Access to Information and Freedom of Information Requests

Abstract

Access to information (ATI) and freedom of information (FOI) mechanisms are now relevant features of governments in many liberal democracies today. Citizens, organizations, and permanent residents in several countries across the globe can request unpublished information from federal, provincial, state, county, and municipal government agencies. However, most qualitative researchers appear to be unfamiliar with ATI/FOI or write it off as an approach used by journalists rather than as a way to systematically produce qualitative and longitudinal data about government practices. In this article, the authors discuss the use of ATI/FOI requests as a means of data production. The authors show how the use of ATI/FOI requests intersects with issues such as reflexivity, the Hawthorne effect, interviewing, and discourse analysis. The study objective is to foster a multidisciplinary discussion of the strengths and weaknesses of ATI/FOI requests as a data production tool.

Keywords

access to/freedom of information data production reflexivity Hawthorne effect

Introduction

Access to information and freedom of information requests (hereafter ATI/FOI) are overlooked means of data production in sociology, political science, criminology, and related disciplines. ATI/FOI legislation recognizes a qualified right on the part of citizens, permanent residents, and organizations operating in a given jurisdiction to request access to records held by various levels of government. As many as 80 countries now have laws that facilitate the public right to request information from governments. Sweden’s Freedom of the Press Act of 1766 is the oldest existing form of access law. One need not be a citizen or a permanent resident to make a request in the United States, where ATI/FOI legislation was passed in 1966. Canada followed suit in 1983. Similar legislation was introduced in Australia, in 1982, and more recently by the United Kingdom, in 2000. In addition to substantive differences between ATI laws, a number of factors account for differences across ATI regimes. These include differences in the approach to transparency expressed by political leaders, the existence of multiple governmental cultures with varying approaches to secrecy (see Thomas, 2010a), differential investment in the infrastructure and bureaucracy of access, and the degree of power and influence exerted by ATI ombudsperson offices. In Canada (where we conduct our research), a multilayered access regime operates through provincial, municipal administrative, and legal frameworks, in addition to a separate federal ATI system. At the federal level, using ATI involves submitting a written request and 5 dollars (Canadian) to the ATI/FOI office at the agency in question. The premise (or promise) of ATI/FOI law is that citizens can request information that has not previously been made a matter of the public record and that requests of this kind facilitate information access in a participatory and democratic manner and reinforce government accountability.

Investigative journalists often use ATI/FOI requests to produce records that form the basis of major news stories. For journalists, ATI/FOI requests offer an opportunity to dig deeper into government agency work and supplement the scripted accounts of official spokespeople or unreliable insider sources and orchestrated “leaks” (Cribb, Jobb, McKie, & Vallance-Jones, 2006; Rosner, 2008). The association of ATI/FOI requests with news-breaking journalism may be one of the reasons that qualitative researchers have yet to fully capitalize on the value of these mechanisms as a means of data production.

Qualitative researchers are missing out. ATI/FOI requests can be fruitfully used to explore the activities of federal, provincial, county, state, and municipal governments. Shadowing government employees (see McDonald, 2005) may provide the most in-depth data about how workers work in government agencies and how organizations change over time. However, if shadowing is not possible for lack of entry, or when dealing with agencies that do not allow researchers entry (such as some security and intelligence agencies), ATI/FOI requests present a viable means of producing textual data. Moreover, as government agencies keep a close eye on in-progress ATI/FOI requests to control the visibility of their work and manage potentially “disruptive disclosures” (see Goldsmith, 2010), ATI/FOI offers a unique means of studying official information management and public relations activities. Both the products of ATI/FOI requests and the information-brokering process itself are valuable sources of research data.

In what follows, we outline the procedures by which qualitative researchers can use ATI/FOI. We argue that ATI/FOI requests can be used to produce data representing three levels of analysis. We then discuss the intersection of ATI/FOI research with issues that are germane to qualitative inquiry, such as reflexivity and the Hawthorne effect. We discuss three moments of reflexivity related to ATI/FOI requests. Following this, we explain how ATI/FOI requests can be used in conjunction with interviewing, and we explore the overlap between ATI/FOI requests and use of discourse analysis as a way of formulating qualitative research design. ATI/FOI requests can be used to generate longitudinal data sets, which also allows for sequential data production where the results of ATI/FOI requests are triangulated with interviews and analysis of official organizational discourses. In this way, use of ATI/FOI requests can be a part of what Lather (2010, p. 65) calls “smart mixed methods.”

By elaborating how qualitative researchers can use ATI/FOI requests as a means of data production and demonstrating how use of ATI/FOI intersects with issues germane to qualitative inquiry, we hope to encourage more prominent usage of ATI/FOI in research on government agencies. We conclude by discussing Bourdieu’s (2010) notion of the collective intellectual in relation to ATI/FOI requests and collaborative research efforts. A focus on internal texts in critical and poststructural analysis of government policy and activities can be fruitful (see Lather, 2010), and use of ATI/FOI requests should be a key strategy in such work. ATI/FOI requests can allow scholars to produce data about government activities and collectively disrupt government discourses, policies, and practices that they find alarming.¹

Investigating Government Agencies in Action: Putting ATI/FOI Requests to Work

Investigative journalists are a small but active and influential subset of regular ATI/FOI users. These researchers often use ATI/FOI requests to dig under the surface of governmental communications. ATI/FOI requests are a way of moving beyond official discourse, which Burton and Carlen (1979) define as carefully prepared, managed, and articulated messages in mass media (and, we would add, on government websites). Journalists often go beyond official discourse to understand how government decisions are made and the impact of these decisions.

ATI/FOI requests are associated with the breaking of a big story that is the golden goose of investigative journalism, and it is perhaps for this reason that many qualitative researchers shy away from using ATI/FOI to produce data. Qualitative researchers have been missing out on a remarkable means of producing data about government agencies and their activities. We use the term “data production” with the intent of pointing to the active role of the researcher (and other people, such as the ATI/FOI coordinator) in shaping the outcome of the ATI/FOI request as well as the constructed nature of texts themselves. In our experience, other qualitative researchers often regard ATI/FOI research as a straightforward and noninteractive method—something like a formalized version of a library database search. But systematic ATI/FOI research is neither straightforward nor unobtrusive. The way that we use ATI/FOI requests in our research on policing and security (see Larsen & Piché, 2009; Monaghan & Walby, in press; Piché & Walby, 2010; Walby & Monaghan, 2010, 2011) requires a commitment to rapport building and negotiation with government ATI/FOI coordinators and the occasional use of confrontational tactics to work around denials of access.

This approach is informed by Gary Marx’s (1984) comments on “dirty data.” Marx describes dirty data as texts that are submerged by governmental agencies since the information therein would be discrediting to the agencies in question. We extend this approach to dirty data with a focus on ATI/FOI requests as a means of accessing texts that could be discrediting or controversial if made a matter of the public record. It is important to note that official government information classifications (Protected, Secret, Top Secret, etc.) and the categories of information that can be accessed or exempted through ATI/FOI mechanisms often intersect but do not necessarily overlap. Although the former governs access based on clearance level and “need to know,” the latter is based on a presumption of access with limitations based on record content and destination. Thus, it is possible to access classified dirty data through ATI/FOI. In addition to such records, the textual trails that we access through ATI/FOI requests contain broader—and from the perspective of government, less contentious—streams of information that reveal much about the everyday practices of government agencies.

Our ATI/FOI-based research focuses on issues of interest to criminologists and legal scholars. Yet the ramifications of ATI/FOI are larger and of potential interest for any researcher who conducts any inquiry related to any governmental agency. Researchers who focus on health agencies, educational agencies, financial agencies, or any other kind of governmental agency that produces texts and that does not make them all a matter of the public record could benefit from using ATI/FOI. ATI/FOI can be put to work to make those records public, and then those records can be analyzed using qualitative data analysis techniques.

There is a small but growing literature on the use of ATI/FOI requests, and less still on the interpretation of resulting data. Most existing literature (such as Roberts, 2004, 2005; Thomas, 2010a) focuses on the administration of ATI/FOI legislation. Gary Marx’s (1984) comments on dirty data are formative but do not develop a methodological program for the use of ATI/FOI requests as data production. In Canada, policing and security scholars, including sociologists and historians, have started to use ATI/FOI requests as a means of data production, which has led to a string of publications on backstage processes of various government agencies. Kinsman and Gentile (2009) have used a combination of archival and ATI/FOI research to obtain records relating to the construction of gays and lesbians as national security threats by the Canadian government during and after the Cold War. Larsen and Piché (2009) have used ATI/FOI requests to explore the post–September 11, 2001, interagency management of security certificate (“secret trial”) detention in Canada, with the creative blurring of organizational mandates allowing for this practice. Larsen (2008) also used ATI/FOI data to explore the relationship between politicians’ statements about security certificates and the texts that govern the certificate process. Piché and Walby (2010) have used ATI/FOI requests to look at how tours of federal prisons are scripted by prison administrators. Walby (2009) has used ATI/FOI requests to explore how conservation officers regulate homeless persons and men having sex with men in parks in the city of Ottawa. Walby and Monaghan (2011) have explored how federal security intelligence agencies such as the Royal Canadian Mounted Police (RCMP) and Canadian Security Intelligence Service (CSIS) coordinate with municipal police for the purposes of suppressing social movements. Walby and Monaghan (2010) have also explored how federal security intelligence agencies coordinate with provincial and municipal police in rural areas as concerns the militarization of areas around nuclear power plants in the province of Ontario. And Hewitt (2002) has used ATI/FOI requests to develop a history of the RCMP’s clandestine surveillance activities on Canadian university campuses.

ATI/FOI requests can be rewarding in producing data about what various government agencies do. But ATI/FOI requests are not limited to use for exploring policing and security work. They might even be better put to use for exploring the work of government agencies related to health, finances, and education, as these government agencies are less likely to invoke exemption clauses that pertain to criminal investigations and national security. Any other government agencies that create policy and enact governance through texts in any other governmental sphere could be investigated using ATI/FOI requests.

Before elaborating on how ATI/FOI requests intersect with debates germane to qualitative inquiry, we offer an overview of some of the key conceptual issues regarding government agencies and how ATI/FOI requests can be used to better understand them. We break these conceptual issues into three elements—texts, work, and organizations (see Smith, 2001). With ATI/FOI requests, the most important element is texts. ATI/FOI requests are used to broker the disclosure of texts (Walby & Larsen, in press). Government agencies are in the business of producing, amassing, collating, archiving, and circulating information in textual form. As Dorothy Smith (1999, 2005) has constantly reminded qualitative researchers, the relationships that go on in governments as well as the relationships between government agencies and the subject populations that they govern are textually mediated. With ATI/FOI requests, we are trying to get at texts used in government processes or texts produced as part of governing.²

There are several layers of texts in government agencies that are important to account for. First are those texts that are used to directly govern the work of government employees such as administrative rules and regulations, codes of conduct, standing orders, and directives. A second level of texts includes those used to govern subject populations directly or make up the arrangements that allow for the governance of subject populations. Some of these texts have peculiar names that might not be well known to most scholars. Of particular interest to us are memorandums of understanding (MOUs) or letters of understanding (LOUs). These documents establish relationships of authority, capacity, and resource distribution between two or more government agencies. MOUs are rarely made publicly available through proactive disclosure, but the arrangements that they establish can modify and extend agency mandates and allow for new techniques of governance (see Larsen & Piché, 2009). ATI/FOI requests are a reliable means of accessing this kind of text. Also of interest are the unofficial texts that are never intended for public circulation, such as the notes and the internal memos and the emails of government employees that comprise the knowledge work that make up those first- and second-order forms of governing texts. In addition, it is possible to access prior iterations of public speeches made by officials, “Qs & As” documents that outline approved talking points for spokespeople, and briefing notes that provide condensed summaries of specific issues. Finally, there are texts such as internal correspondence and communications generated in response to the making of an ATI/FOI request. The making of an ATI/FOI request generates textual material within government agencies, including outputs from ATI/FOI management software, correspondence between ATI/FOI coordinators and the “office(s) of primary interest” (OPIs) within agencies deemed most likely to have the information being sought, and correspondence between ATI/FOI units and government communications officials who work to mitigate the impact that the disclosure of dirty data may have. These are the layers of texts that we get at using ATI/FOI requests. These texts illuminate governmental agency activities better than reliance on official discourse or the carefully managed stories that government agencies themselves release.

The second element of this conceptualization of government agencies is work. Ericson and Haggerty (2005, p. 553) make the observation that workers in knowledge-based occupations who produce and distribute information need to get the job done and also produce and distribute information about their own work. This certainly holds true for ATI workers. Texts are important to government agencies, but they do not do anything by themselves. Texts need to be worked with and activated by government workers (Smith, 1999). Texts need to be penned or typed. Texts need to be sent and circulated and modified, interpreted, summarized, vetted, cleansed, and approved. They need to be discussed at meetings and organized according to internal information management protocols. Such texts encode certain messages about what government agencies are doing or what they should be doing—and how they should be doing it. These texts can be directives for shaping the work that government agencies do or the way that they govern subject populations. Backstage texts accessed through ATI/FOI mechanisms can also help researchers to explore work that occupies the space between a given agency’s official protocol and the informal operational code that governs day-to-day activities (Punch, 2009). Whereas some aspects of the informal “negotiated reality of internal institutional and occupational practices” (Punch, 2009, p. 3) are strictly nontextual, other deviations from the official paradigm take the form of recorded communications, reports, and meeting minutes. By focusing on these texts that we access using ATI/FOI, we take the work of government agency employees as an object of analysis.

The third element of this conceptualization of government agencies is networked organization. The texts that are produced and the work that is done by government agencies are rarely the domain of one organization alone. Instead, government agencies are networked, crossing over into one another. They are always demanding approvals and check marks from one another on various texts and kinds of work. Interagency relationships involve collaboration and sharing as well as competition for position and resources. It is difficult to inquire into the texts that one government agency produces or the work that is done with texts in one government agency without understanding how that work and those texts are organized in a network with other agencies within the government sphere—and often with private sector and nongovernmental interests. The network is something that is material; the material link between these organizations is textual and it is formed through the work of the employees. The texts and the work that government employees do with the texts actuate or enact those networks between various organizations (see Turner, 2006), creating various textual trails that the ATI/FOI researcher can explore using ATI/FOI requests.

ATI/FOI requests can be used to produce data representing three levels of analysis: the level of texts themselves, the level of work, and the level of networks between organizations. However, the right to request information does not necessarily translate into the timely or comprehensive release of records. A range of in-built governance mechanisms—from the application of various legal exemption and redaction clauses, to the design of information management and access software, to the response tactics of ATI/FOI coordinators and resistance from OPIs—mitigate against full disclosure. ATI/FOI coordinators stonewall requests through various techniques such as amber lighting and red filing (see Roberts, 2006; Walby & Larsen, in press). Amber lighting refers to the tagging of a request or a requester as politically contentious. Red filing refers to requests that are stonewalled either by the Minister or by the Prime Minister’s Office (or the equivalent in countries other than Canada), who receive a weekly inventory of tagged requests. This is why a crucial element of making an ATI/FOI request is following up with the agency to investigate how the request itself was managed (an issue we discuss in the following).

Given the potential of ATI/FOI requests to play a significant role in contributing to scholarly understandings of governmental agencies, we will now put the issue of ATI/FOI requests into context within debates germane to qualitative inquiry. To make the link, in what follows we discuss reflexivity, the Hawthorne effect, interviewing, and discourse analysis. Our goal is to foster a multidisciplinary discussion of ATI/FOI requests as a data production tool.

ATI/FOI Requests and Qualitative Inquiry: Making the Link

Reflexivity and the Hawthorne Effect

Reflexivity has been a staple topic of debate in the social sciences for at least three decades. Definitions of reflexivity are slippery, but the general meaning of reflexivity is that the researcher provides an account of how they produced his or her data and the knowledge resulting from a study (see Mauthner & Doucet, 2003). There is (or should be) a self-critical mirror facing the researcher in all phases of the project, from the formulation of a research question, to data production, to data analysis, through to writing and audience reception and response related to a publication. Reflexivity is not simply a moment for researchers to vent about the challenges of doing research (Finlay, 2002; Haggerty, 2003; Lynch, 2000) or what Bourdieu (with Wacquant, 1992, p. 72) critiques as “self-fascinated observation of the observer’s writings and feelings.” Instead, a major facet of reflexivity is as a tool that researchers constantly use to assess how they do what they do in terms of knowledge production.

Reflexivity intersects with the strategy of using ATI/FOI to produce data in at least three ways. First, reflexivity relates to the wording of a request. Second, reflexivity relates to the negotiation of terms of requests with ATI/FOI coordinators in government agencies. Third, reflexivity relates to the necessity of following up with additional ATI/FOI requests to investigate management in the production of texts about the initial ATI/FOI request. We explore each of these moments in what follows.

The wording of the request is what the ATI/FOI coordinator in any government agency uses to task an office of primary interest to search for records. For instance, asking for “all records related to” a given topic will create a large workload for the ATI/FOI coordinator since he or she needs to, if interpreted literally, search for all emails, memorandums, correspondence, and so on. If the requester asks for something narrow, such as all email communication corresponding to a very small date range, the ATI/FOI coordinator will not be forced to cast a wide net. The issue with the wording of requests gets trickier, since the terminology that academics associate with government practices does not always reflect the internal vocabulary of such agencies. Canadian Security Intelligence Service uses the term “multi-issue extremists” to refer to various kinds of activism and political dissent—not everyday terminology for referring to social movement participants. If a researcher did not know those key words, he or she would have a hard time digging up information about the issues at hand. In this way, reflexivity cuts across the initial moment of making an ATI/FOI request insofar as the requester needs to know something about the terminology used by government agencies to whom he or she is asking to provide information. Records obtained through ATI/FOI can be analyzed and coded for internal keywords, and these keywords can form the basis of follow-up requests (see Larsen & Piché, 2009). Carefully documented ATI/FOI research can become more effective and precise over time. Conversely, unexamined problems with request wording can, down the line, lead to avoidable lacunae in release packages. When formulating requests, it is also important to know something about the location of the records being sought within intergovernmental networks, as records that fall under the auspices of several agencies will be subject to delay related to consultations (Walby & Larsen, in press). Such records are often of particular interest to the researcher, but since a given request tends to move at the pace of its most complex component, it can be useful to critically examine request wording, bracket off records likely to be subject to consultation, and seek them through a separate request. The requester must have a research diary (see Mills, 1959) in which he or she makes field notes and write about his or her formulation of request wording and keep track of the terminology used to refer to activities that government agencies are involved in. Without keeping an account of how one formulates and files the ATI/FOI request, the requester can make mistakes at this initial stage and the inquiry can go off track.

The second moment of reflexivity in making an ATI/FOI request is negotiation with ATI/FOI coordinators. The initial wording of the request is almost always revised and critiqued by ATI/FOI coordinators, sometimes with the intent of scuttling the request, but most often because they themselves or the office of primary interest is unsure about how to respond. Establishing rapport is a key issue in qualitative inquiry (see Mazzei & O’Brien, 2009), not the least in these negotiations with ATI/FOI coordinators who act as information brokers for the agency. Rapport and trust can be gained through professional conduct, knowing the pertinent legislation, and knowing about the file structure of the agency. Preliminary research regarding the agency’s record of ATI/FOI compliance and the existence of factors such as systemic delays or high rates of ATI/FOI personnel turnover can assist in the establishment of rapport. This relationship with the ATI/FOI coordinator is important to provide an account of, since it can evolve over time as a requester makes more requests with the same agency. The negotiations with ATI/FOI coordinators are important for the researcher to keep track of and to be self-critical of, since this process of brokering access to the texts can make or break the investigation. Researchers can feel pressured to compromise on scope of the request, to revise wording, or to otherwise limit their requests. Not all negotiations are harmful; conversations with some ATI/FOI coordinators can produce invaluable information about the internal dynamics of the agency (some police agencies have sworn officers as ATI/FOI coordinators, which is good to be aware of) and possible areas of tension between the coordinator and other agency personnel. Some coordinators act as go-betweens, allowing both the requester and the office of primary interest to ask clarifying questions that can open up new avenues for research.³ Keeping careful notes on the request process can also help to catch request mistranslation. This occurs when a request’s original wording is incorrectly summarized or modified by the analyst responsible for communicating with the OPI. In a recent case (RCMP ATI GA-3951-3-0), a request by an investigative journalist colleague for records relating to an “operational undercover training course” was tasked to the OPI as a request for records relating to an “operational underwater training course.” Although the error was caught, it added to the processing time for the request. Meticulous documentation of the request process is also important in the possibility that the researcher opts to formally appeal an agency decision to an ATI/FOI ombudsman. Appeals or complaints processes vary between jurisdictions, but they are always initiated by the submission of a formal complaint, which should include a detailed account of milestones in the request history, interactions between the requester and the ATI/FOI coordinator, and decisions made about alterations to the request wording. Again, a research diary is important to keep; field notes should be collated at the end of the life of a request; the interactions between the ATI/FOI coordinator and the decisions made can be reflected later on as part of the overall process of inquiry. A good example of this kind of reflexivity can be found in Gentile’s (2009) work on “queer(ing) archives,” which provides an account of her time spent in the archival maze of official records related to antihomosexual security purges by the RCMP. Drawing on interactions with ATI coordinators and an analysis of formal ATI processes, Gentile (p. 154) contends that the Access to Information Act in Canada is a way for researchers to acquire “information heretofore unknown” and, simultaneously, a procedure by which officials try to “protect the government and create impenetrable obstacles.”

The third moment of reflexivity is following up with additional ATI/FOI requests to investigate how the initial request was managed within the agency. Each ATI/FOI request is assigned a file number. Subsequent requests can be made for copies of all correspondence concerning the file number within the agency (or between agencies), especially between the ATI/FOI coordinator and the office(s) of primary interest. The expanding automation of aspects of the access process through specialized software like ATIPflow and AccessPro Case Management means that agencies increasingly produce metainformation about requests; follow-up requests for printouts from these software platforms can reveal whether and when a request was flagged for special review, when consultations occurred, and which risk profile the agency assigned to the requester.⁴ This is perhaps the most important moment of reflexivity in ATI/FOI research, for if the goal is to understand how government agencies work with texts and manage information there is no better opportunity to investigate those issues than exploring the textual trails forged in relation to a previous request. It is crucial to follow up on how a request was managed by an ATI/FOI coordinator and an office of primary interest. To understand the limits of ATI, and how certain texts are proofed against disclosure, the ATI/FOI requester must follow up with an ATI/FOI request to see how his or her initial request was managed. If qualitative inquiry is marked by a humility that holds in check the scope of its claims (see Lather, 2010), noting the limits of data production when using ATI/FOI is a key part of being humble about what one does as an academic.

These three moments of reflexivity emerge in response to what Giddens (1991, p. 20) highlights as the tendency of contemporary organizations to constantly revise their messaging and operations based on incoming information. This is a dynamic understanding of organizations, which emphasizes the uncertainty and potential risks associated with knowledge. Reflexivity not only affords an opportunity to examine the position of the researcher and the way they phrase requests and engage with ATI/FOI coordinators but also provides a chance to investigate how the responses of the ATI/FOI coordinator and the subsequent responses of the requester shape the outcome. Reflexivity is also a way of evaluating the integrity of the research strategy, allowing for revisions to be made to subsequent requests and negotiations with ATI/FOI coordinators. Crucially, it also facilitates the development of craft knowledge that can be exchanged with other researchers in a collaborative milieu. Later we return to this issue of collaboration.

Use of ATI/FOI requests as a means of data production involves reflexivity during three moments. We suggest that the issue of ATI/FOI also intersects with the Hawthorne effect. The idea of the Hawthorne effect emerged from behavioral and management sciences and connotes a postpositivist concern for measuring situations where people alter their conduct as a result of an awareness that they are being watched. However, the idea has been adapted by qualitative researchers interested in the ethics of research participants’ awareness of being subjects in a study (Monahan & Fisher, 2010) and the idea of researcher field effects, which has ramifications for use of ATI/FOI in producing data about government agencies. Early debates about the Hawthorne effect in social science research were informed by the idea that it was desirable to attempt to “hold the world still” while researching it. This aversion to contributing to change through research was rejected as being a “diversionary,” undesirable, and ultimately impossible objective by researchers committed to a radical praxis (Taylor, Walton, & Young, 1975, p. 26). We propose that understanding how ATI/FOI relates to the Hawthorne effect is important, not because we seek unobtrusiveness or the minimization of impact but because ATI/FOI mechanisms are interwoven with the knowledge systems that they seek to explore, and this should be accounted for.

We approach the issue of the Hawthorne effect in ATI/FOI research at the particular and general levels. At the particular level, the issue concerns how government employees bury traces of their text production, work, and communications when they know they are or are likely to be subject to ATI/FOI requests. This is especially the case when government agencies know an ATI/FOI requester is interested in a particular set of topics related to their agency or when ATI/FOI material is used by researchers or journalists to inform public debates about a given topic. We have obtained many briefing, training, and issue management texts that include notes about the presence of active ATI/FOI requests in a certain area, and we read these notes as cautions intended to govern further textual work around these issues. Through following up on requests with the Royal Canadian Mounted Police, the second author obtained a copy of an internal training document that includes the heading “How you can avoid an ATIP request.” Whether this text reflects the actual content of the training session or a tongue-in-cheek reference to the prohibitions against breaching information management and access laws, we suggest that it speaks to a general attitude toward ATIP in government at all levels across Canada, and indeed, generally. Once again, use of ATI/FOI requests to produce data and go after evidence of information management is obtrusive research, as it is bound to lead to further modification of government messaging and operations.

At a more general level, the issue is how government employees manage everyday text production and work in light of the potential for information disclosure. The Hawthorne effect is often associated with the idea of behavior modification based on the awareness that one is being studied, implying a direct observer effect (Monahan & Fisher, 2010). In the absence of this direct awareness, the potentiality of ATI/FOI can still contribute to acts of self-censorship, the careful management of paper trails, and a tendency toward off-the-record kinds of communication. To an extent, ATI/FOI mechanisms act as systems of surveillance, at least from the perspective of government workers, who are the targets of a public gaze. They respond by governing work in relation to shifting formal information management and retention protocols as well as the range of decisions based on an awareness of the “ATIPability” of the texts that they work with. Such actions of self-censorship, “covering ass,” and deliberate nonproduction of records create absences within or otherwise cut off textual trails (see Manning, 1980). Government workers are expected to manage sensitive information, which requires an understanding of formal policy as well as familiarity with the “informal, unwritten “rules of the game” for handling and disclosing information” (Thomas, 2010b). However, this information management is never total because workers in organizations do not know the full extent of information held.

Nevertheless, critics of the principle of ATI/FOI (see Savoie, 2003) contend that access laws lead to “fishbowl” cultures in government, where public servants are reluctant to commit frank advice to paper for fear of creating a textual trail that could embarrass officials. As debates in political science suggest (see Roberts, 2006), government employees in various countries started to scale back their production of certain kinds of texts when ATI/FOI legislation was adopted. An awareness of the Hawthorne effect should lead ATI/FOI researchers to critically examine the data they produce and to ask questions about how even the most secret backstage texts may be crafted with an uninvited public audience in mind. It follows that backstage texts should not be treated as uncovered truth but rather as internal documents subject to mediation.

Rather than abandoning the use of ATI/FOI requests because of this Hawthorne effect, as if a researcher using ATI/FOI could become “invisible” to the organization he or she is studying, part of the process of reflexivity is to use ATI/FOI requests to understand how the Hawthorne effect is happening in any government agency. The occurrence of a Hawthorne effect within a government agency in relation to a previous ATI/FOI request is important to provide an account of, as it leads to a better understanding of how texts are produced, how work is managed, and how organizations communicate and coordinate with one another.

As Monahan and Fisher (2010) note in the context of ethnographic research, the occurrence of an “observer effect” does not somehow render data inauthentic. Rather, it offers researchers an opportunity to study negotiated meanings and moments of impression management (see Goffman, 1959). This holds true for ATI/FOI; ATI/FOI research is not something that can be done in an unobtrusive way.⁵ Making an ATI/FOI request creates various kinds of information ripples, enacting texts within the target agencies and generating awareness of scrutiny. If the requester attempts to be unobtrusive or minimize their negotiations with the ATI/FOI coordinator, they may end up unnecessarily restricting their own access. At all times, negotiations with the ATI/FOI coordinator need to be accounted for in a research diary and also in the write-up of the data.

Interviewing and Discourse Analysis

The final issue we discuss is how ATI/FOI requests overlap with more conventional qualitative ways of producing and making sense of data. The use of ATI/FOI requests can be part of a sequential data production strategy that employs ATI/FOI requests in conjunction with interviews as well as discourse analysis.

The literature on qualitative interviewing emphasizes how to formulate questions prior to the encounter and how to pose those questions during the encounter, rapport building, and the difficulty of sustaining dialogue (see Arendell, 1997; Best, 2003; Coffey, 1999; Thapar-Björkert & Henry, 2004). There are two ways that ATI/FOI requests can be used in conjunction with interviews. First, the information produced through ATI/FOI requests can be used as prompts later on in interviews with the government employees that are in some cases named in the text or with individuals who have been subjected to governance practices discussed in or made possible by a text. This information is not used in an accusatory way, but in an exploratory way, such as “Can you tell me about what this reference in document 00-12-34 refers to?” In this way, the researcher is not solely reliant on the results of ATI/FOI requests and can begin to triangulate these data and build on them sequentially. This conjoining of ATI/FOI requests with interviews is vital to understanding organizational change over time.

To provide an example, Walby (2009) used ATI/FOI requests in conjunction with interviews to study the National Capital Commission (NCC) in Ottawa, Canada. NCC conservation officers are engaged in policing public sex and homeless people in public parks. Walby wanted to understand how NCC conservation officer policing is organized and how the work of conservation officers intersects with municipal police, RCMP, and private security work. The initial stage of data production involved ATI requests; it evolved into an amassing of more than a 1,000 occurrence reports representing a 10-year time span, which provide officer narratives concerning their daily work and their policing of Ottawa parks. Walby used those initial results to create questions for interviews. The information from the interviews was used to further investigate the work of the NCC through additional ATI requests.

This is only one example of a research design that conjoins ATI/FOI requests with interviews in sequential data production. There are other examples of ATI being used to study organizations over time. Walby and Monaghan (2011) used ATI requests with CSIS and the RCMP to understand the morphing character of security intelligence threat assessments pertaining to social movement activists over a 7-year time span. Another example can be found in Kinsman and Gentile’s study of the organization of national security regulation of sex and sexuality. By combining a critical analysis of official security texts obtained through ATI/FOI with interviews with former security operatives and individuals targeted by national security campaigns, they were able to explore “the midst of the rupture between official accounts and experiences” (Kinsman & Gentile, 2009, p. 13) and to offer a detailed reading of national security narratives over a time span of three decades.

Second, information gained from interviews or texts can be used to phrase future ATI/FOI requests. It is important to ask about texts when conducting interviews (see Devault & McCoy, 2002; Walby, 2005, 2007) because texts connect government workers within and across organizations. The discussions that the requester has with the ATI/FOI coordinator offer an opportunity to ask questions about file structure, agency-specific terminology, and process within the organization in question. This is the kind of “secret interview” (Hilbert, 1980) that participant observer researchers use to develop a nuanced understanding of argot and emic categories, although there is no need to be secret in the context of ATI/FOI. In fact, more overt, interview-based research with ATI/FOI coordinators should be conducted. Just as ATI/FOI requests are a neglected form of data production, the work of ATI/FOI coordinators in shaping access regimes has yet to be explored in earnest.

ATI/FOI requests can also be used in conjunction with discourse analysis. Discourse analysis focuses on those carefully managed and circulated texts that government agencies use for public relations management (see Alvesson & Karreman, 2000; Burton & Carlen, 1979, on varieties of discourse analysis). Discourse analysis alone would create an incomplete picture of what government agencies do. At the same time, reliance on ATI/FOI data alone would not provide an account of how government agencies try to provide a public face for themselves and spin public understanding of the activities that they are engaged in (Roberts, 2005). A comparison of data produced through ATI/FOI requests and media material treated through discourse analysis can provide a more complete picture of how organizations manage information and their public image.⁵

It is perhaps the case that qualitative researchers have settled for analysis of proactively disclosed official discourse more often because it is easier to get at compared to the lengthy and uncertain process of trying to produce data using ATI/FOI. Using ATI/FOI requests in conjunction with discourse analysis can be done sequentially. The information gleaned from media and discourse analysis can be used to produce the wording of ATI/FOI requests, which can then be analyzed in light of future government messaging on a particular topic. This would lead to an examination of how official discourses are crafted within the organization. For instance, the requester may request previous iterations of a speech delivered in public by a politician, or the minutes and notes and email correspondence pertaining to a particular meeting where some speech or other form of official discourse was crafted. Alternatively, requests for drafts of media lines and briefing notes for approved spokespeople in relation to a particular event can provide invaluable information for an analysis of the careful shaping of messages by government. One could begin with a request for records relating to media inquiries about a particular topic and within a particular timeframe as well as existing briefing material related to that topic. Follow-up requests could track the textual trail between initial interactions with the media and the production of official communication scripts as well as records related to routine and ad hoc media analysis by government agencies that feed back into future public relations efforts. ATI/FOI thus can be used to study the production of official discourse and changes in the production of discourse in the agency.

Using ATI/FOI requests in conjunction with interviews and discourse analysis shows that ATI/FOI requests can be fruitfully positioned as a part of longitudinal research design that involves sequential data production and triangulation (on the importance of longitudinal, qualitative research see Mcleod, 2003). This approach to using ATI/FOI requests in qualitative research emphasizes a sequential and longitudinal design to explore the work of government agencies over time. The use of ATI/FOI requests, interviews, and discourse analysis can be staggered; information gleaned from one module of data production can inform future data production efforts.

Discussion and Conclusion: ATI/FOI and the Collective Intellectual

We are not arguing that ATI/FOI disclosures provide access to “pure” or complete data. Certainly we are not trying to “repositivize” qualitative research either (see Lather, 2010). ATI/FOI regimes are subject to systemic and endemic flaws and have been usefully critiqued as part of government apparatuses of information management (see Gentile, 2009; Roberts, 2006; Walby & Larsen, in press). Access is only ever partial, so our claims must always be qualified. Release packages often contain less information than researchers seek and more than government agencies would prefer to disclose. The ATI/FOI disclosures that we treat as data are subject to all sorts of mediation; for this reason it is important to understand how information management concerning our data production has happened. Hence, we have argued that ATI/FOI requests can and should be used to investigate how previous ATI/FOI requests have been managed. In this way, the use of systematic ATI/FOI requests as a means of data production intersects with the issue of reflexivity.

The examples we have used of the successes of ATI/FOI requests relate to policing and security, yet ATI/FOI may be better put to work in exploring health and financial agencies or educational agencies and the kinds of textual work that goes on within those agencies as well as the relationships between those agencies. Certainly more conventional methods of data production can be used to study those processes, interviews being among them. However, we have suggested that there are limits to what interviews can convey about what goes on in government agencies. Discourse analysis is limited too, insofar as it tends to focus only on official messaging. Thus, for some agencies within government, ATI/FOI provides a direct means of conducting a systematic investigation. This goes for intelligence and security agencies such as the RCMP and CSIS in Canada more than for any other, since security agents and employees of these security agencies are not in the business of sitting down for a coffee and chatting about their work. At minimum, we can no longer afford to have ATI/FOI requests neglected as a means of data production if we wish to understand how government agencies do the work that they do. At best, we can begin to use ATI/FOI requests in triangulation with other means of data production to provide a longer-term picture of organizational change. This responds to Lather’s (2010, p. 65) call for “smart mixed methods.”

ATI/FOI requests provide partial entrance into a little known realm of texts that are crucial to understand how government organizations operate. All work in government organizations is connected by texts (Smith, 2001). Much of what is said and done in government organizations is written down or otherwise documented, and despite a range of limitations, and barriers to access, much of this material is accessible through ATI/FOI. Getting at this realm of backstage texts provides the researcher a chance to go beyond the study of cautiously prepared public relations texts and official discourse propagated by governments. There is another layer of texts (e.g., memorandums of understanding, email correspondence between government employees) that can be revealing; these texts are never meant for public disclosure and only become a matter of the public record through use of ATI/FOI.

Use of ATI/FOI provides a unique perspective for scholars who are trying to conceptualize how government agencies work in action. Yet scholars are not the only people interested in this information. Social movement activists and lawyers and journalists are often interested in the same information, so can work in teams using ATI/FOI, breaking down barriers between universities and activist communities. The use of ATI/FOI requests is best conducted in teams, where multiple requesters seek access to particular facets of broader processes. By pooling resources, findings, and analyses, a group of researchers can work around barriers to access and produce an anthology of data for collaborative exploration and circulation to interested parties. This collective approach compounds the importance of reflexivity in the research process and also places contestation at the forefront of critical research strategies.

The professionalization of social science research has led to a predicament among academics, where they feel that their inquiries are best done by themselves and in their name alone (Burawoy, 2005; Lather, 2010). But the textual trails that we explore are often of such a scope that it is necessary and desirable for researchers to work in teams to explore them. And it is necessary for academics to work with activists, lawyers, and journalists to provide a better understanding of what these agencies do, the implications of government actions, and opportunities for change and resistance (see Kramer, Michalowski, & Chambliss, 2010). The notion of ATI/FOI itself runs contrary to protective approaches to data. After all, it is only by invoking a legal right to know qua member of a public that ATI/FOI researchers are able to access backstage texts. Furthermore, the texts that we obtain through ATI/FOI requests become a part of the public record. A release package that represents the culmination of a careful process of request formulation and negotiation by an ATI/FOI researcher can be accessed without difficulty by subsequent researchers who ask for it by file number. Summaries of all requests processed by a given agency over the course of a year or more can be obtained through ATI, allowing for targeted follow-up on any completed requests of interest. There is always a public aspect to ATI/FOI research, and every ATI/FOI requester is, by implication, an open government activist.

We suggest that ATI/FOI requests as a means of data production can be part of an approach to qualitative inquiry that breaks down barriers between academics and other knowledge communities. We also think that this team-based approach to ATI/FOI reflects what Bourdieu (2010) calls the collective intellectual. Bourdieu’s notion of the collective intellectual is a critique of Sartre’s notion of the universal intellectual who appears to know everything, who comments on everything. Bourdieu valorizes a team-based research approach that extends beyond the walls of the university to intellectual communities, including activists, lawyers, and journalists who have similar interests.

Keeping in mind how the issues of reflexivity, the Hawthorne effect, and triangulation intersect with ATI/FOI requests, ATI/FOI requests are best used in scenarios that reflect a collective approach to scholarship. Team-based scholarship raises interesting questions about reflexivity and how to arrive at collective decisions about research orientations, topic selection, and strategies for analysis (Mauthner & Doucet, 2008; Siltanen, Willis, & Scobie, 2008). Yet group-based research that uses ATI/FOI as a way of producing data provides a more robust sense of the work carried out in government agencies; it also offers a way of doing critical research and struggling against the individualization of scholarly inquiry at the same time we investigate injustice stemming from the activities of government agencies.

Footnotes

Acknowledgements

Thanks to Sean P. Hier, Tia Dafnos, Karine Côté-Boucher, Seantel Anaïs, and Dale Spencer for their comments. The authors contributed equally to the development of this article.

Notes

References

Alvesson

Karreman

(2000). Varieties of discourse: On the study of organizations through discourse analysis. Human Relations, 53(9), 1125-1149.

Arendell

(1997). Reflections on the researcher-researched relationship: A woman interviewing men. Qualitative Sociology, 20(3), 341-368.

Best

(2003). Doing race in the context of feminist interviewing: Constructing whiteness through talk. Qualitative Inquiry, 9(6), 895-914.

Bourdieu

(2010). Sociology is a martial art: Political writings by Pierre Bourdieu. New York, NY: New Press.

Burawoy

(2005). For public sociology. American Sociological Review, 70(1), 4-28.

Burton

Carlen

(1979). Official discourse: On discourse analysis, government publications, ideology and the state. London: Routledge & Kegan Paul.

Coffey

(1999). The ethnographic self: Fieldwork and the representation of identity. London: SAGE.

Cribb

Jobb

McKie

Vallance-Jones

(2006). Digging deeper: A Canadian reporter’s research guide. London: Oxford University Press.

Crow

Wiles

Heath

Charles

(2006). Research ethics and data quality: The implications of informed consent. International Journal of Social Research Methodology, 9(2), 83-95.

10.

DeVault

McCoy

(2002). Institutional ethnography: Using interviews to investigate ruling relations. In Gubrium

J. F.

Holstein

J. A.

(Eds.), Handbook of interviewing: Context and method (pp. 751-775). Thousand Oaks, CA: SAGE.

11.

Ericson

R.V.

Haggerty

(2005). The policing of risk. In Newburn

(Ed.), Policing: Key readings (pp. 550-564). Cornwall, UK: Willan Publishing.

12.

Finlay

(2002). Negotiating the swamp: The opportunity and challenge of reflexivity in research practice. Qualitative Research, 2(2), 209-230.

13.

Gentile

(2009). Resisted access? National security, the Access to Information Act, and queer(ing) archives. Archivaria, 68, 141-158.

14.

Giddens

(1991). Modernity and self-identity: Self and society in the late modern age. Stanford, CA: Stanford University Press.

15.

Goffman

(1959). The presentation of self in everyday life. New York, NY: Anchor Books.

16.

Goldsmith

A. J.

(2010). Policing’s new visibility. British Journal of Criminology, 50(5), 914-934.

17.

Haggerty

(2003). Review essay: Ruminations on reflexivity. Current Sociology, 51(2), 153-162.

18.

Hewitt

(2002). Spying 101: The RCMP’s secret activities at Canadian universities, 1917-1997. Toronto, Ontario, Canada: University of Toronto Press.

19.

Hilbert

(1980). Covert participant observation: On its nature and practice. Journal of Contemporary Ethnography, 9(1), 51-78.

20.

Kinsman

Gentile

(2009). The Canadian war on queers: National security as sexual regulation. Vancouver, British Columbia, Canada: University of British Columbia Press.

21.

Kramer

R. C.

Michalowski

Chambliss

(2010). Epilogue: Toward a public criminology of state crime. In Chambliss

W. J.

Michalowski

Kramer

R. C.

(Eds.), State crime in the global age (pp. 247-261). Devon, UK: Willan Publishing.

22.

Larsen

(2008). Governing non-citizens as security threats: Canada’s security certificate regime. In Ayyash

Hendershot

(Eds.), Violent interventions: Selected proceedings of the Fifteenth Annual Conference of the Centre for International and Security Studies (pp. 21-38). Toronto, Ontario, Canada: Centre for International and Security Studies.

23.

Larsen

Piché

(2009). Exceptional state, pragmatic bureaucracy, and indefinite detention: The case of the Kingston Immigration Holding Centre. Canadian Journal of Law and Society, 24(2), 203-229.

24.

Lather

(2010). Engaging science policy: From the side of the messy. New York, NY: Peter Lang.

25.

Lynch

(2000). Against reflexivity as an academic virtue and source of privileged knowledge. Theory, Culture & Society, 17(1), 26-54.

26.

Manning

(1980). The narc’s game: Organizational and informational limits on drug law enforcement. Cambridge: MIT Press.

27.

Marx

(1984). Notes on the discovery, collection, and assessment of hidden and dirty data. In Schneider

Kitsuse

(Eds.), Studies in the sociology of social problems (pp. 78-113). Norwood, NJ: Ablex.

28.

Mauthner

Doucet

(2003). Reflexive accounts and accounts of reflexivity in qualitative data analysis. Sociology, 37(3), 413-431.

29.

Mauthner

Doucet

(2008). Knowledge once divided can be hard to put together again: An epistemological critique of collaborative and team-based research practices. Sociology, 42(5), 971-985.

30.

Mazzei

O’Brien

(2009). You got it, so when do you flaunt it? Building rapport, intersectionality, and the strategic deployment of gender in the field. Journal of Contemporary Ethnography, 38(3), 358-383.

31.

McDonald

(2005). Studying actions in context: A qualitative shadowing method for organizational research. Qualitative Research, 5(4), 455-473.

32.

Mcleod

(2003). Why we interview now: Reflexivity and perspective in longitudinal study. International Journal of Social Research Methodology, 6(3), 201-211.

33.

Mills

C. W.

(1959). The sociological imagination. London: Oxford University Press.

34.

Monaghan

Walby

(in press). Making up “terror identities”: Security intelligence and Canada’s Integrated Threat Assessment Centre. Policing & Society, 21(4).

35.

Monahan

Fisher

(2010). Benefits of “observer effects”: Lessons from the field. Qualitative Research, 10(3), 357-376.

36.

Piché

Walby

(2010). Problematizing carceral tours. British Journal of Criminology, 50(3), 570-581.

37.

Punch

(2009). Police corruption: Deviance, accountability and reform in policing. Devon, UK: Willan Publishing.

38.

Roberts

(2004). Treatment of sensitive requests under British Columbia’s Freedom of Information Law. Freedom of Information Review, 10(9), 2-4.

39.

Roberts

(2005). Spin control and freedom of information: Lessons for the United Kingdom from Canada. Public Administration, 83(1), 1-23.

40.

Roberts

(2006). Blacked out: Government secrecy in the information age. Cambridge, UK: Cambridge University Press.

41.

Rosner

(2008). Behind the headlines: A history of investigative journalism in Canada. Oxford, UK: Oxford University Press.

42.

Siltanen

Willis

Scobie

(2008). Separately together: Working reflexively as a team. International Journal of Social Research Methodology, 11(1), 45-61.

43.

Smith

D. E.

(1999). Writing the social: Critique, theory, and investigations. Toronto, Ontario, Canada: University of Toronto Press.

44.

Smith

D. E.

(2001). Texts and the ontology of organizations and institutions. Cultures, Organizations, and Societies, 7(2), 159-198.

45.

Smith

D. E.

(2005). Institutional ethnography: A sociology for people. Walnut Creek, CA: AltaMira.

46.

Taylor

Walton

Young

(1975). Critical criminology in Britain: Review and prospects. In Taylor

Walton

Young

(Eds.), Critical criminology (pp. 6-62). London: Routledge & Kegan Paul.

47.

Thapar-Björkert

Henry

(2004). Reassessing the research relationship: Location, position and power in fieldwork accounts. International Journal of Social Research Methodology, 7(5), 363-381.

48.

Thomas

(2010a). Advancing access to information principles through performance management mechanisms: The case of Canada (World Bank Institute Governance Working Paper Series). Washington, DC: International Bank for Reconstruction and Development.

49.

Thomas

(2010b). Who is getting the message? Communications at the centre of government. In Forcese

(Ed.), Public policy issues and the Oliphant Commission: Independent research studies (pp. 77-130). Ottawa, Ontario, Canada: Minister of Public Works and Government Services Canada.

50.

Turner

(2006). Mapping institutions as work and texts. In Smith

(Ed.), Institutional ethnography as practice (pp. 139-162). Oxford, UK: Rowman & Littlefield.

51.

Walby

(2005). Institutional ethnography and surveillance studies: An outline for inquiry. Surveillance and Society, 3(2-3), 158-172.

52.

Walby

(2007). On the social relations of research: A critical assessment of institutional ethnography. Qualitative Inquiry, 13(7), 1008-1030.

53.

Walby

(2009). “He asked me if I was looking for fags . . . ” Ottawa’s National Capital Commission Conservation Officers and the policing of public park sex. Surveillance & Society, 6(4), 367-379.

54.

Walby

Larsen

(in press). Getting at the live archive: On access to information research in Canada. Canadian Journal of Law and Society.

55.

Walby

Monaghan

(2010). Policing proliferation: On the militarization of Police and Atomic Energy Canada Limited’s nuclear response forces. Canadian Journal of Criminology and Criminal Justice, 52(2), 117-145.

56.

Walby

Monaghan

(2011). Private eyes and public order: Policing and surveillance in the suppression of animal rights activists in Canada. Social Movement Studies, 10(1), 21-37.