Copyright Sociological Research Online, 2002


Martin Innes and Nigel Fielding (2002) 'From Community To Communicative Policing: 'Signal Crimes' And The Problem Of Public Reassurance '
Sociological Research Online, vol. 7, no. 2, <>

To cite articles published in Sociological Research Online, please reference the above information and include paragraph numbers if necessary

Received: 26/6/2002      Accepted: 19/8/2002      Published: 31/8/2002


In this paper the concept of 'signal crimes' is proposed to capture the social semiotic processes by which particular types of criminal and disorderly conduct have a disproportionate impact upon fear of crime. Drawing upon the wider social scientific literature on risk perception, a sense of how and why different crime types might be possessed of different signal values is provided and some of the implications for current police practice outlined.

Disorder; Fear Of Crime; Policing; Risk; Social Semiotics


Police officers and police forces throughout Britain are currently grappling with a new problem. After successive decades where recorded crime had been rising, since about 1995 overall levels of recorded crime have been falling.[1] The problem is that the fall in crime has not been matched by a corresponding decline in reported fear of crime. For example, between 1999-2000 the number of offences reported to the British Crime Survey fell by 12%, and whilst there were small reductions in reported levels of fear among some groups in the sample, over 50% of the respondents believed that crime had risen over this period (Kershaw et al., 2001). Although collectively the objective risk of being a victim of crime has decreased, people do not necessarily believe this to be the case. In their lived routines and everyday experiences many individuals and groups have not been reassured that things are getting safer. In this sense, both crime and fear of crime continue to be significant social problems.

In this conceptually oriented paper we have two objectives. Firstly, we seek to provide a diagnosis of the problem outlined above. After reviewing some of the keynote studies that explore the relations between crime and disorder, our argument focuses upon developing the concept of 'signal crimes'. This concept is proposed in order to capture the ways in which different types of crime are important not just in terms of the harm done to the victim, but also in terms of what they signify and communicate to a wider audience. It is argued that certain crimes are perceived as warning signals by those who experience them either directly, or through media reportage. Knowledge of these acts may cause a social reaction on the part of the audience, motivated by a desire to manufacture a greater sense of individual or collective security. This logic suggests that fear of crime needs to be understood as a complex, adaptive, social response to a process of signification, as well as to the actual nature of any criminal or disorderly act performed. Developing such a conceptualisation serves to articulate both how and why certain crimes may be disproportionately influential in terms of de-stabilizing a sense of social order, generating fear and anxiety by undermining 'organic' mechanisms of community based social control. Having developed our diagnosis, we then move on to consider the implications that this understanding has for police practice and how policing strategies need to develop in light of the contemporary situation. In so doing, we comment on the social psychology of fear of crime and argue that a social semiotic dimension to the analysis of crime-induced fear can make an important and valuable contribution to our understandings of such matters.[2]


The concept of fear of crime was established as a key component in the criminological lexicon with the development of mass-victimization surveys in the late 1970s and early 1980s. These surveys consistently demonstrated that citizens were extremely worried about the risks to their safety posed by criminal events and were actively engaging in risk avoidance measures in an attempt to reduce their exposure to perceived threats and hazards. But perhaps the key finding of these surveys was that this 'fear of crime' was paradoxical. In that those groups whom the data showed to be least at risk of victimisation tended to be the most worried, whilst those who were actually most at risk, tended, if anything, to under-estimate the threats to their safety (Hale, 1996). Although, something of a policy problem all the time that crime remained high, there was at least an obvious and logical explanation for why many groups of people expressed high fear of crime. Indeed, as Crawford (1997) has discussed, from a Neo-Liberal political perspective, fear, in as far as it encouraged individuals and communities to take protective measures to try and ensure their safety, had positive social effects. By routinely enacting 'the criminologies of everyday life', people were encouraged to become 'responsibilised' members of a partnership coalition between state and citizens to reduce crime (Garland, 1996). However, when crime started to fall, the expectation was that this would soon be accompanied by commensurate reductions in fear. As yet though, this has not happened and, where formerly the public attitude appeared 'rational', it now seems that worry about crime bespeaks existential concerns that may be harder to address.

In response to this situation a number of recent studies have sought to better understand the cognitive and affective processes involved in fear of crime. The critique of fear of crime encompasses both methodological and conceptual concerns. The methodological critiques centre on the instruments typically used to measure respondents' fear of crime. For example, by deconstructing the meaning of some of the key terms within the standard survey questions used to measure fear of crime, Farrall and Ditton (1999) show that respondents may not necessarily be afraid, but are rather thinking about crime, or feeling anger, outrage or annoyance about it. What researchers have understood as expressions of fear, may then, actually be a more neutral form of awareness about crime, or be related to the petty grievances and annoyances that they have recently experienced (Ditton et al., 1999).

Developing alongside the methodological critiques there has been important questioning of the conceptual adequacy of the established definition of fear of crime. It is becoming increasingly apparent that people use fear of crime as something of a 'dustbin concept'. That is, when people talk about fear of crime, in addition to appraising their probability of victimization, they also tend to articulate nebulous and diffuse anxieties that they have about the state of society and their position within it (Girling, Loader, Sparks, 2000). Furthermore, it is argued by Lupton and Tulloch (1999), that fear of crime studies have for the most part failed to engage with the 'hermeneutic reflexivity' that is central to how fear, as a dynamic mode of perception grounded in everyday experience, is constructed. This is an approach with seeming affinities to the emerging field of cognitive sociology[3], which in turn builds upon the work of Mary Douglas (1992), seeking to uncover the ways in which particular cultural contexts result in the fabrication of shared cognitive frames. Overall then, important questions are being raised about the analytic precision of the concept of fear of crime.[4]

For police organizations this situation represents something of a double problematic. Firstly, it has been confirmed that fear of crime is a problem in its own right that needs to be addressed by the police in providing an effective service to communities. Secondly, it has been shown that whilst the concept of fear of crime can provide a useful metric for analysing the distribution of fear amongst different sections of the population, it is rather less successful at identifying how people construct their fears, what precisely they are fearful of and why they fear it.

Criminogenic Risks

Fear of crime is a product of specific social processes. In order to deconstruct these generative processes, we can usefully conceive of expressed fear of crime as the outcome of how individuals and groups identify, interpret and construct perceived risks and threats to their safety. In effect then, we are looking at the processes of criminogenic risk perception. Therefore, by drawing upon the wider social scientific research literature on risk perception we can gain some initial analytic purchase upon the causal mechanisms that link people's situated realities and their articulation of perceived threats, hazards and dangers.

In a similar fashion to what has occurred throughout the social sciences, the concept of risk has become increasingly influential in shaping some of the key contemporary debates in criminology (cf. Ericson and Haggerty, 1997; Feeley and Simon, 1994). As implied previously though, much of the established criminological effort in respect of fear of crime that has sought to integrate risk as an analytic resource, has focused upon mapping the distribution of the fear of crime. It has been less concerned to explain how and why certain people come to see particular things as risks to their security, than in taking crime simply as one of a number of similar threats to perceived security. The result being that relatively little effort has been invested in identifying how the particularities of crime act as a source of perceived risk that might affect the cognitive processes associated with the negotiation of threats to personal security. Premised upon the assumption that 'risk' and 'fear' are of generic cognitive form, and crime merely one of several phenomena invoking a generic threat- avoidance reaction, the agenda has been to examine patterned regularities in its incidence, rather than considering how risk/fear perception may distinctively work in the case of crime and thus account for the patterns that become manifest. Curiously, then, although informed by social psychology the approach adopted appears to have pursued a customarily sociological agenda, developing metrics for population-level incidence of fear-of-crime, rather than what one might take to be the core psychological ground, documenting the mental process in risk perception. Furthermore, the dominant construction of the cognition of risk is one based not on an appreciation of the semiotic but the calculative rationale of threat/response. Decision-making in negotiating risk is attributed to the use of probabilistic calculations by individuals assessing likely threat and best response, that is, by reference to the kind of rational-actor theory derived from economics rather than psychology.

In contrast to the epistemologies and ontologies of what Renn (1992) dubs 'purely technical risk analyses' that seek objective measures of exposure to hazards, Douglas urges that 'risk is always a social product.' It is an approach that stresses,

...that the reality of the dangers is not at issue...This argument is not about the reality of the dangers, but about how they are politicised. (Douglas, 1990: 8)
Risks are thus to be conceived of as embedded within and an enactment of a surrounding belief system.

Accordingly, perhaps the key finding of the cultural theory of risk literature is that risks are not defined and classified by people according to a rational and objective formula. This echoes the issue noted earlier in relation to the fear of crime paradox (Jefferson and Hollway, 2000). The perception of risks is inherently subjective and as a consequence shaped by semiotic characteristics of a threat, as well as the extant 'social matrix.'[5] In relation to issues such as health scares, bio-technologies and environmental hazards it has been documented that people tend to be more aware and fearful of dramatic, 'high visibility' and spectacular hazards, than those threats to their security that are comparatively invisible (Cutter, 1993; Palmlund, 1992; Hutter & Lloyd-Bostock, 1990). Moreover, the literature on risk perception strongly suggests that people have a tendency to simplify their understanding of risks and to interpret 'new' hazards in accordance with the understandings afforded to them by their existing cognitive frame (Douglas and Wildavsky, 1983; Cutter, 1993)

In an effort to capture some of these dynamics in the process of risk perception, Slovic (1992) proposed that different risks may have different 'signal values.' By this he means that hazards can have a greater or lesser potential to provide new information about the likelihood of similar or more destructive future mishaps. As he describes,

An accident that takes many lives may produce relatively little social disturbance (beyond that caused the victims' families and friends) if it occurs as part of a familiar and well understood system (e.g., a train wreck). However, a small accident in an unfamiliar system (or one perceived as poorly understood), such as a nuclear reactor or a recombinant DNA laboratory, may have immense social consequences if it is perceived as a harbinger of further (and possibly catastrophic) mishaps. (Slovic 1992: 124-5)

This sense that certain behaviours may be construed by a social audience as 'warning signs' that are generative of anxiety about future threats is obviously relevant to our understanding of the perception of criminogenic risks. The implications of this will be worked out more thoroughly in due course. For the moment though, we can note by drawing upon Slovic's (1992) conceptualisation, that public understanding of the seriousness of a risk is not defined solely by the characteristics of the event itself. Rather, it is the nature of the risk, it's semiotic properties, together with the context in which it occurs, that shapes how it is interpreted and understood. Thinking about criminogenic risks, this suggests that the most important incidents in shaping popular fear and anxiety may not necessarily be only those that are traditionally defined by juridical discourses as serious crimes. At a local community or neighbourhood level, what might appear to be relatively trivial occurrences from the point of view of criminal justice professionals, may have significant impacts upon the local community order.[6] As such, we need to examine the links between crime and disorder.

Crime and Disorder

Arguably the most influential contemporary theory on the connections between criminal and disorderly behaviour is Wilson and Kelling's (1982) 'broken windows theory'. The fundamental premise of which is that if physical and social disorders occur in a location and do not receive an adequate response, then this will cause further crime to occur. Despite being little more than a hypothesis devised by Wilson and Kelling (1982), the common-sense 'folk logic' underpinning this approach, has contributed to it becoming something of a taken-for-granted orthodoxy of many contemporary criminal justice policy analyses. A degree of empirical support for the logic expounded by Wilson and Kelling's hypothesis was subsequently provided by Skogan's (1990) study, in which he introduced the notion of 'decay spirals.' Based upon analysis of quantitative survey data, Skogan described how the toleration of, or failure to act against seemingly trivial acts of disorder, could lead to an increasing number of serious crimes being committed in that area. Thus instigating a process of 'community tipping' from low to high crime levels. Skogan found that levels of disorder tended to be highest in areas with low neighbourhood stability, high poverty and a high ethnic minority population.[7]

According to Skogan's explanation, criminal and disorderly behaviours are as important for their communicative properties as the actual harm or damage caused. Developing the logic implied by Wilson and Kelling's formulation, he suggests that the commission of a crime can lead law-abiding community members to perceive themselves to be at greater risk of victimization. As a result of which they withdraw from using public space, thus decreasing levels of informal social control in an area. Concomitantly, the commission of the crime and the lack of an effective response to it, can act as a signal to potential motivated delinquents and criminals that this is an area where there is a decreased chance of being apprehended or challenged if you engage in delinquent or criminal behaviour. As such, Skogan suggests that the above process could itself be a direct cause of more serious crime. More definitely, he identified that the development of such a situation could lead to selective out-migration by certain groups of residents, thus further attenuating the resources for maintaining informal social control by the community concerned. This latter aspect of his argument on out-migration as an engine of neighbourhood change coheres with the work of Schuerman and Kobrin (1986) and Taub et al. (1984). The latter study is particularly important in that it espouses a 'threshold model' of how residents and potential residents judge the quality of an area in terms of its residential desirability. According to this model the signs of early deterioration will cause some residents to move away, but some groups (i.e. the elderly) will not move whatever anyone else does, whilst others in between these groups will be influenced by the opinions and actions of others. Therefore, from these studies we get a sense of how and why neighbourhoods may 'tip' from being comparatively stable and ordered, into high crime areas.

To date though, the most sophisticated analysis of the links between crime and disorder has been provided by Sampson and Raudenbusch's (1999) Systematic Social Observation studies of neighbourhoods in Chicago. Whereas the broken windows hypothesis sees disorder as a fundamental cause of crime, Sampson and Raudenbusch argue that the presence of crimes and physical and social disorders, are all symptoms of a combination of comparative structural disadvantage together with a deficit in 'collective efficacy'. This latter term reflects the fact that Sampson and Raudenbusch recognise that structural disadvantage alone does not determine the fate of communities. Rather it is socio-economic deprivation in combination with a lack of social cohesion, mutual trust and informal social control on the part of the residents or users of an area that provides the most potent combination. This analysis thus keys into contemporary debates about the benefits that can be accrued by communities through mutual investment in the generation of social capital (Putnam, 2000).[8] Thus disorder is part and parcel of crime itself. Graffiti does not cause robbery, but a lack of informal social control is a cause of both.

Methodologically Sampson and Raudenbusch's (1999) study provides a more sophisticated treatment of the crime and disorder nexus. Nevertheless, it shares with the other two theories, several important limitations. Although Wilson and Kelling (1982) suggest that one element of the role of a police officer is to regulate the behaviours of the disorderly 'insiders' and exclude the troublesome 'outsiders' in an area, none of the studies pays sufficient attention to the question of who is committing the disorderly and criminal acts within an area. As a result, they tend to be unwittingly founded upon the notion that the majority of crime and disorder results from the actions of 'outsiders' and the premise of a clear-cut distinction between the disorderly and orderly segments of the population. However, particularly when we talk about social disorders, these often result from the actions of residents of an area, rather than outsiders. For example, in their qualitative study of a comparatively wealthy area of England, Girling, Loader and Sparks' (2000) data suggests that many residents were concerned about the activities of young people 'hanging around the streets.' But a significant proportion of those who expressed such concerns also recognised that these were the local children and that the problem was a symptom of a lack of investment to provide alternative activities for them. Similarly, Jock Young (1999), drawing upon a wealth of criminological evidence, has discussed how in urban areas 'the slow riots of late-modernity' typically involve communities 'imploding', gradually destroying the fabric of their own social and physical environments. Furthermore, as Matza (1964) persuasively argued, involvement in low-level delinquent and deviant acts is typically far better explained by a sense of people 'drifting' between conformity and non-conformity, than by a notion of the 'pathological outlaw.'

Related to this criticism of the three theories, is a recognition of the extent to which they presume the presence of shared norms in a community. At the very least, the idea should be entertained that different members of a community may understand the presence of physical and social disorders in different ways. We cannot presume that all the audience to a new 'broken window' reads it as a warning of community decline, resulting in a withdrawal.[9] For some it may be taken as an indicator that there is a need for action. Indeed, this point is recognised by Wilson and Kelling (1982) in their original formulation, although is frequently missed in subsequent treatments. Nevertheless, conceptually this means that we need to ask whether it is the broken window or the state of the community in which the window is broken that should be taken as the independent variable in any analysis.

Signal Crimes

In each of the three conceptual frameworks reviewed above, it is either explicitly or implicitly presumed that the presence of crime and disorder in a neighbourhood functions as a form of signal to both residents and outsiders, that is 'read' and used to inform beliefs about an area, and actions taken in respect of it. Significantly though, drawing upon the wider literature on risk perception reviewed earlier, and in particular Slovic's theory of the differing 'signal values' of different risks, we would argue that some crimes and disorders are especially influential in shaping the formation of beliefs about an area. These 'signal crimes' manifestly shape the beliefs that are constructed by individuals and groups about the degree of safety and security in an area.

Rather than suggesting that crime or disorder as a generic set of behaviours is generative of further crime, our argument is that certain crimes or disorderly behaviours are construed as 'signal crimes' and 'signal events' by individuals and communities. A signal crime / event can be defined as an incident that is disproportionately influential in terms of causing a person or persons to perceive themselves to be at risk in some sense. In effect, the crime or incident is 'read' as a warning signal by its audience(s) that something is wrong or lacking, as a result of which they might be induced to take some form of protective action. In addition, the presence of this signal will shape how the person or groups concerned construct beliefs concerning other potential dangers and beliefs.

Analysis of the 2000 British Crime Survey data provides some indications about the nature of the signalling that takes place. When questioned about their perceptions of levels of disorder in their area, of those living in comparatively prosperous areas, only a minority of respondents reported high levels of disorder, but significantly more respondents believed that the presence of this disorder had a negative impact upon their quality of life. In less affluent areas, a significantly larger proportion of respondents perceived there to be high levels of disorder, and even more thought that the presence of disorder had a negative impact upon their quality of life. Such findings are rendered even more important when linked to expressed concerns about being a victim of crime. Of those living in areas with high levels of perceived disorder 39% of respondents were 'very worried' about burglary, 34% were very worried about being mugged and 41% about theft of a car. In perceived low disorder areas these figures were 15%, 12% and 14% respectively (Budd and Sims, 2001).

Developing Vaughan's (2002) recent discussion of how signals function in organizational contexts, it is possible to distinguish between 'strong' and 'weak' signal crimes. Strong signals tend to result from those incidents that are of sufficient gravity and seriousness to generate a significant degree of public awareness about the individual event. However, shifts in people's belief systems and actions can also result from a cumulative exposure to a succession of weak signals. Thus although some low level disorders may not of themselves signal decline, a combination of weak signals may do this.

As a theory of communicative action, the concept of signal crimes does not assume that everyone will interpret the signal in the same manner. Key variables such as social class, age, gender, ethnicity, previous victimization and lifestyle may shape how the signal is interpreted and the meaning that is constructed on the basis of this. For example, Jefferson and Hollway (2000) note a gendered dimension to the construction of fear, in that for women fear of crime tends to connote sexual assault, whereas for men it is physical assault. Thus rather than assuming that there will be a homogenous interpretation of the signal, the conceptualisation reflects the fact that the characteristics of audience members, together with the situated context in which the signifier is located, shapes the construction of meaning. This recognition of the significance of context allows the concept of signal crimes to be sensitive to the role of social structure in shaping social reaction to a signal crime. This coheres with Sampson and Raudenbusch's (1999) findings that the strongest associations were between measured disorder, and structural disadvantage and attenuated collective efficacy.

This is important in that what is read as a signal crime by the residents in one area may not be interpreted in the same way by the residents in a different area. That is in a comparatively affluent and ordered neighbourhood a spate of graffiti may be seen as an indicator of potential problems because it has 'high dissonance' value.[10] Whereas in a chronically deprived neighbourhood, where there are other more serious behaviours taking place on a regular basis, the addition of more graffiti may be barely noticed.

Evidence in support of this argument can be gleaned from the Chicago Alternative Policing Strategy programme that has been running throughout the 1990s. In repeated surveys of Chicago residents across a number of different neighbourhoods, it emerged that for the poorer neighbourhoods, gang violence and drug related problems were consistently viewed as the most significant and prevalent problems over the decade. Nevertheless, in all neighbourhoods high levels of concern were consistently expressed about levels of social disorder and physical decay (CCEP, 2000).

Context is also important in terms of the nature and outcomes of action. The occurrence of a signal crime cannot automatically be assumed to have a negative impact on the community. One could predict under certain conditions where there are strong trust networks between residents and high levels of social capital, that they could be galvanised by the signal and encouraged to engage a variety of informal social control strategies. Thus rather than the presence of disorder negatively impacting upon community ordering mechanisms, the more plausible explanation is that the extant order will shape the nature of the reaction provided, which in turn will shape the efficacy of the social control mechanisms.

By focusing upon the concept of signal crimes what is made apparent is that not all events are assumed to have the same 'signal value'. In accordance with the wider literature on risk perception, certain disorderly behaviours or objects are held to have a disproportionate influence in shaping both individual and collective perceptions of risk. As such, communities may be able to tolerate certain kinds of disorders but not others. This susceptibility to particular crimes may reflect the aggregate characteristics of the community. Equally, it seems likely that there are certain disorders or crimes whose persistent presence is likely to degrade the quality of life for all communities.

A further advantage of the concept of signal crimes in analysing the crime-disorder nexus is that it avoids assuming that within a community there will be a given consensus as to what are troublesome or worrisome behaviours. What for young people is merely part of their normal routine and lifestyle (i.e. hanging around on street corners during the evening), may be construed by older residents as signals that they should not go out at night, in order to avoid threats to their safety.

Conceptual Elaboration

On the basis of the previous discussion, it is possible to identify three components to criminogenic risk perception arranged in a triadic relationship, that are consequential in the construction of a signal crime. These are:

  1. Object: Action / actor / location (AAL)
  2. Perception
  3. Reaction

The 'object' is the focus of a risk perception, it is the person, place or behaviour that is seen by an audience as signalling the presence of a threat. Thus the signal can be constructed around a particular action or type of action, a specific social actor or social group, or a place. Perception is the construed risk and the dimensions that inhere in it, in terms of how it is constructed and understood. Reaction refers to the ways in which the perceptions are used to inform a particular reaction to the actions / actor / or location that is perceived as constituting a risk. Importantly though, the relations between these three components are reflexive, with each involved in the articulation and apprehension of the other. Furthermore, they are moulded by a situated context.

Figure 1 below provides an illustrative description of the pathways of influence that are present.

Figure 1 Pathways of Influence in Constructing Risks

According to this model we can identify a dominant pathway and two subordinate pathways of influence in the construction of risk perceptions and thus the interpretative process of manufacturing a signal crime

The dominant pathway is that a person will encounter a deviant action or actor, or will come across a location where there is evidence of deviance being or having been committed. On the basis of their interpretation of this object they will infer a perception that there is a risk to their security, which will inform some form of reaction. In respect of signal crimes this will involve some re-framing of the person's 'cognitive map' of the social world. The social reaction may involve notifying the police or appropriate authorities, or engaging in some form of risk avoidance or risk minimization strategy. It should be noted though, that in some cases people will utilise 'techniques of risk neutralization' to persuade themselves that the perceived risk is less than consequential.[11] The engagement of techniques of risk neutralization will depend upon the context in which the three components are located as well as the person concerned.

However, from the point of view of reassuring the public about crime risks it is important to recognise that there are other pathways opened up. For example, it is likely that the perception of a signal risk at one time will have implications for any future interpretations of similar objects and subsequent reactions. Similarly, reacting in a particular way on one occasion may influence how other potentially risky actions/actors and locations are perceived and appraised.

By differentiating between object, perception and reaction in the process of constructing criminogenic risks we can effectively start to un-pack the elements that are involved in explaining how and why people will identify certain behaviours, people and places as representing threats to their safety. Equally importantly, we will gain some insight into the processes by which such understandings may inform their reported fear of crime.

In Figure 1 the triadic relations are framed by a context. This recognises that the perception of risks may be manifestly shaped by a number of 'context effects'. This reflects the previously noted situation where, for example, people in comparatively well-ordered middle-class areas may be sensitive to any disorders because such acts have a 'high dissonance value' in such a context. Whereas in an area with significant physical and social disorders already present, people may become 'de-sensitized' to any similar new occurrences.

Situated and Disembedded Signals

So far within this paper we have been discussing the construal and reaction to signal crimes at a local and parochial level. This reflects the general tenor of the literature on fear of crime, where it is direct experience of being a victim of crime that is identified as the key factor in predicting levels of fear (Box et al., 1988). But it is important to acknowledge that in late-modern societies individuals and communities do not acquire most of their information about crime from the local area. Mass media is an important, if not the most important source of information that people have about crime (Reiner, 1997; Sparks, 1991).[12] As such, we can distinguish between 'situated signal crimes', which are those incidents that people become aware of through co-present and proximate personal experience, and 'disembedded signal crimes' where knowledge is wholly media dependent.

According to such a distinction the sorts of processes set out by Cohen (1980) and Hall et al. (1978) in their respective studies of moral panics could be construed as the reactions to 'signal crimes.' In both of these analyses of the influence of media on criminal justice politics, the crimes that became the focus of media reportage were used to articulate wider popular fears about the state of society and its citizens. More recently, cases such as those of Damilola Taylor, Sarah Payne and James Bulger have all been the subject of extensive and extended media coverage, which in the process has provoked a degree of cultural introspection about the state of our society and the workings of the criminal justice system. The representations of these incidents and others like them, effectively serve to give expression to and perform many parents' fears concerning the safety of their children and thus may contribute to shaping parental assessments of risk. Furthermore, the media coverage in these cases has been imbricated in 'moral entrepreneurship' campaigns (Becker, 1968). In the language of the current analysis, these cases have been used as vehicles to signal that something is wrong in society and action is required to correct the situation becoming worse. This is not purely a contemporary phenomenon of course. Drawing upon historical accounts of crime and the nature of social reactions to crime, we can identify that perceived rises in the amount of deviant behaviour, or the arrival of 'new' forms of deviance, are periodically enacted as warning signs about the conditions of the extant social order (cf. Pearson, 1983).

Therefore, we can identify that there are a range of different crimes and disorderly behaviours operating at different levels of social organization that perform a signalling function. Analyses of the fear of crime phenomenon have repeatedly documented that certain offences, often the more serious kinds, frequently involving some form of interpersonal violence, are central components of how people articulate their anxieties about crime. These offences have this function even though many of the individuals who are concerned are unlikely to have directly experienced such crimes. The concept of signal crimes recognises the importance of these offences in shaping beliefs and worries, but supplements them with a range of what have frequently in many official crime control discourses been identified as less serious and consequential forms of behaviour. For it is these types of crimes that Wilson and Kelling (1982) correctly identify as impacting upon people's quality of life. Thus at a parochial and neighbourhood level, physical and social disorders can also operate as signals. The signalling function is a quality of the social reaction to the event, rather than being inherent to the act itself.[13]

The Implications for Policing

Signal crimes provoke a reaction on the part of a social audience. Situated signal crimes function at the local level through direct experience and community networks, whereas the more disembedded forms of signal, operate at a regional, national and international level through processes of mediated communication. The publicity afforded to the latter type sometimes has repercussions for the conduct of social control. The affective and emotive coverage provided to this relatively small number of cases, often involving serious forms of violence, has played an important role in the politics of criminal justice and the desire of sections of the public for more and better forms of security (Garland, 2001). In a less immediately obvious and dramatic way, similar social processes are being enacted throughout communities and neighbourhoods. Signal crimes involving seemingly trivial physical and/or social disorders serve to remind people of the risks to which they are potentially exposed, encouraging them to desire and make demands for more security, often in the form of a greater police presence.

In addressing this local level, the discussion has sought to draw upon and synthesize aspects of broken windows theory, Skogan's logic of amplificatory 'decay spirals' and Sampson and Raudenbusch's notion of collective efficacy. Central to the concept of signal crimes is the sense that particular criminal or disorderly acts and not others may be especially potent causes in de-stabilizing levels of organic community based informal social control. The research evidence is not clear as to whether the signalling functions as a direct cause of further crime. We can be more confident though that given certain conditions, the presence of signal crimes is more likely to stimulate a decline in levels of social cohesion and collective efficacy by degrading levels of mutual trust and confidence. In effect then, signal crimes adversely shape people's criminogenic risk perceptions, encouraging higher levels of fear and anxiety.

This formulation has important consequences for the problem outlined at the start of this paper. If the police want to tackle the seemingly intransigent fear of crime, then they may need to try and identify which signal crimes are especially generative of public anxiety. In turn this could inform an examination of what community policing strategies can be used to tackle the problems posed by these disproportionately influential signal crimes. For whilst in its original formulations community policing (CP) was designed to be responsive to locally-defined problems by employing the long-term beat assignment residents find reassuring, subsequently, as the discourse of CP has spread, so its implications for the delivery of policing services have somewhat dissolved (Fielding, 1995).

In a paper published in the journal Criminology in 1974, Bahn noted that police foot patrols could produce what he termed the 'reassurance factor.' That is although, as demonstrated by the Kansas City Preventive Patrol Experiment, police patrols are comparatively ineffective in terms of preventing or solving crimes (Kelling et al., 1974), patrol may provide other less tangible and measurable benefits to communities. In particular, high visibility community based police foot patrols may communicate a sense of 'guardianship' to members of a community, providing a focal point for the display of processes of informal social control. In order to produce this reassurance though, police have to start thinking about processes of symbolic communication, impression management, and the ways in which communities interpret crime and policing on a routine basis. For as Skogan and Hartnett (1997) have shown in Chicago, public reassurance is not generated simply by having more police patrols, the crucial factor seems to be what officers do when they are on patrol. Moreover, the measurement of the impacts of different strategic and tactical policy innovations upon levels of public reassurance will require the development of new and different forms of performance measuring instrument.

In his discussion of problem-oriented-policing, Goldstein (1990) suggests that one of the key failings of contemporary policing strategies has been their inability to recognise that all the 'business' that the public brings to the police should be 'police business.' Beset by increasing public demands for their services, police managers have increasingly foregone responding to low-level disorders in an effort to deal with what they perceive to be the more 'serious' incidents. However, although the public wants a response to these one-off incidents, they also want police to be concerned with and solve the everyday low-level problems that they regularly experience. In effect then, what has tended to happen is that police and public are attuned to different signal crimes. This gap has been reinforced and exacerbated by the machinations of the Home Office's performance management regime and the Home Secretary's annually revised priority crimes, which serve to direct police attention to the more serious crimes. As Fitzgerald et al.'s (2002) recent survey of policing in London showed, both the public and frontline police are of the opinion that the quantitative performance management regime has led to a deterioration in police responsiveness to public anxiety. One of the key concerns for the public was the development of a style of policing that was more responsive to local problems and local needs.[14] Indeed, in recent years where it is perceived that the public police have failed to adequately engage with the public's concerns about low level signal crimes, the public have demonstrated an increased willingness to accept these tasks being performed by private policing services.

If there is a connection between crime and disorder at an area level, this suggests that the police should develop ways of capturing data on the low- level signals that are influential in shaping how communities construct their sense of security. Furthermore, if there is a more robust linkage between the presence of disorder and particular crime types in areas, the capture of data about low-level disorders may function as a form of strategic 'open source' intelligence that can be used to identify developing problem hotspots.

In sum the capture of data on the nature and distribution of crimes with particular signal value may assist police in identifying the specific 'fear triggers' that exist for different communities. In turn, this will provide an opportunity to identify the strategies that will produce the best yield in terms of trying to reassure different sections of the public about their safety and thereby effect reductions in the fear of crime, and possibly subsequently crime itself.


In this paper we have proposed a conceptual toolkit for thinking about how and why certain offences and incivilities are constructed as 'signal crimes', which people use to guide and inform their wider understandings and actions about crime and the threats to which they are exposed. Signal crimes are important in as much as they may encourage the taking of preventative action, but they also inform and shape a wider belief system. In effect then, signal crimes influence how people perceive and react to potential threats to their safety. Thus the logic employed herein is that if it is possible to identify the physical and social disorders in an area that are particularly de-stabilizing to an individual's and community's sense of security, these would be key impact points for the police to address. Tackling these signal factors might be expected to produce disproportionate impacts in terms of improving public confidence, reducing fear and thus inducing a sense of reassurance. This represents an innovation in respect of how we seek to understand and respond to the 'fear of crime' problem. Rather than treating crime and disorder as an undifferentiated mass, it is proposed that tackling some physical and social disorders and some forms of crime may be more important than others. Importantly, this may not simply be an issue of tackling what are traditionally understood as the more 'serious' types of crime.


1Kershaw et al. (2001) records that between 1995-2000 the amount of crime reported to the British Crime Survey fell by about one third.

2The current paper sets out a theoretical framework for ongoing data collection designed to empirically test signal crimes theory, which will be reported in subsequent publications. Data is being collected via a small number of in-depth qualitative interviews with members of the public, incorporating a number of mnemonic techniques associated with cognitive interview instruments (see Geiselman et al., 1986). The incorporation of the latter dimension to the interviews is important in terms of addressing the known problems associated with encouraging respondents to articulate often complex emotional and psychological constructs (see for example Farrall and Ditton, 1999).

3See the collection of essays in Cerulo (2002).

4A further analysis of the problems with the concept can be found in the Police Foundation's recent submission to the Parliamentary Select Committee on Home affairs which can be viewed online at <http://www.parliament.the-stationery->

5The term 'social matrix' is taken from Ian Hacking's (2000) discussion of social constructionism.

6Indeed, data from the 2000 British Crime Survey suggests that between 1992 and 2000 there was a marked increase in the percentage of the population who saw various types of disorder as a problem (Budd and Sims, 2001).

7It should be noted that Harcourt (2001), in discussing his attempts to replicate Skogan's analysis, identifies a number of significant flaws in the latter's research design. Nevertheless, there is a strong empirical tradition underpinning this sort of argument stretching back to some of the keynote Chicago School studies.

8For an overview of the concept of social capital and some suggestions as to its potential application in relation to crime and disorder see The Performance and Innovation Unit's 2002 paper.

9This assertion is supported by elements of Jefferson and Hollway's (2000) analysis.

10Such an assertion is broadly supported by the BCS data presented previously.

11In using this term we deliberately echo Sykes and Matza's (1957) famous phrase.

12Ditton and Duffy (1993) express concern that the techniques and technologies of reporting associated with crime stories can induce in susceptible members of the audience, a distorted perception of crime risks.

13This deliberately echoes a Symbolic Interactionist epistemology and in particular W.I. Thomas's famous dictum that things that are defined as real are real in their consequences.

14Importantly from the point of view of the argument developed in this paper, the survey data shows a marked decline in public confidence in the police, and rising levels of concern about disorderly teenagers, litter, graffiti and drug dealing.


Bahn, C. (1974) 'The reassurance factor in police patrol' Criminology, (12) 338-45. BOX, S., C. Hale & G. Andrews (1988) 'Explaining fear of crime', British Journal of Criminology (28/3) 340-56.

BUDD, T. & L. Sims (2001) Antisocial Behaviour and Disorder: Findings from the 2000 British Crime Survey. London: Home Office.

CERULO, K. (ed.) (2002) Culture in Mind: Toward a Sociology of Culture and Cognition. New York: Routledge.

CHICAGO COMMUNITY POLICING EVALUATION COMMITTEE (2000) Community Policing in Chicago: Year Seven. Chicago: Illinois Criminal Justice Information Authority.

COHEN, S. (1972) Folk Devils and Moral Panics (2nd ed.). London: Paladin. CRAWFORD, A. (1997) The Local Governance of Crime. Oxford: Clarendon Press.

CUTTER, S. (1993) Living With Risk. London: Edward Arnold.

DITTON, J. & J. DUFFY (1983) 'Bias in the newspaper reporting of crime news', British Journal of Criminology (23/2) pp. 159-165.

DITTON, J., J. BANNISTER, E. GILCHRIST & S. FARRELL (1999) 'Afraid or angry? Recalibrating the 'fear of crime'', International Review of Victimology (6) pp. 83-99.

DOUGLAS, M. (1990) 'The forensic concept of risk' Daedalus (119/4).

DOUGLAS, M. (1992) Risk and Blame. London: Routledge.

DOUGLAS, M. & A. WILDAVSKY (1983) Risk and Culture. Berkeley: University of California Press.

ERICSON, R. & K. HAGGERTY (1997) Policing the Risk Society. Oxford: Clarendon Press.

FARELL, S. and J. DITTON (1999) 'Improving the measurement of attitudinal responses: an example from a fear of crime survey', International Journal of Social Research Methodology (2/1) pp. 55-68.

FEELEY, M. & J. SIMON (1994) in D. Nelkin (ed.) The Future(s) of Criminology. London: Sage.

FIELDING, N. (1995) Community Policing. Oxford: Clarendon Press.

FITZGERALD, M., M. HOUGH, I. JOSEPH & T. QUERISHI (2002) Policing for London. Cullompton: Willan.

GARLAND, D. (1996) 'The limits of the sovereign state: strategies of crime control in contemporary society', British Journal of Criminology (36/4) 445-71.

GARLAND, D. (2001) The Culture of Control. Oxford: Oxford University Press.

GEISELMAN, R., R. FISHER, D. MACKINNON and H. HOLLAND (1986) 'Enhancement of Eyewitness Memory with the Cognitive Interview', American Journal of Psychology (99/3) pp. 385-401.

GIRLING, E., I. LOADER & R. SPARKS (2000) Crime and Social Change in Middle England. London: Routledge.

GOLDSTEIN, H. (1990) Problem-Oriented Policing. New York: McGraw Hill.

HACKING, I. (2000) The Social Construction of What? London: Sage.

HALE, C. (1996) 'Fear of crime: a review of the literature', International Review of Victimology (4) 79-105.

HALL, S. et al. (1978) Policing the Crisis. London: Macmillan.

HARCOURT, B. (2001) Illusion of Order: The False Promise of Broken Windows Policing. Cambridge, Ma.: Harvard University Press.

HUTTER, B. & S. LLOYD-BOSTOCK (1990) 'The power of accidents' British Journal of Criminology (30/4) 409-22.

JEFFERSON, T. & W. HOLLOWAY (2000) 'The role of anxiety in fear of crime', in T. Hope & R. Sparks (eds.) Crime, Risk and Insecurity. London: Routledge.

KELLING, G., A. PATE, D. DIECKMAN & C. BROWN (1974) The Kansas City Preventive Patrol Experiment. Washington D.C.: Police Foundation.

KERSHAW, C., N. Chivite-Mathews, C. Thomas and R. Aust (2001) The 2001 British Crime Survey: First Results. Home Office Statistical Bulletin 18/01.

LUPTON, D. and J. TULLOCH (1999) 'Theorizing fear of crime: beyond the rational/irrational opposition', British Journal of Sociology (50/3) pp. 507-23.

MATZA, D. (1964) Delinquency and Drift. New York: Wiley.

PALMLUND, I. (1992) 'Social drama and risk evaluation', in S. Krimsky & D. Goulding (eds.) Social Theories of Risk. Westport: Praeger.

PEARSON, G. (1983) Hooligan: A History of Respectable Fears. London: Routledge.

PERFORMANCE AND INNOVATION UNIT (2002) Social Capital: A Discussion Paper. <http://www.cabinet->

PUTNAM, R. (2000) Bowling Alone. New York: Simon and Schuster.

REINER, R. (1997) 'Media made criminality', in M. Maguire, R. Morgan & R. Reiner (eds.) The Oxford Handbook of Criminology (2nd ed.). Oxford: Oxford University Press.

RENN, O. (1992) 'Concepts of risk: a classification' in S. Krimsky & D. Goulding (eds.) Social Theories of Risk. Westport: Praeger.

SAMPSON, R. & S. RAUDENBUSCH (1999) 'Systematic social observation of public spaces: a new look at disorder in urban neighbourhoods', American Journal of Sociology (105/3) 603-51.

SCHUERMAN, L. & S. KOBRIN (1986) 'Community careers in crime' in A. Reiss & M. Tonry (eds.) Communities and Crime. Chicago: University of Chicago Press.

SKOGAN, W. (1990) Disorder and Decline: Crime and the Spiral of Decay in American Neighbourhoods. New York: Free Press.

SKOGAN, W. & K. HARTNETT (1997) Community Policing Chicago Style. New York: Oxford University Press.

SLOVIC, P (1992) 'Perceptions of risk: reflections on the psychometric paradigm' in S. Krimsky and D. Goulding (eds.) Social Theories of Risk. Westport: Praeger.

SPARKS, R. (1991) Television and the Drama of Crime. Milton Keynes: Open University Press.

SYKES, G. & D, MATZA (1957) 'Techniques of Neutralization', American Sociological Review (22).

TAUB, R. et al (1984) Patterns of Neighbourhood Change. Chicago: University of Chicago Press.

VAUGHAN, D. (2002) 'Signals and interpretive work: the role of culture in a theory of practical action', in K. Cerulo (ed.) Culture in Mind: Toward a Sociology of Culture and Cognition. New York: Routledge.

WILSON, J. Q. & G. KELLING (1982) 'Broken Windows', The Atlantic Monthly (March Issue) 29-38.

YOUNG, J. (1999) The Exclusive Society. London: Sage.

Copyright Sociological Research Online, 2002