Ethics review and freedom of information requests in qualitative research

Abstract

Freedom of information (FOI) requests are increasingly used in sociology, criminology and other social science disciplines to examine government practices and processes. University ethical review boards (ERBs) in Canada have not typically subjected researchers’ FOI requests to independent review, although this may be changing in the United Kingdom and Australia, reflective of what Haggerty calls ‘ethics creep’. Here we present four arguments for why FOI requests in the social sciences should not be subject to formal ethical review by ERBs. These four arguments are: existing, rigorous bureaucratic vetting; double jeopardy; infringement of citizenship rights; and unsuitable ethics paradigm. In the discussion, we reflect on the implications of our analysis for literature on ethical review and qualitative research, and for literature on FOI and government transparency.

Introduction

More than a few scholars across the United Kingdom and Australia have contacted us asking if the ethical review boards (ERBs) at our Canadian universities were requiring us to complete an ethics application for research involving freedom of information (FOI) requests. These scholars asked how we would approach such an application for ethical review. The ERBs at their respective institutions have begun to treat FOI requests as risky and in need of independent ethical review. Given the mission creep of ERBs, what Haggerty (2004) has labelled ‘ethics creep’, we have no doubt that some ERBs will try to regulate FOI use, especially those with ERB members who have little experience using this approach. The reason that these scholars contacted us is presumably because we have published on the methodological challenges of using FOI, proposed conceptual frameworks for framing FOI processes, and have secondarily addressed the theme of ethics and FOI (Brownlee and Walby, 2015; Luscombe and Walby, 2017; Luscombe et al., forthcoming; Walby and Larsen, 2011, 2012; Walby and Luscombe, 2017). Assuming a pattern, one that may eventually creep into our own universities in Canada, has prompted us to address this perhaps good-natured, but utterly misled, trend.

FOI laws exist in over one hundred countries. These laws allow citizens to request information that would not otherwise be publicly accessible from government offices (Brownlee, 2015a; Jiwani and Krawchenko, 2014; Oltmann et al., 2015). Most FOI regimes limit the right to citizens, but in some countries, such as the United States, anyone can make a request regardless of citizenship. Often there is a fee associated with requesting the information and processing the disclosure. Documents are reviewed by FOI analysts, and where necessary redacted prior to being released. FOI laws contain sections that detail the reasons that a government agency can withhold or redact information. The benefit of using FOI requests when studying government agencies is that the researcher is able to directly access insider government records. There is less spin and bowdlerization compared to political speeches, government website material or interviews with government officials. As Savage and Burrows (2007) observe, the validity of surveys and interviews as a source of data is increasingly being called into question, and for good reason when it comes to research on government offices. Government agents may take interviews as opportunities for impression management, whereas the internal records disclosed through FOI are created for use in government by bureaucrats who need these texts to carry out their everyday activities.

Along with others in the FOI research community, we employ FOI requests as a way of ‘rethinking the repertoires of empirical sociology’ (Savage and Burrows, 2007: 895) and the social sciences more broadly. In our research, we use FOI requests in Canada and the United States to produce data on public police, security and correctional agencies (Lippert et al., 2016; Luscombe and Walby, 2014, 2015; Monaghan and Walby, 2012). More and more researchers are doing the same (Bows, 2017; Johnson and Hampson, 2015; Monaghan, 2014; Sheaff, 2016). FOI users in the social sciences are guided by the belief that researchers have a moral, ethical and professional duty to ‘study up’ (Dafnos, 2011; Nader, 1974), empirically documenting the workings of elite power, authority and governance, a project which requires a different relational and professional ethic than ‘studying down’.

Although some existing literature has examined the implications of FOI for university researchers, particularly researchers subject to FOI requests from outside the university (Wilson, 2011), no works have reflected on ERB regulation of FOI requests. Canadian ERBs have not yet attempted to regulate research using FOI requests, presumably because of this statement in the Tri-Council Policy Statement on Ethical Conduct for Research Involving Humans:

Research that relies exclusively on information that is publicly available, or made accessible through legislation or regulation, does not require REB [Research Ethics Board] review. Exemption from REB review for research involving information that is legally accessible to the public is based on the presence of a legally designated custodian/steward who protects its privacy and proprietary interests (e.g., an access to information and privacy coordinator …). (Canada, 2014: 16)

However, ethics policy is a fluid construction, and this policy in Canada could easily change as ERBs in Canada seek to emulate developments in other countries (Kitchin, 2003; Lowman and Palys, 2000, 2013; Stone, 2002). As the final sentence of this passage indicates, data generation through mechanisms such as FOI is excluded from ERB regulation on the grounds that procedures for ethical (not just legal) review are a part of the process. In other countries where the governing documents for university research ethics do not contain such a clause, or where governing documents for ethics do not exist or only exist at the university level and therefore vary, ERBs may now be trying to regulate FOI research, as those who have contacted us from the United Kingdom and Australia indicate. Below we present four arguments explaining why use of FOI requests in the social sciences should not be subject to ERB oversight. We intend these claims to be broad enough to apply to any jurisdiction where a social researcher could use FOI requests as a data production tool.

The four arguments can be summarized as follows. First, there is already a bureaucratic regime in place for managing requests and mitigating risks in every country and at every level of government where FOI laws are in operation. Because of this regime, once released these records are considered published material, rendering ERB review of research involving FOI redundant and unnecessary. Second, it would not be fair for ERBs to create a situation of ‘double jeopardy’ for FOI researchers in which they are being tried twice for the same facts, thereby violating the criterion of procedural fairness. Third, ERB review of research involving FOI would infringe upon citizenship rights if it prevented a researcher from using FOI, and could create legal liability for the university and its ERB should it rule that FOI requests are too risky to use in scholarly research. Fourth, the kind of ethical review that ERBs currently undertake represents a foreign paradigm incommensurate with how qualitative researchers conceptualize and understand their research. As Librett and Perrone (2010) contend, using the biomedical and corporate model of risk/ethics management to govern qualitative research is like comparing apples and oranges (also see van den Hoonaard, 2014). Regulation of FOI according to the dominant ethical paradigm upheld by ERBs would only further undermine the advancement of qualitative research.

We begin by reviewing relevant literature on ethical review in the social sciences that informs our work, notably literature on ethics creep and works that are critical of the current paradigm of ERB regulation. Second, we unpack our four arguments against ERB review of research involving FOI requests. In the conclusion, we reflect on what this intervention means for literature on ethical review and qualitative research as well as for literature on FOI, government transparency and accountability.

Literature on ethical review

There is a sentiment amongst qualitative researchers that ethical review conducted by ERBs does not correspond with constructionist, interactionist and interpretivist conceptions of the research process, and that ERBs limit qualitative research. As Hamilton and van den Hoonaard (2016) argue, the idea that ethics review protocols are standardized and effective at keeping people safe are mythical and should be challenged. Despite the many differences between research methods and paradigms today, research ethics review in North America and beyond ‘trawls all research involving humans into the same net’ (Hamilton and van den Hoonaard, 2016: 6).

Ethical review in the social sciences and humanities should be conducted differently than in the biomedical and natural sciences (Doyle and Buckley, 2017). Although ethical review has certainly ‘evolved from specific biomedical concerns’ (Burr and Reynolds, 2010: 128), the preoccupation with risk continues to create a situation of overreach and intense governance of research. Haggerty (2004) argues that university researchers today are working in an era of ethics creep in which forms of risk management taken from policing and insurance have come to govern ethics in universities. There is a fetishization of rules and categories of risk in ERB procedures which makes the process of applying for ethics especially agonizing, and which unreasonably limits the possibilities of research design. Wilson and Hunter (2010) suggest that riskiness is often imposed by ERBs on people and research practices that do not warrant it. Colnerud (2015) reports that many qualitative researchers feel ERBs limit the scope and relevance of their research questions and design. Moreover, qualitative researchers cannot always anticipate what the next step in data collection will be (also see Guillemin and Gillam, 2004). Given the importance of ‘surprise’ and adaptability in qualitative research design and theory development, being unable to anticipate and plan is not a limitation but a strength (Tavory and Timmermans, 2014). Prefiguring the design would limit the scope and depth of the project and undermine the process of discovery. Yet ERBs treat research as if the steps can all be predetermined as in a medical experiment. To the neoliberal university (Brownlee, 2015b), not knowing all the steps is treated as a source of risk and liability. To the researcher, it is exactly this potential for surprise and ability to adapt to changing circumstances that represents the ethos of good qualitative research (Tavory and Timmermans, 2014). By going overboard with risk management in ethical review, universities undermine their commitment to research and knowledge creation. Fitzgerald (2004) likens the ethical review process to a moral panic. She suggests there is a need to intervene and educate those who set research policy and who sit on ERBs about new research innovations so that these techniques are not inappropriately regulated and discouraged. Based on ERB observations, Hedgecoe (2008: 882) has suggested that ERBs ‘are not inherently hostile to social science research, especially qualitative research’. The problem could be that ERB members lack methodological training and knowledge of qualitative research, and this may lead to unreasonable assessments of levels of risk (Hunter, 2007).

Although things are changing – for example, in the United Kingdom health research now has its own ethical vetting system, and qualitative research is receiving a warmer reception – this literature suggests that ERB review is not designed for qualitative researchers and may be limiting their work in unnecessary, unreasonable ways. There is a general concern that ERB members are using a standardized and inappropriately devised concept of risk to manage forms of research they do not fully understand, reinforcing a bias against newer, qualitative forms of inquiry to the detriment of knowledge creation and discovery. ¹ If ERBs intended to regulate FOI, as they already have at some universities, this would no doubt be the reason why. Below we extend these lines of thought and articulate four arguments for why ERB regulation of FOI requests is not needed and would harm researchers and the public good.

Ethics and freedom of information

FOI already involves a bureaucratic vetting process

FOI does not require ERB regulation because wherever it exists FOI already involves a lengthy set of bureaucratic tests and procedures that seek to uphold ethical values of informed consent, anonymity, confidentiality, and protection from harm. These tests and procedures are entrenched in FOI law and in the training and professional responsibilities of FOI coordinators, who receive requests from members of the public, task the appropriate departments with the request, and manage the redacting of the request according to FOI law and policy. This argument contains three claims.

First, FOI disclosures are considered published by the government that releases them. This is the simplest argument against ERB regulation of FOI. There is no need for ERBs to regulate access to already published material.

Second, FOI law contains sections that stipulate protections for government workers and other persons described in government records. To the extent that many of these provisions seek to uphold values at stake in ERB review, they are not only legal checks but ethical ones. In the Canadian federal Access to Information Act (ATIA), for example, section 13, titled ‘Exemptions’, lays out numerous circumstances in which government must withhold – or protect – information from public disclosure. These include files that a government agency subject to an ATIA request claims to have received ‘in confidence’ (e.g. from a foreign institution). Under these circumstances, and assuming the information has not already been made public, the government is required to either withhold the information from public disclosure or, importantly, seek the third party’s consent to make the disclosure (section 13(2)). Disclosures that could harm employees in government, undermine national security, or put the general public at risk, are withheld under FOI law and at the agency’s discretion. Under the ATIA, both individuals and institutions are protected from harm. If a disclosure is ‘expected to threaten the safety of individuals’, under section 17 the government is required to refuse to publicly release this information. Even in less extreme circumstances, information about individuals, under the Canadian Privacy Act, is protected from disclosure unless the government agency or requester has first obtained the individual’s consent to make the information public (section 19(2)(a)). There are dozens of additional sections outlining rules for document disclosure, redaction and exemption. These rules vary by country and jurisdiction. All FOI laws contain similar rules, and these rules already protect government workers and constitute a rigorous bureaucratic vetting process that is meant to uphold not only legal principles, but also ethical ones.

The third claim is about informed consent by individuals. It has been argued that informed consent for individuals who are research participants is an important part of research ethics (Colnerud, 2015). On the one hand, FOI already includes procedures geared toward protecting and informing individuals of a potential disclosure. On the other hand, the fact that FOI disclosures primarily concern the actions and communications of offices and organizations as a whole renders the idea of ‘individual’ consent somewhat a misnomer. The axiom of informed consent and others related to it, although a central part of the Tri-Council Policy Statement on Ethics that governs research in Canada and equivalent policies elsewhere, do not apply to FOI disclosures in their published format. The purpose of using FOI requests is to study an organization and its processes rather than individuals and their isolated behaviour. Information that identifies specific individuals is generally exempt from disclosure; where it is not, this person has expressly agreed to waive their right to privacy, whether through direct consultation or by virtue of their public position. Individuals are not the subjects of FOI research, and the researcher does not use this information to contact specific individuals (even if they could). The depersonalized nature of these records is also such that individual government workers do not own them or have any kind of right over them. The situation is akin to a government worker being a primary holder of secondary data that the FOI request provides access to. Few would suggest that ERBs should get into the business of regulating how researchers access secondary data, in this case data that rarely if ever contain personal identifying information in the first place. If the FOI coordinator misses something, the record is considered published, like a newspaper article or government report, though the researcher may voluntarily choose to contact the coordinator and return the records for re-processing in the event of an error.

Public officials have these protections in place in the form of administrative law because they are expected to give up other protections by virtue of their office. For Galliher (1980), those in superordinate positions (teachers, lawyers, CEOs etc.), by virtue of the public accountability of their roles, surrender some of their rights to informed consent, confidentiality and anonymity. Those in publicly accountable positions must recognize that a ‘lack of privacy necessarily comes from a public role’ (Galliher, 1980: 303). As Galliher (1980: 306) continues: ‘only by considering the public’s right to know about the activities of those in positions of public trust is the analysis of otherwise private details generally considered acceptable and even necessary in democratic states’. In his critique of ERB regulation, Haggerty (2004) has gone as far as to suggest publicizing names in qualitative research as investigative journalists do. However, we stress that the goal of using FOI requests in qualitative inquiry is not typically to name and shame corrupt individuals. Rather, the objective is to investigate organizational practices and processes. Who exactly is behind this or that decision revealed through FOI is irrelevant to broader research questions about processual workings of government at the institutional level. In FOI records, evidence of corruption, collusion, embezzlement, authority, abuse, brutality, malpractice and malfeasance is revealed in organizational processes. It is the identification of a process and pattern that draws the attention of the FOI researcher, rather than the tracking down, naming and shaming, and prosecution of specific bureaucrats, the latter of which is not even really possible through the use of FOI. The goal is to ‘study up’ (Nader, 1974) by empirically documenting the organizational processes and dynamics of the governing elite.

As we have shown, there are a number of bureaucratic tests and procedures to protect these organizations and the individuals within them, who must also give up certain rights through occupation of a publicly accountable office. ERBs would be duplicating these efforts, undermining their own effectiveness and creating an undue burden for researchers who decided to use this innovative data collection tool.

ERBs should not create a situation akin to double jeopardy

Humphreys et al. (2014) have used the notion of double jeopardy to capture the dual assessment of scientific research standards during ethics review. According to their findings, scientists sitting on ERBs review not only the ethics implications of a proposal, but also the substantive scientific content. Although legal principles cannot always be applied to research ethics in a straightforward manner (Stone, 2002), we use the notion of double jeopardy in a more conventional legal manner than Humphreys et al. (2014). Double jeopardy is a legal principle in most Western countries with common law legal systems. In criminal law, the principle of double jeopardy provides a defence against being tried twice for the same crime (Coughlan, 2012). In more technical terms, the principle of double jeopardy prevents a person from being tried using the same facts or evidence deliberated in a previous trial.

Applied to FOI, we invoke the notion of double jeopardy as an analogy to illustrate how it would be procedurally unfair to ‘try’ or evaluate a research method twice, when no other form of research is subject to review by two oversight bodies. Much like the principle of double jeopardy in criminal law, ERB regulation of FOI would result in a situation in which FOI research is being reviewed twice despite being comprised of the exact same ‘facts’ in both cases. This is because FOI is not solely a legal mechanism, but a quasi-ethical one that takes into consideration many of the same principles and values assessed by an ERB. Between ERB regulation and the application of FOI’s inbuilt mechanisms of bureaucratic vetting and review, nothing in the use and operation of FOI changes. The ‘case’, or the matter, is exactly the same. For ERBs to regulate FOI requests would create a situation of double jeopardy that is procedurally unfair, eroding due process in ethics and university governance.

ERBs cannot infringe on a citizenship right

In his account of modern citizenship, Marshall (1950: 8) suggests that ‘the inequality of social class may be acceptable provided the equality of citizenship is recognized’. For Marshall, the modern polity has been shaped by the development of civil, political and social rights, which together constitute citizenship rights. Civil rights refer to equality before the law. Political rights refer to the extension of franchise and the right to organize political parties. Social rights refer to welfare rights, defined as ‘the provision of “social services”, unemployment benefit and sickness pay’ (Giddens, 1981: 227). Marshall saw social rights as a way of ‘eroding the most glaring inequalities of modern capitalism and of building social solidarity beyond the confines of ethnic or class particularisms’ (Brodie, 2002: 380). FOI is not some side show or gimmick in Western democracies, but rather would be considered a civil right in Marshall’s scheme. This means it cannot be taken away without creating inequality, and the extent to which it is realized works to level inequality. FOI is a citizenship right that social movements past and present have fought for (Jenkins and Goetz, 1999), and has been described as a human right by the United Nations and the Council of Europe, among other international bodies. It is now a core right in over one hundred countries around the world (Byrne, 2003; Savage and Hyde, 2014; Worthy, 2017).

The right to know, also called the right to information, is one set of rights within the ambit of what Caidi and Ross (2005) refer to as information rights, which also includes privacy and freedom of expression. In Canada, FOI is not a constitutional right (Clement, 2015), although in some countries like New Zealand it is. Although these rights can be curtailed in liberal democracies for legally articulated reasons, national security perhaps being the most pervasive one, a body such as the ERB that has no standing or jurisdiction over these rights cannot curtail or infringe on them. The right to know can be amended and be more or less direct, depending on the letter of the law (Bovens, 2002), but an entity without jurisdiction over these rights cannot make that decision for a citizen or on behalf of any level of government. Even if an ERB invoked a law of the land claim, arguing that the law of the land trumps an ethics ruling (e.g. to protect sources and confidentiality) and academic freedom (Lowman and Palys, 2013), in many countries FOI is a constitutional or civil right. Any effort by an ERB to block a citizen’s right to know could be challenged under administrative and constitutional law as a violation of those rights.

There is also the question of how far information rights should extend. Roberts (2001) has explored the appetite for reform and amendment of FOI laws to include government-owned corporations, quasi-governmental organizations, government contractors and the private sector. Although FOI laws and regimes cover some of these entities, meaning they are subject to FOI, in most cases government-owned corporations, quasi-governmental organizations, government contractors and private organizations are not. In this sense, there is a limit to FOI as a citizenship right, and that limit is often met at the boundary between public and private, corporate entities (with quasi-public, private government contractors and other greyer categories falling into the latter). As noted above, the right to access already works in balance with codified restrictions that protect other parties, their information and their rights. Roberts (2001) also commented on the erosion of FOI law and rights in Canada and elsewhere, a situation that has continued (see Luscombe et al., forthcoming). It would be in the interest of universities to push to expand FOI law and policy to cover government-owned corporations, quasi-governmental organizations, government contractors and the private sector. Where better can the right to know be exercised and advocated for if not in the university (but see Lowman and Palys, 2000)?

Apples and oranges, again: Wrong paradigm

In our view, efforts by ERBs to regulate FOI, a relatively new research innovation in academia, are part of the misinformed trend toward the regulation of qualitative research using an inappropriate paradigm of ethical review. For Librett and Perrone (2010), qualitative inquiry has only become subject to ethical review because of a mission creep in ERBs. As they put it, ‘Review boards often function as prospective oversight mechanisms, which run beyond their original jurisdiction, and so can be construed as becoming influenced by mission creep’ (Librett and Perrone, 2010: 734; also see Haggerty, 2004). Subjecting FOI requests to ethical review would qualify as further ERB mission creep. ERB mission creep creates further problems when it not only regulates new domains (in this case FOI), but does so haphazardly through application of inappropriate categories, tools and paradigms of ethical assessment. As Librett and Perrone (2010) argue, ERBs in many cases use categories of risk and harm that make little sense in the context of qualitative research. These authors argue specifically that ethnography should not be subjected to the same ethical review procedures as animal testing because the forms of inquiry are fundamentally different. We make a parallel claim regarding FOI requests. The use of FOI requests in qualitative research typically follows an open, inductive research design. The iterative, unpredictable nature of FOI is such that the researcher cannot anticipate what the next requests will entail, how the scope of a request will change, or which agency the coordinator will send the request to. The request and the disclosure frequently go in unexpected directions, not least because of the actions and decisions of the FOI coordinator, who is empowered with a high degree of decision-making discretion (see Luscombe and Walby, 2015). Moreover, the current approach to ethics assumes the main risks are to the research subject, which in the case of FOI is some government worker whose identity is usually already protected from disclosure under privacy law. Moreover, as noted, the subject of FOI research is typically an organizational practice or process, not an individual. To the extent that current approaches to ethical review individualize risk (Prior, 2010), the predominant paradigm fails to comprehend the focus and object of analysis of much qualitative research, including that which uses FOI requests.

An additional shortcoming in the current precautionary ethics paradigm – one relevant to FOI – concerns the assumption that the researcher is in a position of power-over. In contrast to biomedical research, from which the current paradigm of ethics review was developed, qualitative research involves a more variable power relationship between researcher and participant. In FOI research, the relationship between researcher and coordinator does not fit a biomedical model of a hierarchical, contractual relationship between doctor and subject (Burr and Reynolds, 2010: 130). Rather, the researcher, as empowered citizen, is requesting access to records from a state representative, the FOI coordinator, who not only occupies a publicly accountable government position, but also holds the upper hand in the relationship insofar as they are buoyed by logics and technologies of governance that work against rather than for the researcher-as-citizen (Luscombe and Walby, 2015; Monahan and Fisher, 2015).

Discussion and conclusion

Bracketing broader questions about the necessity of ERB regulation of qualitative inquiry, we have explained why FOI should not be regulated by post-secondary ERBs. Intending for these claims to be relevant to all jurisdictions where FOI laws are in effect, we have made four general arguments about why FOI requests should not be subject to ERB scrutiny. ²

First, we claim that FOI, in contrast to other forms of scholarly research (e.g. biological, psychological experiments), does not require ERB regulation because it is already regulated through in-built legal and bureaucratic mechanisms, which renders the files already published and therefore a matter of public record. Second, it would not be procedurally fair to make research involving FOI requests undergo two forms of review, both of which seek to uphold common ethical principles, when no other forms of qualitative research are subject to this kind of ‘double jeapardy’. ³ Third, we have argued that precautionary reviews by ERBs in relation to FOI would infringe upon citizenship rights and could comprise a legal liability for universities. A researcher whose citizenship ‘right to know’ was violated by the university could potentially take the university to court. Finally, the reviews that ERBs currently undertake represent a foreign paradigm with incongruous categories that hinder the kind of progressive, intellectually stimulating research that universities are meant to promote.

There is much at stake in the potential hindrance of a researcher’s use of FOI to conduct qualitative research on government. FOI not only helps researchers better understand the processual and organizational dynamics of public bodies, it also allows academics, as active citizens, to help hold those in positions of power accountable for their actions. Researchers who use FOI to study issues such as state corporatization, surveillance or social movement policing are using research to create a critical dialogue about these issues in the interests of improving institutional trust, government legitimacy and public accountability. As Rainwater and Pittman (1967: 366) argued long ago: ‘[S]ociologists have the right (and perhaps also the obligation) to study publicly accountable behaviour … One of the functions of our discipline, along with … intellectual pursuits generally, is to further public accountability in a society whose complexity makes it easier for people to avoid their responsibilities.’ FOI requests allow us to do this. ERBs, of all bodies, should not hinder attempts to achieve this goal.