Outsider Inspections of Closed Institutions: An Insider Ethnographic View of Institutional Display

This article questions the value of internal inspections of closed institutions by external agencies, drawing on my unanticipated experience of being deeply immersed as a researcher inside a Secure Children’s Home at the time of an inspection. I describe how an ethnographic approach enabled me to see a dramatic change in the staff–young people relations – from adversarial to cooperative – in the presence of outside inspectors. I make sense of this change through an original application, and novel extension, of Goffman’s theorising. I conceptualise the staff and young people as insiders of a ‘total institution’ working together to perform a misleadingly harmonious ‘institutional display’, motivated by a shared sense of institutional identity. I argue that although the potential for insider misrepresentation can be acknowledged, the extent of it cannot be known by outsiders. This finding is of significance for social policy as closed institutions accommodate vulnerable populations and cases of institutional abuses attest the need for external monitoring. This article calls for recognition of the inherent limitation of external face-to-face inspection processes, and research into new methods of assessment.


Introduction
Secure Children's Homes (SCHs) are locked institutions in England and Wales that accommodate children who have been deprived of their liberty. In an attempt to ensure standards of quality and care, SCHs are routinely inspected by the Office for Standards in Education, Children's Services and Skills (OFSTED). Remarkably, SCHs have evaded sustained academic attention, and little is known about what happens inside this type of institution. The limited research on OFSTED inspections has focused on methodological measures of reliability and validity, which I argue overlooks a fundamental issue with the inspection process itself. This article makes a significant original contribution to knowledge in both substantive areas by exploring what happens to social interaction between staff and young people in the presence of OFSTED inspectors inside a SCH in England.
Empirically, I show that for the duration of the inspection, the staff and young people came together in an 'institutional display' (Goffman, 1959) of cooperation and contentment, which is a misleading representation of the nature of the day-to-day adversarial relationship between these two groups of social actors and masks their shared dissatisfaction with the institution. My rare in-depth ethnographic account as to how the staff and young people changed in the presence of outsiders makes an important and original contribution to the limited existing literature surrounding this type of institution and this type of inspection.
Conceptually, through a unique application and novel extension of Goffman's (1959Goffman's ( , 1961 theorising on microsocial interaction and closed institutions, using Campbell's (1958) concept of 'entitativity', I argue that OFSTED inspectors see collaborative artificial performances by insiders rather than objective representations of institutional functioning. In critiquing the value of the OFSTED inspection practice in a SCH, I question the value of observations of closed institutions by external agencies more generally.
My findings are significant because closed institutions accommodate the most vulnerable groups in society. Historical cases of institutional abuses of children in care, such as those detailed in The Waterhouse Inquiry (Waterhouse et al., 1999), and contemporary cases of abuses in residential care homes for older people (Matthews-King, 2019) and custodial institutions for young offenders (Allison and Hattenstone, 2019;Jay et al., 2019) are among countless other scandals which demonstrate the significance and urgency of finding better ways to enable outsiders to see inside closed institutions.
I begin by outlining the nature of SCHs before providing an overview as to the function and objectives of OFSTED inspections. I then offer a critical review of existing research on this type of inspection. I briefly describe how the data presented in this article were generated through a larger ethnographic study which was designed to explore everyday life inside this type of institution. Moving on to discuss the findings, I first detail different motivations the staff and young people had for a positive inspection outcome. I then theorise how the behaviours and interactions I observed on the day of the inspection changed. In doing so, I demonstrate a unique application of Goffman's (1959Goffman's ( , 1961 theorising, using it as he intended to explore face-to-face social interaction within a closed establishment. To address the critiques of Goffman's work, and explain collective motivation for the change I observed, I use Campbell's (1958) notion of 'entitativity' which illustrates why, in this given social situation, the staff and young people came to share an insider institutional identity. I conclude by (a) underscoring the importance of recognising that the view that outside inspectors can gain of life inside closed institutions will always be, to some extent, artificial and (b) calling for research on new methods of assessment to gain more authentic insight, while acknowledging the challenge of orchestrating the ethnographic vantage point achieved in this research.

Secure Children's Homes
SCHs are small locked ('closed') institutions for young people aged between 10 and 17 who have been deprived of their liberty. They are registered children's homes that form one part of the secure estate in England and Wales. Young people in England and Wales can be detained in a SCH on welfare grounds, on the basis of Section 25 of the Children Act 1989, which means that they are considered likely to abscond from non-secure settings, risking significant harm to themselves and/or others. Young people who have committed offences and who have been sentenced on the grounds of criminal justice legislation can be accommodated in a SCH, a Secure Training Centre (STC), or a Young Offenders' Institution (YOI). So-called 'young offenders' are typically held in a SCH if they are aged under 15 and/or considered especially vulnerable. SCHs have been described as 'both incarceration and an alternative to incarceration, a form of control imposed in order that care can be provided' (Harris and Timms, 1993: 4).

OFSTED inspections
OFSTED inspects and regulates education providers and childcare institutions, including SCHs. The purported goal is to 'achieve excellence in education and skills for learners of all ages, and in the care of children and young people' (OFSTED, 2017a). The results of these inspections are considered indicators of the quality of institutional performance and are significant to a wide range of stakeholders (Richmond, 2019). OFSTED (2011) inspections of children's homes are unannounced so that, in their words, 'we can see what the home is really like and what happens at the home on an ordinary day, and staff cannot prepare for our inspection' (p. 4). From 1 April 2017, OFSTED launched the Social Care Common Inspection Framework (SCCIF) to improve consistency across inspections for children's social care providers. Children's experiences and progress are said to be central to the new SCCIF, and inspectors are guided to be present after the school day, to engage with the young people and gather their views (OFSTED, 2017b). OFSTED inspections of children's homes take place over 1 day, if it is an interim inspection, or 2 days, in the case of a full inspection. Judgements are based on a detailed examination of paperwork and case records in addition to observations and conversations. A children's home can be judged as 'Excellent', 'Good', 'Requires Improvement to be Good', or 'Inadequate' (OFSTED, 2017c). Poorly performing children's homes risk enforced closure.
Knowledge about OFSTED inspections generally -and methodological issues in particular -is limited, which is somewhat surprising given the salience of the results across education and childcare institutions. Richmond (2019) is critical of the fact that OFSTED itself has not published any research since its inception to attest that their judgements are accurate indicators of educational quality. Matthews et al. (1998) suggested that OFSTED inspections are largely reliable, based on their findings in primary and secondary schools in England. They found that when two inspectors were observing the same lesson, 97% arrived at a grade within the same grade boundary, and 66% arrived at the same grade. However, as Richmond (2019) points out, this means that inspectors observing the same lesson awarded different grades in 33% of cases. Moreover, there are grounds to question whether this is even likely to be an underestimate of grade disparity between inspectors.
Matthews et al.'s findings were based on 173 pairs of observations, but the inspectors were self-selecting, and likely therefore to be those who were the most confident and experienced. Also, Matthews and two of his three co-authors were employed by OFSTED at the time of this research (and the third worked for the Dutch Inspectorate), which raises the possibility of bias.
Shortly after Matthews et al.'s (1998) publication, other researchers raised concerns about the reliability of OFSTED inspections, particularly on the grounds of inconsistencies in the application of judgement criteria (Campbell and Husbands, 2000;Sinkinson and Jones, 2001). There had been little in terms of follow-up in the 16 years or so since, until OFSTED (2017d) published its own research on the reliability of inspections. Using the same methodology as Matthews et al., this study sought to evaluate how frequently two inspectors -inspecting the same school on the same day -reached the same judgement. In 22 of 24 inspections, the inspectors agreed on the outcome, suggesting high reliability. However, these findings have been critiqued, by independent commentators as well as the National Audit Office, on three grounds. First, the sample size of 24 inspections in one type of institution is too small. Second, the sample is unrepresentative, because all the lead inspectors in this study were Her Majesty's Inspectors (HMIs). HMIs are more senior and experienced than OFSTED inspectors -because they are directly employed to carry out inspections, as compared to OFSTED inspectors who are infrequently contracted -but they are less likely than OFSTED inspectors to lead inspections. Third, there was a lack of monitoring as to how autonomously decisions were reached (National Audit Office, 2018; Richards, 2017;Richmond, 2019). Independent research has also raised concerns over the reliability and validity of the judgements made by OFSTED (Allen, 2015;Hutchinson, 2016;Strong et al., 2011). Research recently carried out by OFSTED shows an engagement with these issues, but little in terms of progress (Muijs, 2019).
This article argues that this predominant focus on reliability and validity in existing research overlooks a fundamental issue with the OFSTED inspection process itself; that inspectors -as outsiders -can only access the version of everyday life that insiders present. On this basis, the OFSTED grade awarded is a measure of the institutional representation afforded to the inspectors, rather than an objective assessment of service quality. I argue that this new insight is only possible because of my experience of being an immersed ethnographer inside a SCH at the time of an OFSTED inspection, as I explain next.

Method
The data on which this article is based were not generated through research designed to address a predetermined question. Rather, the findings arose fortuitously, in the course of an in-depth interpretivist ethnographic study of the mix of legal placements inside a SCH, 1 undertaken as part of an Economic and Social Research Council-funded piece of doctoral research. 2 The SCH that acted as the case is referred to using the pseudonym 'Woodside'. 3 Although I do not disclose the details of the SCH or the date of the inspection in order to protect the anonymity of the institution, it is significant to note that I considered myself to be immersed in the field at the time the OFSTED inspectors entered. This immersion afforded me a 'privileged insight' (Fountain, 1993) and the opportunity to see what the OFSTED inspectors could not see: a change in the behaviours and interactions of the staff and young people on the day of the inspection. Next, I detail the everyday experience of Woodside I had come to know by this time, before contrasting this with the 'institutional display' (Goffman, 1959) afforded to the inspectors.

Inside Woodside
Everyday life at Woodside was experienced as a battle for control between the staff and young people. Hours of observations revealed how most of the young people acted in violation of the rules and how the staff struggled to maintain order. There were relatively frequent violent outbursts, between the young people, and/or directed at staff, and/or self-inflicted. The residential units became increasingly stark over the duration of the fieldwork, as items that could be used to inflict injury, such as canvas pictures on the walls, or items that had been vandalised, such as televisions, were removed. The panic alarm sounded frequently as members of staff called for assistance from colleagues to handle escalating situations, as much as eight times during one particularly bad 8-hour shift during my fieldwork. The staff themselves were divided and critical of each other, while the young people appeared to align with those who shared the same legal status as them -either 'welfare' or 'criminal' as colloquially expressed -and actively oppose those who did not. There was a shared feeling among the staff and some of the longer term residents that the institution was going through a very troubled period. The young people were open with me about their complaints and critical of their treatment and experiences. Similarly, many of the staff revealed how they were disparaging of the way the institution was operating, and that they were struggling with low morale and a lack of teamwork. With trust came the sharing of quite personal confidences from the staff. Even the manager, Neil, shared highly sensitive information about staffing problems and poor behaviour management within the institution with me in the context of an interview.
It was this closeness to the research participants -my insights into their thoughts and perceptions, and my familiarity with life in the setting -that gave me a privileged perspective into how the young people and staff acted differently on the day that OFSTED came to visit. I sat with the care staff in one of the offices as they were told by the managers what was expected of them. Case files were to be checked immediately, and all records brought up to date. As it was a weekday, all the children were to be in 'education' -the colloquial name given to the set of classrooms -and correctly dressed in their uniforms. Tasks such as these were allocated among the team. There was a strange feeling of tension, but also of apathy. Sean, the Head of Care, full of nervous energy himself, tried to motivate the staff, by emphasising that a poor result could mean closure. As I walked towards the secure residential area with the staff group, there was a palpable sense of foreboding.
However, once on the unit the behaviour of the staff altered radically. The team worked together, occupying themselves happily with their pre-determined task or activity. And, in turn, the young people -who had been alerted that OFSTED were present by the staff -behaved exceptionally. There was none of the usual bickering, swearing, or fighting between them, and they were not defiant towards the staff. The following extract from an interview with Scott, a young male on a long-term justice placement, demonstrates how I was not the only one to notice how the staff and young people acted differently when the OFSTED inspectors were present: It's funny because when OFSTED's here they try and do everything right, like, do you get what I mean? It's funny how when I come back from school, Tommy was in his corridor and Aaron was in his room, because he got sent back from school. But if OFSTED weren't here, they'd both be sitting in the quiet area. Do you know what I mean? And then everything's just trying to run smooth. And they're trying to like do everything they're supposed to do. But when OFSTED's not here, it's just going back to normal . . . Next, I detail factors that may have motivated individuals to act in line, before theorising the overall unity of the 'institutional display' (Goffman, 1959).

Motivations for positive inspections results
There appear to be several factors that might have motivated the staff to seek a positive outcome from the OFSTED inspection. First, there was a sense that the staff perceived that the outcome as a measure of personal ability. This was demonstrated by Cyrus, a long-standing member of the care team, when he was speaking about the negative results of a previous inspection. The full extract from the interview, where Cyrus makes this point clearly, cannot be included, because he discusses inspection results which could compromise the anonymity of the institution. However, within this discussion, Cyrus reflected on the results of one inspection and explained how he interpreted the results as meaning: 'You are absolute crap at the job you do, that's what it tells me. Don't it ..?
That's what it tells me'. Cyrus explained how a judgement of 'Inadequate' would be interpreted as 'That's an inadequate at doing the job. That's what it tells me'. When asked whether this perspective was shared among the staff team, he explained, 'Personal reaction, and I think a lot of staff's personal reaction'. For Cyrus, and many of the other staff, negative results were experienced as demoralising. My fieldnotes from a staff meeting after the results of a previous poor OFSTED judgement were discussed also reflected Cyrus' view. Several spoke about feeling deflated, and others stated that they had nothing positive to say (fieldnote, date restricted). Furthermore, in an informal conversation, another member of staff shared with me that there was a feeling among the staff that they had worked hard to make changes and the result of the inspection had felt to him, and others, 'as a blow' (fieldnote, date restricted). These data suggest a link between the OFSTED judgement and perceptions of personal performance.
Second, the reliance of the staff on their jobs may have influenced their aspiration for the OFSTED inspectors to view the institution positively. This was made visible by the difference in performance between Martin, a member of the care staff, and the rest of the staff. I came to know Martin quite well through our extended informal conversations. We quickly developed a good rapport, and he told me early in my fieldwork that he appreciated that I was spending time in the evenings and at the weekend so that I would get a 'real sense of the place' (fieldnote, 28 February 2014). Martin was not reliant on his job, and he had spoken many times about leaving. He was deeply unhappy about behaviour management at Woodside. On the day of the visit from OFSTED, Martin and I were momentarily alone in the communal area on the residential unit when one of the inspectors walked through. Martin actively approached the inspector and told him how he felt the young people were out of control and the staff could not manage them. At the same time, he criticised the OFSTED inspector for visiting briefly, during the daytime of a weekday, when most of the young people are in education. Martin told the inspector that he should visit in the evenings or at the weekend, when the staff are struggling to look after the whole resident group and their conflicting needs. Martin departed so markedly from the impression the rest of the staff were giving; it was like he was exposing what everyone else was working hard to hide. Everyone else, however, was, to my knowledge, dependent on their job.
While the staff, aside from Martin, may have been motivated by the perceived personal consequences of a negative judgement, the young people's motivations for a positive judgement can be explained in terms of their institutionalisation, relationships to the staff, and issues of trust, all of which can be linked to their fears of leaving Woodside.
All of the young people were critical of their experiences at Woodside, citing boredom (something Bengtsson (2012a) also found in her research), perceived unequal treatment, dissatisfaction with the environment, and levels of violence in particular. However, despite their frustrations, some were fearful of leaving. Rhianna, a slight 14-year-old, accommodated on a justice placement, spoke at length about her worries in her interview. The following extracts capture how she did not want to leave and gives a sense of her fears: Rhianna: I don't even want to go now Caroline: You don't want to go home? Rhianna: No, not really . . . . . . .

Rhianna:
Yeah, it's a bit paranoia when you're about to leave as well. . . because I'm starting to get a bit worried where I'm leaving, because where I haven't been out in so long, I've got a fear that I'm going to get jumped and everything like that. . . so you're on your own yeah, staff won't protect you anymore, so all of a sudden you're just going to get jumped out of nowhere one day.
In addition, the young people shared stories about other SCHs, STCs, and YOIs, which made some of the young people fearful of going to a different institution. In relation to another STC that Rhianna had never visited she said, 'It's bigger kids in there, people that want to start, and its proper lock down' and 'people burn your rooms down'. Among young people on welfare orders, stories of psychiatric units and medical restraints thrived.
Beth, a young person on her second welfare placement at Woodside, also became increasingly anxious about leaving as the end of her placement approached. Beth's conflicted feelings were captured when she said in her interview, 'Woodside is my home, but I hate it'. Within days of her departure, she arrived back at Woodside, having run away from her new open residential placement. Beth, by this time aged 16 and no longer under a court order for secure accommodation, was not eligible for a bed in a SCH. Beth refused to leave and spent the night on the chairs in reception. There was a sense then that these young people were institutionalised, given their 'acceptance of institutional life', and linked to their perceptions of 'an inability to cope on the outside' (Townsend, 1976: 263). Such institutionalisation could reasonably encourage these young people present their situation in a positive way, despite their frustrations with their everyday experiences.
In addition, the young people formed relationships with the staff. Young people in SCHs are likely to have had adverse childhood experiences, with significant disruption, loss of contact, or even death of a close relative (Ellis, 2012;O'Neill, 2001;Sinclair and Geraghty, 2008;Walker et al., 2006). Sadly, these experiences were familiar to many of the children in this research. Not all the staff were popular -and some were certainly not -but many of the young people formed strong attachments with certain individuals. To quote Rhianna again, she said, 'I've got parents here that I call "Mum" and "Dad"'. Reika, who was serving a custodial sentence for a violent offence, explained to me that she was so upset after having punched Fred, a member of the care staff, in the stomach, because he was like 'a second dad' to her (fieldnote, 20.02.17). In addition to these attachments, there was, for some of the young people, a sense of loyalty to the staff, even if it was expressed begrudgingly. For example, Beth said, 'I sort of have respect for them, even the ones I don't really like, I've got respect for them'. These feelings of loyalty and attachment could also explain why the young people acted to show Woodside in a positive light.
Finally, issues of trust with vulnerable young people can also explain why they might not have disclosed any of their concerns. As mentioned above, it took time for me to develop rapport with the young people. Emmel et al. see trust between a researcher and a study population as being influenced by the degree of social separation. For them, '. . . relationships of credibility and trust are built through immersion at the research site' (Emmel et al., 2007: 2). As a university-educated adult female in my early 30s with no previous experience of residential childcare, I was socially distant from the young people, but I was immersed and able to develop trust over time. The OFSTED inspectorsolder males, perhaps on the verge of retirement, wearing formal business suits and carrying corporate-looking black folders -were arguably even more socially removed than I, and they had no time to build relationships with the young people. To illustrate this point, Scott, the young male introduced above, was very open in his interview with me and confided his perceptions such as '. . . they don't care . . . the managers . . . they really don't see what the problems are on the unit'. He spoke in detail about what he saw as inconsistent and preferential treatment by the staff, a lack of resources, a poor environment, and a lack of safety. However, as illustrated below, he told me that he would not share these views with an OFSTED inspector: Lee-Treweek (2002) describes the importance of others' experience of those to be trusted as influencing trust. The staff talked openly, and negatively, about OFSTED in front of the young people. The view that OFSTED could close Woodside if there were further negative inspections was commonly cited among the staff and the young people. If 'Trust is developed in situations where we trust that individuals or institutions will commit actions that will be favourable to our needs and interests' (Emmel et al., 2007: 3) and the young people feared leaving Woodside despite their frustrations, it is not so surprising that Scott, like all of the other young people, complied in presenting Woodside in a favourable light to the OFSTED inspectors.
However, while the staff and young people may have had these individual motivations for their behaviour, something more is needed to theorise the collective behaviour of the staff and young people in the presence of outsiders. This can be done, as I show next, by a unique application, and extension, of elements of Goffman's theorising.

Theorising the institutional display
Goffman was fascinated by the intricacies of microsocial interaction. Although his work on face-to-face interaction and impression management in The Presentation of Self in Everyday Life (Goffman, 1959)

is often treated as distinct from his work on institutional life in his later book, Asylums Essays on the Social Situation of Mental Patients and
Other Inmates (Goffman, 1961), it is possible to trace his thinking between these two influential publications. In the former, Goffman set up his dramaturgical perspective, using the metaphor of the theatre to explore the way social actors present themselves to each other within social establishments. Within this, Goffman (1959) set up a distinction between actors inside and outside of 'relatively closed' establishments (p. 232) and stratified groupings within social organisations. In the latter publication, Goffman (1961) popularised the ideal-typical concept of the 'Total Institution' (TI) which also makes a clear distinction between insiders -'cut off from the wider society for an appreciable period of time' (p. 11) -and outsiders, as well as between staff and inmate groups. Before delineating and applying these ideas in more detail, it is important to note the relevance of my tracing of his thought in this way. First, while Goffman's dramaturgy has been widely applied to other studies of social life (inter alia McQuarrie et al., 2013;Pounders et al., 2016;Ward, 2015), it is arguable that Goffman intended that his dramaturgical metaphor be applied to the study of interaction within closed institutions. My combined application of these elements of Goffman's (1959Goffman's ( , 1961 theorising to the inspection process within a closed institution is in line with this intention and is, to my knowledge, original. Second, this example of the links between his ideas can be used to refute critics who suggest that Goffman produced 'diverse yet not explicitly interrelated conceptual frameworks' (Smith, 1999: 11).
In The Presentation of Self in Everyday Life, Goffman (1959) theorised as to how a social 'actor' presents herself or himself to her or his 'audience' through language and actions, with a key distinction made between 'the expression that he gives, and the expression that he gives off' (p. 14, emphasis in original). The former refers to information conveyed orally, the latter the actions and symbolisms that accompany this communication. Goffman focused mostly on the impression an actor 'gives off' and conceived of the actor as attempting to control the impression others have of her or him.
Pursuing the theatrical metaphor, Goffman theorised that the presentation an individual 'gives' and 'gives off' will depend upon the stage she or he is acting. Goffman distinguished between the 'front region' ('front-stage') where an individual performs for an audience, providing her or his definition of the situation and paying particular attention to how she or he engages with the audience ('matters of politeness') and her or his comportment ('decorum'). In the 'back region' ('back-stage'), she or he is not in view of the audience and she or he does not have to perform. She or he may be among others, but not those for which a 'front-stage' performance is necessary (Goffman, 1959: 109-110). Goffman (1959) also drew attention to the 'outside' region of a social establishment, neither 'front' nor 'back', which restricts the access of 'outsiders' for whom a special type of performance might be required (p. 135). These concepts are metaphorical, applicable in different ways to the same physical space, to explore different social 'encounters' (Goffman, 1959: 26).
Goffman developed his sociological perspective of social interaction beyond the performances between an individual and her or his audience, to dialogue between groups, which he referred to as 'performance teams' ('teams'). These teams 'cooperate in staging a single routine' (Goffman, 1959: 85). Goffman called this 'dramatic interaction' and emphasised that in most cases, each team tries to sustain the definition of the situation the other is trying to portray. Within teams, members must act together, preventing the disclosure of any 'destructive information' that might discredit their performance (Goffman, 1959: 141). These teams are not fixed, but rather form and re-form in different contexts. According to Scott's (2015) reading, 'The reference here to a "single routine" reminds us that these formations are contextual: fellow actors may be supportive teammates in one situation but . . . adversaries in another' (p. 7).
Moving to his later work, Goffman (1961) described a TI as a 'place of residence of work where a large number of like situated individuals. . . together lead an enclosed, formally administered round of life' (p. 11). Goffman (1961) described an incongruity between how TIs present themselves to the public -'as effective machines for producing a few officially avowed and officially approved ends' (p. 73) -and how they are experienced by the staff and 'inmates' within. Goffman (1961) wrote about 'watchdog agencies' (p. 75) that would monitor standards inside TIs and said that when there is a visit from a representative of 'an agency responsible for controlling a whole class of institutions; then we can expect the preparation of a display to be especially elaborate ' (p. 97).
Within the literature on the secure estate, the conceptualisation of SCHs as TIs has been made previously because of their closed nature (Bengtsson, 2012b;Ellis, 2012;Kelly, 1992;O'Neill, 2001). This literature describes how young people deprived of their liberty eat, sleep, play, and are educated -akin to Goffman's notion of 'work' -among their peers within the same confined space. I apply the same conceptualisation in this article. However, to my knowledge, Goffman's ideas have not been applied in research on the secure estate beyond this, which is surprising given his focus on social interaction and institutional life. My application of combined elements Goffman's theorising below is therefore original. Goffman's (1959) dramaturgical theorising makes sense of the change in social interaction between the staff and young people at Woodside on the day of the OFSTED inspection, as described above. The staff and young people can be conceived of as actively attempting to control the impression that the OFSTED inspectors formed of the institution. A 'front-stage' performance was executed in the presence of the inspectors, transforming interaction between the staff and young people so that it was addressed to outsiders. The normal tensions and divisions between the staff and young people, as well as within these groups, were put aside, and the actors re-formed as unified 'teams'. In their respective teams, the staff and young people worked together to sustain the definition of the situation the other team was attempting to portray. A clear example of this the extent to which the young people complied unquestioningly with the requests of the staff. The exception to this is Martin's behaviour, as outlined above. However, far from undermining this theoretical conceptualisation, Martin's behaviour can be explained through Goffman's theorising. Goffman (1959) explained that A team-mate is someone whose dramaturgical cooperation one is dependent upon in fostering a given definition of the situation; if such a person comes to be beyond the pale of informal sanctions and insists on giving the show away or forcing it to take a particular turn, he is none the less part of the team. In fact, it is just because he is part of the team that he can cause this kind of trouble. (pp. 88-89) Martin was arguably beyond the 'pale of informal sanctions' because he was not concerned as to the personal consequences of a negative OFSTED judgement. Goffman's (1961) theorising of the distinction between how institutions are privately experienced and publicly presented also chimes with my ethnographic insight into the contrast between how Woodside was experienced every day, and the experience of everyday as presented to OFSTED. OFSTED can be conceived of as Goffman's notion of a 'watchdog' agency. The difference here is that the OFSTED visit is unannounced, and there is little time for preparation. Regardless, the staff and young people slipped seemingly effortlessly into their normative roles.
As I have shown, a combination of Goffman's (1959Goffman's ( , 1961 theorising conceptualises the reformation of the staff and young people within this closed institution from adversarial to cooperative groups in the presence of outside inspectors, demonstrating the heuristic value and contemporary relevance of his ideas. However, Goffman has been criticised for focusing on the 'how' of social order to the neglect of the 'why' of social action (Scott, 2015: 18-19), alternatively described as failing to account for the motivations of human action (Giddens, 1984). The counter to this critique is that Goffman's focus was on 'how social actors present, perform and strategically manage different versions of themselves in different situations' and on 'observing rather than explaining . . . the mechanics and dynamics of social interaction' (Scott, 2015: 11, 19). His was a 'sociology of "co-presence", of what happens when people are in one another's presence' (Gouldner, 1971: 379). Goffman set out to demonstrate what happens in social interaction, rather than elucidate why, and to criticise him for the latter is to misunderstand his intentions.
To explain why the staff and young people changed their behaviour as they did, I look to other conceptual insight. As a result, I make a novel extension of Goffman's theorising drawing on Campbell's (1958) concept of 'entitativity'. Entitativity relates to the notion of a group as a meaningful entity, as opposed to just a collection of persons (Hamilton, 2007). The importance of the group to the group members depends on the extent of interaction and a sense of shared goals and outcomes (Lickel et al., 2000). In the everyday life of the institution, the goals of the staff and young people are largely in opposition to each other, as the young people act to demonstrate their agency and the staff attempt to maintain order. The outcomes are not shared either; the young people face consequences in terms of sanctions, whereas the staff -their emotions notwithstanding -do not generally experience tangible outcomes based on the events of each day. However, in the presence of the OFSTED inspectors, the staff and young people come to share goals and outcomes; they all (with one exception) want to show the institution in a positive light, for varying reasons as outlined above. While the young people and staff arguably still have a sense of being separate groups -given their distinct roles and status within the institution -they also come to see themselves having more in common with each other when outsiders come inside, fostering a rare sense of a shared institutional identity. This shared group identity motivates the harmonious 'institutional display' (Goffman, 1961) afforded to the OFSTED inspectors. Extending Goffman's (1959Goffman's ( , 1961 theorising using Campbell's (1958) concept of 'entitativity' in this way enables me to conceptualise how the social actors reform in the presence of outsiders, as well as the motivation for their reformation. This theoretical extension addresses the -albeit unfounded -critiques of Goffman.

Concluding remarks
In conclusion, this article finds that the observational component of OFSTED inspections of SCHs captures an 'institutional display' (Goffman, 1959) rather than 'what the home is really like and what happens at the home on an ordinary day' (OFSTED, 2011: 4). The unannounced nature of the inspections, though seemingly strategic in terms of gaining insight, does not hinder the display that the insiders can offer; the staff and young people were easily able to perform their normative roles. In demonstrating this finding, this article has also added to the limited knowledge about life inside SCHs and OFSTED inspection practices.
My original application of Goffman's (1959Goffman's ( , 1961 theorising -as he envisaged it being applied -has added conceptual insight to the transformation from adversarial to cooperative interaction between the staff and young people in the presence of the OFSTED inspectors. My novel extension of Goffman's theorising using Campbell's (1958) concept of 'entitativity' offers an explanation as to why, in this given social situation, the staff and young people were motivated to show the institution in a positive light.
There is no reason for this finding to be limited to OFSTED inspections of SCHs. Rather, the inability of outsiders to see beyond the 'institutional display' insiders offer is arguably applicable more generally to internal inspections of closed institutions, including prisons and locked psychiatric settings. However, as this article illuminates, the extent to which institutional life is misrepresented by insiders will be impossible for outsiders to know.
This limitation to external inspections needs recognition, and new ways of assessing institutions need exploration. A greater reliance on documentary evidence is not considered advantageous because paperwork is prepared for an external audience. Rather, given the motivations for a positive outcome established above, inspectors spending longer inside institutions -attempting to embed themselves -or involving staff from similar establishments -who might better be able to establish trust -in inspections might encourage more openness from insiders. The challenge for researching new inspection strategies is achieving the vantage point needed to see insider misrepresentation; being deeply immersed as an ethnographer in the field at the time of an unannounced inspection in this research was entirely fortuitous, and it would be difficult to orchestrate this opportunity. Still, research in this area must continue, to gain more authentic views of everyday life inside closed institutions.

Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes
1. The data presented in this article were generated through 396 hours of observation over 7 months between January and July 2014, including 24 interviews with staff and 14 interviews with young people. I spent time across the institution, at various times of day and on different days of the week, recording verbal and non-verbal interaction, behaviours and practices, as well as my own interpretations and emotional reactions to the setting. 2. The author is grateful for funding provided by Economic and Social Research Council (ESRC) which enabled this doctoral research. ESRC Funding Award reference 1229184. 3. The purpose of the larger research project was to explore the 'particular conjunction of the therapeutic and the penal' (Harris and Timms, 1993: 171) in Secure Children's Homes (SCHs); how the mix of young people deemed to be 'at risk', accommodated alongside those who pose 'a risk' (O'Neill, 2001), plays out in everyday life. In line with ethical commitments to protect the anonymity of Woodside and the research participants, details of the geographical location, the name of the local authority, the size of the institution, the ratio of welfare and justice beds and even the date and outcome of this, or any other OFSTED visit, are not disclosed.