Improving private security–youth relations in quasi-public spaces: control, care, cooperation

ABSTRACT The global growth of the private security sector is connected to the increasing number of quasi-public spaces policed as private ones but used as public ones. The intensive policing of young people in shopping centres has, however, led to problematic private security–youth relations. This article focuses on a new method in which youth workers train security guards to improve these relations in shopping centres in Finland. The study uses qualitative content analysis to explore unique data – security guards’ daily reports. The encounters displayed control, care and cooperation: security guards controlled the youths’ activities but also talked with them and helped them. The study argues that the role of security guards is even more plural and blurred than previously understood, including a “social” orientation, and cooperation with public, private and third sector agents. Finally, the study highlights the importance of developing procedural justice, both in theory and in practice.


Introduction
The pluralisation of policing suggests that the distinctions between public and private policing have become blurred and that private security increasingly involves policing private, public and quasipublic spaces. It has been suggested that the global and rapid rise of private security indicates a radical shift: from a criminal justice system monopolised by the state and its police, to a pluralised policing system in which the private sector also participates in delivering security (e.g., Shearing & Stenning, 1981;Garland, 2001;White & Gill, 2013;Schuilenburg, 2015;Paek et al., 2020). For instance, shopping centres, defined as quasi-public spaces, are used as public spaces but policed intensively by the private security industry (Von Hirsch & Shearing, 2000;Wakefield, 2003;Manzo, 2004). The role of private security in various locations, as well as the cooperation and collisions between public and private policing are under active discussion (e.g., Michael, 1999;M. Nalla & Hummer, 1999;Hummer & Nalla, 2003;White & Gill, 2013;Kim et al., 2018;Gurinskaya & Nalla, 2020;Meerts, 2020;Nøkleberg, 2020;Paek et al., 2020). Very little is known about the cooperation of private security with other agents, such as the third sector, on which this article focuses. Crawford and L'Hoiry (2017) stress that among policing professionals, inter-disciplinary collaborations with public, private and third sector providers, called boundary crossing, is a means of promoting selfreflection and reflexivity.
Whether policing agents act in a procedurally just manner, treat citizens fairly and respectfully and avoid misuse of their authority, excessive measures and violence, particularly towards vulnerable groups and minorities, is again a highly topical issue. As private security has been growing rapidly worldwide and is increasingly conducting police tasks and social control, they are important to consider. Researchers have been increasingly interested in citizens' general perceptions of private security as well as young people's and minorities' experiences (e.g., Gabbidon, 2003;Keskinen et al., 2018;Moreira et al., 2015;Moreira & Cardoso, 2020;M. K. Nalla & Lim, 2003;Saarikkomäki & Alvesalo-Kuusi, 2020). Studies conducted in different countries, including Finland, where this study takes place, have demonstrated that young people often perceive security guard interventions as negative, too frequent, unfairly targeting them, moving them on for no reason, lacking respect, or as coercive and escalating conflicts (Matthews et al., 2000;Fine et al., 2003;Lampela et al., 2016;Saarikkomäki, 2016Saarikkomäki, , 2018. Fine et al. (2003) found that young people who had many adverse encounters with policing agents felt unwelcome in public spaces and mistrusted by adults. Despite some positive findings of security guards chatting or acting fairly, young people question the importance of security guards' work, trust and legitimacy because of unfairly perceived treatment (Saarikkomäki, 2016(Saarikkomäki, , 2018Saarikkomäki & Alvesalo-Kuusi, 2020).
Thus, there is a pressing need to continuously find new ways to improve relations between young people and policing agents. Despite a handful of studies that have focused on young people's views, there is a lack of knowledge on private security guards' work and role in their encounters with young people. A few ethnographic studies investigating security guards in shopping centres suggest that their work partly focuses on moving young people and marginal groups on (Wakefield, 2003;Manzo, 2004). However, the intensity of policing varies, and some managers wish to have young people around and view them as important future customers, while others do not (Wakefield, 2003;Manzo, 2004;Kim et al., 2018).
In recent years in Finland, youth workers have cooperated with private security and shopping centres to enable young people to spend time in these spaces. As a result, a new work methodyouth-focused security guard -has emerged. Youth workers train security guards to interact more positively with young people and involve young people in, for example, making common rules for the spaces (Lampela et al., 2016). As data, this study uses daily reports that youth-focused security guards completed after each work shift in 2019 about their encounters with young people. To our knowledge, no type of security guards' reports have been used before as research material. Furthermore, the context in which the youth-focused security guards' work enables ideas from youth work (public sector and non-profit) but still involves private security (private sector) and thus provides an interesting context to study the pluralisation of policing and the diverse roles security guards can have.
In this article, we first present our theoretical concepts related to our interest in studying how to improve relations between policing agents and citizens. Second, we introduce this new work method and the context of the study. Third, we broaden our study framework by discussing the pluralisation of policing, giving particular focus on quasi-public spaces. After this, we present our data and findings, and draw conclusions. Our key findings show that these encounters simultaneously involve control, care and cooperation. Firstly, this article suggests that it is crucial to assess procedurally just policing from new angels and in relation to private policing. Also, conventional ways of seeing crime prevention are challenged. Secondly, our findings indicate new insights for the plural policing literature by demonstrating how the role of private police is again even more blurred than previously understood.

Improving relations between control agents and citizens
Procedural justice theory has provided understandings on how to improve police-citizen relations. It has, however, largely focused on public police and ignored the pluralisation of policing. In recent years, research on plural policing and procedural justice has however grown (e.g., Fagan & Tyler, 2005;Gurinskaya & Nalla, 2020;Mopas & Stenning, 2001;Moreira & Cardoso, 2020;Paek et al., 2020;Saarikkomäki, 2016;Saarikkomäki & Alvesalo-Kuusi, 2019). Procedural justice involves: a) citizens' perceptions of fair, neutral and unselective criminal justice processes, b) perceptions of policing agents treating people with politeness and respect c) trustworthy motives, and d) participation, involving citizens in decision-making processes (e.g., Tyler, 1990;Tyler & Lind, 1992;Fagan & Tyler, 2005;Murphy, 2015). The theory stresses that citizens' perceptions of fair treatment are the primary influencer of trust in the legitimacy of the system (Tyler, 1990). Voluntary compliance and accepting decisions against one's own interests depend on the belief that the authority views one as a full member of society, as well as on trust in the authority's ethics and neutrality (Tyler & Lind, 1992). Young people appreciate it when public and private policing agents treat them in a procedurally fair, friendly and respectful manner; show emotions (vs. instrumental moving on); manage negative emotions; act professionally; or explain the reasons for their interventions (Saarikkomäki, 2016). The procedural justice model is largely based on quantitative research and has not paid enough attention to nuances, emotions and interaction.
The actions of policing agents have a substantial effect on how people respond to them. Van Stokkom (2011) argues that if police agents get carried away with emotions or in contrast, practise frozen behaviour, this can result in citizens losing their temper or the authority of policing officers being compromised. Aiming to understand security guards' many possible roles and how to improve their relations with young people, Lampela and Tani (2015) have conceptualised four ideal types of security guards based on Baumrind's (1971Baumrind's ( , 1991 parenting styles. A) The "Pedagogue" is a security guard who is a warm but firm adult. They engage in discussions with young people, negotiate the rules, understand young people, and use democratic power. They encourage young people to act more responsibly and inspire participation. B) The "Disciplinarian" is highly controlling, even aggressive, and promotes provocation and other non-desired responses from young people. C) The "Passer-by" is not interested in young people and does not engage in conversation, which might result in young people using drastic ways to get the adult to notice them. D) The "Buddy" is permissive and has no authority, which results in young people not taking responsibility for their own actions (Lampela & Tani, 2015).

Youth-focused security guard and the study context
In Finland, shopping centres are important locations for young people to spend their time, where they often encounter security guards (Saarikkomäki, 2018;Pyyry & Kaakinen, 2019). Finland is a Nordic country of about 5,5 million people, in which private security sector has grown rapidly. There were 9 000 active licenced guards in 2015 (Source: CoESS, 2015), outnumbering police officers (7 300 in 2021). In 2019, there were 111 shopping centres, and the industry has a 30-year-old history (Finnish Council of Shopping Centers, 2020). The shopping centres are increasingly becoming town centres and include public services, leisure time activities and other non-retail elements, such as public transport, libraries and health services, restaurants, children's playgrounds, youth clubs, hotels and residences. Shopping centres are a very common feature in all Finnish cities and retail parks and hypermarkets are maybe even more common.
In this study, we focus on the six (rather typical) shopping centres which have adopted the new "youth-focused security guard" work method. This method has been developed by a Finnish NGO (Youth Service Association, Nuorten Palvelu ry), in cooperation with the worldwide private security company (Securitas) and Nordic shopping centre owner corporation (Citycon). The idea is to cross boundaries and merge the professional skills and ideologies of youth work with security guards' work. This cooperation aims to enhance youth involvement and their possibilities to use city spaces, which are crucial objectives of youth work and youth research (Lampela et al., 2016). In Finland, youth work is largely carried out by municipalities (public sector) and NGOs (third sector), and financing is authorised by the government according to the Youth Act. The first youth-focused security guards began to work in 2016 and the number is increasing.
To our knowledge, even globally, this kind of cooperation between youth work and private security is scarce or non-existent. The Youth Service Association monitors, supports and trains the youth-focused security guards in three separate three-hour sessions, with at least two weeks in between to practise, and guards receive additional training on a yearly basis. The trainee gains knowledge of youth from a biological, psychological and social perspective and learns to understand different situations, for instance, young people who spend a lot of time in shopping centres and who do not necessarily have other reliable adults around. While recruiting, attention is paid that the person is motivated to work with young people. The Youth Service Association helps the security guards cooperate with key partners (e.g., police, municipal youth workers and other youth organisations, social workers and shopping centre management).
Even though youth-focused security guards have a special focus in their work, they still perform normal security guard duties. In Finland, the operations of shopping centre security guards are widely regulated (Button, 2007) through the Private Security Services Act. There are different types of security guards with different powers depending on their tasks and training. In this article, we refer to shopping centre security guards who are appointed and licenced by the police to maintain public order and safety on the application of the owner of the shopping centre. In comparison to security guards merely guarding private property, shopping centre security guards have a more public duty: they are actors of the private sector and have duties both to the public and the property owner.
Even though the public and private policing have blurred, there are differences in policing agent's roles and responsibilities. In Finland, the private security sector as a whole is supervised and managed by the police and the National Police Board. Therefore, on learning of a crime, security guards have a duty to report citizen arrests to the police and cooperate with them (Private Security Services Act, The Criminal Code of Finland). Finland differs for instance, from North America, where the police plays a more significant and visible role in many large shopping centres. In Finland, police officers are not usually present in shopping centres, which makes security guards a sort of primary policing agent and the first point of contact for the public. Police officers are called upon when needed, and the security guard's task is to assist the police or hand over the situation to the police. Police officers and security guards cooperate in different ways, for instance, in crime situations and police may also ask help from security guards.

Plural policing of quasi-public space
Shopping centres are increasingly dominating city spaces worldwide and because they blur the distinction between public and private spaces they are defined as quasi-public spaces (Shearing & Stenning, 1981;Von Hirsch & Shearing, 2000;Wakefield, 2003). This obscure nature of quasipublic space and the use of public and private policing makes it an interesting context to study (e.g., Gray & Gray, 1999;Von Hirsch & Shearing, 2000). One explanation for the rise of private security is linked to the "mass private property thesis", which refers to an increase in quasi-public spaces, that are privately owned, and thus policed as private spaces, but used as public spaces and host public services (Shearing & Stenning, 1981).
For young people, the shopping centre can present a warm, safe and easily accessible place to spend their free time, where adult supervision is different (Pyyry & Kaakinen, 2019). However, young people, ethnic minorities and marginalised groups are often seen as a threat or as loitering, not consuming, and they feel unwelcome and excluded in city spaces (e.g., Gabbidon, 2003;Wakefield, 2003;Keskinen et al., 2018). The discussion of citizen's right to use quasi-public space versus property owner's possibilities to exclude people or to prevent certain behaviour is an important one, negotiated and regulated somewhat differently in different countries (e.g., Shearing & Stenning, 1981;Gray & Gray, 1999;Von Hirsch & Shearing, 2000;Button, 2007). Critique of an overly strict control has been presented particularly when shopping centres host public services and are used as public spaces, which means that exclusion may result in loss of access to important services (Gray & Gray, 1999;Von Hirsch & Shearing, 2000). In Finland, shopping centre security guards do not have the right to deny access to the common areas of the premise, but they can remove people from the premise in certain situations defined in the law (Private Security Services Act). The shops inside the shopping centre instead have somewhat more power to decide who is allowed in their property. Nevertheless, the accepted ways to use the shopping centre space are under continuous negotiations also in Finland. For instance, making noise is an acceptable reason to remove people from the space, however, "noise" is a very subjective concept.

Blurred roles of public and private policing
Similar tasks and responsibilities between police and private security, as well as their differences, are highlighted in plural policing literature (e.g., Michael, 1999;M. Nalla & Hummer, 1999;Garland, 2001;Hummer & Nalla, 2003;Button, 2007;White & Gill, 2013;Kim et al., 2018). Furthermore, studies have provided valuable information on the relations and cooperation between private and public policing agents, and with other actors, in different countries and contexts (e.g., M. Nalla & Hummer, 1999;Hummer & Nalla, 2003;White & Gill, 2013;Søgaard et al., 2016;Meerts, 2020;Nøkleberg, 2020;Paek et al., 2020). The relationship and cooperation between public and private policing agents is complex and continuously evolving (M. Nalla & Hummer, 1999;Hummer & Nalla, 2003). There exist studies on understandings of guards' role compared to the police's, for instance, as "police officer wannabes", having lower educational level, lower salary and status, or security guards highlighting their multisided roles and not being only "police officer wannabes" (e.g., Michael, 1999;M. Nalla & Hummer, 1999;White & Gill, 2013). A recent study found that along with traditional ideal types of security guards "parapolice orientation" (active use of legal tools, enforcement and confrontations) and "watchmen orientation" (avoidance of danger and unactive), there exist a third, "servicemen" orientation (Kim et al., 2018). Kim and colleagues (2018, 93) found that security guards focused also on "service work", which meant giving directions and helping customers to find lost property, picking up garbage, or helping elderly (see also Wakefield, 2003;Manzo, 2004).
Even though positive views of private security's role are addressed (M. K. Nalla & Lim, 2003;Kim et al., 2018;Moreira & Cardoso, 2020), the legitimacy of private security is often questioned in Western countries (e.g., Mopas & Stenning, 2001;White & Gill, 2013). In Finland, there is no equivalent study to gain information on police-security relations or public views of private security. However, existing studies suggest that young people regard the police as a more trustworthy actor as compared to private security (Saarikkomäki, 2018). Studies have indicated that the security industry gains legitimacy through training, regulation through licencing, legal regulation, uniforms that resemble those of the police, and cooperation with the police (Mopas & Stenning, 2001;Thumala et al., 2011). Finally, White and Gill (2013) have suggested that when public and private policing systems collide with one another it creates, not only new roles but also new mixed rationalities for policing, for instance, private agents aimed to emphasise their role as securing safety and public order and downplay connections to market-based affiliations. Crucially, White and Gill (2013) argue that due to the blurred rationalities of public and private policing, the plural policing landscape is even more blurred than before.

Data and methods
As data, we use reports (n = 1139) completed by ten youth-focused security guards after each work shift from January to December 2019. The security guards were male and female, mostly 20 to 25 years of age (one was over 30) and worked in six shopping centres around Finland. During the study time, they were the only youth-focused security guards. We use the term "security guard" to mean youth-focused security guard, unless otherwise indicated.
The security guards were asked to write an open-ended description of their encounters with young people. In our data, for the most part, one encounter is one case, a data unit. However, some reports mention several cases of different youth activities in just one sentence, and these were calculated as one data unit (but coded in different sections, this is explained later). There were sometimes several encounters in one day and several reports per day. The length of one case varied from a couple of words to 27 sentences, typically being either couple of sentences or one paragraph. Permission to use the reports were received from the guards themselves, their managers, the shopping centres' managers, the private security companies (Securitas) and the shopping centre owners' (Citycon) managers. The descriptions were carefully anonymised to guarantee the anonymity of the security guards, young people and the shopping centres. The reports were not initially intended for research purposes but to monitor the quality of the encounters. Security guards in Finland also report their work in their own respective report systems (not used here). To our knowledge, prior studies have not used any kind of security guard reports as data, making this a unique experiment.
Obviously, there are limitations to our approach as our findings cannot be generalised without caution. Also, the reports are based on the security guards' own viewpoints. The data is representative in studying youth-focused security guards work (in Finland) as it includes all the shopping centres and all the youth-focused security guards. However, it is a sort of a case study of those six shopping centres that have incorporated the new work method, which means generalisations are difficult (see also Kim et al., 2018). This applies even to all the shopping centres in Finland, as those studied may, for instance, be more positive towards young people. As the study is qualitative, generalisations are theoretically oriented rather than data oriented. Our aim is to provide insights of the new roles and occupation cultures of security guards and ways encounters between them and young people could be enhanced.
We used qualitative content analysis (QCA) to categorise our data. While coding, we analysed manifest content, used an abductive approach, and analysed these categories further by discussing our theoretical approaches and prior research and by focusing on the context (e.g., Graneheim et al., 2017). We present the findings in quantitative and qualitative parts. We coded a) the nature of typical encounters, i.e., the reason for the security guards' intervention and b) whether the intervention involved talking with young people and c) whether it involved cooperation with other actors. We first read the security guards' reports and planned an initial coding list. To verify the reliability of the coding, both authors coded the whole data set independently in Excel using the pre-planned coding list. After this, we discussed and decided on how to continue with the cases that we had coded differently (some were simply mistakes whereas others required interpretation). In this phase, we combined some sub-categories and removed codes that had no cases. Based on this, we created the final set of five main categories and sub-categories (Table 1). We added "other" sub-categories when we coded a variety of situations that occurred less often. One report paragraph could be coded into several main or sub-categories. We previously analysed these types of reports collected between June 2016 and March 2018 in an article published in the Finnish Journal of Youth Research (Lampela & Saarikkomäki, 2019). In the current article, we use new data that have not been analysed before, partly using a new analysis strategy and new theoretical approaches.

Descriptive findings
The most common encounter between security guards and young people was related to chatting and deeper discussions with youths (41% , Table 1). (The rest of the reports had no specific mention of discussions; this does not mean that it did not take place.) Greeting or chatting without a specific topic were the most common interactions. Discussions were also related to delinquency: the security guards reminded the youths of the common rules of the shopping centre (7.3%), intervened in their crime or delinquency (5.9%) or substance use (2.3%). Positive feedback was also given, and there were discussions of the future plans, or issues that worried youth.
The second largest category (33%) was situations in which the security guards controlled hanging out and various other youth activities. The most common reasons were youth hanging out in groups or in places intended for other uses and blocking the passageways, as well as skateboarding or cycling indoors. Littering, smoking in front of the entrance doors and making noise were also intervened in. In 16.7% of the situations the security guards intervened because they suspected youth of crime, delinquency or substance use. Within this category, the most common reason was substance use (typically minors drinking alcohol), fights, shoplifting, and supplying alcohol.
Interestingly, we also found that the security guards helped young people in many ways. Around every tenth (12.7%, Table 1) encounters were related to helping young people, most commonly *Passages can be coded into several main and subcategories, except in the "Other (not belonging to above categories)" category. Therefore, the total N of each main category is not equivalent to the total amount of each subcategory. **Percentages are calculated from the total number of reports (n = 1139) because the security guards were concerned about their wellbeing. Practical help was also offered, such as opening phone charger boxes or finding lost items. In 1% of all the situations, the young person was a victim of bullying or security guards resolved other types of conflicts between young people. Less frequently, but worth mentioning, is the security guards helping self-destructive young people or victims of harassment.
We coded mentions of the security guards cooperating with other people (18.3%). This cooperation was usually with youth workers (9%). The cooperation between the Youth Service Association and security guards was not reported, as the data were produced for them. In 7% of all the situations reported, the security guards cooperated with police officers. Interestingly, they also contacted the youths' parents (3%) and in 1.5% of cases received requests to look for youths who had run away from foster care or other places. We also included a "school" category in our coding list, but there were no cases of cooperation with schools. Finally, we coded a separate main category called "Not specified/other/no youths to encounter" (14.5% of all cases), in which the reports typically only stated that it was a quiet shift with few youth encounters but gave no more details.

Controlling youths' hanging out and activities
A large part of the youth-focused security guards' work consisted of controlling youths' hanging out and activities (total 33.1% N = 377, Table 1). The young people used the shopping centre spaces in various ways for their free time activities and sometimes this was seen as a disturbance by shop or shopping centre owners, other customers or the security guards. This category included subcategories of hanging out, blocking passageways, littering, making noise, using offensive language towards other customers or security guards, smoking, entering forbidden or closed spaces, and putting their feet on tables (see Table 1 for percentages and subcategories). The security guards intervened using different measures: moving young people away from the shopping centre or to a different place in the centre. In this category, our findings partly reflect prior research, noting that youths' hanging out is often controlled by security guards and youths are often moved on (e.g., Matthews et al., 2000;Fine et al., 2003;Manzo, 2004;Saarikkomäki, 2018).
During the evening, we had to heavily intervene many times in youths blocking passageways, making noise and generally goofing around. We also had to move many youth groups on, who, despite being told to, continued inappropriate behaviour in the shopping centre premises. (Shopping centre 5) However, deviating from prior research, it seemed that the security guards also allowed some youth activities and did not always want to only move young people on. Furthermore, the guards talked to the young people, involving them, or asking them to clean up their mess or stop making noise or skateboarding indoors. The security guards took the time to explain the common rules of the shopping centre, trying to make the youths understand why certain behaviour is not allowed, such as talking too loudly or hanging out in places intended for other uses like restaurant tables or radiators. We noticed that the security guards also used "pedagogue" and not only "disciplinarian" techniques (Lampela & Tani, 2015). The procedurally fair interaction situations were appreciated by the young people: The other security guard removed four youths from the shopping centre because of disturbing behaviour. The youths said they thought it was unfair, so I went to talk with them about the situation. In the end they understood the reason for moving them on. (Shopping centre 1)

Controlling delinquency and substance use
A notable finding was that most of the encounters took place in situations not related to controlling crime and substance use. In our data, 16.7% of the situations occurred because the security guards suspected the young people of delinquency (Table 1, N = 190). When the youths were suspected of shoplifting (2.2%), the store staff usually informed the security guards, and the guards took the youth(s) to the detention room to resolve the matter and/or wait for the police to arrive. These situations did not always go well, as in the following example that escalated into a conflict. However, these were exceptions in our data.
A girl was removed because of earlier conduct. Right after, the salesperson of a store ran after and pointed at the same girl and said that the anti-theft alarm at the door had gone off. We caught the young person outside. (. . .) Resistance continued in the detention room and also verbal insults and threats. (. . .) We opened the door and then the youth threw something towards me, hitting my shoulder. We had to get her on the floor and handcuff her. (. . .) We rang the emergency number again about the changed situation, then the police arrived. We found alcohol on her and the girl was very drunk. (Shopping centre 4) The reports most often mention intervening in substance use if a young person was suspected of being under the influence or of possessing alcohol or drugs (5.9%), or of supplying alcohol or drugs (2.2%). Interestingly, these control situations also involved care. Surprisingly many reports indicate the security guards noted whether worrying reasons were behind the substance use, as in the extract below. The security guards checked whether the youths were underage and if not, usually they did not follow up the matter, only asked them to move on. If they were underage or in bad condition, the parents and/or the police were often informed. The security guards did not typically just move the problem away, which has been critiqued in previous literature on situational crime prevention (Von Hirsch & Shearing, 2000), but they tried to discuss and resolve the youth's situation. In this way, one could argue, they blurred the boundaries of typical security work and youth work and participated in boundary crossing (Crawford & L'Hoiry, 2017).
We got a report of a young, but not underaged girl, that other youths were worried about because she is going around intoxicated near the shopping centre. (. . .) Our chat revealed she was dealing with lots of problems in her life and that now she treats, for example, her anxiety by drinking alcohol. We gave her support and advice on what she ought to do so that it would not get any worse. (. . .) (Shopping centre 5) The reports sometimes involved a young person behaving aggressively towards other people or towards the security guards themselves. Situations involving fighting or assaults, or suspicion of a fight breaking out typically involved young people, and consequently, the security guards dealt with both the suspect and the victim. Vandalism included drawing on walls, graffiti, or other ways of damaging shopping centre property. To conclude, these types of intervention situations, carry the risk of young people experiencing the interventions unfair and procedurally unjust. Particularly moving on can be experienced as unjust if the young people do not see a clear reason for the intervention and it merely targets hanging out or making noise. In many reports, the security guards perceived that the young people understood the rules and did not want to break them again after they were discussed. However, a challenge in security guards' work is to balance being a relaxed and approachable adult and a controlling authority (Lampela & Saarikkomäki, 2019).

Caring for young people: casual chatting and deeper discussions
The most common of all encounters was security guards casual chatting and deeper discussions with youths, as is the intention of the work method (41%, N = 467, Table 1). The subcategories included chatting, greeting, discussing the common rules, substance use, talking about worrying issues, plans, delinquency, and giving and receiving positive feedback (see Table 1 for percentages). We argue that these all include caring for young people and it is notable that the security guards actively approached the young people with no intention to intervene. Procedural justice theory emphasises a respectful and friendly approach, listening and fair treatment as a means of enhancing trust (Tyler, 1990). Calmness and tone of voice is important to young people (Saarikkomäki, 2016). The reports indicated that all of these were present in the work of the youth-focused security guards. It appeared that the young people and security guards knew each other and could build a more caring relationship.
The security guards gave positive feedback to young people mostly when they respected the rules of the shopping centre. Interestingly, the young people usually gave positive feedback to the security guards about chatting, not being too strict, behaving calmly, in friendly way, or having a humorous approach towards them. Young people appreciated improved relations and mutual respect. Here then, the young people seemed to compliment the security guards for being "pedagogues" (Lampela & Tani, 2015); they also appreciated the security guards not always intervening in minor disturbances.
Talked with the gang of "regular older teens" and praised them for how these days they mostly behave in a smart and decent manner; the young people also gave us positive feedback. They think it's nice that we chat with them and don't always get angry about little things. We told them that they had deserved pleasant and proper behaviour from the security guards with their own behaviour and even though there were some things that needed pointing out sometimes we won't move them on immediately if they don't act like idiots. The youths told us they don't need to and that they don't want to irritate the security guards on purpose now that relations are good, and they are happy at the shopping centre now since we are on good terms. (. . .) (Shopping centre 1) In line with the "pedagogue" role or involving citizens in decision-making processes (Tyler, 1990), the security guards involved the young people in resolving the problem. It seems the security guards believe that discussing matters improves young people's understanding of their own behaviour and thus deters delinquency or for instance, collecting trash if they had been littering. The young people usually responded by understanding and even thanking the security guards rather than getting angry with an adult restricting their activities. Sometimes they did not respond well, but these situations were rare. The security guards believed situational crime prevention was ineffective: they suspected that a young person would only continue to do the same elsewhere or later. Based on the extracts, the understanding was that talking to young people, trying to make them understand the rules, would be more effective.
In addition, the security guards discussed young person's behaviour in situations they defined troublesome. The security guards expected the young people to be more capable of reflecting afterwards, or that youths who were familiar to the security guards would listen better. Interestingly, positive relations seemed to encourage the young people to take the initiative to talk about previous situations; for instance, if they were not sure about what had happened, or why or to encourage youth to report to security guards worrying situations.
A familiar young person informed me that a boy near them was under the influence of something. The boy was not familiar. (. . .) He started to cry almost straightaway, and it was obvious he was really upset. (. . .) He was really worried getting into more trouble for us intervening (. . .) We told him that was not our aim here and we just wanted to know how he was doing and we were really worried about him. The last extract contains many subcategories. The initial contact was due to another young person's concern about the substance use of their friend. The security guards did not use only situational tools, such as moving on, but aimed to find out the reasons behind drinking. Then, the contact moved to other issues and the boy felt safe enough to show emotion and to cry even though he said he had trust issues. The guard also talked to the young person for a relatively long time (one hour) and they offered themselves as a trustworthy adult. These examples show how security guards can have many roles or tasks at the same time: controlling and caring about the young people.

Caring and helping youths
About every tenth (12.7%, N = 145, Table 1) encounter consisted of security guards helping young people. These most distinctly express the youth work or caring for young people aspect of their work or what is defined as security guards' new "servicemen" orientation (Kim et al., 2018), which suggest that work roles are indeed blurred. In their training, the security guards gain knowledge about different situations in which young people might be in distress or need support from adults. Most frequent in this category were the encounters in which security guards were concerned about a young person (4.7%), worrying topics arose during other actions, or the security guards received information from store staff, customers or the young people themselves. For instance, in one example (from shopping centre 1) the security guards were called to check on a 11-year-old girl who had injured her back playing around in the children's playground. During the episode, the security guards called an ambulance and became concerned about the girl's mother being so indifferent even though the girl was truly in pain. In the end, they inform the child protection authorities.
The second most frequent way in which security guards helped young people, and got positive feedback, was concrete, opening storage or charger boxes. Not so frequent but important cases were encountering young people who were victims of bullying, resolving different conflicts between young people, and severe cases of self-destructive youths. In this category, the young people were victims and sought help if they were being harassed, bullied, assaulted, or had lost their friends.
During the conversation she told us she was being bullied at school. This was somewhat surprising because she has always been really cheerful. Although we had recently noticed she'd been bit down but we thought it was because of her relationship or something like that since she had been accompanied by a boy in these situations. (. . .) There was a lot of other youths around, so we didn't press the issue. If at some point later we meet her alone, we will return to the subject if the moment seems right. (Shopping centre 5) These conversations show mutual trust, and the security guards had to be sensitive to notice implications of bullying. Deviating from prior research of instrumental intervening (Saarikkomäki, 2018), the security guards aimed to assess whether a young person had a history of being bullied, trouble at home, depression or other background issues that could affect why they were acting in a way defined as troublesome or delinquent.

Cooperation
Prior research has mainly studied cooperation between security guards and the police (Michael, 1999;M. Nalla & Hummer, 1999;White & Gill, 2013;Søgaard et al., 2016;Meerts, 2020;Paek et al., 2020). However, we found that security guards cooperated not only with the police but also with different other actors (total 18,3%, N = 209, Table 1). Seven per cent of all the reports involved cooperation with the police (Table 1), typically in situations that involved suspicions of crime, such as shoplifting or fights, substance use or otherwise worrying behaviour or young people as crime victims. Security guards in Finland are obliged to inform the police if they apprehend a person. The data included situations in which youths were suspected of shoplifting and taken to the detention room to wait for the police. Getting caught might be a very frightening situation for young people, as is visible in the extract below. The security guards had to manage the girl's feelings and calm her down while they were waiting for the police to arrive.
A 16-year-old girl shoplifted [name of the store removed] several items. She was completely hysterical and crying all the time. I calmed the girl down and gave her tissues to blow her nose and a glass of water. (. . .) I had to be near her all the time because she said she was scared. The police arrived at the detention room to deal with the matter. The girl was also afraid she would be banned from the shopping centre, I told that she is welcome as long as she does not shoplift again. This she promised. (Shopping centre 4) In addition to this kind of "formal" interaction there was also "informal" exchange of knowledge about the young people and attempts to solve problems together with the police. Some situations indicated that the security guards knew the police well. In Finland, police departments in several regions have multi-professional teams consisting of professionals from the police, social services, health services, and youth services. These teams are called "anchors" and their aim is to intervene in criminal behaviour among minors at an early stage, assess a young person's life circumstances, and refer them for appropriate help and support services.
The police anchor team came to make a round [name of shopping centre removed]. We went around the shopping centre with them and encountered familiar youths. We talked with the anchor team about youths with whom we had common dealings. The anchor team gave positive feedback to us that our reports to them have been useful and we have managed to concretely improve youths' life situations through different support measures. (Shopping centre 5) Perhaps surprisingly, the security guards sometimes cooperated with the youths' parents when they were suspected of substance use, delinquency or when they were victims. Based on the extracts, the guards did not only detain young people and inform the police; they also asked why the person had broken the rules, in case background issues were cause for concern, and discussed these with the parents. It is also notable that the police sometimes contacted the security guards for cooperation, as is visible in following example. The boundaries between public and private policing can become blurred when security guards help the police, which can be interpreted as net-widening of policing (also Søgaard et al., 2016).
The young people had organized a back-to-school party on a nearby beach. We talked with several young people about alcohol use and told them off about making trouble. We detained seven youths for drinking alcohol on the request of the police and sorted out the matter with the parents and the police. (Shopping centre 2) The most common reported form of cooperation was with youth workers (9%), which is understandable as the goal of this new work method is to improve relations with young people. Usually, this cooperation was with municipal youth workers, but sometimes also with non-profit or other organisations. Cooperation with the Youth Service Association is not in the figures, as the reports were produced for them. Most of the reported cooperation was described by a brief mention that the security guards had informally visited youth centres to chat with young people, youth workers and drink coffee: "I went several times during the shift to a youth centre [name anonymised] to meet young people and to exchange information with youth workers." (Shopping centre 5). In addition, the youth workers of a nearby youth centre contacted the security guards because they were worried about a young person, they were trying to find a runaway youth, or the security guards contacted the youth workers if they needed to discuss something. The reports indicated multiprofessional cooperation, which is also a part of this work method, when the security guards had caught a young person and informed the police and youth workers. In a few situations, the youth workers contacted the security guards if they needed help with young people who were misbehaving at the youth centre.
In addition, the security guards cooperated with foster carers or others who informed them of runaway youths, or also directly with social workers. The "other" category (3%) involved situations such as calling an ambulance, contacting security guards from other shopping centres, cooperating closely with store staff (including the security guards of a particular store), or other types of networking meetings. The security guards used discretion to assess whether they would inform and cooperate with the police or not, so cooperation should not be over-emphasised, since all the cases did not come to the attention of the police (see also Meerts, 2020). In conclusion, these cooperation situations involved care and/or control. Based on our observations, the roles of the youth-focused security guards were diverse, and they crossed and blurred the boundaries of different sectors in their daily work, cooperating with the public, private and third sector.

Discussion
This study focused on relations between young people and security guards in shopping centres from security guards' perspective. We analysed these relations using unique data: self-reports from "youth-focused security guards" in Finland, who have received special training to work with young people. Our main findings indicate that these relations consist of more than mere control; they are also about caring. Our results differ significantly from prior research on young people's negative experiences with security guards (Fine et al., 2003;Lampela et al., 2016;Manzo, 2004;Matthews et al., 2000;Saarikkomäki, 2016Saarikkomäki, , 2018Saarikkomäki & Alvesalo-Kuusi, 2020). This suggests that cooperation and blurring the boundaries of the private, public, and third sector (nonprofit) offered interesting outcomes, and the aim of this article is to provide implications for procedural justice theory, for plural policing literature and for improving the relations between security guards and objects of policing in practice.
Security guards succeeded in creating trust while interacting with young people and getting to know them but also by intervening in a friendly manner. Perhaps surprisingly, the reports demonstrate that the young people talked about personal issues and showed their emotions, indicating trusting and familiar relations. It appears that the encounters were beneficial for both parties: the young people had more freedom in using the quasi-public space, were able to discuss various issues, and got help. The security guards, for their part, were able to create more positive relations, learn new procedures, and prevent situations from becoming difficult and coercive. Procedural justice research has addressed the importance of fair treatment and participation (Tyler, 1990;Tyler & Lind, 1992). However, the procedural justice literature has largely ignored nuances, such as the nature of interaction and intense emotions in control situations, as well as the pluralisation of policing: aspects we found important. Our data showed that these security guards were "pedagogues" rather than "disciplinarians" or "passers-by" (Lampela & Tani, 2015), who engage in discussions with young people, negotiate the rules, aim to understand young people, and use democratic power. However, we also found control aspects that restricted young people's space, access and activities.
Based on our findings, the new role and occupation culture incorporated control, care and cooperation. This study can provide new understandings of policing agents' work roles (e.g., Michael, 1999;M. Nalla & Hummer, 1999). Kim et al. (2018) have developed traditional roles of security guards "parapolice orientation" and "watchmen orientation" to include also "servicemen" orientation (see also Wakefield, 2003;Manzo, 2004). Our findings similarly indicated that along with traditional roles of controlling, using legal tools and "looking over", servicemen orientation was observable. Security guards participated in service work and helped in practical ways, such as giving directions, opening phone charger boxes, finding lost items, or incentivising young people for cleaning up their trash. Furthermore, our findings interestingly suggest that there was a fourth orientation: "social orientation". This meant that security guards were concerned with youth issues, aimed to help or to direct them to other professionals. Although we assume that this is largely because of their training and recruiting, it is not impossible that other security guards may include this type of social orientation. In fact, there are some indications, for instance, of nice security guards who came to chat with young people or offered help (Saarikkomäki, 2018, pp. 8-9). However, in this work format, the goal was to make it systematic and not dependant on personal characters. Similar to White and Gill (2013), our findings highlight that the plural policing landscape is even more blurred than previously understood. Youth-focused security guards are blurring the roles and rationalities of security work (private) and youth work (third sector, non-profit). In addition, they cooperate formally and informally on one hand with the police and other security guards (in stores and other shopping centres), and on the other hand, with social workers, youth workers and youths' parents. Plural policing literature discuss similarities and differences in public and private policing and suggest that the relationship and cooperation is complex and changing in nature (Michael, 1999;M. Nalla & Hummer, 1999;Hummer & Nalla, 2003;Button, 2007;White & Gill, 2013;Søgaard et al., 2016;Meerts, 2020;Nøkleberg, 2020). The positive sides of "boundary crossing" were highlighted, such as learning from other professions and viewing one's own work critically (Crawford & L'Hoiry, 2017).
Our findings suggest some critique of the trends of moving young people away from public and quasi-public spaces and viewing them as a risk, as well as of situational crime prevention and traditional security work (Gray & Gray, 1999;Von Hirsch & Shearing, 2000;Wakefield, 2003). Only moving young people on might cause them to continue doing the same thing in other locations or later. Furthermore, instrumental intervening and moving on can put youths in danger if they are moved on to unsafe spaces, if they are intoxicated, or if the deeper issues behind worrying behaviour are ignored. Our findings suggest it was more efficient harm prevention to talk about common rules, aiming to better resolve issues together and see the larger context of the potential causes of disturbances, for instance, difficult situations in family and at school. Prior research conducted in Finland has suggested that young people perceive police as more procedurally just than security guards, one reason being that the police were seen to tackle issues more deeply, whereas security guards were criticised for only moving people on from the area they patrol (Saarikkomäki, 2018). Some criminologists have critiqued ignoring structural contexts and issues in crime prevention (e.g., Garland, 2001). Understanding youth activities can be beneficial, and shopping centres vary in how they accept young people and how coercive their private policing practices are (Manzo, 2004;Kim et al., 2018).
Our study had limitations. The reports were guards' descriptions, and it is possible they highlight positive sides. However, the reports consisted of situations that had not gone well (e.g., using force), indicating that also negative aspects were reported. Some insights of this study, both practical and theoretical, can hopefully be transformed to other shopping centres and countries' policing contexts, however with caution. As we focused on the security guards who had received special training, there are limitations to generalising our findings to apply to other guards and contexts, for instance, in shopping centres where young people are not welcome (Manzo, 2004). To our knowledge, Citycon intends to spread the work and has local cooperation projects in other Nordic countries, however, this cooperation is not as coordinated and widespread as in Finland.
Boundary crossing and blurring of the boundaries can also have pitfalls. For instance, security guards may encounter difficulties when facing the diverse issues of young people and when balancing the role between control and care. The larger question is what the rise of the private security sector means: it can be a good thing for young people to have more caring adults around them, but it can also be negative if there is more control and restriction of youth access and activities. It is also viewed critically that these developments, increasing use of quasi-public space and pluralised policing, have led to city spaces becoming more intensively policed (e.g., Von Hirsch & Shearing, 2000;Wakefield, 2003). Ultimately control and cooperation with police and other agents always involves a risk of labelling youth; multi-professional networks and net-widening of policing should be therefore also examined critically (Garland, 2001). Our study demonstrates that it is crucial to understand youth issues more profoundly, both in practice and in academic discussions. Indeed, youth work and youth research aspects can provide new angles for criminology and security studies.