‘I’m not sick!…Are you?’ Groupthink in police services as a barrier to collecting mental health data

Despite the high prevalence of mental disorders among Canadian police officers, treatment-seeking is lower than expected. Toward understanding how mental health services can be tailored for higher utilization by police, we aim to uncover factors that contribute to stigma and barriers to use, specifically within the context of group dynamics between officers. Nine semi-structured focus groups and one interview were conducted with civilian and non-civilian police service employees in Ontario, Canada. Data were coded to allow for themes to emerge from the transcripts. Participant voices (n=33) revealed the presence of three characteristics of Janis’ groupthink: high group cohesion, conditions that create high stress and low self-esteem, and operating under directive leadership; each creating pressures that serve as barriers to treatment-seeking [Janis IL (1972) Victims of Groupthink: A Psychological Study of Foreign-Policy Decisions And Fiascoes. Boston: Houghton Mifflin]. Groupthink offers a potential explanation about why police, despite a high prevalence of mental disorders and access to mental health services, do not seek treatment as expected. Janis’s theory of groupthink is supported by police officer dialogue in focus groups. Understanding police group interactions can better inform prevention and treatment programs, ultimately leading to better access and use of existing mental health services, a reduction in stigma associated with treatment-seeking, and a healthier police workforce.


Introduction
The psychological strain inherent to the police profession and the effect of this strain on officers have come to the forefront of police research (Crank, 2004;Duxbury et al., 2015;Rabe-Hemp, 2009;Ricciardelli, 2018). Understanding and treating psychological strain is an objective of police researchers, care providers, stakeholders, and the federal government in Canada (Oliphant, 2016). In a recent prevalence study of mental disorders among Canadian public safety personnel, 36.7% of municipal and provincial officers and 50.2% of members of the Royal Canadian Mounted Police (RCMP; federal) screened positive for at least one mental disorder (Carleton et al., 2018a). 1 Carleton and colleagues (2018b) also noted that rates of lifetime suicidal ideation were 20.5% for provincial and municipal police officers and 25.7% for RCMP officers. Undoubtedly, too many police staff, including officers and civilian members, experience compromised mental health and an understanding of the sources driving this is valuable. Yet, to truly serve police there is a paramount need to understand the barriers to treatment-seeking behaviors (Reavley et al., 2018;Ricciardelli et al., 2018) and the scope of the stigma around treatment-seeking (Edwards and Kotera, 2020;Ricciardelli et al., 2018), and to include the voice and views of officers in shaping solutions. Only with such knowledge can systemic barriers (e.g., concerns about anonymity and career impacts) (Fox et al., 2012) that may prevent the creation, integration, and adoption of effective assistance and treatment programs be addressed.
We aim to explore how mental health services can be tailored for police, including civilian employees, increasing accessibility and reducing stigma. Via focus groups, capturing the unique needs and preferences of police staff when discussing mental health and treatment-seeking, we unpack the influence of group dynamics between officers and civilian employees, including those in different ranks and positions. Accordingly, we framed the article within the theory of groupthink (Janis, 1972), shedding light on how the variant yet existent institutional culture within police services shapes what constitutes appropriate behaviors and presentations of self among officers. We analyzed focus group dynamics and the expressions of officers and civilian employees (i.e., public safety communicators) within the social identity maintenance (SIM) model of groupthink (JC Turner et al., 1987). By understanding the nuanced police and civilian group dynamic, our intention is to provide insight that will help to design welcomed and acceptable mental health treatments within the group dynamics of police services, toward informing effective infrastructure for treatment-seeking and mental health stigma reduction.

Police culture
Conceptually, police culture developed from theories of occupational culture that characterize how occupations, through informal occupational norms, shape the incumbent's view of their occupation and perpetuate a lens through which the world is interpreted within the social group (J Chan, 1996;O'Neill et al., 2007). Police culture has become integral to understanding police practice and actions (Reiner, 2017). Researchers suggest that police culture explains positive qualities associated with strong systems of organization and hierarchy (Manning, 1977;Punch, 1983;Reuss-Ianni and Ianni, 1983), a sense of policing as a way of life or higher mission (Crank, 2004;Reiner, 2010;Skolnick, 1966), and a strong sense of internal solidarity, which collectively ensues powerful intragroup dynamics (Fielding, 1994;Reiner, 2010;Skolnick, 1966). Yet, police culture has also been used to explain negative qualities associated with police, such as a marked cynicism of the world (Reuss-Ianni and Ianni, 1983), suspicion (Crank, 2004;Manning, 1977;Reiner, 2010;Skolnick, 1966), an attitude of machoism and conservatism (Reiner, 1992)-all of which promote officer isolation and distance from other citizens (Fielding, 1994;Reiner, 2010;Skolnick, 1966).
Police culture is also informed by police operating within a paramilitary structure (Chappell and Lanza-Kaduce, 2010), domination by traditional hypermasculine ideals and a focus on an ideal of crime control (Fielding, 1994;Herbert, 2001). Consequently, in-group dynamics combined with an exceptionalist focus on crime control is credited with forging bonds among officers that enact an ethos of othering (Reiner, 2010).
Working under this paradigm, the different levels of police inform their own cultures and in-group dynamics in light of their shared goals and exposure to one another (Paoline, 2003). Lines are drawn within services based on unit, intragroup conflict, and experience, thus suggesting that police services are not cohesive groups, but instead are shaped by battling ideologies, factions, units, and ranks (JBL Chan, 1997;Herbert, 1998;Reuss-Ianni and Ianni, 1983). However, promotional structures with the hierarchical organizational design also serve to compromise and inform officer actions and impose an overarching culture; making the ability to breach trust among officers or violate culture norms at lower levels of the hierarchy a true possibility for those intending to seek promotion. In this sense, a greater organizational culture does exist and is reinforced by the overarching paramilitary organization of police services. Despite these intragroup conflicts and divisions, shared culture can create cohesive in-group dynamics rooted in shared experience and interpretations of the world (Crank, 2004;Skolnick 1966). Millar's (2002) work examining police help-seeking behavior illustrates that these dynamics are situational and changing. Beyond officers' perceptions of being perceived poorly because of helpseeking behaviors, Millar's (2002) participants described being supported by colleagues (rather than judged or stigmatized) for sharing their mental health and intervention-seeking experiences. The understanding that treatment-seeking is stigmatizing, despite Millar's findings, further supports that police culture contributes to the construction of barriers around mental health and treatment-seeking. Police culture, then, is dynamic, situational, and layered, as well as contested (Chan, 1997;Reiner, 2017). It shapes officers behaviors, including those around prevention, recognition, and mental health treatment-seeking.

Police and mental health
Police are exposed to potentially psychologically traumatic events (Crank, 2004;Territo and Vetter, 1981), experience strain from organizational norms (Rabe-Hemp, 2009), and encounter other occupational challenges (e.g., lack of supports from management, high job demands). Given officers' experiences with these stressors-some occupational (e.g., workload, job duties, civilian and victim encounters) and others not (e.g., family, friends, social stressors)-there is ample opportunity for their work to negatively impact their mental health (Adams and Buck, 2010). Stressors can lead to negative mental health outcomes including compromised psychological wellbeing, emotional distress, and burnout (Adams and Buck, 2010; Karaffa and Koch, 2016), as well as alcohol abuse (Charles et al., 2007), post-traumatic stress disorder (Martin et al., 2009), and depression (Violanti et al, 2008).
As previously noted, in a recent large-scale survey of Canadian public safety personnel, Carleton and his colleagues (2018a) found that officers screen postive for mental disorders at prevalences much higher than those seen among the general population, and screen for higher prevalances of suicide ideation and attempts (Carleton et al., 2018b). However, despite the high prevalence of mental health needs, workplace barriers continue to prevent treatment-seeking among police as well as other public safety personnel ; these barriers include stigma, abandoning colleagues to work in increasingly understaffed conditions, and occupational repercussions (e.g., inability to promote) (Edwards and Kotera, 2020;Ricciardelli et al., 2019;Rose and Unnithan, 2015).
In response, police have been found not to avail themselves of help or seek treatment even in cases where the organization strives to make mental health support normative (Reavley et al., 2018). In fact, when asked what types of mental health support they would access, police report higher endorsement for talking to their spouse or friends rather than mental health professionals (Carleton et al., 2020). The barriers then to treatment-seeking are both organizational (e.g., in some police services, officers have less access to supports) and perceptual (e.g., even when the police service has a positive relationship with mental health provision, aid-seeking behavior is low) (Reavley et al., 2018).
Together, the emphasis on individuality (necessary for promotion), silence (stigma preventing the healthy sharing of information), group cohesion (among officers), and distrust (toward civilians and between ranks) creates conditions that minimize the promotion of mental health and associated treatment-seeking behaviors (Bell and Eski, 2016;Reavley et al., 2018). To this end, Fox and colleagues (2012) suggested that for officers to seek treatment, the assurance of confidentiality is required, perhaps best actualized by locating treatment outside the service to offer some degree of privacy for those who do seek treatment. Although officers in Canada have access to the federally supported and subsidized Canada Employment Assistance Programs (EAPs), peer support, and critical incident stress management (CISM), the Road to Mental Readiness (R2MR) program and the Working Mind (adapted for police by the Mental Health Commission of Canada), alongside diverse private coverage, officers still report not accessing services and experiencing significant mental health challenges (Fox et al., 2012). Overall, despite the intersection of the policing occupation and mental health, barriers to care, particularly those emerging from police culture or those reinforced by the organization hierarchy, require further investigation.

Groupthink
The social group-however it originates (e.g., at work, at home, in sport)-is a way to explain how people transition from individuals to unified groups with shared interests, attitudes, and behaviors (see JC Turner et al., 1987 for a comprehensive review and reconceptualization of the social group). Thus, collective thinking is common among social groups and even expected among organizations like the armed forces and paramilitary services, including the police (Chappell and Lanza-Kaduce, 2010). Janis (1972) developed groupthink to explain instances where social groups made faulty decisions leading to very public blunders (e.g., Kennedy's Bay of Pigs invasion in 1961). He considered cases in which the likelihood of a negative outcome seemed obvious to those outside decision-making groups, and yet those inside the groups, individuals deemed competent in their positions, made errors in decisionmaking that resulted in seemingly avoidable negative outcomes. Janis proposed that these instances could occur when requisite conditions were met, specifically: high group cohesion, insulation from outside experts, limited search and appraisal of information, and operation under directive leadership; conditions that create high stress with low self-esteem, and the belief that there is no better solution to a problem than the one provided by the group leader. The requisite conditions, Janis proposed, create a solidified in-group, resulting in groupthink, which is characterized by delusions of invulnerability, alienation of and rejection of the criticisms of out-groups, and faith in the believed inherent moralities of in-group decisions.
Groupthink has been used in a variety of fields such as communications (Flowers 1977; for a comprehensive review of the historical use of groupthink see Turner and Pratkanis, 1998b). The generalizable capability evidences the truly interdisciplinary application of groupthink as a theory and concept, although relatively few studies have attempted to empirically apply or measure its impact (Turner and Pratkanis, 1998b). Nonetheless, groupthink has appeared contentious in applied research too, resulting in diverging views of how to apply Janis' theory (ME Turner et al., 1992).
To address these contentions, ME Turner et al. (1992) first added the SIM model to Janis', providing a space for his model to become more grounded as an applicable social psychological principle. The SIM model reframes instances of groupthink by directing the focus of the group to maintaining a positive and cohesive group image, applied prior to the occurrence of a negative outcome. Groupthink, then, comes into effect when the group feels their image is directly threatened. In these situations, the group shifts priorities, creating a situation that can result in the faulty decision-making processes characterized by Janis.
We propose that the SIM model of groupthink is a tool for the social psychological analysis of groups, including for police who exhibit a strong collective culture as characterized by Janis' conditions. As such, the model can be used to explain seemingly unpredictable group behavior such as why officers do not seek mental health treatment despite the availability of such treatments or evidenced mental health needs. Therefore, understanding police culture using groupthink within the SIM model provides a less critical explanation of group dynamics that focuses attention on the social relations developed from working in close-knit groups. Despite the applicability and potential insight that the implications of groupthink in police contexts can provide, such processes have seldom been explored. Only two studies to date have included police as the population of interest. In the first, Rosander and colleagues (1998) tested the validity of a questionnaire measuring groupthink tendencies they labelled as omnipotent (characterized by bias and narcissism) and depressive (characterized by failure and helplessness) in several different authentic work organizations, including two police agencies. They reported that the police participants responded in a more ambiguous manner-that is, neither omnipotent nor depressive (Rosander et al., 1998). De Vries et al. (2004) used a more applied methodology to study groupthink among police detectives in South African police services. Using a simulated scenario called an 'escalation dilemma', the researchers asked groups to make a decision about whether to sustain commitment to a project or terminate investment following a series of choices that intensify in complexity. De Vries et al. (2004) found that officers were inclined to fall into groupthink behaviors during times of high stress, and they had very high ingroup cohesion and very poor perspectives of those that they perceived to belong to the out-group, which suggests that groupthink can explain unexpected behaviors (like a reluctance to seek mental health support) in organizations like the police that exhibit strong cultural norms.

Current study
Despite high prevalance of mental disorders among police service members (Carleton et al., 2018a) and readily available mental health treatment and services for officers (e.g., EAPs), usage of such services remain low. Factors such as shiftwork and limited availability of providers (see Fox et al., 2012;Reavley et al., 2018) have been tied to these to low rates of help-seeking behaviors. Other factors such as a strong culture of a machoism (Chan, 1996), distrust of police leadership, and exceptionalism are documented barriers to mental healthcare service usage (e.g., Fox et al., 2012); all factors that can be collectively characterized as rooted in groupthink (see Janis, 1972). In the current study, we seek to demonstrate the relevance of the SIM model of groupthink among police service members. We include both civilian members, specifically persons working in communications, and police officers, recognizing that both experience potentially psychologically traumatic events (e.g., communicators dispatch police to calls for service and provide a 'front line' for care by phone) and civilian members are seldom considered in scholarship. We explore the intersection of groupthink and mental health, integrated approaches, and in particular the SIM model of groupthink among sworn and civilian members of a Canadian municipal police service to provide insight into the infrastructure shaping participant treatment-seeking. We look at how group processes may encourage the stigmatization of mental disorders and thus the perception that voicing support for treatment-seeking for mental wellbeing poses a threat to positive group image and thus, such actions become barred by groupthink. Groupthink, then, if unaccounted, may affect both research (e.g., compromising data quality and usability if the group context is ignored while collecting data) and treatment outcomes with police.

Methods
We conducted semi-structured focus groups with sworn (i.e., officers) and civilian (i.e., communicators) employees of a police service. Focus groups were facilitated by a qualitative researcher (RR), and when possible, she was joined by a clinician (KM) to provide both clinical and qualitative interpretation to the knowledge produced in the focus group context. Focus groups were employed because our interest remains in the knowledge produced in dynamic group interactions, specifically, knowledge agreed upon by participants across rank and roles within the group (Stewart et al., 2007). We draw on "personal meanings" (Livingstone, 2003) mined from collectively accepted interpretations, as put forth and experienced by participants as they contemplate mental health needs and treatment-seeking. Researchers examining sensitive topics or populations have used focus groups to elicit "a level of frankness that is seldom achieved through survey questionnaires" (Madriz, 1997: 3). Yet, focus groups, to our knowledge, have not been used in Canada to understand police service staff's understandings of mental health and treatment-seeking behaviors.
Focus group participants were members of a municipal police service in Ontario, Canada, and included civilian communicators (n¼8; two of whom identify as special constables) and sworn police officers (n¼25) across ranks (e.g., from constables to sergeants). One of our participants self-identified as Indigenous, whereas the other 32 selfidentified as white. Participants had all completed at least high school, 16 had a college diploma, and the highest level of education reported was a postgraduate degree. Most participants (n¼20) were married, and the vast majority (n¼27) had children. Years of experienced ranged from 1 day (i.e., first shift) to 32 years; the average being 16 years of experience.
Recruitment was coordinated by the health and wellness manager of the service through an email, circulated through the organizational email listserv, informing potential participants about the study, and inviting voluntary participation. Focus groups were conducted onsite during work hours and each group consisted of either only sworn staff or sworn and civilian staff. Prior to the focus groups, participants provided informed consent and then completed a demographic self-report survey. Focus groups sizes ranged from three to six participants, and one interview was also conducted given some complications that would have otherwise prevented the officer's participation. Each ranged in length, but averaged approximately 90 minutes in duration and always followed conversational paths as put forth by participants. Focus groups were voice-recorded and participants were entered into a draw where one participant would win a gift certificate for a local coffee shop. Ethics approval was granted for the current study.
A semi-structured guide was used to lead focus groups if conversation felt strained but, more commonly, it served as a checklist of topics that required coverage (see Appendix 1). The focus group structure was intended to ensure participants could propose and discuss topics that they deemed relevant and as they felt comfortable. Given the focus on treatment-seeking practices, questions of a personal nature were never posed and participants only shared personal experience by their own volition.
Focus groups were transcribed verbatim, and participant names and identifying information were replaced with pseudonyms (or omitted when necessary). The transcribed focus groups formed the corpus of data used for our analyses. Transcripts were imported into NVivo Pro 11 (2016) and coded. A codebook was first constructed a priori based on the interview guide, then developed using an approach consistent with a semi-constructed grounded theory (Ricciardelli et al., 2010), which provided opportunities for emergent theme analysis with considerations of theoretical understanding (Charmaz, 2014;Glaser and Strauss, 1967). The approach was semi-grounded in that we did not suspend theoretical knowledge during the data analysis process, and although themes emerged from the data, theory did not, as would be the case in a fully grounded theory approach as per Glaser and Strauss (1967). The data were first coded for concepts and themes regarding treatment-seeking and mental health across experiences and interpretations. An emergent theme was considered when multiple participants shared perspectives or when patterns arose in how participants engaged in discussions across focus groups. As the concept of collective thought processes continued to emerge from the data, the data were recoded for theoretical constructs that reflected Janis' (1972) groupthink theory, explicitly unpacking the possibility of groupthink as an influencer in police service staff mental health treatmentseeking.

Results
We conducted nine semi-structured focus groups in 2018 with 33 participants. From these focus groups, we examined how participants respond to questions posed by the facilitators(s) and how they interacted with one another when formulating their responses. Specifically, the data support that Janis ' (1972) groupthink concept is at work among police service staff. We use participant voices to present the three key characteristics of groupthink that emerged in the focus groups: (a) high group cohesion; (b) conditions that create high stress and low self-esteem; and (c) operating under directive leadership.

High group cohesion
In-group and out-group dynamics are integral to addressing group cohesion, a key prerequisite for groupthink. Within hierarchical or paramilitary organizational structures, like policing, strong in-groups emerge and function because members legitimately like, respect, or/and believe in one another (see Janis, 1972). The shared experiences among those working in occupational roles embedded in the need for confidentiality, the inability to share experiences with non-coworkers, and that encourage a distrust toward other members of society (De Vries et al., 2004;Fox et al., 2012) create a dynamic that informs the positioning of the 'other'. The symbolic 'other' is one who is differently minded, who creates strain for the in-group and of whom police may direct cynicism. The result of such a proscriptive social lens is the creation of out-groups and an associated symbolic labeling of the out-group as the 'other' (Fodor and Smith, 1982;ME Turner et al., 1992).
Yet, to have an out-group an in-group must exist. The police services staff and officers in our study confirm the existence of strong in-group dynamics. In our sample we have two in-groups divided by occupation: one composed of the civilian staff (i.e., communicators), and the second composed of sworn staff (i.e., officers). The police in-group and communication in-group share experiences; however, the experiences are approached from different standpoints and with diverse occupational expectations and roles. The communicators have a unique role in all police service provision; they are the officers' lifeline during deployment for calls for service. 2 However, the role of the communicators diminishes as the role of the officers get increasingly underway, thus their positions overlap and are interconnected but diverse. In fact, interpersonal engagement can be actively discouraged by the organization, such as instances in which officers are not permitted to enter the communications' space and staff are forced to use occupation-specific physical education training gyms. To illustrate the boundaries between the in-groups, we reflect here on discussions around a new gym at the police service station that was designed for only communicators to access in light of their need: K: You're isolating, you're segregating one group from the other group and we're all supposed to be a team. I mean when you start making those little rules and stuff like that, you're increasing the stress of the officers and the communicators and it's creating more stress and mental health issues like-A: Creates a lot of tension in the workplace, right? That doesn't need to be there (Karen and Andrew, Focus Group 9) The participants', two police officers, words reveal that some participants disapproved of the exclusion from using the gym directed toward specific members of the police service. However, those that reside within the police in-group do not appear to recognize the validity of the discomfort expressed by the communication in-group regarding working out; in response they appear to blame to some extent the communication in-group for their resulting differential status. The accessibility of the gym represents a central way in which boundaries between in-groups are defined and negotiated. However, in both in-groups members are able to confide in one another, share experiences (including those encouraging distrust), and an occupational lensincluding if that lens shapes their self-presentation. 3 Across focus group discussions, a pressure to maintain group cohesion was observable. A count of agreement or disagreement tied to each topic within each focus groups reveals that when participants engaged in dialog and contributed their ideas with other participants, the groups came to a quick consensus. Agreement was noted on 151 topics and themes, while disagreement (or confrontation) ensued for only eight topics and themes. This is particularly interesting when recognizing that focus groups included both communicators and police officers with diverse ranks and years of experience. although quantifiably obvious, it was not only the frequency of agreement, but rather its nature across the groups that theoretically supported the groupthink idea of high group cohesion. Interactions between participants always followed the same pattern: a participant voiced an idea and if support for the idea was voiced by another (particularly by a higher-ranking officer), this idea became the norm, yet if there was not support, the idea was either immediately abandoned or, in a select few cases, countered by another officer. Thus, generally intergroup dynamics created space where the acceptable idea could take priority rather quickly.
To illustrate a case of disagreement, after Derek explained that his friend's experience of a treatmentseeking program was rather futile, Edward in his focus group firmly retorted, "I had a complete opposite experience [ . . . ] The issue of it, however, for me, how it worked for me, is that, based on my work schedule, finding time to go, after hours . . . ". Derek, rather than agreeing with Edward, explained that Edward was fortunate: "You're lucky you picked the right one." He continued by noting his friend's experience, "the guy we just talked about, who had the suicide attempt, he tried to get into a program and it was takin forever, and he said 'I've got to go for something, I've got to get some mental health help'". Finally, he noted that his friend was desperate and "tired of waiting" and that he found a provider outside those covered by his employer programs and "paid for it [treatment] out of his own pocket, seven grand. Submit the receipts it was like, 'yeah, these therapists can't be paid for'." Derek's response was affirmative, and the direction in their focus group (Focus Group 1) changed from treatment being helpful at their police service to treatment being insufficient and few providers being covered by insurance or EAP.
Thus, group cohesion was valued over personal experience; Edward outright disagreed with Derek, but when he was presented with the group's perspective, he fell back into place within the group dynamic and ceased the disagreement. Moreover, in the small number of disagreements captured in the focus groups, there were no escalations in moods. Instead, in-group processes allowed for the avoidance of arguments, by having those who established leadership in the focus group, to reaffirm the group perspective on the issue. In our sample, the in-group interjection was never an outright contradiction, but rather a correction back in line with the group, as we saw Derek do in his focus group. In most cases of disagreement, disagreement ceased and the in-group perspective was reassumed. In a small number of cases (n¼9), after a number of failed attempts to realign the disagreement, they appeared silenced, neither agreeing nor disagreeing, but maintaining cohesion with the group perspective.

Conditions that create high stress and low self-esteem
In-group membership is first created and then reinforced through situations of high stress (e.g., responding to calls for service) and, potentially, low self-esteem (e.g., when self-presentation fails to reflect interpretations of experience). The high stress can be understood as responding to or listening to calls for help from those in distress and being in a position to be accountable for such instances, as James (Focus Group 4) describes: And quite often it feels like the public and our management are holding us to an impossible standard, because we're human beings, we make mistakes, we have a reaction, an emotional reaction, to a stressful situation, and when you don't feel you have that kind of support internally, then obviously that weighs on the mental health of the frontline officers.
James' words show the high stress he experiences in performing his occupational responsibilities.
High stress, according to our participants, also stemmed from the intensive nature of working within the service and from demanding schedules that isolated staff from other professions. Perhaps, at least in part, as a result, our participants noted that many people with whom they had relationships with were also part of the police service. For instance, Edward, (Focus Group 1) stated that: "Yeah, I think it really comes down to, um . . . peer support. You're in uniform, you're riding with the same guy for several months, you start talking about things, we all have to have a partner". The excerpt here shows first the closeness among some colleagues, but also the concept of exceptionalism in practice. The in-group dynamic, in this case a close bond between 'friend/colleagues' who understood each other, enacted to further isolate police service staff and members who contextualized their profession as distinct from others and, therefore, consider all others who lack the experience and understanding derived from the occupational participants to be in the out-group (see Reiner, 2010). Although interpretation of such in-group dynamics can and has been attributed to rather negative qualities within police culture (machoism, "thin blue line", etc.), such interpretations are not generalizable, overly accurate, or representative. In addition, they downplay the fact that, as evidenced among our participants, such strong in-group understandings create a more accessible space to find support. Officer participants described feeling they could relate to co-workers more than with others: I find that the best support I have is the relationships I have with my peers. If you're in a unit where you're close-knit or you're like-minded and people think along those lines of "I'm here for you, you're here for me, we're here for each other". I find that's the best that I've had going through the years I've been here. And then, conversely, if you're in a situation where you maybe don't feel supported by your supervisor or something like that, or there's friction there then it's got the adverse effect (Darren, Focus Group 2) Darren's words reveal the powerful pull, actualized by comradery and like-mindedness, of the in-group, and how the out-group is established as being individuals who lack this like-mindedness and are thought unable to understand their in-group's point of view. As such, we see how the ingroup is defined around unit and power.
In addition, across all focus groups, participants noted that peer support programs tied to well-being exist to offer a venue for mental health support or even as way to debrief among peers within the police service. Rachel (Focus Group 3) explained, So, we have a pillar committee, that I'm actually part of; we meet once a year and it's more voluntary, for example, if I know Rhonda and we've worked together in communications and she's having issues, if I know about it I can reach out to her, or she knows I'm on the pillar, she can reach out to me. While Patrick, (Focus Group 8) spoke of his experience with peer support, noting that, [ . . . ] the support system that we get from our peers is paramount and almost immediate. And if I see someone that I think might be suffering that might not be close to me, I can reach out to John and say so-and-so might have a problem. Or he saw something horrific, or she saw something horrific and reach out to them that way.
As evidenced across these officers, speaking to peers was valued as a source of insight, support, and even understanding. Said another way, groupthink insulates the opportunity for peer support as a mental health strategy. Participants explained that within the nature of their job, peers could provide social support because they understood the complexities of their work differently from what could or is offered by out-group members.
Yet just because peer support was valued and available did not immediately translate into its effectiveness or use. Instead, barriers, such as the work required to establish a functional partnership between co-workers and those of confidentiality, remain. Not all participants spoke appreciatively about the program; a select few participants in the focus groups described feeling the pillar program was not effective. For example, James (Focus Group 4) said: Yeah, and again, it's just . . . but there's a big enough list that I probably would have found someone that I probably would have felt comfortable talking to. I was surprised when I looked at the list and I went "Uh, no, no, no, no, no". Just I don't have a comfort level, I'm not passing judgment on them, just that I would not go to them.
There are two key findings here: within the focus group there always tended to be a central view of the program, and that the groups collectively either appreciated the program or felt it was ineffective (but never were groups split). A perspective of the pillar program, either as somewhat positive or negative, appeared to dominate conversational paths. Those promoting the more negative positioning, evidence in Focus Groups 4, 5, and 8, possibly create the tensions around the functionality of the program because they do not feel they are part of the in-group, and as such, are less committed to the collective sharing and caring. In this way, groupthink may also exacerbate challenges tied to help-seeking mental health behaviors. The conditions that create low self-esteem among the ingroup include experiences of embarrassment or feeling inadequate. For example, our participants noted that situations in which one needs to make a decision or engage in an action that has an adverse effect on colleagues or investigative outcomes can have a significant effect on one's self esteem. Derek (Focus Group 1) elected to describe one such experience: He was going to do it, we got the text and we sent guys over to the house and no one saw it coming, but even when we got the text, when the guy's in the state . . . I'm like "Holy fuck", I know what I have to do, but I'm like "Oh my God what does this mean to him?" If I do this I have to call the duty inspector and this guy's career has taken a huge right turn, and the time it's going to take to recover, and even knowing that this guy was on the verge of doing it, I still was like, "God, do I gotta make this call?" Because you worry . . .
Derek illustrates how the consequences of someone facing a mental health challenge become the concern of not only the individual, but anyone who is involved. When an officer was contemplating suicide, his only choice was to take action, and yet the knowledge that he would be affecting both the individual's and his own career weigh heavy, showing that in-group dynamics can also result in exacerbated stress. Moreover, there was significant stress associated with being investigated by or tied to the professional standards department, which functions to investigate internal complaints and issues within the service. Being the subject of an investigation leaves an officer stressed as they face judgment from other service providers and management as well as their peers. Officers who were part of the professional standards unit, also felt their occupational responsibility to investigate their fellow colleagues was a source of occupational stain and stress, often transitioning their positioning in the organization out of the in-group. Not only do such practices evidence the defensive nature and apprehension that exists between service members of similar rank, but it also creates stress for all persons involved-the investigators and the investigated as well as peers more generally. Dylan (Focus Group 7) who was part of the professional standard department described his experiences: I had training this morning, and it's always interesting because people talk to you, but they don't really wanna talk to you, and so there was seven in training and we had to pair up. So of course, 2, 2, 2, and then so I paired up with the instructor because nobody wants to pair up with [professional standards].
The officers' words reveal a feeling of transitioning to a variant of an internal out-group, and despite their prior relations and positions within the in-group they were now viewed as others. Also demonstrated here is how the ingroup bonds by their oppositional view of the professional standards department; they connect through their concerns about the potential of, and justness of, an investigation. In effect, in-group bonds strengthen when perceived threat emerges. Said another way, a degree of group cohesion is created that allows for the service to function as an ideological unit. The outcomes found resonate with the findings of prior scholars; officers consider management [here the investigative department] to be more of a stressor than a resource and detached from understanding the nuances of the police role (Paoline, 2003;Reuss-Ianni and Ianni, 1983). In this way, the professional standards department members and management exert pressure, which creates stress for the in-group. Overall, an apprehension toward specific departments or management reemphasizes the ingroup dynamics for service members. The outcome is the maintenance of a positive group image and the promotion of non-interference within the group.
Individuals may feel that having or disclosing their mental health needs counters what they have learned to be the appropriate self-presentation for those in their occupational role. Among our participants, many expressed concerns about how others would view them if perceived as having a mental health concern, for example: It's reputation it's the rumor mill circuits, if you take a few blocks off cause you need to take a breather and then you come back to work and then all of a sudden people are like "hey man, like are you alright? And I heard this, I heard that". And it's not factual, it's whatever . . . It's not really something people want to talk about cause it's personal, but now you almost feel like you're obligated to kind of save your reputation; "no it was nothing to do with work" or "I'm good to handle calls, it was a private thing or maybe it's a health issue or whatever". There's such a massive stigma, maybe because we're paramilitary? But . . . . (Kyle, Focus group 7) In this case, even the idea that Kyle might take a day off to process something traumatic or just to rest is met with a need for defense. The perception that service members may, possibly, have a mental disorder poses damage to their reputation and, as a result, discourages care-seeking behavior. Additionally, some described fear of being unable to advance or promote if they came forward about possible mental illness, like Thomas (focus group 1) explained: " . . . if I'm personally thinking that I'm struggling, I'm a year and a half in, but, and I want to say something that I, and Caitlyn interjected in agreement, "How am I ever going to get promoted?". Thomas continued to confirm this challenge: "Exactly, so I've got to think about how long my thirty years is going to be". In Focus Group 2, participants reinforced these views but added a twist, when the qualitative researcher asked, "What consequences are there [to coming forward with a mental illness]?" "Allan replied" "Just people knowing your business, or maybe you won't get a certain position, things like that. Maybe people won't want to work with you".
Collectively, participants' words suggest that there remains a common belief that an admission of mental health needs would affect their job or highlight the pressing fear that their peers would think poorly of them if they learned they sought treatment. Consistent with the SIM model of groupthink (ME Turner and Pratkanis, 1998a), such fears are rooted in and exacerbated by the in-group dynamics. As Allan's (Focus Group 2) words show, the perception of differential treatment is real, leading to the erection of barriers that encourage sameness and an incentive not to seek treatment. Participants positioned themselves against speaking about mental health or treatment-seeking behaviors because of the perceived characteristics of the in-group, specifically, ideas within police service cultures (J Chan, 1996) that discourage open discourse about vulnerabilities. In consequence, expressing vulnerabilities tied to mental well-being created a fear of being removed from the in-group and even possibly pushed to the out-group (despite the membership in the occupational service), which our participants also noted would result in a loss of status and potential social ostracization. Given the reliance on colleagues for support and personal relationship, the idea of being shunned by peers was stressful and served as a significant barrier to communicating about mental health and accessing treatment.

Operating under directive leadership
During our focus groups, those with greater seniority typically spoke first, tended to have their comments supported, and listened to others. Officers with experience of more than 5 years dominated the focus groups. For example, the average number of comments made by those with 5 or more years of experience was 52, whereas an employee employed for only a year put forward only 15 comments, consistent across other focus groups with similarly positioned participants.
Conversely, officers with less experience tended to be less vocal during focus groups, particularly when officers of higher rank or with more experience were present. They also tended to qualify their opinions based on years in their positions. For instance, Nick (Focus Group 3) speaking about sick leave, explained that, "I haven't felt that pressure [to take time off] but, again, I'm pretty new, and I guess it hasn't affected me . . . I haven't felt any pressure myself to not take time off". Nick, in this case, was proceeded by participants who felt as though they had been pressured not to take time off, such as Rachel (Focus Group 3): Again the culture, right? And back then, I don't even know if we got paid out for it back then, I can't remember, I guess we did. Just the conversations that would be had. I'll just talk comm [communications] again, if "so-and-so called in sick again" when it could be a mental health day because they had that bad call yesterday.
This chain of events helps to show that, even when they might have different experiences, individuals may defer to those they see as being more experienced by using qualifiers to soften their claims. Like Nick, others justified their positions or deferred to others in the focus groups, by stating, "Well I've only been here for a year, so I can't really speak to it" (Allan, Focus Group 2). In doing so, more senior or higher-ranking members were able to take a leadership role, direct the conversational path, and, thus, promote group cohesion. Such practices evidence the culture of obedience fostered by the paramilitary structure of police services (Chappell and Lanza-Kaduce, 2010).
Directive leadership can serve as a strengthening agent of in-group relations. Within groupthink, it reveals why the opinions, responses and decisions of some individuals (i.e., leadership) are favored over others and even adhered to despite other opinions possibly feeling more appropriate. Directive leadership can allow for the silencing of dissent within the in-group by reminding members of consequences (e.g., being investigated by the professional standards department) to promote conformance with the in-group. The organizational structure and associated leadership promote compliance, cohesion, and bonding in relationships: through the bonds developing among those of equal status or rank (e.g., who share their experiences on duty), and those across the larger organizational infrastructure (e.g., who impose compliance), rather than just social ties. This is especially relevant in police services where the occupational forces are the uniting factor, rather than comradery or friendship, as is the case in self-assembled political organizations as originally analyzed by Janis (1972).

Discussion
Our objective was to help identify some of the barriers that inhibit police from accessing mental health services by introducing the concept of groupthink (Janis, 1972) to explain why police and civilian members may not access these services despite being seemingly well-positioned and supported to do so. Although groupthink has been documented as a difficult social concept to practically apply, the responses we observed by focus group participants were congruent with the groupthink literature. An apparent and rather definitive agreement emerged among almost all participants within each focus group and this phenomenon prompted further investigation, leading to a theoretical interaction with groupthink as a force of socialization and cohesion. As such, analyses of transcripts focused on group agreement only when participants led the discussion there. In framing our findings within the SIM model of groupthink, we propose that the hierarchical norms and social culture of the police service we worked with contributed to this observation.
Researchers who aim to collect data to inform and improve mental health infrastructure and services for law enforcement members, necessarily need honesty, transparency, and individual thought from study participants to ensure that the improvements are suitably effective. Our findings, explained through the theory of groupthink, suggest that group interactions, and the systems from which they emanate, obscure the perceptions and expressions of many participants in favor of the opinions of those who are dominant or in positions of authority. As such, we see these group processes as limiting the capability of certain demographics (i.e., new and less experienced staff) to share freely their mental health needs and preferences. Similarly, higher ranking or senior staff may have their own views about mental health and treatment-seeking, but may feel forced to toe the company line by keeping up the favored opinion and maintaining the status quo. Both of these findings could reduce the accuracy and quantity of information shared and therefore seriously impair efforts to design mental health services to adequately support and enhance wellness.
We propose that high group cohesion, labelling of the out-group, and systems of hierarchy inherent in law enforcement all serve to enhance the risk for groupthink. We know that people did not feel they have the freedom to identify and share their own ideas about what health services should look like for the organization, thus leaving management and clinicians forced to make those decisions without true co-design from the majority of staff. The sensitivity of the topic was a frequent theme in focus groups, thus, due to fear of ostracization, cultures of suspicion, distrust of management, and how participants contextualize the path of "moving up the ranks", participants may be less likely to seek out mental health aid within the organization, and this barrier may remain a problem in attempts to provide additional services.
Our findings of groupthink, we hope, will prompt further examination of the social phenomenon and necessitate acknowledgment by researchers and law enforcement agencies as they navigate development of mental health services for police service staff. In general, researchers and agencies should look to the groupthink literature and implement evidence-based strategies such as employing a devil's advocate (Chen and Lawson, 1996;MacDougall and Baum, 1997) or using accountability (Kroon et al., 1991) to reduce groupthink when acquiring information about mental health needs and preferences. Similarly, members could be separated by rank, or organized into diverse groups from different areas of the agency (Greitemeyer et al., 2009) and even questioned individually or anonymously to provide additional insight into their mental health treatment-seeking needs. Alternatively, individuals in positions of authority should refrain from offering their opinion and instead encourage debate and different styles of thinking (Nemeth et al., 2001). It may even be helpful to bring in a mental health expert to contribute to group discussions, normalize mental health, and offer guidance (Smith, 2004).

Limitations
Our data are limited by the sample size. Given the possible identification of participants if we analyzed variations in groupthink by rank, we would be unable to provide a detailed analysis of how the hierarchical structure of policing may direct responses by participants (e.g., officers may not want to go against their supervisor). Potentially, this constitutes an area of caution for focus group analysis, thus we encourage future researchers to look at how and if rank suppresses free thought and willingness to be open and honest among lower-ranking officers in focus groups. Moreover, given that much of the previous literature looks only at police usage and not necessarily the role that civilians play in accessing care, we included civilian members in the current study. The inclusion of civilians could possibly skew the results away from the perceptions of frontline personnel to include the narratives of someone from other fields. As such, future work that includes separating civilian and police members in focus groups is warranted despite the fact that both groups in our sample evidenced susceptibility and conformity to groupthink. In addition, as with all qualitative data, we caution against the generalizability of the findings.
Finally, groupthink is characterized, in its final stage, as a faulty decision facilitated by requisite conditions. In this case, no faulty decision-making has been specifically measured. It could be argued that the possibility of false reporting on mental health services is a possible example of a faulty decision, however, this would not result in the sort of fiasco that Janis (1972) said characterized groupthink. Further, it is not possible to judge possible false reporting against the results we have collected, and so, while we can evaluate a propensity for groupthink, actual observation of the phenomenon, while suspected, cannot necessarily be confirmed. Instead, this serves as a theoretical expansion of possible barriers to care for police staff, such that this can be considered when developing research projects, as well as when providing or building mental health intervention systems.

Conclusion
Applications of groupthink to police are long overdue, as understandings of groupthink may allow for the navigation of nuanced behaviors of service members that, otherwise, may not be considered in research design or execution. The makeup of police services allows for a significant susceptibility to groupthink, and police culture encourages the development of groupthink systems that generate powerful in-group dynamics and systems of hierarchy that, while not innately harmful, present significant barriers to outside understanding. We ultimately hope that understandings of groupthink within police services will help researchers and law enforcement agencies better understand the intricacies of collected data, ultimately helping inform better prevention and treatment programs, and better utilization of existing supports.

Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article. This project was funded by the Ontario Ministry of Labour, Occupational Health, Safety and Prevention Innovation Program.

Rosemary Ricciardelli
https://orcid.org/0000-0002-0905-8968 Notes 1. Carleton et al. (2018a;2018b) included civilian members of municipal police services and the RCMP in their studies. 2. Given the similarities across in-group and out-group processes, we do not differentiate between the two occupational in-groups unless describable differences exist. In the focus groups, we did find that across in-groups a clear central voice always emerged and consensus reigned. 3. Comments here are meant to illustrate how in-groups and out-groups create divisions, not to make comments on generalizable distrust or disdain between communicators and other service members.