Social disorganization is not an exceptional phenomenon limited to certain periods or certain societies; some of it is found always and everywhere …
Social disorganization is not an exceptional phenomenon limited to certain periods or certain societies; some of it is found always and everywhere …(Thomas and Znaniecki 1920: 3)
While contemporary sociology and criminology associate social disorganization with urban areas, the concept actually emerged from studies of rural Europe. In The Polish Peasant in Europe and America, Thomas and Znaniecki (1920) explored how industrialization and immigration created disorganization among the traditional Polish peasant community and its rural culture. Despite these roots, applications of social disorganization in urban areas were commonly framed in the Chicago School’s depiction of the urban as pathological (Wirth 1938) and the rural as idyllic. This rigid distinction is not only untrue empirically – violence rates are consistently higher in urban areas, though these large-scale classifications obscure substantial variation in urban and rural crime rates, and data from Weisheit and Donnermeyer (2000) and Berg and Lauritsen (2015) suggest a decline in the urban-rural crime rate ratio from the early 1970s to 2010 – but is likely partially responsible for leading criminologists to neglect the distribution and causes of crime across rural places.
This neglect is problematic. First, if we recognize safety as a basic need, right or commodity, then rural citizens and areas should not be omitted from harm reduction attempts. Indeed, one out of six US citizens live in rural areas. Second, rural crime rates vary and we must discover why. The causes of this rural variation may be similar to or different from urban areas but it is too early to tell. Social disorganization is one logical place to begin this exploration given its dominance in structural criminology, which forms the aim of this chapter.
Explanations for crime across place are strongly influenced by the early Chicago School research (e.g. Thomas and Znaniecki 1920; Park and Burgess 1924; Wirth 1938; Shaw and McKay 1942). Cities across the United States experienced rapid population growth in the late 1800s due partially to massive waves of immigration. This influx of a variety of racial and ethnic groups into the urban core was characterized by industry, poverty and instability, and it sparked interest among scholars of crime. Of particular interest to Thomas and Znaniecki (1920) was how Polish immigrants successfully settled into the urban core given the seeming incompatibility between their traditional experience in rural Europe and modern urban Chicago. They discovered, however, that the Polish peasant community was changing long before immigration. Modern ization and industrialization had influenced traditional rural European culture and resulted in weakening kinship and neighborhood ties, and thus greater social disorganization, which they defined as ‘the decrease of the influence of existing social rules of behavior upon individual members of the group’ (Thomas and Znaniecki 1920: 2). This reduction of external control meant individuals had greater freedom to engage in unconventional behavior such as crime.
Building on this discovery, other scholars from the Chicago School applied social disorganization to the distribution of crime across urban neighborhoods, but neglected rural areas. For example, Shaw and McKay (1942) applied social disorganization as an extension to the social ecology model developed by Park and Burgess (1924). Park and Burgess (1924) had shown that immigrant populations in Chicago primarily settled into the urban core, but once they were able to adjust to the conditions of the city and gain enough resources they transitioned outward into zones with improved structural characteristics. Finding that delinquency was highest in the urban and industrial core, Shaw and McKay (1942) suggested this was caused by higher levels of social disorganization within the inner core, and also argued the physical decay and low economic status of these areas caused disorganization among the community and family as units of social control. Neighborhoods in the outer zones, with lower rates of outmigration and higher economic status, experienced higher levels of organization and control.
Other scholars suggested that urban areas are inherently pathological. Wirth (1938) argued increases in population size, density and heterogeneity resulting from urbanization contribute to an erosion of collective understanding that is replaced by increased interdependence. This argument creates an idyllic image of rurality. Similarly, Clinard (1942) suggested that rural communities are distinct from urban communities because they deter crime through personal relations and informal control. He wrote that ‘as long as there exists a predominant measure of personal relations and informal social control in the farm and village areas, it will be impossible for a separate criminal culture to exist as is characteristic of large urban areas’ (Clinard 1942: 213). Although these arguments no doubt relied to some extent on those of Durkheim, they ignored his clear belief that violence rates would be highest where the strength of sentiments about the collective was strongest (DiCristina 2004; Durkheim 1957, 1979: 356–8, 368–9; Pridemore and Kim 2006).
While Shaw and McKay’s (1942) model of social disorganization challenges the notion that urban areas are inherently disorganized and supportive of criminal behavior, Wirth’s (1938) conception held overwhelming support through the 1970s (Bursik and Grasmick 1993: 12) and this limited research attention on rural crime. Revitalizations and extensions of the social disorgan ization model remain primarily urban (e.g. Sampson and Groves 1989; Bursik and Grasmick 1993; Sampson et al. 1997). Beyond creating an idyllic image of rurality, this neglect further promotes the assumption of rural homogeneity (Wells and Weisheit 2004). There is substantial evidence, however, that structural characteristics and crime rates in rural areas vary widely just as they do in urban areas (Jobes et al. 2004; Wells and Weisheit 2004), and thus scholars have begun to consider if social disorganization can explain variation in crime across rural places.
Social disorganization theorists argue the relationship between structural characteristics and crime rates operates via the ‘inability of a community structure to realize the common values of its residents and maintain effective social controls’ (Thomas and Znaniecki 1920; Kornhauser 1978: 120; Bursik 1988: 521). Several structural antecedents have emerged as sources of variation in crime rates via community social disorganization. Due to lack of access to money and resources, for example, low-status communities have lower levels of organizational participation and youth supervision (Sampson and Groves 1989). Low economic status is also associated with high population turnover and ethnic heterogeneity (Bursik and Grasmick 1993). Because communities with a low economic status are less desirable, residents routinely emigrated from the industrial core as soon as economically possible. Meanwhile, racial and ethnic groups immigrated to the newly vacant homes.
According to the systemic model of social disorganization, residential instability and heterogeneity are the most salient causes of disorganization within communities (Bursik and Grasmick 1993). Residential instability is hypothesized to be associated with higher crime rates through its effect on residents’ interest in the community and in the development of social networks, while ethnic heterogeneity is hypothesized to be associated with higher crime rates due to the effects of cultural differences on communication. Residentially unstable and heterogeneous communities are less capable of developing prosocial relationships and less effective at crime control (Kornhauser 1978; Bursik 1988). Sampson (1987) identified family disruption as an additional antecedent of social disorganization because it may increase crime rates through weakening informal social control and supervision and reducing organizational participation. Sampson and Groves (1989) included urbanization as a final structural antecedent. Returning to a pathological image of urban areas, they hypothesize urbanization to be associated with weakened primary and secondary networks and less frequent organizational participation.
Scholars have also elaborated the theoretical mechanisms that capture social disorganization. These systemic factors include social ties, informal supervision and control, and collective efficacy (Kubrin and Weitzer 2003). The emphasis on social ties primarily emerged from Kasarda and Janowitz’s (1974) work on community attachment. According to Kasarda and Janowitz (1974: 329), ‘the local community is viewed as a complex system of friendship and kinship networks and formal and informal associational ties rooted in family life and on-going socialization processes’. Building from this view, social disorganization can be measured as the absence of social networks and collective supervision (Sampson and Groves 1989). More recently, social ties have been examined as three distinct dimensions: friendship networks, neighboring activities such as helping and sharing behaviors, and organizational participation (Bellair and Browning 2010).
Bursik and Grasmick (1993) extended the systemic model by noting that ties are only important insofar as they promote social control and socialization to conventional values. Weak or nonexistent social controls at private, parochial and public levels (Hunter 1985), and ineffective socialization serve as the most proximate causes of high neighborhood crime rates. Sampson et al. (1997) clarified this control mechanism embedded within social networks by introducing the concept of ‘collective efficacy’, which is the combination of social cohesion with shared expectations of control. Within neighborhoods with solidarity, mutual trust and high expectations for social control, residents tend to be more willing to acknowledge and confront neighborhood problems such as crime.
Some scholars suggest social disorganization theory may explain the differences in urban and rural crime rates because the latter are routinely characterized as having greater levels of cohesion, shared values and informal social controls (Websdale 1995; Barnett and Mencken 2002; Bouffard and Muftić 2006). Moreover, rural communities tend to have more family networks, stable populations and greater homogeneity (Gardner and Shoemaker 1989; Websdale 1995; Bouffard and Muftić 2006). Yet, rural communities are diverse and this heterogeneity may help explain variation in rural crime rates. In addition to clear exceptions to the rural idyll in the US, such as Southern Appalachia (Andreescu et al. 2011), Jobes et al. (2004) conducted a cluster analysis of rural Australian communities and found significant variation in rural community types regarding the structural antecedents of social disorganization and crime rates. Just as crime rates vary across rural communities, social disorganization also likely varies.
Freudenburg’s examination of rural ‘boom towns’ coincides with the social disorganization perspective (Freudenburg 1986; Freudenburg and Jones 1991). Freudenburg (1986) highlighted how rapid population growth in rural areas greatly damages existing networks resulting in less effective socialization and a breakdown of formal and informal controls. Similarly, drawing on Granovetter’s (1973) conceptualization of strong and weak ties, Wilkinson (1984b) noted that the heavy reliance on strong ties (i.e. familial networks) in rural areas can be problematic because weak ties (i.e. secondary networks) play an important role in supervision and socialization. Finally, Kaylen and Pridemore (2011) argued that sweeping events such as the 1980s farm crisis and more specific but widespread events such as the opening of big box stores can close businesses, disrupt social networks, and otherwise lead to social disorganization in rural communities.
Early research on crime across rural places focused on the structural antecedents of urbanization (Wilkinson 1984a, 1984b; Kowalski and Duffield 1990). Challenging the idyllic image of rural communities, results suggest that rurality is not a consistent predictor of crime rates. For example, while some early research found a negative relationship between rurality and homicide rates (Kowalski and Duffield 1990), in an examination of 299 counties in northeastern US, Wilkinson (1984a) found rural counties have significantly higher homicide rates than urban counties and argued that in modern societies, rural areas may present a barrier to community integration.
Recognizing that rural areas have significant variation in crime rates, scholars then began to examine rural as a scope condition. Initial tests were hindered by either small sample sizes or brevity (see Osgood and Chambers 2000: 84). For example, Arthur (1991) found that the percentage of the population that was black, of families receiving aid, and of the population living in poverty had effects on rural violent and property crime rates, but his analysis included only 13 rural counties in the Central Savannah River area of Georgia. Petee and Kowalski (1993) utilized a significantly larger sample size and found that residential mobility, racial heterogeneity and family disruption, but not poverty (in contrast to Arthur 1991), had significant positive relationships with robbery and assault rates.
To address the inconsistencies in the poverty-crime association in rural areas, Barnett and Mencken (2002) examined official violent and property crime rates in all nonmetropolitan counties in the continental US and found residential instability may moderate the effect of poverty on crime rates. Calling it the population stability hypothesis, they suggested that in growing nonmetropolitan counties, resource disadvantage has a stronger positive effect on crime because population growth disrupts the local network structure. Contrary to their prediction, Barnett and Mencken (2002) found that resource disadvantage had a stronger positive effect on crime in nonmetropolitan counties that were losing population. Nevertheless, these findings highlight that population change disrupts rural communities and may condition the effect of poverty on rural crime rates.
While the effects of economic measures (as indicators of social disorganization) on rural crime are inconsistent, results generally show other structural antecedents of disorganization such as residential instability, ethnic heterogeneity and family disruption have significant associations with crime rates. These findings tend to generalize to rural counties in different regions of the US (Osgood and Chambers 2000; Bouffard and Muftić 2006; Li 2011) and in Australia (Jobes et al. 2004). However, Bouffard and Muftić (2006) did not find a significant relationship between heterogeneity and violent crime rates. Because their analysis was conducted using 221 nonmetropolitan counties in four Midwestern states (North Dakota, South Dakota, Minnesota and Wisconsin), this finding may imply some potential regional differences in the role of the structural antecedents of social disorganization in predicting rural crime rates.
Before concluding that social disorganization theory is generalizable to rural areas, we must consider the conceptual and methodological limitations and the empirical evidence against this notion. First, scholars must carefully consider the definition of community. While the research on social disorganization and crime in urban areas often defines ‘community’ as neighborhoods (or administrative units such as block groups), similar research in rural areas usually uses large amorphous units such as counties for their analysis. Scholars should explore how residents of both urban and rural areas define community in order to more properly assess community-level theories.
Second, we should consider alternative causes of social disorganization in rural areas and even entirely different models of crime. For example, it may be that social forces other than the traditional structural antecedents operate to create disorganization in rural areas. Kaylen and Pridemore (2011) suggested events such as the 1980s farm crisis and the entrance of large corporate businesses may have unique disruptive effects on rural communities. Supporting their assertion, Wolfe and Pyrooz (2014) found new Walmart stores were associated with higher levels of crime. Additionally, the relative dependence on agriculture and absence of governmental officials in many rural areas may be important factors. In an analysis of rural Mexico, Villarreal (2004) found crime rates to be significantly associated with land distribution, collective organization around agricultural production and the presence of the state. That many rural communities are characterized by poverty (Tickamyer and Duncan 1990; Lee et al. 2003; see also Bouffard and Muftić 2006: 57) also presents a theoretical problem when attempting to generalize social disorganization theory to these areas.
Third, it is possible that social organization itself is qualitatively different in rural areas. Organization in rural communities may be strong enough to suppress the effects of structural antecedents on crime (Kaylen and Pridemore 2013a), suggesting a moderating model, or may even facilitate crime (Barclay et al. 2004; Donnermeyer et al. 2006). Finally, there may be alternative conceptualizations of social organization in rural communities. Lee’s (2008) work on the civic community perspective offers a unique set of factors, including the strength of local business and level of engagement in local and religious institutions. While distinguished from the social disorganization perspective, the conceptual similarity of civic community and social organization factors suggests that a theoretical integration may serve as a better model for explaining the variation in crime across rural communities (Wells and Weisheit 2012).
Research on social disorganization and crime in rural areas faces important measurement and methodological challenges. Measuring crime by official statistics hinders empirical tests of social disorganization in rural areas (Kaylen and Pridemore 2013a, 2013b). Beyond having significant measurement error in county-level UCR data (Maltz and Targonski 2002, 2003; Pridemore 2005), differences in reporting and recording practices may exist in urban and rural areas. Kaylen and Pridemore (2011, 2013a) argue that: (1) rural victims or witnesses may be more hesitant to report crime due to fear of disrupting small town social networks; and (2) rural police officers may be more hesitant to record crimes or make an arrest because they may know the offender or may want to work things out informally so as not to disrupt the community. The authors point out that both instances would bias estimates in favor of supporting social disorganization theory. Further, Berg and Lauritsen (2015) recently discovered that recording practices of rural police services (not the reporting practices of rural residents) explain much of the disparity between UCR and NCVS crime rates. Moreover, the authors found that while poverty is not associated with rural violent crime rates when measured by the UCR (a finding consistent with the traditional research on social disorganization and rural areas), the two are associated when rural violent crime rates are measured by the NCVS. These findings underline the use of official arrest data in tests of social disorganization as problematic (Kaylen and Pridemore 2013b).
With one main exception (Kaylen and Pridemore 2013a), most studies that concluded the association between social disorganization and crime rates generalizes to rural areas did not actually test the social disorganization model. Instead, they only examined the impact on crime of the structural antecedents of social disorganization. Social disorganization, however, is only one of many pathways through which these structural antecedents can influence crime rates. Thus, it is premature to conclude social disorganization generalizes to rural areas.
Recent research has challenged the generalizability of social disorganization theory to rural areas. First, the effects of poverty and related economic measures on rural crime rates are routinely non-significant (Petee and Kowalski 1993; Osgood and Chambers 2000; Jobes et al. 2004; Bouffard and Muftić 2006; Li 2011). In their analysis of 264 nonmetropolitian counties (populations ranging from 560 to 98,000 residents) in Florida, Georgia, South Carolina and Nebraska, Osgood and Chambers (2000) found poverty and unemployment were not associated with violence rates. They also noted rural poverty and residential instability were negatively related, which is inconsistent with research in urban areas.
Second, Kaylen and Pridemore (2011) conducted an analysis of youth victimization rates in 106 rural Missouri counties using hospital records and found that the percentage of female-headed households was the only measure of social disorganization to be significantly associated with assault victimizations. They found no association with violence of other structural antecedents of social disorganization such as poverty, residential instability and ethnic heterogeneity. To determine if their conclusions about the generalizability of social disorganization to rural areas, which were inconsistent with the conclusions drawn by others, might be due to various measurement or methodological limitations, Kaylen and Pridemore (2013b) analyzed four separate data sets that differed on either: (1) the measurement of the dependent variable (aggravated assault arrest data or hospital victimization data); or (2) the sample (their Missouri nonmetropolitan counties or a multistate sample comparable to that utilized by Osgood and Chambers 2000). By comparing the models from each data set, Kaylen and Pridemore (2013b) identified the measurement of the dependent variable as the most likely explanation for the different results between Osgood and Chambers (2000) and Kaylen and Pridemore (2011).
Third, Kaylen and Pridemore (2013a) recently conducted a test of the full social disorganization model in 320 rural areas by including measures of intervening mechanisms through which the structural antecedents are predicted to influence crime. The authors utilized questions from the British Crime Survey to replicate the intervening mechanisms used by Sampson and Groves (1989), which allowed them to examine the mediating effects of the structural antecedents via three measures of social disorganization: local friendship networks, problematic teenage groups and organizational participation. Their results suggest social disorganization in rural communities is only minimally influenced by the structural antecedents. Many of the relationships between the structural antecedents and the intervening mechanisms were non-significant or significant in the direction opposite prediction. For example, ethnic heterogeneity was positively related and residential stability negatively related to local friendship networks. While they did find that some associations were consistent with social disorganization predictions, the social disorganization model explained a very small proportion of the variance in rural crime rates. In whole, Kaylen and Pridemore (2013a) concluded that the full social disorganization model does not operate to explain crime in rural areas in the same manner as it does in urban areas.
There is substantial inconsistency regarding the generalizability of social disorganization theory to rural areas. While early rural research claims support for the social disorganization model (e.g. Osgood and Chambers 2000; Bouffard and Muftić 2006), recent rural research challenges the effect of the traditional structural antecedents on social disorganization measures and even the effect of social disorganization measures on crime rates (Kaylen and Pridemore 2011, 2013a, 2013b). Continued research, with a focus on the conceptual and empirical limitations, is necessary. For example, scholars might consider unique aspects of rural social organization or consider alternative causes of social disorganization in rural areas.
Scholars also should explore alternative strategies of crime measurement in rural areas. Analyses examining crime via victimization surveys or hospital data in addition to official police data may clarify the effect of social disorganization measures on crime. Furthermore, future research should conduct tests of the full social disorganization model. Tests of the full social disorganization model within the US are required because it may be the case that national contexts influence the role that social disorganization variables play in predicting crime. Tests may also examine alternative intervening dimensions of social disorganization, including neighboring, control or collective efficacy (Sampson et al. 1997). Reisig and Cancino (2004) found that collective efficacy is associated with higher levels of perceived signs of decay and disorder in nonmetropolitan communities, though research has yet to examine the effect of collective efficacy on rural crime.
Research on crime and place in rural areas has been overshadowed by research in urban areas, and thus is a relatively new exercise. In recent decades, significant advancements have been made towards the theoretical frameworks and empirical understanding of the distribution of social disorganization and crime rates across rural counties. Conceptual and methodological limitations continue to stunt the growth of this topic, however, and we urge scholars to consider additional theoretical specifications and empirical methods for future research on variation of crime across rural place.