Abstract
Much policy discourse concentrates on the contribution police make to keeping people safe. Often, this means minimizing fear of crime. Yet, more expansive accounts stress the extent to which deeper-rooted forms of security and belonging might also be important ‘outcomes’ of police activity. Using data collected from a survey of residents of a mid-sized English town, Macclesfield in Cheshire, we consider the extent to which evaluations of policing are associated with (1) a ‘shallow’ sense of security—roughly speaking, feeling safe—and (2) a ‘deeper’ sense of security—being comfortable in, and with, one’s environment. Focussing more accurately on the forms of safety and security police can hope to ‘produce’ opens up space for consideration of the ends they seek as well as the means they use.
INTRODUCTION
Much current policing discourse focusses on the production of safety and keeping people free from harm. The London Metropolitan Police’s self-professed mission is to ‘Keep London Safe for everyone’; Greater Manchester Police seeks to ‘continually strive to fight, prevent and reduce crime and harm and keep people safe so that Greater Manchester can be a safer and welcoming place’.1 The Police and Crime Commissioner (PCC) covering the location of the current study has a stated aim of working with police to make the county ‘an even safer place to live, work and visit’.
These are not unreasonable policy aims. It clearly is the job of the police, alongside a range of other actors, to work towards increasing public safety. Yet, to ‘be safe’ only really means something if the individual or group concerned feels, subjectively, this to be the case (Innes 2004). That people should feel safer is now also widely acknowledged as a commitment of police, community safety partnerships and other security providers. Many contemporary policing strategies—community, neighbourhood or reassurance policing, problem-oriented policing, the growing emphasis on vulnerability and trauma-informed policing—aim to enhance both ‘objective’ and ‘subjective’ security. Indeed, the current policing ‘moment’ seems to involve a shifting set of ideas concerning what the police and other agencies are for and what they should seek to achieve. Foregrounding of the concept of ‘harm reduction’ as a distinct discursive element alongside ‘crime’, the content of many acknowledged current priorities such as exploitation, cybercrime and domestic and gender-based violence indicates an increased emphasis on, or at least recognition of, the role of police in generating expansive, positive, forms of security. Yet, such developments are almost invariably counterposed against pressures towards more restrictive policing styles that concentrate on ‘fighting crime’ and the aggressive assertion of order; of the kind, for example, contained in provisions of the Policing, Crime, Sentencing and Courts Act (2022), or recent (and recurring) pressure from government to increase the use of stop and search powers.
Loader (2006) argues that there are significant tensions in the way the apparently fundamental injunction for police to ‘keep people safe’ unfolds in practice. Because such policy aims often imply activities associated with the demonstration of effectiveness, reach, presence and control, they point towards pervasive modes of policing that seek to demonstrate instrumental effectiveness in the ‘fight against crime’ that, ultimately, entail encroachment on the rights and liberties of those who are the particular targets of police activity—with the attendant risk of making them feel less secure. Modes of policing that seek to assert order and control are often experienced as unfair by those exposed to them, serving to marginalize, exclude and indeed harm (Tyler and Fagan 2008; Geller et al. 2014). When policing is experienced as unfair, people are less likely to feel they belong to and are included within the wider social groups police represent (Bradford 2014; Madon et al. 2016), discouraging engagement with and within these groups and ‘pro-social’ activity in support of them (Tyler 2009). All this may, in turn, diminish their objective and subjective safety.
Policing that might make some people or groups feel safer may thus make others feel less safe, such that choices in the deployment of resources and priorities can be more contentious than might first appear. Two further issues are, first, that ‘keeping people safe’ assumes public demands for safety can be adequately met by police and associated actors within current resource and ethical constraints. This may not actually be the case. Second, the fundamental premise that there is indeed a positive link between experiences of policing and people’s sense of safety and security is possibly mistaken, or at least overstated. The evidence that a significant section of the population will feel safer if they perceive policing to be present, active and/or effective is less clear than might be expected (Zhao et al. 2002; Scheider et al. 2003; Gill et al. 2014; Crowl 2017).
Underlying all this, Loader (2006) suggests, is a misunderstanding of what policing can and arguably should produce. Rather than merely a ‘wide but shallow’ sense of safety and security—‘freedom from’ fear—police activity can also be linked to deeper and more productive forms of community and inclusion that constitute the ‘freedom to’ act that stems from a sense of secure belonging. Our contribution in this article is to explore these issues as they play out in one particular place, the town of Macclesfield in Cheshire, England. Using data from a local crime and security survey that provides fine-grained data on local concerns, issues and experiences, we consider whether perceptions and understandings of policing—broadly speaking, whether people trust police to be effective and fair—are associated with residents’ feelings of security. Overall, we find little evidence for an association between trust in police and feelings of safety or worry about crime. But we find a strong link between greater trust in police fairness, specifically, and a sense of belonging and attachment to place that, we argue, represents a deeper sense of security.
POLICING AND THE PROVISION OF SECURITY
It seems axiomatic that policing in contemporary Britain is about providing safety and security. ‘Dealing with’ or ‘fighting’ crime, but also preventing it from happening in the first place and securing the general order that underpins social, economic and cultural life—these are all aspects of how police see their mission (Bowling et al. 2019), and, by and large, how the public see policing (Girling et al. 2000; Higgins 2019). In this article, we leave to one side the contribution police activity may or may not make to the objective safety of individuals, communities and the state to focus on the link between perceptions of policing and subjective feelings of safety and security among the policed. How, and to what extent, might evaluations of police activity—which for brevity’s sake we label trust in the police—generate, reproduce or undermine subjective safety and security?
To answer this question, we focus on two rather contrasting notions of security. The first aligns with the wide but shallow sense of safety that Loader (2006; Loader and Walker 2007) associates with pervasive policing styles, and the second with a deeper sense of secure belonging and social embeddedness that should, arguably, be seen as a normatively more appropriate outcome of police activity.
‘Wide but shallow’ security
On the first account, policy-makers and practitioners take an expansive view of what policing can and should provide protection from—ranging from serious violence to shop-lifting, terrorism to anti-social behaviour—and seek to insert policing in as many situations and contexts as possible to make people feel safe from these threats. While there are of course debates about where police should focus attention, concerning most obviously the need for prioritization in the face of resource constraints (Higgins 2019), this rarely takes the form of a discussion of the appropriate boundaries of policing (debates about ‘defunding’ the police have largely failed to take off in the United Kingdom, although see Fleetwood and Lea 2022). Instead, the police are positioned as the best and most appropriate way to provide for a sense of security and safety across a wide range of potential threats. By demonstrating effectiveness, they can both deter crime and make ‘the law-abiding’ feel safer.
Many policies and strategies of recent decades have placed a significant emphasis on reassuring the public (Tuffin et al. 2006; Quinton and Tuffin 2007), improving quality of life by making neighbourhoods look and feel safer and more orderly (Kelling and Coles 1997; Harcourt 2005), and/or invoking other actors to help create more secure physical and social environments (Bullock et al. 2021). More recently, police have begun to focus on targeting—crime hotspots (Weisburd et al. 2012), prolific and high harm offenders (Sherman 2019), the ‘vulnerable’ (Keay and Kirby 2018). Still, though, such efforts are framed within the need to provide for the safety of the wider community. Moreover, many involve increased police presence, surveillance and intervention in particular places or on particular individuals, meaning ambient policing is significantly increased in some contexts compared to others.
Yet, evidence on the association between perceptions and experiences of policing, on the one hand, and feelings of safety and security, on the other, is decidedly ambivalent. There is certainly evidence that direct, physical, police presence reduces fear and feelings of unsafety (Zhao et al. 2002; Doyle et al. 2016)—but police cannot of course always be present. More widely, many studies have focussed on ‘community-oriented policing’ (COP), broadly defined, and some have identified an association between trust in, and other perceptions of, police and feelings of safety and/or fear of crime (Roh and Oliver 2005; Cho and Park 2019; Lee et al. 2020; Carter and Wolfe 2021). However, a meta-analysis by Gill et al. (2014) concluded that COP interventions were ‘associated with only a small, non-significant improvement in citizens’ feelings of safety’ (p. 18) (see also Scheider et al. 2003; Crowl 2017; Rukus et al. 2018). Despite the sustained policy focus on making people feel safer through policing, then, the evidence of such an effect is rather weak, and indeed contradictory.
One reason for these contrasting findings may be the wide variety, and often poor conceptualization and operationalization, of measures of ‘safety’ and ‘fear of crime’. The former is frequently measured by single survey items covering feeling safe while walking after dark or during the day (e.g. Tuffin et al. 2006), which have been criticized as imprecise, mixing fear and risk, and as only applicable to certain people. Fear of crime is often measured either by items referencing ‘worry’ (e.g. Lee et al. 2020), or by the same ‘walking after dark’ type questions used to measure safety (e.g. Torres and Vogel 2001). Overall, the concept of ‘worry’ seems to be increasingly preferred, and is argued to capture ‘both evaluations of immediate situations and anxiety-producing thoughts about future events’ (Buil-Gil et al. 2021: 277; see also Jackson and Gouseti 2016).
In this article, we utilize measures of both subjective safety and worry about crime. The first uses multiple survey items referencing specific events or behaviours that may make people feel unsafe in their immediate social and physical environment—respondents’ direct assessments of (un)safety in the place they live. The second replicates measures of worry about crime used in many other studies, to enable comparison with earlier research and to broaden our consideration to concern about the potential risk of victimization and the anxiety this may cause.
Security as attachment and belonging
The second notion of security references a deeper and more constitutive sense of attachment and belonging. Our starting point is research within the procedural justice paradigm that has recently begun to describe associations between people’s perceptions and experiences of policing and their sense of inclusion and belonging within social categories the police are said to represent (Kyprianides et al. 2021; Murphy et al. 2022). Drawing on the basic social psychological underpinning of procedural justice theory (Tyler and Blader 2000; Blader and Tyler 2009), and sociological accounts of the symbolic meaning and importance of police (at least in a country such as England: Loader 2006; Bowling et al. 2019), this work has argued that: (1) police represent important social categories, such as nation, state and community; (2) many people feel some sense of affiliation with and belonging to these categories, and are attuned to messages about their status in relation to these groups; (3) the behaviour of police, as ‘proto-typical group representatives’ (Sunshine and Tyler 2003), communicates status, inclusion and value in relation to them and (4) procedural justice generates and enhances a sense of shared group membership between police and public that is closely linked to trust and legitimacy, and which motivates compliance and cooperation.
The procedural justice of police activity—whether people feel they are treated with dignity and respect, afforded a voice, and that police make decisions in a neutral, unbiased and transparent fashion—thus shapes processes of inclusion or exclusion. And the identities and categories involved (citizenship, community) suggest not just a ‘passive’ sense of inclusion and the feelings of security this might bring, but also a more active sense of membership, and the agency that comes with feeling one is recognized as a rights-bearer embedded in appropriately constituted legal and political frameworks (Justice and Meares 2014; Meares 2017). Procedural injustice, in contrast, induces a sense of exclusion among those who experience unfair policing—who may also feel unsafe specifically as a result of such experiences (Epp et al. 2014; Soss and Weaver 2017).
Research on procedural justice often starts from a premise of personal, face-to-face interaction, and considers how the quality of such encounters feeds into trust, legitimacy and other outcomes (e.g. Mazerolle et al. 2013). However, a wide range of studies have found that general perceptions and evaluations of police fairness are also associated with people’s sense of who they are and their relationships with authorities (see Walters and Bolger 2019). In this article, we concentrate on the general evaluations people make of police, which we label trust. We measure people’s general evaluations of police fairness, and also their evaluations of police effectiveness: ‘perceptions of police trustworthiness’ might therefore be a better although less concise term (see Hamm et al. 2017).
The social categories associated with police included in prior research have tended to revolve around the idea of the ‘law-abiding citizen’ (e.g. Bradford et al. 2014) or the nation state (e.g. ‘Australianess’, Murphy et al. 2022). We focus here, though, on the extent to which relationships with police may be associated with more local and embedded understandings of inclusion and belonging. Bradford (2014) found that general perceptions of police procedural justice among young ethnic minority men living in London were associated with feelings of belonging in that place, and we hypothesize that trust in police fairness will be associated with a sense of belonging to and comfort within the places people live—feelings that resonate with the ‘deep’ notion of security described by Loader (2006).
We have two reasons for forming this hypothesis. First, it is plausible that police do in fact represent more local senses of identity and community than are covered by concepts such as ‘nation’ and ‘citizen’. Research has shown, for example, that trust and confidence in police is heavily shaped by the quality of local environments, implying that police seem efficacious and successful when community is felt to be strong and effective (see below). The police do not represent merely an abstract or distant ‘state’—more local and embedded forms of identification are also in play. Similarly, albeit to reverse the hypothesized relationship, a number of studies have shown that trust in the police is linked to greater collective efficacy (Sampson et al. 1997) in local areas. Higher levels of collective efficacy rely on the existence of social bonds between residents and a perceived ability to rely on (i.e. trust in) authorities to intervene if need be (see Yesberg and Bradford 2021). In sum, there seems to be an affective link between police and ‘neighbourhood’ that suggests relationships with the former might shape how the latter is experienced, evaluated and integrated into one’s sense of self.
The second reason relies less on processes of identification and categorization, although it may be closely related to them. Put simply, it may be that if one feels one is governed by legal and other institutions that are fair, trustworthy and legitimate, one is more likely to feel safe and at ease in one’s social and physical environment. Certainly, research in policing contexts rather different to our study site has found that experiences of over-weening, unfair and aggressive policing are linked with higher levels of emotional and physical stress, symptoms of the almost constant need to be on guard against police that is experienced by, for example, many young black men in the United States (Geller et al. 2014; Stutts and Cohen 2022). More widely, perceptions of procedural justice, and the experiences of policing to which they are linked, may be implicated in the production and reproduction of generalized trust relations, where trust signifies a capacity to both repose confidence in the predictable continuity of social relations and expect a degree of predictability in interactions between agents and systems (Giddens 1991). As Giddens and others (e.g. Möllering 2006) have argued, trust provides for the navigation of risk, the reduction of uncertainty and the experience of social contexts as coherent, cohesive and navigable.
Perceptions of (trust in) police could therefore be associated with the ‘deeper’ senses of security outlined above. To operationalize this notion of security, which wraps up both identification with a place and feeling safe and secure within it, we adopt a concept from human geography: sense of place, the affective bond between people and a specific location (Žlender and Gemin 2020: 2). The literature on this issue is large, complex and rather contradictory, but Žlender and Gemin identify three central ideas (also found in associated work on place attachment in environmental psychology, e.g. Hidalgo and Hernandez 2001; Lewicka 2011) that are of particular use to us here. The first is sense of place itself, people’s happiness, and identification with, place, and, crucially, their understanding that they can be themselves in that place (Jorgensen and Stedman 2001). Second is place appreciation, affective responses to a place and the ability to use and enjoy it. Third, place involvement relates to cognitive understandings of a place and one’s place within it, which can also be termed belongness or rootedness. Sense of place thus provides a conceptual bridge between the forms of identification covered in the procedural justice literature and the concept of security developed by Loader (2006) and others. It references identification with place alongside a sense of secure belonging—not freedom from harm, but freedom to act within a particular context in which one feels at ease and able to behave according to one’s wishes.
The importance of context
Research into fear of crime, feelings of safety and perceptions and experiences of policing has frequently stressed the importance of the local, specific and often bounded contexts within which they are generated, motivated and/or played out. The types of places people live, and how they experience them, can be central to shaping their views of policing, crime and safety. The current article needs to take this into account.
Focussing on individual’s perceptions and understandings of their local environment, two inter-related ‘neighbourhood concerns’ are central to much of this literature. The first is low-level disorder. Issues such as public drinking or drug use, litter, vandalism and often simply ‘teenagers hanging around’ seem to loom large in people’s concerns about wider and ‘bigger’ questions of crime, safety and risk, not least because ‘disorder’ may signal ‘danger’ (Girling et al. 2000). This well-established association provides a link between the two notions of security outlined above, cautioning against seeing them as entirely distinct. Perceptions of disorder, fear of crime and related constructs can serve an expressive function, a way for people to talk not only about disorder and crime per se, but also both the state of their immediate social environment and a wider set of concerns about social cohesion and the nature of social change (Jackson 2004; Farrall et al. 2009). These concerns resonate with notions of (in)security that involve not simply more or less immediate threat or risk, but also a deeper sense of attachment and ease within a social context (or the absence thereof).
Perceptions of disorder have also been shown to be strongly associated with trust in the police across multiple contexts (e.g. Jackson et al. 2013). Here, interpretations often involve that same sense of danger, and the fact that low-level disorder may signal both a failure of policing in a direct sense, and failures of the community that become failures of policing (because police and community are associated with each other in many people’s minds). Perceptions of disorder feature in what has been termed a ‘neo-Durkheimian’ model of attitudes towards police (Jackson and Sunshine 2007). Here, people look to the police to defend and uphold the moral structure of society, and ‘day-to-day’ concerns about social order and control, rather than crime per se, are central to their willingness to trust police. They lose faith and confidence in the police when community values, beliefs and norms are seen to be deteriorating.
The second neighbourhood concern is also included in the neo-Durkhemian model—collective efficacy. Collective efficacy is conceptualized as a task-specific property of groups and/or neighbourhoods (Sampson et al. 1997; Hipp and Wickes 2017); specifically, a combination of ‘social cohesion among neighbours combined with their willingness to intervene on behalf of the common good’ (Sampson et al. 1997: 918). Social cohesion relates to the ties between neighbours and mutual trust, whereas informal social control relates to individuals’ or neighbours’ willingness to act to address neighbourhood problems; for example, to break up fights or intervene if children are skipping school. We concentrate here on individual level perceptions of collective efficacy, which have consistently been linked to trust in the police (Jackson et al. 2013; Nix et al. 2015; Kochel 2018). The argument is, again, that low perceived collective efficacy indicates community failings that become police failings. When people believe their community cannot regulate itself, they seem inclined to ascribe this failure, at least in part, to the police. Perhaps unsurprisingly, collective efficacy has also been associated with fear of crime and related variables (Ferguson and Mindel 2007; Brunton-Smith et al. 2018).
RESEARCH QUESTIONS
Two research questions guide our analysis. First, is there an association between trust in the police and a sense of security as freedom from risk, or as worry about crime? If so, we would expect trust in police effectiveness to be most salient. Second, is there an association between trust in police and a deeper sense of security, or people’s sense of place? If so, we would expect trust in police fairness to be more important.
Figure 1 shows a conceptual model that presents one way to summarize the potential relationships between the concepts we have been discussing. Other relationships are of course possible, but the figure displays a plausible set of associations to guide analysis.
Conceptual map.
Starting on the left of Figure 1, we position perceptions of ASB and collective efficacy as in some sense foundational, neighbourhood concerns that comprise people’s apprehensions of their social and physical environment and the state of social relations within it, which seem likely to underpin the way they think about questions of crime, policing, safety and place. In turn, trust in the police, and concerns about safety and crime, are shaped by perceptions of ASB and collective efficacy. We also assume that trust in the police influences the extent to which people have concerns about crime and their personal safety. Finally, we assume that sense of place is, to at least some degree, a product of all these factors. People’s ease in, happiness with and sense of belonging to the place they live, that is, is shaped by their concerns about crime and disorder, their perceptions of social cohesion and collective motivations and intentions, and their sense of trust in an institution fundamentally associated with these social processes and outcomes—the police.
DATA AND MEASURES
Macclesfield is a mid-sized town in the north of England with a population of around 53,000 people. It is in some senses typical of many towns in its region and across the country. The remnants of old manufacturing industries sit alongside newer types of workplace, while a significant number of residents commute to nearby conurbations to work. There are pockets of significant deprivation, but it is generally considered a relatively affluent place (although less so than other nearby towns). The population is largely white. This paper is part of a 're-study' of Macclesfield - see Girling et al. (2000) for details of the earlier study, and Loader et al. (2023) for more on the current project.
The opinion survey company ORS was commissioned to conduct a face-to-face survey in the summer of 2021. Addresses were sampled randomly from each of Macclesfield’s 35 LSOAs (Lower Super Output Areas—see below). Pre-alert letters were sent to the preselected addresses; interviewers subsequently making up to three visits to each address to secure an interview. Interviewees were selected at random from the people living at each address. The response rate was relatively low, at 20 per cent. The timing of the interviews, and feedback from the ground, suggested that concerns about Covid-19 were an important factor supressing responses. Legal restrictions on indoor gatherings were still in place as the fieldwork commenced (the interviews were conducted in such a way as to be ‘Covid compliant’). The structure of the sample is shown in Table 1.
Structure of the sample
| Per cent | Per cent | ||
|---|---|---|---|
| Age | Highest qualification | ||
| 18–24 | 7 | Higher degree | 13 |
| 25–34 | 12 | Degree or equivalent | 26 |
| 35–44 | 13 | NVQ Level 3 or below | 14 |
| 45–54 | 17 | A level, AS level or equivalent | 11 |
| 55–64 | 19 | GCSE/O level/CSE | 17 |
| 65–74 | 18 | Apprenticeship | 6 |
| 75 or over | 14 | No qualifications | 10 |
| Refused | 1 | Other/Refused/Don’t know | 3 |
| Gender | Housing tenure | ||
| Female | 56 | Owner/occupier | 68 |
| Male | 44 | Social renter | 13 |
| Private renter | 14 | ||
| Ethnicity | Other | 5 | |
| White British | 93 | ||
| Other | 7 | Total (n) (=100%) | 427 |
| Per cent | Per cent | ||
|---|---|---|---|
| Age | Highest qualification | ||
| 18–24 | 7 | Higher degree | 13 |
| 25–34 | 12 | Degree or equivalent | 26 |
| 35–44 | 13 | NVQ Level 3 or below | 14 |
| 45–54 | 17 | A level, AS level or equivalent | 11 |
| 55–64 | 19 | GCSE/O level/CSE | 17 |
| 65–74 | 18 | Apprenticeship | 6 |
| 75 or over | 14 | No qualifications | 10 |
| Refused | 1 | Other/Refused/Don’t know | 3 |
| Gender | Housing tenure | ||
| Female | 56 | Owner/occupier | 68 |
| Male | 44 | Social renter | 13 |
| Private renter | 14 | ||
| Ethnicity | Other | 5 | |
| White British | 93 | ||
| Other | 7 | Total (n) (=100%) | 427 |
Structure of the sample
| Per cent | Per cent | ||
|---|---|---|---|
| Age | Highest qualification | ||
| 18–24 | 7 | Higher degree | 13 |
| 25–34 | 12 | Degree or equivalent | 26 |
| 35–44 | 13 | NVQ Level 3 or below | 14 |
| 45–54 | 17 | A level, AS level or equivalent | 11 |
| 55–64 | 19 | GCSE/O level/CSE | 17 |
| 65–74 | 18 | Apprenticeship | 6 |
| 75 or over | 14 | No qualifications | 10 |
| Refused | 1 | Other/Refused/Don’t know | 3 |
| Gender | Housing tenure | ||
| Female | 56 | Owner/occupier | 68 |
| Male | 44 | Social renter | 13 |
| Private renter | 14 | ||
| Ethnicity | Other | 5 | |
| White British | 93 | ||
| Other | 7 | Total (n) (=100%) | 427 |
| Per cent | Per cent | ||
|---|---|---|---|
| Age | Highest qualification | ||
| 18–24 | 7 | Higher degree | 13 |
| 25–34 | 12 | Degree or equivalent | 26 |
| 35–44 | 13 | NVQ Level 3 or below | 14 |
| 45–54 | 17 | A level, AS level or equivalent | 11 |
| 55–64 | 19 | GCSE/O level/CSE | 17 |
| 65–74 | 18 | Apprenticeship | 6 |
| 75 or over | 14 | No qualifications | 10 |
| Refused | 1 | Other/Refused/Don’t know | 3 |
| Gender | Housing tenure | ||
| Female | 56 | Owner/occupier | 68 |
| Male | 44 | Social renter | 13 |
| Private renter | 14 | ||
| Ethnicity | Other | 5 | |
| White British | 93 | ||
| Other | 7 | Total (n) (=100%) | 427 |
Constructs and measures
We used Confirmatory Factor Analysis in the package Mplus 7.2 to derive and validate our key measures. See Tables A1 and A2 for full wordings of the survey items used in this analysis.
Dependent variables
We have three primary dependent (endogenous) variables. First, sense of place is measured by seven items measuring sense of place, place attachment and rootedness (e.g. ‘I feel happiest when I am here’ and ‘I feel as if I’m able to move freely in this place’). Second, feeling unsafe was measured by eight items tapping into respondents’ sense that when using public space in Daleview they felt unsafe due to the behaviours of others, including ‘People drinking alcohol in the streets’ and ‘A lack of police presence’. Third, worry about crime was measured by six items assessing respondents concerns about crime (e.g. burglary, violent crime).
A three-factor model measuring these three constructs, with no cross-loadings, observed indicators set to ordinal, and Full Information Maximum Likelihood (FIML) estimation proved to be an adequate fit to the data (Chi2 = 384.4; df = 181; p < 0.0005; RMSEA = 0.05; CFI = 0.97; TLI = 0.96; SRMR = 0.06). All factor loadings were above 0.45 (see Table A1).
Independent variables
There are two exogenous variables. First, perceptions of disorder, measured by respondents’ perceptions of ‘how much of a problem’ behaviours like vandalism, noisy parties and littering were in their areas. Second, perceptions of collective efficacy, measured by six items tapping into social cohesion (e.g. ‘people in this neighbourhood can be trusted’) and collective action (e.g. ‘The people who live here can be relied upon to call the police if someone is acting suspiciously’).
Two further endogenous variables can also be construed as independent variables. Trust in police fairness was measured by five items tapping into perceptions of police procedural justice and of the fairness and appropriateness of the relationship between police and local communities (e.g. ‘They would treat you with respect if you had contact with them for any reason’ and ‘The police in your area listen to the concerns of local people’). Trust in police effectiveness was measured by five items measuring perceptions of the effectiveness of police in dealing with and preventing crime, and providing a visible presence.
A four-factor model measuring these four constructs, with no cross-loadings, observed indicators set to ordinal, and FIML proved to be an adequate fit to the data (Chi2 = 583.3; df = 225; p < 0.0005; RMSEA = 0.07; CFI = 0.97; TLI = 0.96; SRMR = 0.06), with all factor loadings above 0.40 (see Table A2). In this model, the latent variables trust in police fairness and trust in police effectiveness were highly correlated (r = 0.84, see Table 2), which raises concerns about discriminant validity. We re-estimated the model, combining the fairness and effectiveness constructs for a three-factor solution. Fit statistics from this model indicated a marginally worse fit (Chi2 = 803.9; df = 227; p < 0.0005; RMSEA = 0.08; CFI = 0.95; TLI = 0.94; SRMR = 0.07). We proceed with the four-factor model on conceptual and policy-related grounds—it seems important to retain a distinction between police fairness and effectiveness if possible.
Area-level factors
The perceptual indicators described above are unlikely to be ‘free-floating’. Neighbourhood concerns in particular will be tied to, and in a sense emergent from, the objective characteristics of the places in which people live. To take some account of area-level characteristics we use the 2019 Index of Multiple Deprivation (IMD), where deprivation data are available at the LSOA level (LSOAs are area units tied to the decennial Census). While they vary widely in physical size, across England and Wales the minimum population of an LSOA is 1,000, and the mean around 1,500. The IMD is the ‘official measure of relative deprivation’ in England (MHCLG 2019: 2). It is constructed from statistics covering eight domains: income; employment; health and disability; education, skills and training; crime; barriers to services and housing and living environment. Each LSOA is assigned an overall score and then ranked across the entire country, from 1 to 32,844, such that the value 1 represents the most deprived LSOA.
All LSOAs in Daleview are represented in the dataset, with an average of 12 respondents in each.2 We use the national IMD ranking of these LSOAs divided by 1,000 to make reporting results easier. It is important to underline that high scores on the IMD scale indicate lower deprivation.
Descriptive statistics
Table 2 presents a correlation matrix for the variables described above. Sense of place is moderately to strongly correlated with all the measures of trust in police and neighbourhood concerns, while feelings of safety and worry about crime are particularly strongly correlated with perceptions of disorder (and each other). There are moderate correlations between IMD and both disorder and collective efficacy; however, IMD is barely correlated with other variables.
Correlation matrix of latent variables
| Mean | SD | Min | Max | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Sense of place (1) | 0.00 | 0.64 | −2.88 | 1.70 | 1 | |||||||
| Police fairness (2) | −0.02 | 0.77 | −3.07 | 1.89 | 0.72 | 1 | ||||||
| Police effectiveness (3) | 0.00 | 0.75 | −3.07 | 1.92 | 0.62 | 0.84 | 1 | |||||
| Disorder (4) | 0.04 | 0.60 | −1.42 | 3.05 | −0.63 | −0.44 | −0.53 | 1 | ||||
| Collective efficacy (5) | −0.02 | 0.61 | −2.83 | 1.56 | 0.73 | 0.59 | 0.62 | −0.65 | 1 | |||
| Worry about crime (6) | 0.04 | 0.68 | −1.32 | 2.75 | −0.38 | −0.26 | −0.36 | 0.66 | −0.40 | 1 | ||
| Feeling unsafe (7) | 0.03 | 0.37 | −0.64 | 1.60 | −0.27 | −0.23 | −0.28 | 0.73 | −0.33 | 0.62 | 1 | |
| IMD (8) | 19.79 | 9.47 | 3.00 | 32.63 | 0.03 | 0.03 | 0.06 | −0.15 | 0.20 | −0.04 | −0.07 | 1 |
| Mean | SD | Min | Max | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Sense of place (1) | 0.00 | 0.64 | −2.88 | 1.70 | 1 | |||||||
| Police fairness (2) | −0.02 | 0.77 | −3.07 | 1.89 | 0.72 | 1 | ||||||
| Police effectiveness (3) | 0.00 | 0.75 | −3.07 | 1.92 | 0.62 | 0.84 | 1 | |||||
| Disorder (4) | 0.04 | 0.60 | −1.42 | 3.05 | −0.63 | −0.44 | −0.53 | 1 | ||||
| Collective efficacy (5) | −0.02 | 0.61 | −2.83 | 1.56 | 0.73 | 0.59 | 0.62 | −0.65 | 1 | |||
| Worry about crime (6) | 0.04 | 0.68 | −1.32 | 2.75 | −0.38 | −0.26 | −0.36 | 0.66 | −0.40 | 1 | ||
| Feeling unsafe (7) | 0.03 | 0.37 | −0.64 | 1.60 | −0.27 | −0.23 | −0.28 | 0.73 | −0.33 | 0.62 | 1 | |
| IMD (8) | 19.79 | 9.47 | 3.00 | 32.63 | 0.03 | 0.03 | 0.06 | −0.15 | 0.20 | −0.04 | −0.07 | 1 |
Correlation matrix of latent variables
| Mean | SD | Min | Max | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Sense of place (1) | 0.00 | 0.64 | −2.88 | 1.70 | 1 | |||||||
| Police fairness (2) | −0.02 | 0.77 | −3.07 | 1.89 | 0.72 | 1 | ||||||
| Police effectiveness (3) | 0.00 | 0.75 | −3.07 | 1.92 | 0.62 | 0.84 | 1 | |||||
| Disorder (4) | 0.04 | 0.60 | −1.42 | 3.05 | −0.63 | −0.44 | −0.53 | 1 | ||||
| Collective efficacy (5) | −0.02 | 0.61 | −2.83 | 1.56 | 0.73 | 0.59 | 0.62 | −0.65 | 1 | |||
| Worry about crime (6) | 0.04 | 0.68 | −1.32 | 2.75 | −0.38 | −0.26 | −0.36 | 0.66 | −0.40 | 1 | ||
| Feeling unsafe (7) | 0.03 | 0.37 | −0.64 | 1.60 | −0.27 | −0.23 | −0.28 | 0.73 | −0.33 | 0.62 | 1 | |
| IMD (8) | 19.79 | 9.47 | 3.00 | 32.63 | 0.03 | 0.03 | 0.06 | −0.15 | 0.20 | −0.04 | −0.07 | 1 |
| Mean | SD | Min | Max | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Sense of place (1) | 0.00 | 0.64 | −2.88 | 1.70 | 1 | |||||||
| Police fairness (2) | −0.02 | 0.77 | −3.07 | 1.89 | 0.72 | 1 | ||||||
| Police effectiveness (3) | 0.00 | 0.75 | −3.07 | 1.92 | 0.62 | 0.84 | 1 | |||||
| Disorder (4) | 0.04 | 0.60 | −1.42 | 3.05 | −0.63 | −0.44 | −0.53 | 1 | ||||
| Collective efficacy (5) | −0.02 | 0.61 | −2.83 | 1.56 | 0.73 | 0.59 | 0.62 | −0.65 | 1 | |||
| Worry about crime (6) | 0.04 | 0.68 | −1.32 | 2.75 | −0.38 | −0.26 | −0.36 | 0.66 | −0.40 | 1 | ||
| Feeling unsafe (7) | 0.03 | 0.37 | −0.64 | 1.60 | −0.27 | −0.23 | −0.28 | 0.73 | −0.33 | 0.62 | 1 | |
| IMD (8) | 19.79 | 9.47 | 3.00 | 32.63 | 0.03 | 0.03 | 0.06 | −0.15 | 0.20 | −0.04 | −0.07 | 1 |
RESULTS
To address our research questions, we use Structural Equation Modelling (SEM) in Mplus 7.2. SEM allows us to estimate all hypothesized relationships simultaneously. In this model there are again no cross-loadings, observed indicators are set to ordinal, and FIML estimation is used. The measures of trust in police are regressed on collective efficacy and ASB, and worry about crime and feelings of safety are regressed on trust and neighbourhood concerns. Sense of place is regressed on all the other latent constructs shown. Finally, all latent constructs were regressed on IMD.
Figure 2 shows results from this analysis. For visual ease, only significant paths (p < 0.1) are shown, and IMD and associated regression paths are omitted. Starting on the left-hand of the model, we find, first, that trust in the police is strongly predicted by neighbourhood concerns, with collective efficacy appearing the most important factor. Neighbourhood concerns are also very strongly associated with worry about crime and feeling unsafe, although here perceptions of disorder are dominant (conditional on this, the associations between collective efficacy and these indicators are small and non-significant). In contrast, there are no significant conditional correlations between either measure of trust in the police and the measures of unsafety and worry.
Results from an SEM with sense of place as the ultimate response variable.
Arriving at the right-hand side of the model, we find that sense of place is strongly predicted by the other variables, which jointly explain 64 per cent of the variance in this indicator. Specifically, we find strong, positive, conditional correlations between sense of place and trust in police fairness, collective efficacy and perceptions of disorder. Note that the trust measure has the largest regression weight. In contrast, there is no significant partial correlation between sense of place and either trust in police effectiveness or worry about crime. Finally, there is a somewhat surprising positive conditional correlation between feeling unsafe and sense of place.
While there were only very weak bivariate correlations between area-level deprivation and most of the other variables, in the multivariate analysis these correlations tended to strengthen. As might be expected, when deprivation was lower, perceptions of disorder were also on average lower (β = −0.20, p < 0.001), while perceptions of collective efficacy tended to be more positive (β = 0.22, p < 0.001). Yet, the sign of other correlations was the reverse of what might be expected: Trust in both police fairness (β = −0.13, p = 0.01) and effectiveness (β = −0.11, p = 0.04) tended to be somewhat higher when deprivation was higher, and sense of place tended to decrease as the level of deprivation fell (β = −0.14, p = 0.003).
Recall that the measures of trust in police fairness and effectiveness correlated strongly. As a robustness check we re-estimated the model shown in Figure 2, but this time using a combined ‘trustworthiness’ construct measured by perceptions of police fairness and effectiveness. Results from this model are summarized in Figure 3. Note that model fit is only marginally worse than in Figure 2, reinforcing the basic inter-changeability of the two models. Results correspond closely to those of the previous model: trust is predicted by perceptions of disorder and particularly collective efficacy; disorder, but not collective efficacy, predicts worry about crime and feelings of safety; there is no association between police trustworthiness and worry or safety; and sense of place is predicted by neighbourhood concerns and police trustworthiness.
Results from a second SEM with sense of place as the ultimate response variable.
DISCUSSION
Our research questions queried, first, whether there is an association between trust in the police and a ‘shallow’ sense of security, which we operationalized as worry about crime and feelings of (un)safety. The answer here seems definitively negative—in Macclesfield, at least, there is very little to suggest that believing the police to be more effective (or fairer) is associated, on average, with feeling less worried about crime and/or safer.
Second, we asked whether there is an association between trust in police and people’s sense of place, which we argue is indicative of a secure sense of belonging. Here, the answer is positive—we find evidence of a strong association between trust in police fairness, specifically, and sense of place. If we take two individuals living in areas with the same level of deprivation, who perceive similar levels of disorder and collective efficacy in their physical and social environment (and who have similar levels of concern about crime and personal safety), the person with more trust in police fairness would be much more likely to feel they belong in, and to, the town. In contrast, trust in police effectiveness seems to have little independent association with sense of place. To reiterate, two-thirds of the variance in sense of place was explained by the other variables in the model. While other factors must also important, we seem to be capturing much that is relevant to respondents’ sense of place.
These results suggest that, on the one hand, the demonstration of effectiveness on the part of the police, and indeed other aspects of their performance, may have little impact on public concerns about crime and perceptions of risk. While it remains moot whether people even notice, for example, marginal changes in the ability and success of police in reducing or managing crime, even if they do this might have little effect on the extent to which they feel safe and secure. On the other hand, while it has been argued that people may not notice changes in the fairness of police either (Nagin and Telep 2020), at the very least it seems that people who perceive the police to be fairer also tend to experience a stronger sense of place. This resonates with a growing body of research within the procedural justice paradigm that has stressed the association between the fairness of police activity and the extent to which those experiencing it feel they belong to and are included in wider social categories.
The precise nature of the association between trust in police fairness and sense of place described above remains rather unclear, however. One might assume that perceptions of police fairness are founded most importantly in personal interactions with officers, such that experiences of policing feed directly in people’s sense of place. This may well be the case. But it might also be that our measure of trust in police fairness wraps up wider perceptions of and concerns about the institutional contexts that shape people’s lives. Here, the argument would be closer to that made by Giddens (1991) and others: an understanding that police act fairly is linked to and informs a wider sense that social relations are appropriately ordered, predictable and supportive of life. Seeing the police as unfair, conversely, may have the opposite effect, since this signifies an inappropriate relationship between the individual and an important state institution, indicating uncertainty, a lack of support and a sense of abandonment. Here, it is possible that trust in the police is acting as a partial proxy for a much wider set of social and institutional concerns.
In congruence with other UK-based studies, our analysis also underlines the centrality of perceptions of low-level disorder and collective efficacy in people’s understandings of the places they live. Perceptions of disorder were strongly associated with worry about crime, feeling unsafe and sense of place, while perceptions of collective efficacy were strongly associated with trust in police and sense of place. The immediate quality of people’s social and physical environments seems to inform their views of ‘belonging’, safety, crime and an institution associated with all three phenomena.
Finally, the ‘objective’ characteristics of local areas also mattered, linking the attitudes and beliefs reported in the survey to more concrete social and economic processes. Respondents who lived in more deprived areas tended to report higher levels of disorder and lower levels of collective efficacy. Yet, we also find relatively small but positive conditional correlations between IMD, trust in police and sense of place. Further work would be needed to unpick what is going on here. But to hypothesize, it may be that, all else equal, people in more deprived areas are more dependent on and have a closer relationship with police, prompting an orientation towards the institution similar to that proposed by system justification theory—that those more dependent on a system are inclined to support it, even if it is failing them in some objective sense (Jost et al. 2004). In contrast, those who live in less deprived areas may be relatively more distant from police and thus, perhaps, more willing to be critical. People living in the more deprived parts of town may also be less mobile and have denser social networks, which could explain why sense of place tends to be higher among them.
Limitations
This study has a number of limitations. Most obviously, we have only cross-sectional data, and cannot address questions of causality. The model summarized in Figure 1 could be specified differently (e.g. sense of place predicting worry about crime) and still fit the data well. Four further points are, first, that the measure of (un)safety is limited to local concerns. It is possible that, for example, perceptions of police effectiveness contribute to feeling safe from other types of threat.
Second, we have no measures of attitudes towards other actors and institutions that may be important in shaping sense of place, such as local government or healthcare. We also have little contextual data on why respondents trusted the police (or indeed experienced disorder or low levels of collective efficacy). While we can surmise from previous studies that personal and vicarious encounters with police are vital for trust, what else about local and other forms of policing might be important? Future studies could profitably explore all these issues and produce a more expansive model of the local context of subjective security.
Third, our analysis is by design limited to only one place. While Macclesfield is not ‘every’ or ‘any’ town, it is arguably representative of a broad swathe of ‘middle England’ medium-sized towns. We might expect similar results in similar sorts of places. But it is certainly not representative of either larger cities or more rural locales. Different relationships between policing and perceptions of safety may pertain elsewhere.
Finally, the fact that this research took place in 2021 during the Covid-19 pandemic must also be pertinent. Methodologically, ‘Covid compliant’ doorstep surveying, even with advance letters, may have introduced a response bias towards, most obviously, more trusting individuals; and a response rate of 20 per cent must be considered low. Substantively, this was a period of significant debate around policing, due to Covid-19, and events such as the Black Lives Matter protests in 2020 and the murder of Sarah Everard by a serving police officer in 2021. While this debate has in fact continued, intensified and arguably consolidated in public discourse (Casey 2023), there may have been something about the wider ‘moment’ within which our data were collected that influenced respondents in arguably unusual ways. Replication and development of the model proposed here using data collected in different times and places would be most welcome.
CONCLUSION
To draw out the broadly normative, or at least policy/political, lessons of our findings we return to the distinction drawn by Loader (2006) between ‘shallow and wide’ and ‘deep and narrow’ senses of security. The former stood for the widespread idea that policing can and should make people feel safer by being present, visible and busy at the surface of social relations—by, for example, leading the ‘fight against crime’. The latter notion was intended to suggest that the police can be an answer to deeper questions about security—‘where do I belong?’ and ‘who cares about me?’. We find that police efforts to make people feeler safer by being, and being seen to be, more effective seem unlikely to achieve very much overall, at least in a place like Macclesfield. By contrast, when police are seen as a fair, attentive and engaged presence in local communities, this does seem to be linked to a sense of secure belonging among residents.
A number of inter-related implications may flow from this finding. One of the innovations of procedural justice theory as originally conceived (e.g. Tyler 2006) was its emphasis on the idea that people view police behaviour as something akin to the personal embodiment of state authority. Whether agents of the state—the police in particular—exercise power fairly and accountably is consequential for people’s assessments of the legitimacy of its rule. Subsequent studies have indicated a wider connection between encounters with the police and judgements about community, stability and order. Our observations here, arising from a local survey, suggest that if we consider the notion of security as having regard to the feeling of living comfortably somewhere, then the extent to which people see that place as being policed equitably and considerately is likely to be a salient dimension of what it is to live there in such a way. This is not to place some unfeasible onus on the police to buttress, let alone supply, people’s needs for ontological security. Rather, it is to note that the police represent and stand for a wider set of state behaviours that come to be coded as just, appropriate and contextually relevant, and this happens, in some part at least, precisely because perceptions of policing link what people think happens locally, and their experiences of what it means to live where they do, with wider sets of feelings and experiences concerning the state, society and community writ large.
To build on the potential contribution of policing to ontological security, it would seem to us that this is not, in fact, a desirable, direct, aim for policy and practice. Explicit efforts to enhance a sense of secure belonging among the policed would seem to risk all the downsides of increased police presence and activity noted above. Rather, just as Molotch (2014) argues that the best way to produce security in public spaces is often not by directly pursuing it, but just by making them better, it may be that the best way for police to contribute to security is not by ‘chasing it’ but rather to see it as an indirect effect of doing other things well, or at least better. Loader (2006) argued that the police can best contribute to security in a deeper sense by sticking to a narrow remit; doing the basics of responsive policing well, in ways that are fair and rights-regarding, to which we can add concepts such as voice from the procedural justice literature, as well the notion of ‘community engagement’ and a willingness to listen to and work with people as members of groups (geographically bounded or otherwise) as well as individuals.
Seen in this light, for police to contribute to what we have conceptualized as a deeper and more constitutive sense of security what is needed is an emphasis on both institutional caring and symbolic power. Police are part of the infrastructure of care in local communities—which has clearly frayed in recent years—and they can contribute to that care by the behaviour of individual officers as they go about their business, by wider engagement with local people, and by providing a postive representation of the wider state and its efforts on behalf of its citizens. Naturally, austerity and increasing demand, which can often seem to undermine the ability of officer and organization to ‘show they care’, and long-standing issues in the relationship between police and some communities, mean that the generalizability and continued relevance of these ideas remains an open question.
Finally, if police are part of the infrastructure of care in local neighbourhoods, which is in turn linked to the subjective security of residents, then it would seem that the current article, and indeed similar earlier efforts, tell only a very partial picture. If people’s sense of security, ‘shallow’, ‘deep’ or otherwise, is an appropriate object of public policy, and if local conditions and actors are so important in shaping it, then we need to know much more about what other institutions, conditions and processes work alongside and with disorder, collective efficacy and trust in policing to make people feel more or less safe and secure in the places they live.
FUNDING
This article is part of the Economic and Social Research Council funded project ‘Place, crime and insecurity in everyday life: A contemporary study of an English town’ (ES/S010734/1). See further: https://securityinplace.org/. The data used in this paper will be made available at the UK Data Service (https://ukdataservice.ac.uk) after a period of embargo.
Appendix
Dependent variables: constructs and measures
| Std. factor loading | Item R-square | |
|---|---|---|
| Sense of place | ||
| To what extent do you agree or disagree with the following statements…? | ||
| I feel happiest when I am here | 0.68 | 0.46 |
| This place is my favourite place to be | 0.74 | 0.54 |
| This place makes me feel as if I can be myself | 0.80 | 0.64 |
| This place and its surroundings are good just the way they are | 0.70 | 0.49 |
| There aren’t any features of this place that annoy me | 0.65 | 0.43 |
| I feel as if I’m able to move freely in this place | 0.68 | 0.46 |
| My roots are here | 0.46 | 0.21 |
| Worry about crime | ||
| Please tell me how worried you are about falling victim to the following types of crime…? | ||
| Burglary | 0.78 | 0.61 |
| Mugging/robbery | 0.91 | 0.82 |
| A violent crime | 0.93 | 0.96 |
| A terrorist attack | 0.76 | 0.58 |
| Fraud (computer related fraud/other scams) | 0.49 | 0.24 |
| Being a victim of online crime | 0.53 | 0.28 |
| Feeling unsafe | ||
| When you are using public spaces in Macclesfield, do any of the following things ever make you feel unsafe? | ||
| Cyclists in pedestrian areas | 0.48 | 0.23 |
| People begging | 0.74 | 0.55 |
| A lack of police presence | 0.78 | 0.61 |
| Drug users/visible signs of drug use | 0.77 | 0.59 |
| The way some people drive or park their cars | 0.61 | 0.38 |
| People drinking alcohol in the streets | 0.82 | 0.67 |
| Young people hanging around | 0.69 | 0.47 |
| A lack of street lighting | 0.57 | 0.33 |
| Fit statistics | ||
| Chi-square | 384.40 | |
| Degrees of freedom | 181 | |
| p-Value | <0.0005 | |
| RMSEA | 0.05 | |
| CFI | 0.97 | |
| TLI | 0.97 | |
| SRMR | 0.06 | |
| Std. factor loading | Item R-square | |
|---|---|---|
| Sense of place | ||
| To what extent do you agree or disagree with the following statements…? | ||
| I feel happiest when I am here | 0.68 | 0.46 |
| This place is my favourite place to be | 0.74 | 0.54 |
| This place makes me feel as if I can be myself | 0.80 | 0.64 |
| This place and its surroundings are good just the way they are | 0.70 | 0.49 |
| There aren’t any features of this place that annoy me | 0.65 | 0.43 |
| I feel as if I’m able to move freely in this place | 0.68 | 0.46 |
| My roots are here | 0.46 | 0.21 |
| Worry about crime | ||
| Please tell me how worried you are about falling victim to the following types of crime…? | ||
| Burglary | 0.78 | 0.61 |
| Mugging/robbery | 0.91 | 0.82 |
| A violent crime | 0.93 | 0.96 |
| A terrorist attack | 0.76 | 0.58 |
| Fraud (computer related fraud/other scams) | 0.49 | 0.24 |
| Being a victim of online crime | 0.53 | 0.28 |
| Feeling unsafe | ||
| When you are using public spaces in Macclesfield, do any of the following things ever make you feel unsafe? | ||
| Cyclists in pedestrian areas | 0.48 | 0.23 |
| People begging | 0.74 | 0.55 |
| A lack of police presence | 0.78 | 0.61 |
| Drug users/visible signs of drug use | 0.77 | 0.59 |
| The way some people drive or park their cars | 0.61 | 0.38 |
| People drinking alcohol in the streets | 0.82 | 0.67 |
| Young people hanging around | 0.69 | 0.47 |
| A lack of street lighting | 0.57 | 0.33 |
| Fit statistics | ||
| Chi-square | 384.40 | |
| Degrees of freedom | 181 | |
| p-Value | <0.0005 | |
| RMSEA | 0.05 | |
| CFI | 0.97 | |
| TLI | 0.97 | |
| SRMR | 0.06 | |
Dependent variables: constructs and measures
| Std. factor loading | Item R-square | |
|---|---|---|
| Sense of place | ||
| To what extent do you agree or disagree with the following statements…? | ||
| I feel happiest when I am here | 0.68 | 0.46 |
| This place is my favourite place to be | 0.74 | 0.54 |
| This place makes me feel as if I can be myself | 0.80 | 0.64 |
| This place and its surroundings are good just the way they are | 0.70 | 0.49 |
| There aren’t any features of this place that annoy me | 0.65 | 0.43 |
| I feel as if I’m able to move freely in this place | 0.68 | 0.46 |
| My roots are here | 0.46 | 0.21 |
| Worry about crime | ||
| Please tell me how worried you are about falling victim to the following types of crime…? | ||
| Burglary | 0.78 | 0.61 |
| Mugging/robbery | 0.91 | 0.82 |
| A violent crime | 0.93 | 0.96 |
| A terrorist attack | 0.76 | 0.58 |
| Fraud (computer related fraud/other scams) | 0.49 | 0.24 |
| Being a victim of online crime | 0.53 | 0.28 |
| Feeling unsafe | ||
| When you are using public spaces in Macclesfield, do any of the following things ever make you feel unsafe? | ||
| Cyclists in pedestrian areas | 0.48 | 0.23 |
| People begging | 0.74 | 0.55 |
| A lack of police presence | 0.78 | 0.61 |
| Drug users/visible signs of drug use | 0.77 | 0.59 |
| The way some people drive or park their cars | 0.61 | 0.38 |
| People drinking alcohol in the streets | 0.82 | 0.67 |
| Young people hanging around | 0.69 | 0.47 |
| A lack of street lighting | 0.57 | 0.33 |
| Fit statistics | ||
| Chi-square | 384.40 | |
| Degrees of freedom | 181 | |
| p-Value | <0.0005 | |
| RMSEA | 0.05 | |
| CFI | 0.97 | |
| TLI | 0.97 | |
| SRMR | 0.06 | |
| Std. factor loading | Item R-square | |
|---|---|---|
| Sense of place | ||
| To what extent do you agree or disagree with the following statements…? | ||
| I feel happiest when I am here | 0.68 | 0.46 |
| This place is my favourite place to be | 0.74 | 0.54 |
| This place makes me feel as if I can be myself | 0.80 | 0.64 |
| This place and its surroundings are good just the way they are | 0.70 | 0.49 |
| There aren’t any features of this place that annoy me | 0.65 | 0.43 |
| I feel as if I’m able to move freely in this place | 0.68 | 0.46 |
| My roots are here | 0.46 | 0.21 |
| Worry about crime | ||
| Please tell me how worried you are about falling victim to the following types of crime…? | ||
| Burglary | 0.78 | 0.61 |
| Mugging/robbery | 0.91 | 0.82 |
| A violent crime | 0.93 | 0.96 |
| A terrorist attack | 0.76 | 0.58 |
| Fraud (computer related fraud/other scams) | 0.49 | 0.24 |
| Being a victim of online crime | 0.53 | 0.28 |
| Feeling unsafe | ||
| When you are using public spaces in Macclesfield, do any of the following things ever make you feel unsafe? | ||
| Cyclists in pedestrian areas | 0.48 | 0.23 |
| People begging | 0.74 | 0.55 |
| A lack of police presence | 0.78 | 0.61 |
| Drug users/visible signs of drug use | 0.77 | 0.59 |
| The way some people drive or park their cars | 0.61 | 0.38 |
| People drinking alcohol in the streets | 0.82 | 0.67 |
| Young people hanging around | 0.69 | 0.47 |
| A lack of street lighting | 0.57 | 0.33 |
| Fit statistics | ||
| Chi-square | 384.40 | |
| Degrees of freedom | 181 | |
| p-Value | <0.0005 | |
| RMSEA | 0.05 | |
| CFI | 0.97 | |
| TLI | 0.97 | |
| SRMR | 0.06 | |
Independent variables: constructs and measures
| Std. factor loading | Item R-square | |
|---|---|---|
| Perceptions of disorder | ||
| For the following things I read out, can you tell me how much of a problem they are in your area? | ||
| Noisy neighbours or loud parties? | 0.70 | 0.49 |
| Teenagers hanging around on the streets? | 0.78 | 0.61 |
| Rubbish or litter lying around? | 0.70 | 0.49 |
| Vandalism, graffiti and other deliberate damage to property or vehicles? | 0.84 | 0.71 |
| People being drunk or rowdy in public places? | 0.82 | 0.68 |
| Badly parked cars? | 0.40 | 0.16 |
| Homeless people living on the streets | 0.54 | 0.29 |
| Collective efficacy | ||
| To what extent do you agree or disagree that…? | ||
| You can see from the public spaces here in the area that people take pride in their environment | 0.64 | 0.41 |
| People in this neighbourhood can be trusted | 0.89 | 0.79 |
| People act with courtesy to each other in public spaces in this area | 0.84 | 0.70 |
| If I sensed trouble whilst in this area, I could get help from people who live here | 0.79 | 0.63 |
| The people who live here can be relied upon to call the police if someone is acting suspiciously | 0.76 | 0.57 |
| If any of the children or young people around here are causing trouble, local people will tell them off | 0.51 | 0.26 |
| Trust in police fairness | ||
| To what extent do you agree with these statements about the police in your area? | ||
| They can be relied on to be there when you need them | 0.84 | 0.70 |
| They would treat you with respect if you had contact with them for any reason. | 0.77 | 0.60 |
| The police in your area treat everyone fairly regardless of who they are | 0.79 | 0.62 |
| They are dealing with the things that matter to people in this community | 0.92 | 0.85 |
| The police in your area listen to the concerns of local people | 0.92 | 0.84 |
| Trust in police effectiveness | ||
| How effective do you think police in your area are at…? | ||
| Responding to emergencies quickly | 0.86 | 0.73 |
| Dealing with property crimes such as burglary | 0.79 | 0.62 |
| Dealing with violent crimes | 0.91 | 0.83 |
| Preventing crime | 0.84 | 0.70 |
| Maintaining order in public spaces | 0.82 | 0.67 |
| Fit statistics | ||
| Chi-square | 583.32 | |
| Degrees of freedom | 225 | |
| p-Value | <0.0005 | |
| RMSEA | 0.06 | |
| CFI | 0.97 | |
| TLI | 0.96 | |
| SRMR | 0.06 | |
| Std. factor loading | Item R-square | |
|---|---|---|
| Perceptions of disorder | ||
| For the following things I read out, can you tell me how much of a problem they are in your area? | ||
| Noisy neighbours or loud parties? | 0.70 | 0.49 |
| Teenagers hanging around on the streets? | 0.78 | 0.61 |
| Rubbish or litter lying around? | 0.70 | 0.49 |
| Vandalism, graffiti and other deliberate damage to property or vehicles? | 0.84 | 0.71 |
| People being drunk or rowdy in public places? | 0.82 | 0.68 |
| Badly parked cars? | 0.40 | 0.16 |
| Homeless people living on the streets | 0.54 | 0.29 |
| Collective efficacy | ||
| To what extent do you agree or disagree that…? | ||
| You can see from the public spaces here in the area that people take pride in their environment | 0.64 | 0.41 |
| People in this neighbourhood can be trusted | 0.89 | 0.79 |
| People act with courtesy to each other in public spaces in this area | 0.84 | 0.70 |
| If I sensed trouble whilst in this area, I could get help from people who live here | 0.79 | 0.63 |
| The people who live here can be relied upon to call the police if someone is acting suspiciously | 0.76 | 0.57 |
| If any of the children or young people around here are causing trouble, local people will tell them off | 0.51 | 0.26 |
| Trust in police fairness | ||
| To what extent do you agree with these statements about the police in your area? | ||
| They can be relied on to be there when you need them | 0.84 | 0.70 |
| They would treat you with respect if you had contact with them for any reason. | 0.77 | 0.60 |
| The police in your area treat everyone fairly regardless of who they are | 0.79 | 0.62 |
| They are dealing with the things that matter to people in this community | 0.92 | 0.85 |
| The police in your area listen to the concerns of local people | 0.92 | 0.84 |
| Trust in police effectiveness | ||
| How effective do you think police in your area are at…? | ||
| Responding to emergencies quickly | 0.86 | 0.73 |
| Dealing with property crimes such as burglary | 0.79 | 0.62 |
| Dealing with violent crimes | 0.91 | 0.83 |
| Preventing crime | 0.84 | 0.70 |
| Maintaining order in public spaces | 0.82 | 0.67 |
| Fit statistics | ||
| Chi-square | 583.32 | |
| Degrees of freedom | 225 | |
| p-Value | <0.0005 | |
| RMSEA | 0.06 | |
| CFI | 0.97 | |
| TLI | 0.96 | |
| SRMR | 0.06 | |
Independent variables: constructs and measures
| Std. factor loading | Item R-square | |
|---|---|---|
| Perceptions of disorder | ||
| For the following things I read out, can you tell me how much of a problem they are in your area? | ||
| Noisy neighbours or loud parties? | 0.70 | 0.49 |
| Teenagers hanging around on the streets? | 0.78 | 0.61 |
| Rubbish or litter lying around? | 0.70 | 0.49 |
| Vandalism, graffiti and other deliberate damage to property or vehicles? | 0.84 | 0.71 |
| People being drunk or rowdy in public places? | 0.82 | 0.68 |
| Badly parked cars? | 0.40 | 0.16 |
| Homeless people living on the streets | 0.54 | 0.29 |
| Collective efficacy | ||
| To what extent do you agree or disagree that…? | ||
| You can see from the public spaces here in the area that people take pride in their environment | 0.64 | 0.41 |
| People in this neighbourhood can be trusted | 0.89 | 0.79 |
| People act with courtesy to each other in public spaces in this area | 0.84 | 0.70 |
| If I sensed trouble whilst in this area, I could get help from people who live here | 0.79 | 0.63 |
| The people who live here can be relied upon to call the police if someone is acting suspiciously | 0.76 | 0.57 |
| If any of the children or young people around here are causing trouble, local people will tell them off | 0.51 | 0.26 |
| Trust in police fairness | ||
| To what extent do you agree with these statements about the police in your area? | ||
| They can be relied on to be there when you need them | 0.84 | 0.70 |
| They would treat you with respect if you had contact with them for any reason. | 0.77 | 0.60 |
| The police in your area treat everyone fairly regardless of who they are | 0.79 | 0.62 |
| They are dealing with the things that matter to people in this community | 0.92 | 0.85 |
| The police in your area listen to the concerns of local people | 0.92 | 0.84 |
| Trust in police effectiveness | ||
| How effective do you think police in your area are at…? | ||
| Responding to emergencies quickly | 0.86 | 0.73 |
| Dealing with property crimes such as burglary | 0.79 | 0.62 |
| Dealing with violent crimes | 0.91 | 0.83 |
| Preventing crime | 0.84 | 0.70 |
| Maintaining order in public spaces | 0.82 | 0.67 |
| Fit statistics | ||
| Chi-square | 583.32 | |
| Degrees of freedom | 225 | |
| p-Value | <0.0005 | |
| RMSEA | 0.06 | |
| CFI | 0.97 | |
| TLI | 0.96 | |
| SRMR | 0.06 | |
| Std. factor loading | Item R-square | |
|---|---|---|
| Perceptions of disorder | ||
| For the following things I read out, can you tell me how much of a problem they are in your area? | ||
| Noisy neighbours or loud parties? | 0.70 | 0.49 |
| Teenagers hanging around on the streets? | 0.78 | 0.61 |
| Rubbish or litter lying around? | 0.70 | 0.49 |
| Vandalism, graffiti and other deliberate damage to property or vehicles? | 0.84 | 0.71 |
| People being drunk or rowdy in public places? | 0.82 | 0.68 |
| Badly parked cars? | 0.40 | 0.16 |
| Homeless people living on the streets | 0.54 | 0.29 |
| Collective efficacy | ||
| To what extent do you agree or disagree that…? | ||
| You can see from the public spaces here in the area that people take pride in their environment | 0.64 | 0.41 |
| People in this neighbourhood can be trusted | 0.89 | 0.79 |
| People act with courtesy to each other in public spaces in this area | 0.84 | 0.70 |
| If I sensed trouble whilst in this area, I could get help from people who live here | 0.79 | 0.63 |
| The people who live here can be relied upon to call the police if someone is acting suspiciously | 0.76 | 0.57 |
| If any of the children or young people around here are causing trouble, local people will tell them off | 0.51 | 0.26 |
| Trust in police fairness | ||
| To what extent do you agree with these statements about the police in your area? | ||
| They can be relied on to be there when you need them | 0.84 | 0.70 |
| They would treat you with respect if you had contact with them for any reason. | 0.77 | 0.60 |
| The police in your area treat everyone fairly regardless of who they are | 0.79 | 0.62 |
| They are dealing with the things that matter to people in this community | 0.92 | 0.85 |
| The police in your area listen to the concerns of local people | 0.92 | 0.84 |
| Trust in police effectiveness | ||
| How effective do you think police in your area are at…? | ||
| Responding to emergencies quickly | 0.86 | 0.73 |
| Dealing with property crimes such as burglary | 0.79 | 0.62 |
| Dealing with violent crimes | 0.91 | 0.83 |
| Preventing crime | 0.84 | 0.70 |
| Maintaining order in public spaces | 0.82 | 0.67 |
| Fit statistics | ||
| Chi-square | 583.32 | |
| Degrees of freedom | 225 | |
| p-Value | <0.0005 | |
| RMSEA | 0.06 | |
| CFI | 0.97 | |
| TLI | 0.96 | |
| SRMR | 0.06 | |
REFERENCES
Footnotes
This suggests the possibility of a multi-level modelling approach. However, the small number of LSOAs (i.e. level 2 units) militates against this.
