School fortresses : private security in two South African high schools

(1)

Master International Development Studies

School Fortresses

Private security in two South African high schools

Author: Kevin Boers, 10001716

Supervisor: Tessa G. Diphoorn Second reader: Christine Richter

Date: 28-11-2014

(2)

1

Abstract

Violence in schools can cause truancy, dropouts and psychological and physical problems among students and teachers. Many studies have shown how this leads to decreasing school performances and less educational opportunities that hamper the development of students. Fear of violence negatively affects the personal safety, well-being and development of persons. In South Africa, violence and crime are highly relevant social problems, where fear of crime is rife, primarily through the abundance of “talk of crime”. This fear has led to a securitization of society, where numerous measures are deployed for the safety of oneself and one‟s property. The most common security measure is the employment of private security for protection against the dangers of the outside world. School owners are increasingly installing more security and employing private security companies to protect their students and the school‟s property. Yet there is almost no scholarly research into the (private) security of the schools, and whether that leads to a safe learning environment, where students can safely learn without fear of violence. There is only a debate among American academia about private security in USA schools, but not in other countries. This research aims to fill this empirical gap by analyzing the security of two high schools in Durban, South Africa, to discover whether the private security has an effect on the safety of the learning environment. This research is conducted on two high schools in Durban, Eden College and Glenwood High school, where the perceptions of students, teachers and staff on safety, violence, crime and security are analyzed. From the results of both quantitative and qualitative methodology, I firstly conclude that there is a safe learning environment on both schools, as there is almost no fear of violence and crime. Secondly, I conclude that the employed security measures greatly contribute to the safe learning environment. In the view of the respondents, the security measures create a safe haven where they feel protected against the dangers of South African society. Thirdly, I conclude that students and teachers in these two schools do not fear the violence and crime in their school committed by other students and teachers. Instead, they fear the violence and crime committed by outsiders and they want to be protected against that. Fourthly, the security, violence and crime situation greatly differ in the two schools. There is more crime, violence and security on Glenwood than on Eden. Yet despite the differences in security, the feelings of safety do not differ greatly and this is caused by the fact that Eden has never experienced a serious criminal or violent incident. I therefore fifthly conclude that heavily securing the school might create a perceived safe learning environment, but it does not affect incidents of violence and crime inside the schools. Rather, it increases the fear of the outside world among the students and teachers.

(3)

2

Acknowledgements

I owe a lot to the guidance and contacts of my supervisor Tessa Diphoorn; without her contacts, I would not have been able to successfully start this thesis. One of her contacts, professor Monique Marks, brought me into contact with Chris Marcellin, principal of Eden College. Chris Marcellin was very important to my research; he allowed me to conduct my research in his school and aided me in finding respondents and setting up schedules for interviews. He also provided me with some very interesting insights and brought me into contact with various other principals, deputies and school inspectors. One of these was Kevin Jordan, deputy of Glenwood High, who graciously allowed to me to conduct a part of my research on Glenwood High. He assisted me in arranging interviews with students and teachers, showed me around the school and helped me with conducting the surveys. I would also like to thank my various housemates in South Africa, who made my stay very pleasant and taught me local knowledge, norms and values. Lastly, I would like to thank all my contacts, students, teachers and staff of Eden College and Glenwood High.

(4)

3

List of illustrations

Tables

Table 1 Respondents of semi structured interviews 23

Table 2 Experience with crime on Eden 29

Table 3 Categories of crime on Eden 29

Table 4 Amount of crime according to Eden students 30

Table 5 Experience with violence on Eden 32

Table 6 Categories of violence on Eden 32

Table 7 Amount of violence according to Eden students 32

Table 8 Security services Eden 34

Table 9 Trusting security companies Eden 36

Table 10 Victims of crime and violence in Eden 40

Table 11 Concerned for crime and violence in Eden 40

Table 12 Security make you feel safer Eden 40

Table 13 Experience with crime on Glenwood 45

Table 14 Categories of crime on Glenwood 45

Table 15 Amount of crime according to Glenwood students 46

Table 16 Experience with violence on Glenwood 48

Table 17 Categories of violence on Glenwood 48

Table 18 Amount of violence according to Glenwood students 48

Table 19 Security services Glenwood 50

Table 20 Trusting security companies Glenwood 52

Table 21 Victims of crime and violence Glenwood 54

Table 22 Concerned for crime and violence Glenwood 54

Table 23 Security make you feel safer Glenwood 55

Table 24 Concerned for crime and violence in the outside world 61 Table 25 Victims of crime and violence in the outside world 61 Maps

Map 1 Durban and Glenwood 9

Pictures

Picture 1 Entrance gate to Eden with guard house 37

Picture 2 Entrance to car park Eden College 38

Picture 3 Visitor‟s entrance to Glenwood High school 44

Picture 4 Cameras at Glenwood High 52

Picture 5 School entrance Glenwood High 53

Picture 6 Hallway fences Glenwood High 53

Graphs

Graph 1 Effectiveness of security according to Eden students 36 Graph 2 Effectiveness of security according to Glenwood students 51

(8)

7

Chapter 1 Introduction

“On April 12th

, 2014, four students were stabbed in a fight in a school bus in the province of Kwazulu Natal. Paramedics arrived outside the school and found four patients, aged between 16 and 21, sitting inside and outside the bus where the fight had apparently occurred. Two were stabbed in the chest and treated on the scene on Friday. A third had severe facial injuries, and the fourth had been stabbed in the arm. Paramedics were told that the incident occurred when approximately 12 males arrived at the school and started stabbing the learners on the bus” (www.news24.com/SouthAfrica/News/Pupils-stabbed-on-school-bus-in-KZN-20140412).

Violence in or around schools is probably one of the most important issues facing young people in South Africa today (Burton and Leoshut, 2013) and the above example is but one of the many violent incidents that occur in South Africa. Shootings, stabbings, and physical and emotional violence occur frequently in both public and private schools (Akiba, et al., 2002; Zulu, et al., 2004; as quoted in Mncube and Harber, 2012). In a study on school violence in South Africa conducted by Patrick Burton and Lezanne Leoschut (2013), 58.7 per cent of the students mentioned they have been a victim of violence, including bullying and crime. In another study on school violence, only 23 per cent of South African students said that they felt safe at school (Mncube and Harber, 2012).

Every school should aspire to be a safe learning environment (Bickel, 1994), in order to give their students the possibility to study without the risk of becoming a victim of violence. Numerous studies have demonstrated that violence in schools is detrimental to the welfare of students and teachers. For example, Burton and Leoshut (2013) show that violence in schools leads to physical or psychological harm of students or teachers. However, they highlight that the long-term effects of violence in schools are even more important (Burton and Leoshut, 2013). One of these long terms effect is that violence in schools can result in a lot of truancy or drop outs of students (Burton and Leoshut, 2013, Mncube and Harber, 2012), because the students become too scared to attend school (Burton and Leoshut, 2013). This hinders the educational opportunities (Burton and Leoshut, 2013) and academic performance of these students (Mncube and Harber, 2012). Another long-term effect is that victims of school violence often experience a decrease in school performance, as they have trouble focusing on their work due to fear, depression, loss of self-esteem, stress, or fatigue (Burton and Leoshut, 2013). This negatively affects their future school and work career (Burton and Leoshut, 2013). Another effect of violence in schools is that violence committed by older students or teachers can hurt the ability of victims to form healthy, social and trusting relationships with peers and adults (Burton and Leoshut, 2013). Furthermore, victims of violence at a young age often behave violently at an older age, continuing the cycle (Burton and Leoshut, 2013). Another study by Mncube and Harber (2012) shows that school violence can also lead to an increased risk of teenage pregnancy, the transmission of HIV/Aids amongst youth, and community disintegration (Mncube and Harber, 2012). Finally, violence in school can also lead to the loss of hope and optimism of victims in a better future (Burton and Leoshut, 2013). These studies highlight the social relevance of researching violence in schools.

All the effects of violence in schools that are outlined above are interrelated and influence the economic potential of a country. High rates of violence in schools have great effect on the whole country, as the future generation is negatively affected (Mncube and Harber, 2012). It will lead to less development, economic instability, increased crime, higher unemployment and even

(9)

8 an increased chance on civil war (Mncube and Harber, 2012). Consequently, making schools safe should be a high priority of any country. There are a few suggested ways for schools to become safe, including improving management, better teacher training in recognizing bullying and dealing with violence (Mncube and Harber, 2012) and employing a private security company to guard and monitor the school to prevent and deter violence (Maskaly, 2011). In this thesis, I will research how private security measures influence the creation of a safe learning environment, where students can follow their education without fear of violence.

I conducted this research in two high schools in Durban, South Africa. South Africa has one of the highest homicide rates in the world with 31 homicides per 100,000 populations in 2012 (UNODC, 2013). According to Tessa Diphoorn (2013), these rates of criminal violence cannot be separated from the immense growth of various private security initiatives, such as armed response units, private security guards, neighborhood watches and other citizen initiatives (Diphoorn, 2013). As argued by Altbeker (2007 as quoted in Diphoorn, 2013, p. 7), the high rates of crime in South Africa have led to the securitization of society, where fear and security are a huge part of everyday life. South Africa currently has the largest private security industry in the world (Abrahamsen and Williams, 2011) and the ratio of private security officers to police officers is 2.56 to 1 (www.psira.co.za, 2013). In South Africa, the Private Security Industry Regulatory Authority (PSIRA) regulates the private security sector. This regulatory framework determines a range of issues, such as the level of training of private security officers and the type of armament they are allowed to carry. According to PSIRA‟s most recent Annual Report of 2013, there are 9031 registered active security businesses in South Africa, with almost two million registered security officers.

In South Africa, Durban is one of the cities with the highest violent crime rates (Diphoorn, 2013). According to the website Crimestatsa (2014), there were 16,460 reported crimes in Durban Central alone in 2013, of which there were 39 murders and 2336 incidence of drugs related crimes. The Durban Metropolitan Area is the second largest in South Africa and is commonly regarded as the Indian city of South Africa (Diphoorn, 2013), with 22 per cent of the population having Indian/Asian roots (Marx and Charlton, 2003 as quoted in Diphoorn, 2013, p. 33). The other racial groups are Blacks with 63 per cent, whites with 11 per cent and colored with 3 per cent (Marx and Charlton, 2003 as quoted in Diphoorn, 2013, p. 33). In 2000, Durban was enlarged to include several peripheral areas and the name of the local government was changed to eThekwini Municipality (Marx and Charlton, 2003 as quoted in Diphoorn, 2013, p. 33).

According to the website of private schools in South Africa, namely SA Private Schools, there are 35 private schools in the eThekwini Municipality (saprivateschools, 2014) out of a total 577 schools (schools4sa, 2014). The two schools in this research are Eden College, a private high-end school with minimal private security company presence, and Glenwood High school, a public all-boys school with high private security company presence. These two schools are both situated in the neighborhood of Glenwood, one of the neighborhoods in Durban Central and one of the oldest suburbs of Durban (wheretostay.co.za/town/glenwood-durban). On map 1 you can see where the schools are located in Glenwood, Durban: the gold stars are the locations of the schools and my apartment 266 Bulwer road, which was right next to Glenwood High school. For more maps see appendix B. Glenwood falls under the Umbilo police station and of the 16,460 reported crimes in Durban Central (Crimestatsa, 2014), 6410 occurred in Umbilo police‟s jurisdiction. The statistics shown in appendix C give an overview of the crime in Glenwood and

(10)

9 its surrounding neighborhoods and these statistics make Glenwood seem like a rather dangerous part of Durban.

Source: Google Maps

Despite the large coverage of private security and the large amount of violent incidents in schools, no research has been done in South Africa about the role of private security in creating secure and safe schools. Most research regarding private security in schools comes from the USA, where authors like Maskaly (2011) and Leander (2011) have researched the effectiveness of private security in providing security on schools. Their studies show that private security is not that effective in reducing violence in schools (Maskaly et al, 2011, Leander et al, 2011). However, another study from Ben Brown and Reed Benedict (2005) in American schools shows that private security does positively influence the perceptions of safety among the students and teachers, even though private security has no real effect on reducing violence. In this study I aim to contribute to this debate, by researching private security in schools in a developing country and analyzing its effects on reducing violence in schools in a completely different context than the USA. Private security in schools is not only present in the USA, but in other countries as well, and this study will make the debate more international. This research therefore intends to address an empirical gap in literature, by studying the effect of private security on the safety of

(11)

10 the school in South Africa, in the perceptions of the students and teachers. By researching private security on schools in the South African context, I intend to contribute to the debate about private security in schools by discovering whether private security is effective in another context than the USA. I will also provide some criticism on creating a safe learning environment by heavily securing a school with private security. This will also allow me to create recommendations for further research in this field. My research question is therefore:

How do private security measures influence perceptions and incidents of violence on Eden College and Glenwood High school and how does this affect the safe learning environment and the development/education of the students in both schools?

In the following chapter (chapter 2), I will outline my theoretical framework. I will provide an explanation of private security and perceptions of fear and violence and will discuss freedom of fear, fear of violence and a safe learning environment. In chapter 3 I will present the research questions and discuss the research design and methodology. The two chapters after that will present and analyze the data gathered from the two schools, namely Eden College in chapter four and Glenwood High in chapter 5. In Chapter 6, I will make several conclusions, answer the research questions, engage in an empirical and methodological discussion, and make suggestions for future research.

(12)

11

Chapter 2 Theoretical Framework

In this research I analyze the presence of private security companies on high schools, in order to discover whether these companies create a learning environment that is perceived as safe, i.e. a safe learning environment. The link between private security on schools and development studies comes from the human security approach, more specifically, the freedom from fear branch, which states that everyone should be free from fear of violence for their wellbeing, personal safety, development and human fulfillment (Rothshild, 1995, as quoted in Alikire, 2003). I focus on the safety and security aspect of development and then specifically on the freedom of fear of violence aspect. In a safe learning environment students can follow their education and teachers can teach without fear of violence. My research will focus on how the presence of private security can influence the reduction of fear of violence in a school.

In the first section, I will discuss private security in schools. I will begin with a brief introduction of the private security industry, where I present a definition of private security, briefly describe the history of the industry, explain its current prominence, and discuss some criticisms. After that I will explain the concept of safe learning environment and will discuss the effects private security can have on the safety of the school. In the second section I will delve into the concept of freedom of fear of violence. As a concept, “freedom of fear” has existed since president Roosevelt‟s speech of the four freedoms (Alikire, 2003). Yet, I argue that this definition is state centered and I therefore prefer the concept within the human security approach, which is more people centered (Alikire, 2003). After explaining this concept, I will discuss the concept of „violence‟ by analyzing Galtung‟s (1961) approach and the World Health Organization‟s (Krug et al., 2002) definition. In the third section I will discuss fear of violence and crime, and I will use Caldeira‟s (2000) notion of the „talk of crime‟ and James Garofalo‟s (1981) four considerations and six responses when confronted with fear of crime. In the last section I will conclude this chapter and provide a brief summary.

2.1 Private security in schools

In this section I will discuss the private security industry, safe learning environment and private security‟s effect on the safety of schools.

2.1.1 Private security industry

In Strom et al (2010), a few different definitions for private security are provided. In this research I use the definition from Green (1981 as quoted in Strom et al, 2010, p. 14), who defines private security as “those individuals, organizations, and services other than public law enforcement agencies, which are engaged primarily in the prevention of crime, loss, or harm to specific individuals, organizations, or facilities” (Green, 1981 as quoted in Strom et al, 2010, p. 14). The individuals, organizations and services from Green‟s (1981 as quoted in Strom et al, 2010, p. 14) definition can be classified as the private security industry. The private security industry‟s task is to monitor, protect, prevent and deter. Private security provides their clients with a large variety of security measures, such as guards, alarm systems, fences, cameras, metal detectors and vaults (Strom et al, 2010).

Since the Second-World War the private security industry has increased in all layers of our society, including our schools and universities, in both the Western and the developing world

(13)

12 (Shearing, 1983). Until the 1980s the private security industry was only viewed as a supplement to the police and justice system (Shearing, 1983). However, since the late 80s, under the dominant neoliberal discourse (Holmqvist, 2005), private security is increasingly taking over some tasks from the police. This has a major impact on social control in society (Holmqvist, 2005). As in some cases, it is no longer the state, through the police, that is maintaining the law, but private security guards (Shearing, 1983). The monopoly of violence on which, according to Hobbes, the state authority is built, thereby erodes (Shearing, 1983). Other authors (Diphoorn, 2013) argue that the state has outsourced some of its authority to private security companies, but that this does not mean that the state‟s authority erodes, as the state still employs these companies to assert its authority. In addition to the concern of how the rise of private security may lead to the erosion of state authority, other concerns focus on the lack of accountability of private security officers, the profit-making motivations of these companies, and how such companies do not serve the general public (Akcinaroglu, 2013, Diphoorn, 2013).

Since the 1990s, the private security industry has expanded rapidly and Holmqvist (2005) has identified four global trends that have caused the rapid worldwide expansion of private security in the 1990s:

1. The new dominance of post-cold war neoliberal models of state, with strong incentives to privatize traditional governmental tasks, such as security.

2. The global downsizing of national militaries after the fall of Berlin‟s wall and the end of the Cold War. From 1987 until 1996 the number of soldiers in national armies fell with six million worldwide. This has provided a large man pool of potential recruits.

3. The gradual disengagement of the global powers from many parts of the developing world, letting the development countries fetch for themselves.

4. Private security providers are a logical progression of the privatization of armament production, as these companies often provide training of these weapons, maintenance services and even sometimes provide the people who operate the weapons.

South Africa is one of the most striking examples of this historic trend, as the private security industry has become the most prominent security provider in South Africa since the fall of the apartheid regime in the early 1990s (Abrahamsen and Williams, 2011). This makes South Africa a very relevant case study to analyze the role of private security at schools.

Although these four global trends are important, it is important to keep in mind that the privatization of security in each context is defined by local unique circumstances (Holmqvist, 2005). In some developing states, like South Africa, people turn to private security because they assume there are no alternatives. These people are often the local elite who place their trust in private security companies, rather than the state police (Holmqvist, 2005). In countries like the USA, with strong public security provision, private security often serves as a supplement and is highly regulated (Holmqvist, 2005).

To discuss the effect of private security in schools it is first important to explain the concept of a safe school, i.e. a safe learning environment.

2.1.2 Safe learning environment

I use the term safe learning environment to discuss a learning environment where the fear of violence is absent, a learning environment that is perceived as safe by students, teachers and

(14)

13 staff. It is therefore an environment with freedom from fear, where the personal safety and wellbeing of students is protected. This concerns the physical safety of the school (Clapper, 2010). However, there are other usages of the concept of safe learning environment, for example, the one discussed by Clapper (2010), who argues that the safe learning environment as a concept can also be analyzed from a psychological point of view. His idea of a safe learning environment is a place where students and teachers are not afraid to make mistakes, a place where they can be wrong without the fear of being mocked at or exposed as an imposter (2010). I will take this into account, especially considering bullying, but it is not my main focus, which is on physical violence.

According to Bickel (1994), creating a safe learning environment should be one of the main priorities of any school. However there is a debate among academics and courts about how far the school should be responsible for minimizing the risk of injury of students by third parties or by other students (Bickel, 1994). In some cases, judges have determined that schools are responsible for protecting their students against outsiders, yet are not liable for students attacking other students inside the school (Bickel, 1994). However, others argue that schools are responsible for both, according to Marilyn Campbell, Des Butler and Sally Kift (2008), common law states that a school authority “owes a duty of care towards its students” (Campbell et al, 2008, p.7). The duty is „non-delegable‟, meaning that the legal responsibility for creating a safe learning environment always remains with the school owners, even when the everyday responsibility for ensuring that the school is a safe environment is delegated to the principal of the school. As such, it will be the school owners that bear any legal liability when that duty is breached (Campbell et al, 2008). This duty of creating a safe learning environment includes protecting the students from the conduct of other students. The duty not only extends to physical harm but also to recognizable psychiatric issues (Campbell et al, 2008). One of the means in which school owners can create a safe environment is through hiring private security (Maskaly et al., 2011). The creation of a safe learning environment does not necessarily mean that all violent incidents should be removed or dealt with. Yet what is important is that students and teachers perceive their school environment as safe (Brown and Benedict, 2005). In this research I will analyze whether private security can help in achieving that. Unfortunately not much has been written about the security or safety of schools in the “developing world”, yet the debates originating from the USA articles are also applicable in other contexts and are therefore relevant to discuss here.

2.1.3 Private security in schools

In the USA a lot has been written about security and safety in schools and a lot of research has been conducted about various measures that the school can take to create a safe learning environment. In the USA, the presence of private security in schools has increased since the 1980s (Shearing, 1983), primarily through the provision of metal detectors, guards, fences, cameras and alarm systems (Strom et al, 2010). The intent of private security is to protect schools against attacks of third parties and decrease the violence inside the schools (Maskaly et al, 2011).

Since the various recent school massacres, the debate in the USA about the school‟s responsibility for creating a safe learning environment has surged again (Maskaly et al, 2011). Several authors claim that private security should be deployed at every school (Maskaly et al, 2011) to ensure that fear and violence is eliminated from schools. However, authors such as

(15)

14 Leander et al (2011) highlight that private security guards are not that effective in reducing violence at high schools in the USA as opposed to police forces, who are deemed more effective at decreasing violence in high schools (Leander et al, 2011). This might be the case in the USA, where police forces have a reasonable reputation and can command a certain respect (Maskaly et al, 2011), but in a setting where the state police has a rather negative reputation, like in South Africa (Diphoorn, 2013), the deployment of police officers in schools may not be a viable solution. In such cases, private security might be a more reasonable and practical solution to protect the school against violence and to create a safe learning environment. According to one study by Brown and Benedict (2005) in the USA, private security does seem to have a strong effect on how students, teachers and parents perceive the safety of their schools. People tend to view their schools safer with private security present, even though the actual effect on reducing violence is rather limited (Brown and Benedict, 2005). In this case of Brown and Benedict (2005), I argue there is also a safe learning environment, because the students perceive their school as safe, they have freedom of fear violence, regardless of whether violent incidents have actually decreased.

2.2 Freedom from fear of violence

Freedom from fear is part of the human security paradigm, emerged in the early 1990s as an alternative to the then dominant state centered security approach. Rather than focusing on the security of the state, the paradigm focused on the security of the individual and was therefore regarded as a people centered concept (Alikire, 2003). Human security was first defined and championed by Mahbub ul Haq in the 1994 UNDP report, where it is defined as: freedom from fear and freedom from want to protect the personal safety and well-being of the individual (Rothshild, 1995, as quoted in Alikire, 2003). Freedom from fear was understood as free from fear of violence, while freedom of want was understood as freedom from fear of poverty, yet both were viewed as essential for personal safety, well-being and the development of people (Alikire, 2003). From the UNDP definition of human security, two branches of human security have developed, namely the freedom of fear branch and the freedom of want branch (Alikire, 2003). The freedom of fear branch was based on the Canadian‟s government report on human security, which focused more on direct threats to personal safety and well-being (Bajpai, 2000). Direct threats include violent crime, child abuse, abuse of women, government repression, terrorism, genocide, transnational crime, worldwide drugs trafficking, drugs use, dehumanizing “bullying”, and the proliferation of small arms (Bajpai, 2000). The freedom from want branch focused more on the indirect threats to personal safety and well-being. Authors in support of the freedom of want branch, like King and Murray (2001), removed the threat of violence from the human security definition, and focused on the indirect threat of poverty for personal safety and well-being (King and Murray 2001), other indirect threats include lack of food, water, and primary health care, and unemployment (Bajpai, 2000).

This research will focus on the freedom of fear branch of human security, which focuses on the protection of people from direct violent threats to personal safety and well-being (Bajpai, 2000 and Alikire, 2003). I am primarily concerned with the direct violent threats to schools and the personal safety and well-being of students and teachers, and how private security can impact these threats. From the list of Bajpai (2003), I will primarily look at threats that are most likely to occur in schools, like violent crime, child abuse, abuse of women, proliferation of small arms, drugs use and dehumanizing bullying. Although direct violent threats like terrorism and genocide

(16)

15 can occur on schools, in the South African context these are not very relevant. I will not look at direct threats like government repression, as that cannot be influenced by private security.

In order to understand the freedom of fear of violence concept, one has to understand the concept of violence. To analyze violence I will use Johan Galtung‟s and the World Health Organization‟s definition. According to Galtung (1969), violence “is present when human beings are being influenced so that their actual somatic and mental realizations are below their potential realizations” (p. 168). Although this definition of violence might be a bit outdated, I believe it does capture the essence of violence: when one falls victim to violence, your potential realization might be endangered, because you are physically or mentally injured, lose property or are disadvantaged in relation to others. In his definition of violence, Galtung (1969) makes a distinction between direct and structural violence, which is similar to the direct and indirect threats as described by Bajpai (2000). Direct violence is personal and observable, such as the act of killing someone (Galtung, 1969) and also includes crime, such as burglaries. Structural violence is indirect and sanctioned by society as a whole, such as income inequality between classes or races (Galtung, 1969).

In my view, private security‟s main aim is to protect their clients from direct violence. As such, my main focus is on how private security influences direct violence in schools. However, I also recognize that private security can influence structural violence as well, both negatively and positively. An example of a negative effect on structural violence is that the presence of armed security guards might create an atmosphere of oppression where students cannot fulfill their potential realization, like in Colombia, where the private security guards that were hired against attacks on the universities were the previous attackers of the university (Novelli, 2014).

Having discussed Galtung, I would like to present another more practical definition of violence by the World Health Organization (WHO), which defines violence as “the intentional use of physical force or power, threatened or actual, (against oneself), another person, or against a group or community, that either results in or has a high likelihood of resulting in injury, death, psychological harm, maldevelopment or deprivation” (Krug et al., 2002). This definition is more practical than Galtung‟s, as it highlights the actual or threats of physical force or power and the effect that can have on victims. This WHO definition avoids crime and can also apply to non-criminal violence that is present at schools, like bullying. Many forms of bullying, which are common at school, do not constitute a crime but are violent in nature and result in substantial harm to the victims (Burton, 2013).

In this research I acknowledge that crime is part of Galtung‟s direct violence and I recognize that the terms crime and violence are often interchangeably used in theories of fear of crime or violence and as such I use them interchangeably as well in section 2.3. In fact I use the theory of Galtung to justify that crime and violence can be used interchangeably, as crime falls under his definition of direct violence (1964). However, I want to highlight that I differentiate between crime and violence in my data gathering and presentation. The reason for this is practical: separating crime and violence makes the questions clearer to students and teachers. Questions that solely asked about direct violence would have raised questions among the respondents. When I ask about crime, I specifically ask for activities which are not necessary violent but forbidden by law, like theft, underage drinking, drugs and smoking in school. When I ask about violence I use WHO‟s definition, use or threats of physical force which causes harm (Krug et al., 2002), like fights between students, attacks and bullying.

(17)

16 2.3 Fear of crime

To explain fear of violence, I will use theories about fear and fear of crime that can also apply to violence in general. After all, crime is according to Galtung (1969) part of direct violence. In order to understand freedom of fear of violence, one needs to understand fear. Fear in itself is often understood as an emotion (Diphoorn, 2013), yet it is constructed by society and its norms, structures and logics (Beck, 1992, in Diphoorn, 2013 and Tellegen and Wolsink, 1998).

According to Garofalo (1981), fear of crime is an emotional reaction characterized by a sense of danger and anxiety from the threat of physical harm. Diphoorn (2013) argues that fear of crime is created by society and often fueled by stories and experiences with crime. It is a collective fear, often experienced by the entire community, like in South Africa, where most people generally answer crime when asked what they are afraid of (Diphoorn, 2013). According to Garofalo (1981), the position in social space1 of a person is crucial for the amount and nature of information about crime to which a person is exposed. Garofalo (1981) argues that there are three sources of information about crime: direct experience (as a victim or a witness), interpersonal communication about the direct or indirect experience of others, and the mass media.

These three sources of information about crime are related to what Caldeira (2000) refers to as the „talk of crime‟. Victims and relatives of victims use this „talk of crime‟ to reorder the world they live in (Caldeira, 2000). Caldeira (2000) notes that experience with a violent incident can change your entire vision of society; the fear can stay with you for the rest of your life. As Susana Rotker and her co-authors argue (2002), logic and morality are turned upside down in the face of violence, and victims of violence anchor themselves in emotions and subjectivity, often by dehumanizing the offender. Caldeira (2000) contents with this, arguing that people who have experienced or know someone who has experienced crime have trouble understanding the situations they have to deal with and because of that they use simplifications of criminals such as „nordestino‟, „people from cortiços‟ or „favela‟s‟ in their „talk of crime‟ (Caldeira, 2000). These stereotypes rely on the creation of oppositional categories, the most important of which are good and evil (Caldeira, 2000). In South Africa, “bergies”, black people from Nigeria, and people from townships are generally viewed as “the other”, a potential criminal with evil intentions2

. The „talk of crime‟ therefore creates barriers, not only in discourse, but also in practice, like the walls around property, the suspicion and investigation at entrances and not admitting people belonging to a stereotype criminal group (Caldeira, 2000). In other words, the abandoning of rationality in favor of subjectivity and emotion after experiencing violence only feeds the fear and paranoia in society (Rotker et al, 2002).

Garofalo (1981) argues that when people perceive the risk of crime, they generally have four considerations. Individuals do not make these considerations in a clear and rational assessment of risk, but they are made unconsciously or indirectly. These considerations are:

1. Prevalence (amount of crime) 2. Likelihood of becoming a victim 3. Vulnerability

4. Capability of coping with potential loss as a result of an offense

1_{Social space is one‟s lifestyle and a person‟s past learning experiences and future life‟s chances (Garofalo, 1981).} 2_{From own personal experience and conversations with South Africans.}

(18)

17 An individual can consider the risk of crime low, even if crime is very prevalent, because the vulnerability or likelihood is considered low. For example, if a young, strong guy is walking around in a neighborhood with high crime rates, this person might consider the risk of crime low and feel safe, as his physical built may lead him to think that he is less likely to become a victim of crime. This might differ from an old woman walking around in the same neighborhood, she will consider the risk of crime high, because her physical built may lead her to think that she is very likely to become a victim. In addition to the four considerations, Garofalo (1981) also outlines six general responses that people can have due to fear of crime:

1. Avoidance, removing oneself from, or avoiding a, position or situation where risk of crime is perceived high, such as moving to a safer neighborhood.

2. Protective behavior, increasing your resistance against crime, such as hiring private security companies.

3. Insurance behavior, insuring your property to minimize costs of crime.

4. Communicative behavior, sharing emotions related to crime with others, feeding the „talk of crime‟.

5. Participation behavior, actions with others in society motivated by crime, such as a neighborhood watch.

6. Information gathering, actively gathering information from the media or other victims about crime and criminals, and looking around in the environment for cues that might indicate threats of crime.

These general responses will help to classify and analyze the behavior of the students, teachers and staff in the face of fear of crime and violence. The measures taken to protect oneself against crime and violence differ depending on the considerations and behavior of the individual.

According to Diphoorn (2013) and my own research data, South Africans have often experienced crime personally or know somebody who has experienced it. Consequently, South Africans often perceive their society as insecure and have a need to protect themselves against the dangers in their society (Diphoorn, 2013). As protective behavior, many South Africans turn to private security for protection against crime (Abrahamsen and Williams, 2011; Diphoorn, 2013). People feel a need to protect oneself and one‟s property against the dangers posed by “the criminal” (Caldeira, 2000). According to Lemanksi (2006 in Diphoorn, 2013), fear of crime is often about fear of becoming a victim, rather than fear of being a victim. As such, fear of crime is often associated with victimization prevention, rather than crime prevention. Consequently, employing private security companies is deemed essential to not become a victim of crime (Diphoorn, 2013). The fear of violence or crime is present in all of society, so also in the schools. Schools try to deal with fear of violence by creating spheres where students are free from fear of violence. The school‟s strategy will differ depending on the school owners‟ considerations and responses to fear of violence or crime.

2.4 Conclusion

In this chapter a theoretical justification is given for this research about violence and protection by private security in schools. The freedom of fear concept explains that it is important to protect people against violence, because the fear of violence negatively affects the personal safety, well-being and development of persons (Bajpai, 2000, Alikire, 2003). I use the term safe learning environment to describe a school where the wellbeing, personal safety and development of students and teachers are protected against the fear of violence and direct violent threats. In other

(19)

18 words, in a safe learning environment, there is freedom from fear: people are not afraid to be harmed by violence. Creating a safe learning environment is the responsibility of every school (Bickel, 1994), it is not only about protecting the school against outside threats but also includes protecting the students from the violence of other students (Campbell et al, 2008).

I primarily focus on direct violence, as described by Galtung (1964). In my questions I split direct violence in separated categories of crime and violence, to make it clearer to my respondents. The reason for primarily focusing on direct violence is that I focus on the role of private security in creating a safe learning environment and private security is primarily concerned with protecting people against direct violent threats.

The debate on private security in USA schools shows that private security‟s effect on reducing direct violence can be minimal (Leander et al., 2011), but that private security does have a great effect on the perceptions of safety (Brown and Benedict, 2005). In this study I will research whether that is also the case in South Africa, where the fear of violence is very prevalent, primarily due to the talk of crime (Caldeira, 2000). The behavior of the students, teachers and staff in the face of fear of crime and violence can be classified and analyzed using Garofalo‟s (1981) six behaviors and four considerations. Somebody‟s strategy will differ depending on one‟s considerations and responses to fear of violence or crime.

(20)

19

Chapter 3 Research Design

3.1 Research questions

In this research, I aim to explore how the students and teachers on two distinctive schools perceive violence and crime in their school, how they value the security against acts of violence and crime, and how this impacts the ability of students to follow their education. In order to explore these issues, I have formulated the following research question:

How does private security measures employed by the schools influence perceptions and incidents of violence on Eden College and Glenwood High school and how does this affect the safe

learning environment in both schools and the development/education of the students? In order to answer this research question, I have developed five sub questions:

1. How do students, teachers and staff members perceive the presence of private security in their schools?

2. How do the students, teachers and staff members of both schools perceive criminal activities and violent incidents in their school?

3. How do the students, teachers and staff members of both schools perceive the safety of their schools and the area in which it is located?

4. Does the security make the teachers, staff and students on the school feel safer and what are the reasons for this?

5. How do students, teachers and staff perceive that a safe learning environment is aiding them in their education/achieving their graduation?

(21)

20 3.2 Operationalization

Concept Dimensions Variables Indicators Security Protection Monitoring Prevention Deterrent Private security companies Guards Armed response Cameras Fences Metal detectors Barbed wire School system of security3 Gates Barred windows/gates High walls Violence in schools

Direct violence General acts of violence/disturbance

Fights (between students, teachers or parents) Bullying

Attacks by outsiders

Beatings by teachers or parents

Crime Theft

Underage drinking Drugs (use and dealing) Smoking Freedom from fear Perception of fear Perception of safety Fear in learning environment Safety in learning environment

Concern for violence and crime Feelings of safety in learning environment

Reasons for feeling (un)safe

3_{This is security that is already part of the building or premises and not built or provided by private}

(22)

21 Direct

violence in schools

Crime General acts of

violence/disturbance

3.3 Conceptual scheme

This conceptual scheme shows the aims of this research. I analyze direct violence, security and perceptions of safety and fear of violence in schools. Security in the school can protect against, prevent, deter, and monitor violence. In doing so, the security in schools can influence the perceptions of fear of violence in the schools, as well as the perceptions of safety in the schools. The security can create a learning environment without fear of violence, a safe learning environment, which will be positive for the learning achievements of students. There are other factors that contribute to the safety of a learning environment beside the security of the school, and if I can find these factors I will name them in my empirical chapter. However, these other factors are not my main focus.

Free of fear of violence in a safe learning environment Perceptions of safety in schools Perceptions of fear of violence in schools Positive for learning achievements of students

Security in

schools

Protect Monitor Deter Prevent

(23)

22 3.4 Units of analysis

My units of analysis are the students, teachers and staff of Eden College and Glenwood High school. My core units of analysis are the students and teachers, because I want to research how they perceive security, crime and violence in their schools. I have included members of the staff, by which I mean principals, administrator staff, janitors and receptionists, to discover the reasons for employing private security companies, what they think of the security of their school and to find their view on the safety of the school. I also interviewed some employees of a private security company to discover the daily routine of the guards and their experiences at the school, the guards themselves are not a unit of analysis. I have not included the students‟ parents as units of analysis in my research for two reasons. The first is that I am interested in the perceptions of security, violence and safety in the actual school and the parents are not present there. The second reason is that this would make my research population too large for the given research period. However, I do recognize that they are important players and that their opinions may have prompted the contracting of private security firms by the schools. For these reasons, I have interviewed a few of them, but I have used these interviews strictly to build on the context of the two schools. The (sub-)research questions are thus directed at the people who are actually present in the schools, namely teachers, students and staff members.

3.5 Methods and techniques

My research is inductive as I have not used an existing theory about private security on schools, nor will I form a hypothesis that I aim to test. Instead, I have examined both schools from a background of theories of private security, freedom of fear, fear of violence and crime, and the safe learning environment. I used a sort of grounded theory approach, a theory where data collection, analysis and eventual theory stand in close relationship with each other (Strauss and Corbin, 1998, as quoted in Bryman, 2008) and I analyze and generate my data at the same time. According to Bryman (2008), grounded theory is not necessarily a theory, but more of an approach to the generation of theory out of data.

For my research I have utilized both qualitative and quantitative research methods, a mixed method approach. I used quantitative methods in order to say something about the general opinions of students about security, violence and crime in their school. By using quantitative methods, such as a self-completing questionnaire, I can discover trends and opinions of a larger group of respondents and generalize my findings for the entire school. It can also allow me to compare the schools, if applied properly.

I used qualitative methods to find the underlying reasons for opinions on security and violence in the schools and to gather in-depth information about how students and teachers perceive fear and safety. I had to know their elaborate opinion to be able to say something about their perceptions of fear and safety. A qualitative method allowed me to answer the how and why questions, e.g. why do students have a concern for theft on the school? Or how do students perceive their safety? These questions are not answerable by a self-completing questionnaire. However, the semi-structured interviews are not suited to find correlations between variables, nor generalize findings, nor compare the schools adequately. As such, I used mixed methods; I can only answer my research questions by using both self-completion questionnaires and semi-structured interviews/ participant observations. A pure qualitative or quantitative approach would not provide me with the necessary information that I need to answer my research questions.

(24)

23 3.5.1 Qualitative methods

For this research I conducted 23 semi-structured interviews, 11 for Glenwood High school and 12 for Eden College. This numerical difference is due to practical reasons, as I was not able to interview as many teachers on Glenwood High as on Eden. The interviews on both schools were with teachers, staff en students. On Glenwood High I interviewed eight students, one staff member and two teachers. On Eden I was able to interview six students, one staff members and five teachers.

Table 1 Respondents of semi-structured interviews

School Students Teachers Staff

Eden College 6 5 1

Glenwood High 8 2 1

These semi-structured interviews were interviews where I asked the same open-ended questions to all respondents, I recorded all these interviews and transcribed them at a later date. I

developed an interview guide for these semi-structured interviews, derived from my operationalization, which covered all the main themes of my research. However, due to the nature of the semi-structured interviews, respondents provided a great deal of leeway in how to reply (Bryman, 2008, p. 438). Consequently, every interview was different.

The semi-structured interviews gave me in-depth knowledge of the underlying structures and perceptions of violence, crime and security. I was also able to use the interviews to develop an understanding on whether the learning environment is considered safe by the students and teachers and what effect security has on creating this environment. See Appendix D for the interview guide. In the text semi structured interviews are referred to as: (Interview roman number, date) e.g. (Interview XI, 30-7-14). In appendix A, a full list of respondents is presented, with corresponding roman numbers.

I also used two forms of observation, namely unstructured observation and participant observation. My unstructured observation entailed me walking around both schools and recording as much detail as possible of the behavior of participants (Bryman, 2008, p. 257), to get a general sense of the school and whether it is an unsafe environment or if it is perceived as such by my respondents in their daily work. I used minimal participant observation by accompanying the principal of Eden and the deputy of Glenwood on a walk through the school. This allowed me to observe first-hand how students react to their authority and presence, and analyze how they are regarded. However, as this was just three times on Glenwood and once on Eden, the information gathered here is rather limited, but valuable.

In addition to the semi-structured interviews, I also gathered a lot of information through informal conversations with housemates, the principal of Eden, a professor at DUT, the deputy of Glenwood and other local contacts. According to Bryman, (2008) conversation analysis is the analysis of talk as it occurs in interaction in naturally occurring situations (Bryman, 2008, p. 494). According to Burges (1984 in Bryman, 2008), conversations are very similar to open or unstructured interviews, an interview that starts with maybe just one question, but the respondent is then allowed to respond freely, the interviewer can then follow up on points that he views as appropriate or relevant. It is entirely up to the respondent and the interviewer which direction the interview takes (Bryman, 2008). I did not record these talks with a recording device, as I felt that this would have undermined the natural occurring situations. For example, the talks with my

(25)

24 housemates occurred on a daily basis and relied on mutual trust. In other cases, recording the conversation was simply impossible, like in the case of one of my conversations with the principal, which happened while he was driving me somewhere, the noise of the car would have made recording impossible. As such I have no detailed transcriptions of these talks, but I did write down as much as I could remember immediately after each conversation. Naturally this meant that some details were lost, however, by immediately taking notes after or during the conversations, I was able to gather a lot of information from these talks. Information from the open interviews/informal conversations is referred to as: (informal conversation roman number) e.g. (Informal conversation IV), see appendix A for a list of interview and conversation contacts.

3.5.2 Quantitative methods

For this research I created a survey for both schools to distribute among students, in order to discover their opinions about private security, feelings of safety, fear of violence and their perception on violence and crime. As I wanted to generalize my findings across the entire school, my sample needed to be representative and have as little bias as possible (Bryman, 2008, p. 168-169). As such, I intended to use a probability sample, specifically a simple random sample, where every student had the same chance of being included into the sample. In random sampling, the method to select respondents is entirely mechanical, thereby eliminating the opportunity for human bias. In my case this meant that the respondents of the survey were supposed to be randomly selected from the student records. Another feature of random sampling is that the process is not dependent on the respondent‟s availability or presence, instead the surveys are sent to the names from the random selection of the student list and as such respondents do not have to be near me to be included in the sample (Bryman, p. 172-173). If this had failed, I would have had to use a non-probability method of sampling, namely standing in a central place/class room and interview people that are present. This approach would have contained a lot of bias, as I probably would have selected the respondents based on their appearance, i.e. whether they looked friendly or approachable, and their proximity to me (Bryman, 2008, p. 168-169).

I based my survey on the themes and questions from my operationalization and created some questions on some of the answers from the already conducted semi-structured interviews, see appendix E for the complete survey. After I created the surveys I sent them to my contacts, Chris Marcellin, principal of Eden College and Kevin Jordan, Deputy of Glenwood High school, to acquire their permission to distribute the surveys among their students. After I was granted their permission, the two contacts also offered to help me distribute and collect the surveys. I informed them that the distribution had to be as random as possible, so I asked them to compile a list from the student records and select random names on that list and then give the survey to these selected students.

From what I gathered, my contact at Eden College handed the surveys out to random people from the student records, as I had requested. I distributed 30 surveys at Eden College with a student population of 300, which meant that the sampling fraction, the probability to be

included in the sample (Bryman, 2008, p 172) was

(26)

25 However there was a non-response4 in Eden and I only recovered 22 of my surveys. This means that the real sampling fraction is

n/N = 22/300 = 0.073, so 1 in 13.

This means that one out of every 13 students had the chance of being included in the survey. The deputy of Glenwood did something different, he handed the surveys out to random people from several of his classes of all different age groups. Unintentionally, this resulted in a form of stratified sampling (Bryman, 2008), as he selected students from three of his classes, which were of different age groups. As he limited it to his own classes, not all of the students of Glenwood High had an equal chance to be included in the survey. However, the eventual sample of students was distributed in the same way as the school population in terms of the unintentional stratifying criterion, age (Bryman, 2008 p 173). However, I only found out about his strategy after he handed the survey forms back to me, which was a week before I left. I was therefore not able to conduct another more random survey and decided to use the forms that had already been completed. I argue that I can still use the results, but that comparing the schools will be difficult, a good explanation and reflection will be required with each comparison. At Glenwood High, I distributed 50 surveys out of a student population of 1150, which means that the sampling fraction, the probability to be included in the sample is

n/N = 50/1150 = 0.4, so 1 in 25, where n is the sample and N is the population.

I was not able to gather more than 50 surveys on both schools due to time and financial constraints. However, due to the small size of Eden compared to Glenwood, I thought it wise to not make the sampling fraction of Eden larger than 1 in 10. This had the fortunate result of making the overall operation cheaper. I did not make the sampling fraction the same on both schools. If I had done so, I would have only had 12 students complete the survey at Eden, and according to Bryman (2008), the relative size of the sample does not matter as much as the absolute size of the sample. I thought a survey with just 12 people would not have any meaningful results, so I choose to do 30 instead.

3.6 Data Analysis

As I have used both quantitative and qualitative research methods, my data analysis is both quantitative and qualitative. For the quantitative research data, I used descriptive statistics (Bryman, 2008) to describe the differences of the private security coverage/availability on both schools, the perceptions of security, violence and crime among the students and teachers, and the actual rates of acts of violence and crime. I also used correlation analysis (Bryman, 2008) to find out correlations between different variables, for example, if the rates of violence are related to the perceptions of fear or if the coverage/availability of private security correlates with the perceptions of security and rates of violent acts and crime. I used the program SPSS to do these analyses.

For my qualitative data analysis, I primarily used thematic analysis (Bryman, 2008) that is based on the topic list I created for my semi-structured interviews. This thematic analysis allowed me to discover certain important themes in the answers of the different respondents and

(27)

26 allowed me to compare them with each other. I also applied the thematic analysis to the

observations that I made. Through my thematic analysis, I tried to discover certain themes or customs that occurred on the schools and then analyzed whether there are differences in habits or behavior of both students and teachers. However, as the interviews are semi-structured, which provides room for other questions and leeway in the answers; I also used open coding (Bryman, 2008). I have found some themes in these answers as well, but I have not actively steered the conversations to discover certain themes outside my interview guide. I transcribed all of my interviews, some in the field, and others after my return to the Netherlands. To analyze the data from the qualitative method, I have used the software Atlas-ti. With this program, I was able to assign codes, discover themes and make connections between the interviews.

3.7 Epistemological/Ontological discussion

According to some authors, quantitative research is carried from a positivist epistemological stance and from an objectivist ontological stance, while qualitative research is carried out from an interpretative epistemological stance and from a constructionism ontological stance (Bryman, 2008, Smith 1983). Smith (1983) argues therefore that combining qualitative and quantitative research in one mixed method research is undesirable, as that means you are combining two irreconcilable views on how social reality works and should be studied. I do not agree with this, rather I concur with Bryman (2008) that research methods do not inherently accompany with them fixed epistemological or ontological stances. As my research has a mixed method design, the research does not have a fixed epistemological or ontological stance. Instead, this research has a bit of both sides. It has both elements of an interpretative epistemological stance and a positivist epistemological stance. I do not believe reality is solely formed on laws of causation, as the positivists argue, nor do I believe reality is only formed by every individual‟s social construction, as the interpretativists argue. Rather, I think I can say that I am a realist. According to David Abbott (2010), the debate between interpretative and positivist is rather old, instead most social scientists are now realists. As David Abbott (2010) explains it, “realists acknowledge that scientific methods are not foolproof and agree that humans are reflective. They argue that sociologists can be pragmatic and use whatever methods are appropriate for particular

circumstances. Social reality is complex and to study it, sociologists can draw on both positivist and interpretivist methods” (Abbott, 2010). I want my respondents to construct their perceptions of violence, crime and security on their school, but at the same time find the correlations between certain variables which also influence their reality, such as the relation between age and the feelings of safety, or the relation between the safety of their own home environment and their view on the safety of the school environment. As such I want to use both interpretative and positivist methods to answer my research question.

3.8 Ethics

One of the most important ethical considerations when conducting research is to do “no harm” (Goodhand, 2000). Harm can, according to Bryman (2008), be classified as physical harm, which refers to the harm caused to participants‟ development, loss of self-esteem, and stress. During my research, I tried to prevent causing any harm to the people I interviewed. I therefore offered every respondent the option to be anonymous and out of all my interviews, only one respondent requested anonymity. Some of the questions that I asked in the semi-structured interviews could

School fortresses : private security in two South African high schools