes faced by those researching elites within corporations and other organisations. As the preceding sections indicate, discussions of ‘researching the powerful’ do feature in debates about the application of ethics and feature prominently in codes of ethics. Indeed, it is common for the codes of ethics produced and promoted by the key professional organisations that represent the interests of social researchers to at least acknowledge institutional and organisational power. But as we have also seen, such forms of power are represented in very particular ways.

It is in the failure to present a nuanced understanding of how power shapes the entirety of the research process that codes of ethics underestimate the degree to which there are contradictions in the research process. As we have seen in the preceding discussion, there is some distinction made across different types of research subject in reference to a general notion of power. But those problems of power are not specified or developed in codes of ethics. The danger is, therefore, that a moral and practical equivalence is established between a range of research subjects that are fundamentally unequal in character. We question the proposition that elites and the organisations they work for should be given the same protections as vulnerable research participants. Institutions with significant control and influence over our lives cannot, when they are research subjects, expect the same protections as victims of crime, or homeless people, or users of a leisure service, for example.

This assumption of moral and practical equivalence between relatively powerless and relatively powerful research subjects also fails to recognise fundamental differences in the substantive concerns of research into elites. Those of us who research the powerful are generally not interested in their personal lives. Vulnerable research participants are generally included in a particular study because of an experience, or experiences, they have had in their lives: drug users, prison inmates, residents of a particular housing area, users of social services, or whatever it might be. When they are research respondents, the researcher wants to know about experiences that are directly connected to their personal lives. We want to know about things that are, by definition, personal. Relatively powerful respondents tend (and this is a tendency, not an iron rule) to be included in studies because of their professional role (as corporate executives, politicians, public servants, etc.). This is an important observation to make because in this context, research ethics end up protecting the professional life, rather than the personal life of the powerful research participant (while we recognise that revelations of the former can affect the latter; Lancaster, 2017: 99–100).

And precisely the same effect occurs in the context of the researcher. After all, the researcher is also acting in a professional capacity. This is of crucial significance to how we conceptualise the ethics of research since codes of ethics and rules are not merely about establishing validity for research conduct, they are also about how the researcher’s profession is validated and how the researcher makes sense of her enterprise. Indeed, as the examples recited above note, the sets of ethical rules established in sociology have the effect of establishing a professional autonomy that is based on the separation of the working life of the researcher from the rest of researcher’s life. By doing so, a separate set of rules of conduct are established for ‘life’ and ‘work’.

It is for this reason that it becomes crucial to ask: what kind of public interest is it that we want beyond ‘the job’ of doing research in a university or another institution? What are the ethical principles that should guide us generally, and not just guide our profession? In working through those questions, we argue for a public interest that supports a socially sustainable and socially equal world. In other words, our understanding of the public interest is wedded to collective and universal interests rather than narrow individual interests (Bozeman, 2007). From this perspective, we would argue that because of their social impact on us, it is very clearly in the public interest for corporate activities, decisions, and strategies to be as widely known as possible. Similarly, preventing corporations from hiding their most socially damaging activities—indeed doing everything we can to ‘know’ about those activities—is in the public interest.

Therefore, we argue for a code of ethics that is collective and universal: one that does not distinguish between the academe and the community, or between those within a particular profession and those outside it. Our collective interest in researching and knowing about corporate activities is clear. Corporations have the potential to defraud us, maim, and kill us on a huge scale. Because of the power that corporations hold over our lives and our life chances, we all have an interest in research that seeks to uncover what corporations do and how they do it. Knowledge of corporations is far too important for researchers to keep to themselves. Our research must be made public, and therefore, our ethical approach to the study of corporations should not be one that applies only to the narrow confines of the university but to all that have an interest in our research. Our key task as researchers of power is not to seek to further our narrow institutional or profession interests but the wider collective interest. An ethics of solidarity, rather than a narrow professional ethics, is, therefore, the most viable way to reconstruct a meaningful and pragmatic ethics of researching the powerful.

What we mean by an ethics of solidarity is an ethics that reaches out and connects to the communities that we live in as well as the communities we work in. Our ethical standards should, therefore, be consistent with the ethical standards upheld by the people and the organisations that we work with to expose power. In our work with trade unions, campaign groups, and other social movements, we have much more in common with their ethical commitments than we have with the ethical commitments of our own profession.

In contradistinction to Emmerich’s (2016) contribution to this journal, we, therefore, do not see the solution to the ethical dilemmas that we develop in this article in the expansion of the professional autonomy of social science (although we do support his attempt to challenge the domination of biomedical ethics). Rather, we argue that as social scientists we will need to cede some of our professional autonomy in order to break down the barriers between the ethical standards we embrace in our role as social scientists and the ethical standards we embrace as members of society. Therefore, we call for the reconstruction of the ethical codes designed to guide social scientists, based on a collectivist understanding of the ‘public interest’ that is not exclusively defined for and by the academy but connects to all groups who want to shed light on the closed-off worlds of the powerful.