Studies examining the relationship between neighborhood disorder and neighborhood drug arrests are underrepresented and denote a significant gap in existing knowledge. Neighborhood disorder refers to a lack of order and social control within the neighborhood and is often measured by visible cues within the neighborhood environment (e.g., graffiti, broken windows, abandoned housing) (1). Some early research suggested that the relationship between neighborhood disorder and crime may be spurious – explained better by levels of collective efficacy – but more recent literature suggests that although collective efficacy may mediate the relationship between the two factors, there is no evidence that the relationship is spurious (1,2). Two more recent studies have examined the relationship between neighborhood disorder and juvenile offending (3,4). Fite et al. (3) examined the risk factors that might explain the increased rates of juvenile arrests among African-American males compared with Caucasian males. The authors found that neighborhood disorder (self-reported caretaker reports of vandalism, assault, etc.) was the most important factor accounting for the discrepancy in violence-related arrest rates between African-American and Caucasian males. However, there was no relationship between neighborhood disorder or any other risk factors (externalizing behaviors, parent–child relationship, etc.) and drug arrests. In an earlier study, researchers found that neighborhood disorder moderated the relationship between impulsivity and juvenile offending; the relationship between impulsivity and juvenile offending was stronger in low-income neighborhoods (4).

Although these earlier studies were informative, they relied on self-report and census measures to characterize the neighborhood environment. Census data and self-report data have been useful and widely used in neighborhood research, but there are inherent limitations to both sources of data (5,6). Census data are only collected once a decade, and geographic units based on census data (e.g., census-tract, census block group) may not match residents’ perceptions or experiences of neighborhood boundaries. Self-reported perceptions of neighborhood environment may result in flawed inferences due to reporting bias (7).

This study addresses this limitation in previous research by using an objective measure of neighborhood disorder based on indicators from a valid and reliable neighborhood instrument (8,9) that was administered in the same year as the arrest data. In our previous work, we found that objective measures of neighborhood disorder are associated with precursors to juvenile drug arrests (JDAs), including marijuana use, risk-taking propensity, and internalizing problems (10-12). Furr-Holden et al. (10) found that the growth of neighborhood disorder, as measured by the presence of abandoned buildings, predicted subsequent drug use in a sample of predominately African-American young adults. Another study found that neighborhood disorder was associated with laboratory measures of risk-taking propensity during late childhood (11). However, it is unclear whether neighborhood disorder is only associated with drug use and general risk-taking or weather it is also associated with drug arrests. This is a critical gap in the literature because drug use is not necessarily associated with arrests for drugs. For example, African-American youths are far more likely to be arrested for marijuana offenses than Caucasian youths, despite similar rates of marijuana use among African-American youths compared with Caucasians (13,14). This investigation will examine the relationship between objectively measured neighborhood disorder and neighborhood counts of JDAs in a predominately African-American city.

The descriptive statistics are included in Table 1. The mean population for the neighborhoods included in this investigation was 2,662 and ranged from 168 to 16,792. The neighborhoods were predominately African American (64.9%), which is consistent with the racial composition of the entire city (63.7%African American). The majority of the drug arrests were possession charges (38.4%). Possession with intent accounted for 35.2% of the drug arrests and distribution accounted for 23.4% of the drug arrests. The aggregated neighborhood disorder scale ranged from 0 to 7.8.

There was a positive relationship between neighborhood disorder and the count of JDAs (Table 2; β = 0.34, p<0.001); for each unit increase in neighborhood disorder, juvenile drug arrests increased by a factor of 40% (e^β=1.40). When the outcome was stratified by charge, the estimates for distribution and possession with intent to distribute were the same (β = 0.40, p<0.001; e^β=1.50). For each unit increase in neighborhood disorder, juvenile drug arrests for possession increased by a factor of 29% (β = 0.26, p<0.001; e^β=1.29). Other aggregate measures of neighborhood disorder (i.e. median and maximum) were used as predictors and the results were positive and statistically significant. The sample of 919 blocks used for the neighborhood disorder measure included randomly selected block faces (n = 447) and block faces of the residents of a sample of young adults (n = 472). There was stronger likelihood of a juvenile drug arrest on a block where we know a young adult lives. To determine whether this influenced our results, we examined the relationship between neighborhood disorder and JDAs using only the randomly selected blocks and the results were similar (β = 0.32, p<0.001 for randomly selected blocks; β = 0.34, p<0.001 for all blocks).

Indicators of socio-economic status were unavailable at the neighborhood-level used in this study, but we were able to control for the percentage of African American residents. The percentage of African American residents by neighborhood was strongly associated with count of JDAs (β = 2.10, p<0.001), which is consistent with earlier investigations (3). When the JDAs were stratified by charge type, the relationship between the percentage of African Americans and distribution charges had the strongest association (β = 2.72, p<0.001) compared to possession and possession with intent to distribute. Adjusting for the percentage of African Americans attenuated the relationship between neighborhood disorder and JDAs (Table 3). While holding the percentage of African American residents constant, juvenile drug arrests increased by a factor 27% for each unit increase in neighborhood disorder (β = 0.24, p<0.001; e^β=1.27). The percentage of African American residents in the neighborhood also reduced the strength of the association between neighborhood disorder and JDAs when stratified by charge type although the regression model remains statistically significant.

Ecological and epidemiological models highlight the significance of neighborhood context to health and well-being (20); further, geographic methods support the creation of neighborhoodbased measures to capture the spatial and structural nature of risk environments. In this preliminary investigation, there was a positive and significant relationship between the count of JDAs and neighborhood disorder. These findings are consistent with earlier studies that found that concentrated poverty and observable physical decay at the community level might explain variation in levels of arrests and specific involvement in drug/alcohol-related activities (10,21,22). The strength of this study is its use of geographic methods and a reliable, valid and objective environmental assessment tool of disorder to examine associations of disorder and drug-related arrests in an urban setting. This represents a noteworthy advancement over previous research that relied on economic and structural factors from census data, rarely providing opportunities for targeted interventions or neighborhood improvement. Although prior research has shown associations between disorder and crime, using a validated measure of neighborhood disorder provides initial steps in understanding the aspects of disorder amenable to interventions. This preliminary investigation supports a growing body of research exploring the association between neighborhood-level disorder and drug-related arrests and helps to better align the conceptualizations of neighborhood-level disorder and drug activity, thereby enhancing intervention planning.

This work should be interpreted in view of several limitations. First, we examine data only at one point in time, so causality could not be determined. For example, longitudinal studies are needed in determining whether neighborhood disorder leads to drug arrests or whether both drug arrests and neighborhood disorder are better explained by a third variable, such as drug activity, in the neighborhood. Second, juveniles who recidivate will inflate the JDA count for their neighborhood. Third, there was no measure of socioeconomic status of the residents or of the JDA youth. However, we were able to control for percent African-American living in the neighborhood, a sociodemographic variable previously found to be related to drug arrests (3,13). Despite these limitations, this study provides preliminary evidence of the co-occurrence of community disorder and youth drug arrests. Additionally, we found that increases in the percentage of African-Americans in the neighborhood were associated with increased community disorder and youth drug arrests. We are cautious in noting the racial findings observed in this investigation as racial segregation is often confounded by poverty. Future investigations should examine the interplay between racial segregation, concentrated poverty, and neighborhood disorder, as racial segregation of African-Americans is often associated with concentrated poverty, which is a leading indicator of crime and enforcement activities of police and may account for the increased rates of JDAs. Approaches to reduce drug-related crime by addressing the larger neighborhood and social context in which drug involvement and crime occurs warrant further consideration.