Strategies for Obtaining Access to Secretive or Guarded Organizations

Abstract

Establishing contacts and gaining permission to conduct ethnographic or qualitative research can be time-consuming and stressful processes. Gaining access can be especially challenging when representatives of prospective research sites see their work as being sensitive and would prefer to avoid outside scrutiny altogether. One result of this dynamic is that many organizations that exert a profound influence in governing populations and regulating individuals’ access to basic needs are relatively invisible to the public and shielded from meaningful public accountability. Therefore, it is vital to effectively study secretive or guarded organizations and fill out the empirical record, which in turn could create the conditions for greater public awareness and debate. To that end, this paper draws on our collective research experience and the scholarship of others to present nine strategies that we have found to be especially effective for securing access to secretive organizations.

There are many challenges to performing robust field research. Depending on the area of inquiry, scholars might find that initiating the project is the most difficult step of all. Establishing contacts and gaining permission to conduct ethnographic or qualitative research can be time-consuming and stressful processes that require researchers to be creative problem-solvers. In some cases, researchers simply change their focus of study entirely when they encounter persistently closed doors at their selected field sites.

While some domains have historically been easier for researchers to access, the worlds of technoscientific knowledge production are notoriously difficult to study. As science and technology studies scholars, we are cognizant of the unique methodological obstacles to investigating scientific or high-tech organizations. Part of the challenge lies in overcoming dominant societal beliefs about impartial scientific facts and neutral technological artifacts. When scholars cultivate a sensibility that sees technoscientific domains as subcultures in their own right, engaged in value-laden practices, this in turn opens these domains up to critical inquiry and cultural critique (Helmreich 2009; Latour and Woolgar 1986; Traweek 1988). The greater challenge, though, is usually the much more pragmatic one of obtaining access to sites whose practitioners, for a variety of reasons, would prefer to avoid outside scrutiny. ¹ Our own research projects provide examples of this dynamic in action. In the case of organizations engaged in surveillance, practitioners are often keenly aware that any negative representation of their activities could invite public backlash, legal action, or dissolution of their organizations. In the case of private companies producing new technologies or pharmaceuticals, imperatives to guard trade secrets and protect brand images can compel such companies to seal themselves off completely from outside researchers. One outcome of these protective measures is that organizations that exert a profound influence in governing populations and regulating individuals’ access to basic needs are relatively invisible to the public and shielded from meaningful public accountability.

Therefore, it is vital to study secretive organizations effectively and expand the empirical record, which in turn could create the conditions for greater public awareness and debate (Nader 1972). To that end, this paper draws upon our collective research experience to present strategies that we have found to be especially effective for securing access to well-guarded field sites. Some of the sites to which we have gained access include Department of Homeland Security (DHS) “fusion centers,” intelligent transportation centers, hospitals, and companies conducting pharmaceutical clinical trials. Because our own research projects have centered on technoscientific organizations in the United States, we filter our discussion of secretive and guarded field sites through those experiences, but we also reference the work of colleagues who have used similar strategies to access a range of organizations in other countries.

In the pages that follow, we provide an overview of some of the scholarly literature on access before outlining a series of specific strategies for gaining access to reticent field sites. While the emphasis will be on offering practical tips for researchers to use and develop to meet their own needs, our larger goal is to encourage creative responses to the myriad challenges researchers face in the field.

Perspectives on Access

It is unfortunate that many qualitative research books have surprisingly little to say about obtaining access to field sites or recruiting interview subjects (e.g., Becker 1998; Charmaz 2006; Stark and Roberts 2002). In the wider literature, there is some consensus that one should work through existing networks to try to find a “known sponsor” or “orienting figure” who could provide referrals or facilitate entry into the field (Patton 2002; Vallance 2001; Weiss 1994) and that negotiating access is an ongoing process (Burgess 1991), but the assumption seems to be that because field sites vary radically, specific approaches might not be transportable from one site to the next (Emerson 2001).

When it comes to gaining access to secretive sites or “elite” informants, issues of power dynamics and differentials come to the fore. ² One practical difficulty is that individuals in positions of relative power may be more likely to refuse to participate in research studies (Berg 2009; Cochrane 1998; Hertz and Imber 1993). ³ Organizations typically have a range of gatekeepers as well, such as receptionists or assistants (Patton 2002), and some of them may functionally serve as professional gatekeepers, such as public relations staff (Thomas 1994), whose job is not just to filter outside requests but to respond to them, making it difficult for researchers to assert that their inquiries were not taken seriously. It also is not uncommon for informants to conduct searches on researchers before deciding whether to respond to requests (Ostrander 1993), and, as we have found, this often entails them reading the requestor’s publications or website material to ascertain whether the researcher’s politics are aligned with their own or their organization’s mission.

Other difficulties can arise in the agreements researchers make with the organizations they study. Once initial contact is established, researchers can engage in “commitment acts” (Feldman, Bell, and Berger 2003) to gain goodwill, or they increasingly consent to assist organizations in some demonstrable way as a condition (or precondition) of access. This “reciprocity model” (Patton 2002) of gaining entry can take many forms: interning, writing reports or grant proposals, giving lectures, conducting analyses of pre-existing data, introducing informants to others, or otherwise working on projects that are not a core part of the researcher’s study. While there can be benefits from reciprocity arrangements, they also introduce a host of risks to research participants as well as the researcher, including ambiguities with informed consent, unclear lines between what does and does not count as research, competing claims to intellectual property, co-optation of the research project, self-censorship, and exploitation of researchers, many of whom may be in vulnerable positions with respect to their universities or the job market.

The very notion of studying influential organizations or people in positions of relative power calls into question the applicability of other innovative modes of collaborative research that are intended to empower research participants. As Bruce Berg (2009) writes, “critical ethnography is an orientation where the researcher has a concern about social inequalities and directs his or her efforts toward positive change . . . and empowerment for participants” (199). Likewise, participatory action research has a strong social justice inflection and is typically mobilized by researchers in partnership with communities working to contend with and alter systems of oppression (Eubanks 2011; Fine and Torre 2004; Guishard 2008). Because of this tension between the goals of critical collaborative research methods and the relative security and influence of “elites,” some scholars have suggested that such empowering approaches may be inappropriate for studies with these populations (Bradshaw 2001). Certainly if some groups are actively endangering communities or violating the civil liberties of individuals, it would be unethical to empower them to execute those tasks more efficiently. Others contend that it is possible and desirable to use collaborative methods to reorient and “empower” elites to bring about meaningful social change (Kezar 2003). It is clearly true that power differentials exist within organizations too, so a more nuanced and contextual understanding of power is necessary to make any determination about the role and potential impact of one’s research (Mumby 2005; Smith 2006).

Covert research provides another potentially fraught approach to gaining access to secretive sites. Ethics review boards often have a broad definition of “covert research” that can even include observation of public behavior involving anonymous participants. Here we are referring to projects in which researchers join an organization as an employee, volunteer, or other integrated role in order to conduct their research clandestinely. ⁴ Assuming that there is a vital social need for the research, all other approaches have failed or would be likely to, and ethical considerations have been addressed and ethics review board approvals obtained, deceptively passing oneself off as an insider—or as a “potential participant” (Tewksbury 2001)—could be an effective way to secure access to a field site. Because informed consent would be impossible under such circumstances, it is questionable whether this type of deception in research could ever be fully ethical even if the research is socially important; still, there are instances where it may be necessary, and there is certainly a colorful history of such research projects (e.g., Balch and Taylor 1977; Humphreys 1970; Leo 1995). ⁵ In the context of “studying up,” especially of technoscientific domains, covert research has a strong chance of backfiring, as subjects will likely discover that the research has occurred (or is occurring), could publicly challenge and possibly sue the researcher and his or her institution, and could file complaints with ethics review boards, funding agencies, and professional societies, all of which could potentially destroy the career of the researcher. Another argument against covert studies is that researchers have an ethical obligation to those who come after them and should not salt the fields, thereby making future research impossible.

Depending on the area of study and existing networks, access to technoscientific realms can range from relatively easy to intractably difficult. If the research is occurring at one’s own university, securing access may be a relatively straightforward process of asking for permission and obtaining it, perhaps with some of the reciprocity conditions outlined above (Zenzen and Restivo 1982). This does not mean that such projects are easy to conduct, as the degree of technical literacy required for such research can be daunting, ⁶ which introduces further challenges for researchers to carve out space for critical social analysis when the impulse might be to relay technical details (as discussed by Hess 1997; Martin 1993; Shostak 2013; Woodhouse et al. 2002). At the other end of the spectrum, especially for secretive sites, researchers may never obtain access in the traditional sense and must devise alternative strategies for collecting data, such as the model of “polymorphous engagement” developed by Hugh Gusterson to conduct ethnographic research on nuclear weapons scientists by “interacting with informants across a number of dispersed sites . . . and collecting data eclectically from a disparate array of sources in many different ways” (Gusterson 1997). In Gusterson’s case, he drew upon multiple sources, including activist accounts, media representations, and specialized journals, but he also became a part of the community of weapons scientists in their non-work environments (e.g., churches, hiking groups, bars) and developed friendships with them, even if he could not observe them in their labs (Gusterson, 1997, 2004).

Clearly, some of the best qualitative research is defined and enabled by creative improvisation, persistence, social commitment, and ethical sensitivity. It is in that spirit, and with the goal of assisting others working in difficult-to-study areas, that we now turn to some viable strategies for obtaining access to secretive or guarded organizations.

Strategies for Obtaining Access

In this section, we offer some proven, less obvious approaches to obtaining access to secretive field sites or “elite” informants. Of course, personal contacts, referrals, and snowball sampling are also important techniques to arrange field sites or schedule interviews, but we know these traditional methods of access have their limits and can fail, especially with secretive organizations. Many of the suggestions come from our own experiences in the field, although we do include methods deployed successfully by our colleagues as well. Because the purpose is to provide possible starting points for researchers embarking on new projects or struggling with existing ones, the idea is that researchers would necessarily build upon and modify the strategies that they find most appropriate for their projects. Needless to say, this is not a comprehensive list but rather an initial contribution that we hope will serve as a resource for others.

Attend Industry or Government Conferences

While it may seem impossible to access some informants in their primary work environments, conferences serve as amazing venues for connecting with them, establishing initial rapport, explaining one’s project, conducting interviews, and—in the best of situations—receiving permission to visit other field sites. Additionally, unlike meeting potential informants in other social settings, at conferences people often embody their professional identities: performing for peers, conveying information, and telling stories, making such events ideal sites for data collection (Ortner 2010). Depending on the nature of the conference and the expectations of participants, it is even possible to receive approval from ethics review boards to treat presentations as data, especially if a persuasive case can be made that presenters are consciously performing for a semi-public audience and are aware that people are taking notes about and possibly recording their talks.

Furthermore, at industry and government conferences, it is common for organizations, especially technology companies, to set up booths displaying their products or services. Representatives of these organizations often must remain in their booths and effectively are a captive pool of potential subjects for researchers. Because representatives are paid to represent their organizations and liaise with other interested parties, it is typically quite easy to strike up conversations with them. Indeed, these representatives are likely to make the first move, so to speak, and introduce themselves, hoping to pitch their products or services to whomever pauses in front of their booths.

Even if one’s research project is not on technology vendors or their products per se, such people often have personal contacts in other organizations and are willing to provide names, phone numbers, and email addresses of people to which researchers might otherwise not be able to get access. Perhaps best of all, vendors then become “known sponsors,” allowing researchers to mobilize their names when they contact other individuals, thereby increasing the likelihood that one will be responded to because the new contact may feel a sense of social obligation to the referrer. For researchers studying the use of technological systems, such vendors may also be able to arrange for demonstrations of their systems in organizational sites to which it might otherwise be difficult to access.

We have executed this strategy for research on hospital tracking systems and DHS fusion centers (Fisher and Monahan 2011, 2012; Monahan and Regan 2012). The first step is to conduct Internet searches and review trade magazines to find possible conferences where desired organizational representatives will be present. With industry conferences in particular, the registration fees can be exorbitant, but sometimes one can get this price reduced or waived if researchers can make an argument that they are representing nonprofit educational institutions, that their interest in attending is “research” only (not business-related), and that they are students (if applicable). The key is finding a contact person, usually listed on the conference website, prior to registering and asking that person to look into what other registration possibilities exist. For government or combined industry and government conferences, it may be unclear whether one is permitted to attend, which is something we have found for security-related conferences. Even if they are restricted to government employees or individuals with security clearances, we have found some success in obtaining permission to attend. Some persuasion may be necessary, such as making a case that as an employee or student of a public university, the conference organizers could choose to view you as a government representative, even if that is not what they had in mind when they formulated the restriction.

Find the Names and Make Cold Calls

Because gatekeepers in organizations can be so effective, one key for getting past them is intoning the right password, which is generally the name of the person with whom you would like to speak. At times, especially as one’s familiarity with the research site grows, another way to gain access to the right person is by knowing their job title, but a lack of organizational symmetry across companies and agencies can make one’s best guesses at job titles ineffective. As we mentioned, vendors or presenters at conferences could provide names and contact information, but where does one begin if such connections are missing? For starters, we recommend conducting thorough searches of all available public sources on one’s specific area of study. For example, search Google News or LexisNexis to find recent media stories on the topic and then note which organizational representatives are quoted; usually that will give you a name and affiliation of someone who is comfortable talking with outsiders. Also, as long as sources are not confidential, journalists will sometimes be willing to provide you with names and contact information of individuals they have quoted—or even others whom they have not quoted. Moreover, in the best of circumstances, reciprocal relationships can be established between journalists and researchers so that the two groups can keep each other informed of new developments.

Other places that one could look for names include conference programs, shareholder reports, government documents, technical reports, newsletters, and grant proposals. Many of these sources, if they exist, can be found online. For instance, the website Public Intelligence has a trove of government documents related to security and surveillance. ⁷ Some of these documents have been leaked, some openly distributed by agencies, and others obtained through freedom of information requests. For research on DHS fusion centers, one of us (TM) has used these documents to find the names of potential informants, whether they are the authors of reports or individuals named in them. For research on the clinical trials industry, one of us (JF) has relied on a paid subscription to an insider newsletter, CenterWatch Weekly, which routinely publishes interviews with company representatives as well as stories about company initiatives in which employees are usually quoted. In one instance, after many failed attempts to make it successfully through the phone tree of a large corporation because no one seemed to know how to direct the call, JF happened upon an advertisement for the company that included a photograph of its team of vice presidents. With the photo helping to identify the correct person to contact to negotiate access to the desired research site, it was remarkable how using the right name completely changed the interaction between the researcher and the large organization.

Snowballing can be an effective recruitment strategy, but the ball sometimes needs a good nudge from helpful others. It is usually fruitful to ask for references from other researchers working on the topic or related ones. Although some researchers can be territorial and dismissive, as we have unfortunately discovered, most are willing to assist if they can and see more scholarly attention directed toward a subject they care about as positive for the field as a whole. One limitation to relying on other researchers is that their promise of confidentiality to their own informants can preclude them from providing a list of contacts as well as make other forms of introductions difficult. Operating without these restrictions, activists and civil society groups, like journalists, may have a ready list of contacts and be willing to share them with others. Finally, we always ask at the end of interviews whom else we should speak with, and on more than one occasion people have printed out a list of contacts saying, “you didn’t get this from me.” Those names, once vetted for relevance to the project, then become new passwords to try on gatekeepers, and—hopefully—new informants for the research study. At the same time, it is critical to be attentive to power dynamics in this type of sampling. Specifically, the ethics of snowball sampling has been questioned because researchers may not be fully protecting the confidentiality of participants already enrolled in the study (i.e., if the names of participants are used with others) and new contacts may feel coerced into participating if others recommended them (Brace-Govan 2004). As with the other strategies, one should approach such exchanges carefully so as to minimize ethical risks.

Making cold calls can be a very uncomfortable and difficult activity for anyone. It is also sometimes a necessary step in securing access to informants and field sites. Having the name or title of the person one is trying to reach helps tremendously but is not always an option. Likewise, once you do reach the person you are trying to, you may have only a short grace period to make your “pitch,” which should be well rehearsed. One way to warm up such cold calls is to provide advanced notice to informants that you will be calling them. In such a message, you could give information about your study and make a case for why their contribution would be valuable. Some scholars recommend doing this through email (Stephens 2007), but that could lead to informants quickly dismissing you by responding that they are not the right person or that they are not interested. Others suggest that sending a physical letter first, along with accompanying material, can have the effect of communicating the seriousness with which researchers are taking the potential relationship with that person (Weiss 1994). We have also found that it can be useful to prepare in advance a follow-up email that provides a summary of the details of the research study and can be sent immediately after a cold call; this can be especially effective when the potential informant wants to consult with colleagues, her manager, or a company’s legal team before making a decision. Although the success rate or appropriateness of each strategy will no doubt vary depending on the project and the personalities involved, the more one can do to establish credibility and rapport, the better.

Persistence is also crucial. Gaining access to field sites is often a multi-staged process that requires the researcher to make contact several times. In our experience, a cold call often leads to the potential informant instructing us to call back in a set number of days (or sometimes weeks or months). Even when we have gotten the distinct impression that the contact was likely a dead end, we have continued to follow up with contacts until they have agreed to participate in our studies or they have explicitly declined. While difficulty connecting with informants by phone or email can be read as an implied “no,” it is important to remember that these contacts are busy and cannot be expected to prioritize returning a researcher’s call or email. Robert Weiss (1994) tells a story about a phenomenal phone recruiter who would—respectfully—not take “no” for an answer. In order to try to figure out why she was so successful when others were failing, he role-played with her, with him acting as the potential interviewee:

I said, “No, I’d rather not participate.” Mrs. Adams said, “Yes, of course, I understand, but I want to tell you why the study is being done and who is doing it.” And she went on to tell me about the sponsorship of the study and the kinds of questions that would be asked and how important it would be to have my perspective. . . . She said that I would find the interview interesting and that it would be held whenever and wherever suited me. By now I was intrigued by the study and flattered to be so wanted, as well as being just a bit exasperated by being unable to escape. I said, “All right, let’s set a time.” (36)

One should be respectful of informants and potential informants, but it could be, as in this example, that if they better understood why their participation was so important that they would consent, even with initial reluctance. In cases where individuals communicate that they are not interested at the moment, we ask for permission to follow up with them in a few months, and because there is no commitment, but there might be some social awkwardness to refuse, it is rare for anyone to say, “No, I’d prefer it if you didn’t.” Naturally, we do follow up, and because we are then more familiar to them, we tend to have a better success rate the second time around.

Communicate Legitimacy

It can be helpful to know that informants may have been burned in the past by journalists or researchers. Gaining the trust of such informants, if possible, may require an intuitive, sympathetic understanding of that history and response to it, even if they do not communicate such concerns directly. We have found that recruitment is much more successful and interview data more detailed when we first explain our approach to social science research and differentiate our project from those of hypothetical others. This is a mechanism of asserting legitimacy as committed researchers who are sincerely interested in finding out answers to questions, not simply seeking to produce exposés. We are certainly dedicated to invested and engaged research that brings about progressive social change, and following the tenets of science and technology studies (STS) and surveillance studies, we do not believe that research is a neutral or impartial activity (Ball and Haggerty 2005; Barnes and MacKenzie 1979; Haggerty 2009; Haraway 1991; Monahan 2011; Restivo 1988). Indeed, we have argued that researchers should be pursuing more interested, reflexive projects oriented toward solving social problems (Monahan 2008; Monahan and Fisher 2010). That said, informants need to know that you are not out to get them, that you are truly curious about their worlds, and that you are open to and interested in unanticipated findings that could alter your initial take on the larger issue being studied.

Beyond that, there are several other related ways to communicate legitimacy, which could in turn assist with access. First, most academic researchers have an institutional affiliation that should be referenced explicitly (e.g., “I am a doctoral student at X university” or “I am a professional researcher working for Y research organization.”). ⁸ Whenever possible, communicate that institutional legitimacy through its various symbolic trappings, such as letterhead, email signatures, and even email addresses (e.g., if one has a choice to use a university’s own email system instead of Gmail, the address with the university’s name will carry more weight). Second, if an external sponsor is funding the research, such as the National Science Foundation (NSF), then that should be stated in an explicit but accurate way (e.g., “I am working on a study that I conceptualized and am directing with financial support from NSF.”). Third, any previous personal experience or that of one’s research mentor in the area of study could be mentioned as well, provided that it would not alert potential informants to controversial work that could give them pause. In some instances, one may have a background in the area apart from the role of researcher, such as working previously in a similar setting, in which case mobilizing the specialized argot of such practitioners could strengthen legitimacy and assist one in becoming an “indigenous ethnographer” (Patton 2002).

Finally, while such approaches do not sit easy with us, self-delegitimizing efforts can sometimes succeed too. For instance, one could draw upon popular misconceptions of academic research and the less-threatening status of student to diminish the potential risk of participating in research (e.g., “I’m [just] a doctoral student working on a research project.”) On occasion, although admittedly less often lately, even as faculty members our informants have mistaken us for students working on class projects and needed to be corrected. Usually these misperceptions came to light only after conducting an interview or spending time with the informant, however, so the delayed correction, combined with the presumed innocuousness of our studies and the self-professed generousness of informants, probably improved our access. There is an acknowledged, uneasy tradition of “playing dumb” in the field to establish rapport and elicit less-directed informant accounts (Chong 2008; Lofland and Lofland 1995; Scott et al. 2012), but we must also confront the power and gender implications of such moves—they position researchers as naive supplicants in need of guidance from wise others, which is especially evident in cases where women researchers have found increased success by pretending to be less knowledgeable than they really were, particularly when interviewing older men (McDowell 1998). ⁹ While it is not the focus of this paper, it is relevant to ask whether de-legitimizing practices are worth the access they afford, given that they could negatively shape public perceptions about the value of social science research.

Reduce Perception of Threat

Secretive or guarded organizations are likely to perceive researchers as introducing a threat to their operations. Unlike “underdog” organizations that might believe that a researcher could help publicize or legitimate their cause or business, ethnographic investigations of powerful organizations can produce anxiety and suspicion that the research will unearth information that will damage their reputations. ¹⁰ As mentioned above, explaining how social science research is different from journalism—especially in that the goal is not to produce an exposé—can minimize an organization’s perception that the researcher introduces a threat. Understanding the more complex and multifaceted ways in which potential informants might perceive the threats of participating in research is helpful for successfully negotiating access.

One common pitfall for researchers during their interactions with potential informants is relying on terminology that has a negative connotation for the organization and signals a threat. While it is not always possible to know the best way to frame a new project when seeking access, researchers should be attentive to descriptions of their studies that may have a chilling effect. As one example, most social science scholars perceive the word “ethics” to be fairly neutral, indicating perhaps a field of study, a mode of engagement, or even procedures to be followed. In the field, however, the word itself is often negatively valenced. Regardless of what the researcher intends, to tell an organization that one is interested in “ethics” can be tantamount to saying that the study is designed to uncover ethical breaches or other insalubrious practices. Often the objectives of the study are not what produce the threat but the framing alone. In one of our areas of research (JF’s), clinical research organizations tend to be quite interested in research questions regarding the informed consent process, trial participants’ motivations to enroll in drug studies, and physicians’ relationships with the pharmaceutical industry. All of these topics fit nicely within the frame of “ethics” within academia, but for industry, it is better to ask questions about the routine practices of those being studied. As another example, in the domain of surveillance, TM has found that “privacy” is the threatening term that makes access difficult or is a nonstarter in interviews. It is not that potential informants are unwilling to discuss their practices surrounding data sharing and retention, but rather that for them the word privacy is encoded with a politics that makes them feel vulnerable to criticism. In both of these settings, the use of alternative, indigenous frames (e.g., informed consent or data sharing) more directly conveys the purpose of the research and reduces the unease of subjects. Every research project uses terminology that can raise red flags, and it is the responsibility of researchers to discover the best—and least threatening—way to communicate with the organizations they want to study.

Even when potential informants are interested in or supportive of the research being conducted, there are other obstacles to gaining access that hinge on minimizing other perceived threats of bringing an outsider into an organization. For technoscientific organizations in particular, informants often voice serious concerns about how researchers will handle information they glean relating to trade secrets, identities of the organization’s clients, and other legitimate threats to their operations. It will not always be obvious in advance what perceived threats will be articulated, but researchers should develop responses to some of these concerns so they are prepared should they arise. For example, in JF’s projects on clinical trials, many of her prospective field sites wondered whether participating in her research would jeopardize their clients’ (i.e., pharmaceutical companies) proprietary information. Because the studies she has conducted have not hinged on specific pharmaceuticals, she was able to quickly dispatch with their concern by assuring them that she would not collect that information as part of the research.

Providing confidentiality to your field sites can help minimize the concerns expressed by organizations’ various representatives, but researchers should not take for granted that potential informants will understand what this means for them personally or their organizations. To maximize the benefits to the research, we recommend being concrete about what confidentiality means in practice, including how information about the organizations will be reported or modified to protect their identities. At times, signing a nondisclosure agreement (NDA) can convince the organization’s representatives that the site will be protected because specific treatment of its identity or proprietary information has been detailed. ¹¹ In other cases, researchers have agreed to allow organizations to review any subsequent manuscripts produced from the research. Neither of us has personally agreed to such a condition of access, but if one does so, clear information should be given about the expected time frame for organizations to review those publications before the researcher has the right to proceed (so the publications are not held hostage by busy or hostile organizations), as well as caveats that the organization has the right only to request changes to the manuscript for accuracy and does not have the authority to stop publication altogether.

Negotiating access to a specific field site can also involve conversations with multiple people from the organization, with each often representing a different division and voicing unique concerns about being involved in a research project. Cultivating an advocate in the organization can be an enormous help, especially if that person can persuade others that the research poses little risk. We have learned that even when advocates are willing to talk with others in their organizations, it is usually more effective to insist on being included in those conversations. Otherwise, access may be denied if the advocate cannot answer questions or address concerns raised by his or her colleagues. This is especially true when an organization’s legal group reviews the research request. When researchers can become individuals with whom members of the organization interact, these personal communications facilitate trust and help reduce perceptions of threat.

Coordinate Coincidence and Make Barely Announced Visits

In almost all of our projects, there have been important sites that simply resist granting access. Rather than provide definitive rejections, these sites normally will delay making any decision, saying things like, “Now’s not a good time” or “I’ll have to check with someone else,” or they simply will not respond at all. Counterintuitive as it may sound, for some reason when we give sites an open calendar and let them know that we will come whenever is convenient for them, this creates a sense of inertia, where our flexibility effectively translates into permission to postpone indefinitely. Normally we do start out being completely adaptable, within reason, to the scheduling that would work best for our informants, but when this fails, offering a narrower window often helps (e.g., “For the purpose of the project, we really need to conduct the site visit sometime in the next month.”) Then again, it can backfire—if the next month is not feasible for them, they could more easily decline to participate. This has led us to develop an alternative strategy that leaves open the possibility of future visits.

Making coincidental or barely announced visits seems to succeed in achieving access to secretive or guarded sites when all other approaches fail. The way they appear to work best is when the researcher resides in another geographical location, the farther away the better, and schedules a trip to the city or region of the prospective research site, then lets contacts at the site know that he or she will happen to be in town during those days and would love to stop by. Being able to provide a solid rationale for one’s travel to the area, such as an academic conference, certainly helps but is not necessary. Then leading up to the trip, we continually communicate with informants at the research site, saying things like, “I’ll be in town for only these two days. Your site is known for doing some very innovative work, so it’s important for us to include you in the study. Could you find any time in those two days to meet with me?” Even if there is no clear agreement, if we really do travel to the area, we call and email our contacts a few times a day when we are in town. Then, if sites still do not respond at all, we inform them that we will be showing up on a certain day and time and that we hope they will be able to meet with us, which is a claim that one of us once had to follow through on, with success, but usually informants grant permission for a visit before it comes to that. The more difficult it is to get permission from a site using this strategy, the more likely that the access the researcher can hope to gain will be limited. For instance, it is very difficult to pull off a barely announced visit to perform an ethnography of an organization, but this approach tends to succeed for conducting interviews and for short-term observations. Thus far, for those few holdout research sites, this strategy has not failed us; it has worked to get us access to DHS fusion centers (Monahan and Regan 2012), phase I clinical trials facilities (Fisher, 2015), and hospitals (Fisher and Monahan 2012; Monahan and Fisher 2011).

Mobilize Indirect Access

There are some organizations that are simply impossible to gain permission to study. It might be difficult to convince representatives of secretive or guarded organizations that the research is not a threat to their operations or that they can gain something meaningful from their participation. In some cases, there might be other organizations that will grant access, and the research can move forward. In other cases, a specific site is vital to the success of the research, and it becomes necessary to envision other means to collect data about that organization. As we described above, Hugh Gusterson (2004) conducted research on nuclear weapons scientists by integrating himself into their nonwork communities. More recently, Kenneth MacLeish mobilized similar ethnographic methods to study the military community of Fort Hood, Texas, by interacting with soldiers when they were not on the military base (MacLeish 2013).

In addition to immersing oneself in the community surrounding a research site (which might not be effective for some organizations and locations), there are other strategies of indirect access that can provide the researcher with data about an organization. While the researcher might have to forego ethnographic or observational methods, data about organizations can often still be obtained from interviews with other outsiders who do have access to those sites and can provide their first-hand experiences. In technoscientific realms, these individuals who have personal knowledge of companies, laboratories, federal agencies, and so on have many roles, including technology vendors, contractors, inspectors, and clinical trial participants. These data may be primary to the research, or they may become important supplements to provide broader trends or context on the field sites included in a study. For example, in our collaborative research on hospital tracking systems, we gained direct access to more than a dozen hospitals, but through our interviews with technology vendors we were able to collect stories and information about twice as many sites. Technology vendors also proved critical for TM’s project on DHS fusion centers, which are much more secretive organizations than are hospitals. Faced with difficulty getting direct access to the desired number of fusion centers, technology vendors and civil society groups helped to provide critical information about the types of information systems in use, as well as their limitations. Additionally, outsiders who have limited access to multiple organizations in a research area can shed light on why some sites are more willing than are others to provide the researcher with access. For example, in JF’s study on Phase I clinical trials, the healthy trial participants she recruited frequented many research clinics in their regions and as a result were able to describe each clinic as well as its practices and culture in detail. JF was able to quickly ascertain that the research clinics that had denied her request for access were by and large the facilities about which the trial participants had the most complaints or negative experiences. In this particular research project, the participants were able to help minimize the extent to which the sites that participated in the study biased the results because participants provided data on a larger sample of clinics. This underscores how indirect access can remain an important methodological tool for secretive or guarded research sites, even when the researcher is able to secure direct access to a sample of organizations.

File Freedom of Information Requests

Many governments throughout the world have provisions that allow citizens and noncitizens to access documents produced by government agencies. The provisions go by a number of different names, such as Freedom of Information (FOI), Access to Information (ATI), or Open-Records (on the level of U.S. states), but the general idea is that transparent governments are more democratic and accountable, and therefore less likely to abuse their authority. For instance, the U.S. strengthened and expanded its 1966 Freedom of Information Act (FOIA) in the period following the Watergate and COINTELPRO scandals in the 1970s with the goal of discouraging future large-scale government abuses; since then, however, especially during the 1980s and 2000s, FOIA was weakened, making access to information more difficult (Price 1997, 2010).

Like other strategies of indirect access, freedom of information requests can be especially effective for studying secretive government sites. Kevin Walby and Mike Larsen (2011) explain that beyond security- or intelligence-related research, such requests can also support research 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” (33). And while there are restrictions, such as not being able to request information about living individuals other than oneself through FOIA, there are innumerable points of contact between groups, organizations, and government agencies, each of which affords opportunities for information requests. One important consideration, though, is that FOIA requests can sometimes take years to be fulfilled, and when they are, the files might not be that revealing, or might be redacted beyond the point of being able to say very much about them (Price 1997), so we do not recommend this as the sole method of data collection, especially not for doctoral students needing to complete dissertations or junior faculty needing to secure tenure.

Practically speaking, filing freedom of information requests is fairly straightforward. David Price, who is one of the leading proponents of this method, explains: “all that is needed is a letter addressed to the FOIA officer at the government agency of interest, specifying exactly what records are sought” (Price 1997, 13). Depending on the agency, there could be a fee involved, ranging from nominal to potentially prohibitive (e.g., 10 cents a page to $500 for a basic search regardless of whether anything is found), although fee waivers are possible for scholars (Price 1997). Some tips for success include finding the right agency and FOIA officer, making requests as specific as possible, using the agency’s terminology for the group or activity being studied, and following up to ensure that the request was processed correctly (Price 2010; Walby and Larsen 2011). Fortunately, it is not too difficult to find online sample request letters ¹² and contact information for FOIA officers. ¹³

As a word of caution, just because government documents can be obtained does not mean that it is ethical to divulge or widely circulate the contents of those documents (Price 2010). For instance, many employees might send emails under the assumption that these messages were confidential, and if these are not directly related to an issue of public importance, personal and professional codes of ethical conduct may dictate that the information should not be circulated beyond the researcher who receives the documents. Relatedly, and to put this in perspective, if research is publicly funded or a researcher is working at a public institution, documents or emails produced by the researcher could also be subject to FOIA requests (Price 2010), as has happened (Charlesworth 2012).

Triangulate Internet Data

Increasingly, data on secretive or guarded sites can be collected from a heterogeneous mix of Internet sites. For instance, whereas freedom of information requests may take years to fulfill, many times researchers can speed up this process by tapping into the online repositories of civil society groups or activists who have already filed such requests and obtained documents. In the surveillance and security domains, Public Intelligence, ¹⁴ the Partnership for Civil Justice Fund, ¹⁵ and the American Civil Liberties Union ¹⁶ are just a few groups that file FOIA requests and circulate received documents online. Relatedly, sites like WikiLeaks ¹⁷ are renowned repositories of vast numbers of confidential and formerly confidential documents, many of which have not been systematically analyzed by social scientists; the organization, its members, and its whistleblower contributors, like Chelsea Manning, have paid high prices to make data public with the objective of ensuring more transparent and accountable governments (Gilliom and Monahan 2013), so we would argue that the research community should do its part to make sense of those data.

Earlier we suggested that shareholder reports, industry publications, and government reports could be mined for the names of potential organizational contacts. These, and related documents, which often can be found online, can also be analyzed for their content. For publicly traded companies, Security Exchange Commission filings, like quarterly reports, can be downloaded from company websites or finance websites (e.g., Yahoo! Finance ¹⁸ ). For government agencies, one can also download budgets, audits, transcripts from hearings, and other reports that could be used, for example, to “follow the money” and see what programs and partnerships are being supported (cf. Hayes 2009) or to track tension and political infighting behind the scenes (e.g., reports produced by the U.S. Government Accountability Office or Congressional Budget Office).

Other underutilized resources include the many company and government insider communication platforms that are open to the public. These could include blogs, RSS (real simple syndication) feeds, wiki pages, listservs (and their archives), and others. For instance, the Department of Homeland Security runs a number of RSS feeds ¹⁹ that provide insight into stories and events about which it thinks its practitioners would like to know; these feeds could assist researchers in staying up-to-date on current homeland security developments and finding events to attend, but they could also serve as metadata or contextual data about the culture(s) of homeland security and its public face. Similarly, many local jurisdictions run their own RSS feeds that are far less polished, and therefore more interesting, than their federal-level counterparts. Simply by searching “rss feed,” “blog,” or “listserv”—along with the keywords of one’s research area—should turn up some sites of interest.

A final related possibility is to download the entire website of an organization you are studying. Oftentimes you, and perhaps individuals at the organization, would be surprised just how many documents this could turn up: draft reports, PowerPoint presentations, minutes of meetings, photographs, old site content, directory information, and much more. Although legal guidelines pertaining to such access may differ across countries, and professional ethical standards may dictate whether you utilize this strategy or how you mobilize the data you find, it can be a highly effective way to collect raw data. From our perspective, this does not constitute hacking if all the files are openly accessible without passwords, not hidden behind security firewalls. Some of the current software for downloading entire sites includes “Sitesucker” ²⁰ for Mac platforms and “HTTrack” ²¹ for PC and Linux platforms, but many other possibilities exist.

Initiate and Follow-Up on Multiple Leads Simultaneously

Because gaining access to secretive and guarded organizations is challenging, it is quite difficult to predict the outcome when making contact with potential field sites. Many scholars use a sequential approach to pursuing leads, especially when they perceive certain organizations as more valuable or desirable for their research. The risk of this approach, however, is that a lot of time can be lost on an organization that eventually denies or reneges on the access request. In order for studies to commence as quickly as possible, we recommend contacting several organizations at once. Staggering initial contact with each potential field site can be helpful in order to learn how best to discuss the study and interact with representatives from the organizations. ²² Waiting for a final determination on a request, however, can cost the project months of effort and leave the researcher back at square one. If several organizations approve access, researchers are in the enviable position of being able to select their preferred field site or expanding the scope of their projects. It is also important to avoid being derailed by rejections from specific organizations; working multiple leads simultaneously is often a useful mechanism to maintain momentum and morale on a project.

Concluding Thoughts

We have detailed nine strategies for obtaining research access to secretive and guarded organizations, and we would be remiss not to offer a final note on research ethics. Underlying all of our suggestions for researchers’ interactions with organizations is the desire to gain access to a field site but to do so in way that is respectful and authentic. We assume that researchers are not trying to play “gotcha” with their field sites but are instead mobilizing robust qualitative methods to study organizations and social phenomena. Qualitative research—and ethnographic methods in particular—can involve heightened ethical responsibility to study participants. Forming what are often temporary but nonetheless close relationships with informants is problematic when those informants might not be fully aware of or consenting to data being generated from informal conversations and observations of their daily practices. In some instances, informants might even experience a betrayal of trust when the data collection has concluded and the researcher begins publishing (Bosk 2001). While it might not be possible to avoid these misunderstandings, researchers do have the obligation to remain open minded and fair in their representations of the individuals and organizations they study.

Accurately and meaningfully representing research to potential informants can be difficult. Scholars should see their interactions with representatives at conferences or through cold calls as opportunities to create alternative frames for their projects that can help establish common ground with organizations so they can more fully understand the value of participating in social science research. Likewise, we see our recommendation to coordinate coincidence and make barely announced visits as a way to help a busy organization prioritize an outsider’s schedule, project timeline, and resources; participation in such research should always be voluntary. Finally, even when data about organizations are obtained through indirect access—through third-party informants, freedom of information requests, or Internet sources, there is still an ethical obligation to consider when and how it is appropriate to name organizations or keep their identities and communications confidential.

We decided to highlight the access strategies we have used successfully because we know firsthand how challenging it can be to conduct ethnographic and qualitative research on secretive and guarded organizations. We also know how intellectually rewarding and socially relevant it can be to investigate companies, government agencies, and other organizations that tend to eschew outside scrutiny. While there is no one formula for gaining access that will be effective with all organizations, our nine strategies should underscore both the need to persevere and be creative in trying different approaches with potential informants.