Integrating Individual Risk and Social Exposure to Violence: A Multilevel Victimization Perspective

Abstract

We test two major hypotheses in this article: (a) macrolevels of school disorganization and individual levels of low self-control will be directly, and positively, linked to victimization and (bi) low self-control will have the largest impact on exposure to victimization (ETV) when it interacts with negative environments consistent with a social enhancement perspective, or (bii) low self-control will have a weaker impact on ETV when it interacts with negative environments consistent with saturation or social push models. The data for the current study were collected as part of the second International Self-Report Delinquency Study (ISRD-II). A total of 49,685 individuals from 30 countries are nested within 1,427 schools. We use multilevel generalized linear regression models with violent victimization (robbery and assault) regressed on demographic, family, school, and neighborhood variables. Multiplicative interaction terms are included in separate models to examine key moderation effects consistent with expectations drawn from the victimization literature. Analyses reveal that low self-control and perceptions of school disorganization are both associated with an increase in the odds of experiencing victimization. Interactions between low self-control and school disorganization are also found to be consistent with saturation/social push models. Our regulation approach offers a foundation for theorizing about ETV and provides a testable model for future research. However, elements of the regulation model are in need of further refinement and testing before the perspective can be moved toward a broader theory of victimization.

Keywords

disorganization integration schools self-control victimization

Introduction

Theorizing about offending and victimization is currently in a dystopian state. It has long been recognized that the “theoretical landscape” is cluttered (Bernard, 1990), with theories seeking to advance understanding of behavior and exposure to victimization (ETV) proliferating, while older theories remain in the mix. Further, existing theories do not explain much of the variation in offending and victimization in the population (Weisburd & Piquero, 2008). This suggests a real need for additional theoretical approaches and refinement of theories that can better predict outcomes of interest to criminologists and describe connections between individuals and their social contexts. Many of the efforts to address these shortcomings have taken an integrated approach—combining one or more existing theories into a more holistic theoretical model (refer to Thornberry, 2012). However, many integrated theories have been difficult to test and are met with limited, or no, support (refer to Krohn & Ward, 2015). Additionally, most of these theories are applied only to offending (Messner et al., 1989; but refer to Miethe & McDowall, 1993).

Recently, new theoretical models and statistical approaches have begun to advance knowledge about crime causation and ETV. One of the most fruitful recent approaches has been to consider both social structural factors and individual differences together. For instance, person-in-context research has investigated how individual differences interact with social environments to predict negative outcomes (Magnusson & Stattin, 2007; Vogel, 2016; Vogel & van Ham, 2018; Zimmerman, 2010). Biosocial models stress how the combination of the physical environment and genetic predispositions leads to offending and ETV (Wright & Boisvert, 2009). Moreover, developmental criminology continues to explicate the social and biological processes that lead to poor life outcomes across life stages (Farrington, 2003). Through these types of integration, a better model for explaining life outcomes can be posited and more effective intervention approaches can be implemented (Rocque & Welsh, 2012).

The current study expands victimization research and theory by considering both macro-level factors (perceptions of school disorganization) and individual differences (low self-control) as well as their interactive effects to explain ETV, it is likely that both individual propensity to engage in risky activities and spending time in risky environments will increase victimization. Opportunities to engage in risky activities are more prevalent in disorganized environments, but individuals can seek out these opportunities even in organized environments. Similarly, even those who refrain from engaging in risky activities might find themselves in risky situations when they live or spend time in disorganized environments (Sampson & Wooldridge, 1987; Zimmerman, 2010). Therefore, there is an intersection between personal characteristics and social environments that lead to negative social outcomes. Using a unique school-based sample from a dataset with information from 30 countries, we use multilevel models to examine these novel victimological questions.

Theoretical Foundation

With the rise of theoretical integration, cross-level theorizing emerged to show how macro and micro factors combine to explain offending (Miethe & Meier, 1994). With few exceptions (refer to, e.g., Rountree et al., 1994), victimization theories remain single-level perspectives, with lifestyle or opportunity accounts dominating these efforts (refer to Cohen & Felson, 1979; Hindelang, 1978; Schreck & Fisher, 2004). These theories suggest that victimization is a function of being in the wrong place, at the wrong time (e.g., when the risk of criminal events is high). Certain research has shown the relevance of individual differences, such as self-control (Piquero et al., 2005; Pratt et al., 2014; Schreck, 1999) and psychopathy (Flexon et al., 2016) to being victimized. To date, however, few have integrated individual differences and macro factors to more fully understand victimization.

Macro-level Social Control

Macro-level theories of individual outcomes have received much attention in the social sciences, particularly from sociology and social psychology (Messner & Rosenfeld, 2012). At their core, macro-sociological theories suggest that human behavior and mental/physical health outcomes are the results of environmental factors, such as neighborhood poverty and disorganization. Often, these factors are used to explain varying rates of antisocial behavior across macro units.

In the early 1900s, the Chicago School put forth theories that considered the city, or neighborhood, as a vital context in which to consider crime and violence (Shaw & McKay, 1942). Social disorganization theory and, later, collective efficacy theory (Sampson & Groves, 1989), built on social regulation ideas and argued that areas which could not regulate the behavior of its residents would suffer high rates of crime and violence. Criminologists Steven Messner and Richard Rosenfeld have continued to argue for a macro-level control theory which suggests that imbalance in social institutions at the macro-level will lead to a more criminogenic society through dysregulation of individual behavior (Messner & Rosenfeld, 2012). These theories of social regulation are particularly important for the current study in that the ability of the physical environment to permeate the individual and influence behavior is an essential component of a regulatory theory of victimization. It is also an integral aspect of the person-in-context theoretical approach.

More recent research has shown that the macro-level also matters for individual outcomes. For example, much work has shown that neighborhood factors predict individual behavior (Elliott et al., 1996; Rotger & Galster, 2019; Xie & Baumer, 2019). At the same time, however, macro-level perspectives are incomplete, given that most people in a particular area are not engaged in deviance and are not the victims of violent crime.

Micro-level Social Control

Micro-level theories have found a slightly different audience than macro-level theories across the decades (although micro-sociological approaches have been used in explaining criminal behavior including Hirschi’s, 1969, general theory of delinquency). Popular among psychologists and biologists, micro or individual theories of negative mental and physical health outcomes argue that individual differences that exist among people, such as levels of self-regulation (Steinberg, 2010), intelligence (Beaver et al., 2016; Murray & Farrington, 2010), and empathy (Jolliffe & Farrington, 2004; Posick et al., 2014), are the major drivers of behavior and other negative social outcomes.

Today, many scholars argue that individual differences play a prominent role in leading to poor behavior and exposure to violence. Just some of these differences include IQ/intelligence, psychopathic personality, empathy, genetics, and self-control (Beaver et al., 2014; DeLisi, 2009; Moffitt et al., 2011). Pertinent to the current framework, self-control remains one of the most powerful predictors of criminal behavior (Pratt & Cullen, 2000) and ETV (Pratt et al., 2014). Studies from various fields substantiate the idea that low self-regulation places individuals at risk for behaving delinquently as well as exposes them to a host of maladies including violent and non-violent victimization (Posick, 2013; Schreck, 1999; refer to Rocque et al., 2016 for a review).

With respect to victimization, individual-level factors such as self-control have recently been examined. Schreck (1999) first attempted to show how Gottfredson and Hirschi’s (1990) theory, which claimed that self-control is the primary cause of offending, could account for victimization as well. Much research since then has shown that self-control is an important predictor of victimization (Pratt et al., 2014). At the same time, individual factors do not operate in a vacuum. This knowledgebase highlights the importance of both macro and micro factors in understanding individual outcomes. Given the empirical support for low self-control as an individual-level predictor of ETV and the role of negative environments on a host of poor health outcomes, this study focuses on the role of self-control and perceptions of school disorganization on ETV.

Macro–Micro Integration and the Current Theoretical Framework

A regulatory theoretical framework for exposure to violence.

In Wellford’s seminal piece on theoretical integration (1989, p. 123), he argued that a truly interdisciplinary theory would “accept the following meta-theoretical principles:” first, that levels of analysis are independent, implying that “no level of analysis can be totally explained, nor can its effects be totally explained by the operation of any other level of analysis.” In other words, solely examining one level of analysis will be woefully inadequate. Second, to develop a true theory (as opposed to a multi-factor approach), it is essential to examine how levels of analysis interact.

Research has been instructive concerning how individual-level factors interact, or depend on, higher-level factors at the meso- and macro-levels. From a risk perspective, those with individual-level characteristics that increase the chances of becoming a victim, who also live in areas with higher risk of victimization, should be most likely to be victimized. In the biosocial literature, this line of thinking is termed the “diathesis-stress” or amplification hypothesis (Belsky & Pluess, 2009; Simons et al., 2011). Other work has indicated that risky individual-level traits may only influence behavior in disadvantaged contexts; that is, environmental-level characteristics are conditioned by individual-level factors. For example, in a ground-breaking study, Caspi et al. (2002) found that maltreated children who carried a particular variant of MAOA were more likely to become antisocial than those who did not have the variant. Genetic and biological predispositions (ETV included) are often, but not always, dependent on the environment (Beaver et al., 2007; Duncan & Keller, 2011).

Disadvantaged neighborhoods also have the potential to impact neurological functioning leading to a host of maladies both behavioral and physical (Umbach et al., 2017). Even country-level contextual variation in societal functioning has been shown to impact the effects of individual-level variables (Posick & Gould, 2015). While still understudied, evidence suggests that some personal factors might integrate with the environment to influence behavior and other outcomes (for a useful example of a multilevel approach to studying victimization consult Miethe & Meier, 1994).

Pertinent to the current exposition of ETV, self-control may rely, at least partially, on the environment in its impact on behavior and situational outcomes. The effects of impulsivity, a facet of low self-control, have been shown to be amplified in areas with high socioeconomic status and collective efficacy (Zimmerman, 2010; but also refer to Vazsonyi et al., 2006 for contrary results). The effects of self-control also vary by the community moral context. Individuals with low self-control who reside in neighborhoods that are characterized by low morality are at elevated risk of exhibiting criminal behavior (Zimmerman et al., 2015). Other work has found that self-control interacts with opportunity such that with more opportunities comes more delinquency (Longshore, 1998; Longshore & Turner, 1998).

Another seminal multilevel study on victimization also focused on opportunity along with a person’s lifestyles and routine activities. Drew upon both routine activities theory (macro) and lifestyles theory (micro), to show that both levels of analysis matter, and that the effect of individual-level factors is dependent on macro-level characteristics (refer to also Wilcox et al., 2003). This type of multilevel analysis has become common in victimization studies, which have recognized the importance of both the micro and macro environments (refer to Velez, 2001).

Yet individual differences, such as personality or genetic factors, have been less utilized in these integration attempts. There is reason to believe such individual-level factors are conditioned by the environment in predicting victimization. For example, Brendgen et al. (2008) found that peer victimization interacted with genes in leading to aggression for girls. In a later study, Brendgen et al. (2015) found that genetic variants putting individuals at risk for aggression influenced victimization only in certain school classroom contexts. If the classroom norms frowned upon antisocial behavior, such genes increased the likelihood of victimization; in other classroom environments, such genes decreased the likelihood of victimization. Similar work with genetic factors is needed in the future.

In this study, we propose that victimization be considered under the rubric of social and self-regulation. Individuals who are exposed to violent victimization and, effectively, become victims, are those who are often in contexts that promote violence (or at least appear to condone violence) and who lack the ability to adequately diffuse potential violent interpersonal conflict. This is certainly not to suggest that all victimization is due to these two factors and how they interact, but, specifically with violent crime, the inability to regulate the physical environment coupled with difficulty in self-regulation can place an individual at heightened risk for victimization. In other words, control operates on (at least) two levels of analysis—the social/macro and the micro levels. When control is absent or deficient in one of these levels, it must be “made up” in the other. When both are lacking, victimization is likely to occur.

Figure 1 highlights the perspective advanced here. The horizontal line (x-axis) represented self-regulation or individual vulnerability to victimization, while the vertical line represents environmental-regulation or situational ETV. Individuals with high dysregulation (on the right side of the matrix) are often vulnerable to violence either due to with whom or where they associate and/or their inability to appropriately negotiate social interactions which makes them a target of violence. Individuals who spend time in dysregulated environments are more likely to be exposed to violence that happens within areas that are not able to regulate the behavior of its inhabitants. Where both the individual and environment are dysregulated (top right quadrant), exposure to violence is expected to be greatest whereas exposure is expected to be lowest where the individual and environment are well regulated (bottom left quadrant).

Figure 1.

Regulation-exposure theoretical perspective of victimization.

While research tends to treat the “environment” as communities or neighborhoods, when examining outcomes related to adolescents, the school environment may be the most relevant. In fact, during the academic year, adolescents spend more time in school than in any other setting (Steinberg, 2000). It is also at school where the majority of students learn attitudinal and behavioral norms that will shape their behavior (Harris et al., 2002). Prior research supports the notion that the school context can moderate the effects of other variables such as exposure to violence (Zimmerman & Posick, 2014). When testing regulatory theories of behavior and ETV, it is essential to model environments that are most important to the sample being studied. Therefore, we use the school setting as the focal macro/environmental factor in this study.

Using this literature as an empirical backdrop, we capitalize on countries and hypothesize that: (a) macro-levels of school disorganization and individual levels of low self-control will be directly, and positively, linked to victimization. Borrowing from the person-incontext literature, it is also hypothesized that: (bi) low self-control will have the largest impact on ETV when it interacts with negative environments consistent with a social enhancement perspec-tive, or (bii) low self-control will have a weaker impact on ETV when it interacts with negative environments consistent with saturation or social push models. These hypotheses are tested using data from the second wave of the International Self-Report Delinquency Study.

Data and Methods

Sample

The data for the current study were collected as part of the second International Self-Report Delinquency Study (ISRD-II). The ISRD-II sampling procedure was fairly complex and complete information on the design can be found in Marshall and Enzmann (2012). Generally, the intent was to have standardized survey instruments, sampling plans, and data entry across participating countries. In this effort, the ISRD-II countries used a “flexible” standardization method whereby each country used the same questionnaire but was allowed to add questions to the end of the survey if necessary.

The participating countries attempted to obtain a sample which included three subpopulations: (a) a large metropolitan area that served as an economic center (population of 300,000 or greater); (b) a mid-size city (100,000–299,999 inhabitants); and (c) two small cities or townships (10,000–99,999 inhabitants). The three subpopulations would have a total sample size of at least 700 students. Most countries included a large or medium-sized city as required (except Aruba), but the selection of these cities was based more on convenience (given the connections of researchers and the feasibility of sampling) than on representativeness of the country or comparability to other cities from other countries.

The sampling strategy required that each country randomly select 7th-, 8th-, and 9th-grade classrooms in each selected city—making the ISRD a school-based sample and well-suited for the current research study. First, a listing of all schools in the research cities was created that included all 7th, 8th, and 9th-grade classrooms. Second, classrooms were randomly drawn from the original list of classrooms containing 7th–9th graders. The sample includes a mix of public, private, vocational, technical, and academic schools.

The hierarchical structure of the data is presented in Table 1. There are a total of 49,685 individuals from 30 countries spread across 1,427 schools.¹

This was the final sample size after listwise deletion of individuals with missing cases and dropping schools with under six individuals. While more males were missing than females, the final results remained unchanged when using imputation through chained equations.

Each country has a substantial sample size to contribute to the overall dataset (on average 1,713 students) and, on average, each country contains a sample of about 49 schools with almost 44 students per school on average.

Table 1.

Descriptive Characteristics of Countries and Schools.

Country	Sample Size	Number of Schools	Averagen per School
United States	1,722	14	123
Ireland	1,145	37	31
Iceland	464	25	19
Norway	1,245	41	30
Sweden	1,572	71	22
Finland	1,316	44	30
Netherlands	1,986	39	51
Austria	2,461	45	55
Germany	2,890	68	43
Switzerland	3,090	70	44
France	1,737	21	83
Belgium	1,629	41	40
Portugal	1,639	56	29
Cyprus	1,578	16	99
Italy	4,169	94	44
Spain	1,348	79	17
NLAntilles	1,165	32	36
Aruba	467	10	47
Suriname	1,671	83	20
Venezuela	1,121	45	25
Estonia	1,580	89	18
Lithuania	1,544	47	33
Poland	1,034	41	25
Czech Republic	2,576	89	29
Hungary	1,644	96	17
Slovenia	1,745	41	43
Bosnia & Herzegovina	1,397	37	38
Russia	1,867	41	46
Armenia	1,883	15	126
Total	49,685	1,427	–
Averages	1713.3	49.2	43.6

Variables

Dependent variable.

The focus of the present study is to explain serious personal victimization. Therefore, the dependent variable reflects a measure of violent victimization which includes whether or not the respondent was ever victimized by robbery or assault in the past 12 months (0 = no; 1 = yes). The sample is relatively low risk, and this is reflected by a fairly low prevalence of violent victimization (6.43%).

Independent variables.

Two study variables are the focus of the current study. The first is a measure of school disorganization. This measure was constructed from four questions on the questionnaire including whether the respondent has witnessed: (a) stealing; (b) fighting; (c) vandalism; and (d) drug use in their school (ranging from 1 = not at all true to 4 = very true). This measure was converted to a percent of maximum possible (POMP) score using the method by Cohen et al. (1999) theoretically ranging from 0 indicating no disorganization at all to 100, which indicates the highest possible level of disorganization on each item (α = 0.75). The scores were then averaged over schools to measure school disorganization.

The second independent variable is low self-control. Low self-control is measured by combining 12 items from the Bursick et al. (1993) scale which covers the temperament, self-centeredness, risk-seeking, and impulsivity domains. Each item ranges from (1) Fully Agree to (4) Fully Disagree for questions such as “I act on the spur of the moment without stopping to think.” The scale is reverse coded, and converted to a POMP score, so that higher scores represent less self-control (α = 0.82).

Control variables.

Demographic characteristics are included as controls in the study including: sex (males are included in the model and females are the reference category), grade (includes 7th and 8th graders in the models as dummy variables with 9th graders considered the reference category), and nativity status (non-natives are captured in the models, while natives are the reference category). Two other individual-level characteristics, drug/alcohol use, and time spent with friends, are included in the models. Drug/alcohol use is measured by the number of times the individual drank beer and/or wine, drank spirits, and/or smoked hash (marijuana) in the past month (ranging from 0 to 64 times in the past month). Time spent with friends is measured by one question on the survey asking the respondent how much time each day they spend with friends ranging from 1 = no time to 6 = four or more hours.

A series of theoretical constructs are developed and modeled in the current study. Neighborhood disorganization is very similar to the school disorganization variable and taps similar activities but in the neighborhood including whether the respondent agrees or disagrees with the following statements: (a) There is a lot of crime in my neighborhood; (b) There is a lot of drug selling; (c) There is a lot of fighting; (d) There are a lot of abandon and empty buildings; and (e) There is a lot of graffiti. Each item ranges from 1 = not at all true to 4 = very true. These items are combined into a POMP score (α = 0.81).

Attitudes toward violence capture the extent to which the respondent agrees with certain behaviors/attitudes including: (a) A bit of violence is part of the fun; (b) One needs to make use of force to be respected; (c) If someone attacks me, I will hit him/her back; (d) Without violence everything would be much more boring; and (e) It is completely normal that boys want to prove themselves in physical fights with others. Each item ranges from 1 = disagree fully to 4 = agree fully. These items are combined into a POMP score (α = 0.70).

Negative life events represent a series of stressful situations that might strain a young person. This measure includes questions about a death or illness of a close relative or friend and problems with parents in the home (e.g., divorce). Each item has response options of 0 = no and 1 = yes. This measure does not represent a latent construct but does represent an accumulation of serious negative events that children may be exposed to. Regardless, it is converted into a POMP score for easy interpretation.

Family bonding is a construct consisting of four items on the questionnaire including: (a) How do you usually get along with the man you live with (father, stepfather…) (4-point scale; I get along just fine to I do not get along at all); (b) How do you usually get along with the woman you live with (your mother or stepmother) (4-point scale; I get along just fine to I do not get along at all); (c) How often do you and your parents (or the adults you live with) do something together, such as going to the movies, going for a walk or hike, visiting relatives, attending a sporting event, and things like that? (6-point scale; More than once a week to Almost never); and (d) How many days a week do you usually eat the evening meal with (one of) your parents (or the adults you live with)? (8-point scale; Never to Daily). Each item was standardized prior to converting to a POMP score given different response categories (α = 0.55).²

The reliability of this measure is low. However, the items theoretically group together and have been used in prior research (refer to Posick, 2013). Therefore, we have included this control variable. The effect of this variable may be weaker than expected given the low reliability.

Finally, a composite measure of the delinquent activities that the respondent’s friends engage in was developed from five questions on the questionnaire asking the respondent if they have friends who: (a) use hard drugs; (b) steal; (c) burglarize; (d) rob; and (e) assault others. The measure ranges from 0 = no activities to 5 = each activity. Descriptive statistics of the variables included in this study are presented in Table 2.

Table 2.

Descriptive Statistics (Individual n = 49,685; School n = 1,427).

Variables	Mean/%	Standard Deviation	Minimum Value	Maximum Value
Dependent variable
Violent victimization	6.43%	–	0	1
Focal independent variables
School disorganization	36.54	24.44	0	100
Low self-control	37.83	19.70	0	100
Control variables
Sex (male)	47.44%	–	0	1
Sex (female)^r	52.56%	–	0	1
Grade 7	32.51%	–	0	1
Grade 8	32.93%	–	0	1
Grade 9 ^r	34.56%	–	0	1
Non-native	7.22%	–	0	1
Native ^r	92.78%	–	0	1
Drug/alcohol use	1.12	3.40	0	64
Time spent w/friends	4.13	1.65	1	6
Neighborhood disorganization	19.76	22.59	0	100
Attitudes toward violence	32.10	21.41	0	100
Negative life events	18.75	14.98	0	100
Family bonding	81.59	16.56	0	100
Delinquent peer activity	0.74	1.13	0	5

Note. n = Sample size; ^r = Reference group in regression analysis.

Analytic Strategy

The main analysis used to examine the proposed micro–macro integration is a multilevel logistic regression where individuals are represented in level 1 and schools are represented in level 2. The unconditional, random intercept model (not shown) indicated that the intercept (i.e., level of victimization) significantly varied across schools, and that about 10% of the variation in victimization was due to the school context. This necessitates a multilevel technique capable of calculating the appropriate coefficient estimates and standard errors (Raudenbush & Bryk, 2002).

Since only variables with complete information on all of the study variables were used in the analysis, each model has the same sample size. This allows for the easiest and most appropriate comparison of the log-likelihood statistic (Hedeker & Gibbons, 2006). As a sensitivity test, all models were re-run using imputation through chained equations. The results indicated no substantive differences. All continuous variables were standardized by computing z-scores. This increases the interpretability of regression coefficients as well as reduces collinearity, particularly when including the interaction terms. There was no collinearity evident among any of the independent variables according to variance inflation factors.

Results

Table 3 presents the results from the multilevel logistic regression models regressing personal victimization on school disorganization and self-control and control variables. In Model 1, both school disorganization and low self-control are significantly related to victimization (b = 0.15, p < 0.001) supporting the first hypothesis. This corresponds to a 16% increase in the odds of victimization for each standard deviation increase in school disorganization and low self-control. Each of the control variables is also significantly related to victimization in the expected direction except for attitudes toward violence which is found to decrease the odds of victimization. This may reflect a true effect where these attitudes insulate against offending (e.g., people do not want to victimize someone who harbors these violent attitudes or who might retaliate) or it may be that those individuals who have positive attitudes toward violence are unwilling to admit or disclose their experiences with victimization.

Table 3.

Multilevel Regression Models Examining the Impact of School Disorganization and Low Self-control on Victimization (Individual n = 49,685; School n = 1,427).

	Model 1		Model 2
	b	SE	b	SE
Focal independent variables
School disorganization	0.15***	0.02	0.17***	0.02
Low self-control	0.15***	0.02	0.17***	0.03
Control variables
Sex (male)	0.85***	0.04	0.84***	0.04
Grade 7	0.29***	0.05	0.30***	0.05
Grade 8	0.18***	0.05	0.19***	0.05
Non-native	0.25***	0.07	0.25***	0.07
Drug/alcohol use	0.01**	0.00	0.01**	0.00
Time spent w/friends	0.03*	0.01	0.03*	0.01
Neighborhood disorganization	0.16***	0.02	0.16***	0.02
Attitudes toward violence	–0.05*	0.02	–0.05	0.02
Negative life events	0.30***	0.02	0.30***	0.02
Family bonding	–0.14***	0.02	–0.14***	0.02
Delinquent peer activity	0.18***	0.02	0.18***	0.02
Low self-control × School disorg.	–	–	–0.05**	0.02
Intercept	–3.93	0.08	–3.92	0.08
Wald chi²	1,843.98***		1,835.48***
Log likelihood	–10,745.72		–10,740.26
Random components
Intercept	0.59	0.03	0.59	0.03
Rho	0.10	0.01	0.10	0.01

Model 2 of Table 3 incorporates an interaction among self-control and school disorganization. The interaction is significant and negative (b = –0.05, p < 0.05). The negative coefficient for the interaction term indicates that the effects of low self-control decrease as perceptions of school disorganization increase. This result is contrary to the Hypothesis bii which expects that negative environments will increase, or amplify, the effect of self-control consistent with the diathesis/stress or amplification theory. However, it stands that individuals with low selfcontrol and who attend disorganized schools have higher levels of victimization..

Figure 2 graphically depicts this relationship. Looking at the right side of the graph, it is clear that among those who perceive high levels of social disorganization, low self-control has an equally powerful influence on ETV. The left side of the graph suggests that low self-control has a more powerful impact on those who perceive less social disorganization.

Figure 2.

The effect of low self-control on victimization according to levels of school disorganization.

Discussion

Advances in theoretical perspectives have shown demonstrably that single levels of analysis are not adequate to understand behavioral outcomes. Work in the area of victimization has also indicated that macro- and micro-level factors are necessary to take into consideration. However, this work has rarely taken individual-level factors into account, such as self-control. Given that research has strongly supported the association between this trait and victimization, micro–macro perspectives should incorporate it into theoretical accounts. The current study proposed a new cross-level theoretical perspective and tested it with a sample of school-aged youth across the world.

Theoretical Implications

The results of the study point to some fruitful areas for future theorizing. The direct effects of self-control on victimization are consistent with prior studies (Pratt et al., 2014; Schreck, 1999). Regardless of the person and the context they are in, low self-control increases the risk of victimization. However, similar to a study by Posick and Zimmerman (2015), school context did moderate the link to victimization. Among those who perceived less school disorganization, low self-control had a larger impact on victimization. As discussed by others (refer to Wright & Fagan, 2013), this may represent a saturation effect whereby those individuals who spend time in disorganized schools are already exposed to so many risk factors that the total effect of any one factor is diminished by the others (refer also to Raine, 2002). There may also be some desensitization effects where disorganization is just a normal part of life for some adolescents and they do not seem to believe that there is much to be concerned about or report.

In a theoretical and empirical sense—context matters. Like previous studies by other social scientists (refer to Posick & Zimmerman, 2015; Zimmerman, 2010; Zimmerman & Messner, 2011), this study confirms the role that context plays in calibrating the effects of specific variables—here, low self-control. Given the rich history that context plays in criminology, victimology should likewise seek to place the person within their context to understand ETV. Moving forward, additional individual and contextual factors can be examined to elaborate victimization theories and more fully integrate multilevel insights.

Finally, interactive effects hypothesized by current theories may need to be revisited. For example, in some studies, disadvantaged areas increase the effect of self-control (Zimmerman et al., 2015), decrease the effect of self-control (Zimmerman, 2010), or have no effect on self-control (Vazsonyi et al., 2006). Asking “which is it?” might be the wrong question. More likely, the relevant question is which individuals, within which contexts, are most impacted by individual-level characteristics? There appears to be room for nuance and further investigation into this ripe theoretical territory. Intersectional research, which considered the multiple roles and experiences of diverse populations, will certainly contribute to a better understanding of the “who” and “where” of ETV. One thing is clear, however, from this and previous work: theoretical perspectives that focus on one level of analysis are almost assuredly incomplete.

Practical Implications

It is much too early to suggest definitive policies and programming related to the theoretical and empirical framework presented and tested in this study. However, there are a few take-away points that emerge not only from the current study but also with prior research. First, self-regulation and self-control, once again, are clearly linked to a negative social outcome—here, violent victimization. Violence prevention and intervention programs would do well to capitalize on this knowledge and address deficits in self-control and use empirically supported strategies to increase self-regulation. Research has strongly supported the notion that self-control, in spite of the arguments of Gottfredson and Hirschi (1990) can be increased through programming (Piquero et al., 2016). While these programs may generally be thought to help prevent antisocial behavior, our results along with previous work indicate that they may prevent victimization as well. Given that self-control was found to be more strongly related to victimization in socially disorganized environments, our findings point out contexts in which such programs may be more effective.

Considering the environmental exposure components of this research, schools can likely play a part in decreasing victimization by providing safe activities for youth with appropriate guardianship. Environments offering structured activities with supervision by prosocial adults are much more capable of regulating the social environment than unstructured and unsupervised areas. Schools and local communities that likewise provide clean, organized, and safe places for youth to learn and play can decrease violence by signaling to individuals that any disruption of the local area will be addressed. This idea is in line with a long line of research showing an association between unstructured socializing and delinquency for youth (Osgood & Anderson, 2004). Places with organized safe spaces are not only able to keep youth busy during high-risk times but are also likely to establish stakes in conformity for those spending time in those areas.

Further, assessments should not ignore either individual characteristics or contextual factors in adolescent’s lives that likely effect social, physical, and mental well-being. A long line of research highlights the important role of social context for well-being and this study echoes this literature. However, the exact ways in which environments influence other factors are still up for debate and it is too early to suggest how policies and programs should use this information. Future research should continue to integrate multiple factors, in a theoretically consistent manner, when investigating victimization as well as consider other phenomena such as self-selection into particular environments that might influence the outcome of research studies.

Limitations and Future Research

A few non-trivial limitations of this study provide avenues for future research. First, the data are non-experimental and cross-sectional which limits the ability to identify causality and the time ordering of events. This is particularly problematic with the misperceptions of school disorganization variable because it is impossible to know if perceptions changed after variable because it is impossible to know if perceptions changed after being victimized or if individuals self-selected into more disorganized areas of school leading to their victimization. While experimental methods would help establish causality, it is not ethically possible given random assignment into potentially harmful environments. Therefore, longitudinal data collected from multiple contexts is needed to expand this perspective. Longitudinal data could help address causal order to ensure the factors identified here are temporally prior to the dependent variable. Longitudinal data could also be used to test reciprocal relationships. Causal order remains a very valuable issue to have clarity on moving forward with this line of research.

Second, there are more individual differences that should be accounted for including personality traits from the “dark triad”³

The “Dark Triad” includes three individual personality differences: (a) Psychopathy, (b) Machiavellianism, and (c) Narcissism. Refer to Paulhus and Williams (2002) for more information.

and biological differences—all of which have been shown to be linked to exposure to violence (Flexon et al., 2016; Furnham et al., 2013). In fact, low self-control itself plays a prominent role in the dark triad. Low self-control is prominent in psychopaths who are the most likely to exhibit antisocial behavior (Jones & Paulhus, 2011). Again, self-control might be one component of psychopathy that needs to be integrated with other variables (refer to DeLisi, 2009) including biological factors, such as genes, which have also been shown to increase exposure and resiliency to victimization (Beaver et al., 2007; Beaver et al., 2011; Boutwell et al., 2013).

Finally, the dampening role of school disorganization on the effect of self-control may be due to the low prevalence of victimization in the sample and/or the result of having such a young, relatively non-deviant sample. This may explain some of the contradictory findings with previous research. How the results would change with a system-involved or high-risk sample will have to wait until similar research studies are carried out with different samples. This could include youth who are on probation or who are incarcerated along with those who have been identified as at-risk in their school system. Not only would this increase the prevalence of victimization outcomes but the variety of exposures as well. Additionally, the results may change when examining different dependent variables, with various prevalence rates, including property crime.

Final Remarks

This study is among the first to examine the risk of victimization by combining individual-trait level variables as well as information from one’s school. While the results should not be taken as the final word on the perspective, they do indicate promise in using a multilevel theoretical framework to explain victimization and to conduct empirical tests of the risk of victimization. Approaches that take the individual as well as context into account may be particularly useful as the perspective specifically incorporates individual differences into theoretical explanations of behavior while also recognizing contextual factors that might influence and/or interact with personal characteristics. As researchers are continuing to collect information at the individual level as well as contextual data on families, schools, and neighborhoods (as well as international data) the potential to refine and expand empirical research on this victimization theory is vast.

Footnotes

Acknowledgments

We thank Kristina Thompson for useful comments on earlier versions of this article.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Chad Posick

Author Biographies

Chad Posick, PhD, is an associate professor and the Graduate Director at Georgia Southern University. He teaches in the areas of victimology, statistics, and quantitative research methods. His research focuses on the causes and consequences of victimization and violence prevention. He is a member of the Scholars Strategy Network and the Health Criminology Research Consortium.

Michael Rocque, PhD, is an associate professor of Sociology at Bates College in Lewiston, Maine. His research interests include race and crime, life-course criminology, and corrections. He is a member of the Scholars Strategy Network.

References

Beaver

K. M.

, Mancini

, DeLisi

, & Vaughn

M. G.

(2011). Resiliency to victimization: The role of genetic factors. Journal of Interpersonal Violence , 26(5), 874–898.

Beaver

K. M.

, Nedelec

J. L.

, Schwartz

J. A.

, & Connolly

E. J.

(2014). Evolutionary behavioral genetics of violent crime. In Shackelford

T. K.

& Hansen

R. D.

(Eds.), The evolution of violence (pp. 117–135). Springer.

Beaver

K. M.

, Nedelec

J. L.

, Barnes

J. C.

, Boutwell

B. B.

, & Boccio

(2016). The association between intelligence and personal victimization in adolescence and adulthood. Personality and Individual Differences , 98, 355–360.

Beaver

K. M.

, Wright

J. P.

, DeLisi

, Daigle

L. E.

, Swatt

M. L.

, & Gibson

C. L.

(2007). Evidence of a gene X environment interaction in the creation of victimization: Results from a longitudinal sample of adolescents. International Journal of Offender Therapy and Comparative Criminology , 51(6), 620–645.

Belsky

, & Pluess

(2009). Beyond diathesis stress: Differential susceptibility to environmental influences. Psychological Bulletin , 35(6), 885–908.

Bernard

T. J.

(1990). Twenty years of testing theories: What have we learned and why? Journal of Research in Crime and Delinquency , 27(4), 325–347.

Boutwell

B. B.

, Franklin

C. A.

, Barnes

J. C.

, Tamplin

A. K.

, Beaver

K. M.

, & Petkovsek

(2013). Unraveling the covariation of low self-control and victimization: A behavior genetic approach. Journal of Adolescence , 36(4), 657–666.

Brendgen

, Girard

, Vitaro

, Dionne

, & Boivin

(2015). Gene-environment correlation linking aggression and peer victimization: Do classroom behavioral norms matter? Journal of Abnormal Child Psychology , 43(1), 19–31.

Brendgen

, Boivin

, Vitaro

, Girard

, Dionne

, & Pérusse

(2008). Gene-environment interaction between peer victimization and child aggression. Development and Psychopathology , 20(2), 455–471.

10.

Caspi

, McClay

, Moffitt

T. E.

, Mill

, Martin

, Craig, I. W., Taylor

, & Poulton

(2002). Role of genotype in the cycle of violence in maltreated children. Science , 297(5582), 851–854.

11.

Cohen

L. E.

, & Felson

(1979). Social change and crime rate trends: A routine activity approach. American Sociological Review , 44(4), 588–608.

12.

Cohen

, Cohen

, Aiken

L. S.

, & West

S. G.

(1999). The problem of units and the circumstance for POMP. Multivariate Behavioral Research , 34(3), 315–346.

13.

DeLisi

(2009). Psychopathy is the unified theory of crime. Youth Violence and Juvenile Justice , 7(3), 256–273.

14.

Duncan

L. E.

, & Keller

M. C.

(2011). A critical review of the first 10 years of candidate gene-by-environment interaction research in psychiatry. American Journal of Psychiatry , 168(10), 1041–1049.

15.

Elliott

D. S.

, Wilson

W. J.

, Huizinga

, Sampson

R. J.

, Elliott

, & Rankin

(1996). The effects of neighborhood disadvantage on adolescent development. Journal of Research in Crime and Delinquency , 33(4), 389–426.

16.

Farrington

D. P.

(2003). Developmental and life-course criminology: Key theoretical and empirical issues-the 2002 Sutherland Award address. Criminology , 41(2), 221–225.

17.

Flexon

J. L.

, Meldrum

R. C.

, Young

J. T.

, & Lehmann

P. S.

(2016). Low self-control and the Dark Triad: Disentangling the predictive power of personality traits on young adult substance use, offending and victimization. Journal of Criminal Justice , 46, 159–169.

18.

Furnham

, Richards

S. C.

, & Paulhus

D. L.

(2013). The Dark Triad of personality: A 10-year review. Social and Personality Psychology Compass , 7(3), 199–216.

19.

Gottfredson

M. R.

, & Hirschi

(1990). A general theory of crime . Stanford University Press.

20.

Grasmick

H. G.

, Tittle

C. R.

, Bursik

R. J.

, & Arneklev

B. J.

(1993). Testing the core empirical implications of Gottfredson and Hirschi’s general theory of crime. Journal of Research in Crime and Delinquency , 30(1), 5–29.

21.

Harris

K. M.

, Duncan

G. J.

, & Boisjoly

(2002). Evaluating the role of “nothing to lose” attitudes on risky behavior in adolescence. Social Forces , 80(3), 1005–1039.

22.

Hedeker

, & Gibbons

R. D.

(2006). Longitudinal data analysis . John Wiley and Sons.

23.

Hindelang

M. J.

, Gottfredson

M. R.

, & Garofalo

(1978). Victims of personal crime: An empirical foundation for a theory of personal victimization . Ballinger.

24.

Hirschi

(1969). Causes of delinquency. Berkley . University of California Press.

25.

Jolliffe

, & Farrington

D. P.

(2004). Empathy and offending: A systematic review and meta-analysis. Aggression and Violent Behavior , 9(5), 441–476.

26.

Jones

D. N.

, & Paulhus

D. L.

(2011). The role of impulsivity in the Dark Triad of personality. Personality and Individual Differences , 51(5), 679–682.

27.

Krohn

, & Ward

J. T.

(2015). Integrating criminological theories. In Piquero

A. R.

(Ed.) The Handbook of Criminological Theory (pp. 318–335). John Wiley & Sons, Inc.

28.

Longshore

(1998). Self-control and criminal opportunity: A prospective test of the general theory of crime. Social Problems , 45(1), 102–113.

29.

Longshore

, & Turner

(1998). Self-control and criminal opportunity: Cross-sectional test of the general theory of crime. Criminal Justice and Behavior , 25(1), 81–98.

30.

Magnusson

, & Stattin

(2007). The person in context: A holistic-interactionistic approach. In Lerner

R. M.

& Damon

(Eds.), Theoretical models of human development (pp. 400–464). Wiley.

31.

Marshall

I. H.

, & Enzmann

(2012). Methodology and design of the ISRD-2 study. In Junger-Tas

, Marshall

I. H.

, Enzmann

, Killias

, Steketee

& Gruszczynska

(Eds.), The many faces of youth crime (pp. 21–65). Springer.

32.

Messner

S. F.

, & Rosenfeld

(2012). Crime and the American dream . Cengage Learning.

33.

Miethe

T. D.

, & McDowall

(1993). Contextual effects in models of criminal victimization. Social Forces , 71(3), 741–759.

34.

Miethe

T. D.

, & Meier

R. F.

(1994). Crime and its social context: Toward an integrated theory of offenders, victims, and situations . State University of New York Press.

35.

Moffitt

T. E.

, Arseneault

, Belsky

, Dickson

, Hancox

R. J.

, Harrington

, ... & Sears

M. R.

(2011). A gradient of childhood self-control predicts health, wealth, and public safety. Proceedings of the National Academy of Sciences , 108(7), 2693–2698.

36.

Murray

, & Farrington

D. P.

(2010). Risk factors for conduct disorder and delinquency: Key findings from longitudinal studies. The Canadian Journal of Psychiatry , 55(10), 633–642.

37.

Osgood

D. W.

, & Anderson

A. L.

(2004). Unstructured socializing and rates of delinquency. Criminology , 42(3), 519–550.

38.

Paulhus

D. L.

, & Williams

K. M.

(2002). The dark triad of personality: Narcissism, Machiavellianism, and psychopathy. Journal of Research in Personality , 36(6), 556–563.

39.

Piquero

A. R.

, Jennings

W. G.

, Farrington

D. P.

, Diamond

, & Gonzalez

J. M. R.

(2016). A meta-analysis update on the effectiveness of early self-control improvement programs to improve self-control and reduce delinquency. Journal of Experimental Criminology , 12(2), 249–264.

40.

Piquero

A. R.

, MacDonald

, Dobrin

, Daigle

L. E.

, & Cullen

F. T.

(2005). Self-control, violent offending, and homicide victimization: Assessing the general theory of crime. Journal of Quantitative Criminology , 21(1), 55–71.

41.

Posick

(2013). The overlap between offending and victimization among adolescents: Results from the second international self-report delinquency study. Journal of Contemporary Criminal Justice , 29(1), 106–124.

42.

Posick

, & Gould

L. A.

(2015). On the general relationship between victimization and offending: Examining cultural contingencies. Journal of Criminal Justice , 43(3), 195–204.

43.

Posick

, & Zimmerman

G. M.

(2015). Person-in-context: Insights on contextual variation in the victim–offender overlap across schools. Journal of Interpersonal Violence , 30(8), 1432–1455.

44.

Posick

, Rocque

, & Rafter

(2014). More than a feeling: Integrating empathy into the study of lawmaking, lawbreaking, and reactions to lawbreaking. International Journal of Offender Therapy and Comparative Criminology , 58(1), 5–26.

45.

Pratt

T. C.

, & Cullen

F. T.

(2000). The empirical status of Gottfredson and Hirschi’s general theory of crime: A meta-analysis. Criminology , 38(3), 931–964.

46.

Pratt

T. C.

, Turanovic

J. J.

, Fox

K. A.

, & Wright

K. A.

(2014). Self-control and victimization: A meta-analysis. Criminology , 52(1), 87–116.

47.

Raine

(2002). Biosocial studies of antisocial and violent behavior in children and adults: A review. Journal of Abnormal Child Psychology , 30(4), 311–326.

48.

Raudenbush

S. W.

, & Bryk

A. S.

(2002). Hierarchical linear models: Applications and data analysis methods . Sage Publications.

49.

Rocque

, Posick

, & Piquero

A. R.

(2016). Self-control and crime: Theory, research, and remaining puzzles. In Vohs

K. D.

& Baumeister

R. F.

(Eds.), Handbook of self-regulation: Research, theory, and applications (pp. 514–532). Guilford Publications.

50.

Rocque

, Welsh

B. C.

, & Raine

(2012). Biosocial criminology and modern crime prevention. Journal of Criminal Justice , 40(4), 306–312.

51.

Rotger

G. P.

, & Galster

G. C.

(2019). Neighborhood peer effects on youth crime: Natural experimental evidence. Journal of Economic Geography , 19(3), 655–676.

52.

Rountree

P. W.

, Land

K. C.

, & Miethe

T. D.

(1994). Macro-micro integration in the study of victimization: A hierarchical logistic model analysis across Seattle neighborhoods. Criminology , 32(3), 387–414.

53.

Sampson

R. J.

, & Groves

W. B.

(1989). Community structure and crime: Testing social-disorganization theory. American Journal of Sociology , 94(4), 774–802.

54.

Sampson

R. J.

, & Wooldredge

J. D.

(1987). Linking the micro- and macro-level dimensions of lifestyle-routine activity and opportunity models of predatory victimization. Journal of Quantitative Criminology , 3(4), 371–393.

55.

Schreck

C. J.

(1999). Criminal victimization and low self-control: An extension and test of a general theory of crime. Justice Quarterly , 16(3), 633–654.

56.

Schreck

C. J.

, & Fisher

B. S.

(2004). Specifying the influence of family and peers on violent victimization: Extending routine activities and lifestyles theories. Journal of Interpersonal Violence , 19(9), 1021–1041.

57.

Shaw

C. R.

, & McKay

H. D.

(1942). Juvenile delinquency and urban areas . Chicago University Press.

58.

Simons

R. L.

, Lei

M. K.

, Beach

S. R.

, Brody

G. H.

, Philibert

R. A.

, & Gibbons

F. X.

(2011). Social environment, genes, and aggression: Evidence supporting the differential susceptibility perspective. American Sociological Review , 76(6), 883–912.

59.

Steinberg

(2000). Adolescence . McGraw-Hill.

60.

Steinberg

(2010). A dual systems model of adolescent risk-taking. Developmental Psychobiology , 52(3), 216–224.

61.

Thornberry

T. P.

(1989). Reflections on the advantages and disadvantages of theoretical integration. In Messner

S. F.

, Krohn

M. D.

& Liska

A. E.

(Eds.), Theoretical integration in the study of deviance and crime: Problems and prospects . State University of New York Press.

62.

Thornberry

T. P.

(2012). Criminological theory. In Loeber

& Welsh

B. C.

(Eds.), The future of criminology (pp. 46–54). Oxford University Press.

63.

Umbach

, Raine

, Gur

R. C.

, & Portnoy

(2017). Neighborhood disadvantage and neuropsychological functioning as part mediators of the race–antisocial relationship: A serial mediation model. Journal of Quantitative Criminology , 34(2), 481–512.

64.

Vazsonyi

A. T.

, Cleveland

H. H.

, & Wiebe

R. P.

(2006). Does the effect of impulsivity on delinquency vary by level of neighborhood disadvantage? Criminal Justice and Behavior , 33(4), 511–541.

65.

Velez

M. B.

(2001). The role of public social control in urban neighborhoods: A multilevel analysis of victimization risk. Criminology , 39(4), 837–864.

66.

Vogel

(2016). The modifiable areal unit problem in person–context research. Journal of Research in Crime and Delinquency , 53(1), 112–135.

67.

Vogel

, & Ham

M. van.

(2018). Unpacking the relationships between impulsivity, neighborhood disadvantage, and adolescent violence: An application of a neighborhood-based group decomposition. Journal of Youth and Adolescence , 47(4), 859–871.

68.

Weisburd

, & Piquero

A. R.

(2008). How well do criminologists explain crime? Statistical modeling in published studies. Crime and Justice , 37(1), 453–502.

69.

Wellford

C. F.

(1989). Towards an integrated theory of criminal behavior. In Messner

S. F.

, Krohn

M. D.

& Liska

A. E.

(Eds.), Theoretical integration in the study of deviance and crime: Problems and prospects (pp. 119–127). State University of New York Press.

70.

Wilcox

Kenneth

C. Land,

& Scott

A. Hunt.

(2003). Criminal Circumstance: A Dynamic Multicontextual Criminal Opportunity Theory . Aldine de Gruter.

71.

Wright

E. M.

, & Fagan

A. A.

(2013). The cycle of violence in context: Exploring the moderating roles of neighborhood disadvantage and cultural norms. Criminology , 51(2), 217–249.

72.

Wright

J. P.

, & Boisvert

(2009). What biosocial criminology offers criminology. Criminal Justice and Behavior , 36(11), 1228–1240.

73.

Xie

, & Baumer

E. P.

(2019). Neighborhood immigrant concentration and violent crime reporting to the police: A multilevel analysis of data from the National Crime Victimization Survey. Criminology , 57(2), 237–267.

74.

Zimmerman

G. M.

(2010). Impulsivity, offending, and the neighborhood: Investigating the person–context nexus. Journal of Quantitative Criminology , 26(3), 301–332.

75.

Zimmerman

G. M.

, & Messner

S. F.

(2011). Neighborhood context and nonlinear peer effects on adolescent violent crime. Criminology , 49(3), 873–903.

76.

Zimmerman

G. M.

, & Posick

(2014). Detecting specialization in interpersonal violence versus suicidal behavior. Journal of Adolescent Health , 55(6), 810–816.

77.

Zimmerman

G. M.

, Botchkovar

E. V.

, Antonaccio

, & Hughes

L. A.

(2015). Low self-control in “bad” neighborhoods: Assessing the role of context on the relationship between self-control and crime. Justice Quarterly , 32(1), 56–84.