Leiden University | Faculty of Governance and Global Affairs |
MSc Crisis and Security Management
The Effectiveness of
Neighbourhood Watch
On the Causes of Success
By Robin Kuijpers BA | 2467771 Date | 11 January 2020
1
I | General information
Thesis title
The Effectiveness of Neighbourhood Watch: On the Causes of Success
Number of ECTS Date
15 12 – 01 – 2020
Student Study Program
Name | R. (Robin) Kuijpers BA Leiden University
Number | 2467771 Faculty of Governance and Global Affairs
MSc Crisis and Security Management
Contact details Supervisor
Westpoint 206 Dr. Leticia Elias Carrillo
5038 KG Tilburg i.l.elias.carrillo@fgga.leidenuniv.nl +31 (0)6 425 333 64 + 31 70 800 9506 robinkuijpers@live.com Second reader Dr. mr. Ernst Dijxhoorn e.e.a.dijxhoorn@fgga.leidenuniv.nl + 31 70 800 9082
2
II | Abstract
This thesis focussed on the question which variables were the cause of Neighbourhood Watch’s (NW) success in reducing home burglary numbers in the city of Tilburg. Two teams were tracked: Goirke/Hasselt/Bouwmeester (GoHaBo) and Noordhoek. The results were gathered through thirteen interviews and the observation of seventeen patrols comprising 48 interactions with citizens. The results were quantified, and content analysis was performed on the transcribed interviews. The observations were coded and used for illustration and triangulation purposes. These two methods led to the conclusion that personal enthusiasm and respect towards citizens are absolutely essential for NW to work properly. Social control was given a third place, and deterrence a fourth. These are the two main mechanisms through which NW works. A fifth and last place was given to communicative skills of the volunteers. As such, the hypothesis that deterrence would make it to the top of the ranking was proven wrong.
3
III | Foreword
This thesis marks the end phase of my Master’s, and thereby likely the end phase of my study period as well. During my studies, I had an amazing time; both at Tilburg University and at Leiden University. I was given many opportunities, and especially Leiden offered me a field that I can really see myself working in. Writing this thesis would have been significantly harder without the help of some people.
Of course, all people that I interviewed during my thesis, or went on patrols with for participant observations, were a tremendous help. I am not going to name all of them, as it would become quite a long list. If one of them is reading this, they will know I am talking about them.
Besides the people that helped me gather my data, I would especially like to thank Leticia (Lety) Carrillo, my supervisor. She truly helped me in my writing process. This is my first Master’s thesis, so obviously I did not know about all the things that should or should not be included. Lety’s feedback was always very clear, and she took a lot of time out of her agenda to explain everything. I was expecting sessions of 30 minutes, but one and a half hours were not uncommon. Thank you for your effort.
Stefan Berrelkamp, my lovely friend: as always, I owe much to you. Whenever I got stuck you helped me. Like during my Bachelor’s thesis, while at the time I could not always appreciate your feedback, of course it was for the better and you helped me get forward. You made me focus. You were and are like my personal anchor; always bringing me back when I get distracted by everything life throws at us. Thank you for your steadfastness.
Yvonne Simons, while I barely know you, through you I got in touch with the Tilburg Neighbourhood Watch schemes. Without you, this thesis would have an entirely different topic and I probably would not have finished it on time. Thank you for giving me inspiration.
And lastly, Ernst Dijxhoorn. So far, I have never met you, but I hope to do so during my defence. Still, you take the effort to be my second reader. Thank you for agreeing to do so. Thank you for helping out a stranger.
4
III | Table of contents
I General information 1 II Abstract 2 III Foreword 3 IV Table of contents 4 1 Introduction 6 1.1 Topic 6
1.2 Justification of the topic 6
1.3 RQ and focus 8
1.4 Relation to CSM 10
1.5 Aim and hypothesis 11
1.6 Methodology summarised 12 1.7 Relevance 12 1.7.1 Academic relevance 12 1.7.2 Societal relevance 12 1.8 Reading Guide 13 2 Theoretical framework 14 2.1 How NW operates 14 2.2 Literature on NW 15
2.2.1 Concepts from the literature 15
2.2.1.1 Objective and subjective effectiveness of NW 16
2.2.1.2 Physical and digital NW 17
2.2.1.3 Social control and deterrence 19
2.2.1.4 Personal enthusiasm 23
2.2.1.5 Communication skills 24
2.2.1.6 Respect 24
2.3 Theory testing 25
2.4 Conceptual model and conceptualisation 25
3 Research design 28
3.1 On observation and generalisability 28
3.2 Tilburg as a case study 29
3.3 On sources and data 30
3.3.1 Sources 30
3.3.2 Data collection 30
3.4 The neighbourhoods under review 32
5 3.5.1 Content analysis and participant observation 34
3.5.2 Observing structurally 35
4 Data collection 40
4.1 Interview protocol 40
4.1.1 Script before interview starts 42
4.1.2 Script after interview 46
4.2 The coding scheme 46
4.2.1 The codebook explained 46
4.2.2 Elaboration on the unit of analysis 47
4.2.3 Codebook 48
4.2.3.1 Coding rules 49
5 Results and Analysis 51
5.1 Interview results 51
5.2 Participant observation results 54
5.3 Analysis 57 5.3.1 On personal enthusiasm 58 5.3.2 On respect 61 5.3.3 On communicative skills 63 5.3.4 On social control 66 5.3.5 On deterrence 70 5.3.6 On WhatsApp NW 73 6 Conclusion 75 6.1 Answering the RQ 75 6.2 Limitations 77 6.3 Discussion 77
6.4 Recommendations for further research 79
V References 80
6
1 | Introduction
“Roughly 80 percent of all arrests due to home burglary are (…) made possible because of the alertness of inhabitants in the neighbourhood” (NOS, 2017).
1.1 | Topic
Eighty percent is an enormous amount. Apparently, citizens are quite concerned about suspicious activities in their neighbourhood, which stands in contrast to cities sometimes being characterised as areas lacking high levels of social control. This, for instance, shows in Lyerly & Skipper’s research (1981), in which they look at the effects of social control in urban and rural areas. That people do not even know their own neighbours anymore has become a common phrase reflecting this trend (Zito, 1974). An exception to this development is the rise of Neighbourhood Watch schemes, in Dutch called Buurtpreventie. Neighbourhood Watch (NW) is a form of co-production of public services. State and citizens work together to tackle a problem; in this case security in the neighbourhood, with a special focus on reducing home burglaries (Brandsen & Honingh, 2015). When a neighbourhood has a NW team, it means that citizens are walking the streets (often in the evening) and keep an eye out for suspicious activities to ensure the security of the neighbourhood. Originally, NW started out in the late 1960’s in the United States (US), in Queens (NY), after the murder of Catherine Genovese (Rasenberger, 2006). These days, NW has found its way into Dutch society as well; especially during the last five to ten years (Lub, 2016).
1.2 | Justification of the topic
A significant amount of research has been carried out to find answers to the question “Is NW effective in preventing crime?”. A vast meta-analysis of all this research was published in Security Journal in 2009 (Bennett, Holloway & Farrington). This meta-analysis took the results of eighteen NW evaluations into account, together combined into twelve research projects; meaning that some of those projects looked at more than one evaluation. Fourteen of these evaluations focussed on the United Kingdom (UK), three on the US, and one on Canada. This comprised all the evaluations in light of effectiveness in the English language on this topic at that time. The analysis, indeed, found NW to be effective in preventing home burglary. However, all this research was quantitative in nature. This kind of methodology was best suited to answer such a closed question as whether NW was effective or not. Their quantitative
7 approach consisted of, for example, comparing the crime figures in a certain area before and after the introduction of NW, or by comparing those numbers in a NW area with a similar area that has no NW scheme(s).
In studies like the evaluations mentioned above, it is not common to adopt a qualitative approach. That is because the underlying reasons for the (in)effectiveness are simply not the focus of these studies; the focus has mainly been on evaluation. Nevertheless, not every single research was quantitative in nature, as the example in the next paragraph shows.
Van der Land (2014), performed ethnographic research regarding the status of physical NW in the Netherlands. In this research, he also looked at NW’s effectiveness. Van der Land held interviews in sixteen different NW teams, covering multiple municipalities. One of the results he found was that the people he interviewed were convinced of the positive effects of NW regarding the countering of home burglary.
“Burglaries were reduced by 56 percent. Last week we received the figures on this. I do think that is because of the team, for in the city centre criminality increased by 40 percent, and in another neighbourhood even by 60 percent.” says a volunteer in Sliedrecht (Van der Land, 2014, p. 64).1
Several variables that will be covered in the Theoretical framework are estimated to have an impact on NW’s effectiveness, although the empirical validations of these estimations do not cover all variables mentioned in the literature. While all were chosen for their relevancy, the studies from which they were drawn sometimes did not focus on home burglary. This is also clarified in the Theoretical framework chapter. Because not all concepts were tested with regards to home burglary, there is more work to be done in light of academic publications regarding NW. That way, also the variables that are estimated to be relevant, but so far have not been validated, can be empirically tested. It would be a waste to ignore the qualitative side of one of the major themes in the expanding subfield of coproduction of security and only focus on numbers. Without discrediting quantitative methods, qualitative methods are more practical than quantitative methods in shedding light on the impact of the presumably relevant variables. It can be difficult to see the whole picture when one focuses merely on numbers.
1 Own translation by the author. Original Dutch quote: “Inbraken zijn met 56 procent gedaald. Vorige week hebben we hier de cijfers over gekregen. Dat ligt denk ik wel aan het team, want in het centrum was criminaliteit met 40 procent gestegen en in een andere wijk zelfs met 60 procent.” says a volunteer in Sliedrecht.”
8 1.3 | RQ and focus
In light of the above, this research will evolve around the question “What variables explain the effectiveness of Dutch Neighbourhood Watch schemes regarding the prevention of home burglaries?” As NW originated as an American concept, the United States has, logically, also been the main context used to research NW. The Dutch focus in this paper is therefore chosen to both diversify the geographical context, and to be able to gather qualitative data easier. It is not just research focussed on the US and UK, like the meta-analysis described above, that show the effectiveness of NW. Dutch research shows this success as well. The amount of property crimes, for instance, has lowered since the introduction of NW, as shows in Figure 1 below. This is the same type of crime as that which was focussed on in Bennett, Holloway & Farrington’s research (the meta-analysis). This is because home burglary belongs to the category of property crime. Not all property crimes are burglaries, but all burglaries are aimed at property crime. The decrease of property crime shows in Lub’s research called “De burger op wacht” (2016).
Figure 1: Development of NW and property crimes (taken from Lub, 2016, p. 27).
Blue line: Annual development of municipalities with new NW teams (left number) Brown line: Annual development of property crimes (right number)
9 Figure 1 looks at new teams as to show the accelerating pace with which new teams are being set up annually. It is true that the above figure 1 could show a correlation instead of a causation, although viewed in combination with the 2009 meta-analysis mentioned above, these numbers suggest that Dutch NW is indeed effective. This is because apparently a rise in NW either sometimes lowers but at least keeps steady the amount of property crimes. It no longer rises.
In the Netherlands, NW is not used to prevent or tackle serious crime; that is not among NW’s goals and it is not what NW volunteers are trained for. Tackling and preventing serious crimes (armed robberies, drug trafficking, murder, etcetera) is considered to be the responsibility of the police only, whereas burglary (and property crime in general), is not.
“Serious crime, people often say, is kind of difficult to detect for inhabitants, while they are noticing nuisance way more often and are also better able to guess whether that decreases or not” – Van der Land (2014, p. 63).2
Therefore, this research is limited to the prevention and tackling of home burglaries solely, as overall, reducing residential burglaries scores highest among NW’s goals in the Netherlands. This shows in Figure 2 below. NW teams work with a goal in mind; something they would like to achieve in their neighbourhood. On average, 88 percent of the Dutch NW schemes has included the prevention and tackling of home burglary in their goals (in Dutch called woninginbraken). This type of crime is therefore, by far, the most prevalent shared goal among NW in the Netherlands.
2 Own translation by the author. Original Dutch quote: “De harde criminaliteit, zegt men vaak, gaat min of meer langs de dagelijkse leefwereld van bewoners heen, terwijl zij van overlast veel meer merken en ook beter kunnen inschatten of die afneemt of niet.”
10 Figure 2: the goals of Dutch NW teams (taken from Lub, 2016, p. 23).
As Lub states: “The largest part of neighbourhood watch schemes in the Netherlands is focussed on the prevention of home burglary” (Lub, 2016, p. 31). That this has effect is, for example, portrayed in the fact that between 2016 and 2017, Dutch home burglary numbers lowered by 10% (Rijksoverheid, 27 December 2017). As a reason for this trend, the national government mentioned prevention to be to be significant. As such, new initiatives regarding NW were stimulated in 2018. (Rijksoverheid, 27 December 2017).
1.4 | Relation to CSM
In order to provide others with security, one must have a certain degree of power. This idea is nicely elaborated upon in the famous book Leviathan (Hobbes, 1651). Throughout the ages, such power has certainly not always been with the state or a similar authority. One only has to look at present-day Somalia, where the power of the government does not reach far
11 beyond the borders of its capital. The contemporary system of states possessing the monopoly on the use of violence is actually rather unique; it is a result of the Peace of Westphalia, signed in 1648, which turned the state into the official sovereign entity inside its own borders (Croxton, 1999). The system is also subject to change as, for instance, can be seen in the rise of private security and military companies (Krahmann, 2008).
This changing nature of the status quo shows in the rise of the co-production of security as well; the state and its inhabitants are increasingly looking to tackle security-related issues as a team where both are valued. According to Van der Land (2014), the last three decennia have been characterised by a growing number of citizens being involved in security provision. This has been done through the voicing of their opinions on policies, together with active participation through production via, for instance, NW. The growing entrenchment of co-production can therefore be said to be a symptom of our current Zeitgeist; the spirit of the times. “NW … is … mainly a corollary of the securitising Zeitgeist of the society” (Lub, 2014, p. 31). This growing entrenchment of coproduction is the central topic in CSM’s 2019 elective course “Privatisation of CSM”. “Privatisation” has multiple definitions, but here means “the act of reducing the role of government or increasing the role of the private institutions of society in satisfying people’s needs; it means relying more on the private sector and less on government” (Savas, 2000, p. 2). Privatisation is increasingly commonly referred to as “public-private partnership”, while this is not exactly the same. Public-private partnerships are, rather, a form of privatisation. While every public-private partnership is a form of privatisation, not every form of privatisation can be called a public-private partnership. “Privatisation of CSM” divides these partnerships into (1) partnerships between the state and private companies, and (2) the state and local communities. Because of the number two, it is important for the field of security management to increase its knowledge regarding the variables that are at play in NW schemes, as NW is the most prominent example of a public-private partnership between the state and local communities.
1.5 | Aim and hypothesis
This research will be among the earliest studies dedicated to this topic specifically. As such, and due to it having the format of a Master thesis instead of a larger research over a longer time period, it is not realistic to expect a 100% clarity on the mechanisms beneath NW being effective. Providing percentages regarding the variables’ effectiveness is, therefore, not the aim.
12 One thesis is too small for that. The aim is to rank the variables in relation to their importance in terms of impact on the effectiveness of NW. Currently, it is expected that one variable has quite an impact: deterrence. This is due to the high expectations that were raised regarding this variable in the research of Boutasmit & Snel (2016). As such, the hypothesis during this thesis will be that deterrence will make the top of the list of most relevant variables for the prevention of home burglaries in Dutch NW schemes. Nevertheless, research on the variables in terms of theory-testing is quite minimal, because as mentioned above, the focus of this research field has been more on the question of effectiveness of NW instead of the mechanisms underlying that effectiveness. In order to turn the assumptions into facts, this research wants to look at those assumed variables and either corroborate or falsify them. How large the impact of these variables is, is also not known at all.
1.6 | Methodology summarised
In order to answer the RQ and see whether the hypothesis developed is indeed correct, interviews will be held with three actors: members of NW teams, police officers, and the municipality. In order to double-check the answers that are going to be given, the author shall participate in a NW scheme for several weeks himself. The combination of the interviews with the observations will make it possible to check for inconsistencies in the data gathered. This methodology is further elaborated upon in Chapter 3: the Research design.
1.7 | Relevance
1.7.1 | Academic relevance. Studying these presumably important variables is relevant for academic for three reasons. (1) Because it can falsify or corroborate the results found in the studies in the literature review in the next chapter. (2) It adds an extra layer of complexity to that by ranking the variables in terms of impact as well. (3) Besides that, it helps to encourage the practise of studying the phenomenon of NW in a different national context than that of the US or Great Britain.
1.7.2 | Societal relevance. From a societal perspective, the results could be beneficial to local governments and local police. If the results would indicate that one of the variables found in the literature review is, according to the data gathered in this research, not relevant, the need to devote time and resources that are scarce to this seemingly irrelevant variable
13 becomes less pressing. While the outcomes of this thesis shall also have to be corroborated by others, it is hoped that these cumulative outcomes eventually will be able to start providing extra focus to this form of coproduction. After all, money and time can only be spend once. Extra focus on the variables that really make a difference could enhance the efficiency of NW schemes.
1.8 | Reading guide
Chapter 2 will take a look at the Theoretical framework used, discussing the link to the field of Crisis and Security Management (CSM), what the literature has to say, the concepts used, the operationalisation of those concepts, and the conceptual model. Chapter 3 looks into the Research design. It focusses on what kind of choices were made with regards to type of research, data collection, and analysis and justifies them. It also looks at operationalisation. Chapter 4 looks at Data collection. It contains the interview protocol, a list containing all interviewees, a codebook for content analysis, as well as the coding rules that are adhered to in the process. Chapter 5 starts by portraying the actual Results of the interviews and participant observations in a quantitative manner. During the Analysis afterwards, qualitative data is added to put the numbers into perspective and give meaning to them. To support this process, situations encountered during the participant observations are described as well, after which a ranking of all the variables is made. The chapter ends by giving some attention to the digital variant of NW. Chapter 6 is the Conclusion. It contains a summary of the main findings of the Analysis, as well as an answer to the research question. Limitations to the research will also be talked about in the conclusion. Chapter 7 is the last chapter, the Discussion, in which ideas raised by the interviewees themselves are discussed. The most promising ideas will be recommended for further research.
14
2 | Theoretical framework
2.1 | How NW operates
Figure 3: organisation of a NW scheme (made by author)
Above, in Figure 3, one can see how NW operates. The main task of the team of NW volunteers is to walk around the neighbourhood and be watchful of any suspicious activities. While this thesis only looks at reducing home burglaries, NW volunteers in general focus on more than that. They might also do something about littering, talk to people who are making nuisance, walking elders home at night, etcetera. What is relevant for this research, though, is that NW’s main goal generally is the reduction of home burglaries. The volunteers report information to both the police and the municipality, depending on the topic (for example, information about littering will be reported to the municipality, while information about someone stealing bicycles will be reported to the police). The police provides information to the team of volunteers as well, when they expect certain areas to be of more interest due to, for instance, calls they received from inhabitants. The municipality provides the team with material, such as jackets, searchlights, flyers, etcetera (Municipality of Tilburg, 2019).
15 2.2 | Literature on NW
A significant amount of US focussed studies have been carried out to research the effectiveness of NW. Some of this research has also been used in a meta-analysis discussed in the introduction (Bennett, Holloway & Farrington, 2009), as earlier studies could not agree on whether NW was effective or not. A meta-analysis does not study something “in the real world”, but makes an analysis of the results of a large number of separate studies to come to an overall conclusion. This all-quantitative meta-analysis answered this positively, and now serves as an important argument in this discussion, due to its vastly extensive nature. In the Netherlands, authors that built upon this research include, but are not limited to, Van Eijk (2013;) and Van der Land (2014). Van Eijk looked at whether NW made the participants feel more secure or not. It was found that this mainly depended on external variables; elements other than whether they participated or not, such as contact with other inhabitants. This resulted even in different perspectives in the same neighbourhood. An example of such an external variable would be the amount of information on security-related issues the volunteer has access to. As such, nothing inherent to NW was found that resulted in an increase in subjective feelings of safety. Van der Land’s study rather focussed on the status of NW in the Netherlands; what kinds of co-production there are, what the results are, etcetera. On top of that, multiple students addressed co-production in general in their theses since 2010. Examples include Stokes (2010) and Breugel (2015). Stokes looked at to what extent the level of participation influenced the expectations of police and participating civilians of their common projects. This extent was found to be quite large, although there were other variables of influence as well. Breugel looked at the motivation of volunteers to participate, while making use of Self-Determination theory. In the end it was found that, to best enhance motivation, volunteers needed to be given enough space. The role of the municipality should therefore be facilitating instead of rigid in their guidance on how to do things.
2.2.1 | Concepts from the literature
In this section, the literature review will be extended. This is done specifically to derive the variables from the literature that could explain the effectiveness of NW, as well as to learn more about the larger theories surrounding these variables. Not every concept in the literature was given a definition by its author. Nevertheless, these concepts might still be considered relevant variables for this thesis due to them being likely to explain NW’s effectiveness. As
16 such, these definitions will be created by the author. Justifications for those definitions will be provided each time.
2.2.1.1. | Objective and subjective effectiveness of NW. The main concept that is used in this research is the effectiveness of NW. This has both an objective and a subjective component to it. Even more so in the public sector than in the private sector, performance is usually measured by both. Nevertheless, even in the public sector, objective effectiveness is viewed as the golden standard, for it is regarded to “reflect the real world accurately and minimize discretion” (Andrews, Boyne & Walker, 2006, p. 16). Objective effectiveness is defined as the actual result in light of a decrease in petty crime figures, because this is something that municipalities indeed keeps track of. Subjective effectiveness is defined as having a positive influence on the perception of security as felt by citizens in the respective neighbourhood. This definition is chosen because “security” is a broad category that can encompass many different things. This is done on purpose, as by definition a subjective feeling of something is different for each person, and therefore should not be narrowed down to merely mean specific things. The concept of home burglary, corresponding to “objective effectiveness”, is defined as (attempted) thieving from a residential building, but also from corresponding garages, gardens, etcetera. This is the definition that the Dutch CBS (Central Bureau for Statistics) adheres to when it keeps track of home burglaries in the Netherlands.
That NW generally leads to objective effectiveness has by now been accepted by most scholars in the field, due to the extensive amount of research that has led to this conclusion in the meta-analysis discussed earlier (Bennet, Holloway, & Farrington, 2009). Whether it also leads to subjective effectiveness is a more contested debate. That NW also has an effect on subjective security is known, but what this effect exactly is still remains unknown. (Rosenbaum, 1987). Although it is not the main aim of the study, it is has to be acknowledged that the subjective element plays a part. This is because in some instances, subjective effectiveness and objective effectiveness are intertwined. This is elaborated on in the paragraph below, which is the only space this topic will be given.
Does NW make people feel more secure or not? On the one hand, there are people that agree with this. When personally observing one of the NW patrols in Noordhoek (Tilburg), a woman came up to us saying that she was glad that she found out NW was present in her neighbourhood. She mentioned that it made her feel safer at night. On the other hand, the
17 argument can be made that when NW volunteers are walking around, this might be because it is necessary. While most neighbourhoods where NW initiatives start in fact have relatively low crime numbers (Lub, 2016), that argument might not be convincing when talking about feelings of (in)security. Objective and subjective security are intertwined concepts. When people see a crime happening, chances are high that this will affect how secure or insecure they feel. That, of course, works both ways: when one feels truly secure, (s)he might not even notice or pay attention to home burglary on the other side of the street. Such a person could dismiss a crime like that as an anomaly in an otherwise secure environment.
The woman in the above example mentioned she felt “safer”. In this research, an emphasis is placed on “security”, and not on “safety”. Nevertheless, people might talk about “safety”. This, however, will be out of ignorance of the difference between the terms and is, therefore, not intentional. Safety denotes protection from unintentional harm, while the term security places its focus on intentional harm (Pietre-Cambacédes & Chaudet, 2010). NW thus aims to increase security. An example clarifies the difference between the two: NW volunteers might warn inhabitants about a burglar being active in that respective area. Burglary is done on purpose; it is intentional, and therefore a focus for NW schemes. When a car in that same neighbourhood accidentally hits a child that is consequentially hospitalised, this may cause the same amount of harm (or perhaps even more). Nevertheless, it was an accident; it was unintentional, thereby making it a safety-issue and not a security-issue. As such, this last example would not be part of NW’s focus.
2.2.1.2 | Physical and digital NW. The concept of NW itself is divided into the physical teams on the street on the one hand, and the WhatsApp groups (Vollaard, 2016) on the other. This differentiation is important, as these two forms of NW are rather different in what they do. WhatsApp NW group users pay closer attention to out-of-the-ordinary activities in their neighbourhood and warn the other group members and/or the police about such activities when they take place (own observation, April & March 2019). While physical NW teams do the same, they do so by actively patrolling the streets, instead of communicating via WhatsApp. Their physical presence goes beyond merely being part of the police’s network of eyes and ears. The central distinction is the contact with the people in the neighbourhood. This research’ focus is on the physical teams. The difference between the two phenomena is considered too large to thoroughly look at them in the same thesis. It is, however, not entirely dismissed. The author acknowledges digital NW’s value in light of reducing home burglary figures. As such, while
18 not the main focus, digital NW will to a certain extent be addressed in all the interviews that will be conducted.3 For participation in WhatsApp NW, proactiveness of the participants is much less required than in its physical counterpart. This also corresponds to the fact that WhatsApp NW requires significantly less time from the volunteers. Also, less coordination is required. This lesser extent of coordination for example shows in the fact that physical teams usually receive training and equipment from the municipality, whereas digital teams do not (Municipality of Tilburg, 27 September 2019). A last significant difference is related to the amount of social cohesion in the group. Whereas in a physical NW team people actually have to know each other (to get started as a team, and to walk the patrols), this is not a prerequisite in a digital team (Lub, 2019, p. 9).
In 2018, roughly 3000 WhatsApp NW groups existed, circa 2000 of which were created after 2014, according to Lub. In total, 653 physical teams were in place. Of those 653 teams, 190 teams were set up after 2014 (Lub, 2019). These numbers show a significant growth in both the digital and physical forms of NW. As such, it is not the case that the digital teams are replacing the physical ones. The one is not the next-generation version of the other; they are complementary and both increasing, as can be seen in Figure 4 below.
Physical NW Digital NW
Figure 4: NW groups created after 2014, as percentage of the total amount (taken from Lub, 2019).
Akkermans & Vollaard (2015) performed research on the digital variant in Tilburg. They suggest that both forms of NW are relevant in the reduction of home burglary, and that it is likely that both have a positive effect on one another in terms of acquisition of new members. However, the authors did not research the effect of physical NW; their focus was on the digital
19 variant. In this same research, they show that between 2013 and 2015 home burglary numbers lowered by 40% in the neighbourhoods of Tilburg that were having WhatsApp NW groups. The authors admit that it was difficult to distinguish these results from the other four measures the municipality was taking, one of them being the physical version of NW. They managed to rule out the results of three of those measures.
However, in their dataset, in order to distinguish the effect of the WhatsApp groups from other measures, they assumed physical NW started at the same time in every neighbourhood: January 2012. This is incorrect. In Tilburg, physical NW is not a city-wide initiative. Like in most municipalities, NW is arranged on a neighbourhood level (Zijl, 2019). As such, years of origin differ per team, even though the starting date of the first NW initiative may indeed have been January 2012.
The neighbourhoods this research will look into actually started in 2014, and not in 2012. Therefore, that 40% reduction in home burglaries could also be interpreted as a result of physical NW. Likely, it is a combination of both physical and digital NW. The point is that the effect of multiple measures on home burglary figures is difficult to distinguish. The decrease should not be said to be the consequence of one single measure as long as the effect of other measures cannot be sufficiently ruled out.
2.2.1.3 | Social control and deterrence. The concepts “social control” and “deterrence” are often mentioned in the literature as to be potentially relevant for the reduction of home burglaries through NW (Perry, 1984; Lub, 2016; Boutasmit & Snel, 2016). For social control, Lub could not corroborate this in his participant observations. Deterrence, however, was indeed deemed relevant by the NW teams that he followed. Let us see how the deterrence doctrine and social control theory relate to each other. This is depicted in Figure 5 on the right. In this thesis, social control is conceptualised as “the capacity of social
groups to effect self-regulation”. Its goal is that people conform their behaviour to norms, such as person X locking her door at night, not because someone forces her, but because she has
Deterrence Social control theory
Figure 5: Relationship between social control and deterrence (made by author for visual clarification).
20 been conditioned to genuinely believe this to be a smart idea. This is quite a broad category, and when deterrence is defined as “discouraging an action or event through instilling doubt or fear of the consequences” (Oxford dictionary), we see that the latter concept is more narrowed down and focussed. This is because the deterrence doctrine is, in fact, part of social control theory. However, in order to be able to more precisely pinpoint which part of social control theory is relevant for NW, in this thesis an analytical distinction is made between the two. Deterrence is just one of the ways through which self-regulation can be effected in a group; in this case the neighbourhood. Another way through which this could be achieved according to social control theory is, for example, the internalisation of common norms. This is, however, not relevant for NW. While NW indeed focusses on prevention, it is not a social institution aimed at norm internalisation like schools are, or families, churches, mosques, etcetera. The only relation NW has with norm internalisation is in increasing civilians’ awareness regarding
security related issues (Municipality of Tilburg, 26 September 2019). An example would be reminding elders of what in Dutch is called “babbeltrucs”: the practise of people trying to get access to another’s home by pretending to be someone they are not, like a technician from an energy company, putting them in an easier position to commit theft (Police, n.d.).
When talking about social control, there has been discussion on whether or not to include intentionality in the concept. While some scholars say that to include intentionality is to provide more focus, thereby making the concept more relevant, others stick to the idea that social control is, in fact, any way in which a human being affects the behaviour of another. As such, the inclusion of intentionality excludes certain acts that, according to them, should fall under the header of social control (Gibbs, 1977). Such a conceptualisation is, however, too large to be of any real analytical value. Especially for NW, “attempt” is vital and should be read in the conceptualisation of social control as well. This is because not every encounter between civilians and volunteers on their patrols will have the desired effect, while nevertheless the attempt of volunteers to change the behaviour of people in a certain way is still important. NW volunteers will not get everything right all of the time, but their attempts to do so are still making a difference on average. As Gibbs mentions, “the very notion of an attempt entails the notion of intentionality” (Gibbs, 1977, p. 412), and that is why for NW, intentionality is important to the concept of social control (and thereby to deterrence as well).
21 An example of social control being relevant is the following. Social control in light of family ties is believed to be of major influence (Scholte, 2008). Social control is conceptualised as the capacity of social groups to effect self-regulation (Janowitz, 1975). However, this is not to say that this concept of Scholte works as strongly as is suggested in the study in which it was previously applied (Boutasmit & Snel, 2016), due to the different cultural context of the case used in this current paper. Their research was focussed on the Dutch Moroccan community, which is influenced by Arabic culture, especially in light of family ties. In such a context, family ties are often considered more important than in all-Dutch families. An Arab saying exemplifies this: I and my brother are against my cousin, I and my cousin are against the stranger (Naguib, 2008). If Arab culture is not strongly present in a neighbourhood, social control may be of less value as a variable. It could perhaps also manifest itself in a different way.
The deterrence doctrine comes in play in the argument that when NW volunteers walk their patrols and people see them on the streets, this already has a deterring effect on criminal behaviour. That deterring effect is something that is specifically mentioned during the trainings provided to the NW volunteers in Tilburg (Municipality of Tilburg, 2019).
Cohen and Felson (1979) performed research on the causes of crime. The idea that other theories of crime and of crime prevention were insufficient, as they did not match their data on crime numbers, led them to design a framework consisting of three essential variables that explain criminal behaviour. Their work was influenced by rational choice theory (Felson, 1986), as the variables show: (1) motivated offenders, (2) suitable targets, and (3) the absence of capable guardians. According to them, a lack of any of these three will cause a reduction in crime rates (Cohen & Felson, 1979, p. 604). The NW volunteer would in casu be that “capable guardian”. In a later work, Felson (1995) goes more in-depth on this “capable guardian” idea. He extends this idea by creating a matrix consisting of twelve categories of crime discouragement: three types of supervisor and objects of supervision times four levels of responsibility.
Before seeing all the forms a “capable guardian’ can take, it is important to address why these forms matter for NW volunteers. There is no national NW policy, meaning that each municipality has to figure out on its own how to deal with NW schemes. While most schemes indeed look alike, it can happen that some volunteers might view their tasks in a certain way, and others in another. In Tilburg, the municipality provides basic training for each NW team where this question is addressed. Nevertheless, other municipalities might not have that same
22 policy. As such, it is important to single out the responsibility of NW volunteers, for different views in this regard could affect the way volunteers go about their work. An example could be that volunteer A has the idea that she should monitor likely offenders, for instance youth in the local park at night, while volunteer B believes it is about monitoring the entire neighbourhood at all times. The consequence would be that volunteer A’s focus is mainly on the park during the evenings, while volunteer B also goes on patrols during the afternoon and not just in the park.
In Table 1 down below, one can see that C2 (managers, assigned), C3 (managers, diffuse) and C4 (managers, general) are marked in red. These three categories are the ones that NW volunteers can be argued to fall into. Let us first look at the A-B-C differentiation. While type of supervisor is a category of no true value (for it does not matter what it is named), the category object of supervision is indeed relevant. C stands for amenable places, and is the correct category for NW. After all, NW volunteers patrol the streets of a certain neighbourhood, an area; they do not monitor suitable targets (A) or likely offenders (B). Whether their level of responsibility should be branded as assigned, diffuse, or general is up for debate. One could argue that the correct category should be C2, as while they are volunteers, they do have certain tasks (depending on the NW scheme they are in). On the other hand, they can be characterised as C3 as well, for NW volunteers have a large amount of leeway in how they perform their tasks and what kinds of things exactly they pay attention to during their patrols. Lastly, an argument can be made for C4, as C4 does not treat NW volunteers as employees; which they indeed are not. The fourth category, whether it falls under A, B, or C, assumes an equal power relation between individuals. This makes for a more legalistic viewpoint. According to Dutch law, NW volunteers do not have any rights which make them stand above other civilians, because that is what they are: civilians. As such, legalistically, there is indeed an equal standing. The drawback of that argument is that when NW volunteers walk in groups and are uniformed, they will still have a certain appearance that makes it difficult to think of them as being on complete equal standing with other civilians. Therefore, C2 and C3 seem more likely to fit NW.
Table 1
For NW marked version of Felson’s Table of Crime Discouragement (taken from Felson, 1995, p. 59).
23 A. Guardians (monitoring suitable targets) B. Handlers (monitoring likely offenders) C. Managers (Monitoring amenable places) 1. Personal (owners, family, friends) A1 B1 C1 2. Assigned (employees with specific assignment) A2 B2 C2 3. Diffuse (employees with general assignment) A3 B3 C3 4. General (strangers, other citizens) A4 B4 C4
Now that the origins of the variables of social control and deterrence were both discussed, let us summarise the core. As mentioned before, the deterrence doctrine is part of social control theory, although in this thesis, for analytical purposes, a distinction between the two is made. If an act of social control cannot be characterised as the instilment of doubt or fear of the consequences, then it is assumed to fall under the social control variable and not under the deterrence variable. Social control also comprises attempt, and thereby intentionality should be read into the definition. NW volunteers monitor amenable places and their level of responsibility can be called either assigned, diffuse, or general. Each categorisation has its merits, and the way NW volunteers perceive themselves has consequences for their performance.
2.2.1.4 | Personal enthusiasm. Another concept possibly explaining the effectiveness of NW is that of personal enthusiasm of the individuals working and volunteering in NW schemes. Van Essen (2015) corroborated that this was a majo external influence in his own research, which was related to co-production between local governments (municipalities) and citizens. Therefore, it is expected to have a large impact on the effectiveness of NW as well. It
24 is defined as to what extent the NW volunteers enjoy what they are doing, for the author feels enjoyment to be most central to the concept of enthusiasm.
An example of personal enthusiasm impacting NW’s effectiveness is the amount of times a week a volunteer goes on patrol. NW is done on an entirely voluntary basis. This means that when someone does not want to go on a patrol, (s)he simply does not go. As such, volunteers are not obliged to do anything. The amount of involvement really comes down to volunteers’ personal enthusiasm for NW.
2.2.1.5 | Communication skills. Boutasmit & Snel identify the communication skills of the volunteers to be of great importance to NW’s effectiveness; more specifically, the cumulative positive or negative experiences that people in the neighbourhood have with the manner in which NW volunteers communicate with them. This was corroborated in their study, although this study did not focus on home burglary specifically. Communication skills are conceptualised as the extent to which NW volunteers are able to level with the civilians in the neighbourhood during conversations. This definition is chosen due to the connection that Boutasmit & Snel make between communication skills and civilians; it is not about internal communication, not about communication with the general public via media, but about the actual civilians that NW volunteers communicate with during their patrols.
Think of a volunteer that, for instance, does not know how to level with the respective inhabitant she is talking to. She might have more difficulty with conveying a certain message than another volunteer that knows how to show just the right amount of empathy. Without that right amount of empathy, telling a civilian that he, for example, left his front door open at night, could come across as negative interference.
2.2.1.6 | Respect. Boutasmit & Snel mention one, quite plausible, independent variable as well: respect. Respect is conceptualised as whether civilians are being approached in a decent manner, because “approached” includes verbal and non-verbal communication. Both are important in conveying the right message. Respect was found to be an important variable in all communication with citizens, although again it is stressed that this research did not solely focus on home burglary.
25 When a NW volunteer talks in a perfectly decent words, while at the same time his tone of voice or his stance communicate that he, for example, feels like he is standing above the person he is talking to, then respect is still lacking. This will then result in the volunteer’s message, again, not being taken seriously.
2.3 | Theory testing
It is important to keep in mind that the independent variables discussed above are variables that are presumed to have an impact on NW’s effectiveness in reducing home burglaries They were found in the literature review, but are not (yet) corroborated in multiple studies. As such, they form a starting point, but are not taken as absolutes. Therefore, what is done in this research can be characterised as theory testing.
2.4 | Conceptual model and conceptualisation
Figure 6: Conceptual model (made by author)
The largest box in the middle of the conceptual model in Figure 6 above portrays the variables that this paper expects to be of impact. Placed on the left, at the start, are the NW projects. These projects come in digital form (WhatsApp groups), as well as physical (volunteers on the streets). Especially physical NW is focussed on, as was elaborated upon earlier. As one moves to the right, (s)he comes across a larger box. This box is filled with the concepts that were discussed above. These are the independent variables. They are, based on the scarce literature there is on this topic, what is believed to make for the effectiveness of
26 Dutch NW projects. They are “personal enthusiasm”, “social control”, “communicative skills”, “deterrence”, and respect. The ranking here is chosen at random. All of the above finally leads to “effectiveness”; the dependent variable. When looking down, one can see that this variable is divided into objective and subjective effectiveness. As mentioned before, the author acknowledges that both are relevant due to them being intertwined to a certain extent, although subjective effectiveness is not focussed on in this research. In the conclusion, the reader will find an adapted version of this conceptual model. This adapted version will be based on the rank the respective variables were awarded in the Analysis chapter.
Below, in Table 2, all concepts that were discussed above are summarised, together with their respective definitions.
Table 2
An overview of all the definitions of the concepts discussed above (made by author).
Concept Definition
Objective effectiveness The decrease in home burglary figures.
Home burglary (Attempted) thieving from a residential
building, but also from corresponding garages, gardens, etcetera
Physical Neighbourhood Watch Teams of citizens patrolling the streets that are in contact with the police.
Personal enthusiasm (independent variable) To what extent the NW volunteers enjoy what they are doing.
Communicative skills (independent variable)
To what extent the NW volunteers are able to level with the civilians in the
neighbourhood during conversations. Social control (independent variable) The capacity of social groups to effect
self-regulation.
Deterrence (independent variable) Discouraging an action or event through instilling doubt or fear of the consequences.
27 Respect (independent variable) Whether civilians are being approached in a
decent manner.
The concepts that have “independent variable” behind it between brackets are the ones that are further operationalised at the end of chapter three.
28
3 | Research design
3.1 | On observation and generalisability
This research is an explanatory study. This is because it aims to explain the causal relationship between NW and the lowering of home burglaries (Sekaran & Bougie, 2013, p. 98). This relationship is also quantified in the section “the neighbourhoods under review” in this chapter. In order to answer the research question, a mixed methods strategy will be used, combining interviews with structured uncontrolled participant observation in the context of a single case study. There are four types of participant observation, as Table 3 below shows. Unstructured observation means that the observation is not focussed. What will be observed and how has not been predetermined yet. On the contrary, a structured observation starts with mutually exclusive categories, the instances of which can be coded by use of a coding scheme (Sekaran & Bougie, 2013, p. 139-141). With uncontrolled observation, one means “observation … conducted in a natural setting” (Sekaran & Bougie, 2013, p. 131). The opposite, controlled observation, means observation within an experimental setting where the variables under study are systematically manipulated (Sekaran & Bougie, 2013, p. 131). As mentioned before, this study makes use of structured, but uncontrolled participant observation, as the natural environment and concealed nature of the observation rule out manipulating the variables through experiments. Summarising, a structured uncontrolled participant observation will be used. In this case, this is a form of participant observation in which the interactions of NW volunteers with people in the neighbourhood during their patrols will be coded into mutually exclusive categories that are defined before the start of the observations.
Table 3
Possible forms of participant observation (made by author).
Controlled Uncontrolled
Structured X
29 Of course, a case study can sometimes be a questionable design to work with, due to the often inherent low external validity, or generalizability, that accompanies it. As Moravcsik mentions: “It is difficult to draw generalizations from a small set of successful cases; the result tends to inflate the number of apparently necessary conditions” (2013, p. 77). While due to their uniqueness this may be true for many cases, the author does not believe this to be of large negative effect on this research specifically. This is because NW schemes are numerous and because most look alike. They might organise themselves slightly different: for instance, some teams will work with rosters, whereas others do not. Also, in some teams, each member has the direct telephone number of the community police officers, whereas in others, one of the volunteers is appointed as the “coordinator” who then keeps in touch with the police. Nevertheless, all physical teams in their core consist of volunteers that patrol the streets and are in contact with the police and the municipality (Municipality of Tilburg, 2019). Still, although the organisation is similar, the neighbourhoods in which these teams operate are not. Generally speaking, this will not cause real problems in light of generalisability. However, some neighbourhoods deviate significantly from the “average” Dutch neighbourhood. Think of the neighbourhood in Boutasmit & Snel’s (2016) research mentioned earlier, where the Dutch Moroccan community was severely overrepresented. The neighbourhoods in this thesis, that will be elaborated upon later in this chapter, are relatively “average” neighbourhoods. They are not very wealthy, not really poor, do not overrepresent certain minority cultures, etcetera. Therefore, the author acknowledges that there may be special or more deviant contexts in which the relationships found in this thesis do not apply, or at least not to the same extent.
3.2 | Tilburg as a case study
In the Netherlands, hundreds of NW teams are active. Let us take the city of Tilburg as an example, as this will be the city that is focussed upon in this research; over 400 volunteers, together comprising many different teams, are currently patrolling the streets in roughly the same format. (Municipality of Tilburg, 2019). Results obtained in this thesis will therefore not be unique to just the NW teams under study in the city of Tilburg, as it will be easy to use them in similar cases. Similar cases can be other neighbourhoods in Tilburg that make use of NW, but also different neighbourhoods in other Dutch municipalities in which NW is active. Tilburg was chosen because in Lub’s research “De burger op wacht”, he mentions that De Reeshof (an area of Tilburg) is nationally renowned as a “best-practice” in light of its organisation and coordination. (Lub, 2016, p. 75). By now, NW has spread across the rest of Tilburg as well.
30 Lub talks about how civil servants from this municipality give workshops in other towns on how to best organise and coordinate NW schemes. Therefore, Tilburg is expected to be an interesting case study from which much can be learnt.
3.3 | On sources and data 3.3.1 | Sources
The information needed to get a clear picture of this case will be extracted mainly through method triangulation, drawing on primary sources. As Moerman states “If you are doing an interview and you are observing, you are using two methods. And if you are focussing that on the same issue … then you can triangulate; even with two methods. So for triangulation, you do not need three methods, you can use two methods” (Moerman, 2018, video 3.2). Primary sources are instances in which the data is obtained first-hand by the researcher, whereas secondary sources already exist; the data has already been gathered by other people (Sekaran & Bougie, 2013, p. 113). The author prefers primary sources over secondary sources for two reasons. First, the method through which the data will be derived can more easily be changed from the usual quantitative methods in research on NW to qualitative methods. This better serves the research goal that was mentioned in the introduction. As mentioned, a significant amount of the research done so far was quantitative in nature. By using data obtained first-hand, this research can view NW from a different angle, without neglecting qualitative methods. Second, for a still young subfield of research, care must be taken in order to avoid relying too much on secondary sources. That would make for a weak foundation for further research in the field. In other fields, like the study of terrorism, the overreliance on secondary material is a major issue (Schuurman & Eijkman, 2013). While this does not have to be an issue per se, we simply do not know yet. When the literature around NW becomes more numerous, time will tell whether this becomes an issue. Until then, it is better to be on the safe side, as it would be a shame to repeat the mistakes from another field instead of learning from them.
3.3.2 | Data collection
The data will be obtained via thirteen interviews with people involved in NW, together with participant observations in two NW teams in the city of Tilburg. The participant observation will be concealed, making for a non-contrived setting with minimal researcher
31 interference; in other words, the author will aim to influence the patrols as little as possible. The choice for a non-contrived setting with minimal researcher interference is made for two reasons. First, it is a logical consequence of concealed observation, as it is difficult to manipulate the variables at play while hiding that one is doing research at the same time. Whereas in experiments variables can easily be manipulated, this is not the case when one is trying to hide the fact that (s)he is doing research. Second, it seems disrespectful to interfere with the work of volunteers, thereby perhaps affecting the results of their work in a negative manner. It is true that a concealed observation has the potential drawback of people feeling “used” without them agreeing to it. “Concealed observation raises ethical concerns since it may violate the principles of informed consent, privacy, and confidentiality” (Sekaran & Bougie, 2013, p. 134). This argument, however, is considered to be of less weight than the other potential drawback of reactivity by the people under observation. Reactivity could endanger the validity of the results and is, therefore, something that should be minimized. It is defined as “the extent to which the observer affects the situation under observation” (Sekaran & Bougie, 2013, p. 133). In order to neutralise the negative effects of concealed observation, no personal data of any kind shall be written down. The participant observation will be a structured observation. As such, although indeed no personal data will be written down, there will be coding schemes involved like the one depicted in on page 35 to keep track of the independent variables. During each interaction with people during the patrols, the author shall check whether the independent variables are present or not and note this down quickly on his phone.
The use of interviews is preferred to a literature study, due to interviews being able to better capture the experiences of NW volunteers on a personal level. On top of that, there is a lack of relevant data available in documents and academic articles on this thesis’ topic specifically. When articles are written about NW, the two main topics are usually evaluations of NW’s effectiveness in certain areas as was mentioned in the introduction, as well as what motivates people to participate in NW schemes4. The interviews will be of a semi-structured type. This term is quite an empty container, as it can mean so many different things (Moerman, 2018, video 4.2). Therefore, it is stressed that in these interviews the amount of control will be high in the sense that most questions are fixed, but low in the sense that there is room for additional comments and questions. The order in which the questions are addressed is not rigid, as long as all topics in the interview protocol on pages 43-46 are discussed eventually. If the
32 interviewee in answering one question also answers another question from the protocol, the other question will no longer be asked.
The participant observation is added because it can corroborate or weaken the data obtained from the interviews by providing an extra, more objective, point of view. Through this, the reliability of the data should be enhanced (Sekaran & Bougie, 2013, pp. 228-229), because one would expect to find the same results in the observation as in the interviews (the same level of respect being practised as was professed, for example). These observations will give the thesis a higher internal validity as well, since the author will join the Tilburg NW scheme himself for several weeks. Thus, the author can personally see whether there are alternative variables that could make for NW’s effectiveness, and whether these have merit. There will also be some secondary data gathered on the objective effectiveness specifically related to Tilburg (like home burglary numbers), as to further clarify the impact that NW has had on the security in the neighbourhoods in Tilburg. These numbers were written down in the next section “the neighbourhoods under review”.
3. 4 | The neighbourhoods under review
The single case study will focus on NW in the neighbourhoods of Goirke, Hasselt, and Bouwmeester (GoHaBo), as well as Noordhoek. The last neighbourhood has a new team, the first three neighbourhoods are all in one older team together, as is shown in Table 4.
Table 4
NW teams (made by author).
Team name Year of establishment Neighbourhood(s)
GoHaBo Since 2014 Goirke, Hasselt,
Bouwmeester Noordhoekpreventie
(control group)
33 Observing/interviewing two teams seems better than one in light of access to the relevant data. There is a need for a team that has been participating in NW for quite some years, as in this area we can see the home burglary numbers per 1000 inhabitants declining rapidly like is shown in Table 5. Noordhoek, having a relatively new team, serves as a control group to the outcomes of GoHaBo. A control group is necessary, for if the study would look at only one neighbourhood, it would make it difficult to say whether a decline in burglary numbers would be due to NW or not. While the control neighbourhood was apparently safer to begin with, its home burglary numbers are also slightly declining. Nevertheless, this trend is significantly less steep than compared to the GoHaBo area that has been having an operational NW team for years already. The numbers from 2010 to 2018 below in Table 5 show that this is indeed the case. Noordhoek, being the control group, is the only one of the four neighbourhoods that did not experience large relative gains when comparing the start and end results of the measurement. This becomes even clearer when converting the start and end results to percentages, as is also done in the calculation in Table 6 overleaf.
Table 5
Home burglary numbers per 1000 inhabitants in four Tilburg Neighbourhoods (taken from CBS, 2019).
Orange = year in which NW was introduced Green = NW at full operational capacity
2010 2011 2012 2013 2014 2015 2016 2017 2018 Noordhoek 4.8 6.9 6.4 6.2 9.3 4.3 4.0 3.9 4.1 Bouwmeester 7.1 7.2 8.9 8.1 7.1 6.4 3.0 4.5 2.9 Hasselt 7.3 10.0 8.0 10.2 12.1 10.6 5.0 4.5 5.5 Het Goirke 10.5 4.6 5.0 8.4 7.2 10.4 7.0 4.5 3.3
The green numbers are more relevant than the orange ones. This is because during the green years, the NW teams are roughly at full operational capacity. For initiatives like NW, the first year can be difficult. The team has to be set up, the number of members needs to be
34 increased, people have to get to know each other, the habit of walking patrols needs to be cultivated, etcetera. As such, the green years show a more reliable image of the effect of NW.
Table 6
Home burglary decline rate with and without NW (made by author, based on Table 5).
Neighbourhood Calculation Outcome With or without
NW Noordhoek (4.1 – 4.8) / 4.8 x 100 - 14.6% Without NW Bouwmeester (2.9 – 7.1) / 7.1 x 100 - 59.2% With NW Hasselt (5.5 – 7.3) / 7.3 x 100 - 24.7% With NW Het Goirke (3.3 – 10.5) / 10.5 x 100 - 68.6 % With NW 3.5 | On analysis 3.5.1 | Content analysis and participant observation
After transcribing the interviews, the themes found in the answers turn into the unit of analysis as the responses related to the independent variables will be coded and categorised by use of content analysis. “Content analysis is a research technique for making replicable and valid inferences from texts (or other meaningful matter) to the contexts of their use” (Krippendorff, 2004, p. 18). The ideal use of this research technique allows other researchers to derive the same results from the same cases, by providing an elaborate coding scheme. Relevant parts of the participant observations will also be singled out and written down, as to provide context to the interviews. This is done by giving examples of the situations the author comes across during the observations, when they pertain to the answers given in the interviews. What is deemed relevant depends on whether the passage involves the independent variables of personal enthusiasm, communication skills, social control, deterrence, or respect in a sense that clearly portrays why these variables have an effect on the dependent variable: effectiveness of NW. Content analysis shall only be applied to the transcribed interviews; not to the passages taken from the observations. The latter would require elaborate notes being taken during the patrols, which is not possible when doing concealed participant observation in order to avoid
35 reactivity. The results of the content analysis and the participant observations will be compared with one another, so that conclusions from this can be drawn. The codebook for the content analysis can be found in Chapter 4 on Data collection.
The variables cannot solely rely on the interview results. The interviews are used in order to establish the importance of the variable. Since it is expected, but of course not sure yet, that the interviewees will consider some of the variables to be more important than others, this is followed up on by a structured observation (Sekaran & Bougie, 2013, p. 139). It is hoped that such an observation provides more insight on the importance of the variables, making it easier to rank them.
3.5.2 | Observing structurally
To see the importance of the variables in practise, NW is observed while performing its patrols. Let us take the variable “respect” as an example. Respect is measured by use of the following coding scheme: each time a volunteer starts a conversation with a citizen during the patrols, this will be noted down. After every conversation, the interaction will be assigned either a green code or a blue code, where blue stands for respect towards the citizen and green stands for disrespect. An example is given in Table 7 below. For the variables of respect and communicative skills, this will be coded per conversation. For the others, it will be coded whenever it occurs. In the Analysis, two tables will be made for the results according to this distinction. The tables will be coloured to make them easier to read.
Table76 (made by author)
Example of a structured observation coding scheme to measure “respect”.
Question: was the person approached in a decent, respectful manner? Blue = yes / Green = no
Conversation 1 2 3 4 5 6 7 8 Etcetera
Code Etcetera
The reason for the separation into two tables is that the latter three variables could manifest themselves in interactions with people, but they do not necessarily have to. Other
