MAOA,Drug Selling,and Violent Victimization

Abstract

Involvement in drug markets is a significant risk factor for criminal victimization. Separately, the monoamine oxidase A (MAOA) gene has been identified as correlating with risky and antisocial behaviors and moderating the effects of environmental risk factors on antisocial behaviors. Using a sample drawn from the National Longitudinal Study of Adolescent to Adult Health (N = 8,860), we explore whether MAOA genotype moderates the effect of drug selling on violent victimization. Results show that drug selling increases violent victimization among males, but not females. Additionally, the effect of drug selling on violent victimization among males is greater among the carriers of the 2R/3R alleles of MAOA, providing evidence of Gene × Environment interaction. These results appear despite a number of controls that potentially make the drug selling–violent victimization relationship spurious. Implications of the findings are discussed.

Keywords

victimization drug selling MAOA gene–environment interactions

Engagement in drug markets through drug selling is a risk factor for criminal victimization. This is due in large part to the unregulated nature of illicit drug markets, which means that informal social control, in particular violent self-help, is a normative way to deal with interpersonal disputes (Anderson, 2000; Bourgois, 2003; Jacobs, Topalli, & Wright, 2000; Jacques & Wright, 2011; Jacques, Wright, & Allen, 2014; Topalli, Wright, & Fornango, 2002). Yet despite the documented link between drug selling and risks for victimization, little research to date has attempted to identify the factors that might moderate this relationship. In particular, research is lacking that has examined the role biology plays in the relationship between drug selling and victimization.

The current study integrates the literature on drug selling and victimization with a biosocial framework, wherein negative outcomes result from an interaction between an individual’s environment (here, selling drugs) and their genotype. This type of study is needed, as scientific research in a number of fields has shown that genetics matter for a number of life outcomes when combined with the right environmental triggers (Caspi et al., 2002; Guo, Roettger, & Cai, 2008). Specifically, we draw on data from the National Longitudinal Study of Adolescent to Adult Health (Add Health) to examine whether the monoamine oxidase A (MAOA) gene moderates the effect of drug selling on violent victimization while controlling for a number of important variables that potentially make the drug selling–violent victimization relationship spurious. Prior research has shown that MAOA is related to a host of negative outcomes, usually by moderating the effects of key variables in the criminology and victimology literatures (Beaver et al., 2010; Caspi et al., 2002; Foley et al., 2004; Kim-Cohen et al., 2006; Nilsson et al., 2006). We examine these effects separately for males and females for two reasons. First, we anticipate, based on prior research, that males will be much more likely to report drug selling (Adler, 1993) and violent victimization (Lauritsen & Heimer, 2008). Second, this split is important because females carry two copies of the MAOA gene, while males carry one, meaning that any gene–environment interaction (G × E) effects involving MAOA are likely to differ by gender (Caspi et al., 2002). In the following sections, we review the literature on drug selling and victimization and how MAOA fits in with this literature.

Drug Selling and Victimization

There is much support for an association between drug selling and both violent offending and violent victimization (Bourgois, 2003; Curtin & Wendel, 2007; Goldstein, 1985; Small et al., 2013).¹ There is, in fact, an expectation of violence by individuals involved in the illegal drug trade (Wright & Decker, 1997). Goldstein (1985) posits that violence is inherent to the unregulated illicit drug trade. His typology classifies three types of violence in relation to drug markets. These classifications include psychopharmacological, economic-compulsive, and systemic. The first, psychopharmacological, focuses on the pharmacological effects of illicit drugs and how those effects can lead to violence. The second, economic-compulsive, involves the violent ventures that are conducted to obtain the finances needed to procure illicit drugs. Lastly, systemic violence is related to the unregulated nature of illicit drug markets, where there is no form of legitimate conflict resolution available (Curtis & Wendel, 2007; Goldstein, 1985). Systemic violence includes violent action that derives from disputes between distributors, dealers, and buyers. Importantly, dealers may initiate and act as offenders but may also be attacked by others, including buyers, competition, an authority figure in their organization, or an outsider.

Due to the violence implicit in the illegal drug trade, drug dealers are at increased risk for both violent offending and violent victimization. When drug dealers violently offend or are targeted for violence, there is a higher level of danger and risk of lethal violence because of the need to engage in violence as a form of social control. Blumstein (1995a) highlights the relationship between drug markets and violence in association with the use of firearms. Individuals engaged in drug markets arm themselves for self-defense. Blumstein (1995a) argues that individuals fall into an “arms race” that actually escalates violence among participants in illicit drug markets. In one study, Blumstein and Cork (1996) find that increases in the illegal drug trade in New York coincided with an escalation of gun-related homicide among juveniles. An additional layer of risk is added for those involved in the drug trade because robbers and burglars consider drug dealers attractive targets since they are known to have both money and drugs and are considered very unlikely to contact law enforcement to report their victimizations (Wright & Decker, 1997). This latter point means that the drug dealer is reliant on informal social control when they have a grievance, where the response and deterrent is retaliation, which is often violent in nature (Anderson, 2000; Bourgois, 2003; Jacobs et al., 2000; Jacques & Wright, 2011; Jacques, Allen, & Wright, 2014; Topalli et al., 2002). Such retaliation puts drug dealers under the threat of further returned victimization. Through this process, violence becomes a normative mode to the drug dealer, as it develops through common interactive means (Burgois, 2003; Burgois, Prince, & Moss, 2004).

As mentioned above, these risks for a drug dealer are further compounded by the fact that when victimized, they are limited in their options, as reporting a victimization will increase the chances that their own illicit activities are discovered by law enforcement. Furthermore, drug dealers often don’t believe police will be of any assistance in their case (Moskos, 2008). Even if reporting was a more viable option, it would put the drug dealer under threat of further violence from the street element because they could be labeled a “snitch” and face violent victimization due to the application of this label (Rosenfeld, Jacobs, & Wright, 2003). This process of social control is known by those who wish to take advantage of the criminal opportunity by stealing from a drug dealer, thus further increasing the risk of victimization for the dealer. Due to the few legitimate options available for dealing with victimization, those involved in illicit drug markets become attractive targets who respond to violence with violence, forming a vicious cycle of victimization that is hard to break out of. The next section will illustrate how genetics fit in to the drug selling–violent victimization relationship.

MAOA and G × E

Research examining the genetic basis for disparate life outcomes has tended to focus on genes involved in the regulation of neurotransmitter activity in the brain. One gene that has received much attention in the literature is the MAOA gene. The MAOA gene encodes the MAOA enzyme, which metabolizes neurotransmitters, including norepinephrine, serotonin, and dopamine, and plays a key role in regulating behavior (Belsky & Pluess, 2009). Since the MAOA gene is found on the X chromosome, males have only a single copy while females have two copies. Research to date has suggested that it may only be males who are affected by MAOA genotype in regard to risky and antisocial behaviors (Beaver et al., 2010; Simons et al., 2011a).² The findings concerning MAOA and risky and antisocial behaviors have been so consistent that MAOA has been given the moniker of “the warrior gene” (Beaver et al., 2010; Holland & DeLisi, 2014).

Studies, discussed below, looking at the relationship between MAOA and antisocial outcomes tend to find that environmental factors have their most pronounced effects among those carrying the low-activity version of MAOA (the 2R and 3R alleles). In a seminal study looking at G × E effects and MAOA, Caspi et al. (2002) find that young males, in a sample of New Zealanders who are the carriers of the 2R and 3R alleles, were most effected by child maltreatment in regard to their later antisocial behavior and aggression. The males in the sample with high-activity versions of MAOA who had also been the victims of child maltreatment displayed substantially less antisocial behavior later in life. It is important to note, however, that other researchers have failed to replicate the findings of Caspi and colleagues concerning MAOA, child maltreatment, and antisocial behavior utilizing other samples (Haberstick et al., 2005; Young et al., 2006). In another study looking at child maltreatment and MAOA, Kim-Cohen et al. (2006) find that 7-year-old abused boys who carry the low-activity variants of MAOA were rated by their mothers and teachers as having more attention deficits than their abused peers with the high-activity variants of MAOA. More recently, a meta-analysis conducted by Byrd and Manuck (2014) found that across 27 peer-reviewed English-language studies, the low-activity variant of MAOA consistently and significantly moderates the effect of child maltreatment specifically on antisocial outcomes across many different samples.

A number of other studies also reveal a significant interaction between MAOA and environmental adversity. In a large longitudinal study of twin adolescent boys, Foley et al. (2004) find that boys who carry the low-activity variants of MAOA are more likely than their high-activity carrying peers to be diagnosed with conduct disorder when exposed to high levels of adversity during childhood. Nilsson et al. (2006) report similar results in a cross-sectional study, finding that maltreatment and living arrangement experiences were most related to criminal behavior among carriers of the low-activity variants of MAOA. Other studies have produced similar findings concerning MAOA and environmental adversity (Ducci et al., 2008; Widom & Brzustowicz, 2006).

More recent studies have produced further evidence of G × E effects on antisocial outcomes involving MAOA. Simons and colleagues (2011b) find in a sample of African Americans that MAOA (along with 5-HTTLPR and DRD4) moderates the effects of family and community diversity on the adoption of the “street code” and aggressive behavior. Fergusson, Boden, Horwood, Miller, and Kennedy (2011, 2012) find in two separate studies that MAOA moderated the effects of child maltreatment and school failure on property and violent offending and number of criminal convictions in later adolescence. Beaver, DeLisi, Vaughn, and Wright (2010) find in a study of White males in the Add Health sample that the effect of verbal ability on self-control and delinquency is moderated by MAOA genotype. In a study looking at desistance, Beaver, Wright, DeLisi, and Vaughn (2008) find that several genes, MAOA among them, interact with marital status to predict the patterns of desistance among males in the Add Health sample. Watts and McNulty (2014) find in the Add Health sample that the effect of the parent–child relationship on the levels of self-control and criminal behavior is moderated by MAOA (as well as DAT1). Armstrong and colleagues (2014) find in an incarcerated sample that parental criminality interacts with MAOA genotype to predict the self-reports of serious criminal behavior for both property and violent arrest rates. Finally, Ouellet-Morin and colleagues (2015) find in a sample of Canadian youths that MAOA moderates the effect of violence exposure on violence perpetration.

In summation, studies show that the 2R/3R alleles of MAOA moderate (i.e., amplify) the effects of environmental adversity on antisocial outcomes. But why do genes have this moderating effect on risky behaviors and antisocial outcomes? Belsky and Pluess (2009) observe that genes like MAOA are related to the dopaminergic and serotonergic system related to reward and punishment sensitivity. Thus, individuals are more or less sensitive to their environment when it comes to experiencing and reacting to pleasure and displeasure based in part on their genetics. In the current study, this means that engagement in drug markets by selling drugs is a risky behavior that is more or less risky in its impact on violent victimization depending on individual genotype, specifically MAOA genotype.

Concerning MAOA genotype more specifically, it has been hypothesized that MAOA shows its effects because it is involved in the regulation of emotion and cognition in the limbic system (Beaver et al., 2010). In particular, the low-activity version of MAOA has been shown through functional MRI analyses to relate to increased amygdala arousal, which is related to aggressive behavior, and diminished activity of the regulatory prefrontal cortex, which is related to behavioral inhibition (Beaver et al., 2010). So, in a context of drug selling, a person with the low-activity version of MAOA might, in challenging circumstances, be more likely to behave and react to others in a way that puts them at greater risk for violent victimization. This is due in part to drug markets being populated by individuals who, compared to individuals not in the drug market, are generally more violence prone (Anderson, 2000; Blumstein, 1995b).

The Present Study

The current study seeks to expand on the existing literature concerning drug selling and victimization. This will be accomplished by examining the effect of drug selling on violent victimization in a nationally representative sample of youths. More importantly, the current study will also test whether the effect of drug selling on violent victimization is moderated by genetics, specifically the MAOA gene. This represents a merging of the literatures on drug selling and victimization and the G × E literature. In the current study, drug selling is seen as a risk factor for victimization that may be moderated by genotype. Therefore, people of different genotypes may, when exposed to the same environmental risk factor, experience different victimization outcomes.

Prior research on the drug selling–victimization link and in the G × E literature led to the development of two hypotheses that are tested in the current study. Hypothesis 1 predicts that drug selling at Wave I will increase violent victimization at Wave II, net of important controls that could potentially make this relationship spurious. Hypothesis 2 predicts that MAOA genotype will moderate the drug selling–violent victimization relationship, such that drug-selling carriers of the 2R or 3R alleles will report more violent victimization at Wave II than drug sellers who are not the carriers of the 2R or 3R alleles of MAOA.

Data and Method

Sample

The current study uses data from the Add Health. Add Health is nationally representative, consisting of a sample of adolescents who were first recruited during 1994–1995 when respondents were in Grades 7–12 (Harris et al., 2003; Udry, 1998). Add Health acquired a nationally representative sample of adolescents by employing a multistage stratified sampling process to select 80 high schools and 52 middle and junior high schools for inclusion in the study. Over 90,000 students filled out in-school self-report surveys, and of this group, a subsample was randomly chosen for the Wave I in-home component of Add Health. In total, 20,745 adolescents and 17,700 of their primary caregivers participated in the Wave I in-home component of Add Health (Harris et al., 2003). Wave II data collection occurred approximately 1–2 years after Wave I. Wave III data were collected approximately 7 years after Wave I when respondents were between 18 and 26 years old, and Wave IV data were collected during 2007–2008 when respondents were between 24 and 32 years old.

During Wave IV in-home interviews, Add Health took saliva swabs from all respondents for DNA analysis. In conjunction with the Institute for Behavioral Genetics in Boulder, CO, Add Health genotyped Wave IV interviewees for a set of genetic markers of interest to biological and biosocial researchers. The current study includes respondents interviewed at Waves I, II, and IV, who had complete weighting data and were not missing data concerning MAOA genotype.³ The final sample for analysis includes information gathered from 8,860 respondents.⁴ That Add Health is a large and nationally representative data set that contains variables measuring both genetics, and the social environment makes it well suited for the present study.

Measures

Dependent variable

Violent victimization

The dependent variable violent victimization is drawn from Wave II, when the average respondent was approximately 16 years old. At Wave II, respondents were asked a number of questions about their experiences witnessing, perpetrating, and being the victim of violence in the past 12 months. Among these items were questions that asked respondents if in the past 12 months, they had been jumped, cut or stabbed, or shot. For all three questions, respondents had the option to answer never, once, and more than once. Very few respondents report any violent victimization in the prior year at Wave II (approximately 10%; see Table 1), with frequent violent victimization being extremely rare. We thus treat violent victimization conservatively, creating a single, dichotomous measure where a score of 1 indicates that a respondent was jumped, cut or stabbed, shot, or any combination of the three at least once in the past year.⁵

Table 1.

Descriptive Statistics.

Variables		Full Sample
	Range	(N = 8,860)
	Range	Mean (SE)
Dependent variable
Violent victimization W2	0/1	0.10 (.00)
Independent variables
Drug selling W1	0/1	0.07 (.00)
MAOA genotype (1 = 2R/3R)	0/1	0.53 (.01)
Controls
Sex (1 = male)	0/1	0.45 (.01)
White	0/1	0.55 (.01)
Hispanic	0/1	0.15 (.00)
Black	0/1	0.20 (.00)
Native American	0/1	0.03 (.00)
Asian	0/1	0.07 (.00)
Other	0/1	0.01 (.00)
Age W2	11–23	16.18 (.02)
Parent’s education	0/1	0.28 (.00)
Parent employed W1	0/1	0.91 (.00)
Minor delinquency W1	0–9	1.88 (.02)
Fighting W1	0/1	0.31 (.00)
Gut feeling	1–5	2.97 (.01)
Evaluate outcomes	1–5	2.20 (.01)
Affiliation with delinquent peers	0–9	2.37 (.03)
Cocaine W1	0/1	0.03 (.00)
Other drugs W1	0/1	0.07 (.00)

Note. Because these statistics are weighted and adjusted for survey design, standard errors are produced rather than standard deviations. MAOA = Monoamine oxidase A.

Independent variables

Drug selling

The independent variable drug selling is drawn from Wave I. As part of the delinquency questionnaire in Wave I, respondents were asked how often in the past 12 months they had sold marijuana or other drugs. Respondents could answer never, 1 or 2 times, 3 or 4 times, or 5 or more times. Very few respondents report any involvement in drug selling at Wave I (approximately 7%; see Table 1), with frequent drug selling being rare. We thus transformed this variable by dichotomizing it, with a score of 1 indicating that a respondent sold marijuana or other drugs at least once in the year prior to Wave I.

MAOA genotype

The past research on MAOA strongly suggests that two low-activity versions of this gene (2 repeat and 3 repeat) are associated with negative behavioral and mental health outcomes, particularly among males (Belsky & Pluess, 2009; Caspi et al., 2002; Kim-Cohen et al., 2006). Following the past research, we code MAOA to reflect the nonpresence (0) or presence (1) of either the 2R or 3R allele. We lump the 2R and 3R alleles together because there are very few 2R allele carriers in the Add Health data set, limiting statistical power when trying to make comparisons based on a three group coding of the gene (no 2R/3R vs. 2R vs. 3R). Based on our coding, about 53% of the full sample are carriers of either the 2R or 3R allele. The Hardy–Weinberg equilibrium test confirms that the distribution of MAOA among the females in the sample does not differ significantly from that predicted on the basis of simple Mendelian inheritance.⁶

Control variables

We include several general controls in all of the regression models: Age, dummy variables for Hispanic, non-Hispanic Black, Native American, Asian, and Other (with non-Hispanic White as the reference category), target parent’s education (1 = 4 year degree or more), and target parent’s employment status Wave I (1 = employed). About 55% of the sample is non-Hispanic White, 20% non-Hispanic Black, 15% Hispanic with the remainder comprising Asian (7%), Native American (3%), and members of other racial/ethnic groups (1%). About 28% of respondent’s parents are college graduates, while 91% of respondent’s parents were employed at Wave I.

We additionally control for several more variables in all of the regression models as a check for spuriousness in the drug selling–violent victimization relationship. Implicit in our analysis is a view of drug selling as a risk factor that increases risks for violent victimization. However, the population heterogeneity approach to understanding the offending–victimization overlap would argue that both drug selling and violent victimization are the result of some individuals having simultaneously greater risks for both offending and victimization due to a number of factors, such as self-control and substance use (Fisher, Sloan, Cullen, & Lu, 1998; Gover, 2004; Regoeczi, 2000; Schreck & Fisher, 2004; Schreck, Stewart, & Fisher, 2006). If this is the case, then any findings we produce showing a positive relationship between drug selling and violent victimization may conceal a spurious relationship driven by these other factors.

To control for this possibility, we include in all of the regression models a number of theoretically important variables that correlate with both offending and victimization. Affiliation with delinquent peers is represented by 3 items that measure a respondent’s association with substance-using friends. Add Health respondents were asked at Wave I how many of their three closest friends smoke cigarettes, drink alcohol, and smoke marijuana. We summed these 3 items to create a measure of association with delinquent peers (α = .76). Higher scores on this measure indicate more involvement with substance-using friends. Self-control is represented by two separate items, wherein respondents were asked to agree or disagree on a Likert-type scale with questions that asked if when dealing with problems they follow their “gut feelings” without considering consequences and whether they evaluate outcomes after the fact.

We additionally control for minor delinquency and violence at Wave I. Minor delinquency consists of three combined items that asked respondents how often in the past year they had damaged other’s property, lied to their parents about where they were or who they were with, and acted loud, rowdy, or unruly in public. Lastly, we control for drug use at Wave I. Violence is represented by a measure of fighting where respondents were asked if they had gotten into a physical fight in the past year (1 = yes). Two separate drug measures asked whether respondents had ever in their lifetime used cocaine or other drugs like LSD, heroin, mushrooms, and so on (1 = yes).

Analytic Strategy

Given that the dependent variable is dichotomous, we utilize logistic regression techniques. These models test whether drug selling increases violent victimization in the presence of controls that potentially make the drug selling–violent victimization relationship spurious (Hypothesis 1) and whether this relationship is strengthened by the presence of the 2R/3R alleles of MAOA (Hypothesis 2). As previously mentioned, we run separate models for males and females because of expected gender differences in drug selling and violent victimization and because males carry one while females carry two copies of the MAOA gene. We utilize the appropriate weight, cluster, and strata variables in all analyses to account for the complex Add Health survey design.⁷ Tests using variance inflation factors show that multicollinearity is not a problem in any of the presented equations.

Results

Descriptive Statistics

Table 1 presents the descriptive statistics for the full sample. As can be seen in Table 1, about 10% of the full sample report experiencing a violent victimization between Waves I and II, while about 7% of the full sample report selling drugs in the year prior to Wave I in-home interviews. The full sample is almost evenly split by MAOA genotype, with 53% of the full sample being carriers of the 2R/3R alleles.

Table 2 presents the descriptive statistics and mean comparisons by gender. Numerous sex differences in the sample can be noted. There is significantly more violent victimization and drug selling among the males in the sample. Females in the sample are significantly more likely to be carriers of the 2R/3R alleles than are males. This makes sense due to the fact that females carry two copies of the MAOA gene, while males carry only one. Among the controls, the male respondents are slightly older, and males report greater involvement in minor delinquency and more involvement with delinquent peers.

Table 2.

Descriptive Statistics and Mean Comparisons by Gender.

Variables		Males	Females
	Range	(n = 3,999)	(n = 4,861)
	Range	Mean (SE)	Mean (SE)
Dependent variable
Violent victimization W2	0/1	0.15 (.01)	0.05 (.00)**
Independent variables
Drug selling W1	0/1	0.10 (.00)	0.04 (.00)**
MAOA genotype (1 = 2R/3R)	0/1	0.41 (.01)	0.63 (.01)**
Controls
White	0/1	0.56 (.01)	0.54 (.01)
Hispanic	0/1	0.16 (.01)	0.15 (.01)
Black	0/1	0.18 (.01)	0.21 (.01)**
Native American	0/1	0.02 (.00)	0.03 (.00)
Asian	0/1	0.07 (.00)	.06 (.00)**
Other	0/1	0.01 (.00)	0.01 (.00)
Age W2	11–23	16.29 (.03)	16.08 (.02)**
Parent’s education	0/1	0.30 (.01)	0.27 (.01)**
Parent employed W1	0/1	0.92 (.00)	0.91 (.00)
Minor delinquency W1	0–9	1.96 (.03)	1.82 (.03)**
Fighting W1	0/1	0.41 (.01)	0.23 (.01)**
Gut feeling	1–5	3.06 (.02)	2.90 (.02)**
Evaluate outcomes	1–5	2.20 (.01)	2.20 (.01)
Affiliation with delinquent peers	0–9	2.46 (.04)	2.30 (.04)**
Cocaine W1	0/1	0.03 (.00)	0.03 (.00)
Other drugs W1	0/1	0.07 (.00)	0.07 (.00)

Note. Because these statistics are weighted and adjusted for survey design, standard errors are produced rather than standard deviations. MAOA = Monoamine oxidase A.

**p < .01, mean one-way analysis of variances denote significant gender comparisons.

Mean comparisons by genotype for the full sample (not presented) reveal one important difference between 2R/3R carriers and carriers of other MAOA alleles. Carriers of the 2R/3R alleles report significantly less violent victimization than do the carriers of other MAOA alleles. There is not a significant difference in drug selling between the two genetic subgroups in the full sample.

Table 3 presents the correlation matrix for the study variables for the full sample, which serves as a check for gene–environment correlations (rGE). rGE refers to a nonrandom distribution of environments among different genotypes (Simons et al., 2011a), which potentially confounds G × E effects (Guo et al., 2008; Guo, Tong, & Cai, 2008). As can be seen in Table 3, drug selling does not correlate with MAOA genotype. There is, however, an extremely weak, but significant, correlation between violent victimization and MAOA genotype, with the 2R/3R alleles correlating with less violent victimization. As expected, drug selling and violent victimization have a positive, statistically significant relationship.

Table 3.

Correlation Matrix for the Full Sample.

Variables	1.	2.	3.	4.	5.	6.	7.	8.	9.	10.	11.	12.	13.	14.	15.	16.	17.	18.	19.	20.
1. Violent victimization W2	X
2. Drug selling W1	.18**	X
3. MAOA genotype (1 = 2R/3R)	−.03**	−.02	X
4. Sex (1 = male)	.16**	.11**	−.22**	X
5. White	−.09**	−.03**	−.15**	.01	X
6. Hispanic	.08**	.05**	−.02	.01	−.55**	X
7. Black	.04**	.00	.15**	−.05**	−.55**	−.47**	X
8. Native American	.04**	.03**	.01	.00	−.18**	−.21**	−.08**	X
9. Asian	−.02	−.03*	.08**	.03**	−.29**	−.07**	−.13**	−.05**	X
10. Other	.00	.01	.01	.00	−.13**	−.11**	−.06**	−.02	−.03**	X
11. Age W2	.03**	.10**	−.01	.06**	−.07**	.01	−.02	−.01	.06**	−.01	X
12. Parent’s education	−.03**	–.02*	.00	.04**	.03*	.07**	.05**	−.03*	.10**	.00	−.06**	X
13. Parent employed W1	.02	.02	−.01	.02	−.04**	−.15**	.05**	−.01	.02	.00	.01	.05**	X
14. Minor delinquency W1	.17**	.30**	−.01	.03**	.00	.03*	−.05**	.03*	.02	.00	.06**	.02	.03*	X
15. Fighting W1	.27**	.17**	−.03**	.19**	−.09**	.04**	.08**	.03**	−.02	.00	−.02	−.06**	.00	.26**	X
16. Gut feeling	.09**	.07**	−.02	.07**	−.03*	.03*	−.13**	.01	−.02	−.01	−.05**	−.08**	.01	.34**	.13**	X
17. Evaluate outcomes	−.02*	.04**	−.02*	.00	.11**	−.04**	−.02	.01	−.03**	−.02	−.07	.02*	−.01	.13**	.01	.04**	X
18. Affiliation with delinquent peers	.18**	.34**	.00	.03**	.07**	.01	−.08**	.04**	−.05**	.00	.27	−.09**	.00	.10**	.20**	.13**	.06**	X
19. Cocaine W1	.08**	.29**	−.01	.02	.00	.05**	−.06**	.02	−.01	.03**	.07**	−.02	.00	.15**	.08**	.07**	.02	.22**	X
20. Other drugs W1	.09**	.36**	−.01	.00	.07**	.03**	−.10**	.01	−.03**	.01	.11**	−.04**	.01	.25**	.09**	.06**	.04**	.35**	.45**	X

Note. N = 8,860. MAOA = Monoamine oxidase A.

Multivariate Analysis

Table 4 presents the logistic regression models with violent victimization at Wave II regressed on drug selling at Wave I, MAOA genotype, and controls. Separate models are run for males and females. As can be seen in Model 1 of Table 4, the key independent variable of interest, drug selling, significantly increases violent victimization among males, but not females. Thus, Hypothesis 1 is supported among males, but not females.⁸ MAOA genotype does not directly shape violent victimization for either males or females. Among the more theoretically important controls, minor delinquency, fighting, and affiliating with delinquent peers increase violent victimization among both males and females. Relying on gut feelings increases violent victimization among males, but not females.

Table 4.

Violent Victimization W2 Regressed on Drug Selling W1, MAOA Genotype, and Controls.

Variables	Model 1		Model 2
	Males	Females	Males	Females
	OR (SE)	OR (SE)	OR (SE)	OR (SE)
Environment and genetic variables
Drug selling W1	1.47 (.26)*	1.10 (.31)	1.06 (.26)	0.88 (0.43)
MAOA (1 = 2R/3R)	0.82 (.10)	1.13 (.22)	0.72 (.10)*	1.09 (0.23)
Two-way interaction
Drug selling W1 × MAOA			2.07 (.67)*	1.40 (0.80)
Controls
Hispanic	1.89 (.33)**	2.49 (.64)**	1.88 (.32)**	2.48 (0.64)**
Black	1.69 (.32)**	2.81 (.66)**	1.70 (.33)**	2.81 (0.67)**
Native American	2.06 (.83)	3.98 (1.30)**	1.98 (.83)	4.08 (1.29)**
Asian	1.11 (.40)	2.49 (1.33)	1.16 (.42)	2.50 (1.34)
Other	1.50 (.55)	0.94 (.59)	1.53 (.54)	0.96 (0.59)
Age W2	1.06 (.04)	0.87 (.05)**	1.06 (.04)	0.87 (0.05)**
Parent’s education	0.80 (.12)	0.84 (.17)	0.80 (.13)	0.85 (0.17)
Parent employed W1	1.26 (.33)	1.08 (.36)	1.25 (.33)	1.09 (0.37)
Minor delinquency W1	1.08 (.04)*	1.21 (.06)**	1.08 (.04)*	1.21 (0.06)**
Fighting W1	3.09 (.51)**	5.84 (1.29)**	3.08 (.50)**	5.83 (1.30)**
Gut feeling	1.13 (.06)*	1.01 (.09)	1.13 (.06)*	1.01 (.09)
Evaluate outcomes	0.86 (.07)	0.87 (.10)	0.86 (.07)	0.87 (0.10)
Affiliation with delinquent peers	1.17 (.03)**	1.10 (.04)**	1.17 (.03)**	1.10 (0.04)**
Cocaine W1	1.18 (.42)	1.16 (.52)	1.22 (.43)	1.14 (0.51)
Other drugs W1	0.96 (.25)	1.32 (.38)	1.02 (.27)	1.32 (0.38)
Constant	0.01 (.01)**	0.08 (.09)*	0.01 (.01)**	0.08 (0.09)*

Note. N = 8,860. Non-Hispanic White is the reference category for all race/ethnic groups. This table includes odds ratios (OR; linearized standard errors) from logistic regression models. MAOA = Monoamine oxidase A.

*p < .05. **p < .01.

In Model 2 of Table 4, the interaction between drug selling and MAOA genotype is introduced. This interaction variable was produced by simply multiplying the 0/1 drug selling measure by the 0/1 MAOA genotype measure. This interaction is significant and in the expected direction for males, but not females, among whom the interaction is insignificant. Like with Hypothesis 1, support for Hypothesis 2 is found only among males. Table 5 uses the regression results from Model 2 in Table 4 to present the predicted probabilities of violent victimization at Wave II among the different categories of individuals based on the drug selling–MAOA interaction for males and females separately. These predicted probabilities further highlight the nuanced effect of the interaction between drug selling and MAOA genotype on violent victimization. Focusing on males, the group with the lowest predicted probability of violent victimization at Wave II is those who carry the 2R/3R allele of MAOA, who do not sell drugs. Per what is seen in Model 2 of Table 4, the group among males with the highest risks for violent victimization based on the predicted probabilities is those who carry the 2R/3R allele of MAOA and sell drugs. The predicted probability of violent victimization at Wave II for this group is twice that of nondrug sellers who carry the 2R/3R allele of MAOA.

Table 5.

Predicted Probabilities of Violent Victimization W2.

Drug selling by MAOA combinations	Predicted Probabilities of Victimization
Males
NO drug selling, NO 2R/3R	.109
NO drug selling, 2R/3R	.081
Drug selling, NO 2R/3R	.115
Drug selling, 2R/3R	.162
Females
NO drug selling, NO 2R/3R	.030
NO drug selling, 2R/3R	.033
Drug selling, NO 2R/3R	.027
Drug selling, 2R/3R	.040

Note. All other covariates held at their means. All presented probabilities sig. at p < .05.

Discussion

This article sought to examine the relationship between genetics, drug selling, and victimization. Specifically, we utilized data from Add Health to test whether selling drugs influenced later violent victimization while controlling for a number of variables that potentially make the drug selling–victimization relationship spurious and whether the relationship between drug selling and violent victimization is moderated by MAOA genotype. Results differed by gender. Among males but not females, drug selling increases later violent victimization, net of controls, supporting Hypothesis 1 among this group. Additionally, the drug selling–violent victimization relationship among males is moderated by MAOA, such that male, drug-selling carriers of the 2R or 3R alleles of MAOA reported significantly more violent victimization that male drug sellers who carry some other MAOA allele. Like with Hypothesis 1, Hypothesis 2 is supported among males only.

There are several meaningful conclusions to be drawn from these results. First, drug selling increases violent victimization among males but not females, net of controls. This seems to suggest that engaging in drug markets is riskier for males than females in terms of exposure to violence. Prior qualitative research has suggested this is the case, with females using gender stereotypes and their gendered roles within criminal organizations to decrease their likelihood of experiencing victimization (Miller & Decker, 2001). Future research and theorizing on drug selling as a risk factor for victimization must continue to account for gender and how it conditions this relationship.

Second, the drug selling–violent victimization relationship among males is moderated by the 2R/3R alleles of MAOA, such that male carriers of the 2R/3R alleles of MAOA are at increased risks for violent victimization when they sell drugs compared to males who carry some other allele for MAOA. This finding matches up with prior biosocial research, which has found that behaviors and environmental stressors that increase the likelihood of antisocial outcomes are exacerbated among male carriers of the 2R/3R allele (Beaver et al., 2010; Simons et al., 2011a). The question remains, why is this the case? The descriptive results suggest an absence of rGE, so male carriers of the 2R/3R alleles aren’t self-selecting into drug selling more often, so what part of engaging in drug markets puts male carriers of the 2R/3R alleles more at risk for violent victimization? Given the moniker of the “warrior gene” that has been bestowed upon MAOA, it could have something to do with the levels of aggression on the part of these individuals. Perhaps once they are involved in drug markets, they are more likely to use aggression to gain successful outcomes, whether this aggression is provoked or not. Looking into this possibility is beyond the scope of the current study, but more research is needed.

Lastly, it is worth noting that the results not presented show that the drug selling–violent victimization relationship among females is a spurious relationship driven by the combination of self-control, affiliation with delinquent peers, delinquency, and drug use. Another question that arises is why is the drug selling–violent victimization relationship gendered in this way? Why is there a direct relationship among males, while among females this relationship is driven by factors that shape both offending and victimization risks? It could be that males and females differ in their motivations for selling drugs, and this seems like an area where qualitative research would be quite informative.

Before the implications of this study are discussed, the key limitation of sample attrition should be addressed. Due to decisions made by Add Health concerning who to reinterview at Waves II and IV, as well as losing respondents without sampling weight information, the final analytic sample in the current study is considerably smaller than the original Wave I sample. There is also some further attrition because some respondents refused DNA swabs at Wave IV (less than 5% of respondents). On this point, it is worth noting that reports published by Add Health suggest that, at the very least, those who refused DNA swabs at Wave IV do not differ from those who participated concerning race/ethnicity (Smolen et al., 2012). The Add Health team has published reports in the past arguing that as long as researchers utilize the sampling weight data they provide in the correct manner to account for the project’s sampling design when conducting analyses utilizing more than one wave of data, the coefficients and standard errors produced by statistical models should remain unbiased (Chen & Chantala, 2014). Still, the loss of a large, nonrandom portion of the original Add Health sample when conducting analyses utilizing multiple waves of data must be noted as an important limitation of the data set and the current study.

Two practical implications of the current study for criminal justice systems should be noted. First, the results suggest that among males, drug selling is a risk factor for violent victimization. While youths are actively discouraged from engaging in drug markets based on many different kinds of reasoning, that reasoning does not often include informing youths that they are increasing their risks for serious injury and/or death by selling drugs. Alongside the other warnings, this line of reasoning could further discourage youths from selling drugs. Second, the results suggest that among males who carry the 2R/3R alleles of MAOA, there is an even higher risk of injury. This result suggests that individualized programs that target youths who are considered at-risk, both socially and genetically could result in greater decreases in antisocial outcomes (Wright & Boisvert, 2009). These targeted interventions would be more efficient by focusing on offenders who are the most high-risk, and thus program success rates could be improved (Gajos, Fagan, & Beaver, 2016).

In summary, this study illustrates, like many before it, that genetics condition the effects of risky behaviors on antisocial outcomes. Future research must continue this trend by focusing on other candidate genes and risky behaviors and the antisocial outcomes they shape.

Footnotes

Declaration of Conflicting Interests

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

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Notes

References

Adler

P. A.

(1993). Wheeling and dealing: An ethnography of an upper-level drug dealing and smuggling community. New York, NY: Columbia University Press.

Anderson

(2000). Code of the street: Decency, violence, and the moral life of the inner city. New York, NY: W. W. Norton & Company.

Armstrong

T. A.

Boutwell

B. B.

Flores

Symonds

Keller

Gangitano

D. A.

(2014). Monoamine oxidase A genotype, childhood adversity, and criminal behavior in an incarcerated sample. Psychiatric Genetics, 24, 164–171.

Beaver

K. M.

DeLisi

Vaughn

M. G.

Barnes

J. C.

(2010). Monoamine oxidase A genotype is associated with gang membership and weapon use. Comprehensive Psychiatry, 51, 130–134.

Beaver

K. M.

DeLisi

Vaughn

M. G.

Wright

J. P.

(2010). The intersection of genes and neuropsychological deficits in the prediction of adolescent delinquency and low self-control. International Journal of Offender Therapy & Comparative Criminology, 54, 22–42.

Beaver

K. M.

Wright

J. P.

DeLisi

Vaughn

M. G.

(2008). Desistance from delinquency: The marriage effect revisited and extended. Social Science Research, 37, 736–752.

Belsky

Pluess

(2009). Beyond diathesis stress: Differential susceptibility to environmental influences. Psychological Bulletin, 135, 885–908.

Berg

M. T.

Loeber

(2015). Violent conduct and victimization risk in the urban illicit drug economy: A prospective examination. Justice quarterly, 32, 32–55.

Blumstein

(1995a). Violence by young people: Why the deadly nexus. National Institute of Justice Journal, 229, 2–9.

10.

Blumstein

(1995b). Youth violence, guns, and the illicit-drug industry. The Journal of Criminal Law and Criminology (1973–), 86, 10–36.

11.

Blumstein

Cork

(1996). Linking gun availability to youth gun violence. Law and Contemporary Problems, 59, 5–24.

12.

Bourgois

(2003). In search of respect: Selling crack in El Barrio (Vol. 10). Cambridge, MA: Cambridge University Press.

13.

Bourgois

Prince

Moss

(2004). The everyday violence of hepatitis C among young women who inject drugs in San Francisco. Human Organization, 63, 253–264.

14.

Byrd

A. L.

Manuck

S. B.

(2014). MAOA, childhood maltreatment, and antisocial behavior: Meta-analysis of a gene-environment interaction. Biological Psychiatry, 75, 9–17.

15.

Caspi

McClay

Moffitt

T. E.

Mill

Martin

Craig

I. W.,

… Poulton

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

16.

Chen

Chantala

(2014). Guidelines for analyzing Add Health data. Chapel Hill, NC: Carolina Population Center, University of North Carolina.

17.

Curtis

Wendel

(2007). “You’re always training the dog”: Strategic interventions to reconfigure drug markets. Journal of Drug Issues, 37, 867–891.

18.

Ducci

Enoch

M. A.

Hodgkinson

Catena

Robin

R. W.

Goldman

(2008). Interaction between a functional MAOA locus and childhood sexual abuse predicts alcoholism and antisocial personality disorder in adult women. Molecular Psychiatry, 13, 334–347.

19.

Fergusson

D. M.

Boden

J. M.

Horwood

L. J.

Miller

A. L.

Kennedy

M. A.

(2011). MAOA, abuse exposure and antisocial behaviour: 30-year longitudinal study. The British Journal of Psychiatry, 198, 457–463.

20.

Fergusson

D. M.

Boden

J. M.

Horwood

L. J.

Miller

Kennedy

M. A.

(2012). Moderating role of the MAOA genotype in antisocial behaviour. The British Journal of Psychiatry, 200, 116–123.

21.

Fisher

B. S.

Sloan

J. J.

Cullen

F. T.

(1998). Crime in the ivory tower: The level and sources of student victimization. Criminology, 36(3), 671–710.

22.

Foley

D. L.

Eaves

L. J.

Wormley

Silberg

J. L.

Maes

H. H.

Kuhn

Riley

(2004). Childhood adversity, monoamine oxidase a genotype, and risk for conduct disorder. Archives of General Psychiatry, 61, 738–744.

23.

Gajos

J. M.

Fagan

A. A.

Beaver

K. M.

(2016). Use of genetically informed evidence-based prevention science to understand and prevent crime and related behavioral disorders. Criminology & Public Policy, 15, 683–701.

24.

Goldstein

P. J.

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

25.

Gover

A. R.

(2004). Risky lifestyles and dating violence: A theoretical test of violent victimization. Journal of Criminal Justice, 32(2), 171–180.

26.

Guo

Roettger

M. E.

Cai

(2008). The integration of genetic propensities into social-control models of delinquency and violence among male youths. American Sociological Review, 73, 543–568.

27.

Guo

Tong

Cai

. (2008). Gene by social context interactions for number of sexual partners among white male youths: Genetics-informed sociology. American Journal of Sociology, 114(S1), S36–S66.

28.

Haberstick

B. C.

Lessem

J. M.

Hopfer

C. J.

Smolen

Ehringer

M. A.

Timberlake

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, 135B, 59–64.

29.

Harris

K. M.

Florey

Tabor

Bearman

P. S.

Jones

Udry

J. R.

(2003). The National Longitudinal Study of Adolescent Health: Research Design. Retrieved from http://www.cpc.unc.edu/projects/addhealth/design

30.

Holland

N. R.

DeLisi

(2014). The warrior gene. In DeLisi

Vaugn

M. G.

(Eds.), The Routledge international handbook of biosocial criminology (pp. 179–189). New York, NY: Routledge.

31.

Jacobs

B. A.

Topalli

Wright

(2000). Managing retaliation: Drug robbery and informal sanction threats. Criminology, 38, 171–198.

32.

Jacques

(2010). The necessary conditions for retaliation: Toward a theory of non-violent and violent forms in drug markets. Justice Quarterly, 27, 186–205.

33.

Jacques

Allen

Wright

(2014). Drug dealers’ rational choices on which customers to rip-off. International Journal of Drug Policy, 25, 251–256.

34.

Jacques

Wright

(2011). Informal control and illicit drug trade. Criminology, 49, 729–765.

35.

Jacques

Wright

Allen

(2014). Drug dealers, retaliation, and deterrence. International Journal of Drug Policy, 25, 656–662.

36.

Johnson

Golub

Dunlap

(2000). The rise and decline of hard drugs, drug markets, and violence in inner-city New York. In Blumstein

Wallman

(Eds.), The crime drop in America (pp. 164–206). New York, NY: Cambridge University Press.

37.

Kim-Cohen

Caspi

Taylor

Williams

Newcombe

Craig

I. W.

Moffitt

T. E.

(2006). MAOA, maltreatment, and gene–environment interaction predicting children’s mental health: New evidence and a meta-analysis. Molecular Psychiatry, 11, 903–913.

38.

Lauritsen

J. L.

Heimer

(2008). The gender gap in violent victimization, 1973-2004. Journal of Quantitative Criminology, 24, 125–147.

39.

Miller

Decker

S. H.

(2001). Young women and gang violence: Gender, street offending, and violent victimization in gangs. Justice Quarterly, 18, 115–140.

40.

Moskos

(2008). Cop in the hood: My year policing Baltimore’s Eastern District. Princeton, NJ: Princeton University Press.

41.

Nilsson

K. W.

Sjoberg

R. L.

Damberg

Leppert

Ohrvik

Alm

P. O.,

… Oreland

(2006). Role of monoamine oxidase a genotype and psychosocial factors in male adolescent criminal activity. Biological Psychiatry, 59, 121–127.

42.

Ouellet-Morin

Côté

S. M.

Vitaro

Hébert

Carbonneau

Lacourse

É.,

… Tremblay

R. E.

(2015). Effects of the MAOA gene and levels of exposure to violence on antisocial outcomes. The British Journal of Psychiatry. doi:10.1192/bjp.bp.114.162081

43.

Regoeczi

W. C.

(2000). Adolescent violent victimization and offending: Assessing the extent of the link. Canadian Journal of Criminology, 42, 493–505.

44.

Rosenfeld

Jacobs

B. A.

Wright

(2003). Snitching and the code of the street. British Journal of Criminology, 43, 291–309.

45.

Schreck

C. J.

Fisher

B. S.

(2004). Specifying the influence of family and peers on violent victimization. Journal of Interpersonal Violence, 19(9), 1021–1041.

46.

Schreck

Stewart

Fisher

(2006). Self-control, victimization, and their influence on risky lifestyles: A longitudinal analysis using panel data. Journal of Quantitative Criminology, 22(4), 319–340.

47.

Simons

R. L.

Lei

M. K.

Beach

S. R. H.

Brody

G. H.

Philibert

R. A.

Gibbons

F. X.

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

48.

Simons

R. L.

Lei

M. K.

Beach

S. R. H.

Brody

G. H.

Philibert

R. A.

Gibbons

F. X.

(2011b). Social adversity, genetic variation, street code, and aggression: A genetically informed model of violent behavior. Youth Violence and Juvenile Justice, 10, 3–24.

49.

Sjoberg

R. L.

Nilsson

K. W.

Wargelius

Leppert

Lindström

Oreland

(2007). Adolescent girls and criminal activity: Role of MAOA-LPR genotype and psychosocial factors. American Journal of Medical Genetics Part B: Neuropsychiatric Genetics, 144B, 159–164.

50.

Small

Maher

Lawlor

Wood

Shannon

Kerr

(2013). Injection drug users’ involvement in drug dealing in the downtown eastside of Vancouver: Social organization and systemic violence. International Journal of Drug Policy, 24, 479–487.

51.

Smolen

Whitsel

E. A.

Tabor

Killeya-Jones

L. A.

Cuthbertson

Hussey

J. M.,

… Harris

K. M.

(2012). Add health wave IV documentation: Candidate genes. Chapel Hill, NC: National Longitudinal Study of Adolescent Health, Carolina Population Center, University of North Carolina at Chapel Hill.

52.

Topalli

Wright

Fornango

(2002). Drug dealers, robbery and retaliation. Vulnerability, deterrence and the contagion of violence. British Journal of Criminology, 42, 337–351.

53.

Udry

J. R.

(1998). The National Longitudinal Study of Adolescent Health (Add Health). Chapel Hill, NC: Carolina Population Center, University of North Carolina at Chapel Hill.

54.

Watts

S. J.

McNulty

T. L.

(2014). Genes, parenting, self-control, and criminal behavior. International Journal of Offender Therapy and Comparative Criminology. doi:10.1177/0306624x14553813

55.

Widom

C. S.

Brzustowicz

L. M.

(2006). MAOA and the “cycle of violence”: Childhood abuse and neglect, MAOA genotype, and risk for violent and antisocial behavior. Biological Psychiatry, 60, 684–689.

56.

Wright

J. P.

Boisvert

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

57.

Wright

Decker

S. H.

(1997). Armed robbers in action: Stickups and street culture. Boston, MA: North Eastern University Press.

58.

Young

Smolen

Hewitt

J. K.

Haberstick

B. C.

Stallings

M. C.

Corley

Crowley

(2006). Interaction between MAO-A genotype and maltreatment in the risk for conduct disorder: Failure to confirm in adolescent patients. The American Journal of Psychiatry, 163, 1019–1025.