Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2016 Aug 1.
Published in final edited form as: Drug Alcohol Depend. 2015 May 27;153:180–186. doi: 10.1016/j.drugalcdep.2015.05.022

A Multi-level Analysis of the Impact of Neighborhood Structural and Social Factors on Adolescent Substance Use

Abigail A Fagan 1,✉, Emily M Wright 2, Gillian M Pinchevsky 3
PMCID: PMC4510008  NIHMSID: NIHMS697400  PMID: 26049206

Abstract

Background

This paper examined the effects of neighborhood structural (i.e., economic disadvantage, immigrant concentration, residential stability) and social (e.g., collective efficacy, social network interactions, intolerance of drug use, legal cynicism) factors on the likelihood of any adolescent tobacco, alcohol, and marijuana use.

Methods

Analyses drew upon information from the Project on Human Development in Chicago Neighborhoods (PHDCN). Data were obtained from a survey of adult residents of 79 Chicago neighborhoods, two waves of interviews with 1,657 to 1,664 care-givers and youth aged 8 to 16 years, and information from the 1990 U.S. Census Bureau. Hierarchical Bernoulli regression models estimated the impact of neighborhood factors on substance use controlling for individual-level demographic characteristics and psycho-social risk factors.

Results

Few neighborhood factors had statistically significant direct effects on adolescent tobacco, alcohol or marijuana use, although youth living in neighborhoods with greater levels of immigrant concentration were less likely to report any drinking.

Conclusion

Additional theorizing and more empirical research are needed to better understand the ways in which contextual influences affect adolescent substance use and delinquency.

Keywords: adolescent substance use, neighborhood context, multi-level analysis

1. INTRODUCTION

Adolescent substance use is a public health concern (National Academy of Sciences, 2004). In the U.S., 22% to 35% of high school students report current use of tobacco, alcohol, and marijuana (Kann et al., 2014) and lifetime drug use is even higher (Johnston et al., 2013). Moreover, worldwide estimates of substance use disorders and dependency range from 6% to 16% among adolescents (National Research Council and Institute of Medicine, 2009). These rates are concerning given the immediate and long-term consequences of substance use on public health problems including drug abuse, crime and violence, and physical and mental illness (Hingson et al., 2006; Mrug and Windle, 2009).

The extent and consequences of illegal substance use by adolescents has led to calls for more preventative interventions to reduce use (National Prevention Council, 2011; National Research Council and Institute of Medicine, 2009). Doing so requires a full understanding of the circumstances that place adolescents at risk for substance use. Research has indicated that adolescents’ individual characteristics, peer groups, families, and schools affect their likelihood of smoking, drinking, and illicit drug use (Durlak, 1998; Hawkins et al., 1992). There is also evidence that rates of substance use vary significantly across neighborhood contexts (Bernat et al., 2009; Karriker-Jaffe, 2011; Wilcox, 2003), but the specific structural and social factors which contribute to this variation have not yet been clearly identified (Leventhal and Brooks-Gunn, 2000).

1.1 Contextual influences on adolescent substance use

Social disorganization theories (Sampson et al., 1997; Shaw and McKay, 1942) posit that areas of economic and social deprivation will have more delinquency and crime than affluent and socially organized neighborhoods. Social ecological theories (e.g., Bronfenbrenner, 1979) also emphasize the role of the neighborhood context in shaping development and also recognize the importance of other social influences, such as peer interactions and family processes. Guided by these theories, studies have increasingly examined the impact of neighborhood context on adolescent development (e.g., Ennett et al., 2008; Karriker-Jaffe et al., 2013; Lambert et al., 2004; Sampson, 2012; Tobler et al., 2009; Zimmerman and Messner, 2011).

In contrast to the predictions of social disorganization and social ecological theories, however, much of this literature has failed to show a direct effect of neighborhood factors on substance use, and when significant effects have been evidenced, they have been mixed across studies. For example, a review of 34 studies found that 18% reported a negative relationship between community socioeconomic status (SES) and alcohol use by adolescents and young adults (i.e., drinking was more likely in low-SES areas), 14% of studies showed the opposite effect (i.e., drinking was greater in high-SES areas), and the remainder (68%) did not find a significant relationship (Karriker-Jaffe, 2011). Similarly mixed findings are reported in other systematic reviews of contextual influences on adolescent drinking (Bryden et al., 2013; Hanson and Chen, 2007; Jackson et al., 2014). Studies have shown more consistent direct effects of community SES on smoking, with most showing higher rates of tobacco use in lower-SES communities, but some research has indicated the opposite relationship or a lack of significant effects (Gardner et al., 2010; Hanson and Chen, 2007). Investigations of neighborhood structural factors on marijuana use are too few to draw strong conclusions (Gardner et al., 2010).

Far fewer studies have investigated the impact of neighborhood social processes on substance use. One review (Jackson et al., 2014) found only three studies that analyzed the impact of community attitudes regarding substance use on adolescent drinking, three studies assessing collective efficacy (i.e., social cohesion and efforts to informally control crime or deviance; see Sampson et al., 1997), and five studies examining social capital or neighborhood attachment. Most of these studies indicated null or contradictory effects of social processes on adolescent drug use (e.g., De Haan and Boljevac, 2010; Ennett et al., 2008, 2010; Musick et al., 2008).

1.2 Limitations and gaps in prior research

The reviews cited above concur that more information is needed to better understand if and how contextual influences affect adolescent substance use (Bryden et al., 2013; Jackson et al., 2014; Karriker-Jaffe, 2011). Available literature has focused more on structural factors like SES than social processes such as social capital, collective efficacy, and community norms (Bryden et al., 2013; Jackson et al., 2014). In addition, most research has evaluated either tobacco or alcohol use, with fewer studies evaluating other substances or comparing effects across different substances (Karriker-Jaffe, 2011).

More methodologically rigorous examinations are also needed. Leventhal and Brooks-Gunn (2000) note that relatively few studies have been specifically designed to study neighborhood effects and that few have collected reliable and valid data on neighborhood processes from enough geographical areas and respondents per neighborhood to ensure sufficient variability in constructs and adequate statistical power to find effects. In addition, neighborhood constructs, especially those intended to measure social processes, are typically measured using data from the same adolescents who report on substance use (Wilcox, 2003). However, relying on the same sources to report independent and dependent variables can inflate effect sizes. In addition, individuals’ perceptions of their neighborhood environments are likely influenced by their own experiences and/or psychological characteristics and may not represent actual neighborhood conditions (Sampson and Raudenbush, 1999). To avoid bias, neighborhood characteristics are ideally measured with objective sources like archival data (e.g., U.S. Census data), systematic observations, or surveys of community leaders (Leventhal and Brooks-Gunn, 2000; Sampson and Raudenbush, 1999). It is also important to ensure that “neighborhoods” represent meaningful ecological contexts, but the common use of administrative data (e.g., census tracts) to define neighborhood boundaries may not produce areas that match residents’ views of their neighborhoods (Sampson, 2012).

Neighborhood studies often fail to investigate the impact of community and individual-level factors and to utilize multi-level analyses when doing so. Based on social ecological theories and research indicating that many individual, peer, and family factors influence adolescent substance use (Hawkins et al., 1992), failure to control for these variables could artificially inflate neighborhood effects. (It is also true that analyses which include factors that mediate the effects of neighborhood context on substance use may under-estimate neighborhood direct effects. To avoid this problem in our analyses, our first set of multi-level multivariate analyses includes a limited number of individual-level controls.) Model mis-specification can also occur if factors which affect neighborhood selection (e.g., individual SES; Gardner et al., 2010) are not included. Multi-level statistical techniques should be used when simultaneously investigating the impact of community- and individual-level factors in order to minimize correlated error and heteroskedasticity and to avoid biased hypotheses testing (Sampson and Raudenbush, 1999). Finally, many studies have relied on cross-sectional data which limits causal inferences regarding the impact of contextual influences on substance use (Jackson et al., 2014; Leventhal and Brooks-Gunn, 2000).

1.3. The current study

The current study seeks to address these issues and advance our understanding of how neighborhood context affects adolescent substance use. Analyses draw on data from the Project on Human Development in Chicago Neighborhoods (PHDCN; Earls et al., 2002), a study purposefully designed to examine contextual effects on youth development. We examine the impact of structural and social neighborhood constructs on the three substances most commonly used during adolescence---tobacco, alcohol and marijuana---controlling for many individual-level predictors and using prospective measures from multiple informants. Two research questions are examined: 1) To what extent does adolescent substance use (tobacco, alcohol, and marijuana) vary by neighborhood? 2) What are the direct effects of neighborhood structural and social characteristics on adolescent substance use, controlling for individual-level factors?

2. MATERIAL AND METHODS

2.1 Participants

We draw on data from three sources of data collected in the PHDCN. The first is the Community Survey of adult residents of Chicago neighborhoods. To obtain reliable estimates of neighborhood processes across the city, Chicago’s 847 census tracts were divided into 343 neighborhood clusters (NCs) based on knowledge of existing neighborhoods and geographic boundaries and to ensure homogenous units of analysis (Sampson, 2012). Using a three-stage sampling design, city blocks were then sampled within each NC, dwelling units were sampled within blocks, and one adult resident was sampled within each dwelling unit and interviewed in 1994–1995 regarding neighborhood social processes. To assess neighborhood structural characteristics, data from the 1990 U.S. Census were collected and linked to the 343 NCs.

To examine the impact of neighborhood characteristics on youth outcomes, the 343 NCs were stratified by seven categories of racial/ethnic diversity and three levels of SES, and 80 NCs were selected via stratified probability sampling. Households within these areas with at least one child in one of seven age cohorts (newborns and children ages 3, 6, 9, 12, 15, and 18) were eligible to participate in the Longitudinal Cohort Study. In 1994–1997, wave one interviewers were conducted with 6,228 children and caregivers. Wave two interviews were conducted in 1997–2000.

Because this study focuses on adolescent substance use, participants included youth from Cohorts 9, 12, and 15 who resided in 79 of the 80 NCs (one NC was not included due to missing data from respondents in these cohorts) and who provided data at waves one and two. Wave one included 2,344 youth, while 1,987 youth (85% of the original sample) participated at wave two. After listwise deletion, the analysis samples included 1,657–1,664 youth across the three outcomes. (A comparison of all youth in Cohorts 9–15 at wave one (N=2,344) and the analyses samples showed no significant differences on the primary independent or dependent variables. However, the analyses samples had significantly (p ≤ .05) more Hispanic youth and higher family income compared to the initial sample.) At wave one, youth were a mean age of 12 years, about half were male, 48% reported their race/ethnicity as Hispanic, 34% African American, 14% Caucasian, and 4% as another race/ethnicity (See table 1).

Table 1.

Individual-Level Models Predicting Any Adolescent Substance Usea

Cigarette Use (N=1664)
Alcohol Use (N=1657)
Marijuana Use (N=1660)
b SE b SE b SE
Intercept −2.28** .11 −2.17** .11 −4.40** .21
 Age .21** .05 .38** .05 .51** .06
 Male −.06 .16 −.13 .17 .46 .28
 Socioeconomic status .16* .08 .03 .09 −.04 .14
 African Americanb −.78** .26 −.65** .22 .14 .32
 Hispanicb .22 .27 .13 .24 .08 .38
 Other race/ethnicityb −.58 .47 −.36 .48 −.57 .64
 Immigrant −.25 .27 −.10 .30 −.94** .34
 Low self-control .16* .08 .06 .08 .05 .12
 Peer social support .30** .10 .16 .10 .38** .11
 Family social support −.24** .09 .03 .09 −.23* .11
 Curfew .08 .14 .20 .16 .14 .15
 Routine activities .27** .09 .56** .09 .30** .11
 Perceived harm of drugs −.37** .08 −.28** .09 −.57** .10
 Perceived availability of drugs .04 .12 .27* .12 .57** .15
 Peer substance use .97** .09 .95** .12 .97** .11
χ2 70.22 80.02 92.78
Open in a new tab
**

p ≤ .01

*

p ≤ .05

a

Results are based on Bernoulli models and fixed effects for all variables

b

Compared to Caucasian youth

2.2 Measures

2.2.1 Adolescent substance use

Substance use was measured at wave two based on adolescent reports on three items rating the frequency of tobacco, alcohol, and marijuana use in the past year using a nine-point scale ranging from no use to 200 or more times (National Household Survey on Drug Abuse, 1991). Because substance use was relatively low and few respondents reported frequent use, we created dichotomous outcomes reflecting any tobacco use, any alcohol use and any marijuana use in the past year, which distinguished users (19%, 23%, and 11%, respectively) from non-users.

2.2.2 Neighborhood characteristics

The neighborhood structural and social characteristics were based on information from the same 79 NCs in which youth resided. A principal components factor analysis of items from the 1990 U.S. Census was conducted to create the three neighborhood structural variables. As in prior research (Browning et al., 2005; Maimon and Browning, 2010), neighborhood economic disadvantage included four poverty-related variables (alpha=0.88): the percentage of residents in the NC below the poverty line, receiving public assistance, unemployed, and in female-headed households. Immigrant concentration included two items (alpha=0.70): the percentage of foreign-born and Hispanic residents (Maimon and Browning, 2010; Morenoff et al., 2001). Residential stability was based on two items (alpha=0.76): the percentage of owner-occupied homes and those living in the same home for five years (Morenoff et al., 2001; Sampson et al., 2005).

Four constructs representing neighborhood social processes were created from responses from approximately 40 adults per NC participating in the Community Survey. Neighborhood collective efficacy was based on 10 items (internal consistency=0.85) representing social cohesion and informal social control. Residents rated their agreement with five items assessing trust and support between neighbors (e.g., people around here are willing to help their neighbors) using a five-point Likert scale. Another five items asked about the likelihood (on a five-point scale) that residents would utilize informal social control to help keep the neighborhoods safe (e.g., neighbors would intervene if children were skipping school and hanging out). Following prior research (Browning et al., 2004; Sampson et al., 1997), the ten items were combined using a three-level item response model (IRM), which helps avoid the loss of data from missing item responses (Osgood et al., 2002), and accounts for item severities and respondent characteristics (e.g., sex or race/ethnicity) as covariates (Raudenbush and Sampson, 1999).

The other three social process measures were also created using IRMs. Social network interaction was based on four items (internal consistency=0.73) asking residents to rate how often neighbors do favors for each other, ask for advice, have get-togethers, and visit each other (Browning et al., 2004). Each was rated on a four-point scale from “never” to “often.” Community intolerance of drug use was based on residents’ reports of how wrong they consider teenage smoking, drinking and marijuana use (internal consistency=0.50) using a five-point scale for each of the three items (from “not wrong at all” to “extremely wrong”) (Kirk and Papachristos, 2011; Wright and Fagan, 2013). Legal cynicism was based on levels of agreement (rated on a five-point scale from “strongly disagree” to “strongly agree”) to five statements (internal consistency=0.48) regarding the legitimacy of laws and social norms, such as “laws were meant to be broken” (Sampson and Bartusch, 1998).

2.2.3 Control variables

Statistical models controlled for individual-level factors shown to be associated with adolescent substance use (Donovan, 2004; Hawkins et al., 1992). Control variables were taken from wave one surveys when available; otherwise, wave two reports were used. Demographic characteristics included adolescents’ age, sex, race/ethnicity, immigrant status, and socio-economic status. SES was a factor score based on caregivers’ responses to three items regarding personal or household income, the highest educational level of either parent, and employment status of the primary caregiver. Measures of peer social support (alpha=0.70; e.g., “I have at least one friend I can tell anything to”) and family social support (alpha=0.67; e.g., “No matter what happens, my family will be here for me”) were based on youth responses to nine and six items, respectively (Turner et al., 1983). Youth low self-control was based on 17 responses (alpha=0.75) from caregivers rating children’s impulsivity, decision-making, and sensation-seeking (Buss and Plomin, 1975; Gibson et al., 2010). At wave two, youth reported on their involvement in unstructured/routine activities (4 items, alpha=0.58; e.g., “going to parties”; Osgood et al., 1996), perceptions that drug use is harmful (7 items, alpha=0.76), perceived availability of drugs (3 items, alpha=0.87; e.g., “how easy would it be to get alcohol”), and exposure to peer substance use (4 items, alpha=0.85; e.g., number of friends who use marijuana). These variables were created by standardizing and summing responses to all items. Youth supervision was the sum of three dichotomous items (alpha=0.60) asking caregivers if children had a curfew.

2.3 Statistical Analyses

The statistical analyses relied on hierarchical modeling techniques using HLM 7.0 software (Raudenbush and Bryk, 2014). Hierarchical Bernoulli regression models, analogous to logistic regression models, were utilized to predict the three dichotomous outcomes. Tolerance values were all above 0.40, suggesting that multicollinearity was not a problem (Allison, 1999).

The analyses proceeded in a step-wise fashion. First, unconditional models were estimated to examine if the distribution of substance use varied significantly across NCs (Research Question 1). Next, all individual-level predictors were grand-mean centered and fixed to remove within-NC variation potentially related to substance use and to aid in the interpretation of coefficients. (When conducting the individual-level analyses, the reliability of the intercepts was reduced. To adjust for this situation, the Empirical Bayes estimates were modeled at level-two; Raudenbush and Bryk, 2002, 2014.) The neighborhood characteristics were then added to the models to assess their direct effects on cigarette, alcohol, and marijuana use (Research Question 2). The first set of models analyzed the relationship between each neighborhood variable and each outcome, without the other neighborhood-level variables in the model. These analyses first controlled for child demographic characteristics and low self-control, then all other individual-level variables were added. Because neighborhood social processes have been shown to mediate the effects of structural characteristics (e.g., Sampson et al., 1997), the second set of models estimated the relationship between each of the four social characteristics and each drug use outcome, controlling for the three structural characteristics and all individual-level variables.

3. RESULTS

The unconditional models showed significant (p<.05) variation in all three substance use outcomes across NCs (results not shown). The intra-class correlations (ICCs) for tobacco, alcohol and marijuana use were 0.037, 0.047, and 0.024, respectively, indicating that 3.7%, 4.7%, and 2.4% of the variation in substance use existed at the neighborhood-level.

Before examining the neighborhood factors that might account for this variation, we assessed the relationship between the individual-level control variables and each outcome. The findings, shown in Table 1, are largely consistent with prior research. Individual and peer factors were the most robust and consistent predictors of use. A greater likelihood of substance use was reported by older and Caucasian adolescents compared to younger and African American youth. Use was also more likely among those who spent more time engaging in routine, unstructured activities (without adult supervision), reported that drugs were available to them, and had more substance-using peers. Adolescents who perceived drug use to be harmful were less likely to report any tobacco, alcohol or marijuana use.

The direct effects of the neighborhood characteristics on substance use are shown in Tables 2 and 3. As shown in Table 2, Model 1, none of the neighborhood factors had a significant (p<.05) direct effect on adolescent substance use when controlling for youth demographic characteristics and low self-control. Similarly, as shown in Model 2, no neighborhood variables had a significant effect on substance use when controlling for all individual-level variables.

Table 2.

The Impact of Each Neighborhood Factor on Any Substance Use, with Individual-Level Controlsa

Model 1 (Demographic Controls) Model 2 (All Controls)
Cigarette Use (N=1664)
b SE b SE
Economic Disadvantage −.01 .01 .00 .01
Immigrant Concentration −.01 .01 −.01 .01
Residential Stability .00 .01 .00 .01
Collective Efficacy .12┼ .06 .04┼ .03
Social Network Interaction .12 .08 .04 .03
Intolerance of Drug Use −.24 .17 −.07 .07
Legal Cynicism .11 .21 .02 .09
Alcohol Use (N=1657)
b SE b SE
Economic Disadvantage −.00 .02 −.00 .01
Immigrant Concentration −.01 .02 −.01┼ .01
Residential Stability −.01 .02 .00 .01
Collective Efficacy .04 .07 .02 .03
Social Network Interaction .01 .09 −.02 .04
Intolerance of Drug Use −.03 .19 −.00 .07
Legal Cynicism −.01 .22 −.01 .09
Marijuana Use (N=1660)
b SE b SE
Economic Disadvantage −.01 .01 −.02 .03
Immigrant Concentration −.01 .01 −.02 .03
Residential Stability −.01 .01 −.02 .03
Collective Efficacy .02 .06 −.02 .12
Social Network Interaction .05 .08 .00 .16
Intolerance of Drug Use −.07 .17 .02 .33
Legal Cynicism −.05 .21 −.23 .40
Open in a new tab
**

p ≤ .01

*

p ≤ .05

┼

≤ .10

a

Empirical Bayes analyses predicting any cigarette, alcohol and marijuana use among individuals living in 79 NCs. Coefficients reflect the impact of each neighborhood variable on substance use without the other level-two variables in the model. Model 1 controls for demographic characteristics (age, sex, African American, Hispanic, other race/ethnicity, immigrant status, socioeconomic status) and low self-control; Model 2 includes all individual-level controls shown in Table 1.

Table 3.

The Relationship Between Neighborhood Structural and Social Factors and Any Substance Use, with Individual-Level Controlsa

Model 3 Model 4 Model 5 Model 6
Cigarette Use (N=1664)
b SE b SE b SE b SE
Economic Disadvantage .01 .01 −.00 .01 −.00 .01 .00 .01
Immigrant Concentration −.00 .01 −.00 .01 −.00 .01 −.01 .01
Residential Stability −.00 .01 .00 .01 .00 .01 .00 .01
Collective Efficacy .06 .04 -- -- -- -- -- --
Social Network Interaction -- -- .03 .04 -- -- -- --
Intolerance of Drug Use -- -- -- -- −.08 .08 -- --
Legal Cynicism -- -- -- -- -- -- .00 .09
Alcohol Use (N=1657)
b SE b SE b SE b SE
Economic Disadvantage −.01 .01 −.00 .01 −.00 .01 −.01 .01
Immigrant Concentration −.01┼ .01 −.02* .01 −.01* .01 −.01┼ .01
Residential Stability −.00 .01 −.00 .01 −.00 .01 −.00 .01
Collective Efficacy −.02 .04 -- -- -- -- -- --
Social Network Interaction -- -- −.05 .04 -- -- -- --
Intolerance of Drug Use -- -- -- -- .04 .08 -- --
Legal Cynicism -- -- -- -- -- -- −.01 .09
Marijuana Use (N=1660)
b SE b SE b SE b SE
Economic Disadvantage −.06 .03 −.04 .03 −.04 .03 −.04 .03
Immigrant Concentration −.05 .03 −.04 .03 −.04 .03 −.04 .03
Residential Stability −.03 .03 −.04 .03 −.05 .03 −.04 .03
Collective Efficacy −.16 .18 -- -- -- -- -- --
Social Network Interaction -- -- .01 .17 -- -- -- --
Intolerance of Drug Use -- -- -- -- .16 .37 -- --
Legal Cynicism -- -- -- -- -- -- −.08 .42
Open in a new tab
**

p ≤ .01

*

p ≤ .05

┼

≤ .10

a

Empirical Bayes analyses predicting any cigarette, alcohol and marijuana use among individuals living in 79 NCs. Coefficients reflect the impact of each neighborhood social processes variable controlling for all other neighborhood structural variables on substance use. Models also control for all individual-level controls shown in Table 1.

Table 3 presents the results of models in which all the neighborhood structural variables were assessed simultaneously, along with each of the social variables (entered one by one) and controlling for all the individual-level variables shown in Table 1. These results also demonstrated a lack of significant, direct effects of neighborhood factors on adolescent substance use. The only significant (p<.05) effect indicated that youth from NCs with a higher percentage of residents from immigrant backgrounds were less likely to report alcohol use compared to those in NCs with fewer immigrant residents.

4. DISCUSSION

This study utilized a rigorous research design, prospective data, and multi-level analysis to investigate the direct effects of neighborhood factors on adolescent tobacco, alcohol, and marijuana use. In models controlling for individual-level risk factors, and even in scaled-down models with only demographic characteristics and self-control as controls, direct neighborhood effects were not significant. The only exception was that, in some of the models predicting alcohol use, youth living in areas with higher levels of immigrant concentration were significantly less likely to report drinking in the past year compared to those living in areas with lower immigrant concentration.

These findings are somewhat surprising, as they do not support social ecological (e.g., Bronfenbrenner, 1979) or social disorganization theories (Sampson et al., 1997; Shaw and McKay, 1942), both of which suggest that neighborhood characteristics will have direct effects on youth delinquency. However, the results are consistent with several reviews, which indicate a lack of significant effects in the majority of studies investigating neighborhood predictors of adolescent substance use (Bryden et al., 2013; Jackson et al., 2014; Karriker-Jaffe, 2011). Those reviews all emphasize the need for additional investigation of contextual influences on adolescent substance use, especially studies that measure social processes. Our study sought to fill this gap and provide a comprehensive analysis of the direct effects of a range of structural and social factors on the three most commonly used substances using multi-level analyses and controlling for many individual risk factors for substance use.

The general pattern of results found in the current study and in much past research, that neighborhoods do not have direct, robust effects on adolescent substance use, suggests the need to more carefully consider how neighborhoods influence various forms of delinquency (Jackson et al., 2014). Most social disorganization theories were developed to explain youth violent or property crimes, not substance use, and the role of the community may differ for these behaviors. Unlike predatory and economic crimes, drinking and drug use may be considered less harmful and more of a “rite of passage” for adolescents, which could result in less adult condemnation and regulation (Ennett et al., 2010; Foley et al., 2004). If this is the case, substance use will likely be unaffected by social norms and controls. In addition, drug use often occurs indoors (Maimon and Browning, 2012), making it difficult for adult residents to influence. Lastly, although drug use has been hypothesized as a coping mechanism used to alleviate the stressful conditions of disorganized neighborhoods (Lambert et al., 2004), teenagers often engage in smoking and drinking because they are fun and social activities (Kuntsche et al., 2005). As such, these actions may not be affected by adverse neighborhood conditions.

More nuanced theories that can better specify the ways in which neighborhood context affects youth development and substance use are needed. Such perspective can help direct future research and inform the development of environmentally-focused substance use/abuse prevention programs. Far fewer of these models have been created, tested, and/or shown to affect substance use compared to interventions which target individual-level factors, despite the potential of neighborhood-level interventions to reach large numbers of youth and to reinforce messages communicated in other types of interventions (Fagan and Hawkins, 2012).

Although the current study failed to identify significant direct effects of certain contextual variables, the fact that the unconditional models indicated significant variation in substance use across NCs suggests that some important neighborhood factors may have been omitted. In fact, analyses did not include all possible contextual risk factors for substance use, such as physical or social disorder (e.g., Furr-Holden et al., 2011), tobacco and alcohol outlet density (e.g., Maimon and Browning, 2012; Tobler et al., 2009) and neighborhood levels of crime (e.g., Mrug and Windle, 2009). In addition, analyses did not examine the degree to which neighborhood factors moderated the impact of individual, peer and family influences on adolescent substance use, although other studies have shown such results (Fagan et al., 2014; Lo et al., 2006; Snedker et al., 2009; Wright et al., in press; Zimmerman and Vasquez, 2011). Likewise, we did not assess if neighborhood processes were mediated by influences in other contexts, as also demonstrated in past research (Chuang et al., 2005; Gibbons et al., 2004; Tobler et al., 2009). The current study focused on direct effects given the inconclusive evidence to date that neighborhoods exert such influences (particularly when social processes are examined), but we acknowledge that a more complete understanding of the role of neighborhood context requires specification of both mediating and moderating effects (Leventhal and Brooks-Gunn, 2000). A final limitation of our study is that the outcome variables were dichotomized to represent any substance use and did not capture more frequent or problematic levels of use. It is possible that a different pattern of results would have been evidenced if more serious forms of substance use were assessed and we hope that future research will test this possibility.

Highlights.

  • Neighborhood factors are expected to influence youth substance use

  • This study and others show few direct effects of neighborhood factors on drug use

  • More nuanced theories are needed to explain how neighborhoods affect substance use

Acknowledgments

Role of Funding Source. Funding for this study was provided by Grant R01DA30387-01 from the National Institute of Drug Abuse (NIDA). NIDA had no role in the study design, collection, analysis or interpretation of the data, writing the manuscript, or the decision to submit the paper for publication.

The data used in this study were made available by the Inter-University Consortium for Political and Social Research (ICPSR). The Consortium does not bear any responsibility for the analyses or conclusions presented here.

Footnotes

Conflicts of Interest

The authors have no conflicts of interest to declare.

Contributors

Each of the authors has made substantial contributions to the manuscript. AF and EW jointly conceived of the study’s research questions and the design of the study; AF took the lead role in drafting the paper; EW created the dataset to be analyzed in the study and managed the statistical analyses; GP took the lead role in conducting the statistical analyses; and EW and GP helped to draft the manuscript. All authors read and approved the final manuscript.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Contributor Information

Abigail A. Fagan, Email: afagan@ufl.edu, Department of Sociology and Criminology & Law, University of Florida, 3219 Turlington Hall, P.O. Box 117330, Gainesville FL 32611-7330 United States, Phone: (352) 294-7169; Fax: (352) 392-6568

Emily M. Wright, Email: EMWright@unomaha.edu, School of Criminology and Criminal Justice, University of Nebraska, Omaha, 6001 Dodge Street, 218 CPACS, Omaha, NE 68182-0149 United States

Gillian M. Pinchevsky, Email: gillian.pinchevsky@unlv.edu, Department of Criminal Justice, University of Nevada, Las Vegas, 4505 South Maryland Parkway, Box 455009, Las Vegas, Nevada, 89154 United States

References

  1. Allison PD. Multiple Regression: A Primer. Pine Forge Press; Thousand Oaks, CA: 1999. [Google Scholar]
  2. Bernat DH, Lazovich D, Forster JL, Oakes JM, Chen V. Area-level variation in adolescent smoking. Public Health Res Pract Policy. 2009;6:1–8. [PMC free article] [PubMed] [Google Scholar]
  3. Bronfenbrenner U. The Ecology Of Human Development. Harvard University Press; Cambridge, MA: 1979. [Google Scholar]
  4. Browning CR, Feinberg SL, Dietz RD. The paradox of social organization: networks, collective efficacy and violent crime in urban neighborhoods. Social Forces. 2004;83:503–534. [Google Scholar]
  5. Browning CR, Leventhal T, Brooks-Gunn J. Sexual initiation in early adolescence. Am Sociol Rev. 2005;70:758–778. [Google Scholar]
  6. Bryden A, Roberts B, Petticrew M, McKee M. A systematic review of the influence of community level social factors on alcohol use. Health Place. 2013;21:70–85. doi: 10.1016/j.healthplace.2013.01.012. [DOI] [PubMed] [Google Scholar]
  7. Buss A, Plomin R. A Temperament Theory Of Personality Development. Wiley; New York, NY: 1975. [Google Scholar]
  8. Chuang YC, Ennett ST, Bauman KE, Foshee VA. Neighborhood influences on adolescent cigarette use and alcohol use: Mediating effects through parent and peer behaviors. J Health Soc Behav. 2005;46:187–204. doi: 10.1177/002214650504600205. [DOI] [PubMed] [Google Scholar]
  9. De Haan L, Boljevac T. Alcohol prevalence and attitudes among adults and adolescents: Their relation to early adolescent alcohol use in rural communities. J Child Adolesc Subst Abuse. 2010;19:223–243. doi: 10.1080/1067828X.2010.488960. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Donovan JE. Adolescent alcohol initiation: A review of psychosocial risk factors. J Adolesc Health. 2004;35:e7–e18. doi: 10.1016/j.jadohealth.2004.02.003. [DOI] [PubMed] [Google Scholar]
  11. Durlak JA. Common risk and protective factors in successful prevention programs. Am J Orthopsychiatr. 1998;68:512–520. doi: 10.1037/h0080360. [DOI] [PubMed] [Google Scholar]
  12. Earls FJ, Brooks-Gunn J, Raudenbush SW, Sampson RJ. Project on Human Development in Chicago Neighborhoods (PHDCN) Inter-university Consortium for Political and Social Research; Ann Arbor, MI: 2002. [Google Scholar]
  13. Ennett ST, Foshee VA, Bauman KE, Hussong A, Cai L, Reyes HL, Faris R, Hipp J, Durant R. The social ecology and adolescent alcohol misuse. Child Dev. 2008;79:1777–1791. doi: 10.1111/j.1467-8624.2008.01225.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Ennett ST, Foshee VA, Bauman KE, Hussong A, Faris R, Hipp J, Cai L. A social contextual analysis of youth cigarette smoking development. Nicotine Tob Res. 2010;12:950–962. doi: 10.1093/ntr/ntq122. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Fagan AA, Hawkins JD. Community-based substance use prevention. In: Welsh BC, Farrington DP, editors. The Oxford handbook On Crime Prevention. Oxford University Press; New York: 2012. pp. 247–268. [Google Scholar]
  16. Fagan AA, Wright EM, Pinchevsky GM. The protective effects of neighborhood collective efficacy on adolescent substance use and violence following exposure to violence. J Youth Adolesc. 2014;43:1498–1512. doi: 10.1007/s10964-013-0049-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Foley KL, Altman D, Durant R, Wolfson M. Adults’ approval and adolescents’ alcohol use. J Adolesc Health. 2004;34:345e.317–345e.326. [PubMed] [Google Scholar]
  18. Furr-Holden CDM, Hwa Lee M, Milam AJ, Johnson RM, Lee KS, Ialongo NS. The growth of neighborhood disorder and marijuana use among urban adolescents: a case for policy and environmental interventions. J Stud Alcohol Drugs. 2011;72:371–379. doi: 10.15288/jsad.2011.72.371. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Gardner M, Barajas RG, Brooks-Gunn J. Neighborhood influences on substance use etiology: is where you live important? In: Scheier LM, editor. Handbook Of Drug Use Etiology: Theory, Methods, And Empirical Findings. American Psychological Association; Washington, D.C: 2010. pp. 423–442. [Google Scholar]
  20. Gibbons FX, Gerrard M, Vande Lune LS, Ashby Wills T, Brody GH, Conger R. Context and cognitions: Environmental risk, social influence, and adolescent substance use. Personality and Social Psychology Bulletin. 2004;30:1048–1061. doi: 10.1177/0146167204264788. [DOI] [PubMed] [Google Scholar]
  21. Gibson CL, Sullivan CJ, Jones S, Piquero AR. “Does it take a village?”. Assessing neighborhood influences on children’s self control. J Res Crime Delinq. 2010;47:31–62. [Google Scholar]
  22. Hanson MD, Chen E. Socioeconomic status and health behaviors in adolescence: a review of the literature. J Behav Med. 2007;30:263–285. doi: 10.1007/s10865-007-9098-3. [DOI] [PubMed] [Google Scholar]
  23. Hawkins JD, Catalano RF, Miller JY. Risk and protective factors for alcohol and other drug problems in adolescence and early adulthood: implications for substance abuse prevention. Psychol Bull. 1992;112:64–105. doi: 10.1037/0033-2909.112.1.64. [DOI] [PubMed] [Google Scholar]
  24. Hingson RW, Heeren T, Winter MR. Age at drinking onset and alcohol dependence: age at onset, duration, and severity. Arch Pediatric Adolesc Med. 2006;160:739–746. doi: 10.1001/archpedi.160.7.739. [DOI] [PubMed] [Google Scholar]
  25. Jackson N, Denny S, Ameratunga S. Social and socio-demographic neighborhood effects on adolescent alcohol use: a systematic review of multi-level studies. Soc Sci Med. 2014;115:10–20. doi: 10.1016/j.socscimed.2014.06.004. [DOI] [PubMed] [Google Scholar]
  26. Johnston LD, O’Malley PM, Bachman JG, Schulenberg JE. Monitoring The Future National Results On Drug Use: 2012 Overview, Key Findings On Adolescent Drug Use. Institute for Social Research, University of Michigan; Ann Arbor, MI: 2013. [Google Scholar]
  27. Kann L, Kinchen S, Shanklin S, Flint KH, Hawkins J, Harris WA, Lowry R, Olsen EO, McManus T, Chyen D, Whittle L, Taylor E, Demissie Z, Brener N, Thornton J, Zaza S CDC. Youth Risk Behavior Surveillance - United States, 2013. Surveillance Summaries. 2014 Jun 13;:MMWR 63. [PubMed] [Google Scholar]
  28. Karriker-Jaffe KJ. Areas of disadvantage: a systematic review of effects of area-level socioeconomic status on substance use outcomes. Drug Alcohol Rev. 2011;30:84–95. doi: 10.1111/j.1465-3362.2010.00191.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Karriker-Jaffe KJ, Foshee VA, Ennett ST, Suchindran C. Associations of neighborhood and family factors with trajectories of physical and social aggression during adolescence. J Youth Adolesc. 2013;42:861–877. doi: 10.1007/s10964-012-9832-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Kirk DS, Papachristos AV. Cultural mechanisms and the persistence of neighborhood violence. Am J Sociol. 2011;116:1190–1233. doi: 10.1086/655754. [DOI] [PubMed] [Google Scholar]
  31. Kuntsche E, Knibbe R, Gmel G, Engles R. Why do young people drink? A review of drinking motives. Clin Psychol Rev. 2005;25:841–861. doi: 10.1016/j.cpr.2005.06.002. [DOI] [PubMed] [Google Scholar]
  32. Lambert SF, Brown TL, Phillips CM, Ialongo NS. The relationship between perceptions of neighborhood characteristics and substance use among urban African American adolescents. Am J Community Psychol. 2004;34:205–218. doi: 10.1007/s10464-004-7415-3. [DOI] [PubMed] [Google Scholar]
  33. Leventhal T, Brooks-Gunn J. The neighborhoods they live in: the effects of neighborhood residence on child and adolescent outcomes. Psychol Bull. 2000;126:309–337. doi: 10.1037/0033-2909.126.2.309. [DOI] [PubMed] [Google Scholar]
  34. Lo CC, Anderson AS, Minugh PA, Lomuto N. Protecting Alabama students from alcohol and drugs: a multi-level modeling approach. J Drug Issues. 2006;36:687–718. [Google Scholar]
  35. Maimon D, Browning CR. Underage drinking, alcohol sales and collective efficacy: Informal control and opportunity in the study of alcohol use. Soc Sci Res. 2012;41:977–990. doi: 10.1016/j.ssresearch.2012.01.009. [DOI] [PubMed] [Google Scholar]
  36. Maimon D, Browning CR. Unstructured socializing, collective efficacy, and violent behavior among urban youth. Criminology. 2010;48:443–474. [Google Scholar]
  37. Morenoff JD, Sampson RJ, Raudenbush SW. Neighborhood inequality, collective efficacy, and the spatial dynamics of urban violence. Criminology. 2001;39:517–560. [Google Scholar]
  38. Mrug S, Windle M. Initiation of alcohol use in early adolescence: links with exposure to community violence across time. Addict Behav. 2009;34:779–781. doi: 10.1016/j.addbeh.2009.04.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Musick K, Seltzer JA, Schwartz CR. Neighborhood norms and substance use among teens. Soc Sci Res. 2008;37:138–155. doi: 10.1016/j.ssresearch.2007.02.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. National Academy of Sciences. Reducing Underage Drinking: A Collective Responsbility. National Academies Press; Washington, DC: 2004. [PubMed] [Google Scholar]
  41. National Household Survey on Drug Abuse. National Household Survey on Drug Abuse: Population estimates. National Institute on Drug Abuse; Rockville, MD: 1991. [Google Scholar]
  42. National Prevention Council. National Prevention Strategy. U.S. Department of Health and Human Services, Office of the Surgeon General; Washington, D.C: 2011. [Google Scholar]
  43. National Research Council and Institute of Medicine. Preventing Mental, Emotional, And Behavioral Disorders Among Young People: Progress And Possibilities. National Academies Press; Washington, D.C: 2009. [PubMed] [Google Scholar]
  44. Osgood DW, McMorris BJ, Potenza MT. Analyzing multiple-item measures of crime and delinquency I: item response theory scaling. J, Quant, Criminol. 2002;18:267–296. [Google Scholar]
  45. Osgood DW, Wilson JK, O’Malley PM, Bachman JG, Johnston LD. Routine activities and individual deviant behavior. Am Sociol Rev. 1996;61:635–655. [Google Scholar]
  46. Raudenbush SW, Bryk AS. Hierarchical Linear Models: Applications And Data Analysis Methods. 2. Sage; Newbury Park, CA: 2002. [Google Scholar]
  47. Raudenbush SW, Bryk AS. HLM 6: Hierarchical Linear and Nonlinear Modeling. Scientific Software International, Inc; Lincolnwood, IL: 2014. [Google Scholar]
  48. Raudenbush SW, Sampson RJ. Toward a science of assessing ecological settings, with application to the systematic social observation of neighborhoods. Sociol Methodol. 1999;29:1–41. [Google Scholar]
  49. Sampson RJ. Great American City: Chicago And The Enduring Neighborhood Effect. University of Chicago Press; Chicago, IL: 2012. [Google Scholar]
  50. Sampson RJ, Bartusch DJ. Legal cynicism (and subcultural?) tolerance of deviance: The neighborhood context of racial differences. Law Soc Rev. 1998;32:777–804. [Google Scholar]
  51. Sampson RJ, Morenoff JD, Raudenbush SW. Social anatomy of racial and ethnic disparities in violence. Am J Public Health. 2005;95:224–232. doi: 10.2105/AJPH.2004.037705. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Sampson RJ, Raudenbush SW. Systematic social observation of public spaces: a new look at disorder in urban neighborhoods. Am J Sociol. 1999;105:603–651. [Google Scholar]
  53. Sampson RJ, Raudenbush SW, Earls F. Neighborhoods and violent crime: a multilevel study of collective efficacy. Science. 1997;277:918–924. doi: 10.1126/science.277.5328.918. [DOI] [PubMed] [Google Scholar]
  54. Shaw CR, McKay HD. Juvenile Delinquency And Urban Areas. University of Chicago Press; Chicago: 1942. [Google Scholar]
  55. Snedker KA, Herting JR, Walton E. Contextual effects and adolescent substance use: exploring the role of neighborhoods. Soc Sci Q. 2009;90:1272–1297. [Google Scholar]
  56. Tobler AL, Komro KA, Maldonado-Molina MM. Relationship between neighborhood context, family management practices, and alcohol use among urban, multi-ethnic, young adolescents. Prev Sci. 2009;10:313–324. doi: 10.1007/s11121-009-0133-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Turner RJ, Frankel B, Levin D. Social support: conceptualization, measurement, and implications for mental health. In: Greenley JR, editor. Research In Community And Mental Health. Vol. 3. JAI Press; Greenwich, CT: 1983. pp. 67–111. [Google Scholar]
  58. Wilcox P. An ecological approach to understanding youth smoking trajectories: problems and prospects. Addiction. 2003;98:57–77. doi: 10.1046/j.1360-0443.98.s1.5.x. [DOI] [PubMed] [Google Scholar]
  59. Wright EM, Fagan AA. The cycle of violence in context: exploring the moderating roles of neighborhood disadvantage and cultural norms. Criminology. 2013;51:217–249. doi: 10.1111/1745-9125.12003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Wright EM, Fagan AA, Pinchevsky GM. Penny for your thoughts? The protective effect of youths’ attitudes against drug use in high risk communities. Youth Violence Juv Justice in press. [Google Scholar]
  61. Zimmerman GM, Messner SF. Neighborhood context and nonlinear peer effects on adolescent violent crime. Criminology. 2011;49:873–903. [Google Scholar]
  62. Zimmerman GM, Vasquez BE. Decomposing the peer effects on adolescent substance use: mediation, nonlinearity, and differential nonlinearity. Criminology. 2011;49:1235–1273. [Google Scholar]

RESOURCES