16-U3D1 - Evaluating Broken Windows Theory

Introduction: Canterbury’s Broken Windows and Graffiti ‘Scourge’

Graffiti is regarded a problem in the city of Canterbury. This small, largely wealthy and relatively serene Cathedral city in the south-east of England, with a population of around 70,000 people and a local economy that largely runs on tourism, its night-life and its student populace, seemingly considers itself to be under threat from forms of street art that inhabit the city’s walls (This is Kent, 22/1/2009). Though not intrusive, the sight of graffiti is still part of the everyday experience for those who walk the city, where most appears either in small clusters or isolated tags and is noticed only by those who keep their eyes open for it. Though its graffiti scene therefore bears little resemblance to the street art meccas of New York, London, Bristol, Berlin or Melbourne, the police and local council, however, believe graffiti to be a serious local issue. In an effort to alleviate the apparently ‘disgraceful state of this graffiti on the walls of our city’ (Adam Parsons quoted in The Kentish Gazette, 2/6/2011: 14), authorities have sought to remove graffiti where possible and aggressively police and punish writers in order to deter those responsible for this local concern. This has included the swift removal of graffiti in the city with a Graffiti Busters service (Canterbury City Council, 2012), a free hotline for reporting writing, raids on suspected writers’ homes (This is Kent, 8/4/2011), and the arrest and prosecution of writers at every given opportunity (Claridge, A. and Alston, K., 18/3/2010, This is Kent, 12/2/2010, 19/1/2012). This attempt to send out a ‘firm message that we take this antisocial behaviour seriously’ (This is Kent, 8/4/2011) has also seen a verbal assault launched against graffiti whereby the acts of writers are described as harbingers of social decline and urban decay. Consider this statement on the city council’s website:

‘Graffiti undermines pride in the local community. Graffiti is often the first element in a spiral of decline and creates a negative impression of an area. It also contributes to the fear of crime’ (Canterbury City Council, 2012: 1).

Such a disparaging attitude to graffiti is also shared by many senior members of local communities, some of whom have started their own anti-graffiti campaigns (see Napler, 19/5/2011 and The Kentish Gazette, 2/6/2011: 14), suggesting that the ‘scourge of graffiti’ (Herbert Pragnall quoted in The Kentish Gazette, 25/3/2010: 9) is already damaging the local area:

‘This graffiti is detrimental…because it sends out the signal that no one cares, and that it's okay to throw some more rubbish down…It's not only rubbish which is increasing in the area but dog fouling as well, because it makes the area look rundown’ (This is Kent, 11/3/2011).

Such descriptions of graffiti’s effect on its immediate surroundings clearly echo the logic found in Broken Windows theory, originally articulated in Kelling and Wilson’s (1982) seminal article in Atlantic Monthly. Famously, or perhaps ‘infamously’ (Wasterfors, 2009: 92)), suggesting that crime is likely to ‘occur anywhere once communal barriers—the sense of mutual regard and the obligations of civility—are lowered by actions that seem to signal that "no one cares"’ (Wilson and Kelling, 1982: 33), Broken Windows rationale has seen graffiti writing characterised as a sign of incivility that will attract crime to locations in which it is prevalent (see Sloan-Hewitt and Kelling, 1990). According to Wilson and Kelling, graffiti effects its environment by making it appear ‘uncontrolled and uncontrollable, and that anyone can invade it to do whatever damage and mischief the mind suggests’ (1982: 34). Authorities in Canterbury appear to take such ideas seriously and have adopted elements of so-called ‘Broken Windows policing’ (Hinkle and Weisburd, 2008) in their local war on writing; an approach to crime control born from the logic articulated throughout this thesis. Such a policing programme, based on the central tenets of Kelling and Wilson’s hypothesis, has generally involved an increase in manpower and resources oriented towards policing minor misdemeanours, under the guise that this will herald a reduction in disorder and more serious crime (McLaughlin, 2008): ‘where a window is likely to be broken at any time, it must quickly be fixed if all are not to be shattered’ (Wilson and Kelling, 1982: 37). In Canterbury this has equated to local authorities intervening at the ‘first element in a spiral of decline’ (Canterbury City Council, 2012: 1) represented by graffiti and other signs of low-level disorder in the city. This resulted in a multi-agency approach between the local council and police force, with even traffic wardens being invested with the power to issue on the spot fines for writers or those who litter in the city (Warren, 23/8/12).

After its perceived successes in New York City and across America (Harcourt and Ludwig, 2006), Broken Windows theory and policing appear to have garnered the approval of many police practitioners the world over (Wacquant, 2003). However, the premises of Kelling and Wilson’s argument continues to inspire debate amongst criminologists, many of which remain unresolved. Many scholars remain either sceptical towards or vehemently critical of the theory’s assertions, with its claims that acts of so-called ‘disorder’ both increase fear of crime and criminality in their local context regularly being called into question. Moreover, the effectiveness of Broken Windows policing remains part of on-going debates, where it has been suggested that stories of the policy’s success have been vastly overstated (Harcourt, 1998, Harcourt and Ludwig, 2006). Among many cultural criminologists there remains little doubt that those in political power have utilised Broken Windows as a ‘convenient foundation’ (Ferrell, 2005: 148) for ‘assign[ing] the blame for street crime not to poverty and marginalisation, but to the poor and marginalised’ (Ferrell cited in Snyder, 2009: 6), and implicitly linking disorderly behaviours to an ‘underclass’ way of life’ (Squires, 2008: 15), and in doing so propping up their own popularity (Snyder, 2009: 24). Indeed, this has been suggested to generally be the case with authorities’ approach to graffiti writers, and perhaps more so than to any other deviant group: Austin’s (2001) and Ferrell’s (1996) analysis across New York and Denver respectively highlight the mobilisation of Broken Windows against writers at a time when each city’s mayor stood to acquire political capital by claiming writers to be the source of several social problems. In light of such scepticism, this popular policing approach has been suggested to be based on ‘folk wisdom’ (Matthews, 1992: 20) and ‘the political and economic goals of those who orchestrate’ such approaches (Snyder, 2006: 95): representing ‘a taken for granted assumption’ amongst policy practitioners (Crawford cited in Beckett and Godoy, 2010: 279), rather than a police derived from any robust reasoning or evidence.

The Research

It is therefore important to determine whether Broken Windows does accurately describe graffiti’s impact on its urban surroundings. The combination of the policy relevance of these discussions and the relatively nebulous body of existing evidence on this topic, makes these issues extremely pertinent and relevant to intertwining discussions of theory and policing practice in contemporary criminology. This research, then, seeks to test whether Broken Windows theory provides the best theoretical framework for understanding the relationship between graffiti and wider crime in Canterbury. In doing so, it will also reflect on whether the Broken Windows policing model employed across the city offers the most effective way of reducing the city’s crime rates. Pertaining to issues of graffiti writing, this study is centred on an attempt to bring some clarity to the aforementioned debates by answering two questions:

· To what extent does Broken Windows theory provide an accurate understanding of graffiti writing’s effect on its immediate environment in Canterbury?

· To what extent is a focus on graffiti as part of Broken Windows policing in Canterbury likely to be effective in reducing crime as a whole?

It is hoped that this thesis will make both a theoretical and methodological contribution to the body of literature surrounding Broken Windows theory. As well as shedding light on the aforementioned discussions, this mixed-methods study will utilise a mixed methods approach to the testing of Wilson and Kelling’s hypothesis by combining the quantitative analysis found in most of the literature with a more contextualised, street-based approach (similar to that previously used by Taylor, 1999, 2001). Alongside the use of crime maps and photography of graffiti on the streets of the city to unearth correlations between graffiti and crime across the city, a series of smaller case studies will be analysed in order to develop a richer sense of context around graffiti writing: whether (and why) graffiti is largely visible or hidden, as well as the likely consequences for this in terms of its impact on its immediate environment. The combination of these approaches is intended to provide a broad picture of graffiti and crime’s distribution throughout the entire city, whilst also getting to the root of the specific nature of the relationship between graffiti and crime through these more detailed case studies.

As will be revealed in the upcoming chapters, the findings of this research project give reason to be highly critical of Broken Windows theory; as they demonstrate the theory’s central premises to have little correspondence with the apparent relationship between graffiti and crime throughout the city. Both in the quantitative analysis, from which no meaningful correlation between graffiti and other forms of crime is found, and the analysis of case studies throughout the city, little evidence was found to suggest that the ‘developmental sequence’ towards urban decay described by Kelling and Wilson, in which ‘disorder and crime are… inextricably linked’ (both quotes 1982: 32), had begun or was likely to begin. Deriving from these findings it has also been suggested that there is little empirical foundation for Broken Windows policing in the city, with it being argued that attempts to reduce crime through controlling graffiti are therefore likely to be ineffectual. Alternatively, as the thesis progresses, a variety of cultural criminological arguments, which sheds light on graffiti’s location within mainstream culture, the unintended effects of Broken Windows policing on practices of writing and the pluralistic uses of particular urban spaces, emerge as a much more serviceable theoretical framework for explaining these findings. This prompts a reconfiguration of the way we understand graffiti’s impact on its immediate environment and the effectiveness of attempts to effectively police it through a Broken Windows model, emphasising the need for us to move beyond Wilson and Kelling’s ideas in order to understand, and develop policies towards, graffiti writing and policing in the city of Canterbury.

Chapter II:

Literature Review

Before delving into the findings of this research project, it is important to outline what is currently known about Broken Windows theory and policing, including: the arguments that put forward this theory; tests of its accuracy and the effectiveness of its associated policing practices and, then; some of its main critiques in criminological literature. This literature review will detail each of these bodies of work and will, throughout the chapter, make clear how they relate to existing scholarly knowledge on graffiti writing’s impact on its immediate environment. It will also flag the contributions this study aims to make to existing knowledge on Broken Windows theory and graffiti writing. As this literature review will show, there is no shortage on works commenting on, critiquing or testing the Broken Windows thesis. What we must consider, however, is what these works collectively tell us and whether there remain unanswered questions and gaps in knowledge in this area of study that this research can viably address.

The “Broken Windows” Thesis

Broken Windows theory’s brand of ‘environmental criminology’ (see Bottom and Wiles, 2002), as originally found in Wilson and Kelling’s (1982) seminal article (‘arguably one of the most influential and widely cited articles in North American criminology’ (McLaughlin, 2008: 27)), constructs a model of crime causation that identifies low-level disorder as a cause of more serious offences. Providing what has been described as ‘an alluring and elegant simplicity’ (Willis, 2002: 220) in their portrayal of an area’s sequential descent into unmanageable chaos, Wilson and Kelling’s thesis offers a description of how ‘serious street crime flourishes in areas in which disorderly behaviour goes unchecked’ (1982: 36). Characterising forms of incivility as ‘symbolic invitations to more serious forms of criminality’, this perspective itself suggests that initial disruptions of social order may pave the way to ‘a downward spiral of criminal disorder’ (both quotes Ferrell, 2005: 148) that causes community members to withdraw from order maintenance, as fear of crime increases, and leave the area ‘vulnerable to criminal invasion’ (Wilson and Kelling, 1982: 32). Claiming that such visible signs of disorder ‘create a negative impression of an unmanaged and uncared for environment’ (Department for Transport, 2010: 3), it is argued that they therefore attract offenders from outside the area who sense that it has become a vulnerable and less risky site for crime. Wilson and Kelling continue:

‘it is more likely here, rather than in places where people are confident they can regulate public behaviour by informal controls, drugs will change hands, prostitutes will solicit, and cars will be stripped. That the drunks will be robbed by boys who do it as a lark, and that prostitute’s customers will be robbed by men who do it purposefully and perhaps violently. That muggings will occur’ (Kelling and Wilson, 1982: 33).

Hinkle and Weisburd’s (2008) representation of the sequential decline described in Wilson and Kelling’s Broken Windows thesis.

Amongst the acts of so-called ‘disorder’ that have been characterised as catalysts for such sequential decay by Broken Windows scholars and various public officials, graffiti appears to have assumed a leading role. Examples of this in action have ranged from the spectacular association of graffiti with the most heinous of criminal acts, ‘obviously, murder and graffiti are two vastly different crimes… But they are part of the same continuum, and a climate that tolerates one is more likely to tolerate the other’ (Giuliani, 1998: 1), to less a hyperbolic belief that ‘graffiti not only damages property but actually makes public space more dangerous by turning neutral space into crime space’ (Snyder, 2009: 48). Regularly referred to as a harbinger of bad news for urban locales, as a crime that ‘can ruin lives and create an environment where more serious crime can take hold’ (HM Government, 2010: 2, see also Sloan-Hewitt and Kelling, 1990), graffiti is often characterised as the quintessential threat to social order in the eyes of those who subscribe to the Broken Windows account of social decline in urban space. As one Swedish report argues:

‘Illegal graffiti is the visual impression of an uncaring and indifferent society, where small crimes can lead to bigger crimes. Without exception, failure to identify the whole problem accurately and take a proactive approach in the early stages encourages illegal graffiti to continue to escalate until resources or the cost of effective control is beyond the means of any administrations’ (quoted in Department for Transport, 2010: 5).

The latter stages of the above quote allude to the strategies that have been employed by policing practitioners in an effort to stave off the processes described in Wilson and Kelling’s influential article. So-called ‘Broken Windows policing’ (Hinkle and Weisburd, 2008), also referred to elsewhere as both ‘zero-tolerance’ and ‘quality of life’ policing (McLaughlin, 2008a), extends from the theory’s logic by embracing ‘the order maintenance function of the police’ (Wilson and Kelling, 1982: 35) and aggressively enforcing minor misdemeanour laws in an effort to bring about a reduction in initial symptoms of disorder and, thus, prevent a downward spiral of urban decline (Jang et al, 2008). Simultaneously, the development of new technology to aid the police, including developments such as COMPSTAT and the use of digital technology for surveillance as well as increased manpower, has generally been central to this approach (McLaughlin, 2008a, Snyder, 2009). First applied in William Bratton’s attempt to ‘reclaim the [New York City] subway system’ (Bratton quoted in McLaughlin, 2008: 59) from criminality, Broken Windows policing became ‘an integral part of [the] law enforcement strategy’ (Giuliani, 1998: 1) throughout the whole city in the 1990s once Randolph Giuliani was elected Mayor and Bratton appointed police commissioner. At present, the New York Broken Windows model has since become the default method of policing across America more generally, also being used Chicago and Los Angeles (Harcourt and Ludwig, 2006), and, after perceived success of this approach in reducing crime rates throughout the 1990s, the dominant framework for reducing and preventing crime in urban space across Europe and many places elsewhere (Wacquant, 2003)[footnoteRef:1]. Moreover, this has given the theoretical ideas that underpin this approach legitimacy in the eyes of many policing practitioners and public officials, with the original articulation of Broken Windows theory now standing as one of the most cited articles in policing literature (Manning, 2001). [1: That the scholarship of J.Q. Wilson should have a significant policy impact is of no surprise. Having served on White House task Force on Crime in 1966 and the US President’s Foreign Intelligence Advisory Board (1985-1991), ‘Wilson’s scholarship stood in stark contrast’ (Delis, 2010: 192) critical criminologies that dominated the disciplines throughout the 1960s and 70s. An opposition that Wilson himself was well aware of (see Wilson, 1975).]

Testing the Broken Windows Thesis

Whilst Broken Windows ideas have received acclaim from those working in policy circles (for example Giuliani, 1998, Home Office, 2010), debates have rumbled on regarding the accuracy of the theory’s account of crime causation in urban spaces among criminologists and remain largely unresolved. Indeed, in scholarly circles the theory has received a mixed reaction and has been the subject of longstanding processes of evaluation, often based around various forms of statistical analysis, in an attempt to quantify the accuracy of Wilson and Kelling’s claims. Such tests of Broken Windows theory have, thus far, taken on one of two forms: either looking for correlations after measuring neighbourhood disorder and crime, or measuring of the impact of misdemeanour policing on crime rates (Harcourt and Ludwig, 2006). What follows is a brief account of this literature; the key studies and themes that have emerged in the past 30 years from the testing of Wilson and Kelling’s original claims. Moreover, it will summarise what they ultimately tell us about the accuracy of the Broken Windows thesis.

Alongside a body of literature that lends support to the claim that signs of disorder impact upon residents’ fear of crime and their involvement in the local community (see Pitner et al (2012), Chappell et al (2011)), there are a number of studies that concur with Wilson and Kelling’s claims regarding disorder and more serious crime by and showing Broken Windows policing to have a desirable impact on crime rates. For instance, we may look to Sloan-Hewitt and Kelling’s (1990) claims of success on the policing of the New York underground, as well as Golub et al’s (2003) interviewing of arrestees in New York found that almost half had reported stopping or cutting back on disorderly behaviour because of Broken Windows policing (though obviously not enough to avoid being arrested!). Kelling and Sousa’s (2001) attempt to quantify the impact of Broken Windows policing led them to look at the entire city’s crime rates and, by estimating that for every 28 misdemeanour arrests there was one violent crime prevented, claimed quality of life policing to have prevented 60,000 violent crimes between 1989 and 1998. These studies, and also other works (such as Zimring, 2007, and Corman and Mocan, 2005), have attributed a large share of the New York crime drop in the 1990s to this approach of policing. Outside of New York, Wesley Skogan’s (1990) widely-cited analysis across 6 American cities found similar a chain of fear of crime being linked to disorder, and disorder being a good predictor of robbery victimisation. Similarly, Thomas et al’s (2011) analysis of arson in Michigan found ‘that proxies of social disorder can be used to identify areas of high arson risk, consistent with the Broken Window theory’ (2011: 271). Such studies, however, have often been criticised for the quality and the nature of their analysis. Re-examining Skogan’s data set, for example, Brian Harcourt (1998) found that the conclusions originally offered by the author were skewed by the influence of a few extreme cases and locations, finding instead that there was an insignificant relationship between disorder and serious crime in the vast majority of locations. In addition, Harcourt and Ludwig’s (2006) analysis prompts caution towards Kelling and Sousa’s (2001) findings by noting that their analysis had been skewed by a failing to control for crime rates prior to 1990[footnoteRef:2]. As these cases show, despite the evidence pointing towards Wilson and Kelling’s thesis and its cognate programme of Broken Windows policing being of some value, there remains reason to remain wary of the these claims and the evidence used to support them. [2: Citing mean reversion, Harcourt and Ludwig note that the biggest drops in the data set came from areas that experienced the largest rise in crime in the 1980s. They observe that ‘jurisdictions with the greatest increases in crime during the 1980s tend to experience the largest subsequent declines as well’ (2006: 276) regardless of whether Broken Windows policing was used in these areas or not, and that the effect of such policing was vastly been over-estimated by Kelling and Sousa (2001).]

In addition, other analyses have looked to a context wider than New York City to suggest that the impact of Broken Windows policing may have been overstated by the likes of Kelling and Sousa, and Corman and Mocan (see Barker, 2010, Blumenstein and Wallman, 2006). For example, Eck and Maguire (2006) note that the crime rate fell similarly in San Diego, Seattle and Dallas throughout the 1990s where Broken Windows policing was not adopted. Furthermore, they show that the crime rate in Seattle fell 18% with a 6% decrease in manpower, likewise in Dallas where a 39% reduction in crime came with a 3% decrease in policing numbers. Moreover, Yang’s (2007) discovery of a negative correlation between incidents of violent crime and disorder in Seattle, and Sampson and Raudenbush’s (1999, see also 2001, 2004) finding that correlations between disorder and crime were moderate at best, once more bring the theory’s claims into question and suggest Broken Windows policing to be ‘analytically weak strategy to reduce crime’ (Sampson and Raudenbush cited in Harcourt and Ludwig, 2006: 284) (see also Cerdia et al, 2009, Karmen, 2000).

Reflecting on this collection of work as a whole, including those studies that support and cast doubt on the Broken Windows thesis, it appears that the quantitative studies conducted thus far tell us relatively little in terms of clarifying the merits of this theory. Presenting an ambiguous picture, where various statistics are pitted against one-another without either perspective ultimately disproving the other, this body of literature merely gives us reason to be sceptical of some of the theory’s claims and the hyperbole surrounding claims of success in New York City (where it is claimed that declining crime rates were ‘proving the Broken Windows theory’ (William Bratton quoted in Corman and Mocan, 2005: 239)). It is at this point, then, that the chapter will progress from the detailing of these tests of Broken Windows relevance to understanding urban crime, and towards theoretical and contextually-driven criticisms of Wilson and Kelling’s thesis based on qualitative research. As will be revealed below, it is in this body of literature that a great deal more is revealed about the problems in the Broken Windows thesis.

Critiques of Broken Windows

Before we delve into criticisms of Broken Windows theory that arise from street-level qualitative research, we can see how Broken Windows’ separation of ‘disorder’ and ‘crime’ has been criticised. Arguing that the theory fails to establish meaningful and robust concepts of ‘crime’ and ‘disorder’ (Gau and Pratt, 2010) by re-defining forms of low-level crime (such as prostitution, public drinking and vandalism) as ‘disorder’, a term without any legal implications or inherent suggestion of law-breaking, Gau and Pratt note that Wilson and Kelling’s theory, in actuality, fails to make any meaningful statement about how factors outside of criminal behaviour may encourage criminality (as suggested by Gau and Pratt, 2010). In this sense the theory provides an imagined sequence of progression from ‘disorder’ to crime, whereas in reality there was only ever ‘crime’ all along. This therefore means that the aforementioned studies that have subsequently suggested disorder and crime to be positively correlated often tell us very little; as such terms often mean the same thing, thus undermining their usefulness even further!

In addition, looking back on the body of empirical evidence preceding their work, Gau and Pratt remark that ‘prior studies have linked levels of structural and social disadvantage to disorder and crime in a manner suggesting that disorder may simply be one component of a larger condition of… concentrated socio-structural disadvantage’ (2010: 763). That is, social structure resembles a lingering factor that appears to explain the emergence of both disorder and crime. As Sampson and Raudenbush (1999) put it, disorder is of the same aetiology as crime: ‘rather than conceive of disorder as a direct cause of crime, we view many elements of disorder as part and parcel of crime itself’ (quoted in Harcourt and Ludwig 2006: 285). In light of these critiques, ‘disorder’, as found in Broken Windows theory, is exposed as a red herring: a theoretical disguise for both ‘the structural characteristics of the neighbourhoods, as well as neighbourhood cohesion and informal social control – not levels of disorder – that most affect crime’ (Sampson and Raudenbush quoted in Department for transport, 2010: 6).

These valuable and telling critiques aside, a variety of street-level studies have also been shown to undermine Wilson and Kellings thesis with similar effectiveness. Emerging from cultural criminological critiques of Wilson and Kelling’s theory, and supported by the findings of Taylor (1999, 2001), other scholars have called for a greater appreciation of the multiple meanings that forms of so-called ‘disorder’ can carry than Broken Windows theory offers. After all, if ‘disorder’ is not interpreted as such by those who encounter it, then Wilson and Kelling’s account of urban decline will surely fail to materialise. Perhaps the greatest example of a crime that is often ambiguous in its meaning is indeed graffiti writing, which ‘continues to spread like some mutating double-helix, confounding art, crime and commodity’ (Ferrell et al, 2008: 151) in the late-modern age and actively demonstrates that acts of criminality can be simultaneously be disdained by various individuals, embraced by some and met with indifference by others. While graffiti may indeed seen as a sign of undesirable things to come (by some), for niche and edgy shopping districts graffiti may serve to ‘tell shoppers that this is a cool space’ (Snyder, 2009: 53), and for others a sign of potential urban regeneration:

‘A proliferation of hip-hop graffiti in place of gang graffiti (a distinction ignored by Wilson and Kelling) may likewise suggest a decline in criminal violence – that is, it may lead some neighbourhood residents to understand that gang crime is on the decline – and in fact it may harbinger a less violent social order now negotiated through the very symbols that Wilson and Kelling so tellingly misrepresent’ (Ferrell et al, 2008: 105).

It becomes apparent, then, that the ‘decontextualisation of disorder’ (Gau and Pratt, 2010: 759) within a Broken Windows model that ‘constructs a series of abstract, one-dimensional meanings that it arbitrarily assigns to dislocated images’ (Ferrell et al, 2008: 104) of urban crime is not one that is capable of grasping the complexities and contradictions in graffiti’s, and perhaps other examples of so-called disorder’s, meaning to urban audiences:

‘None [are] self-evident or assured; each is negotiated day-to-day amidst the swirl of urban populations. Because of this, accurately understanding public display as a key nexus between crime and culture requires attentiveness and long-term emersion in everyday situations – not [Broken Windows’] abstract political posturing’ (Ferrell, 2005: 149).

Ferrell’s call for a contextualised analysis that can dig deeper into the subtleties of crime has not only extended to a cultural criminological appreciation of the meanings applied to graffiti and other crimes; it also relates to different approaches to ascertaining what environmental factors encourage particular criminal behaviours. Departing from the detached position adopted by ‘the milieu of social researchers who choose not to look at the world’ Emmanuel David (2007: 251), Ralph Taylor’s studies of Baltimore (1999, 2001) where a mixture of on-site assessments, maps, census, crime figures and interviews with residents and community leaders were used, provide a good example of such an approach. Finding that while some types of incivilities were related to serious crimes and many others were not, Taylor identified a variety of separate problems of crime without a common underlying cause or remedy. In short, the idea of disorder underpinning the spread of urban crime has been suggested by Taylor to be false. Elsewhere, Broken Windows theory’s explanation of criminal behaviour itself, where criminals are presented as rational actors in the process of risk-avoidance whilst committing crime (therefore leading them to low risk ‘disorderly’ areas), has similarly been shown to be equally troublesome. The work of Peter St. Jean is useful here for demonstrating how Broken Windows theory fails to account for the driving forces behind the choice of location for criminal acts. Pockets of Crime (St. Jean, 2007) shows how decisions regarding where to commit crime are often based on characteristics of the area that bear little relevance to disorder; where the area’s impact on the deniability of the crime, footfall, demand for drugs, or ready distractions are found to be more integral to understanding where crime takes place than the prevalence of disorder in an area. Thus while the perception of a low-risk environment created by signs of disorder may attract some crime, St. Jean’s street-based analysis demonstrates that some strategic decisions made by criminals actually lead them away from disorderly areas. Ethnographers of graffiti writing subcultures have identified similar dynamics in the behaviour of writers to those explained by St. Jean, with locations for writing regularly being shown to be influenced less by perceived disorder and more so by an intended target audience. Snyder’s (2009) analysis of New York’s graffiti scene provides a nice example of this and suggests that writers will often target more affluent areas (such Manhattan) or rail-side locations in order to be more widely seen by the people of their city and by other writers. Thus, it appears that it is not only Wilson and Kelling’s understanding of how these acts are seen by city dwellers and the meanings that they carry: the same criticism applies to their understanding of criminal acts themselves. Moreover, it seems then that the kind of criticisms levelled at Broken Windows theory by cultural criminologists (and also Ralph Taylor and Peter St. Jean), by providing examples of how and why the theory is limited in particular regards, tell us significantly more about the accuracy of Wilson and Kelling’s thesis than the aforementioned body of quantitative analysis. What is required then, in order to move knowledge forward in this area, is something that lies closer to a cultural criminological analysis of urban space and crime: one that adopts an interpretive epistemology that allows the scholar to understand the multiple meanings that signs of criminal acts can carry, and is attentive to the context of the criminal acts and the visible evidence it leaves behind.

How This Study Builds on the Literature

It seems then that work in the vein of Snyder, Taylor, St. Jean and also Mitchell Duneier (2000) provides the best means of addressing gaps in knowledge surrounding Broken Windows theory and policing; where the waters remain muddy regarding each’s respective accuracy and effectiveness. As has been outlined throughout this literature review, the current body of knowledge on the accuracy of Broken Windows theory remains relatively unclear: the existing battles between various statistics, data-sets and formulae has failed to give a clear idea of the theory’s strengths and the successes of Broken Windows policing. This requires of criminologists that such tests should be continued to establish greater clarity on the issue, however greater awareness of the contexts in which crime and disorder lie must be taken in order to provide useful answers to such unresolved questions. This research intends to do just this, and to add to an emerging body of literature that tests the claims of Broken Windows theory by testing the links between ‘disorder’ (in this case pertaining exclusively to graffiti writing) and crime. In doing so it also seeks to contribute to understandings of the effectiveness of Broken Windows policing. With its remit being strictly focused on graffiti writing, and the city of Canterbury, it should perhaps be viewed as yet another piece in a larger puzzle of discerning how accurate Broken Windows theory can be considered to be. Given that the ideas expressed in the Broken Windows thesis continue to exert a heavy influence over contemporary criminal justice practice (Becket and Godoy, 2010), it remains criminology’s task to continue to test and critique this theory, as well as offer alternatives where it is shown to be insufficient in explaining the reality of crime and disorder in urban neighbourhoods.

As well as a body of literature that both tests and critiques Wilson and Kelling’s highly influential claims, this study will also add to knowledge related to understanding where graffiti tends to be written, the impact it has on its surroundings and the relationship between these dynamics. In particular it will also contribute to a tentative body of literature that reflects on changes in the behaviour of writers amidst aggressive policing of their activities, the emergence of digital technology being used to share images of writing (see Snyder, 2006, 2009), and shifts in graffiti’s meaning in mainstream culture (Ferrell et al, 2008). Moreover, with this contextualised analysis it will re-evaluate whether Broken Windows, in light of these potential changes, is more or less equipped to explain the effect of graffiti’s impact on its area as a visible sign of ‘disorder’.

The chapters that follow will detail the study’s findings, how they were collected and analysed, and how they relate back to this body of literature. Beginning with a methodology chapter that will detail the research design and explain how it relates to these discussions, the thesis will progress to reveal its findings whilst constantly looking back to the body of literature detailed here as a reference point. Thus, the arguments and studies detailed here should be kept in mind, and will be referred to, throughout the duration of the rest of this thesis.

Chapter III:

Methodology

Having acknowledged the gaps and ambiguities within existing knowledge on Broken Windows theory’s (Wilson and Kelling, 1982) accuracy and the effectiveness of Broken Windows policing, here I will outline the methods used in this project’s attempt to answer its research questions and help establish some clarity on the topic. What follows, then, is the detailing of: the method used; the decision making and interpretation involved in several stages in the research; some of the shortcomings previously identified in the use of such methods; and how such shortcomings have been compensated for within aspects of the methodology engaged here. Following in the vein of Ralph Taylor’s (1999, 2001) use of various methods to come to understand the validity of Broken Windows theory in Baltimore, this chapter will detail how the mixed methods approach adopted here, which utilises examples of both quantitative and qualitative approaches, provides the best means of answering our research questions. As was concluded at the end of the last chapter, studies that have pursued a purely method have both individually and collectively told us relatively little about the accuracy of Broken Windows theory and the success of its associated policing practices. Proceeding in two sections, organised around the two respective stages of quantitative analysis and qualitative research conducted as part of this thesis, this chapter will also outline how the research conducted here provides the broad picture and small-scale specificity required to address this issue.

Quantifying and Mapping Crime and Graffiti: The Search for Correlations

In the first stage of this research, a combination of crime maps populated with recorded crime data and photography will be utilised to first quantify crime and graffiti in Canterbury, before identifying correlations between these variables. Though it is therefore a quantitative method used here, this is not to be confused with a strictly ‘scientific’ process by any means: it instead involves subjective decision-making and interpretation on the part of the researcher at several stages[footnoteRef:3]. Thus, this section will not only reveal the specifics of this methods, but it will also acknowledge where, how and why such judgements have been made and acted upon, as well as how weaknesses in this method have been confronted. It will, of course, also highlight the value of the insights that can be gained through the use of this approach. [3: Furthermore, the subjectivity within crime statistics is recognised. As Lea and Young (1942: 12-13) put it, ‘both the interpretation of whether or not an act is criminal and the resources available for labelling acts as criminal make an enormous amount of difference to the final statistic. Nor is this a result of an imperfect world where, if only we had more tightly written laws and a lot more police, we could arrive at the ‘true’ crime rate. Rather, it lies in the nature of crime itself that there is an important level of human interpretation in the definition of what is criminal’]

Recorded crime statistics were drawn from the newly-launched police.uk website and used to quantify crime within 10 selected areas of the city for the 6 month period of December 2011-May 2012. The designation of these 10 zones was based on the researcher’s decision making, though underpinned by an attempt to replicate the conditions being tested to those described in the Broken Windows thesis. The leading imperative was that each of the ‘zones’ designated for this research project be roughly of equal size and, where possible, their boundaries were selected so that each would represent an identifiable part of the city or a neighbourhood/estate as they are regarded locally. Thus, zones include the Knight’s Avenue Estate, Spring Lane Estate, or the city centre which is separated from the rest of Canterbury by the medieval city walls. This was done in an effort in to replicate, as accurately as possible, the conditions described by Wilson and Kelling in their article; where the emphasis is placed around their being a manifestation of communities or ‘communal barriers’ (1982: 34) that are affected by fear of crime and, then, criminal invasion. By trying to make these zones applicable to areas in the town that are likely to consider themselves communities, such as a self-contained housing estates, it is hoped that this has been successfully replicated somewhat so that the analysis provides the most appropriate framework for testing this thesis and answering the research questions.

The process of quantifying graffiti was similarly based on interpretation, though a logical one, as there appeared no self-evident method of quantifying writing objectively. Indeed such problems in enumerating graffiti have been previously acknowledged to be troublesome. As Greg Snyder notes, the very nature of graffiti makes it hard to quantify exactly the amount in any given area: while some is plainly visible, some remains intentionally hidden in concealed locations and will often evade the sights of local residents, authorities and touring researchers. While it is therefore ‘difficult to compare the amount of graffiti in an exact quantitative fashion’ (2009: 51) in a study of this nature, exploring these areas by foot and documenting writing via photography remains the most viable way of doing so. Bearing this in mind, other scholars have sought to use of photography as a means of quantifying writing in urban spaces; where previous examples include Snyder’s (2009) photography of mailbox graffiti in New York, as well as Heitor Alvelos’ (2004) long term study of urban graffiti using similar visual methods. Marked on the map shown below are, as well as the boundaries of each zone being studied, the locations in which graffiti was found on walking tours of the city. These are represented by black dots. Dots were marked for locations where there was interpreted to be a noticeable and substantial amount of graffiti: locations in which there were one or two relatively small tags were ignored, whilst anything upwards from 3 or 4 tags or any larger pieces and murals (or a combination of these types of writing) visible from the same spot saw the marking of a dot on the map. This was based on a reasoning that again related back to the Broken Windows thesis: it was judged that an area with a more dense collection of writing would be more likely to give the impression of repeated criminality that would inspire fear of crime in residents, thus potentially giving the impression of an ‘uncontrolled and uncontrollable’ environment (Wilson and Kelling 1982: 34), whilst an isolated tag could be more easily dismissed or disregarded by locals. Thus, to categorise a single tag on a par with more conspicuous clusters of writing would both misrepresent the prevalence of writing across the city and see the research divert from the nature of Broken Windows claims. In addition, to avoid marking locations too close together and misrepresenting one pocket of writing as two clusters, if two examples of graffiti were within seeing distance of one another (this often constituted a distance of no more than around 50 metres) then only this only counted towards the marking of one dot.

Following the quantification of each of the variables (‘graffiti’ and ‘crime’) through these 10 areas of the city, analysis of this data was conducted to determine what correlations lie between these two measurements; whether meaningful relationship between graffiti and crime exist across the city. SPSS was used to process the data and to conduct this analysis. This formed the extent of the quantitative element of the project and will serve to test the basic premise of Wilson and Kelling’s thesis by revealing whether, through an observed positive correlation, it appears the graffiti and crime tend populate the same areas (as is implied by Broken Windows logic). The value of such an approach, regarding obtaining answers to our research questions, is that provides a broad picture of the relationship between crime and graffiti across the city: offering basic indications of the theory’s accuracy concerning graffiti’s causal relationship with wider crime, and a testing the empirical foundation on which Broken Windows policing is allegedly based.

The map of Canterbury produced through the first stage of this research, where the demarcations for each of the 10 zones and pockets of graffiti have been plotted

Zone Key: 1 = City Centre, 2 = Wincheap, 3 = Knight’s Avenue Estate, 4 = St. Dunstan’s Area, 5 = St. Stephen’s Area, 6 = Hales Place Estate, 7 = Vauxhall Road Estate, 8 = Military Road Area, 9 = Spring Lane Estate, 10 = Old Dover Road Area

The Shortcomings of Official Data and Crime Mapping

While there is evident value in using both recorded data and forms of mapping to obtain an idea of crime levels across Canterbury, in a process offers that practical advantages of convenience and a lack of expense and gives a broad view of the relationship between graffiti and crime, potential problems do arise from the use of official data and forms of crime mapping that underpins this approach. These problems have, however, been acknowledged and compensated for in the methodological design of this research, meaning that such shortcomings’ impact on our ability to effectively answer the research questions remain limited. The nature of these problems, as well moves to restrict their negative impact and offer solutions to them, will be outlined in this section.

Police.uk’s mapping of recorded crime data across the UK’s streets, which maps crime for each street in the UK and makes data freely available to any individual interested in ‘identifying crimes right on their doorstep anywhere in the country’ (Whitehead, 2011: 1), presents new opportunities for criminologists interested in researching crime rates by allowing researchers (and the public) to chart the distribution of crime across and within an entire city at the click of a few buttons[footnoteRef:4]. Whilst this innovative method of presenting data freely on this internet offers convenience for scholars interested in studying crime rate, pre-existing concerns regarding the validity and reliability of official data dictate that we must continue to exercise caution in how we use such crime statistics in research. Amongst criminologists and statisticians alike, the long-standing criticism of crime figures is that they ‘do not give a clear picture of the social or situational context of crimes’ (Maguire 2002: 343) and remains a powerful check on the use of official data as the base for any research seeking to accurately convey the reality of crime. Certainly, it is widely known amongst criminologists that problems of differing interpretations of criminality that impact upon decisions whether to report record or crimes on the part of several actors, a dark figure of crime, and a variety of other shortcomings severely affect crime data’s ability to do so (see, for example, Mclintock, 1963, Cicourel, 1968). In light of this, the ‘subjective and political nature of the ‘official’ crime statistics’ (Lea and Young, 1984: 16) sees such data regarded as ‘embodiments of police discretion’ and ‘governmental agenda’ (Ferrell et al, 2008: 195-196) rather than hard facts by some. It may then appear that the use of such data may jeopardise the project’s ability to test the Broken Windows thesis with accuracy. That would certainly be the case if they were relied to convey a precise picture of crime throughout the city of Canterbury. However in a basic capacity crime statistics maintain the ability to notify us of increases or decreases in crime, as well as providing a basis on which we can compare levels of criminality across different areas, therefore potentially fulfilling a viable role in a variety of criminological projects. As John Lea and Jock Young remind us, it would be a mistake to dismiss them entirely: [4: One should always keep an eye out for anomalies produced by errors in data, however, as there have been teething problems with the police.uk project. For instance, one street in Preston was recorded as having suffered 152 incidents of recorded crime in a single month, despite only actually having has 3 incidents (Bunyan, 2011)!]

‘We must handle the figures with caution, and, most importantly, that we must develop a sense of realism. We must avoid both the alarmism which takes the figures simply at their face value and the sense of false calm which insists that the same statistics are a mere product of police practices’ (Lea and Young, 1984: 16)

What remains crucial is that criminologists approach official crime data ‘in a critical frame of mind’ (Maguire, 2002: 322), namely by being aware of such failings and accommodating for them in their methodological designs. By and large, crime statistics have proven useful when allowing criminologists to largely discern when crime is increasing or declining and whether particular areas suffer from particularly high or low levels of crime, rather than being relied upon to convey an accurate representation of the nature or a precise measurement of crime and its effects more generally (Lea and Young, 1984). It is in this limited capacity that such statistics are used in this project, where data is merely used to provide a general measurement of crime that allows the research to establish which areas of Canterbury tend to suffer from higher crime rates than others for the purpose of identifying correlations.

A similar (but healthy) scepticism is useful when assessing the usefulness of crime mapping in conveying an accurate picture of crime; a practice intricately connected with the collection of crime statistics as it sees such data plotted on forms of maps (e.g. police.uk). Whilst practices of geographically mapping crime extend back to ‘the progenitors of environmental criminology’ (Hayward, 2012a: 443) in the Chicago school of sociology (see Park, 1925, Burgess, 1925), it is in the last couple of decades that contemporary forms of crime mapping have become an increasingly popular method of presenting official data and regarded as an effective method through which authorities and criminologist alike may come to understand crime and measure the performance of those tasked with reducing it. As Mike Maguire notes in relation to such techniques:

‘the amount of sophistication of analysis will also grow as data from different sources are increasingly combined in ‘data warehouses’, ‘mapped’ via grid references, and so on – developments being actively encouraged by the new regional ‘crime reduction directors’ who have been appointed to facilitate the work of local crime and disorder partnerships’ (Maguire, 2002: 369).

While there is clearly great fervour for the development of crime mapping on the part of administrative criminologists (see also Harries, 1999), authorities across the country and also the public, there have, however, been an abundance of problems identified with developing an understanding of crime purely from the bird’s-eye gaze of crime maps. The most pertinent of which is that crime mapping actually does very little to illuminate understandings of urban criminality by encouraging the criminologist to engage with crime from an abstracted and detached position, therefore making them ignorant to the intricacies of street level conditions in which crime is committed and the complexities of the criminal act itself. The kind of city seen through such maps and designs is referred to by de Certeau (1984) as the ‘Concept-city’; Keith Hayward explains:

‘This is the city as seen by planners, developers, statisticians and, all too often, criminologists. Here the pluralistic fabric and contradictions inherent in urban life… are distilled to leave only quantitative data, demographics and rational discourse’ (2004: 2).

Such an understanding of the city, where one may gaze from a bird’s-eye view to see a city void of human inhabitants, is, however, problematic:

‘The problem… is that the life of the city, the constellation of lives that make a city what it is, the actual experience of the city, in other words, is not contained in the concept of the city. Lives can not be mapped in this way – cannot be read – or even truly rendered readable by maps… something always slips away’ (Buchanan, 2000: 110).

As Hayward puts it in relation to urban crime, ‘the unfortunate thing for… ‘crime analysts’ is that crime, incivility and transgressive behaviour are very complex, multi-faceted, ever-changing socio-cultural phenomena’, and thus maps populated with quantitative data alone are ‘of no use whatsoever in helping us to understand the complex and diverse social and cultural motivations and individual experiences behind a great many criminal offences’ (2004: 110-111). Thus, using crime maps for research purposes does little to tell us about the contexts in which crime is committed and the experiences and motivations behind criminal acts, both of which were identified as integral to understanding the relationship between crime, disorder and space in the literature review. In that sense the maps of police.uk and those created in this research (shown earlier in the chapter), alone, offer relatively little in the testing of the Broken Windows thesis and answering our research questions: they tell us where graffiti and crime exist but do very little in illuminating exactly how one may impact upon the other.

While the solution to the problems associated with official crime statistics was seen in limiting their influence over the project, these issues within mapping practices have been tackled by adding an alternative view of Canterbury’s streets to the analysis. Crime mapping and statistics may provide us with a model through which we can observe the broad distribution of crime across urban areas (an undoubtedly useful tool in assessing the Broken Windows account of urban decay), they, however, only present a computer-generated veneer of city life and the dynamics of crime, disorder and culture that simmer underneath. The following section will now outline how these problems are addressed in the second level of analysis, and how the research takes its cue from De Certeau and Hayward’s arguments regarding the need for to see the city as a duality: the combination of the ‘Concept-city’ perspective alongside an experiential dimension of urban living that mapping fails to capture and which can only be obtained from street level research. This dual mapping, it will be argued, provides platform from which the research questions can most effectively be answered by combining the broad perspective developed here with a close inspection of locations filled with a more vivid and detailed description of how graffiti and crime interact on a street level.

Street Level Analysis

While the value of using official statistics, crime mapping and quantified graffiti writing towards identifying city-wide correlations between graffiti and crime has been emphasised by the broad picture it provides, it is correctly uttered by criminological scholars and statisticians alike that ‘correlation is not causation’ (Hope, 2005: 56, emphasis added). This being true, the correlations drawn from the first level of analysis in this research will clearly not suffice to provide a convincing answer to the issues thrown up by the research questions laid out at the beginning of this thesis. While they offer a valuable, yet basic, insight into the accuracy of Broken Windows theory and the imperative under which Broken Windows policing is advocated, a second level of more detailed analysis is clearly required to flesh out this findings and better explain the relationships lying behind these figures. Such a realisation among others who have embarked on such analysis has, in the past and very much the present, led vast swathes of criminologists further into the depths of regression analysis in attempts to compensate for this shortcoming and reach casual explanations with other forms of statistical sophistication. However, such advanced analytical techniques have previously been found to be ill-equipped to fulfil such an undertaking (Mills, 1959, Young, 2011); as was demonstrated by the collective existence by of an altogether ambiguous statistical assessment of Broken Windows theory laid out in the last chapter. The progression to a second level of analysis in this research, in order to circumnavigate this trading off of statistics and duelling of data-sets and towards something more useful, will instead take to the streets to observe first-hand how dynamics between graffiti, disorder and crime interact. Quantitative analysis is therefore used in the manner outlined by Jock Young in The Criminological Imagination (2011), where the role of this method is both critically viewed and used ‘in a much more limited and circumspect way’ (2011: ix). For Young, this is ‘a question of setting rules and limits… of determining where statistical testing is useful and where it is a distraction’ (2011: 55); words heeded in the restrained use of statistical analysis outlined above. Therefore, the second level of analysis adopted here is a qualitative one that is more attuned to uncovering the subtleties within urban space and writing practices that may impact on the relationship between graffiti and wider crime (in the vein Paul St. Jean and Ralph Taylor’s previous work on this topic).

Specifically, this involved the closer inspection of 3 small-scale studies of areas within the city, developed from the same walking tours and photographs used to quantify the amount of graffiti throughout the city; thus providing a more detailed street-level analysis of graffiti’s impact on its immediate environment. It is through this part of the research that flesh may be added to the bone of the correlations observed in the earlier analysis; that we can add a detailed and theoretically infused critique to the plots on the graph. It is intended that this element of the research project will develop some of the basic insights offered by the quantitative analysis and provide a contextually driven understanding of the relationship to graffiti in Canterbury. Through an attentive gaze, it will also seek to provide a vivid description of examples of either the successes or failures of Broken Windows policing in particular locations and how the effects of such policing approaches are visible in such urban spaces. Thus, while the initial analysis will indicate whether Broken Windows logic and policing appear useful, this stage of the research will shed light on how and reasons why Broken Windows logic may un/successfully explain what is found, and whether Broken Windows policing is successfully policing these areas.

While the use of visual methods to quantify graffiti has already been outlined, the second function of the walking tours taken throughout the city contributes to this second level of analysis and pertains much more to the critical discussions of mapping practice outlined in the previous section. Penetrating below the bird’s-eye-view found on police.uk, and therefore providing a means through which we may better understand the lived reality of these locations in the vein of de Certeau’s duality and Hayward’s cultural criminology, this street-level analysis provides a means of better knowing the full reality of the relationship between graffiti and crime in the locations subject to study here. As Hayward, in his cultural criminological critique of crime mapping practices, states:

‘Any understanding (or ‘mapping’) of urban space must therefore place great store in the multi-layered interactivity that takes place at ‘street level’, the cultural and social dimensions of everyday city-life that enable the formation of a very different interpretive framework’ (Hayward, 2004: 2).

The study of these spaces developed form walking tours and visual analysis being used to ascertain a sense of the street level experience of the city, will indeed be reflected upon to give a more rounded and detailed appreciation of these dynamics and thus directly confront the weaknesses earlier identified police.uk’s mapping method. Earlier evident in the works of Walker Evans (Agee and Evans, 1960) and W Eugene Smith (1998) and more recently been emphasised by several cultural criminologists (see Hayward and Presdee, 2009, Vergara, 1995), such street level visual analysis has the potential to provide the contextualised picture of urban space required for such an undertaking; providing a platform from which the kind of understanding that is regularly found to be lacking in detached forms of crime mapping can be developed. Thus, the use of photography in this project will specifically add a street level view of graffiti and the visible signs of disorder around it, in keeping with a cultural criminological approach, and allow the research to project to develop a context-specific sense of graffiti’s likely impact on its surrounding area. By doing so, the project is well-placed to scrutinise both Wilson and Kelling’s claims and the elaborate on the tentative conclusions reached through the first level of analysis.

Summary

This chapter has outlined the methods used in this project, the nature of their use and how they are tailored to answering the project’s research questions. It has set out to explain how this research will develop over two stages, and how the weaknesses in each of the methods used is compensated by the strengths of the other. In short, this approach combines a broad gaze and narrow, more detailed view that will combine to answer whether (or not) Broken Windows logic and policing succeed within the city, and why this is the case. The mixed-methods approach adopted here sees rudimentary statistical analysis conducted to provide a basic testing of Broken Windows logic. Here the relationship between graffiti and crime across the city is tested to see if the two variables correlate with one another: according to the rationale laid down by Wilson and Kelling, one would expect to observe a positive correlation that shows high graffiti areas to also suffer the highest rate of recorded crime. Whether or not such a correlation materialises will be the initial test of Broken Windows theory, and an indication of whether Broken Windows policing practices are worth pursuing. The weaknesses in this approach, where crime statistics and detached forms of mapping tell us very little about why graffiti may or may not cause crime, have been combatted somewhat by a more detailed analysis of case studies of locations within the city. It has been explained here that visual analysis of street-level locations and photographs of graffiti taken on extensive walking tours of the city will not only to quantify the amount of graffiti in the city, but also provide the kind of contextualised insights into how graffiti interacts with the space around it that forms of crime mapping can not offer. The following chapters will present the findings of this research. Beginning first with the statistical analysis in the next chapter, from which tentative conclusions will be drawn, a more fleshed out analysis of the street-level case studies will follow afterwards.

Chapter IV:

City Wide Correlations: Quantitative Analysis

This chapter will present the results of the statistical analysis conducted as part of this research project. As was explained in the methodology chapter, the scope of this analysis is purposefully limited and its methods are intentionally simplistic and elementary in comparison to other scholarly works that have previously sought to test the Broken Windows thesis and Broken Windows policing impact on crime rates (see Thomas et al, 2011 and Kelling and Sousa, 2001 for example). What follows is a the detailing of the findings in this search for correlations between graffiti and crime rates across 10 areas of Canterbury, as well as a discussion of how these results relate to the research questions posed in the introductory chapter of this thesis. This analysis will shed light on whether or not the sequential decline described by Wilson and Kelling materialises through recorded crime figures, or a prevailing status quo where areas with a prevalence of graffiti also suffer high crime. It will also reflect on the decision of local authorities to commit resources to a programme of Broken Windows policing, under the guise that aggressively controlling graffiti will reduce and prevent other forms of crime; revealing whether this approach’s focus on graffiti is empirically supported and likely to bring about a reduction in general crime rates. The value of this research in particular is that it will help establish a broad coverage of the relationship between graffiti and crime across the majority of the city, however, a second level of examination is required to elaborate on these findings, where a more detailed visual analysis of street-level locations within the city will be introduced to add flesh to the bone of the findings presented here. What follows here, then, is one half of this project’s findings and some brief discussions and soft conclusions drawn from them: the correlations here will be used to offer tentative answers to our research questions whilst, acknowledging that correlation is not causation, a more detailed causal explanation to underpin these results will be developed in the next chapter.

Findings

The correlations that lie between the distribution of graffiti and crime across these 10 zones in the 6 months of December 2011-May 2012, as is shown in the graph below, appear to show a weak or no relationship between these variables (a fit line inserted on to the graph shows there to be a weak negative relationship). This scatter-plot actually reveals there to be little difference in the crime suffered in areas of low and high graffiti. Whilst it is shown that there are indeed some low crime/low graffiti areas of the city, the upwards diagonal line one would expect to see leading away from the intersect of the X and Y axis never materialises: what we instead see, by the positioning of the points on the scatter diagram below, is that there is little difference in the amount of crime suffered in both low and high graffiti areas. Indeed, in some cases it points to a high graffiti / low crime character of parts of the city: for example, the area in which the level of crime is the lowest amongst each of the zones and across this 6 month period also contains the second most pockets of graffiti writing (the St. Stephen’s area). More fundamentally, however, the graph shows there to no meaningful relationship between the two variables across the entire 10 zones: suggesting that both high and low graffiti areas within the city will suffer similar levels of crime.

Furthermore, taking apart the data and inspecting it more closely does little to contradict these findings. As the below image of the data broken down into its component monthly totals illustrates, the correlations between graffiti and crime remain fairly constant over each month within this period. Indeed, a correlation similar to that shown for the entire 6 month period is each shown in five of these graphs. The only exception is shown in February 2012, where it appears that the three zones containing the least graffiti each experienced a decline in crime, whereas the rest remained relatively constant. What this breakdown shows is that the correlation identified above has not been skewed by one or two anomalous figures that do not represent the character of crime distribution of across the rest of this time period; an important factor to consider when we bear in mind Harcourt’s (1998) re-analysis of Wesley Skogan’s datasets that proved such influences to sometimes have a significant effect on such analyses. On the contrary, the below images reveal this 6 month correlation to be based on relatively stable figures showing a lack of a relationship between the variables that tend to repeat themselves over the individual months within this period. Furthermore, where this pattern repeats itself time and time again, with the possible exception of February 2012, this re-affirms what conclusions we may draw from the initial graph and the correlation it offers. With some confidence that these results are not corrupted by anomalous results, we can therefore maintain that there appears to be no relationship between graffiti and crime within these 10 areas of Canterbury.

The correlations between graffiti and crime for each of the respective months within the December 2011-May 2012 period

Interpreting the Figures

What answers, then, do these figures and correlations offer to our research questions? At a simplistic level, these results tentatively suggest that the relationship between graffiti writing and wider crime in Wilson and Kelling’s thesis bears little resemblance to graffiti’s impact on spaces throughout Canterbury. These results then initially indicate that, in answer to our first research question, that a Broken Windows model of understanding graffiti’s impact on crime throughout the city is a fairly inaccurate one. Where we would expect to find a supposed connection between graffiti other forms of crime, which would see at the very least a moderate (perhaps even a strong) positive correlation between the variables emerge, this result indeed prompts us towards rather different conclusions. Thus, what we see is an apparent failure in Broken Windows logic, which would suggest that areas in which there is the most graffiti there will also be the most crime, and its ability to explain the relationship between graffiti and other forms of criminality: areas in which graffiti was most widespread were found to have roughly the same, and sometimes less, crime than areas with a much less writing. Furthermore, through the breakdown of monthly figures shown above, we can also observe that there appeared little evidence of crime rising in high graffiti areas, in the vein of the sequential decline described by Wilson and Kelling, over this 6 month period. In light of this, these figures neither point to a scenario where graffiti has already attracted criminality to the areas in which it is most widespread, nor do they bear any evidence of graffiti causing an increase in crime over this period of time. In short, then, these results suggest that Broken Windows theory is of relatively little use in explaining how graffiti effects its surrounding environment across the city (therefore being comparable to the findings of Yang, 2007, Chappell et al, 2011, Cerdia et al, 2009 Harcourt, 1998, Sampson and Raudenbush, 1999, 2004). They insinuate, then, that factors outside of graffiti’s visible disorder will be more useful in explaining the prevalence and distribution of throughout in the city.

These findings would, by implication, also appear to cast doubt on the effectiveness, and potential for success, of Broken Windows policing. Indeed, this analysis simultaneously offers some answers pertaining to the project’s second research question surrounding policing practices born from the rationale used in Wilson and Kelling’s thesis. If, as these correlations infer, it is indeed true that areas in which disorder (in the form of graffiti writing) is most prevalent tend to suffer marginally less crime than those in which disorder is much less visible, then the value of tackling that disorder as a means of reducing crime must surely be questioned. While these figures offer little in terms of establishing a causal explanation for the distribution of crime throughout the city, they do suggest that graffiti is unlikely to be a catalyst for criminal behaviour. Suggesting graffiti to have little impact on criminality within the city, and therefore undermining the empirical foundation and basis on which Broken Windows policing is advanced against writing, they therefore hint that attempts to reduce crime by reducing and controlling graffiti will be largely futile. It is in this vein that these findings correspond with other studies that have similarly suggested, through statistical analysis and other methods, Broken Windows policing to be ineffectual in reducing wider crime (see Katz et al, 2009, Karmen, 2000). While it is conceded that this analysis can offer no insights into other driving forces behind criminal behaviour within the city, it does intimate that a Broken Windows policing model that will continue to consider graffiti to have a role to play in causing further criminality is unlikely to be successful in its aims. We should therefore continue to look, through further research and the deconstruction of local data, for alternative explanations for the prevalence of crime in areas of the city, from which alternative policing methods that better correspond to the causes of crime can be developed.

While this analysis appears to offer some valuable insights that we may to use to begin to answer the project’s research questions, we must take caution before overstating the strength of these conclusions. As was discussed in the methodology chapter, crime statistics have a limited value in bearing the reality of crime, which as Lea and Young remind us ‘are not hard facts’ (1984: 12), whilst statistical analysis has a restricted potential in revealing the causal factors underlying criminal behaviour (Young, 2011). In this vein, what this kind of analysis does not provide us with is any serviceable insight into why Broken Windows theory’s predictions fail to materialise on Canterbury’s streets; whether these failings, for instance, potentially stem from the meaning of disorder being inappropriately charged on to writing, or perhaps that such disorderly sights having little effect in breeding criminality in their surrounding environment. Furthermore, in relation to Broken Windows policing practices, these figures merely suggest that the basis on which Broken Windows policing is alleged to be successful possesses little relevance to Canterbury and its graffiti writing. Thus, all that we can, at this point, conclude is that writing appears to not have had the impact on the city that Wilson and Kelling’s theory would suggest, much to its discredit, and that Broken Windows policing practices are unlikely to be result in a reduction in crime across the city.

Conclusions

The findings presented in this chapter have corresponded very closely with other attempts by scholars to quantitatively test the Broken Windows hypothesis, both in terms of their findings and usefulness. As have many other examinations of the Broken Windows thesis (see Yang, 2007, Sampson and Raudenbush, 1999, Cerdia et al, 2009), these findings cast doubt on the premises of Wilson and Kelling’s claims by showing, through correlations, there to be no evident relationship between crime and visible disorder (in this case, specifically graffiti writing); which its authors claim ‘are usually inextricably linked, in a kind of developmental sequence’ (Wilson and Kelling, 1982: 31). The figures that were laid out here revealed there to be no meaningful correlation between the variables of disorder and crime across the city: it was shown that a high prevalence of graffiti in an area was not a good predictor of high crime rates. Amidst these correlations, where this relationship was fairly constant across a 6 month period, there was also no evidence to suggest that graffiti has been acting as a catalyst to kick-start a downward trajectory into increased criminality in areas of the city. Crime rates remained consistently low in areas in which graffiti was particularly prevalent. As discussed above, these findings therefore have negative implications for how the accuracy of the theory and its associated policing approaches may be judged.

This chapter has, however, also acknowledged the limited value of this methodological approach when conducting such investigations. Whilst these findings clearly speak to the claims made by Wilson and Kelling (1982), they do little beyond merely tentatively indicate that aspects of the theory appear to possess little relevance to Canterbury and that policing practices developed from its central premise are unlikely to successfully reduce crime throughout the city. As was argued in the literature review and the last chapter, the complex relationships between crime, graffiti and space require criminologists to physically enter such spaces to closely inspect the nature of these connections from an interpretive standpoint. Thus we will now progress to the next chapter, based on a street-level analysis of three case studies within the city, to flesh out this analysis: to move beyond these simplistic findings that offer guidance on the accuracy of the theory more broadly and to a level where we may diagnose the specific failures of Broken Windows theory (and policing) where they are found on the city’s streets.

Chapter V

Breaking Down the Walls: the Street-Level Gaze

The thesis will now progress to its qualitative case study analysis of three locations within the city. As was mentioned in the methodology chapter, the case studies here are based on walking tours and photographs taken (by the author) of areas within the city, as well as various other sources of information (including local campaign literature and local media articles) that provide a greater sense of the area than does the statistical information drawn from the bird’s-eye gaze of police.uk’s crime maps. The findings presented here are, then, an interpretation based on this immersion in the localities in Canterbury with which this project is concerned. Whereas the last chapter offered a broad scope and soft conclusions suggesting that Broken Windows theory fails to explain the relationship between graffiti and wider crime across Canterbury, and that associated policing practices are subsequently likely to fail in their aims, here these discussions are fleshed out with a narrow focus and increased detail. A description of the spaces underlying these figures, and the processes and developments at work within them, is intended to lead us closer to an explanation of exactly how and where the theory succeeds or fails, and ultimately provide more robust answers to the research questions at the centre of this project. Throughout this chapter, the findings presented here will also be placed in on-going discussions of graffiti, crime, urban space and the relationship between them. Through such discussions, cultural criminology emerges as an alternative framework through which we may understand how the relationship between graffiti and crime plays out within Canterbury’s streets.

The photos used in this project offer a window into a variety of street-level settings within Canterbury, providing a medium of penetrating beneath the bird’s eye view found on police.uk (as shown above).

Case Study I: Inside the City Walls

Contrary to what Broken Windows logic might predict, the prevalence of graffiti inside Canterbury’s medieval city walls, which surround the city’s main shopping district, is relatively low whilst crime rates are extremely high. The amount of crime in this central region of the city between December 2011-May 2012, totalling 824 offences, measures alone at over 37% of all recorded crime in the 10 zones of the city covered in the previous chapter. On the city’s high-street and main shopping arcades there is little graffiti in plain sight, whilst much of the graffiti there is exists in small pockets among the narrow roads, alleyways and subways that flow from and around this main parade. They mostly take the form of small tags. Overall, this central area of the city, whose attractions include: the city’s high-street; various shopping arcades; the city’s largest theatre; the historic Canterbury Cathedral; as well as a variety of restaurants, pubs and bars, can reasonably be described as extremely well-kept. Indeed, the variety of hanging baskets, colourful banners along the high-street, plush trees and bushes and clean shop fronts indeed seem to cast a good impression on those who walk around the city-centre. Moreover, the frequent sight of patrolling police officers (both on foot and in squad cars), various litter-pickers and chewing gum removers, as well as the private army of staff hired to maintain and police the Whitefriars’ shopping area, indicate that this is a part of Canterbury in which authorities, developers and property owners intend to protect the impression of security the area gives in order to attract shoppers (and their money) to the city. In light of this sense of orderliness and the lack of the graffiti in the area, it appears to make little sense to attribute the blame for the high crime rates experienced here on visible disorder in the mould of Broken Windows theory.

.

As these photos (above) show, the area in and around Canterbury’s city centre appears well-kept and gives the impression of security and orderliness.

This description, however, only applies to Canterbury city-centre as a day-time hive of shopping, consumption and relaxation, whereas we must look to the use of this same space at night in order to explain why the city-centre boasts such high figures of recorded crime. At night, other sources of revenue and forms of consumption tend to occupy these same streets and transform the city into a somewhat different place: a tale that is told both through recorded crime figures and walks through the city on busy evenings. Like many so-called ‘beer towns’ across the UK (Hobbs et al, 2005), in which the night-time industry ‘spawns criminal violence at every turn’ (Ferrell et al, 2008: 204), much of Canterbury’s local economy runs on night-time consumption of alcohol, in turn, apparently suffers from a high amount of crime committed by night-time drinkers. In light of this, it appears more appropriate to characterise the city’s night-time economy as the catalyst for acts of criminality, rather than visible signs of disorder (as Broken Windows theory would suggest)[footnoteRef:5]. As the figures on police.uk reveal, much of the city-centre’s crime is, unsurprisingly, clustered around the town’s most popular drinking spots; further compounding suggestions that it is indeed the city’s night-time industry that lies at the heart of much of the area’s crime. Furthermore, police.uk is able to reveal that the types of crime committed in this area are largely consistent with those associated with night-time economies by existing literature (see Winlow and Hall, 2006, Hobbs et al, 2005). Take, for instance, the area around the Cherry Tree pub in the month of May 2012 which illustrates such dynamics: of the 18 crimes committed in the nearby area that month, the variety of anti-social, violent, drugs and public disorder offences do just this, and indeed present just one piece of a wider jigsaw of similar locations throughout the city-centre that provide an illustration of the same dynamics. [5: Another explanation for the high rates of crime in this area, though one that appears less influential than night-time drinking, is day-time shoplifting. Given the wide array of shops in the city centre, opportunities for shoplifting here obviously exceed those provided anywhere else in the city. Thus, we should still take care not to claim that all crime in the area is driven by the city’s night-time economy.]

Such findings, told through crime figures and first-hand observation, raise issues of contention against those who promote Broken Windows in the city, and accentuate how the theory’s emphasis on visible disorder as a cause of criminal behaviour fails to relate to particular urban environments. Indeed, this example has shown how graffiti emerges as a non-factor in explaining the prevalence of crime in this area. In contrast it has been demonstrated that the city’s night-time economy, alongside the kind of shoplifting crimes that are an inevitability in such shopping spaces, rather than visible disorder, emerges as a more reasonable explanation for the extremely pronounced recorded crime figures inside the city walls. Ill-equipped to consider the often parafunctional nature of urban space (Hayward, 2012a), what is revealed through this example is that Broken Windows’ lack of reflexivity means that the theory fails to account for how Canterbury’s city-centre can both resemble a secure shopper’s paradise and one of the ‘brutal, instrumental urban site of alcohol-driven excess and violence’ (Ferrell et al, 2008: 106) located in and around the country. Thus, in order to explain crime in this area it is appears necessary to turn to alternative bodies of work, some of which include: cultural criminological literature on the multiple uses of urban space (see Hayward, 2012a); studies of behaviour amidst hedonistic night-time consumption of alcohol and illicit drugs providing a greater understanding of how, why and when crime will be committed in such environments (see Hobbs et al, 2005); as well as cultural criminology’s appreciation of the social, economic and cultural conditions that makes such pursuits popular (see Presdee, 1999), to provide a more relevant wisdom for accounting for crime in this area. Thus this case study, in answer to our first question, demonstrates how Broken Windows logic fails to account for the causes of crime in this area (and this crime’s relationship with visible disorder), and can obscure more useful explanations from high crime rates with its distracting logic. Showing there to be high crime in a low-graffiti and highly ordered environment clearly suggests that the theory is of little use in explaining why crime rates remain high in this zone. Furthermore, this case study suggests that Broken Windows policing will be unsuccessful in its attempts at reducing crime generally: visible disorder is clearly not the cause for much of the criminality in the city-centre, so the policing of graffiti is unlikely to have any significant impact on wider rates of offences in the area.

Case Study II: Kingsmead Road

Kingsmead Road’s relationship with graffiti and crime is a complex one, and perhaps demonstrates best among these case studies the value of taking to the street to closely inspect the complex and nuanced dynamics at work between graffiti, crime and culture in urban space. Alongside a dense collection of graffiti writing, the area hosts: a large supermarket; council-run swimming pool and gym; as well as a medical centre. While there is indeed a greater prevalence of graffiti in this small area than is typically found across Canterbury, alongside an above average rate of crime, one should be cautious before concluding that Broken Windows theory offers the best theoretical framework for explaining the interactions between these phenomena in this area. As will be revealed here, there is more than meets the bird’s eye view in this particular case. Indeed, the Kingsmead area presents a fine example of the varieties in graffiti’s meaning and also emphasises the need to adopt a cultural criminological perspective, rather than Broken Windows rationale, in order to unravel these complicated dynamics within a wider cultural backdrop.

Whilst the prevalence of graffiti in this microcosm of the city may suggest that writing has, in some way, been responsible for its relatively high crime rates, a closer inspection indicates that such blunt and simplistic reasoning fails to account for wider cultural dynamics that influence how it is likely to be perceived and interpreted by local residents. Firstly, however, the issue of the night-time economy appears to raise its head once more as the confounding factor that can more reasonably be held accountable for the levels of crime here. The Penny Theatre, a large premises lying adjacent to two late-night kebab houses on the corner where Kingsmead Road meets Northgate, is one of the city’s most popular drinking spots amongst local students and young adults. As was the case with night-time spots in the previous section, the crime committed in this area can largely be attributed to a number of anti-social behaviour, public disorder and violent offences committed inside or just outside the pub (at their worst, totalling 9 incidences in the month of May 2012). Revealing the majority of crime in this small area to be clustered around the Penny Theatre, rather than the dense collection of graffiti surrounding the nearby youth club, it thus compounds the suggestions made in the last case study that the city’s night-time industry tends to lies at the root of much of criminal disruption throughout the city.

The legally written graffiti surrounding the youth club (above) appears to be more technically sophisticated than the majority of writing across the city, whilst appearing to do little in attract crime and disorder to the area.

The Riverside Youth Centre, on Kingsmead Road, alternatively brings to the fore some of the complications intricately tied into graffiti’s contemporary cultural meaning, and simultaneously throws another spanner in the works of Broken Windows’ portrayal of graffiti acting as a bearer of bad news. Indeed, contrary to such an interpretation of graffiti’s meaning, the writing that decorates the wall of the buildings appears to function as a symbol of inclusion (an invitation for young people to enter the club) and as a celebration of graffiti within mainstream youth culture. At the manager’s request, local writers were invited to write on the walls of the club and subsequently created a variety of colourful murals and pieces along the front of the main building’s walls and also the on-site five-a-side football pitch[footnoteRef:6] (see Tandasatu, 2/4/2010, for a video of this process). That those running this youth club, funded by the local council, saw fit to have it decorated with graffiti, speaks volumes about: the way graffiti is valued within the current cultural context; the alternative meanings that different parts of the local population attach to different examples of writing; and the distinctions criminologists accordingly need to make between different varieties of graffiti. In particular, the painting of the youth club corresponds with wider developments identified by cultural criminologists; where it has been noted that graffiti, amongst other crimes, has been ‘packaged and marketed to young people as a romantic, exciting, cool and fashionable cultural symbol’ by a number of industries. Keith Hayward explains, ‘it is in this cultural context that transgression becomes a desirable consumer choice. Within consumer culture, crime is being aestheticised and stylised, presented in the mass media as on a par with a fashion aesthetic’ (both quotes from Hayward, 2004: 169). Driven by representations in commercial advertising and the use of graffiti’s style in many commodities themselves, through this process, graffiti has evolved from a feature of counter-culture to become part of the mainstream. As a result, graffiti, its visual style and emotional sensations of transgression are now ‘celebrate[d] and very importantly, commodif[ied]’ (Hayward, 2004: 173) by industries selling products to youth markets. Recent examples of this are beyond counting; however they include the use of graffiti within adverts for Adidas clothes and Converse trainers; the use of graffiti (or graffiti’s style) as an advertising medium itself; fashion items that themselves adopt graffiti’s visual style; and even video games (such as Mark Ecko’s ‘Getting Up’) and; now evidently by youth clubs attempting to aesthetically express the style of those who use the clubs and also to attract new members: [6: However, some writing was not welcomed by those who run the club (this mainly consists of a few tags that have been peppered around the club’s walls and also some large containers near the building)]

‘Graffiti style has also gone mainstream; as graffiti has become the lingua franca of hip, artistic rebellion has indeed turned into big money… suggesting of course that whatever cultural authenticity the law can’t kill, the market can’ (Ferrell et al, 2008: 151).

The appropriation of graffiti’s style and urban edginess has been evident in graffiti-based products: such as Marc Ecko’s Getting Up Video Game (above left); the use of graffiti writing’s transgressive sensations in Adidas advertising campaigns (above centre), and the use of graffiti aesthetic for adverts by Sony (above right).

Once we contextualise the graffiti surrounding the Riverside Youth Centre in these cultural processes, it quickly becomes clear that Wilson and Kelling’s blunt understanding of graffiti as a symbol of disorder is both ignorant to these cultural shifts and unable to accurately characterise this graffiti’s meaning to nearby urban audiences. While perhaps the tag continues to represent senseless vandalism to some (Ferrell et al, 2008: 101), the murals decorating the youth club appear to represent something rather different. Indeed, this case study has specifically shown that, within this evolving and mutating cultural context, the youth club’s use of graffiti signals to the community that this is a space for young people and to signal to young people themselves that this is a cool and fun place to be. In doing so, it also relates to wider examples of the appropriation of graffiti to attract people to an area: for example, Snyder’s finding that certain shopkeepers in areas of New York encouraged graffiti writing inside their stores in order to ‘tell shoppers that this is a cool space’, and demonstrate graffiti is indeed ‘good for business’ (2009: 53).

This case study thus demonstrates the value of cultural criminology’s culturally aware and interpretive approach to graffiti writing in comparison to Broken Windows blunt and unwavering characterisation of graffiti as a sign of disorder. Whereas a static Broken Windows logic leads us to assume that the Riverside Youth Club’s graffiti might inspire crime, a reflexive cultural criminological approach sees the lack of crime around the club better explained through the ‘plac[ing of] criminality and its control in the context of culture’ (culturalcriminology.org, 2012); in this case meaning situating Kingsmead’s graffiti within writing’s journey into mainstream youth culture. Suggesting, therefore, that understandings of graffiti as a sign of disorder, in Broken Windows arguments, remain too rigid in light of graffiti’s use as a sign of urban cool to young audiences, the graffiti on the Riverside Youth Club reminds us that criminologists must remain attentive to the diverse and ever-changing meanings that writing can hold and the distinctions between different variations of writing. This case study then, pertaining to the first of our research questions, shows Broken Windows theory spectacularly failing to explain graffiti writing’s impact on its environment through wildly misinterpreting what writing means to particular urban audiences. Moreover, it then implies that Broken Windows policing will be entirely ineffective in reducing wider crime by virtue of this writing both being legal and symbolising mainstream faux-edginess rather than genuine urban disorder.

Case Study III: Westgate

Whereas the previous case study raised pertinent issues regarding the discrepancy between Broken Windows theory’s interpretation of writing and that of Canterbury’s residents, the case of the Westgate area concerns the actions of writers themselves. Providing a fine example of the unintended consequences of Broken Windows policing it, in short, tells a story of crime displacement brought about by methods of crime control used, within a popular model of Broken Windows policing, to counteract the movements of writers. Drawing on a precedent identified in cultural criminological literature, this case study will provide a demonstration that, rather than preventing and controlling writing as they intend, in certain cases such measures ‘will, over time, be circumnavigated or outwitted, by dint of simple force, innovative criminal practice or, that old nemesis of situational crime prevention, crime displacement’ (Hayward, 2004: 186). It therefore, in relation to our second research question, points to a significant failing in this policing approach.

As visitors walk from the Westgate gardens’ entrance (above left), they will pass features of beautiful greenery (above centre) before coming across the steel gate and graffiti at the area’s edge

The entrance to the Wesgate gardens lies adjacent, unsurprisingly, to the West gate of the Canterbury’s medieval city wall. As one enters the gardens and walks by the plush foliage found along the river Stour’s banks within this peaceful pocket of Canterbury, there is little to suggest that within a stone’s throw lies one of the most dense collections of graffiti in the city. Certainly, visitors to the gardens will find themselves a world away from the disordered landscapes populated with drunks and drug dealers and associated with graffiti writing by Wilson and Kelling’s (1982) Broken Windows theory. However, at the far end of this most tranquil and peaceful part of the city, full of open green areas in which people regularly come to read, play, picnic or merely spend an enjoyable lunch hour, the gated entrance to the underside of the road-support bridge (which is arguably unsightly in its own right) at the end of the garden signals a divide between the serene lawns and a supposed symbol of urban decay. Boasting some the city’s most intricate and most technically-sophisticated graffiti, the underside of the bridge contains a diverse range of writing styles exhibited on almost every available inch of the walls and pillars on show. Posing yet another contradiction to Broken Windows logic, they, however, appear to have had very little effect on crime and disorder nearby: not only do the gardens themselves still give the impression of a safe and peaceful haven within the city, recorded crime in the surrounding area of Westgate remains among the lowest throughout the rest of the city.

While the lack of crime surrounding this graffiti once more undermines the premises of Wilson and Kelling’s theory, more significantly, the appearance of writing in such a location points to failings in a Broken Windows policing method of crime control. Indeed, the location of the vibrant collection of murals, pieces and tags found here corresponds very closely with discussions of the displacement of writing identified in other cities by Snyder (2006, 2009) and Ferrell (1996, 1998), among others (Hayward, 2004, 2007), and in doing so offers another contradiction to Broken Windows logic. As has been acknowledged by cultural criminological scholars, displacement has been identified as a regular consequence of situational approaches to crime control (Hayward, 2007), many of which tend to be included in models of Broken Windows policing (Manning, 2001, McLaughlin, 2008a). Specifically towards graffiti in Canterbury these have included: the swift removal of graffiti; the use of anti-graffiti paint to coat surfaces; a variety of surveillance techniques including CCTV (Canterbury City Council, 2012a); on the spot fines for those caught writing (Warren, 23/8/2012); and the frequent prosecution of writers (Claridge and Alston, 18/2/2010). The result, as has been shown in many cases, is that graffiti tends to merely be moved to obscure and less-visible locations rather than prevented or reduced; to locations where writers can avoid surveillance and develop sophisticated and time-consuming pieces of writing without fear of punishment[footnoteRef:7]. What is found in the Westgate gardens serves as an example of these very processes playing out in Canterbury’s urban landscape: where the city centre and other areas of the city are subject to this policing, the edge of the garden offers the kind of unsurveilled location that writers have elsewhere been increasingly shown to favour (see Ferrell, 1996). Furthermore, evidence of this displacement effect is also evident across other areas of the city in which writing is found where much writing is now found in hidden pockets of the city landscape. Much like Jeff Ferrell found that writers in Denver ‘were pushed into real obscure places where’ and left ‘basically… painting for other writers and puttin’ up wallpaper for winos and bums’ (Eye Six quoted in Ferrell, 1990: 11) by similar crackdowns on writing, in Canterbury it has been subways frequently used by the city’s homeless for shelter and begging that similarly offer some of the most dense collections of graffiti in the area. As the Westgate gardens and several other collections of writing located in hidden pockets of Canterbury’s urban landscape then demonstrate, these dynamics equally appear to be playing out in space across this small Kentish city. [7: In other cases, such measures have also been shown to frequently encourage writers for whom the main motivation for writing is risk taking (see Ferrell, 1996, Snyder, 2009, Lyng, 2005, see also Katz, 1988) where, as Ferrell, Hayward and Young put it, such measures have ‘ironically accelerated graffiti’s experiential seductions’ (2008: 150) in the past.]

The underside of the bridge (above) reveals one of the city’s densest collection of graffiti, as well as some of the city’s most technically-sophisticated pieces.

Displacement is evident across the city, both with writing increasingly appearing to be hidden from view (above left), whilst some writers appear to use the example of the Pacman video game to taunt authorities’ inability to control their actions (above right).

We must also consider, in order to fully comprehend the driving forces lying behind these developments in writing, factors external to the cat and mouse interplay between authorities and writers that are driving such changes. The increasing significance of photography to writing and writers that has seen graffiti increasingly shared via graffiti-based media publications, and more recently the internet, is relevant here. This, it would seem as previously suggested by cultural criminological scholars, has also accelerated the displacement effect caused by Broken Windows policing (see Hayward, 2007, 2012) in many locations by leading many writers to prize visibility gained through sharing digital photography above being seen in their own cities . As Greg Snyder writes:

‘Photographs [make] ephemeral graffiti pieces permanent, allowing writers to view the work of others without attachment to a specific place or time. The inclusion of these ‘flicks’ in magazines created a space where graffiti pieces from all over the world could come together to be judged, critiqued, and offered as instruction’ (Snyder, 2006: 93).

Not only does this apply to the magazine publications mentioned by Snyder (such as Getting Up and ON the Go), but also now internet forums and sites that bring together writers from all parts of the globe (Mills, 2011). Where ‘the World Wide Web has added another path to fame that doesn’t include illegally saturating city walls with a name’ (Snyder, 2009: 35), the need for writers to attain visibility in their own cities has been superseded by recognition for quality of work among a global network of peers, where graffiti is now mediated either through print or virtual media (Snyder, 2006). This has resulted in writers seeking out obscure locations to write, in which they can produce better work, and contributed to the displacement of writing in locations where techniques of Broken Windows policing have been utilised. Evidence of this development’s relevance to Canterbury’s writing scene can indeed be found with a few clicks of a mouse, thus compounding these suggestions that the internet and Broken Windows theory have collectively driven graffiti writing to the margins of the city. Aside from whatever content has been beamed into various channels in multiple internet forums and millions of discussion threads, these include videos showing examples of graffiti around the city (see, Chapman-Jones, 2009, Soflesviolence, 2010) as well as a dedicated Canterbury Bombs website.

By looking to these cultural criminological analyses of graffiti culture, ones that takes into account the unintended consequences of Broken Windows policing and the role of new media in criminal sub and counter-cultures, we can therefore explain why, contrary to Broken Windows’ expectations, Canterbury’s graffiti can increasingly be found in locations like the Westgate gardens. It is through these progressions that we can also identify a failure in Broken Windows policing, which appears to fall at the first hurdle in relation to graffiti writing: if it is believed that aggressively policing writing will herald a reduction in crime, a claim that has thus far been suggested to lack an empirical foundation anyway, the inability of this approach to actually to control writing surely undermines it entirely! Rather than consider writers as rational actors who will be deterred by such policing, cultural criminologists’ attentiveness to the supposedly ‘irrational’ (in that they stand contrary to the expectations of rational choice theorists) features of writing practices has proven useful in discussions of these findings (see Hayward, 2007). Once more, where it has been observed that Broken Windows logic and policing have been observed to have failed, cultural criminology has proven useful in explaining how and why writers have changed their writing practices in the face of Broken Windows policing practices.

Conclusions

In light of the findings of these three case studies, and in combination with the elementary insights drawn from this project’s quantitative analysis, Broken Windows emerges as an extremely problematic theoretical framework for understanding the effect of graffiti on its immediate environment. The last chapter’s findings tentatively suggested that the relationship between graffiti and crime across Canterbury more broadly bore little relation to the dynamics described in Broken Windows theory, where graffiti among other crimes is described as a catalyst for urban decline and an increase in crime. They also suggested that there was little empirical imperative under which to advocate a Broken Windows policing model against graffiti writing in the fight against crime more generally; suggesting this policy is unlikely to herald a reduction in wider criminality. Alongside this discovery of a weak relationship between the distribution of graffiti and crime throughout the city, this chapter has ratcheted up this critique of Broken Windows theory by throwing up examples of how and why this theory fails in its explanations of crime’s causes. In relation to our research questions, this chapter has demonstrated how high crime areas may appear highly ordered and contain little graffiti, how graffiti may often not symbolise disorder in the manner Wilson and Kelling claim, and how Broken Windows policing practices fail to control writing practices. As a result, the tentative conclusions reached at the end of the last chapter have been reinforced: the theoretical premises of Broken Windows theory, and the policing practices that have been spawned from its rationale, have each been revealed to have significant flaws. Also pointed to, in the night-time economy, features of the city that appear to have had a greater effect on crime rates than any form of writing or lingering and visible disorder.

Moreover this chapter has, through discussions of how we can better interpret these findings, introduced several elements of cultural criminology that tally with what has been observed here. Be it acknowledgements of distinctions between different variations of writing, awareness of graffiti’s transition into mainstream youth/consumer culture as a feature of selling products to young people, or processes of displacement brought about by Broken Windows policing and online digital culture, the nuanced and reflexive interpretive framework that cultural criminology brings to this topic has clearly been demonstrated to be more useful than Wilson and Kelling’s thesis. These findings and the discussions presented here, thus, suggest that we must move away from Broken Windows theory and reconfigure our understandings of the relationship between graffiti and its surroundings. Not only do they indicate that the theory itself lacks accuracy, they implicitly indicate that cognate practices of Broken Windows policing are ineffective: both when it comes to controlling writing and reducing crime generally by controlling writing. In light of this, this thesis shares the sentiments of Hayward in regarding Broken Windows theory largely as a ‘flawed semiotic interpretation of urban space’ (Hayward, 2012a: 441), and those of Sampson of Raudenbush who, after observing similar faults in this policing approach, refer to Broken Windows policing as an ‘analytically weak strategy to reduce crime’ (cited in Harcourt and Ludwig, 2006: 284).

Chapter VI

Conclusions

As was noted at the beginning of this thesis, Broken Windows logic has been used frequently and fervently by local authorities in the city of Canterbury: both to describe the effect of writing on its immediate environment within the city (see Canterbury City Council, 2012); and to underpin a policing approach that would see the aggressive law enforcement against writers (Canterbury City Council, 2012a). Acknowledging that the value of this logic was not self-evident, and that evidence pointing to its accuracy remained relatively unconvincing, this research sought to test the theory’s claims in relation to graffiti writing. It also sought to examine the effectiveness of methods of policing born from the logic originally employed in Wilson and Kelling’s (1982) seminal ‘Broken Windows’ article (and pursed by the like of Kelling and Coles, 1997). Thus, it progressed around the asking of two related questions:

· To what extent does Broken Windows theory provide an accurate understanding of graffiti writing’s effect on its immediate environment in Canterbury?

· To what extent is a focus on graffiti as part of Broken Windows policing in Canterbury likely to be effective in reducing crime as a whole?

Using a mixed-methods approach, which combined respective levels of quantitative and qualitative analysis, this project has found that the arguments in Broken Windows theory possess little relevance to Canterbury’s graffiti writing. It also found no empirical basis on which it could be suggested that a programme of Broken Windows policing against graffiti has been, or is likely to be, responsible for heralding any noticeable reduction in crime. This research firstly involved statistical analysis of recorded crime data that sought to unearth correlations between crime and graffiti across 10 areas of the city. This analysis found no meaningful relationship existing between these two variables and, therefore, tentatively cast doubt on Wilson and Kelling’s arguments from a broad perspective. It seemed to also condemn its associated policing practices by suggesting that areas in which graffiti was most prevalent did not suffer higher rates of crime than the rest of the city. The qualitative analysis, based on visual analysis of street level locations and photographs, that followed took this one step further and pointed towards examples of specific failings in the Broken Windows thesis and Broken Windows policing model. These included, in relation the theoretical arguments, a noted failure to adapt its interpretation of graffiti in to account for cases in which graffiti writing represents a symbol of acceptable youth edginess within the mainstream consciousness and, in relation to its cognate policing practices, the unintended consequence of displacement of writing. Collectively these findings serve to severely undermine this theoretical perspective and policing approach, whilst simultaneously emphasising the viability of cultural criminology as an alternative theoretical lens through which we may understand graffiti’s impacts on urban space.

As another concluding remark, we may broaden our discussions to note how cultural criminological is also useful in deconstructing local crime concerns about graffiti, which often use Broken Windows logic to characterise graffiti as a social threat, and the interests that tend to drive them to prominence. In particular, we may reflect on how political and economic interests appear to have, despite the insignificance of graffiti’s impact on criminality in the city in comparison to that of the city’s night-time drinking industry, established Broken Windows rationale used against writing as a common feature in authorities’ discussions of local crime. Indeed, the recognition of many central and local governments’ ‘morally dubious relationship with the UK drinks industry’, amidst ‘desperate attempts to keep the drinks industry satisfied and the revenues flowing in’ (both quotes Ferrell et al, 2008: 106), among such scholars appears highly relevant the state of crime and industry within the city. Any suspicions of local authorities willingly ignoring the problems caused by caused by the city’s vibrant night-time economy only appear to be confirmed amidst accusations of ‘deliberate attempts to bury’ (Cllr Nick Eden-Green quoted in Stafford, 12/7/2012: 2) this very issue being levelled against those running the city council. This has apparently included moves prevent evidence of the damage done to the city by night-time drinking being presented at local council meetings; leaving senior councillor Nick Eden-Green to claim that ‘this is the gagging and censorship of a problem that, clearly, Canterbury City Council are desperate to brush under the carpet’ (quoted in Stafford, 12/7/2012: 2). Where previous examples in various cities (see Austin, 2001, Snyder, 2009, Ferrell, 1996) show such local crime discussions being ‘shaped not just by the politics of city government, but by the economics of ownership and enterprise’ (Ferrell, 1996: 110), rather than the nature of the problem itself, there is clearly a precedent in this literature with which we can identify these attempts to simultaneously flag writing as a threat to local tranquillity and play down the damaging effect of the city’s (economically profitable!) night-time economy. It is for this reason that a cultural criminology that considers ‘the ‘constant interplay of moral entrepreneurship, political innovation and transgression’ (culturalcriminology.org, 2012) appears better equipped to unravel how such campaigns, ‘based not sound social research but on the values of the business and political forces’ (Ferrell, 1996: 135), are sustained. In light of this, such concerns should also be considered by those looking to deconstruct the emergence of Broken Windows rhetoric in future research.

How, then, can better we understand graffiti’s impact on its environment? The results of this research indicate that the answer lies in the development of a more reflexive approach towards, and contextualised understanding of, crime and disorder than would stem from Broken Windows logic. This has been evident in the value found in cultural criminology’s interpretive approach to understanding of crime that emerged through the discussions developed in Chapter V; where its proved useful in explaining both how writing may often carry meanings other than that of disorder, and the processes of crime displacement that leave Broken Windows policing against writing as an exercise in futility. Cultural criminology’s appreciation of both immediate spatial and wider cultural context in this regard, as well as that emphasised through Ralph Taylor (1999) and Peter St. Jean’s (2007) previous work on Broken Windows theory, demonstrates the importance of accommodating for the variety of meanings that writing can carry. This research has demonstrated that graffiti’s impact on its environment varies; often being largely dependent a variety of factors, including differing audiences and the styles of the writing itself. Thus, it is a theoretical approach such as a contextually aware cultural criminology, one that moves beyond Broken Windows’ blunt one-size fits all approach, that appears to be best equipped to explain the full extent of the relationship between graffiti, space and, beyond that, culture.

Outside of Broken Windows policing, what other strategies appear to offer a more effective method of reducing crime within the city? The intricacies lying within urban space and graffiti writing that suggest that such an appreciation of setting should be a central feature of approaches used to reduce crime in parts of Canterbury. Proponents of such a programme have previously claimed that Broken Windows policing represents ‘a method to reduce crime and disorder that would work in any city’ (William Bratton cited in Becket and Godoy, 2010: 291), and is capable of tackling a variety of crimes: including looting in post-Hurricane Katrina New Orleans (Woolner, 2003); arson (Thomas et al, 2011) and, of course; graffiti (Sloan-Hewitt and Kelling, 1990). The evidence found here, however, suggests that, in Canterbury, a programme of Broken Windows policing unsuccessfully attempts to reduce crime by controlling graffiti, amongst other signs of so-called ‘disorder’. This is because it appears that graffiti actually has little effect in causing crime in the city. The implications of this for policing in the city are neither to accept, as a rule of thumb, that graffiti has no role at all in encouraging other forms of criminality in urban spaces, nor that it universally does. A more fruitful line of understanding and, subsequently, policing may be for local policing practitioners to pay closer attention to their own cities: the intricacies in the flow of criminal behaviour around their streets and; the policing approaches that can be tailored to these conditions. That is, by acknowledging that ‘incivilities and related problems are highly context-specific and should be addressed accordingly’ (Gau and Pratt, 2010: 758), authorities should be committed to identifying specifically where graffiti is a problem, where it is relatively harmless, and where other factors evidently play a larger role in breeding criminality. This should not mean, however, that we reduce to our understanding of crime purely to the identification of ‘hotspots’ of crime on crime maps: the weaknesses of such a quantitative approach were outlined in Chapter III (also see Hayward, 2004: 99-100). Thus, it is recommended here that police pursue a methodological approach similar that taken in this project: using a mixture of statistical information, local knowledge and first-hand observation to develop a rounded appreciation of where crime flourishes, why, and what can be done to reduce it.

I would also argue that a continued endeavour to understand crime within its contexts, undertaken here and previously emphasised by the likes of Ferrell et al (2008), Gau and Pratt (2010), Peter St. Jean (2007) and Ralph Taylor (1999, 2001) among others, should lay at the heart of future scholarly research concerned either with testing the Broken Windows thesis and policing, or understanding the impact of particular variances of ‘disorder’ on urban space. This thesis has offered a critique of Broken Windows theory that remains limited in its scope, being limited to graffiti and the city of Canterbury. However, it is through a variety of such studies that perhaps heralds the greatest potential for developing a useful body of knowledge in this regard: where future studies should be tasked with similarly adding to a mosaic of strategic understanding which can provide precise and demonstrably accurate insights in to the causes of particular crimes, and the contexts in which they flourish, rather than remain inhibited by a stifling and erroneous Broken Windows logic.

Bibliography

Agee, J. and Evans, W. (1960). Let Us Now Praise Famous Men. New York: Ballantine.

Alvelos, H. (2004). ‘The desert of imagination in the city of signs’ in J. Ferrell et al. (eds), Cultural Criminology Unleashed. London: Glass House.

Austin, J. (2001). Taking the Train: How Graffiti Art Became and Urban Crisis in New York City. New York: University of Columbia Press.

Barker, V. (2010). ‘Explaining the Great American Crime Decline: A Review of Blumstein and Wallman, Goldberger and Rosenfeld, and Zimring’ in Law and Social Inquiry, 35(2), pp.489-516.

Beckett, K. and Godoy, A. (2010). ‘A Tale of Two Cities: A Comparative Analysis of Quality of Life Initiatives in New York and Bogota’ in Urban Studies, 47(2), pp.277-301.

Blumstein, A. and Wallman, J. (2006). The Crime Drop in America, Revised Edition. New York: Cambridge University Press.

Bottoms A. E. and Wiles P. (2002) ‘Environmental criminology’, in M .Maguire, R. Morgan, R. Reiner (eds) in The Oxford Handbook of Criminology, 3rd edition. Oxford: Oxford University Press.

Buchanan, I. (2000). Michel de Certeau: Cultural Theorist. London: Sage.

Burgess, E. W. (1925). ‘The Growth of the City’, in R. Park, E.W. Burgess and R.D.McKenzie (eds), The City. Chicago: University of Chicago Press.

Cerdia, M., Tracy, M., Messner, S.F., Vlahov, D., Tardiff, K. and Galea, S. (2009). ‘Misdemeanour Policing, Physical Disorder, and Grun-related Homicide: A Spatial Analytic Test of “Broken Windows” Theory’ in Epidemiology, 20(4), pp.533-541.

Chappell, A.T., Monk-Turner, E. and Payne, B.K. (2011). ‘Broken Windows or Window Breakers: The influence of physical and social disorder on quality of life’ in Justice Quarterly, 28(3), pp.522-540.

Cicourel, A.V. (1968). The Social Organisation of Juvenile Justice. London: Transaction Publishers.

Corman, H. and Mocan, N. (2005). ‘Carrots, Sticks and Broken Windows’ in Journal of Law and Economics, 48(1), pp. 235-66.

David, E. (2007). ‘Signs of resistance: marking public space through a renewed cultural activism’ in G. Stancak (ed.), Visual Research Methods. Los Angeles: Sage.

de Certeau, M. (1984). The Practice of Everyday Life. Berkeley, CA: University of California Press.

Delis, M. (2010). ‘James Q. Wilson’ in K. Hayward, S. Maruna and J. Mooney (eds), 50 Key Thinkers in Criminology. London: Routledge.

Department for Transport (2010). Vandalism, Graffiti and Environmental Nuisance on Public Transport.

Duneier, M. (2000). Sidewalk. London: Farrar, Straus and Giroux.

Eck, J. and Maguire, E. (2006). ‘Have changes in policing reduced violent crime? An assessment of the evidence’ in A. Blumstein and J. Wallman (eds.), The crime drop in America, revised edition. New York: Cambridge University Press.

Ferrell, J. (1990). ‘Bombers’ Confidential: Interview with Eye Six and Rasta 68 (Part One)’ in Clot, 1, pp.10-11.

Ferrell, J. (1996). Crimes of Style: Urban Graffiti and the Politics of Criminality. Boston: Northeastern University Press.

Ferrell, J. (2005). ‘Crime and Culture’ in C. Hale, K.J. Hayward, A. Wahidin, E. Wincup (eds), Criminology. Oxford: Oxford University Press.

Ferrell, J., Hayward, K. and Young, J. (2008). Cultural Criminology. London: Sage.

Gau, J.M. and Pratt, T.C. (2010). ‘Revisiting Broken Windows Theory: Examining the Sources of the Discriminant Validity of Perceived Disorder and Crime’ in Journal of Criminal Justice, 38(1), pp.758-766.

Golub, A., Johnson, B.D. and Taylor, A. (2003). ‘Quality-of-life policing: do offenders get the message?’ in Policing, 26(1), pp.690-707.

Harcourt, B.E. (1998). ‘Reflecting on the subject: A critique of the social influence conception of deterrence, the broken windows theory and, order-maintenance policing New York style’ in Michigan Law Review, 97(1), pp.1-51.

Harcourt, B.E. and Ludwig, J. (2006). ‘Broken Windows: New evidence from New York City and a five-city social experiment’ in University of Chicago Law Review, 73 pp.271-320.

Harries, D. (1999). Mapping Crime: Principles and Practice. Washington DC: National Institute of Justice.

Hayward, K. (2004). City Limits: Crime Consumer Culture and the Urban Experience. London: Glasshouse Press.

Hayward, K. (2007). ‘Situational Crime Prevention and its Discontents: Rational Choice Theory versus a ‘Culture of Now’ in Social Policy and Administration, 43(7), pp. 232-250.

Hayward, K. (2012). ‘A Response to Farrell’ in Social Policy and Administration, 46(1), pp. 21–34.

Hayward, K. (2012a). ‘Five Spaces of Cultural Criminology’ in British Journal of Criminology, 52(1), pp. 441-462.

Hayward, K. and Presdee, M. (eds.) (2009). Framing Crime: Cultural Criminology and the Image. London: Taylor and Francis.

Hinkle, J.C. and Weisburd, D. (2008). ‘The irony of broken windows policing: A micro-place study of the relationship between disorder, focused police crackdowns and fear of crime’ in Journal of Criminal Justice, 36(1), pp.503-512.

HM Government (2010). Tackling Anti-Social Behaviour.

Hobbs, D. Winlow, S., Hadfield, P. and Lister, S. (2005). ‘Violent Hypocrisy: Post Industrialism, and the Night-time Economy’ in European Journal of Criminology, 2(2), pp.161-168.

Home Office (2010). Tackling Graffiti: One of the Key Drivers of ASB.

Hope, T. (2005). ‘What do Crime Statistics tell us?’ in C. Hale, K. Hayward, A. Wahidin and E. Wincup (eds), Criminology. Oxford: Oxford University Press.

Jang, H., Hoover, L.T. and Lawton, B.A. (2008). ‘Effect of broken windows enforcement on clearance rates’ in Journal of Criminal Justice, 36(1), pp.529-538.

Karmen, A. (2000). New York murder mystery: The true story behind the crime crash of the 1990s. New York: New York University Press.

Katz, C.M., Webb, V.J. and Schaefer, D.R. (2001). ‘An Assessment of the impact of quality-of-life policing on crime and disorder’ in Justice Quarterly, 18(1), pp.825-876.

Katz, J. (1988). Seductions of Crime: Moral and Sensual Attractions in Doing Evil. New York: Basic Books.

Kelling, G. and Coles (1997). Fixing Broken Windows: Restoring Order And Reducing Crime In Our Communities. London: Simon and Schuster.

Kelling, G. and Sousa, W.H. (2001). Do police matter? An analysis of the impact of New York City’s police reforms (Civic Report No.22). New York: Manhattan Institute for Policy Research.

Lea, J. and Young, J. (1984). What is to be done about Law and Order? London: Pluto Press.

Lyng, S. (2005). Edgework: the sociology of risk taking. Oxon: Routledge

Maguire, M. (2002). ‘Crime Statistics: the ‘data explosion’ and its implications’ in M. Maguire, R. Morgan and R. Reiner (eds) The Oxford Handbook of Criminology, Third Edition. Oxon: Oxford University Press.

Manning, P.K. (2001). ‘Theorising policing: The drama and myth of crime control in the NYPD’ in Theoretical Criminology, 5(3), pp.315-344.

Matthews, R. (1992) 'Replacing Broken Windows: Crime, Incivilities and Urban Change' in R. Matthews and J. Young (eds.), Issues in Realist Criminology. London: Sage.

McLaughlin, E. (2008). ‘Broken Windows’ in E. McLaughlin and J. Muncie (eds), The Sage Dictionary of Criminology, second edition. London: Sage.

McLaughlin, E. (2008a). ‘COMPSTAT’ in E. McLaughlin and J. Muncie (eds), The Sage Dictionary of Criminology, second edition. London: Sage.

McLintock, F.H. (1963). Crimes of Violence. New York: St. Martin’s Press.

Mills, C.W. (1959). The Sociological Imagination. Oxford: Oxford University Press.

Mills, M. (2011). Picking up the ‘Pieces’: What the internet tells us about the causes of graffiti writing and the effectiveness of situational crime prevention. Unpublished Undergraduate Dissertation.

Park, R. E. (1925). ‘The City as a Social Laboratory’, in T. V. Smith and L. White (eds), Chicago: An Experiment in Social Science Research. Chicago: Chicago University Press.

Pitner, R.O., Yu, M. and Brown, E. (2012). ‘Making neighbourhoods safer: examining predictors of residents’ concerns about neighbourhood safety’ in Journal of Environmental Psychology, 32(1), pp.43-49.

Presdee, M. (1999). Cultural Criminology and the Carnival of Crime. London: Routledge.

Sampson, R.T. and Raudenbush, S.W. (1999). ‘Systematic social observation of public spaces: A new look at disorder in urban neighbourhoods’ in American Journal of Sociology, 105(1), pp.603-651.

Sampson, R.T. and Raudenbush, S.W. (2001). Disorder in Urban Neighbourhoods: Does it Lead to Crime? Washington National Institute of Justice.

Sampson, R.T. and Raudenbush, S.W. (2004). ‘Neighbourhood Stigma and the Social Construction of “Broken Windows”’ in Social Psychology Quarterly, 67(4), pp.319-342.

Skogan, W. (1990). Disorder and Crime: Crime and the spiral of urban decay in American neighbourhoods. New York: Free Press.

Sloan-Howitt, M. and Kelling, G. (1990), Subway Graffiti in New York City: ‘Getting up’ vs. ‘Meaning It and Cleaning It’ in Security Journal, 1, pp.131–6.

Smith, W.E. (1998). W Eugene Smith: Photographs1934-1975, edited by G. Mora and J.T. Hill. New York: Harry Abrams Publishers.

Snyder, G.J. (2006). ‘Graffiti media and the perpetuation of an illegal subculture’ in Crime, Media, Culture, 2(1), pp.93-101.

Snyder, G.J. (2009). Graffiti Lives: Beyond the tag in New York’s urban underground. New York: NYU Press.

Squires, P. (2008). ‘Anti-Social Behaviour’ in in E. McLaughlin and J. Muncie (eds), The Sage Dictionary of Criminology, second edition. London: Sage.

St. Jean, P.K.B. (2007). Pockets of Crime: Broken Windows, Collective Efficacy, and the Criminal Point of View. London: University of Chicago Press.

Taylor, R. (1999). Crime, grime, fear and decline: a longitudinal look (Research in Brief). Washington, D.C.: National Institute of Justice.

Taylor, R. (2001). Breaking Away from Broken Windows: Baltimore neighbourhoods and the national fight against crime, grime, fear and decline. Boulder, CO: Westview Press.

Thomas, S.T., Butry, D.T. and Prestemon, J.P. (2011). ‘Enticing Arsonists with Broken Windows’ in Fire Technology, 47(1), pp.255-273.

Vergara, C.J. (1995). The New American Ghetto. New Brunswick, NJ: Rutgers University Press.

Wacquant, L. (2003). ‘Toward a dictatorship over the poor? Notes on the penalisation of poverty in Brazil’ in Punishment and Society, 5(2), pp.197-205.

Wasterfors, D. (2009). ‘Book Review: Peter K. B. St. Jean Pockets of Crime: Broken Windows, Collective Efficacy, and the Criminal Point of View’ in Acta Sociologica, 52(1), pp.92-94.

Willis, J. (2002). ‘Breaking Away from Broken Windows: Baltimore Neighbourhoods and the Nationwide Fight against Crime, Grime, Fear, and Decline by Ralph B. Taylor’ in Contemporary Sociology, 31(2), pp.220-221.

Wilson, J.Q. (1975). Thinking About Crime. New York: Basic Books.

Wilson, J.Q. and Kelling, G. (1982). ‘Broken Windows’ in Atlantic Monthly, 249, pp. 29-38.

Winlow, S. and Hall, S. (2006). Violent Night. Oxford: Berg.

Yang, S. (2007). Causal or merely co-existing: a longitudinal study of disorder and violence at places. Unpublished doctoral dissertation, University of Maryland, College Park.

Young, J. (2011). The Criminological Imagination. London: Polity Press.

Zimring, F.E. (2007). The Great American Crime Decline. New York: Oxford University Press.

Media Sources:

Claridge, A. and Alston, K. (18/3/2010). ‘Student caught scrawling tag after night out in pub’ in The Kentish Gazette.

Napler, S. (19/5/2011). ‘Volunteers get stuck in to give path a makeover’ in The Kentish Gazette.

Stafford, L. (12/7/2012). ‘Council Accused of ‘Burying’ Problem’ in The Canterbury Times.

The Kentish Gazette (25/3/2010). ‘Opinion’ in The Kentish Gazette.

The Kentish Gazette (2/6/2011). ‘I’ll clean up the graffiti myself’ in The Kentish Gazette.

Warren, G. (23/8/2012). ‘Traffic wardens take on powers to dish out fines’ in The Kentish Gazette.

Online Sources: