Violence events tend to cluster together geospatially. Various features of communities and their residents have been highlighted as explanations for such clustering in the literature. One reliable correlate of violence is neighborhood instability. Research on neighborhood instability indicates that such instability can be measured as property tax delinquency, yet no known work has contrasted external and internal sources of instability in predicting neighborhood violence. To this end we collected data on violence events, company and personal property tax delinquency, population density, race, income, food stamps, and alcohol outlets for each of Richmond, Virginia’s 148 neighborhoods. We constructed and compared ordinary least-squares (OLS) to geographically weighted regression (GWR) models before constructing a final algorithm-selected GWR model. Our results indicated that the tax delinquency of company-owned properties (e.g., rental homes, apartments) was the only variable in our model (R 2 = 0.62) that was associated with violence in all but four Richmond neighborhoods. We replicated this analysis using violence data from a later point in time which yielded largely identical results. These findings indicate that external sources of neighborhood instability may be more important to predicting violence than internal sources. Our results further provide support for social disorganization theory and point to opportunities to expand this framework.
Funding: Acknowledgements: Drs Thomson and West’s effort on this publication was supported, in part, by the Center for Disease Control and Prevention National Center for Injury Prevention and Control ( cdc.gov ) under Award Number 5R01CE003296 (PI: NDT). The content is solely the responsibility of the authors and does not necessarily represent the official views of the CDC. The CDC played no role in the design or conduct of the study described in this manuscript.
Introduction
Violence is a major public health problem in the United States. Rates of homicide in the US are 7.50 times that of other high-income nations [1]. This problem is more pronounced in Richmond, VA, where the leading cause of death among young people is violence [2]. Epidemiological research has placed an emphasis on the environments in which violence is most likely to occur, revealing that waves of violence occur similarly to viral epidemics, spreading like a contagious disease [3]. A clear picture emerges from this research: low-income and otherwise underserved communities are at the highest risk for being exposed to violence in some way [4]. In the current work we drew from social disorganization theory to examine the impact of various geospatial features on neighborhood-level violence data from Richmond, VA.
Theoretical accounts of violence Several theoretical models have been put forward in the literature to explain observed trends in violence and antisocial behavior. Social disorganization theory (SDT) places a major emphasis on specific elements of the environment a person resides in explaining such antisocial behavior [5,6]. SDT broadly identifies important physical features of communities such as physical disorder (e.g., dilapidated buildings) [7] and meaningful social ties (e.g., with one’s neighbors) [8] as primary factors in predicting antisocial behaviors. Evidence indicates that stronger social ties in a community facilitate a sort of collective efficacy via expanded support networks and other forms of resource sharing which may provide access to a broader array of mediation options when conflicts may arise [9,10]. This element is also theorized to ameliorate the negative effects of societal impacts originating from outside one’s own neighborhood (e.g., poverty; residential instability) [11]. In support of these assertions research indicates that various forms of violence (e.g., intimate partner violence, homicide) are directly linked with poor collective efficacy and low stability (i.e., social disorganization) in a given neighborhood [12,13]. These findings have been replicated in urban and rural communities [14] and other countries such as Brazil and South Africa [15,16]. One major source of such disorganization identified by SDT is that of residential instability.
Neighborhood residential stability and violence Neighborhood residential instability refers to how likely it is for members of a given neighborhood to relocate, typically due to the loss of one’s home. At a conceptual level low neighborhood stability makes good sense as a predictor of violence. Consistent with SDT, neighborhoods with high resident turnover are likely to have poor social bonds among neighbors which can lead to little willingness to work together with others towards common goals (i.e., reducing or preventing violence) [17]. Recent work further indicates that low neighborhood stability is a strong predictor of interpersonal violence [18,19]. Various operational definitions of neighborhood instability have been applied in the literature. Some work has utilized measures of tenure or the proportion of the residents who recently moved into a given neighborhood. However, this measure may be too general to identify the probable sources of such residential turnover which may account for the weak relationships observed between measures of tenure and rates of crime [20]. Given the emphasis SDT places on sources of disorganization (i.e., endogenous versus exogenous disorganization) distinguishing between such sources of residential instability may further our understanding of the influence of this factor on violence. Another measure that has emerged as appropriate for examining social disorganization at the neighborhood level is that of property tax delinquency [21]. The number of tax delinquent properties in a given residential area captures crucial structural differences across neighborhoods which are strongly linked with greater socioeconomic disadvantage, poorer community health, and greater neighborhood disorganization and distress [21,22]. Underscoring these differences, those from high tax delinquency neighborhoods are also more likely to die at an earlier age [22]. Similarly, a recent study found that the number of tax delinquent and vacant properties in a neighborhood also predict greater rates of violent crime [23]. Despite this evidence, no known research has examined the potential for divergent sources of neighborhood instability to differentially predict violence at the neighborhood level.
Contrasting sources of instability: Company and personal tax delinquency In general, there are two forms of tax delinquency that comprise most delinquent residential properties: those owned by companies or landlords (i.e., rental properties) and those owned by the residents themselves. The implications of personal tax delinquency are relatively clear when considered in the context of SDT: owners of run-down homes are less likely to keep up the financial state of their home much like the physical state of their home, thus introducing social disorder and contributing to an environment that is conducive to violence [24]. SDT also indicates that such physical disorder is likely to be intervened upon by other residents as neighborhoods are endogenously self-stabilizing [25]. However, the implications of tax delinquent company-owned properties are less clear. In the case of rental properties, the tenants themselves often have little say in the state of the building they live in and are typically at the mercy of their landlord for repairs or improvements, which may diminish the ability of collective efficacy to address such physical disorder. Tax delinquent rental properties hold unique implications for the housing security of tenants, as many localities have laws allowing properties that have been in delinquency for a given period to be foreclosed and auctioned to other investors. Such transactions are directly linked to eviction rates and thus pose a critical threat to the collective efficacy of a community in turn [26]. For example, in the city of Richmond, VA properties that are tax delinquent for two years are commonly seized by the city and auctioned to recoup lost tax revenues. Richmond (like many other localities) has no official regulation protecting tenants in the case that winners of auctioned delinquent properties decide to evict them. Evidence indicates that real estate companies frequently purchase such delinquent properties, targeting low-income neighborhoods and communities of color, ultimately disrupting any resident-based revitalization or improvement efforts [27]. The effects of company property tax delinquency may thus be distinct from private tax delinquency as company tax delinquency appears to contribute to both physical disorder and residential instability and may thus poses a greater threat to collective efficacy. Although there is no extant empirical literature comparing the impact of company and personal tax delinquency directly, one example revealing the impact of company delinquency does exist. The East Liberty neighborhood had one of the highest crime rates in Philadelphia in 2008. A group of residents decided to form a real-estate investment group, named East Liberty Development Inc. (ELDI), for the purpose of revitalizing many of the dilapidated, tax delinquent properties in their neighborhood in hopes of reducing crime rates [28]. After numerous interviews with East Liberty residents, ELDI leadership concluded that there was a connection between crime rates and slumlords–that is, rental property owners who had not taken care of their rental properties despite having tenants. ELDI thus began its “slumlord buyout” program, wherein they purchased run down rental properties from slumlords in order to renovate them in collaboration with the current tenants. Studies examining the impact of this program revealed that violent crimes (i.e., aggravated assault, sexual assault, homicide) decreased by 49% over a four-year period [29,30]. As such, it appears that rental properties with absentee landlords (e.g., those holding tax delinquent properties) may predict violence in a given neighborhood, but it remains unclear if this variable is a more appropriate predictor of violence than the tax delinquency of personal properties. A proper investigation of these relationships must account for the geospatial clustering commonly observed with violence.
Geospatial features of violence Geospatial analyses have been used in various fields (e.g., epidemiology) to determine patterns and predictors of violence in selected geographical regions. This approach has yielded meaningful insights that can be applied towards effective localized interventions. For example, applying geospatial analysis to violence data can reveal high-risk neighborhoods that can be targeted with intervention efforts [4]. Indeed, severe violence events are most likely to occur within the victim’s own neighborhood [31]. Those living in densely populated, low-income, and racially diverse neighborhoods are at a significantly higher risk of violence than those from predominantly affluent, white neighborhoods [32]. One analytic technique known as Geographically Weighted Regression (GWR) [33] accounts for the spatial heterogeneity of data by allowing the estimated parameters to vary across each designated region while accounting for the influence of neighboring regions. This approach allows researchers to examine the extent to which predictor variables account for a given outcome (i.e., violence) in each region (e.g., neighborhoods) and at the global level (e.g., a whole city). For example, research applying this approach indicates that the density of alcohol outlets predicts levels of violence [32,34]. Population density has also repeatedly emerged as an important geospatial predictor of violence [35]. Other work using GWR indicates that food insecurity is a strong predictor of gun violence [36]. Research reveals that cities historically affected by racist ‘redlining’ housing practices demonstrate spatial heterogeneity in statistical models predicting forms of antisocial behavior due to the protracted effects of the discriminatory policies used to shape the demographics of affected neighborhoods [37–39].