Environment and Crime among Residents in Urban Areas

A Study of Districts in Stockholm

Olof Dahlbäck

• Publisher: Routledge
• Year: 2016
• eBook ISBN: 9781317142416
• Topic: Social Sciences
• Subtopic: Criminology
• Index: Social Sciences

This groundbreaking book by Olof Dahlbäck analyzes the direct effects of the environment and the indirect effects of geographical differentiation of individuals on the offender rates of different urban areas. In order to do this, relationships between crime and independent factors are analyzed in various ways - by considering cross-sectional and longitudinal aspects, linear and non-linear models, point and change data, different time periods, micro- and macro-level interaction, and data for individuals with different patterns of moves. The offender rates analyzed refer to individuals suspected by the police. The directly crime-influencing processes focused upon imply that individuals are affected by social control and social resources. The study makes use of advanced analytical models, novel methods and comprehensive data, and it solves several problems that have hampered research.

Chapter 1
Previous Research

Urban Areas and Crime

The subject of how crime emerges and develops in cities is of great criminological significance. The reasons for this are obvious: crime has shown to be strongly related to urbanization – the urban environment seems to be conducive to its inhabitants engaging in criminal activity – and a very large portion of the populations in many countries lives in cities.

The existence of strong relationships between crime rates of the inhabitants of areas in large cities and other social properties of these areas seems to be a highly general phenomenon. For traditional types of crimes, similar relationships have been found in many Western countries (Wikström, 1998). In the cities studied, there seems as a rule to have been a regional differentiation as to various social factors, and several of these factors have been shown to be related to crime. In particular, it has been found that areas with disadvantageous social conditions have high rates of crime. This finding has been obtained for a variety of measures of crime and for areas of different sizes.

Most of the studies dealing with the relationship between crime rates of the inhabitants of urban areas and other social properties of these areas are methodologically simple. The most common design is probably the cross-sectional type applied to register data. The independent factors studied are typically quite similar: as a rule factors describing economic resources, family/civil status, ethnic status, and mobility are included. To a great extent, this is probably due to the fact that these factors can be easily measured by register data. Another reason may be that the research that was conducted by Clifford Shaw and Henry McKay (see below) has been influential in focusing on these types of factors. Nevertheless, it seems clear, as discussed below, that several of these factors are interesting from theoretical points of view.

Relationships of the types mentioned above – between the crime rates of the inhabitants of urban areas and other social properties of these areas – can be found in the City (that is, the municipality) of Stockholm, the capital of Sweden. However, one might think that there is reason to expect that Stockholm would not exhibit these types of relationships. Its regional social differentiation is smaller than for many other cities in Western Europe and in the USA due to the public housing policy in postwar Sweden, which has aimed at counteracting residential segregation by socioeconomic characteristics. Nevertheless, as will be shown in this book, the regional differences in Stockholm have in fact been substantial and large enough to generate strong relationships between social factors and crime on the area level.

Furthermore, postwar Stockholm is interesting to study due to the emergence and development of a variety of problematic conditions. Substantial changes occurred during the 70s and 80s in the social conditions in the city. For example, the numbers of divorced individuals and of immigrants increased strongly. New housing areas were built, some of which later became the locus of serious social problems.

There is a good deal of previous research on crime in Stockholm that makes use of register data. Per-Olof Wikström is among the more prominent of those who have carried out such research. In what is probably his most important study on districts in this city, Wikström (1991) analyzes a comprehensive body of data. The results show the same type of pattern of cross-sectional relationships between crime rates and various factors as can be found in many other studies. However, due to the nature of the data, the possibilities of performing an advanced causal analysis were very limited in Wikström’s study.

Theoretical Perspectives

The theories most often discussed in the field attribute causes of crime among residents in urban areas to one or more of the following conditions: social disorganization, weak social control, lack of social resources, or subcultures in these areas. Social disorganization theory was developed specifically to explain urban crime. The social control and social resources theories have a more general sociological character, but are interesting because social-control and social-resources factors have been found to vary geographically in large cities and to be strongly related to crime among the residents in the areas. The theory of urban subcultures refers to a variety of deviant behaviors, among these crime.

Social disorganization theory, originally developed by Clifford Shaw and Henry McKay (Shaw and McKay, 1969 [1942]), has played a central role in the research. According to Shaw and MacKay, social disorganization in an urban area is an important cause of crime among the youth living there. By “social disorganization” it is meant that inhabitants have poor social contact with each other and an inadequate organization of activities needed to protect their interests, for example to protect against crime. According to Shaw and McKay, disorganization may be traced back to the lack of economic resources, ethnic heterogeneity, and high mobility of area residents. This theory has been the subject of several different interpretations and revisions (Wikström, 1998), and has also been the target of much criticism (as to criticism focusing on individual-level aspects, see Farrington, 1993). In its modern form, the core of the theory seems to consist of ideas about social control.

According to social disorganization theory, more disorganized urban areas generate more criminal activity. Shaw and McKay maintained that they had shown that registered crime in Chicago was distributed in such a way that the highest level of crime existed in a zone where social organization, measured with a set of indicators, tended to be worst. Moreover, they maintained that this level of crime remained about the same over time, independent of how the population in large was constituted. Some of their empirical results, for example regarding the permanent level of crime, have been questioned or refuted (see, for example, Bursik, 1988). However, their ideas about the significance of social disorganization and its underlying factors have had great impact on criminological research. They belong to the standard repertoire of theoretical ideas in social ecological contexts.

Social disorganization theory has attracted renewed interest in more recent years, largely inspired by an article by Sampson and Groves (1989). These researchers modified the theory in several ways. For example, they introduced factors that they believed describe disorganization more directly than was previously the case. They argued that the theory had never been tested in full, because disorganization had only been measured indirectly with the three factors assumed to underlie it. They conducted an empirical study that they contended showed relationships on the area level between crime and new direct indicators. Social organization was measured in terms of local friendship networks, control of street corner teenage peer groups, and prevalence of organizational participation. Sampson and Groves’ ideas and their study have strongly affected subsequent research about disorganization and crime (for example, Bellair, 1997; Sampson et al., 1997), also research with a critical stance (for example, Veysey and Messner, 1999).

The idea that networks are an important aspect of social organization has played a prominent role in modern research on disorganization and urban crime. For example, Bursick and Grasmick (1993) have argued that disorganization theory should consider not only the local networks among residents, but also the networks that connect neighborhoods with the greater society in which these are imbedded. Connections with schools, churches, the police and other institutions and agencies outside the neighborhood may be important. However, seeing general networks among residents as a factor that keeps crime down has been met with skepticism. It is a fact that there are neighborhoods with relatively dense networks among residents that nevertheless exhibit fairly high crime rates, and it has been argued that while such networks may contribute to neighborhood ability to control crime, they may also provide a source of social capital for offenders (Browning et al., 2004). Furthermore, insofar as networks do affect crime their causal status has been problematized. For example, it has been argued that networks, and local organizations and voluntary associations as well, only have an indirect effect via collective efficacy, measured as residents’ views of the social cohesion and social control (Morenoff et al., 2001). In these and other modern researches testing social disorganization theory, the concept of social capital is important (Portes, 1998). Salmi and Kivivuori (2006) studied how crime on the individual level is related to indicators of social capital and various other independent factors, and they found crime to be negatively related to a number of the indicators.

Social disorganization theory has been cited in many contexts. There is probably some truth in it, but it is unclear what and how much it explains and it has been heavily critized. I myself am critical of it and of much of the research that is based on it. I think that its coupling between the micro and macro levels is unclear and that it is questionable as an explanation of neighborhood crime to the extent that the social organization referred to is assumed to be collective, that is, to exist outside the family (and, in my view, if organization is assumed to refer to the family, the theory has hardly anything unique to contribute). In actuality, tests of the theory have not yielded unequivocally positive results. For example, according to the theory, mobility is expected to have a strongly positive relationship with the crime of the inhabitants on the area level. Certainly, newly arrived inhabitants in an urban area can hardly be socially well organized. The fact is, however, that it has often been difficult to find such an empirical relationship (this is for example the case in the Sampson-Groves study). There is also much to criticize in the research that has been adduced as support for the disorganization theory, particularly regarding the methods used. I therefore believe that the possibilities of drawing conclusions about the validity of the theory from this research are severely limited. Some of my methodological criticisms are presented below.

Much of the modern sociological research on crime is dominated by theories that stress the significance of social control. Many of these theories are revised versions of or are strongly influenced by the “social bonding theory” propounded by Travis Hirschi in his book Causes of Delinquency (Hirschi, 1969). A main idea in this theory is that crime arises when individuals’ bonds to the conventional society are weakened or broken (for a discussion, see Akers, 2000, pp. 105–110). There are several different types of bonds, including personal bonds to individuals not engaged in criminal activities, stakes in conventional work (for example regarding education and occupation), engagement in conventional activities, and belief in conventional values and norms. The Sampson-Laub life-course theory of social control is founded on this perspective (Sampson and Laub, 1990, 1992; Laub and Sampson, 1993). This theory focuses on the influences on crime of the social bonds created by social institutions and major life events (for example, getting a steady job, getting married, becoming a parent) and it deals particularly with how such bonds facilitate desistance from crime.

Social control theories have been supported by many studies (as to the Sampson-Laub theory, see Savolainen, 2009). Family disintegration and attachment to parents are examples of aspects of social bonds that have been found to be associated with crime (Rankin and Kern, 1994; Skarðhamar, 2009).

There are different theories built on the assumption that lack of social resources and inequality cause crime, and they have been used as a basis for a comprehensive body of empirical research (Akers, 2000, pp. 143–162). However, the results of the many studies made on the individual level are not, it seems, as clear as are the results obtained in studies in which social control factors are used, as least not as far as the criminality of young individuals is concerned. The association between social class or socioeconomic status or income, which are properties often used to describe social resources, and the criminality of young individuals turns out to be weak or non-existent in most cases. This has made researchers question the existence of such an association of any greater importance and has led to a long debate about the matter and to attempts to show that social class and similar properties only play a modest role for crime (Tittle et al. 1978; Braithwaite, 1981; Tittle and Meier, 1990; Dunaway et al., 2000; Agnew et al., 2008; see also Becker and Mehlkop, 2006; Ring and Svensson, 2007). Furthermore, there are several questions about the theories, for example, exactly how the lack of resources and the inequality exert an influence on crime.

Thus, the results of the research seem to indicate that the individual-level association between social class or similar properties and crime is weak at most. Furthermore, the influence of social class on crime can be assumed to be indirect, and it has been shown that controlling for various factors supposed to mediate this influence strongly reduces an association found (Fergusson et al., 2004).

However, despite the failure to find a strong relationship between young individuals’ criminality and their social class or similar properties, some researchers have maintained that there is such a relationship. For example, it has been suggested that the failure is due to ignorance of the fact that the tendency to become delinquent is extraordinarily strong among individuals who experience persistent childhood poverty. Jarjoura et al. (2002) have presented empirical support for the existence of such a tendency. Thus, it seems that individuals with very poor social resources should be particularly focused when analyzing the causes of criminality. The finding of a relationship on the individual level between severe poverty and criminality has a parallel on the societal level. On this level, it has been found that the crime rates of communities are better predicted by factors that describe the existence of extreme disadvantage of the neighborhood than by factors that describe a more balanced picture of such disadvantage (Krivo and Peterson, 1996).

Social resources may refer to different things – for example, money and things having economic value, knowledge, status, and social relations. As to the economic dimension, the question may be raised whether it is the absolute or the relative economic value of resources that affects crime. It might seem that there are good reasons for focusing on the relative dimension. However, while there may be a relationship on the individual level that implies that poorer individuals tend to commit more crimes, there may on the societal level be a relationship between economic resources and crime that implies that more resources lead to more crime, because greater affluence of resources results in the fact that more objects having economic value are poorly guarded and are therefore easier to steal.

In the research, the predominating idea of how lack of social resources and inequality cause crime on the individual level is the frustration hypothesis. According to this hypothesis crime may be committed by resource-poor individuals, because poor resources lead to failure to achieve positively valued goals and to a sense of being unjustly treated – consequences that lead to frustration and to a pressure for corrective action, which may be crime. This explanation was early suggested by Richard Cloward and Lloyd Ohlin and by Robert Merton (Cloward and Ohlin, 1960; Merton 1968, pp. 185–248).

Robert Agnew has taken up and revised the frustration hypothesis (Agnew, 1992, 1993, 1999), and his theory has attracted much interest in the research in later years (Froggio, 2007; Rebellon et al., 2009; Froggio et al., 2009; Botchkovar et al., 2009). According to Agnew, the negative emotions that accompany the frustration following lack of social resources and inequality often turn into desires for revenge, which energize the individual for action. These views and emotions may be a motive for committing crime. Furthermore, Agnew has suggested that the strains associated with a lack of social resources and inequality interact with other factors, for example control factors, in the influence on crime (Agnew, 1999; 2001; 2005, pp. 109–120). This interaction could be due to the fact that while lack of social resources provides a motive for committing crime, lack of social control provides a prerequisite for this motive to be realized in action (Agnew, 1993).

Evaluating the results of the tests made of Agnew’s strain theory is made difficult by various methodological problems. As to the core of the theory, the results must be characterized as mixed. As to the idea that strain interacts with social control in the influence on crime, the results are still more problematic, but some findings indicate or can be interpreted to indicate that such interaction exists (Agnew and White, 1992; Mazerolle and Maahs, 2000).

Finally, a fourth type of theory seeks the explanation of crime among residents in urban areas in subcultural conditions (Fischer, 1995). The fundamental assumption here is that norms and values that make people prone to commit crime can be found in their culturally determined attitudes and ideas. Two types of theories dominate. They explain urban crime by a subculture of violence or by a subculture of poverty. Subculture of poverty theories have often focused on crime in ghettos in central cities, but this is a type of environment that does not exist in Stockholm.

Geographical Selection of Individuals and the Influence from Local Environment

The theories just discussed are not total in the sense that they can on their own adequately explain crime rates of urban areas. Instead, they may offer partial or idealized explanations, focusing on particular aspects. This is not surprising, since the social mechanisms that underlie the crime rates of urban areas can be assumed to be complex and difficult to describe in full. A micro-macro view of the matter provides insights into this complexity. Crime rates and other factors of areas can be seen as aggregates of individual data of the inhabitants. The relationship between the rates and the factors may be analyzed by considering two types of mechanisms: mechanisms producing relationships on the individual level between various properties and the generation of crime (crimes committed by the individual or by others), and geographical selection processes that imply that individuals with certain properties and certain tendencies to generate crime are to be found in certain areas. The mechanisms producing the individual relationships are of two types – those that work on the local level only and those that work on the global level. Global mechanisms are those mechanisms that work in the same way everywhere irrespective of geographical location. Local mechanisms, on the other hand, are specific for some particular area or areas. They are dependent on the conditions – the physical conditions and the social make-up – in this area or these areas.

Thus, in order to explain a rate of crimes committed by the inhabitants in an area at least three issues must be considered: 1) the global individual-level relationships between the generation of crime and various properties, 2) the geographical selection of individuals according to some factors and the resulting distribution of properties in areas, and 3) the influences of local properties on the generation of crime on the individual level. There can be no doubt that the first two of these issues are important. Certainly, there are properties that are related globally on the individual level to the generation of crime and that are also involved in geographical selection processes, thereby creating a relationship between the rates of properties and the rate of crime in areas. The selection is above all determined by the economic resources of the individuals and by the properties of the areas, including the dwellings in these. The competition among individuals for attractive dwellings and other local advantages lies at bottom of this process. Thus, individuals’ economic resources are a very important selective factor, and even if these resources do not cause crime directly on the individual level, they are related to properties, for example other resources properties and control properties, that do cause crime, and these crime-affecting properties are therefore geographically differentiated.

Thus, an area’s social make-up may have an indirect influence on its crime because of the selection of individuals with certain tendencies to generate crime globally. But how is it with the direct influence of the social organization of those individuals living in the area? Social organization should then be taken in...

Table of contents

• Cover Page
• Title Page
• Copyright Page
• Contents
• List of Figures
• List of Tables
• Acknowledgment
• Introduction
• 1 Previous Research
• 2 Theory
• 3 Method
• 4 Results I. Cross-Sectional Relationships
• 5 Results II. Longitudinal Relationships
• 6 Summary and Conclusions
• References
• Appendices I-V
• Appendix I Definition of Areas
• Appendix II Populations of Areas
• Appendix III Maps
• Appendix IV Construction of a Measure of Crime in Adjacent Areas
• Appendix V Tables
• Index