APA Article Summary

3-page brief summary in which you summarize the article content, cite the author in text, and provide a reference in APA style. Title page formatted in APA style, one page of text that contains an in-text citation, and a reference page.

Journal of Interpersonal Violence 18(1) 92 –120

© The Author(s) 2013 Reprints and permission:

sagepub.com/journalsPermissions.nav DOI: 10.1177/0886260512448847


448847 JIV18110.1177/0886260512448847Barn es and JacobsJournal of Interpersonal Violence © The Author(s) 2011

Reprints and permission: http://www. sagepub.com/journalsPermissions.nav

1The University of Texas at Dallas, Richardson, TX, USA

Corresponding Author: J. C. Barnes, School of Economic, Political & Policy Sciences, The University of Texas at Dallas, 800 West Campbell Road, Richardson, TX 75080, USA Email: jcbarnes@utdallas.edu

Genetic Risk for Violent Behavior and Environmental Exposure to Disadvantage and Violent Crime: The Case for Gene–Environment Interaction

J. C. Barnes1 and Bruce A. Jacobs1


Despite mounds of evidence to suggest that neighborhood structural factors predict violent behavior, almost no attention has been given to how these influences work synergistically (i.e., interact) with an individual’s genetic propensity toward violent behavior. Indeed, two streams of research have, heretofore, flowed independently of one another. On one hand, criminolo- gists have underscored the importance of neighborhood context in the etiol- ogy of violence. On the other hand, behavioral geneticists have argued that individual-level genetic propensities are important for understanding violence. The current study seeks to integrate these two compatible frameworks by exploring gene–environment interactions (GxE). Two GxEs were examined and supported by the data (i.e., the National Longitudinal Study of Adolescent Health). Using a scale of genetic risk based on three dopamine genes, the analysis revealed that genetic risk had a greater influence on violent behavior when the individual was also exposed to neighborhood disadvantage or when




Barnes and Jacobs 93

the individual was exposed to higher violent crime rates. The relevance of these findings for criminological theorizing was considered.


genetic risk, neighborhood context, violence, gene–environment interaction (GxE)

Biosocial research has blossomed over the past decade (Moffitt, Ross, & Raine, 2011). There are now hundreds of studies that incorporate biological or genetic perspectives into their theoretical propositions (Moffitt, 1993; Walsh, 2002), their statistical models (Barnes, Beaver, & Boutwell, 2011; Burt, 2009; Schilling, Walsh, & Yun, 2011), or their discussion of the future of criminology (Cullen, 2011; DeLisi & Piquero, 2011; Piquero, 2011). As a result, there is now little doubt that biological and genetic risk factors play a key role in the etiology of delinquent, criminal, and antisocial behavior (Raine, 1993). Although biosocial criminology draws on many different, albeit related, perspectives such as evolutionary psychology (Campbell, 2009), biological criminology (Mazur, 2009), and neurocriminology (Raine et al., 2003; Weber, Habel, Amunts, & Schneider, 2008), one focus has gener- ated an impressive body of evidence: behavioral and molecular genetics research (Craig & Halton, 2009; Ferguson, 2010).

As summarized in a number of recent meta-analyses, genetic factors account for a significant portion of the variance in antisocial behavior (Ferguson, 2010; Mason & Frick, 1994; Miles & Carey, 1997; Moffitt, 2005; Raine, 1993; Rhee & Waldman, 2002; Schilling et al., 2011). These analyses are impressive in the consistency with which they estimate the genetic influ- ence on antisocial behavior. To be specific, they reveal that genes are respon- sible for roughly half of the variance in antisocial behavior, with the remaining variance being attributed primarily to the nonshared environment. Spurred by these findings, scholars have begun to explore genetic factors in more detail; researchers are now analyzing the link between specific genes (i.e., genetic polymorphisms) and antisocial behavior (Beaver, DeLisi, Vaughn, & Barnes, 2010; Burt & Mikolajewski, 2009; Craig & Halton, 2009).

The current study seeks to extend contemporary biosocial research by ana- lyzing the link between three genetic polymorphisms and violent criminal behavior. Although this is an important element of the current focus, perhaps just as important is our focus on gene–environment interactions (GxE). The question of whether genetic effects are contingent on environmental



94 Journal of Interpersonal Violence 18(1)

experiences has been under investigation by recent cutting-edge scholarship (Caspi et al., 2002). It is to this literature that we now turn.

Gene–Environment Interaction Behavioral genetics research typically decomposes the variance in a pheno- type into three separate components (i.e., heritability, shared environment, and nonshared environment). As noted above, behavioral geneticists have consistently reported that criminal behavior is around 50% heritable. One limitation of this research, however, is that the specific genetic factors that comprise this heritability estimate are unidentified. In other words, behav- ioral genetics cannot tell us which genes are driving the heritability estimate.

With the recent mapping of the human genome, researchers are beginning to “pull back the heritability curtain” to identify links between measured genes and phenotypic outcomes. This line of research—referred to as molec- ular genetics—has already produced a wealth of insights into these links (Carey, 2003). For example, certain genetic polymorphisms have been linked to various antisocial behaviors such as ADHD (Faraone, Doyle, Mick, & Biederman, 2001), childhood conduct disorder (Beaver, 2009a; Beaver et al., 2007), and adulthood violent behavior (Burt & Mikolajewski, 2009; Craig & Halton, 2009). Perhaps more importantly, molecular genetics research has identified the importance of the environment in triggering genetic effects—a process known as GxE (Shanahan & Hofer, 2005).1 Findings from GxE research show that certain genetic effects are more likely to manifest when combined with environmental risk factors (Beaver, 2008; Cadoret, Yates, Troughton, Woodworth, & Stewart, 1995; Caspi et al., 2002; Moffitt, Caspi, & Rutter, 2006; Thapar, Harold, Rice, Langley, & O’Donovan, 2007).

The logic of GxE notes that the effects of a genetic risk factor on the development of a phenotype (e.g., antisocial behavior) will differ across indi- viduals according to their exposure to environmental risk factors or vice versa. In other words, GxE calls for a nonadditive effect between an environ- mental risk factor and a genetic risk factor in the etiology of antisocial behav- ior. For example, a genetic risk factor may have a small or negligible effect on criminal behavior when a low level of environmental risk is present. However, when environmental risk is increased, the effects of the genetic risk factor are substantially increased.

There is now a sizable body of research examining GxEs in the develop- ment of antisocial phenotypes (Beaver et al., 2007; Caspi et al., 2002; Foley et al., 2004; Haberstick et al., 2005; Jaffee et al., 2005; Kim-Cohen et al., 2006; Vaske, 2009). Caspi and colleagues (2002) were some of the first to test



Barnes and Jacobs 95

the GxE hypothesis—in regards to a behavioral phenotype—with a measured gene. Their findings revealed that males with a particular genotype (i.e., alleles for the MAOA gene linked with low MAOA activity) who were also maltreated as children (i.e., the environmental risk factor) were more likely to have been convicted of a violent offense compared to males who were mal- treated but did not have the genetic risk factor (i.e., respondents who had the high MAOA activity allele). Although only 12% of the sample was exposed to both risk factors (i.e., childhood maltreatment and low MAOA activity allele), these respondents accounted for approximately 44% of all violent convictions in the sample. A recent meta-analysis supported the GxE between MAOA genotype and childhood maltreatment in the prediction of antisocial behavior (Kim-Cohen et al., 2006; Taylor & Kim-Cohen, 2007).

It is worth noting that a slightly different interpretation of GxE findings has been proffered by Belsky, Bakermans-Kranenburg, and van Ijzendoorn (2007). Briefly, Belsky and colleagues (2007) reviewed much of the research into GxE and brought an important point to bear; much of the evidence reveals that individuals with more “risk” alleles are impacted to a greater degree by bad environments. But—and this is where their differential suscep- tibility hypothesis differs from the standard GxE framework—there is also a great deal of evidence revealing that individuals with more risk alleles are impacted to a greater degree by good environments. In other words, Belsky et al. proposed a “plasticity alleles” hypothesis which states that individuals carrying more plasticity alleles (previously referred to as risk alleles) are more influenced by the environment; whether that is for better or for worse. Caspi et al.’s (2002) landmark study reported evidence in support of this hypothesis. Although individuals carrying the low activity MAOA allele dis- played more antisocial behavior when maltreatment was high; they also dis- played fewer antisocial behaviors when not exposed to maltreatment as compared to individuals carrying the high-activity MAOA allele.

Research has tested GxE (and plasticity allele) hypotheses using genetic polymorphisms other than MAOA. One emerging line of research has exam- ined the link between certain dopaminergic genes and antisocial behavior (Beaver, 2009b). Dopamine is a chemical (neurotransmitter) that is found in the brain and is believed to be part of the body’s pleasure and reward center (Beaver, 2009a). Thus, geneticists have hypothesized that polymorphisms in certain dopamine genes may be linked to antisocial behavior via pleasure/ reward pathways in the brain (Rutter, 2006).

Scholars have identified three dopamine genes (DAT1, DRD2, and DRD4) related to antisocial behavior with empirical regularity (Beaver, 2009b; Boutwell & Beaver, 2008; Burt & Mikolajewski, 2009; Guo, Roettger, &



96 Journal of Interpersonal Violence 18(1)

Shih, 2007; Guo, Roettger, & Cai, 2008), and other researchers have begun to identify GxEs with these polymorphisms (Bakermans-Kranenburg & van IJzendoorn, 2006; Sheese, Voelker, Rothbart, & Posner, 2007). For example, Bakermans-Kranenburg and van IJzendoorn (2006) reported that children with the risk allele on DRD4 who were also exposed to insensitive care from their mothers were much more likely to display externalizing behavioral problems compared to other children. Sheese et al. (2007) found that children with the risk allele for DRD4 displayed higher levels of sensation seeking when low-quality parenting was also present. Parenting quality did not affect sensation seeking for children without the risk allele. These findings suggest that the effect of dopaminergic genes in the etiology of antisocial behavior is salient, but that these effects may be contingent on environmental experi- ences. To date, criminological research has not considered the full range of possible GxEs. One group of environmental influences that deserves atten- tion is neighborhood and structural factors.

Structural Factors and Violent Behavior Neighborhood Impacts on Violent Behavior

Criminology has a long history of studying how neighborhood and structural factors affect a person’s propensity for violence (Chauhan, Reppucci, & Turkheimer, 2009; Cloward & Ohlin, 1960; Kornhauser, 1978; Leventhal & Brooks-Gunn, 2003; Sampson, Raudenbush, & Earls, 1997; Shaw & McKay, 1972 [though it must be noted that Shaw & McKay, 1972 spoke only to neighborhood effects and these findings may not apply at the individual level]). Among the most consistent results is that neighborhood disadvantage predicts violent criminal behavior (Pratt & Cullen, 2005). Sampson and his colleagues (1997), drawing on data from Chicago residents and neighbor- hoods, reported that neighborhood measures of concentrated disadvantage were strongly related to an individual’s level of self-reported violent behavior. Importantly, the impact of neighborhood disadvantage on violent behavior was not fully mediated by measures of immigrant concentration, residential stability, or collective efficacy. This finding suggests that neighborhood fac- tors are important in the etiology of violent behavior.

Building on the work of Sampson et al. (1997; as well as a foundation of other research such as Miller, 1958; Shaw & McKay, 1972; Sutherland, 1939), one might expect that individuals who live in areas plagued by high crime rates will have a greater propensity toward violence. This hypothesis is consistent with the above work which has shown that crime rates covary with



Barnes and Jacobs 97

neighborhood disadvantage. Thus, crime rates are associated with neighbor- hood disadvantage, and neighborhood disadvantage has been shown to pre- dict individual-level crime propensity (Sampson et al., 1997). Working backwards from these findings, it is logical to assume that crime rates in a given area will also be predictive of a person’s criminal propensity. Unfortunately, to our knowledge, no study has directly addressed this issue (although it may be argued that Anderson, 1999 highlighted these possibili- ties). Put differently, as obvious as the association appears, criminologists have yet to consider whether aggregate crime rates are predictive of an indi- vidual’s criminal involvement or whether violent crime rates interact with individual propensities to predict criminal behavior. This represents an obvi- ous limitation with the extant literature.

GxE and Neighborhood Influences Despite the consistency with which neighborhood and structural factors have been correlated with an individual’s violent behavior, researchers have long noted that not everyone who grows up in a “bad” neighborhood becomes an offender (Anderson, 1999; Piquero, 2011). Juxtaposing the research on struc- tural factors against the research on GxE, a provocative question begins to emerge (Plomin & Daniels, 1987): Do genetic risk factors interact with neighborhood and county-level risk factors to predict violent criminal behav- ior and if so, what are the underlying mechanisms of this relationship?

Although there is little doubt that neighborhood factors matter, for whom, when, and why they matter is still shrouded in mystery. Biosocial research suggests several mechanisms by which neighborhood factors may interact with genetic risk for offending. Particularly relevant is research that focuses on criminal opportunities and offending likelihood (Cohen & Felson, 1979). In short, a crime can only take place if the opportunity for crime exists. Although it may be argued that criminal opportunities are omnipresent, opportunities for violent crime may be more numerous in neighborhoods that are marked by higher crime rates and greater disadvantage. Typically, these neighborhoods are defined by social structures that may not be as adept at controlling violent crime (Sampson et al., 1997) and may even encourage law-breaking behavior (Keizer, Lindenberg, & Steg, 2008). If individuals are more or less likely to engage in violent behavior as a function of genetic dif- ferences, it is also logical to expect that differences in violent behavior will emerge as a function of the interaction between genetic propensities and neighborhood exposure to opportunities.



98 Journal of Interpersonal Violence 18(1)

As of this writing, only one study offers any insight on this issue. Beaver, Gibson, DeLisi, Vaughn, and Wright (2011) analyzed the effect of two dopa- mine genes (DRD2 and DRD4) on violent delinquency for two sets of respon- dents: (a) respondents who lived in an adequate neighborhood (measured as living at or below the 75th percentile on a measure of neighborhood disadvan- tage); and, (b) respondents who lived in a disadvantaged neighborhood (i.e., living above the 75th percentile). The authors reported that risk alleles on two dopamine genes had a significant influence on violent delinquency, but only if the respondent lived in a disadvantaged neighborhood. The dopamine genes conferred no risk for respondents who lived in an adequate neighborhood. In short, Beaver et al. (2011) identified a compelling relationship between genetic factors and neighborhood disadvantage. Whether this relationship holds when different measures of genetic risk are examined and when different types of environmental measures are utilized remains to be seen.

Current Study In his recent Sutherland Address to the American Society of Criminology, Francis Cullen (2011, p. 311) warned that, “we can no longer pretend that biology is not intimately implicated in human behavior and thus in criminal behavior.” He added that the challenge for traditional criminologists is “to put out the welcome mat to crime scientists and to understand that the future of criminology will be advanced by exploring systematically the nexus between propensity and opportunity—between offender and situation” (p. 315). The current study takes up this challenge by examining whether genetic influ- ences on violent behavior are contingent on exposure to environmental structures that are conducive to violence.

Our efforts advance similar work spearheaded by Beaver et al. (2011) but differ from that research in three important respects. First, we employ an alternate strategy for measuring genetic risk. Unlike Beaver et al.—who examined the individual effects of two genes—we compile a genetic risk index based on three genetic markers. This is important because genetic effects may act in concert to confer an increased risk toward violent behavior; an effect that will not be captured by exploring the genes individually. Second, we specifically examine the interaction effect between genetic risk and violent crime rates. Beaver et al. examined only the interaction between genetic risk and neighborhood disadvantage which is useful but limited because neighborhood disadvantage may not capture the same variance as a measure of local crime rates. It is worth pointing out that no research (to our



Barnes and Jacobs 99

knowledge) has considered the relationship between macro-level crime rates and an individual’s criminal (or violent) behavior. And third, we employed an alternative strategy for testing GxE by using split samples as well as multipli- cative interaction terms. This strategy avoids the problems of multicollinear- ity (by using split samples) and reduced sample sizes (by using the multiplicative interaction). In short, we capitalized on the benefits of both strategies in order to ensure that the findings were not sensitive to one meth- odological approach.

The present study examines three specific hypotheses:

Hypothesis 1: Respondents who carry a genetic risk toward violent behavior will be more likely to report violent behavior as compared to those who do not carry the genetic risk.

Hypothesis 2: The effect of genetic risk on violent behavior will be contingent upon the respondent’s exposure to neighborhood disad- vantage. When disadvantage is high, genetic risk will have a larger effect on violent behavior.

Hypothesis 3: The effect of genetic risk on violent behavior will be contingent upon the respondent’s exposure to violent crime. For respondents who live in areas with high violent crime rates, genetic risk will have a larger effect on violent behavior.

Method Data

The data for this study were drawn from the National Longitudinal Study of Adolescent Health (Add Health; Harris, 2009). These data have been described at length elsewhere (Harris et al., 2009; Kelly & Peterson, 1997) and need not be repeated here. Briefly, the Add Health is a nationally repre- sentative, longitudinal survey of American adolescents who were enrolled in middle and high school during the 1994-1995 academic year. The study began by interviewing all students enrolled in roughly 130 schools across the United States (N ~ 90,000). From this sample of respondents, a subsample of approximately 20,000 were drawn and were administered a more lengthy follow-up survey that was completed in the respondents’ homes (i.e., wave 1). These surveys addressed myriad topics such as the respondent’s health, their personal relationships, and their involvement in delinquent and criminal behaviors. Four waves of data have been collected thus far. Because the cur- rent efforts draw only on wave 1 data, waves 2 through 4 are not discussed.



100 Journal of Interpersonal Violence 18(1)

Two unique features of the Add Health design are utilized by the current analysis. First, the Add Health researchers included a host of neighborhood and county-level measures that can be linked with the individual respondents. These data allow for the examination of contextual effects that may influence violent behavior. The second design feature of the Add Health was the inclu- sion of certain genetic markers for a subset of the participants. Respondents who had a twin or a full sibling participating in the Add Health study were asked to provide buccal cell samples so that they may be genotyped. Originally, 2,574 respondents were genotyped (Cohen et al., n.d.). After elim- inating one twin from each MZ pair (to avoid artificially deflating standard errors), and after eliminating females from the sample (because of males’ over involvement in violence [Beaver et al., 2011]), a final analytic sample of 1,078 was obtained.

Measures Dependent Variable

Violent behavior. During wave 1 interviews, respondents were asked to report the frequency with which they had been involved in a number of vio- lent behaviors over the past 12 months. To be specific, each respondent was asked to indicate how often they had used a weapon to get something from someone, gotten into a group fight, gotten into a serious fight, hurt someone badly enough that they required medical attention, used a weapon in a fight, and taken a weapon to school. Responses to the first four items were coded so that 0 = never, 1 = 1 or 2 times, 2 = 3 or 4 times, and 3 = 5 or more times. Responses to the last two items (weapon fight and weapon school) were coded so that 0 = no and 1 = yes. Factor analysis indicated that all six items hung together on a single factor. Thus, to create the violent behavior scale, responses to the six items were summed together so that higher values indi- cated more involvement in violent behavior (α = .75). Descriptive statistics for this and all other variables utilized in the analysis can be found in Table 1.

Genetic Risk Variable Dopamine risk. A rich line of evidence suggests certain genetic markers

related to dopamine activity are associated with criminal and antisocial behavior (Beaver, 2009b; Craig & Halton, 2009). The Add Health included genotypic information for three dopamine polymorphisms: DAT1, DRD2, and DRD4. To create the dopamine risk scale, a series of four steps was fol- lowed. First, DAT1 is a dopamine transporter gene that has been linked with myriad antisocial behaviors (Schilling et al., 2011). The two most common



Barnes and Jacobs 101

alleles are the 9-repeat and the 10-repeat, with the 10-repeat allele being the risk allele (Gill, Daly, Heron, Hawi, & Fitzgerald, 1997; Rowe et al., 2001). Following prior research (Beaver, Wright, DeLisi, & Vaughn, 2008) we coded each allele so that 0 = 9-repeat allele and 1 = 10-repeat allele. Respon- dents with any other allele were assigned a missing value and were omitted from the analyses (Hopfer et al., 2005).

Second, DRD2 is a dopamine receptor polymorphism that has two differ- ent alleles: the A1 allele and the A2 allele. The A1 allele has been identified as the risk allele (Guo et al., 2007) and was, therefore, coded as 1. The A2 allele was coded as 0. Third, the DRD4 polymorphism is a dopamine receptor gene that has been implicated in antisocial conduct with the 7-repeat allele conferring increased risk as compared to the 4-repeat allele (Faraone et al., 2001; Rowe et al., 2001). Following prior researchers (Beaver et al., 2011), the DRD4 polymorphism was coded so that the 7-repeat allele (along with the 8-, 9-, and 10-repeat alleles) = 1 and the 4-repeat allele (along with the 2-, 3-, 5-, and 6-repeat alleles) = 0.

The fourth step to creating the dopamine risk scale was to sum each respondent’s number of risk alleles for DAT1, DRD2, and DRD4 into a single scale. The polymorphisms were coded codominantly, meaning that

Table 1. Descriptive Statistics for Add Health Males

Frequency Mean SD Minimum Maximum

Violent behavior 1.48 2.23 0 14 Dopamine risk 2.50 1.02 0 6 0 risk alleles 13 1 risk alleles 162 2 risk alleles 371 3 risk alleles 367 4 risk alleles 140 5 risk alleles 22 6 risk alleles 3 Neighborhood disadvantage 0.00 0.94 –1.19 4.61 County violent crime rate 735.19 624.58 24.73 3007.91 Age 15.66 1.69 12 20 Race .18 .38 0 1 Black 201 Non Black 928



102 Journal of Interpersonal Violence 18(1)

the value for each polymorphism reflected the number of risk alleles pres- ent in the respondent. Humans carry two copies of every gene (with the exception of genes located on the sex chromosomes for males—none of the dopamine genes analyzed here are located on a sex chromosome). Thus, when summed together, the dopamine risk scale ranged from a mini- mum of 0 (i.e., no dopamine risk alleles) to a maximum of 6 (i.e., six dopamine risk alleles).

Environmental Variables Neighborhood disadvantage. A compelling line of research has shown that

neighborhood indicators of disadvantage are salient predictors of violent criminal activity (Sampson et al., 1997). To account for these influences, we created an indicator of neighborhood disadvantage that was measured at the block-group level. The block-group level is the smallest level of aggre- gation, making it the most appropriate unit of analysis for estimating neigh- borhood effects. To create the neighborhood disadvantage scale, the following measures (taken from the 1990 U.S. Census) were factor ana- lyzed: the percentage of Black residents, the percentage of female headed households, the percentage of residents with an income under US$15,000, the percentage of residents on public assistance, and the unemployment rate. Factor analysis revealed that the five items were tapping an underlying latent construct. All factor loadings were greater than or equal to .69 and the reliability coefficient was .80. The scale was created using regression scor- ing based on the factor analysis results. Higher values reflected more neigh- borhood disadvantage.

Violent crime rate. The Uniform Crime Reports (UCR) is produced by the Federal Bureau of Investigation each year. These data reflect the total amount of reported crime in each of the 50 states. The Add Health sample included county-level crime rate data drawn from these UCR statistics (1993 data). The violent crime rate variable is a composite variable reflecting the number of robberies, aggravated assaults, rapes, and homicides per 100,000 residents in each county.

Control Variables Age. To control for age differences in violent behavior, the respondent’s

age was included in the statistical analysis. The age variable was coded con- tinuously in years.

Race. To control for any potential race differences in violent behavior, the respondent’s race was controlled with a dichotomous variable coded 0 = non- Black and 1 = Black.



Barnes and Jacobs 103

Analytic Strategy

The analysis unfolded in two interlocking steps. The first step estimated the influence of dopamine risk on self-reported violent behavior, while also including a multiplicative interaction term between the dopamine risk scale and the two environmental measures (i.e., neighborhood disadvantage and violent crime rate). The multiplicative terms were created by mean centering the dopamine risk scale, the neighborhood disadvantage scale, and the violent crime rate and then multiplying the dopamine risk scale by the neighborhood disadvantage scale and by the violent crime rate. Thus, two multiplicative terms were generated: dopamine risk X neighborhood disadvantage and dopa- mine risk X violent crime rate. Two negative binomial models were estimated. The first examined the interaction between dopamine risk and neighborhood disadvantage. The second examined the interaction between dopamine risk and violent crime rate. The coefficient estimate for the interaction terms indicated whether genetic risk was contingent upon environmental exposure to disadvantage/violent crime.

The second step in the analysis also investigated the interaction between the dopamine risk variable and the environmental variables. This portion of the analysis approached the interaction question using a slightly different strategy. To be specific, the effect of dopamine risk on violent behavior was examined after splitting the sample according to scores on the environmental variables. In the first analysis, respondents were split into two groups: those living at or above the 75th percentile on the neighborhood disadvantage scale (i.e., a high degree of disadvantage) and those living below the 75th percentile (i.e., moderate to low levels of neighborhood disadvantage). Once respon- dents were split into the two groups, the negative binomial models were rees- timated, but this time the multiplicative term was omitted. The same strategy was followed in respect to the violent crime rate. The benefit of these models is twofold. First, these models offer a clean way to reconfirm any findings gleaned from the models that employ the multiplicative interaction term. Second, this approach is amenable to a graphical depiction of the effect of dopamine risk on violent behavior at different levels of environmental risk.

Findings Table 2 presents the findings gleaned from the first set of negative binomial models where self-reported violent behavior is the dependent variable. As can be seen, these two models explored the interaction between dopamine risk and the environmental measures by including multiplicative interaction



104 Journal of Interpersonal Violence 18(1)

terms. Model 1 analyzed the interaction between dopamine risk and the neighborhood disadvantage scale. Two findings emerged. First, the coeffi- cient estimate for the dopamine risk scale was positive and statistically sig- nificant.2 The incidence rate ratio revealed, for instance, that a one unit increase in the dopamine risk scale increased the rate of violent behavior by roughly 8% (when neighborhood disadvantage is set to zero—or the mean). The second key finding was that the multiplicative interaction term was moderately (p < .10) significant and the effect was positive. This finding deserves close attention because it suggests an interesting relationship between dopamine risk and neighborhood disadvantage. To be specific, the interaction term indicates that the effect of dopamine risk on violent behavior is contingent on the level of environmental risk that is present. As the envi- ronmental risk increases (i.e., gets more positive), the effect of the dopamine risk also increases.

Model 2 in Table 2 presents the findings from the regression model which explored the interaction between dopamine risk and violent crime rates. Similar to the findings from model 1, model 2 revealed that dopamine risk was positively related to the respondent’s self-reported violent behavior. Importantly, model 2 also revealed that dopamine risk and the violent crime rate interacted such that the effect of dopamine risk on violent behavior was stronger for respondents who lived in high crime counties.

Table 2. Negative Binomial Regression of Self-Reported Violent Behavior on Dopamine Risk and Environmental Risk for Add Health Males

Model 1 Model 2


Age –.04 0.97 .03 –.05* 0.95 .03 Black (=1) .10 1.11 .17 .13 1.14 .13 Dopamine risk .08* 1.08 .05 .07** 1.07 .05 Neighborhood disadvantage .07 1.07 .06 Dopamine risk × neighborhood

disadvantage .07** 1.07 .04

County violent crime rate .0001 1.0001 .0001 Dopamine Risk × county

violent crime rate .0002* 1.0002 .0001

Note: b = unstandardized coefficient; IRR = incidence rate ratio; SE = standard error; Standard errors are clustered by block group in model 1 and by county in model 2. *p < .05, **p < .10 (one-tailed).



Barnes and Jacobs 105

As outlined above, the second step of the analysis involved splitting the sample according to their level of exposure to the environmental risk vari- ables. Presented in Figure 1 is the first set of findings (with the parameter estimates presented in the figure caption) where the sample was split accord- ing to scores on the neighborhood disadvantage scale. The findings from these models are directly in line with the findings from model 1, Table 2. Two points are worth noting. First, the effect of dopamine risk on violent behavior is practically nonexistent for respondents living below the 75th percentile on

0 0 1 2 3 4 5 6








4 Pr

ed ic

te d

Se lf

-R ep

or te

d V

io le

nt B

eh av

io r

Number of Dopamine Risk Alleles

At/Above 75th Percentile

Below 75th Percentile

Figure 1. Predicted scores on the self-reported violent behavior scale for males living at or above the 75th percentile and below the 75th percentile for the neighborhood disadvantage scale Note: Standard errors are clustered by block group; Models control for age, race, and neighborhood disadvantage; Above 75th percentile coefficient estimates: b

Dopamine Risk = .23,

SE Dopamine Risk

= .08, p < .05; Below 75th percentile coefficient estimates: b Dopamine Risk

= .02, SE

Dopamine Risk = .05, p > .05.



106 Journal of Interpersonal Violence 18(1)

the neighborhood disadvantage scale (i.e., moderate to low levels of disad- vantage). Indeed, the coefficient estimate for the dopamine risk scale was not statistically different from zero in this model (see note of Figure 1). The sec- ond finding to note is that the dopamine risk scale had a strong positive effect on violent behavior for respondents living at or above the 75th percentile on the neighborhood disadvantage scale. For respondents living at or above the 75th percentile, there was an approximately 200% increase in self-reported violent behavior between respondents with minimal dopamine risk (i.e., one risk allele) and those with high dopamine risk (i.e., six risk alleles). To be specific, respondents with 1 risk allele were predicted to report 1.16 violent acts while respondents with six risk alleles were predicted to report 3.60 vio- lent acts. A coefficient comparison test (Paternoster, Brame, Mazerolle, & Piquero, 1998) indicated that the coefficient estimate for the dopamine risk scale was statistically different across the two models (z = 2.23; p < .05, one-tailed).

The next set of findings is presented in Figure 2 (with the parameter esti- mates presented in the figure caption). This figure depicts the findings from the analyses where the sample was split according to scores on the violent crime rate measure. As before, respondents were split at the 75th percentile and separate regression models were estimated. The findings in Figure 2 are consistent with the findings from Table 2, model 2. In particular, dopamine risk had no effect on violent behavior for respondents living in moderate to low crime counties (i.e., below the 75th percentile). Dopamine risk did, how- ever, increase violent behavior for respondents living in high-crime areas (i.e., at or above the 75th percentile). On one hand, respondents living in high-crime areas who had no risk alleles were predicted to report less than one violent act (predicted rate = .91). On the other hand, respondents with five risk alleles were predicted to report 3.13 violent acts (there were no respondents with six risk alleles who lived at or above the 75th percentile for the violent crime rate). A coefficient comparison test indicated that the effect of the dopamine risk scale was significantly different across the two models (z = 2.47; p < .05, one-tailed). Each of the findings outlined in this section are placed within the broader theoretical context in the next section.

Discussion The current study tested three hypotheses. The first hypothesis stated that respondents who carry a genetic risk toward violent behavior would be more likely to report violent behavior as compared to those who did not carry the genetic risk. The results of the analysis supported this hypothesis by revealing



Barnes and Jacobs 107

that individuals with more risk alleles on the dopamine risk scale were more likely to report violent behavior. The second hypothesis argued that the effect of genetic risk on violent behavior would be contingent upon the respon- dent’s exposure to neighborhood disadvantage. This hypothesis was tested in two ways. First, a multiplicative interaction was entered into the regression model and the results supported the hypothesis; the interaction term was positive and statistically significant. Second, the sample was split at the 75th percentile and main effects models were estimated. These models (the results








3.5 Pr

ed ic

te d

Se lf

-R ep

or te

d V

io le

nt B

eh av

io r

At/Above 75th Percentile

Below 75th Percentile

0 1 2 3 4 5 6 Number of Dopamine Risk Alleles

Figure 2. Predicted scores on the self-reported violent behavior scale for males living at or above the 75th percentile and below the 75th percentile for the violent crime rate Note: Standard errors are clustered by county; Models control for age, race, and violent crime rate; Above 75th percentile coefficient estimates: b

Dopamine Risk = .25, SE

Dopamine Risk = .08, p < .05;

Below 75th percentile coefficient estimates: b Dopamine Risk

= .003, SE Dopamine Risk

= .06, p > .05.



108 Journal of Interpersonal Violence 18(1)

of which were plotted in Figure 1) also revealed support for Hypothesis 2: respondents who lived in disadvantaged neighborhoods were most likely to report violence if they also had genetic risk factors.

The third hypothesis noted that the effect of genetic risk on violent behav- ior would be contingent on the respondent’s exposure to violent crime. This hypothesis was tested using two different approaches. First, a multiplicative interaction term was used and the results gleaned from this model supported the notion that respondents who lived in violent areas and had a genetic risk toward violent behavior were most likely to report violence. The second way in which Hypothesis 3 was tested was with split sample models (split at the 75th percentile) and the results obtained from these models were also sup- portive of the hypothesis (results are plotted in Figure 2). In summary, the current study revealed evidence to support a GxE between a dopamine risk scale (a scale which indexed the number of risk alleles on three dopamine genes carried by each respondent) and the respondent’s exposure to neigh- borhood disadvantage and county-level violent crime rates.

It is worth mentioning, however, that the current findings did not conform to the differential susceptibility hypothesis (Belsky et al., 2007), although this does not disprove that hypothesis. The differential susceptibility hypoth- esis suggests that a “cross-over” effect will be observed in the data (see Simons et al., 2011, 2012). Although our data reveal a cross-over effect, this effect necessitates a different interpretation: those with the lowest genetic risk who lived in the worst neighborhoods displayed less violence as compared to all other respondents. The differential susceptibility hypothesis expects indi- viduals with the highest genetic risk to display more violent behavior in high- risk environments but less violent behavior in low-risk environments as compared to other individuals carrying less or no genetic risk. This hypothe- sis would have been supported had the main effect term (in the multiplicative models) for the dopamine risk scale emerged as having a negative impact on violent behavior. Alternatively, this hypothesis would have been supported had the dopamine risk scale exhibited a negative effect for individuals living below the 75th percentile and a positive effect for individuals living at or above the 75th percentile. Note, however, that the current study was not intended to be a test of the differential susceptibility hypothesis, so the lack of supportive evidence should not be taken as negative evidence for that per- spective. This is important because the differential susceptibility hypothesis necessitates an environmental continuum ranging from “good” to “bad.” Our environmental measures, however, were operationalized as “bad” (i.e., at or above 75th percentile) and “not bad” (below 75th percentile) which may reflect a different phenomenon.



Barnes and Jacobs 109

Limitations to the analysis must be discussed. First, although a link between dopamine genes and violent behavior has been highlighted by prior work (e.g., Guo et al., 2008), and some scholars have reported dopamine X environment interactions (Beaver et al., 2011), the exact mechanisms under- lying these relationships are not well understood. Research suggests that dopamine is part of the body’s pleasure/reward center (Rutter, 2006), but this does not describe why dopaminergic activity is related to violent behavior. Researchers should prioritize studies that seek to answer this question. A sec- ond limitation is that only three dopamine genes were analyzed. Indeed, the human genome is believed to include approximately 25,000 genes, leaving much to be learned about the integration of genetics into criminology. Finally, we were unable to directly specify the interactional nature of dopamine risk and the two environmental risk measures (Shanahan & Hofer, 2005). In other words, how does exposure to neighborhood disadvantage and violent crime affect genetic risk? Future work must seek to understand the forces behind this relationship. The remainder of our discussion is devoted largely to this issue.

If predisposition to violence is a switch that must be “tripped” by contex- tual factors before it can exert an influence over behavior (Pinker, 2002), the precise mechanism by which neighborhood disadvantage trips this switch is unclear. Prior research suggests that it may be rooted in one of two dynamics: contextual triggering or social context as a social control (Beaver et al., 2011). The contextual triggering explanation holds that stressful environ- ments cause specific genotypes to be expressed. Thus, Caspi et al. (2002) found that severe child maltreatment was associated with particular genetic expressions not discovered among participants who were not maltreated; both the maltreatment and the genotypes were subsequently associated with antisocial conduct. The implication is that the stress of severe maltreatment caused the effects of the particular genotype to surface, which in turn facili- tated the antisocial conduct. The second explanation, rooted in social control, implies that the effects of particular genotypes are inhibited from formation when predisposed persons are adequately monitored and supervised. In the absence of adequate monitoring, the genotype has a greater likelihood of sur- facing, resulting in antisocial behavior.3

Although our research cannot pinpoint the precise mechanism for the observed relationship between violent crime, neighborhood disadvantage, and genotypes, deconstructing the relationship may be more important to developing insights into its etiology than anything else. Only then can researchers better specify the factors that moderate and mediate the relation- ship. We believe microstructural, subcultural, situational, and existential pro- cesses to be implicated.4



110 Journal of Interpersonal Violence 18(1)

Beginning with microstructure, we make the simple observation that dis- advantaged neighborhoods are generally not pleasant places to be. Physical decay abounds. Disorder is visual and widespread. The threat of predation is palpable. Singly and in combination, these and related forces foment fear (see, for example, Winkel, 1998). Fear breeds insularity, and insularity likely encourages residents to retreat to within-network ties generally high in cohe- sion (for an overview of the relationship between social ties and crime fear, see Gibson, Zhao, Lovrich, & Gaffney, 2002). But strong ties are not neces- sarily protective. Some of the most lethal violence the world over is commit- ted between people who know one another, often intimately. As retreat to such ties increases, the frequency and duration of contact increases and so does the likelihood of disputatiousness (on disputatiousness, see Luckenbill & Doyle, 1989). Undermining disputants’ ability to cope is resource depriva- tion (Agnew, 1992), thereby lowering the flashpoint for violence. Risk ver- sions of dopaminergic alleles may amplify this problem by promoting low frustration tolerance, agitation, insensitivity, poor problem-solving skills, and hostile attribution bias (LaHoste et al., 1996). The ability of affected indi- viduals to resolve conflicts nonviolently may be compromised just as the need for peaceful conflict resolution rises.

The link between genotypes, neighborhood disadvantage, and violence is likely also rooted in subculture. Prior research suggests that residents in highly disadvantaged communities feel a profound sense of procedural and distributive injustice (Downing, 2011). More specifically, acute resource shortfalls give rise to perceptions of absolute and relative deprivation. Relative deprivation is especially probative of violence in disadvantaged neighbor- hoods because it promotes displays of one-upmanship to assert status among similarly positioned others (Anderson, 1999). Such displays constitute both a putdown and a provocation to those on the receiving end by casting those persons as inferior (Jacobs, Topalli, & Wright, 2000). Such displays convey double rejection: Not only has one failed in mainstream society, she or he also cannot “measure up” against similarly situated others. Subcultural norms, epitomized by the “code of the street” (Anderson, 1999), take natural sensi- tivity to rejection and amplify it by directing those slighted to respond vigor- ously to all affronts, big and small. Respect is currency in disadvantaged neighborhoods, so one must not only advance respect whenever possible but defend it vigorously whenever it comes into question. Because rejection may promote imbalances in brain chemistry governed partially by the genes explored here (see generally, Way, Taylor, & Eisenberger, 2009), affected individuals may be more likely to engage in compensatory behavior. In dis- advantaged neighborhoods, such behavior frequently involves aggression,



Barnes and Jacobs 111

which permits affected individuals to “slough off” negative affect and restore a semblance of equilibrium, however temporary it may be (Brown & Gershon, 1993).

Situational forces also are implicated in the genes/disadvantage/violence link. Criminologists have consistently found that neighborhoods with high levels of disadvantage present widespread opportunities for predatory vio- lence (Sampson, Morenoff, & Gannon-Rowley, 2002). Part of the reason is rooted in weak informal social control, which creates attractive, guardian- free targets among impulsive offenders looking for a quick score. Many such opportunities emerge serendipitously, and serendipity hinges on the situated recognition and exploitation of novel cues (Jacobs, 2010). The desire for nov- elty seeking appears to be mediated by dopaminergic genes (Zald et al., 2008; but see Burt, McGue, Iacono, Comings, & MacMurray, 2002), which dictates how cues are perceived, how they are assessed, and how they promote action to exploit them. Moreover, persistent need in disadvantaged settings erodes the ability of affected individuals to exercise inhibition over time in response to novelty (see Gottfredson & Hirschi, 1990). This is important because dopamine genes have been linked to deficits in inhibition and conduct disor- der (Beaver et al., 2007; Boutwell & Beaver, 2008; Retz, Rosler, Supprian, Retz-Junginger, & Thome, 2003). Lowered inhibitions coupled with desires for novelty likely reduce the deterrent effect of threatened sanctions (Gottfredson & Hirschi, 1990; Wilson & Herrnstein, 1985), especially in localities where formal authority is not respected. This certainly is the case in many disadvantaged communities, where defiance often is more common than deterrence. Disrespect for authority promotes noncompliance to threat- ened sanctions (Sherman, 1993) which in combination with a target-rich environment, likely liberates a nontrivial amount of predatory violence.

Then there are existential processes. Numerous studies have shown that dopamine levels are correlated with violent behavior (Ferguson & Beaver, 2009). Indeed, violence can promote a seductive thrill that produces measur- able rises in mood-enhancing brain chemicals (Raine, 1993). Some violent offenders operating in disadvantaged communities specifically implicate “the rush” of aggression as a motivating factor in their behavior (Jacobs, 2000). Explosive violence is then applauded and encouraged by the street code, which bolsters its reinforcement potential (Anderson, 1999). The more obvi- ous existential link between dopaminergic genes, violence, and neighborhood disadvantage is indirect—coming through illicit drug use itself. Few socio- logical variables have been more robustly linked to illegal drug consumption than neighborhood disadvantage (Boardman, Finch, Ellison, Williams, & Jackson, 2001; Galea, Ahern, & Vlahov, 2003). The extent to which an



112 Journal of Interpersonal Violence 18(1)

available, accessible, and relatively cheap supply of illicit drugs “creates” drug use is unclear, but in the presence of at-risk dopamine alleles, the nega- tive synergy cannot be ignored. Dopamine imbalances have long been impli- cated in addictive drug use. Scholars suggest that drugs may compensate for biological deficits in dopamine production or uptake for some users (Filbey et al., 2008; Volkow, Fowler, Wang, & Swanson, 2004). Once use begins and escalates, the propensity for psychopharmacological, economically compul- sive, and systemic aggression rises proportionately (Goldstein, 1985). Because most of this violence occurs, by definition, among criminal disputants, griev- ances will tend to be resolved informally through reprisal. Retaliation has a strong tendency to promote counter-retaliation, which promotes conflict spi- rals and a self-reinforcing cycle of instability that compounds neighborhood disadvantage and heightens the importance of reacting violently to conditions that violate (Anderson, 1999). This feedback loop then weakens collective efficacy and leads to yet more instability. Even if a small proportion of affected individuals turn to drug use because of some combination of disad- vantage and genotype, the synergy between drugs and violence and the pro- clivity for discrete disputes to trigger retaliation and then to spread in contagion-like fashion across disadvantaged communities likely promotes a disproportionate amount of serious violent crime.

The link between biological risk factors, neighborhood disadvantage, and violence is multisourced. Whether context or biology triggers them is less important than their symbiotic and contingent relationship. Such factors interact to potentiate violence in multiple and complex ways.


This research uses data from Add Health, a program project directed by Kathleen Mullan Harris and designed by J. Richard Udry, Peter S. Bearman, and Kathleen Mullan Harris at the University of North Carolina at Chapel Hill, and funded by grant P01-HD31921 from the Eunice Kennedy Shriver National Institute of Child Health and Human Development, with cooperative funding from 23 other federal agencies and foundations. Special acknowledgment is due Ronald R. Rindfuss and Barbara Entwisle for assistance in the original design. Information on how to obtain the Add Health data files is available on the Add Health website (http://www.cpc.unc.edu/ addhealth). No direct support was received from grant P01-HD31921 for this analysis.

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.



Barnes and Jacobs 113

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


1. It should be noted that GxEs are not limited to triggering effects. As discussed by Shanahan and Hofer (2005), GxEs can also work by blunting genetic influences, by allowing genetic influences to flourish, and by enhancing adaptation.

2. It is important to point out that statisticians disagree over whether the main effect terms should be interpreted in a multiplicative interaction model (McClendon, 2002). We have included an interpretation of the main effects here because the coefficient estimates were consistent with a model that did not include the mul- tiplicative term (i.e., the coefficient estimate reported in the tables was simi- lar [within .01] to the coefficient estimate retrieved from a main effects only model—although the standard errors were slightly smaller in the interaction model). Also, the bivariate correlation between the dopamine risk scale and the violent behavior scale was r = .06 (p < .10).

3. We acknowledge an anonymous reviewer, who insightfully pointed out that mal- treated individuals could be appropriately monitored—creating the simultaneous operation of contextual triggers and inhibiting forces. Furthermore, it is unclear from previous theoretical statements whether parental maltreatment is concep- tually distinct from inappropriate monitoring (i.e., could parental maltreatment manifest in the form of inappropriate monitoring?). It is our position that parental maltreatment and inappropriate monitoring represent distinct behavioral phenom- ena that, therefore, may occur simultaneously, or independently. We advocate for more refined measures of parental control and maltreatment so that future research- ers can address this more precisely. As noted by the blind reviewer, one could divide a sample into those who were both maltreated and subjected to inappropri- ate monitoring, and those who were subjected to only one or the other in order to disentangle which of these two conditions are more prominent in their effect.

4. We thank an anonymous reviewer for pointing out that these four areas are not necessarily mutually exclusive. We offer them simply to sensitize readers to the nuanced variation in the mechanisms that likely mediate/moderate the GxE rela- tionship. Such nuances permit theoretical refinement by future research.


Agnew, R. (1992). Foundation for a general strain theory of delinquency. Criminol- ogy, 30, 47-87.

Anderson, E. (1999). Code of the street: Decency, violence and the moral life of the inner city. New York, NY: Norton.



114 Journal of Interpersonal Violence 18(1)

Bakermans-Kranenburg, M. J., & van IJzendoorn, M. H. (2006). Gene-environment interaction of the dopamine D4 receptor (DRD4) and observed maternal insensi- tivity predicting externalizing behavior in preschoolers. Developmental Psycho- biology, 48, 406-409.

Barnes, J. C., Beaver, K. M., & Boutwell, B. B. (2011). Examining the genetic under- pinnings to Moffitt’s developmental taxonomy: A behavioral genetic analysis. Criminology, 49, 923-954.

Beaver, K. M. (2008). The interaction between genetic risk and childhood sexual abuse in the prediction of adolescent violent behavior. Sexual Abuse: A Journal of Research and Treatment, 20, 426-443.

Beaver, K. M. (2009a). Biosocial Criminology: A Primer. Dubuque, IA: Kendall/Hunt. Beaver, K. M. (2009b). Molecular genetics and crime. In A. Walsh & K. M. Beaver

(Eds.), Biosocial criminology: New Directions in theory and research (pp. 50-72). New York, NY: Routledge.

Beaver, K. M., DeLisi, M., Vaughn, M. G., & Barnes, J. C. (2010). Monoamine oxi- dase A genotype is associated with gang membership and weapon use. Compre- hensive Psychiatry, 51, 130-134.

Beaver, K. M., Gibson, C. L., DeLisi, M., Vaughn, M. G., & Wright, J. P. (2011). The interaction between neighborhood disadvantage and genetic factors in the predic- tion of antisocial outcomes. Youth Violence and Juvenile Justice, 10, 25-40.

Beaver, K. M., Wright, J. P., DeLisi, M., & Vaughn, M. G. (2008). Desistance from delinquency: The marriage effect revisited and extended. Social Science Research, 37, 736-752.

Beaver, K. M., Wright, J. P., DeLisi, M., Walsh, A., Vaughn, M. G., Boisvert, D., & Vaske, J. (2007). A gene X gene interaction between DRD2 and DRD4 is asso- ciated with conduct disorder and antisocial behavior in males. Behavioral and Brain Functions, 3, 30.

Belsky, J., Bakermans-Kranenburg, M. J., & van Ijzendoorn, M. H. (2007). For better and for worse: Differential susceptibility to environmental influences. Current Directions in Psychological Science, 16, 300-304.

Boardman, J. D., Finch, B. K., Ellison, C. G., Williams, D. R., & Jackson, J. S. (2001). Neighborhood disadvantage, stress, and drug use among adults. Journal of Health and Social Behavior, 42, 151-165.

Boutwell, B. B., & Beaver, K. M. (2008). A biosocial explanation of delinquency abstention. Criminal Behaviour and Mental Health, 18, 59-74.

Brown, A. S., & Gerson, S. (1993). Dopamine and depression. Journal of Neural Transmission, 91, 75-109.

Burt, S. A. (2009). Rethinking environmental contributions to child and adolescent psychopathology: A meta-analysis of shared environmental influences. Psycho- logical Bulletin, 135, 608-637.



Barnes and Jacobs 115

Burt, S. A., McGue, M., Iacono, W., Comings, D., & MacMurray, J. (2002). An exam- ination of the association between DRD4 and DRD2 polymorphisms and person- ality traits. Personality and Individual Differences, 33, 849-859.

Burt, S. A., & Mikolajewski, A. J. (2009). Preliminary evidence that specific candi- date genes are associated with adolescent-onset antisocial behavior. Aggressive Behavior, 34, 437-445.

Cadoret, R. J., Yates, W. R., Troughton, E., Woodworth, G., & Stewart, M. A. (1995). Genetic- environment interaction in the genesis of aggressivity and conduct dis- orders. Archives of General Psychiatry, 52, 916-924.

Campbell, A. (2009). Gender and crime: An evolutionary perspective. In A. Walsh & K. M. Beaver (Eds.), Biosocial criminology: New Directions in theory and research (pp. 117-136). New York, NY: Routledge.

Carey, G. (2003). Human genetics for the social sciences. Thousand Oaks, CA: Sage. Caspi, A., McClay, J., Moffitt, T. E., Mill, J., Martin, J., Craig, I. W., . . . Poulton, R.

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

Chauhan, P., Reppucci, N. D., & Turkheimer, E. N. (2009). Racial differences in the associations of neighborhood disadvantage, exposure to violence, and criminal recidivism among female juvenile offenders. Behavioral Sciences and the Law, 27, 531-552.

Cloward, R. A., & Ohlin, L. E. (1960). Delinquency and opportunity: A theory of delinquent gangs. New York, NY: Free Press.

Cohen, L. E., & Felson, M. (1979). Social change and crime rate trends: A routine activities approach. American Sociological Review, 44, 588-608.

Cohen, M. S., Feng, Q., Florey, F. E., Ford, C. A., Harris, K. M., Hewitt, J. K., Hobbs, M. M., . . .Udry, J. R. (n.d.). Biomarkers in Wave III of the Add Health Study. The Add Health Biomarker Team.

Craig, I. W., & Halton, K. E. (2009). Genetics of human aggressive behavior. Human Genetics, 126, 101-113.

Cullen, F. T. (2011). Beyond adolescence-limited criminology: Choosing our future— the American Society of Criminology 2010 Sutherland Address. Criminology, 49, 287-330.

DeLisi, M., & Piquero, A. R. (2011). New frontiers in criminal careers research, 2000- 2011: A state-of-the-art review. Journal of Criminal Justice, 39, 289-301.

Downing, S. K. (2011). Street justice: A theoretical paradigm. Criminal Justice Review, 14, 1125-1147.

Faraone, S. V., Doyle, A. E., Mick, E., & Biederman, J. (2001). Meta-analysis of the association between the 7-repeat allele of the dopamine D4 receptor gene and attention deficit hyperactivity disorder. American Journal of Psychiatry, 158, 1052-1057.



116 Journal of Interpersonal Violence 18(1)

Ferguson, C. J. (2010). Genetic contributions to antisocial personality and behavior: A meta-analytic review from an evolutionary perspective. Journal of Social Psy- chology, 150, 160-180.

Ferguson, C. J., & Beaver, K. M. (2009). Natural born killers: The genetic origins of extreme violence. Aggression and Violent Behavior, 14, 286-294.

Filbey, F. M., Ray, L., Smolen, A., Claus, E. D., Audette, A., & Hutchinson, K. E. (2008). Differential neural response to alcohol priming and alcohol taste cues is associated with DRD4 VNTR and OPRM1 genotypes. Alcoholism: Clinical & Experimental Research, 32, 1113-1123.

Foley, D. L., Eaves, L. J., Wormley, B., Sillberg, J. L., Maes, H. H., Kuhn, J., & Riley, B. (2004). Childhood adversity, Monoamine Oxidase A genotype, and risk for conduct disorder. Archives of General Psychiatry, 61, 738-744.

Galea, S., Ahern, J., & Vlahov, D. (2003). Contextual determinants of drug use risk behavior: A theoretical framework. Journal of Urban Health, 80, iii50-iii58.

Gibson, C. L., Zhao, J., Lovrich, N. P., & Gaffney, M. F. (2002). Social integration, individual perceptions of collective efficacy, and fear of crime in three cities. Jus- tice Quarterly, 19, 537-564.

Gill, M., Daly, G., Heron, S., Hawi, Z., & Fitzgerald, M. (1997). Confirmation of association between attention deficit hyperactivity disorder and a dopamine trans- porter polymorphism. Molecular Psychiatry, 2, 311-313.

Goldstein, P. J. (1985). The drugs/violence nexus: A tripartite conceptual framework. Journal of Drug Issues, 15, 493-506.

Gottfredson, M. R., & Hirschi, T. (1990). A General Theory of Crime. Stanford, CA: Stanford University Press.

Guo, G., Roettger, M. E., & Cai, T. (2008). The integration of genetic propensities into social control models of delinquency and violence among male youths. American Sociological Review, 73, 543-568.

Guo, G., Roettger, M. E., & Shih, J. C. (2007). Contributions of the DAT1 and DRD2 Genes to Serious and Violent Delinquency among Adolescents and Young Adults. Human Genetics, 121, 125-136.

Haberstick, B. C., Lessem, J. M., Hopfer, C. J., Smolen, A., Ehringer, M. A., Timberlake, D., & Hewitt, J. K. (2005). Monoamine Oxidase A (MAOA) and antisocial behaviors in the presence of childhood and adolescent maltreatment. American Journal of Medical Genetics Part B (Neuropsychiatric Genetics), 135, 59-64.

Harris, K. M. (2009). The National Longitudinal Study of Adolescent Health (Add Health), Waves I & II, 1994–1996; Wave III, 2001–2002; Wave IV, 2007-2009 [machine-readable data file and documentation]. Chapel Hill, NC: Carolina Popu- lation Center, University of North Carolina at Chapel Hill.



Barnes and Jacobs 117

Harris, K. M., Halpern, C. T., Whitsel, E., Hussey, J., Tabor, J., Entzel, P., & Udry, J. R. (2009). The National Longitudinal Study of Adolescent Health: Research Design. Retrieved from http://www.cpc.unc.edu/projects/addhealth/ design

Hopfer, C. J., Timberlake, D., Haberstick, B., Lessem, J. M., Ehringer, M. A., Smolen, A., & Hewitt, J. K. (2005). Genetic influences on quantity of alcohol consumed by adolescents and young adults. Drug and Alcohol Dependence, 78, 187-193

Jacobs, B. A. (2000). Robbing Drug Dealers: Violence beyond the Law. New York, NY: Aldine de Gruyter.

Jacobs, B. A. (2010). Serendipity in robbery target selection. British Journal of Crimi- nology, 50, 514-529.

Jacobs, B. A, Topalli, V., & Wright, R. (2000). Managing retaliation: Drug robbery and informal sanction threats. Criminology, 38, 171-198.

Jaffee, S. R., Caspi, A., Moffitt, T. E., Dodge, K. A., Rutter, M., Taylor, A., & Tully, L. A. (2005). Nature x nurture: Genetic vulnerabilities interact with physi- cal maltreatment to promote conduct problems. Development and Psychopathol- ogy 17, 67-84.

Keizer, K., Lindenberg, S., & Steg, L. (2008). The spreading of disorder. Science, 322, 1681-1685.

Kelly, M. S., & Peterson, J. L. (1997). The National Longitudinal Study of Adolescent Health (Add Health), Waves I & II, 1994-1996: A user’s guide to the machine- readable files and documentation (Data Sets 48-50, 98, A1-A3). Los Altos, CA: Sociometrics Corporation, American Family Data Archive.

Kim-Cohen, J., Caspi, A., Taylor, A., Williams, B., Newcombe, R., Craig, I. W., & Moffitt, T. E. (2006). MAOA, maltreatment, and gene-environment interaction predicting children’s mental health: New evidence and a meta-analysis. Molecu- lar Psychiatry, 11, 903-913.

Kornhauser, R. R. (1978). Social sources of delinquency: An appraisal of analytic models. Chicago, IL: University of Chicago Press.

LaHoste, G. J., Swanson, J. M., Wigal, S. B., Glabe, C., Wigal, T., King, N., & Kennedy, J. L. (1996). Dopamine D4 receptor gene polymorphism is associated with attention deficit hyperactivity disorder. Molecular Psychiatry, 1, 121-24.

Leventhal, T., & Brooks-Gunn, J. (2003). Children and youth in neighborhood con- texts. Current Directions in Psychological Sciences, 12, 27-31.

Luckenbill, D. F., & Doyle, D. P. (1989). Structural position and violence: Developing a cultural explanation. Criminology, 27, 416-436.

Mason, D. A., & Frick, P. J. (1994). The heritability of antisocial behavior: A meta- analysis of twin and adoption studies. Journal of Psychopathology and Behav- ioral Assessment, 16, 301-323.



118 Journal of Interpersonal Violence 18(1)

Mazur, A. (2009). Testosterone and violence among young men. In A. Walsh & K. M. Beaver (Eds.), Biosocial criminology: New Directions in theory and research (pp. 190-204). New York, NY: Routledge.

McClendon, M. J. (2002). Multiple regression and causal analysis. Long Grove, IL: Waveland Press.

Miles, D. R., & Carey, G. (1997). Genetic and environmental architecture of human aggression. Journal of Personality and Social Psychology, 72, 207-217.

Miller, W. B. (1958). Lower class culture as a generating milieu of gang delinquency. Journal of Social Issues, 14, 5-19.

Moffitt, T. E. (1993). Adolescence-limited and life-course persistent antisocial behav- ior: A developmental taxonomy. Psychological Review, 100, 674-701.

Moffitt, T. E. (2005). The new look of behavioral genetics in developmental psy- chopathology: Gene-environment interplay in antisocial behaviors. Psychological Bulletin, 131, 533-554.

Moffitt, T. E., Caspi, A., & Rutter, M. (2006). Measured gene-environment interac- tions in psychopathology: Concepts, research strategies, and implications for research, intervention, and public understanding of genetics. Perspectives on Psy- chological Science, 1, 5-27.

Moffitt, T. E., Ross, S., & Raine, A. (2011). Crime and biology. In J. Q. Wilson & J. Petersilia (Eds.), Crime and public policy (pp. 53-87). New York, NY: Oxford.

Paternoster, R., Brame, R., Mazerolle, P., & Piquero, A. R. (1998). Using the correct statistical test for the equality of regression coefficients. Criminology, 36, 859-866.

Pinker, S. (2002). The blank slate: The modern denial of human nature. New York, NY: Penguin.

Piquero, A. R. (2011). James Joyce, Alice in Wonderland, the Rolling Stones, and criminal careers. Journal of Youth and Adolescence, 40, 761-775.

Plomin, R., & Daniels, D. (1987). Why are children in the same family so different from one another? Behavioral and Brain Sciences, 10, 1-16.

Pratt, T. C., & Cullen, F. T. (2005). Assessing macro-level predictors and theories of crime: A meta-analysis. Crime and Justice, 32, 373-450.

Raine, A. (1993). The psychopathology of crime: Criminal behavior as a clinical disorder. San Diego, CA: Academic Press.

Raine, A., Lencz, T., Taylor, K., Hellige, J. B., Bihrle, S., Lacasse, L., Lee, M., Ishikawa, S., & Colletti, P. (2003). Corpus callosum abnormalities in psycho- pathic antisocial individuals. Archives of General Psychiatry, 60, 1134-1142.

Retz, W., Rosler, M., Supprian, T., Retz-Junginger, P., & Thome, J. (2003). Dopamine D3 receptor gene polymorphism and violent behavior: Relation to impulsive- ness and ADHD- related psychopathology. Journal of Neural Transmission, 110, 561-572.



Barnes and Jacobs 119

Rhee, S. H., & Waldman, I. D. (2002). Genetic and environmental influences on anti- social behavior: A meta-analysis of twin and adoption studies. Psychological Bul- letin,128, 490-529.

Rowe, D. C., Stever, C., Chase, D., Sherman, S., Abramowitz, A., & Waldman, I. D. (2001). Two dopamine genes related to reports of childhood retrospective inatten- tion and conduct disorder symptoms. Molecular Psychiatry, 6, 429-433.

Rutter, M. (2006). Genes and behavior: Nature-nurture interplay explained. Malden, MA: Blackwell.

Sampson, R. J., Morenoff, J. D., & Gannon-Rowley, T. (2002). Assessing “neighbor- hood effects”: Social processes and new directions in research. Annual Review of Sociology, 28, 443-478.

Sampson, R. J., Raudenbush, S., & Earls, F. (1997). Neighborhoods and violent crime: A multilevel study of collective efficacy. Science, 277, 918-924.

Schilling, C. M., Walsh, A., & Yun, I. (2011). ADHD and criminality: A primer on the genetic, neurobiological, evolutionary, and treatment literature for criminologists. Journal of Criminal Justice, 39, 3-11.

Shanahan, M. J., & Hofer, S. M. (2005). Social context in gene-environment interac- tions: Retrospect and prospect. Journals of Gerontology: Series B, 60B, 65-76.

Shaw, C. R., & McKay, H. D. (1972). Juvenile delinquency and urban areas: A study of rates of delinquency in relation to differential characteristics of local communi- ties in American cities. Chicago, IL: University of Chicago Press.

Sheese, B. E., Voelker, P. E., Rothbart, M. K., & Posner, M. I. (2007). Parenting qual- ity interacts with genetic variation in dopamine receptor D4 to influence tempera- ment in early childhood. Development and Psychopathology, 19, 1039-1046.

Sherman, L. W. (1993). Defiance, deterrence, and irrelevance: A theory of the crimi- nal sanction. Journal of Research in Crime & Delinquency, 30, 445-473.

Simons, R. L., Lei, M. K., Beach, S. R. H., Brody, G. H., Philibert, R. A., & Gibbons, F. X. (2011). Social environment, genes, and aggression: Evidence sup- porting the differential susceptibility perspective. American Sociological Review, 76, 883-912.

Simons, R. L., Lei, M. K., Stewart, E. A., Beach, S. R. H., Brody, G. H., Philibert, R. A., & Gibbons, F. X. (2012). Social adversity, genetic variation, street code, and aggression: A genetically informed model of violent behavior. Youth Violence and Juvenile Justice, 10, 3-24.

Sutherland, E. (1939). Principles of criminology (3rd ed.). Philadelphia, PA: J.B. Lippincott.

Taylor, A., & Kim-Cohen, J. (2007). Meta-analysis of gene-environment interac- tions in developmental psychopathology. Development and Psychopathology, 19, 1029-1037.



120 Journal of Interpersonal Violence 18(1)

Thapar, A., Harold, G., Rice, F., Langley, K., & O’Donovan, M. (2007). The con- tribution of gene-environment interaction to psychopathology. Development and Psychopathology, 19, 989-1004.

Vaske, J. (2009). The role of genes and abuse in the etiology of offending. Unpub- lished Dissertation, University of Cincinnati, Cincinnati, OH.

Volkow, N. D., Fowler, J. S., Wang, G-J, & Swanson, J. M. (2004). Dopamine in drug abuse and addiction: Results from imaging studies and treatment implications. Molecular Psychiatry, 9, 557-569.

Walsh, A. (2002). Biosocial criminology: Introduction and integration. Cincinnati, OH: Anderson.

Way, B. M., Taylor, S. E., & Eisenberger, N. I. (2009). Variation in the μ-opioid recep- tor gene (OPRM1) is associated with dispositional and neural sensitivity to social rejection. Proceedings of the National Academy of Sciences, 106, 15076-15084.

Weber, S., Habel, U., Amunts, K., & Schneider, F. (2008). Structural brain abnormali- ties in psychopaths—A review. Behavioral Sciences and the Law, 26, 7-28.

Wilson, J. Q., & Herrnstein, R. J. (1985). Crime and human nature: The definitive study of the causes of crime. New York, NY: Free Press.

Winkel, F. W. (1998). Fear of crime and criminal victimization: Testing a theory of psychological incapacitation of the ‘stressor’ based on downward comparison processes. British Journal of Criminology, 38, 473-484.

Zald, D. H., Cowan, R. L., Riccardi, P., Baldwin, R. M., Ansari, M. S., Li, R., . . . Kessler, R. M. (2008). Midbrain dopamine receptor availability is inversely associated with novelty-seeking traits in humans. Journal of Neuroscience, 28, 14372-14378.


J. C. Barnes is an assistant professor in the Criminology Department at The University of Texas at Dallas. His research seeks to understand how genetic and environmental factors combine to impact criminological phenomena.

Bruce A. Jacobs is a professor in the Criminology Department at The University of Texas at Dallas. He studies violent crime and offender decision-making.