Research in carceral contexts: confronting access barriers and engaging former prisoners

Controlling research access and limiting carceral knowledge

CSC is the Canadian federal government agency that is responsible for administering sentences of two years or more, while provincial/territorial systems administer sentences of less than two years. We focused our PNSP study on the federal prison system because federal policies and programs create a national framework and context which can influence provincial decision-making, and thus support for in-prison programs at the federal level could have a direct impact on program implementation within provincial corrections. Further, we recruited people who have served federal sentences because of the frequency of injection drug use within these prisons, something that has been acknowledged by the federal system itself (Zakaria et al., 2010).

CSC manages 43 federal institutions, 15 community correctional centres, and 91 parole offices across five administrative regions in Canada. ¹ It also conducts research on prisoner populations and correctional programming, including research on institutional operations and security, the needs of women and Indigenous prisoners, and in-prison mental health issues and substance use. ² External researchers and stakeholders wishing to collaborate or conduct their own independent research involving current prisoners or employees of the prison system must obtain approval from CSC in addition to securing any mandatory approval(s) from their university or other institutional research ethics board (REB). Details about how to propose research to CSC are sparse on the agency’s website and interested parties are asked to contact CSC for more information. Research proposals are processed and reviewed by a committee through the centralized Research Branch located in Canada’s capital, Ottawa, while regional research branches may also assess proposals that involve federal correctional facilities in a given geographic area. To the best of our knowledge, and based on recent personal communications with other prison researchers, this procedure has not changed since we conducted our study.

According to CSC’s website, the agency welcomes collaborative research partnerships and widely shares findings from their own studies. In practice, however, this is not always the case. CSC-collected data is not readily accessible to external researchers, and some have documented their experiences of being denied access to CSC information, prisoners, and staff (Martel, 2004; Yeager, 2008). Indeed, a recent case study found that CSC openness to external researchers has become less common over time, with a limited number of collaborations occurring per year; out of 66 project descriptions in CSC’s 2010–2011 research plan, only six mentioned external partners as collaborators, and though an additional 31 project requests came from external researchers during that plan’s timeframe, there is no indication of how many were approved (Watson, 2015).

This nebulous setting of external research approval and cases of denied access are significant since carceral research is already challenging and time-consuming (Castellano, 2007; Goodman, 2011). Importantly, access barriers and the denial of approval for prison- or prisoner-related research can seriously limit the scope and quality of correctional knowledge. Researchers facing study-related restrictions will sometimes undertake efforts such as targeting specific prisons that are more likely to grant approval and/or modifying their proposals and study instruments (actions that we also engaged in), which can ultimately affect the production of prison knowledge and may create skewed data results (see Hannah-Moffat, 2011; Martel, 2004; Spivakovsky, 2011, Watson, 2015). Those wishing to conduct qualitative or critical studies appear to be particularly disadvantaged by CSC’s hegemonic control of research activities. For example, in Martel’s (2004) attempt to interview federally-incarcerated women, CSC’s response to the research request drew a clear line between qualitative and quantitative methodologies, and included a strong recommendation to incorporate the latter. Thus, access barriers not only prevent certain kinds of research from occurring in the first place, but can also shape the kinds of projects that eventually get approved, and therefore can have a substantial impact on carceral knowledge overall.

One way to attempt to overcome these kinds of barriers is through access to information and freedom of information (ATI/FOI) requests (Piché, 2015; Wright et al., 2016; Yeager, 2006). ATI/FOI mechanisms are useful for getting at a range of internal records and reports (Savage and Hyde, 2014; Walby and Larsen, 2011b). However, while these requests can perhaps ‘overlap with more conventional qualitative ways of producing and making sense of data’ (Walby and Larsen, 2011a: 37), they are not designed to produce richly detailed firsthand accounts of lived experience. Given the short timeframe in which we needed to complete our data collection, as required by our funder, our specific data goals (i.e., to speak with prisoners and others about harm reduction programming that does not yet exist in Canadian prisons), and the length of time it generally takes for ATI/FOI requests to be granted or denied, we did not pursue this option. ³ Instead, we worked with key community partners and collected qualitative data through one-on-one interviews and focus group sessions to better understand experiences of health- and drug-related behaviours and services in prison.

Reflections on access barriers and impacts on the research process

Before describing our research process and access barriers, we want to acknowledge that the qualitative study upon which this article is based would not have been possible without collaborative partnerships with two organizations that have been working on issues related to prison health services for decades – PASAN and the Canadian HIV/AIDS Legal Network. ⁴ These organizations had ongoing professional relationships and histories of working together, which proved extremely helpful for creating a cohesive and mutually supportive research team, as well as a high level of trust and comfort for participants in the study. Representatives from these organizations on the team possessed in-depth, professional knowledge of relevant community support services, correctional (especially CSC) practices and policies, prison culture(s), and legislative frameworks. Their leadership was instrumental in the research design and planning process as well as for the development of the ethics applications, for participant recruitment, and for data analysis to ensure more valid and grounded interpretation of the results. Dissemination of some of the research findings has also happened collaboratively. Other local stakeholders and community organizations were likewise involved, to different degrees, in aspects of the research process. ⁵

To capture a wide range of perspectives on in-prison injection drug use and PNSPs, we initially planned to engage multiple participant groups including current and former prisoners, prison healthcare staff, and expert stakeholders. This plan meant that we needed ethics approval from the REB at Ryerson University, where the principal investigator (EvdM) is based, and approval from CSC’s Research Branch. Our familiarity with CSC’s control of research had us expecting challenges in getting approval to hold prison-based focus groups and interviews with staff, especially given the contentious and political nature of our study, as Canada is a jurisdiction with limited prison-based harm reduction programming and no PNSPs (Watson, 2014). As the research team developed the two requisite applications, we decided that we would conduct additional community-based focus groups with former prisoners in the event that CSC denied our request to hold in-prison groups. With the research team’s close connections to organizations that support criminalized and marginalized populations, we did not anticipate that recruitment of former prisoners would present any significant challenges. In other words, we assumed at the time that we would likely confront barriers to gaining approval from CSC, but would not face the same access challenges in the community. In hindsight, we concede that we perhaps did not sufficiently consider the blurred boundaries between ‘prison’ and ‘community’ (see Cohen, 1985), a point we revisit when we discuss parolees below, as we initially focused on gaining access to prisons where, historically and to date, ‘private’ exercises of punishment are carried out (Foucault, 1977).

The two applications we developed were similar but not identical. In the ethics protocol for the university, we included questions about injection drug use in prison, harm reduction programming, and specific PNSP models; whereas in our research application for CSC, we made the questions more general, focusing on prisoners’ access to and perceptions of health-related services, including programming designed to prevent transmission of HIV and hepatitis C in prisons, with PNSPs listed among different examples. That is, we broadened the questions we would pose to current prisoners in the CSC proposal – a decision with the potential to shift the direction of the research or, more precisely, our ability to obtain targeted information about PNSPs from prisoners themselves – thinking that this approach would more likely result in approval from the federal Research Branch. Our creation of the more general application for CSC was not to be disingenuous about the topic of the study, since our application was still focused on infectious disease transmission and health in prison, but was an effort to pre-empt the ‘stalls’ and blockages in carceral research access that are well documented in the literature (Lippert et al., 2016; see also Wright et al., 2016). We see, especially in retrospect, how this effort joins other previously noted strategies to modify research proposals in the hopes of gaining research access, which may limit the development of more comprehensive and authentic carceral knowledge.

The principal investigator sought preliminary, pre-submission feedback from both CSC’s Research Branch and the university REB since the two applications were being developed at the same time and, as such, modifications to one could be incorporated into the other, if applicable. This correspondence proved helpful for amending the applications. Perhaps unsurprisingly, feedback from CSC that emerged during email and phone conversations differed substantially from feedback received from the university REB. A notable comment from the former, for example, included a suggestion for consistency in language throughout the proposal, specifically that our application ‘blend with [CSC’s] current terminology usage’ (email dated 12 September 2014) regarding the term ‘offender’ instead of ‘prisoner’. While we accommodated all other suggestions received from CSC, which were relatively minor in nature, in the final application, we did not comply with using ‘offender’ because research team members agreed that this term is stigmatizing and problematic.

During this process of correspondence with a representative from the Research Branch, who was responsive and supportive, we were verbally informed that we would also need CSC permission to interview people who were currently on parole. We had not anticipated this access restriction and did not see mention of it in any of the research guidelines or policy documents we received from CSC, nor in what we retrieved from their website and other sources. With interest in being granted access to conduct prison-based focus groups, the research team decided to keep the CSC application tailored to current prisoners and healthcare staff only, and to cooperate with the ostensibly unwritten rule about not recruiting people on parole for our community-based groups. This action could be viewed as a further example of our acquiescence to CSC in an attempt to gain access.

Our decision to cooperate with, rather than challenge, CSC’s directive regarding people currently on parole was also informed by comments from the university REB as well as research team members’ level of comfort with potential consequences to participants. As we were updating the CSC application, members of the research team also had meetings with the university REB to discuss the board’s concerns, many of which focused on possible risks to participants. As we discuss below, the university REB posed questions about how we would safeguard both privacy and confidentiality for people who were recruited from community settings. As a team, we agreed that minimizing participant risk was a primary ethical concern. People on parole are an already marginalized population, still under the surveillance of the correctional system, and even if the likelihood of discovery of their research participation by parole officers is very low, we did not want to inadvertently introduce potential harm. This is a prime example of where the boundaries between prison and community are blurred, and where the processes of the former are reproduced to some extent in the latter (see again Cohen, 1985). That said, we recognize there are important and unanswered ethical, legal, and theoretical questions that emerge from this experience, including the question of CSC’s rightful authority to monitor or dictate whether or not people on parole can participate in external studies.

The updated university ethics protocol was submitted to the REB in May 2014 and approval was received six weeks later. In September 2014, we submitted our final research application to CSC. The response from CSC’s research review committee was swift – notice that they would not approve the study was received just over a week after application submission. Despite the quick rejection, it was stated that the ‘committee was extremely impressed with the quality, completeness and professional nature of [the] proposal’ (email dated 26 September 2014). While this correspondence and preliminary feedback on the proposal displayed enthusiasm for the study, the principal investigator was advised that the major reasons for rejection included the following: that CSC already collects data on HIV and hepatitis C transmission and prevalence, and has been doing so since 2000; that CSC monitors when prisoners contract HIV and hepatitis C (i.e., before or during federal incarceration); and that research is underway at CSC to assess risk factors and behaviours, including a recent cross-Canada study that produced results being used to revise programming. CSC’s research review committee did not cite specific concerns over the methodological aspects of our study, ethical considerations, or the general research plan. Effectively, the official response conveyed that CSC conducts their own monitoring and research on infectious disease, ⁶ enough so that external studies on such topics are not needed. This response fits well within an organizational risk management strategy that seeks to protect CSC’s agenda by impeding external researchers (Watson, 2015), and is in keeping with the denial of access that other scholars have documented (Martel, 2004; Yeager, 2008). The rejected application meant that we would not be able to hold focus groups in federal prisons, nor would we be able to consult prison healthcare staff as participants. This presented a significant, although somewhat expected, limitation for our data collection, one that we aimed to mitigate by recruiting a greater number of people who have previous lived experience of incarceration.

Recruitment of ex-prisoners began in October 2014. Managers and staff at numerous organizations – including local harm reduction programs, health services, and Indigenous organizations – agreed to assist with study promotion and recruitment, and appeared highly interested in the study. Electronic and hard-copy recruitment information and flyers were sent through organizational channels and networks across Ontario to reach service users and contacts. Despite the considerable support and effort from the organizations that assisted us, we found recruitment to be challenging, even in a city that has numerous federal prisons in its vicinity. We note below that former prisoners can be a hard-to-reach group, and there are multiple reasons why some eligible persons might be reluctant to participate in a study about how to address in-prison drug use and infectious disease. ⁷ However, we believe the recruitment challenges were, at least in part, due to the abovementioned access restriction introduced by CSC. Following this directive, we included a line on our study flyers that stated, ‘People currently on parole or under supervision by the Correctional Service of Canada are not able to participate in this study.’ In speaking to staff at the organizations that assisted with recruitment, comments we received indicated that this exclusion considerably limited the pool of potential participants ⁸ – another clear way in which access barriers can shape the nature of research.

In light of our recruitment challenges and that we remained unable to find official written documentation that CSC approval is needed to recruit people on parole, the principal investigator took steps to better understand CSC’s research policies. She also aimed to clarify whether a reapplication that included only people on parole as participants, and not current prisoners or prison staff, might be accepted since parolees were not explicitly identified in our original application. In email correspondence with CSC’s Research Branch, she was again advised that formal approval is needed to recruit anyone that falls under CSC’s jurisdiction, which includes people on parole regardless of where they are residing (i.e., CSC-operated facilities, residential facilities, or elsewhere in the community), and that a reapplication on the same topic that focuses solely on people on parole would not be supported.

The access barriers and recruitment challenges we confronted forced a change in our study inclusion criteria. We had initially aimed to recruit participants who had spent time in a federal prison(s) within the last four years, but expanded that criterion to six years, hoping to increase our sample size while still engaging people with recent incarceration experience. Despite the noted difficulties, we completed focus groups and interviews with a total of 30 former prisoners, as well as 10 expert stakeholders, between October 2014 and April 2015. The participants who had been formerly incarcerated comprised a racially and gender diverse group, and represented a range of experiences in terms of injection drug use and where they served their federal sentence(s). Our final sample enhanced our confidence that we engaged a highly knowledgeable group of former prisoners whose insights were invaluable for developing PNSP implementation recommendations. Nonetheless, our concerns about the potential impact of CSC’s access barriers on the scope and quality of our data, with reflection on our retrospective interview/focus group approach, warrant expansion. In the next section, we thus broaden the discussion about research with former prisoners as participants – its benefits and challenges, including ethical challenges – and continue to critically reflect on aspects of our study as learning exemplars for other researchers.

Learning opportunities and challenges with former prisoners as study participants

While some scholars disagree as to whether there is an ongoing paucity of research involving prisons and prisoners in general, many have noted a shortage of qualitative studies in particular (Crewe and Jewkes, 2011; Piché et al., 2014; Reiter, 2014; Wacquant, 2002). Meanwhile, although there is a fairly robust literature on prisoner re-entry experiences (Harding et al., 2014; Maruna, 2001; see also Pager, 2007), research involving former prisoners has been critiqued for often focusing on relatively narrow outcomes related to recidivism and neglecting some key personal realities of post-prison reintegration (Visher and Travis, 2003). In our efforts to learn from others who have conducted qualitative health-related research involving former prisoners, we found there to be a limited number of published studies (a few notable exceptions include Salem et al., 2013; Small et al., 2005; van Dooren et al., 2011). Ample opportunities exist, therefore, to ask new questions about prisoner health with the use of qualitative and collaborative methods. We hope that what we share below sheds further light on data quality considerations and assists other researchers in planning ethical engagement with former prisoner participants.

Concluding thoughts on enhancing research with former prisoners

While most prison-related research undertakings include various challenges and barriers, we found ours to be particularly illustrative of ways in which carceral institutions can shape and limit knowledge (Martel, 2004; Yeager, 2008). As we experienced firsthand when we were advised that we could not recruit people who were on parole without formal CSC permission, access barriers can unexpectedly extend beyond prison walls. Reiter (2014) notes that ‘methodological creativity’ (for example, collaborating and drawing on multiple disciplines and methods) can help researchers overcome access barriers. Our methodological creativity was both informed and practical as we worked closely with community partners to engage a diverse and knowledgeable participant group. Because we anticipated access-related issues, especially given our topic of study, we carefully modified our proposal for CSC and developed different questionnaires for current and former prisoners. Even after learning that approaching people on parole would require CSC permission, we were willing to further modify our research application for this federal agency. Perhaps in other instances similar steps would successfully lead to research access; in our case, it did not. We invite researchers to consider how our acquiescence in this regard may reinforce rather than challenge the monopoly and power that CSC or similar penal systems have over carceral research and knowledge development.

During the CSC application process, we received a response that reveals a strategy the agency uses to maintain control over research – that is, stating CSC already does their own monitoring and evaluation of infectious disease prevention – essentially dismissing our proposed study as potential duplication, and not on the basis of methodological or ethical concerns. It is surely possible CSC’s Research Branch perceived our study as unnecessary based on their own research agenda; it is also possible that they saw our proposal as too controversial or as having the potential to uncover deep-seated policy or operational problems that could threaten CSC’s reputation. We contribute to the literature another example of how access barriers can have multiple layers, including the possibility of organized efforts by correctional authorities to protect their status quo and to stifle or silence potentially critical or contentious research (see again Watson, 2015). In our case, while the access restrictions we experienced were introduced by the carceral system itself, others have documented various concerns regarding how university REBs may engage in similar knowledge control processes (Adler and Adler, 2002; Haggerty, 2004). More work is needed to unpack these concerns in general, while acknowledging the important role that university REBs play in ensuring ethical research.

Interestingly, and unlike the experience documented by Watson (2015: 342) who observed some ‘falling back to official policy’, the rule or policy regarding authorization to conduct research with people on parole could not be located in the guidelines and documents we received from CSC. While this access restriction introduced potential limitations in our study, we decided not to push back against the directive. Our primary concern was possible risks to participants, and we did not want to create a situation where someone on parole might experience retaliation, harm, or any negative consequence for participating in the research. Upon our deeper reflection, we question CSC’s authority to regulate access to potential research participants outside of prison walls, and have yet to find a clear policy or legislative basis for this form of gatekeeping. Perhaps we should have asked more persistent questions about the rationale behind the seemingly unwritten policy or had taken more time to challenge the directive. We strongly support future persistence and richly theoretical investigations into such questions because the lack of qualitative research on prisoner populations widens gaps in carceral knowledge, including much-needed and up-to-date information about the health and service needs of these populations and how prison systems may derive their authority to police research access.

While only able to recruit former prisoners, we found our study’s participants to be extremely knowledgeable and to have had diverse prison-related experiences, thus helping to mitigate the limitations imposed by CSC’s access restrictions. As scholars interested in expanding research-related roles and learning more directly from those who are often treated as discredited voices (see Piché et al., 2014), our research team included members who work directly with and advocate for current and former prisoners. In retrospect, however, it would have been advantageous to also have had former prisoners themselves involved in the research design, data collection, and analysis. As Newbold et al. (2014: 441) argue, ‘the passion engendered by the experience of incarceration can add color, context, and contour both to objective and subjective findings.’ While our research process comprised a community-based undertaking insofar as it was developed with and by representatives from local organizations, having former prisoners more meaningfully involved in all stages of the research could have been be very promising not only for the additional contributions gained from their insider perspectives, but also for supporting the skill development and personal capacity of marginalized individuals (for relevant discussions of community-based and participatory action research, please see Boyd and Boyd, 2014; Brydon-Miller et al., 2003; Cornwall and Jewkes, 1995; Dupont, 2008; Israel et al., 1998; Minkler and Wallerstein, 2008; van der Meulen, 2011). To make our study more inclusive, we should have allocated resources for a former prisoner to participate on the research team, given that they would be doing so as a volunteer whereas all other team members were there as part of their paid employment. Had we planned accordingly, we could have requested an honorarium as part of the initial grant application. We wish to note, however, that this kind of peer engagement in research comes with its own sets of benefits and challenges, an exploration of which is beyond the scope of this article and has been documented elsewhere (Alexander and Richman, 2008; Banks et al., 2013; Guta et al., 2013; Ross et al., 2010).

Overall, our experience demonstrates that it is vital to be reflexive when conducting studies that include former prisoners, and that such qualitative research is greatly needed, despite its challenges, given the important insights that this population can share – and may be more likely to share when researchers employ a collaborative approach designed to enhance participant trust. Research access barriers like the ones we encountered can greatly limit new empirical and theoretical opportunities. We hope that our methodological narrative gives readers new insights that they can use in future and innovative qualitative research in carceral contexts.