Big Brother wants to secure you: a
camera surveillance effect study in
Amsterdam shopping areas
Final thesis
Crisis and Security Management
Name:
Abel Mensink
Student ID:
s1756346
Supervisor: Dr. M.A.J. Ezinga
Second Reader: Dr. M.C.A. Liem
Date:
13 March 2016
Word count: 23.598
2
“Who controls the past, controls the future…
… Who controls the present, controls the past”
3
Abstract
Prior research has shown that the effect of camera surveillance projects on crime rates is ambiguous and dependent on specific features of the project. An important condition for the effectiveness of camera surveillance projects, in both an objective and a subjective manner, is that users of the public space are aware of the presence of cameras. Therefore it can be determined whether this awareness leads to an improved sense of security among visitors of the public space. Research is executed in two shopping areas in Amsterdam, the Netherlands. A survey was conducted with a total of 184 respondents. By means of multiple frequency analyses and a t-test, insights were analysed related to the sense of security of respondents perceiving camera surveillance to be present and the sense of security of those aware of the absence of cameras or lacking knowledge in this regard. The results of the analyses showed that visitors of the two shopping areas generally feel safe, are poorly aware of the presence of surveillance cameras and have a positive stance regarding the execution of camera surveillance. Finally, the perceived presence of cameras has no significant effect on the sense of security.
4 Table of Contents ABSTRACT 3 TABLE OF CONTENTS 4 1. INTRODUCTION 6 1.1 Research objective. 7
1.2 Setting of the research. 9
1.2.1 Policy framework 10
1.2.2 Legal framework 11
1.3 Research outline. 12
2. BODY OF KNOWLEDGE 13
2.1 The occurrence of criminal events. 13
CCTV, target suitability and behavioural adaptation 15
2.1.2 Conceptualization of the sense of security 17
2.2 Towards camera surveillance in public space. 19 2.3 Closed Circuit Television and Crime. 22
2.3.1 Effect of CCTV in international studies 23
2.3.2 Effect of CCTV in Dutch studies 25
2.3.3 Evaluation of camera project ‘Buikslotermeerplein’ 28
2.4 Broader perspective on the effect of CCTV. 29
2.4.1 CCTV and displacement of crime or diffusion of benefits 30
2.4.2 CCTV and privacy issues 31
3. METHODOLOGY 34
3.1 Data collection. 34
3.1.1 Sample selection 35
3.1.2 Location selection 37
3.1.3 Content of the questionnaire 40
3.1.4 Description of the variables 40
3.2 Data analysis. 42
3.2.1 The sense of security 42
3.2.2 The awareness on camera presence 43
3.2.3 The effect of perceived camera presence on the sense of security 43
5
4. RESULTS 47
4.1 The sense of security. 47
4.2 Awareness on camera presence. 48 4.3 The effect of perceived presence of cameras on the sense of security. 49 4.4 Routine activity patterns and the public stance regarding camera surveillance. 52
5. CONCLUSION 56
6. DISCUSSION 60
6.1 Positioning of the results in the current body of knowledge. 60
6.2 Data. 63 6.2.1 Sample selection 63 6.2.2 Locations 64 6.2.3 Questionnaire 64 Analyses 65 6.3 Recommendations. 66
6.4.1 Recommendations for future research 66
6.4.2 Recommendations for the municipality of Amsterdam 67
7. BIBLIOGRAPHY 69
APPENDIX I: QUESTIONNAIRE (IN DUTCH) 73
APPENDIX II: DO-FILE STATASE 14 77
6
1. Introduction
Camera surveillance is, in recent decades, a popular measure to secure the public space. This trend coincides with the technological developments improving the applicability of cameras. Examples include license plate registration, bodycams carried by police officers, speed checks, or static cameras monitoring public space (Flight, 2016a). The idea behind camera surveillance is that it can influence social behaviour in a positive manner (Welsh & Farrington, 2009; Van Rompay, Vonk & Fransen, 2009).Oppositely, the influence of camera surveillance is also feared: in the novel 1984, George Orwell presents his ultimate and futuristic example of the effect of camera surveillance, and introduced the idea of Big Brother. Represented through telephoto lenses in each and every habitat of people’s lives, the government was controlling its citizens (Flight, 2016a). As Flight (2016a) argues, cameras are currently meant to have the same correcting effect as Big Brother in George Orwell’s book, although “correcting” in this sense is more related to countering criminal behaviour than controlling individual citizens.
Camera surveillance projects are used to reduce crime (objective safety) and enhance the sense of security (subjective safety) (Gill & Spriggs, 2005). The extensive growth of the use of surveillance cameras has resulted in national and international debates on its effectiveness in countering crime and improving the sense of security (Gill & Spriggs, 2005; Wittebrood & Van Beem, 2004; Welsh & Farrington, 2002; 2009). Several studies have been executed to determine how successful cameras are in achieving crime rate reduction, of which the results are mixed. An acclaimed evaluation of surveillance projects in the United Kingdom and the United States showed mixed outcomes (Welsh & Farrington, 2002; 2009). Given the period camera surveillance projects are executed, one would expect that the body of knowledge regarding the subjective effectiveness is abundant. Remarkably, sound research regarding the effect of surveillance cameras on the sense of security remains limited. Likewise the objective effect studies, research regarding the subjective effect of camera surveillance provides in contradictory outcomes. Gill and Spriggs (2005) argue that the topic related to the sense of security among citizens is still largely under researched. From a social perspective, more research is required to improve policy decision making and secure public values (Welsh & Farrington, 2009; Piza, 2016).
A possible explanation for the ambiguous findings might be related to familiarization with cameras in public space (Flight, 2016a). The use of communal cameras in the Netherlands,
7
for example, emerged from 1995 onwards. The cameras are placed mainly in shopping centres, entertainment and nightlife areas, but also in residential areas. Flight (2016a) argues that camera surveillance has become such a common feature in the past 20 years, especially in urban areas, that cameras are hardly noticed in people’s everyday life. This observation is striking, particularly because awareness on the presence of surveillance cameras is believed to be necessary to comply with the purpose of its use. For effective camera surveillance two conditions are necessary: 1) the public needs to be aware of the presence of cameras and 2) has to feel saver due to this knowledge (Van Eijk, Kanning, Molenaar, Strijbos & Bernasco, 2006). According to this line of argument, it can be stated that awareness on the presence of cameras is crucial.
Next to concerns related to the effectiveness of camera surveillance, the costs to society should be taken into consideration. It is argued that camera surveillance is an expensive measure paid of public funds (Flight, 2016b). The expenditures include purchase costs, maintenance costs, and costs associated with the storage of camera footage. Beside financial costs, costs in terms of immaterial factors such as privacy should be taken into account (Flight, 2016a; Van Eijk et al., 2006). Even though the potential disadvantages are known, the implementation of projects is still rising. For example in the Netherlands in 2014, an estimated 3.300 communal cameras were in use (Bicknese, 2014). Only in 2015, 40 new communal projects were launched, adding an unknown amount of cameras to the existing 3.300. In the first three quarters of 2016, 27 new projects were launched (Flight, 2016a). Amsterdam is not the leading municipality in the Netherlands in terms of the number of surveillance cameras (Bicknese, 2014). However, it is among the few cities in the Netherlands presenting methodologically correct effect evaluation studies and camera surveillance projects appeared to have a positive effect on crime (Schreijenberg & Homburg, 2010). Additionally, over the past years, the Amsterdam police force has been subject of a number of tests with new camera surveillance techniques (Flight, 2016b). This accumulated communal experience with camera surveillance projects and the standardized evaluation schemes provide in a suitable environment for additional research.
1.1 Research objective.
According to the Amsterdam communal administration, camera surveillance contributes to an improved public safety. Extra eyes on the street increase the effectiveness of enforcement
8
authorities and help to prevent disruptions in public space (Gemeente Amsterdam Bestuursdienst, 2012). The communal administration developed a framework, of which the latest version was presented in 2012, to explain and legitimize the utility of this measure within its city borders. Important aspects to legitimize camera surveillance in Amsterdam are proportionality and subsidiarity. Camera surveillance can only be used when there are serious security problems (proportionality) and those problems cannot be countered with less profound measures (subsidiarity). Therefore, the framework prescribes an evaluation of the on-going projects every two years (Gemeente Amsterdam Bestuursdienst, 2012). The so called effect evaluations entail two domains: a study on the objective changes in crime rates and a study on changes in the perception of crime. The Amsterdam municipality uses a before-after comparison of data from visitors of the area where camera surveillance is executed. Unfortunately, the precise method of the evaluation is not publicly available, making it impossible to tightly control the outcomes of the research. For example, one of the oldest camera surveillance projects executed within Amsterdam is the one at ‘Buikslotermeerplein’. Here, cameras are installed in shopping area ‘Boven ‘t Y’ to increase safety in general at the square and to prevent violent raids and youth nuisance (Homburg & Bleeker, 2015a). The latest evaluation of this camera project revealed that it remains necessary to label the shopping centre as ‘unsafe’ based on the crime rates, thus the project is believed to be proportional. However, the extent to which visitors of the shopping area feel safer as a result of the surveillance cameras or whether other features are at stake remains unclear from the data presented by Homburg and Bleeker (2015a). For example, it is imaginable that the moment of a visit to shopping area ‘Boven ‘t Y’ is partly determined by demographic features like age and gender, making these features of importance in relation to safety feelings of visitors. This conception is derived from the rational choice theory by Cohen and Felson (1979). In the ‘Buikslotermeerplein’ camera surveillance evaluation studies, the effect of routine activity patterns is not included when the public stance regarding safety feelings was determined. Although it is not likely that evaluation studies are manipulated, the outcomes serve as a base for policy decisions. And when measures tend to be costly, in material and immaterial terms, transparency in the decision making process is desirable. If it is not possible to supply in transparency due to operational constraints, it leaves room for independent examination. One of the manners to evaluate camera surveillance projects is to compare empirical research findings in an area where camera surveillance is executed with data gathered in a similar area where surveillance cameras are absent. Shopping area ‘Oostpoort’ in the eastern district of Amsterdam shares multiple
9
characteristics with shopping area ‘Boven ‘t Y’, and is suitable for a comparison due to the absence of surveillance cameras Therefore, the following research questions and sub-questions are investigated in this research:
1. What is the sense of security among visitors of Amsterdam shopping areas ‘Boven ‘t
Y’ and ‘Oostpoort’ and to what extent is this influenced by the perceived presence of surveillance cameras monitoring public space?
1.1. How do visitors of the Amsterdam shopping areas ‘Boven ‘t Y’ and ‘Oostpoort’
asses their sense of security?
1.2. To what extend are visitors of shopping areas ‘Boven ‘t Y’ and ‘Oostpoort’ aware
of the presence of cameras monitoring the public space?
1.3. Is the perceived presence of surveillance cameras leading to a greater sense of
security among visitors of shopping areas ‘Boven ‘t Y’ and ‘Oostpoort’?
2. To what extent can the sense of security among visitors of Amsterdam shopping areas
‘Boven ‘t Y’ and ‘Oostpoort’ be explained by their routine activity patterns?
2.1 What is the public stance in shopping areas ‘Boven ‘t Y’ and ‘Oostpoort’
regarding surveillance cameras monitoring public space?
In doing so, it can be determined whether the policy decision to execute camera surveillance in shopping area ‘Boven ‘t Y’ complies with the requirements of proportionality and subsidiarity. The focus of the sub-questions is on the presumption that the two conditions for effective camera surveillance – awareness on the presence of cameras and improved feelings of safety due to this awareness – are met in the Amsterdam context.
1.2 Setting of the research.
This research is executed within the city of Amsterdam. Therefore, attention is drawn on the local context of camera surveillance projects. First, the policy framework will be further explained, followed by the legal framework.
10
1.2.1 Policy framework
The first policy framework of the municipality of Amsterdam regarding camera surveillance within the city was drafted in 2003. Due to developments in law and technology and the centralization of the surveillance centre in Amsterdam (the ‘Centrale Cameratoezichtruimte’, a place where all the camera footage from municipal cameras is monitored), a new framework was presented in 2012 (Gemeente Amsterdam, 2012). The policy framework is meant to provide a procedure for the execution of camera surveillance in public areas and describes the conditions that have to be met. Moreover, the policy framework entails a guideline for the evaluation phase of camera surveillance projects. According to the framework, it can be stated that camera surveillance projects in Amsterdam contribute to the fostering of security. Three reasons are mentioned: camera surveillance helps to increase the effectiveness of law enforcement authorities, it prevents disorder in public space, and it increases the sense of security amongst citizens (Gemeente Amsterdam Bestuursdienst, 2012). Due to both material and immaterial costs, the Amsterdam communal administration is restricted by regulations before it is allowed to execute Closed Circuit Television (hereafter: CCTV) surveillance. Important aspects mentioned previously in this chapter are 1) proportionality and 2) subsidiarity: 1) the measure can only be used when there are serious security problems and 2) those problems cannot be countered with less profound measures (Gemeente Amsterdam Bestuursdienst, 2012). CCTV is considered to be a last-resource measure. This implies that governmental authorities, business owners and the public must have undertaken prior actions to diminish the risk of disorder and ensure prevention of crime as much as possible (Gemeente Amsterdam Bestuursdienst, 2012). Prior to the implementation of a CCTV system, the local administration has to define the problem of disorder in an analysis. From the problem analysis it must be evident that camera surveillance is a justified and necessary strategy to be followed to minimize public disorder. Furthermore, in the problem analysis, goals have to be drafted applicable to the situation, camera positions have to be determined and, optionally, the time that active monitoring is executed needs to be determined. These conditions are eventually included in a ‘Plan of Action’ (in Dutch: Plan Van Aanpak), initiated by one of the communal districts and drafted in consultation with local police authorities. The Plan of Action eventually serves as a basis for the decision on the execution of a camera surveillance project.
11
Next to the regulations related to implementing CCTV, the existing camera projects within the municipality of Amsterdam have to be evaluated regularly and simultaneously along the same methodological standards. These regulations are developed by and under supervision of the central communal administration of Amsterdam (Gemeente Amsterdam, 2012). Unfortunately, the Plan of Actions for the different camera surveillance projects are not publicly accessible due to sensible police data included in the analysis (G.H.P. Bibbe, personal communication, 24 October 2016). However, summaries of the evaluations are published, enabling the possibility to filter the results of the evaluation of camera surveillance projects.
1.2.2 Legal framework
The city of Amsterdam has around 200 municipal cameras which are monitoring the public space at various locations throughout the city. Those cameras are legally attributed to Article 151C of the Municipal Act, which is in force since 2006 (Gemeente Amsterdam, 2015). The article states: “the council empowers the mayor to regulate, if it is at the best interest to
maintain public order, the execution of fixed cameras for a specified duration for the purpose of monitoring public places as defined in Article 1 of the Act on public events” (Gemeente
Amsterdam, 2012). Based on this article, a mayor can decide to implement camera surveillance in a public area, if that is believed to be necessary to maintain public order. This implies that the municipal council only needs to be informed in case of the deployment decisions. Furthermore, a law was passed recently in the Netherlands which allows municipal authorities to use flexible cameras. Until the spring of 2016, the communal jurisdiction was limited to the installation of static (attached to walls or posts) closed circuit cameras (Gemeente Amsterdam, 2012). The cameras were allowed to be relocated when crime patterns required so, but until recently they had to be installed statically. Since the first of July 2016 municipal authorities are allowed to use flexible cameras. The rationale behind this law was that crime displacement can be countered much faster when it requires little time to meet the standard procedures (Rijksoverheid, 2016). The limitations of camera surveillance in public area are, since 2008, laid down in the Dutch law on police data (Wet PG, 2008). As previously mentioned, it refers to places which can be accessed freely by the public. Some areas are defined as public, but are privately owned. A good example is train stations. If a municipality is willing to execute camera surveillance at train stations, they have to coincide
12
with the owner of the area, which will be the Dutch railway company NS in the example of the train station. However, these companies are legally responsible for safety trough other laws and regulations, thus camera surveillance is often already executed in those areas. Camera surveillance projects are further limited by a footage retention period of four weeks. Amsterdam brought this limit even further down to seven days, since it appears from practice that this is proportionate. Nearly all crime declarations are done within seven days, allowing the police to secure the footage. Longer storage would cost extra money, and as said, appears to be unnecessary (Gemeente Amsterdam, 2012).
1.3 Research outline.
To provide in the necessary background information on the constructs used in this research, the second chapter starts off with an explanation of the occurrence of criminal events and the concept of sense of security. Subsequently, attention is drawn on the emergence of camera surveillance in public space, followed by an explorative analysis on the objective and subjective effect of camera surveillance. Three contexts are discussed: the international, the national (Dutch) and the local (Amsterdam). The chapter concludes with attention for the broader perspective of the effect of camera surveillance. The research methodology will entail a quantitative research. The method involves two steps: data collection and data analysis. Additional information on these steps is supplied in chapter 3. The results of the study are presented in chapter 4. The research questions will be answered in a concluding chapter. Finally, this study will end with a discussion, starting with a reflection on the information supplied in the body of knowledge and how this research can be positioned in the current academic field. Furthermore, remarks and pitfalls are discussed and suggestions for improvement are made up in this section.
13
2. Body of Knowledge
When evaluating crime reducing policy initiatives, it is essential to be consistent in both definition and measurement (Hale, 1996). Thus, when camera surveillance is expected to influence the occurrence of criminal events and the sense of security of visitors in an area where it is executed, questions regarding both constructs arise. Why do criminal events occur? How is sense of security defined? And what does it entail? Insights from criminological theory related to the first two questions are supplied in paragraph 2.1. Thereafter, attention is drawn on the emergence of camera surveillance projects. This occurred along multiple steps of development, which are further explored in paragraph 2.2. Subsequently, camera surveillance projects are believed to have an impact on the objective and subjective safety. Paragraph 2.3 supplies in an overview of national as well as international studies regarding the magnitude of these two effects of CCTV cameras in public space. Finally, apart from the objective and subjective effects of camera surveillance in a restricted area, its implementation has consequences on additional features. Therefore, displacement, diffusion of benefits and privacy issues are explored and discussed in paragraph 2.4.
2.1 The occurrence of criminal events .
A leading theory for the occurrence of criminal events is stating that offences are the result of
“1) motivated offenders, 2) suitable targets, and 3) the absence of capable guardians against a violation” (Cohen & Felson, 1979: p. 589). A sociological paradox formed the motivation
for development of this theory: in the United States, social and economic conditions improved during the 1960s, while, at the same time, crime rates increased. Cohen and Felson (1979) relate this paradox to changes in routine activities, which caused more opportunities for crime. For example, the relation between women emancipation and morning crime rates. The increased women participation in the labour force resulted in new routine activities. Consequently, the amount of unattended households at 8 A.M. significantly increased in American neighbourhoods, providing opportunities for burglary. This is in line with the underlying assumption of the theory by Cohen and Felson (1979) that crime is not randomly distributed over time and space, but is linked to the “places where offenders, targets and
weak guardians converge” (Sherman, Gartin & Buerger, 1989: p. 31). In an extensive spatial
analysis, Sherman et al. (1989) found support for the theory that crime is non-randomly distributed, through a research in which they studied the standards of offenders, targets and
14
(absence of) capable guardians. They found evidence that a small number of places have a large share of crimes reported to the police. By analysing calls for police service, they found out that just over 50% of all the calls in a single year could be attributed to 3.3% of all addresses and intersections in the city of Minneapolis. Apart from the place specific character of crime, Cohen and Felson (1979) argued that many technological developments designed for legitimate purposes may both motivate offenders to engage in criminal events and enable people to successfully act as a guardian, thereby preventing crime. Think of cars, small power tools, mobile telephones or ATMs, on the one hand all ensuring new trends in criminal offences and on the other hand improving guardianship. For example, cars enable offenders to travel to suitable areas for committing a crime, however, cars also provide in a safer commodity for people compared to public transport. For mobile telephones, the example is even more evident: mobile telephones enable people to call for service wherever they are, however, the value of telephones may motivate offenders to engage in a criminal event. Some technological developments, such as security gates in shops and surveillance cameras are designed to increase security and prevent crime and, ostensibly, did not result in new trends in criminal offences. However, whether these developments are able to produce a net reduction in crime rates is, according to Sherman et al. (1989), difficult to determine. Main argument is that is appears to be impossible to account for changes in the amount of motivated offenders, or even determining who these people are. To enhance the criminology of places, focus should be on controlled, longitudinal experiments (Sherman et al., 1989). Thus, the routine activity theory implies that a successful violation requires an offender and a suitable target in absence of capable guardians. It is emphasized that a decline in one of the first two elements or an increase in the third is enough to prevent a criminal act from occurring (Cohen & Felson, 1979). Additionally, more time and place specific experiments are necessary to supply in empirical evidence for this theory (Sherman et al., 1989). Apart from the lack of empirical evidence by the time Sherman et al., (1989) presented their insights, another limitation of the routine activity theory is that it applies better to some offences than others. As argued by Sherman et al. (1989) one can prevent robbery twice as effective by avoiding a certain area than one can prevent sexual violation or car theft. Apparent from these insights is the importance of opportunity, which is described as a ‘root cause’ of crime. There is no such thing as an individual condition explaining the cause for a person committing a crime; additionally, opportunity is the necessary factor for the occurrence of criminological events (Felson & Clarke, 1998). In complementing the routine
15
activity approach, Felson and Clarke (1998) describe how crime can be prevented by reducing opportunities. Back in 1979, crime studies were often focussed on the offenders’ perspective, with the search for conditions explaining the cause for a person to commit a crime. Although the opportunity theory also leaves implications for the offenders’ perspective, additional attention is drawn on the perspectives of (possible) targets and guardians (Felson & Clarke, 1998). For example, camera surveillance is mentioned as a measure to protect possible targets and can be seen as an additional guardian at a location. Therefore, camera surveillance increases the perceived effort of crime through control of access to a target (both person and object) and increases the perceived risk of crime. In other words, it reduces opportunities for offenders, which will be further explained in paragraph 2.3.
First, routine activities itself are discussed. Cohen and Felson (1979: p. 593) define routine activities as “any recurrent and prevalent activities which provide for basic population and
individual needs, whatever their biological or cultural origins.”. As argued by the authors,
these activities may go beyond the basic necessities for a population to prevent extinction as long as the activity is part of everyday life. Examples of routine activities are work, provisioning of food and shelter, commuting and, more generally, social interactions (Cohen & Felson, 1979). These activities may occur in roughly three places: at home, in schools or jobs away from home and finally, in other activities away from home. In post-World War II America, a growing trend of activities is noted outside homes, resulting in an increased probability of motivated offenders converging in time and space with suitable targets while guardians are absent (Cohen & Felson, 1979). This implies that targets are increasingly ‘exposed’ due to changing routine activities, resulting in higher risk of victimization in non-household activities. The theory was confirmed by victimization surveys in the 1970s (Cohen & Felson, 1979).
CCTV, target suitability and behavioural adaptation
Now it is observed that targets are increasingly exposed due to changing routine activities, the question remains in which manner cameras influence the suitability of targets. According to Felson and Clarke (1998), surveillance cameras protect visitors of the public space. However, abundant empirical evidence for this statement is, up to present, lacking. If cameras are not noticed by visitors of the public space, it is unlikely that they change behaviour. Those aware of the cameras, for example offenders, might work their way around it, with
16
which the protecting effect of cameras disappears. Additionally, when looking at personal victimization characteristics, Cohen and Felson (1979) supply evidence proving that age is inversely related to victimization rates. As the age of a person increases, the chance of victimization decreases. This is an important notion to be kept in mind when discussing the results of this study, but it is not related to the execution of camera surveillance. Nevertheless, when the insights of cameras protecting targets and of age being inversely related to victimization rates are combined, it might supply in useful input for analysis.
Two effects are conceivable when cameras are present and noticed: 1) it may induce visitors of the public space to take additional precautionary security measures due to a fear of crime or 2) it may result in a reluctance to take additional measures due to feelings of safety (Zehnder, 2009). Both effects assume a certain level of behavioural adaptation to camera surveillance areas. From the former effect, a statement can be derived that cameras indeed help to protect possible targets. However, the latter reveals a potential problem: when visitors of the public space feel safer due to camera presence, they might actually be more vulnerable for victimization. As Flight (2016) argues, empirical studies regarding these behavioural adaptations show a high level of ambiguity in the outcomes. The used research methodologies are mentioned as a cause of this ambiguity; answers on a questionnaire, for example, often differ from the actual behaviour of a respondent. To study the real effect of camera surveillance on the behaviour of those exposed to it, additional research methods are required (Flight, 2016). Mazerolle, Hurley and Chamlin (2002) did so through an extensive observational study on adaptive behaviour in CCTV areas. Instead of asking visitors of the public space about their behaviour through questionnaires, they observed behaviour of people in three different public areas through hidden cameras: a downtown market, a residential shopping centre and a park surrounded by affordable flats. For the analysis, a distinction between pro- and anti-social behaviour was made. Pro-social behaviour included amongst others riding bikes, greeting and conversing with other individuals, playing games and waiting for buses. Loitering, begging, drinking alcoholic beverages and drug-related activity are amongst behaviours identified as anti-social (Mazerolle et al., 2002). The study revealed that location specific differences are applicable when camera surveillance is executed. In the downtown market area, anti-social behaviour increased whereas people in the park showed an increase in pro-social behaviour. In the residential shopping area, no signs of changing behaviour were observed (Mazerolle et al., 2002). Most appealing conclusion of the study was that any adaptation of behaviour due to camera surveillance lasted for a maximum period
17
of two months, after which the behaviour of people was found to be similar to the situation before camera surveillance was executed (Mazerolle et al., 2002). Regardless the fact that this conclusion does not supply in specific information on the effect of cameras on target suitability, it does explain that, under certain circumstances, visitors of the public space show adaptive behaviour as a result of cameras. When referring back to the notion that age is inversely related to victimization rates, it is possible that elderly adapt their routine activity behaviour more often than youngsters when confronted with cameras in public space. At least, it is worth to explore this observation.
To sum up, changing routine activities resulted in an increased convergence of offenders and targets in the absence of guardians. Multiple studies have been executed studying the personal factors of offenders in relation to crime, however, opportunity is believed to be the necessary factor for the occurrence of criminal events (Felson & Clarke, 1998). The extent to which cameras influence target suitability remains unclear, although adaptive behaviour as a result of surveillance cameras has been found under specific conditions. However, for solid conclusions, additional empirical evidence is required. For a better understanding of the occurrence of crime and the effect of camera surveillance, it is important to take further notion of the personal factors of potential targets. In the following paragraph this is done through a conceptualization of the sense of security among visitors of the public space.
2.1.2 Conceptualization of the sense of security
In literature, camera surveillance and the impact of CCTV projects on visitors of the public space are often related to the narrow description of sense of security, being ‘fear of crime’. It is narrow in the sense that many factors can be identified that have an influence on someone’s feeling of security, of which the majority is not related to crime (Gill & Spriggs, 2005). In a broader description, one could for example point to the feeling of being safe and additionally, refer to worries about becoming a victim of a criminal act. However, being safe means different things to different people. Furthermore, worries on becoming a victim of criminal act vary to a large extent amongst people, making a broader description sense of security not perfect either (Gill & Spriggs, 2005). Nevertheless, it is chosen to use the broader description for this research, to prevent that possibly important information is missed.
As Box, Hale and Andrews (1988: p. 341) argue, there are six factors that contribute to the sense of security: 1) vulnerability, 2) the state of the environment, 3) knowledge on crime and
18
victimization, 4) confidence in the investigation and justice system, 5) the perception of personal risk of getting victimized, and 6) the assessment of impact from various offences. In this framework of factors, vulnerability refers to the assessment a person makes on the physical and emotional recovery time in case of victimization. When the outcome of the assessment is regarded as ‘long’, the person in question is regarded ‘vulnerable’ (Box et al., 1988). Other scientific insights regarding vulnerability relate to demographic characteristics. As explained in the study by Pantazis (2000), the idea in science prevails that women and elderly generally have a larger fear of crime because they perceive the consequences of crime more serious. This leaves interesting implications for further analysis in this study. According to Box et al. (1988), the state of the environment is of importance since it can be perceived as threatening; examples include loiterers, litter, vacancy and graffiti. According to Box et al. (1988), these are factors contributing to a fear of crime. Regarding knowledge on crime and victimization, it is conceivable that victims of crime tend to be more afraid of crime than non-victims. Additionally, if police efforts to apprehend offenders and respond to calls for crime are perceived to be effective, and offenders are punished through a trustworthy system, fear of crime tends to decrease (Box et al., 1988). About the perception of personal risk and the assessment of impact from offences, the authors declare: “Only where citizens feel they are
highly at risk of being victimised seriously is it likely that fear of crime would exist. Thus even where the risks of victimization were thought to be high, this would not generate fear if that particular offence were evaluated as being trivial” (Warr & Stafford, as cited in Box et al.,
1988: p. 342). Given the information on the different factors contributing to a fear for crime, the potential contributing factor of camera surveillance becomes evident. Even the perceived presence of camera surveillance can influence one’s experience regarding the sense of security in public space (see paragraph 2.1.1). Cameras can have an influence on the features vulnerability, confidence in investigation and justice systems and the perception of personal risk, and should therefore be taken into account. From these insights, the following hypothesis can be derived:
Hypothesis 1 (H1): the perceived presence of surveillance cameras leads to a greater sense of
security..
This hypothesis will be taken into account when data from the empirical research is analysed. However, it should be considered that scientific evidence for an effect of camera surveillance on the sense of security is ambiguous. In an extensive research on the subjective effect of surveillance cameras in Glasgow, Ditton (2000) comes up with the general conclusion that
19
CCTV did not improve but worsened the sense of security, at least amongst visitors of the city centre. Additionally, it turned out that crime rates in Glasgow’s city centre increased after implementing the CCTV project (Ditton, 2000). And as previously noted in paragraph 2.1.1, the effect of camera surveillance implementation is found to last for a maximum period of two months (Mazerolle et al., 2002). Thus, as mainly apparent from the study by Ditton (2000), there is no correlation between the sense of security and the absolute feature of actually being safe and secure.
To conclude, when analysing the effect of camera surveillance on crime rates and the sense of security of visitors of the public space, it is important to take notion of the reason behind the occurrence of criminal events. As argued by Cohen and Felson (1979), economic development has been a root cause for increased activities away from home. This resulted in an increase in opportunity for crime, as offenders and targets converged more often in time and space in absence of capable guardians (Cohen & Felson, 1979). Based on the theory, it is expected that gender and age specific differences in sense of security affect routine activity patterns. Likewise, cameras monitoring public space possibly affect the sense of security and routine activity patterns. Therefore, hypothesis 1 has been drafted. Since the perceived presence of surveillance cameras might influence the sense of security of visitors of the public space, this condition is found to be necessary to include in the hypothesis. Now that the occurrence of criminal events is embedded in this body of knowledge, the focus will shift towards the emergence of camera surveillance in public space.
2.2 Towards camera surveillance in public space.
The emergence of camera surveillance in public space dates back to the late 1950’s, when police forces in the United Kingdom started using cameras to operate traffic lights (Williams, 2003). In the following decades, the rise in camera surveillance was largely driven by the commercial availability of cameras itself; however, this was mainly in semi-private spheres, where marketing altered the idea that CCTV deterred shoplifters. It lasted until the mid-70s when cameras in public space slightly emerged, mainly to monitor traffic flows in London’s city centre (Norris, McCahill & Wood, 2002).
The tremendous growth in public space camera surveillance came during the 1990’s, when an increase in recorded crime raised public anxiety, resulting in demanding pressure on the
20
central government to come up with a solution. In this decade, eventually, over three-quarters of the UK Government’s budget on crime prevention was spent on public space CCTV projects (Welsh & Farrington as cited in Norris et al., 2002). Although the United Kingdom appears to stand on itself in this regard, public camera surveillance projects rose in all developed countries from the 90’s onwards. Although initially mainly adapted by the private sector, the 9/11 terrorist attacks resulted for example in a growth of America’s public camera surveillance industry (Norris et al., 2002). On the European mainland, figures tend to vary amongst countries. France, Spain and the Netherlands appear to have followed the British trend. Due to concerns on increasing crime rates and general security concerns in cities, suburbs and nightlife areas, observation of public space by cameras emerged from 1994 onwards in these countries (Hempel & Töpfer, 2002). Other European countries appeared to be restrained with the usage of public space camera surveillance. In 2004, a study recognized 40.000 cameras monitoring public area in the UK, whereas Norway, Denmark, Hungry, Germany and Austria together accounted for fewer than 1000 (European Union as cited in Norris et al., 2002).
In general, it can be stated that the surveillance through cameras has expanded from the private domain into the public sphere. As Norris et al. (2002: p. 118) state:
Thus, Norris et al. (2002) argue that the role of governments is to reduce opportunities for crime and misbehaviour and to manage risks. In this light, camera surveillance appears to be a useful measure. In assuming that the presence of cameras indeed reduces opportunities for crime and is useful to manage risks, governments acted to have found the silver bullet for countering public anxiety. Great events, such as the 9/11 terror attacks in New York and the 2005 London bombings, ensured the emergence of camera surveillance schemes, complete in
“Increasing urbanisation has exacerbated the trend towards anonymity, leading to
concerns over establishing and verifying identity. Increasing mobility, both locally and internationally, have given rise to a global ‘stranger society’, where social control and governance based on intimacy and face-to-face knowledge are increasingly less viable. Risk management has also become the dominant mode of reasoning for both international corporations and governments alike. In the realm of criminal justice, reformist ideals have given way to more modest preventative responses that focus on ‘opportunity reduction’, ‘situational prevention’ and ‘risk management’, …”
21
line with Kingdon’s agenda setting theory, where events create a ‘window of opportunity’ for policy development (Kingdon, 2003).
Norris et al. (2002) further argue that the extension of camera surveillance occurs along four stages. The first stage is remarked by diffusion in the private sector. The second stage is characterized by diffusion in the institutional public domain (for example schools, transportation systems and government buildings). The third stage shows an increase in the larger public area, where camera surveillance is introduced as a crime reducing measure. The camera projects themselves are small scale and focused on local problems. The fourth stage describes the creation of an overwhelming system in which multiple systems, public and private, are integrated and which is able to cover large urban areas for a variety of purposes (Norris et al., 2002). Examples of these fourth stage camera surveillance systems are license plate and facial recognition systems, or smart cameras which are able to monitor suspicious behaviour. Most of the current camera surveillance projects executed by local government authorities in the Netherlands are situated in the third stage. The description of the four stages might imply that the development of camera surveillances projects evolves along a standardized procedure, in which in each stage projects are evaluated and analysed, and when the project is found to be effective, an authority can decide to develop from one stage to the other. However, Norris et al. (2002) conclude that the extent to which camera surveillance projects do what they intend to do, being deterring and preventing crime, is questionable as well as irrelevant. Since psychological, social and political values and conditions are at stake when dealing with crime, the problem to solve it becomes complex. Norris et al. (2002) even argued that the execution of camera surveillance will never be the silver bullet in preventing the occurrence of crime. From a political perspective, there is often need to just be doing something (Norris et al., 2002).
To sum up, risk management efforts of governments together with the commercial availability of surveillance cameras are noted as causes for the emergence of camera surveillance in public space. Increased crime rates and major events further nourished the emergence of camera surveillance projects over the last decades. The extent to which camera surveillance projects succeed in managing risks or, in other words, deter crime, remains unclear. Therefore, the next subject that has to be addressed is the extent to which surveillance cameras are of influence on crime.
22
2.3 Closed Circuit Television and Crime .
As Ditton (2000) explains, two main fields of interest need to be taken into account when determining the influence of camera surveillance on crime: an effect on the objective safety and an effect on the subjective safety in the area where camera surveillance is executed. The objective safety can be measured by studying the crime rates in a specific area, given the assumption that CCTV diminishes the occurrence of criminal activities based on the rational choice theory of offenders. The idea of this theory is that crimes are committed on the premises of accessibility, vulnerability and attractiveness; the decision to commit an offense is then “negatively related to the perceived costs of crime and positively related to the
perceived rewards of crime” (Nagin & Paternoster, 1993: p. 479). In other words, offenders
consider for example the chance of getting caught as a ‘cost’, and when the total costs are higher than the expected revenues it is unlikely that they proceed to the commission of a crime. Moreover, as Bernasco, Elffers and Bruinsma (2006) argue, other targets in the immediate vicinity should be less attractive in terms of this cost-benefit analysis. In this light, the introduction of camera surveillance ensures a higher risk for the offender to get caught. This means a perceived rise of costs in the target area, resulting in an assumed decrease of committed crimes. ‘Assumed’ in the sense that the revenues of the crime might still be high enough to proceed with the criminal activity, or relocation to another area is still less favourable for the offender. In these occasions, crime rates remain equal (Bernasco et al., 2006). The level of subjective safety can be studied by determining how the sense of security amongst visitors of the public space changes when they are exposed to camera surveillance. One could speak of increased subjective safety when the sense of security improved after the introduction of camera surveillance in a specific area, independently from a potential improvement in terms of the objective safety (Bernasco et al., 2006). Similar to this statement, the sense of security might also improve when there is a perceived presence of camera surveillance.
What appears from these introductory remarks is that information on the presence of camera surveillance is crucial: camera surveillance affects the crime rates and the sense of security only when users of the public space, including both offenders and targets, are aware of the presence of cameras (Van Eijk et al., 2006). The question remains how effective camera surveillance projects so far have proven to be. In the following paragraphs, the results of studies on the effect of camera surveillance are discussed, both in the national and international context.
23
2.3.1 Effect of CCTV in international studies
Multiple evaluative studies have been executed on the subject of camera surveillance as a deterrent and preventive strategy against criminal offences. Welsh and Farrington (2002; 2009) supplied in an extensive part of the body of knowledge with their two meta-analytic studies. The main objectives of their studies were twofold: 1) gaining knowledge and providing evidence on the effects of camera surveillance on crime, and 2) gaining information on the preventive utility of camera surveillance in relation to criminal offences. There were four criteria for including studies in their analysis. In a nutshell: CCTV had to be the main focus of intervention, crime was to be the outcome measure, quality of the used methodology was high (i.e. both experiment and control areas needed to be present) and the number of crimes should exceed 20 prior to the intervention with cameras (Welsh & Farrington, 2009). It is important to bear in mind that these methodological criteria are strict, and valuable insights from studies that do not meet these criteria might be lost in this process. However, strict criteria like these enable generalizable conclusions which can be used to better interpret the usefulness of camera surveillance.
In 2002, the authors found 22 camera surveillance evaluation studies, in both the United Kingdom and the United States, which met the strict criteria for analysis. Only half of the studies found a positive effect of camera surveillance on crime rates. Moreover, five studies found a negative objective effect, five studies recorded no change in crime rates and in one case the effect was classified as uncertain (Welsh & Farrington, 2002). Backed by their meta-analysis, Welsh and Farrington (2002) concluded that the effect of camera surveillance on crime rates was small but positive; overall, crime rates went down by 4%. This effect appears to be low compared to some studies excluded from the analysis. In these studies, major diminishing effects on crime rates were reported (Tilley; Chatterton & Frenz; Davidson & Farr, as cited in Welsh & Farrington, 2002: p. 10). However, the extent to which these effects can be attributed to the presence of surveillance cameras remains unclear, due to the inferior methodological standards of these studies compared to the methodology used by Welsh & Farrington (2002; 2009). Additionally, it turned out that camera surveillance has a different effect when looking more specifically, for example at the setting in which it is executed. When looking at studies where CCTV was used in a city centre setting, the effect was rather negligible, whereas evidence was found that camera surveillance used in car parks resulted in a reduction of over 40% in crime rates (Welsh & Farrington, 2002). In 2009, an additional 22 studies were incorporated in a new meta-analysis. The conclusions of this new analysis were
24
slightly different than the ones in 2002. The study, which included more data, revealed that 37% of the studies found a positive objective effect, 56% found no significant effect and 7% of the studies found a negative objective effect of camera surveillance on crime (Welsh & Farrington, 2009). Overall, crime rates went down by 16% in the studies that were subject of the meta-analysis. Remarkably, the majority of the studies in this meta-analysis were carried out in the United Kingdom (83%). When comparing the effect of cameras on crime per country, the United Kingdom showed an overall positive effect of 19%, which is different from the other studies, executed in the United States, Canada, and Scandinavia. Here, no desirable effect was found. It is evident from paragraph 2.2 that the United Kingdom is a frontrunner in the execution of camera surveillance projects. Their extensive experience with these projects might be of importance on the reported effects, although it is difficult to precisely determine the reason of success in specifically this country. As Welsh and Farrington (2009) argue, the success in the British evaluation studies was largely driven by results of camera surveillance in car parks. This is an argument for the importance of context in which camera surveillance projects are executed. An ostensible important factor not addressed in the meta-analysis studies by Welsh and Farrington (2002; 2009) is the difference between actively and passively monitored camera surveillance. Passive camera surveillance involves examination of recorded footage; active camera surveillance entails real time monitoring through security operators (Hollis-Peel, Reynald, Van Bavel, Elffers, & Welsh, 2011). As determined by Hollis-Peel et al. (2011), over three quarter of the studies included in the 2009 meta-analysis of Welsh and Farrington used actively monitored surveillance cameras. When addressing the difference in effect on crime rates between actively and passively monitored CCTV, Hollis-Peel et al. (2011) concluded that actively monitored camera surveillance is favourable. However, there is still no extensive proof of the advantages of actively monitored camera surveillance over its passive counterpart. Recently, Piza (2016) argued that camera surveillance works best in combination with other, evidence-based strategies. This statement is derived from a study where actively monitored surveillance was introduced together with directed police patrols, resulting in diminishing crime rates in the target area (Piza, 2016). Again these findings are debatable, given the outcomes of another recent study regarding the effects of actively monitored camera surveillance. In a quasi-experimental case-study on the situation in Malmö, Sweden, there was no observable effect after similar interventions (Gerell, 2016).
25
With regard to the impact of camera surveillance on the fear of crime, Gill and Spriggs (2005) discuss 12 studies in which a before-after comparison was made. To be more specific on the setting, nine studies were executed in residential areas, three in city centre areas. In one case, cameras were installed when the evaluation phase already started, making it unable to make statements about a before-after comparison. Remarkably, in all the areas which were subject of analysis, worries about being a victim of crime decreased after cameras were implemented. Regarding the setting of the evaluations, no differences were shown between residential areas and city centres: in both areas worries about being a victim of crime reduced. However, in only two studies, the difference with the control area was significant. In those two studies, the average reduction was 8.5% (Gill & Spriggs, 2005). One of the hypotheses in the study of Gill and Spriggs (2005) is that a real effect of camera surveillance can only be taken into account if the public is aware of this measure, as explained in the introduction of this paragraph. Therefore, they analysed whether differences in fear of crime exist between those aware of surveillance cameras and those who are not. Interestingly, it turned out that respondents who were aware of the cameras showed a larger fear of crime than those who were not (the two studies which were meeting evaluation requirements showed an increase of 19% and 10% respectively). Although the evidence from this study is not extensive, it appears that the perceived presence of cameras increases the fear of crime, due to the observation that the public assumes an area where CCTV is necessary as problematic (Gill & Spriggs, 2005). In addition to fear of crime, camera surveillance might have an influence on feelings of safety. The same 12 studies revealed that camera surveillance improved the feelings of safety in nearly all the areas where an evaluation was executed; in only one study there was no effect. However, as previously noted in paragraph 2.1.2, being safe means different things to different people, and it sometimes has nothing to do with crime. Additionally, none of the studies showed a significant difference between the feeling of safety in the experiment area and the control area. Finally, no difference was found in feelings of safety in varying settings: the experience was similar in both residential areas and town centres (Gill & Spriggs, 2005).
2.3.2 Effect of CCTV in Dutch studies
In line with the international trend, multiple evaluation studies have been executed in the Netherlands regarding the effect of camera surveillance on crime. An important meta-analysis
26
on effect studies in The Netherlands was executed by Wittebrood and Van Beem (2004). Apart from camera surveillance effect studies, the evaluation included studies on additional measures to increase public safety, for example increased police surveillance and adjustment of street lighting. In 2004, they found 11 effect studies related to camera surveillance, which were executed in multiple settings. Three of the studies were executed in residential areas, four in public transport and the final four in entertainment and nightlife areas. Of these studies, only two were conducted in an appropriate methodological fashion. Appropriate in the sense that they had a quasi-experimental design with the same strict requirements as presented by Welsh & Farrington (2002). This enables statements about the effectiveness of camera surveillance. One of the appropriate studies was executed in a residential area, where the objective safety increased, and especially the execution of vandalism dropped as a result of camera surveillance (Wittebrood & Van Beem, 2004: p. 48). The other study which complied with the strict requirements was executed in public transport. Here, the objective safety slightly improved. Although none of the studies in entertainment and nightlife areas complied with the standards for conducting a meta-analysis, the effect studies provided interesting insights. In these areas, camera surveillance appears to improve the objective safety. Statements should be made carefully, since there are usually multiple measures simultaneously introduced in these areas. What can be stated is that the amount of violations registered by the police increased in camera surveillance areas. This can mainly be attributed to the actively monitored character of camera surveillance (Wittebrood & Van Beem, 2004: p. 52). Hence, this might again be a reason to favour actively monitored CCTV above its passive counterpart.
Apart from the evaluation by Wittebrood and Van Beem (2004), which studied general measures to increase public safety, there is only one meta-analysis executed with the specific subject of the effectiveness of camera surveillance. In 2010, Schreijenberg and Homburg presented a study in which they found 51 effect studies regarding camera surveillance, of which only eight were suitable enough to be used for a meta-analysis. The picture that emerges from the 51 effect studies is ambiguous: in some studies there is a clear improvement of the objective safety, others show a significant rise in crime rates. This is underlined by the outcomes of the meta-analysis: half of the studies showed a negative effect of camera surveillance on the objective safety, the other half showed a positive effect (Schreijenberg & Homburg, 2010). Exact percentages on the effects were not presented in this meta-analysis, leaving room for imagination. However, the setting in which camera
27
surveillance was executed might provide in interesting observations. Three out of four studies where a negative effect was found on the crime rates were executed in entertainment and nightlife areas. The four studies which found a positive effect on crime were all executed in residential areas, the city centre or non-specified areas within the municipal borders (Schreijenberg & Homburg, 2010: pp. 33-48). This leaves the impression that camera surveillance might not be the best solution to counter crime in entertainment and nightlife areas, although caution is needed since the body of knowledge underlying this statement is fragile. To put it more generally, the amount of effect studies executed in the Netherlands is too little to come up with causal conclusions on the objective effect of camera surveillance (Wittebrood & Van Beem, 2004). Combined with the international studies, it appears that there is a small, positive effect on crime rates, however, this is highly dependent on both the manner and the setting in which camera surveillance is executed. There are reasons to assume, for example, that actively monitored CCTV is more effective in diminishing crime rates, although this is still highly debatable, and more research into this subject is needed.
Varying results can also be derived from evaluation studies regarding the subjective effect of camera surveillance in the Dutch context. In the evaluation study of Wittebrood and Van Beem (2004) there was, beside the effect of camera surveillance on the objective safety, also attention for the subjective effect of CCTV. It remains unclear in how many of the 11 studies on camera surveillance subjective effects were measured, although in both studies which met the quasi experimental guidelines it was part of the research framework. The study executed in the residential area showed a positive effect on the subjective safety, however, Wittebrood and Van Beem (2004) do not mention the exact magnitude of the effect, thus the significance of the improvement remains unclear. The study regarding cameras in public transport initially showed a small improvement of the subjective safety, but it turned out to be non-significant. Therefore, it is argued that there is no effect of camera surveillance on the subjective safety among public transport users (Wittebrood & Van Beem, 2004). In their concluding remarks, Wittebrood and Van Beem (2004) state that it remains unclear which measures are effective in increasing social safety, both in objective and subjective manner. Additionally, in light of the gained insights regarding the effect of camera surveillance so far, it is sensible to assume that measures in general affect both objective crime and subjective sense of security. It is even stated that it requires more effort to improve the subjective safety than only countering crime in absolute, objective manner (Wittebrood & Van Beem, 2004; Ditton, 2000). However, a more recently executed study comes up with different conclusions. From the eight studies
28
that were subject of the meta-analysis by Schreijenberg and Homburg in 2010, five were dealing with the subjective effect of camera surveillance. In these five studies, the sense of security increased, resulting in the observation that it does not matter in which setting camera surveillance is used; it has a positive effect on the sense of security of the public (Schreijenberg & Homburg, 2010).
2.3.3 Evaluation of camera project ‘Buikslotermeerplein’
Additional to the international and national studies regarding the effect of camera surveillance on objective and subjective elements of crime, the municipality of Amsterdam executes evaluations of its camera projects. The purpose of the biannual evaluations is to provide a transparent assessment of the proportionality of the executed surveillance measure, as explained in paragraph 1.3.1. The evaluation serves as a base for the decision on continuation, adjustment or termination of the existing projects in the city. The decision instrument consists of four components: (the development of) the safety in the target area, the usage of camera footage to increase the safety, the sense of security amongst civilians and lastly, potential demographic, social or spatial developments that may influence the necessity of camera surveillance in the near future (Homburg & Bleeker, 2015b). The first three components refer to the values of objective and subjective safety and the evaluation can, therefore, provide extra information for the current research.
The latest evaluation of camera surveillance projects in Amsterdam was presented in 2015. For the objective analysis, crime data from the police of the period between August 2014 and July 2015 was used and compared with the data of the previous year. The total crime rate at the ‘Buikslotermeerplein’, the square where shopping area ‘Boven ‘t Y’ (subject of analysis for this study) is situated, went down by 19 percent in 2015. However, public violence offences such as assaults and vandalism increased, in the same period, by more than 38 percent (Homburg & Bleeker, 2015b: p. 27). A major part of the increase can be explained by a significant increase in the amount of reported quarrels, which account for 50% of the public violence offences. Data for the subjective analysis was gathered by the Research & Statistics department of the Amsterdam communal administration through a survey executed in August and September 2015. A total number of 204 respondents took part in a survey executed in the public space of the ‘Buikslotermeerplein’, of which 104 were residing at or close to the square and the other 100 were visitors (i.e. for shopping purposes) (Homburg & Bleeker,
29
2015b). From the 204 respondents, 19% sometimes or often felt insecure at the site compared to 26% in the previous year. Furthermore, according to 47% of the respondents, surveillance cameras help to improve the sense of security, against 38% the year before. Additionally, nearly 70% of the people believed in the deterring effect of cameras on crime rates. Specific interesting findings in relation to the current research is the awareness on the presence of cameras: 66% of the respondents indicated that they are aware of the presence of cameras at the ‘Buikslotermeerplein’, a small increase in comparison to the 62% that noticed the cameras to be present in 2014. Furthermore, 97% of the questioned people were in favour of camera surveillance and only 10% raised concerns on privacy issues. Interestingly in this light is that still 47% of the people prefer real life police surveillance above its digital counterpart (Homburg & Bleeker, 2015b: pp 45-68).
To conclude: the evaluation showed a decrease in the total amount of committed crimes at ‘Buikslotermeerplein’, resulting in a positive trend of objective safety. Homburg and Bleeker (2015a) argue that relatively few priority offences are observed given the amount of cameras present. In considering the proportionality of the camera project ‘Buikslotermeerplein’, changes in the crime rate numbers did not provide in a clear answer according to the evaluation report, whereas the subjective safety indicated that the project is proportional (Homburg & Bleeker, 2015a). The mayor of Amsterdam found in this enough supporting evidence for the extension of the camera surveillance project at ‘Buikslotermeerplein’ until 31 December 2017 (Gemeente Amsterdam, 2016).
2.4 Broader perspective on the effect of CCTV.
Apart from a positive objective and subjective effect of surveillance cameras, it is imaginable that cameras can have a negative impact too. For example, it can give visitors of the public space a false feeling of security, making them vulnerable for an assault (Welsh & Farrington, 2009). Additionally, the instalment of cameras is a so called opportunity restricting measure in a delimited, target area. National as well as local authorities use this measure along with, amongst others, improvement of street lighting, increased police surveillance or varying closing times of nightlife venues to diminish the opportunities for the execution of crime. As stated in paragraph 2.2, it is assumed by authorities that camera surveillance will result in decreasing crime rates, however, this assumption is often refuted by the idea that these measures result in a displacement of crime (Bernasco et al., 2006; Braga, Papachristos &
