Do Adolescent Drug Users Fare the Worst? Onset Type,Juvenile Delinquency,and Criminal Careers

Abstract

Although substance abuse often accompanies delinquency and other forms of antisocial behavior, there is less scholarly agreement about the timing of substance use vis-à-vis an individual’s antisocial trajectory. Similarly, although there is extraordinary evidence that onset is inversely related to the severity of the criminal career, there is surprisingly little research on the offense type of onset or the type of antisocial behavior that was displayed when an individual initiated his or her offending career. Drawing on data from a sample of serious adult criminal offenders (N = 500), the current study examined 12 forms of juvenile delinquency (murder, rape, robbery, aggravated assault, burglary, larceny, auto theft, arson, weapons, sexual offense, drug sales, and drug use) in addition to age at arrest onset, age, sex, race to explore their association with chronicity (total arrests), extreme chronicity (1 SD above the mean which was equivalent to 90 career arrests), and lambda (offending per year). The only onset offense type that was significantly associated with all criminal career outcomes was juvenile drug use. Additional research on the offense type of delinquent onset is needed to understand launching points of serious antisocial careers.

Keywords

onset type juvenile delinquency substance use criminal career chronic offending

Introduction

Substance abuse is linked to other forms of criminal offending and criminal justice system involvement. Empirical studies guided by an assortment of theoretical/conceptual perspectives, including general strain theory (Sealock & Manasse, 2012; Sharp, Peck, & Hartsfield, 2012), deterrence theory (Gallupe, Bouchard, & Caulkins, 2011; O’Connell, Visher, Martin, Parker, & Brent, 2011), feminist theory (Garcia & Lane, 2012), social support theory (Pettus-Davis, Howard, Roberts-Lewis, & Scheyett, 2011), situational action theory (Gallupe & Bouchard, 2013), self-control theory (De Li, 2005; Higgins, Mahoney, & Ricketts, 2009), psychopathy theory (Magyar, Edens, Lilienfeld, Douglas, & Poythress, 2011), the risk–needs–responsivity approach (Marlowe et al., 2011), and life course/criminal career models (Farabee, Joshi, & Anglin, 2001; Hammersley, 2011), have provided evidence for a causal and/or correlational relationship between drug use and crime. For instance, a recent study using data from the 2009 National Survey on Drug Use and Health found support for social learning theory, social control theory, and strain theory in the prediction of substance use delinquency, specifically the misuse of prescription drugs (Schroeder & Ford, 2012).

Based on multiple waves of data from the Rochester Youth Development Study, Phillips (2012) found that adolescent drug use was predictive of an array of criminal offenses, including violent ones such as attacking another person with a weapon, throwing objects at another person, forcibly taking money from another person, and others. Among a random sample of adult arrestees, DeLisi (2003) reported that individuals who were arrested for drug use violations were significantly likely to be later arrested for violent crimes, property crimes, white-collar crimes, and nuisance offenses. Although the precise nature of the drugs–crime relationship is equivocal (for a recent research overview, see Boyum, Caulkins, & Kleiman, 2011), there is no question why substance use and abuse are frequently part and parcel of delinquent behavior

Because drugs are intimately connected to criminal offending, drug problems also figure prominently in the lives of criminal offenders who are facing treatment and various criminal justice system processes (Boyum et al., 2011; Mitchell, Wilson, Eggers, & MacKenzie, 2012; Vaughn, 2011). Data from recent epidemiological studies are illustrative. For example, in a study utilizing data from more than 43,000 participants in the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC), Vaughn et al. (2011) used latent class analysis to explore groupings of antisocial classes. Among the three antisocial groups in their data, there was clear evidence of comorbid involvement in drug use and versatile types of antisocial conduct. Two of these classes were characterized by severe drug problems and the third group was characterized by moderate drug and crime problems. A follow-up study using the same NESARC data found that abstainers from substance use and antisocial behavior were healthier and more highly functioning than individuals who had varying usage of drugs and alcohol (Vaughn et al., 2011).

Based on information from nearly 40,000 participants from the 2009 National Survey on Drug Use and Health, Vaughn, DeLisi, Beaver, Perron, and Abdon (2012) also compared health characteristics, offending behavior, substance use, and other social correlates between individuals on probation or parole and those in the community. On every measure of substance use, dependence or other problems associated with drugs, criminal justice system clients had significantly more drug problems than persons without criminal justice system status. The problems included use of cocaine, heroin, marijuana, tobacco, crack, inhalants, methamphetamine, pain relievers, drunk driving, and others. In addition to large effect size differences between the groups for polydrug use, parolees and probationers were also significantly less likely than persons without correctional status to view drug use as risky and had more extensive treatment histories for drug problems (Vaughn et al., 2012).

Although substance abuse often accompanies other forms of antisocial behavior, there is less scholarly agreement about the timing of substance use vis-à-vis an individual’s antisocial trajectory (for a recent research overview, see Jennings & Reingle, 2012). Indeed, there is evidence that initiation of drug use predates the onset of other antisocial behavior, evidence that criminal onset unfolds into substance use, and evidence that drug use and crime emerge concurrently (cf. Elliott, Huizinga, & Menard, 1989; Joshi, Grella, & Hser, 2001; Menard, Mihalic, & Huizinga, 2001; Sullivan & Piquero, 2010; Thornberry, Krohn, & Freeman-Gallant, 2006; Ward, Stogner, Gibson, & Akers, 2010; Zhang, Wieczorek, & Welte, 2011). The unfolding of the drug use–crime nexus depends on a host of factors, including psychiatric functioning, emotional problems, family problems, school problems, psychopathology, and others (Elliott et al., 1989; Joshi et al., 2001; Thornberry et al., 2006). In sum, although drugs and crime clearly coincide, the precise mechanisms of the relationship are less clear.

Interestingly, the uncertain relationship between substance use/abuse and the criminal career is also seen in the criminal careers literature in terms of onset type. Although there is extraordinary evidence that onset is inversely related to the severity of the criminal career (DeLisi & Piquero, 2011), there is surprisingly little research on the offense type of onset, or the type of antisocial behavior that was displayed when an individual initiated his or her offending career.

Onset Type

To our knowledge, no extant conceptual model explicitly identifies specific criminal offenses that mark criminal onset. Instead, there is a general trend of early-emerging deficits in self-regulation, temperament, and social cognition that develop and often worsen over time. In psychiatry, for instance, the developmental course of Oppositional Defiant Disorder (ODD) to Conduct Disorder to Antisocial Personality Disorder is an example of the overall development of antisociality from acts of disobedience and emotional liability to incrementally more overtly antisocial acts. For example, a meta-analysis of 44 studies including 28,401 children indicated that behavioral problems in children with ODD generally occur along two dimensions: overt/covert behavior and destructive/nondestructive behavior (Frick et al., 1993). This means that the disorder is characterized by noncriminal, but aversive behaviors that compromise the child’s ability to get along with others and by delinquent behaviors that are in many respects fledgling indicators of the later crimes to come. However, psychiatric models are incapable of identifying the precise delinquent offense that occurs at onset.

Surprisingly, few criminal career studies have researched onset type, but there are important exceptions. In a large-scale study of four birth cohorts from Sweden, Svensson (2002) examined onset type and its relation to chronic offending and found that property offending, especially vehicle theft and theft were most consistently predictive of chronic offending careers. Indeed, the prevalence of property onset offenders who became chronic offenders was 5 times higher than that of violence onset offenders who became chronic offenders. Based on data from serious youthful offenders in Queensland, Australia, Mazerolle, Piquero, and Brame (2010) examined criminal career differences among youth whose first offense was for a violent crime and those whose first offense was for a nonviolent crime. Although there were significant differences among these groups, they were not consistent in terms of severity of criminal career. For some measures, violent onset was associated with a more extensive criminal career, whereas for other measures, violent onset was associated with a less severe career than nonviolent onset offenders. In other words, the relationship between onset offense type and the subsequent criminal career was equivocal.

Recently, Harris (2013) examined the criminal histories of 751 male sexual offenders who had been referred for civil commitment to explore differences among those whose first offense was a sexual offense, a nonsexual violent offense or a nonsexual property offense. In terms of offense seriousness, it was hypothesized that a property onset would lead to the least severity in terms of the subsequent criminal career (compared with those whose delinquent debuts were violent); however, Harris found that property onset offenders had the most severe careers. Compared with those with a sexual or violent onset, property onset offenders had the earliest onset, lengthiest career span, most official charges, and were rated as the most psychopathic. Property onset offenders also had greater prevalence of elementary school problems, antisocial behavior during adolescence, drug problems, and alcohol use prior to index offending.

To date, these studies have primarily focused on potential distinctions between offenders whose onset comprised violence and those whose onset comprised nonviolence, or less serious antisocial behavior. There has been less focus on the potential role of drug offending—particularly drug selling and drug use—along with the other forms of delinquency to evaluate how offense type predicts various facets of the criminal career. The current study seeks to fill this void using a large sample of adult habitual offenders.

Method

Participants and Procedures

Data were derived from extant records of interviews with arrestees by a pretrial services officer or bond commissioner at a large urban jail located in Colorado over a 6-year period. In this jurisdiction, bond commissioners served as judicial officers and worked in conjunction with sheriff deputies within the county jail. Their function was to interview all criminal defendants brought to the jail and to obtain employment, residency, and criminal history for setting bond. Bond commissioners had the authority to release eligible defendants on recognizance bonds. This work experience permitted constant access (the bond commissioner unit was staffed around the clock) to all arrestees who were brought to the jail during this period.

In this jurisdiction, the bond commissioner unit conducted a pilot study to identify the most recidivistic offenders to determine their eligibility for various social service policies (e.g., a program designed to meet the needs of indigent, transient offenders) and prosecutorial efforts (e.g., selective prosecution using habitual offender statutes). Approximately 50 offenders comprised the original “frequent offender” roster, and their criminal histories contained an average of 30 arrest charges. Based on this selection criterion, any offender whose record contained 30 arrest charges was classified as a frequent offender upon approval from the chief district judge and district attorney’s office. Frequent offenders, because of their habitual criminal conduct, were precluded from receiving personal recognizance bonds. From 1995 to 2000, the bond commissioner unit processed 25,640 defendants, of whom 500 (less than 2%) qualified for frequent/habitual offender status. These 500 offenders constituted, in effect, the population of a 6-year census of official criminal offenders processed in this jurisdiction. Importantly, although the offenders were processed at one facility, their criminal activity spanned multiple jurisdictions.¹

Data and Measures

During bond interviews, legal proceedings conducted under oath, defendants self-reported their criminal history, including all police contacts, arrests, court actions, and sentences. Self-reports can yield arrests and other criminal activities that do not appear on official records, arguably rendering them a more accurate reflection of an individual’s true criminal past. The self-report method is problematic with career criminals, however. The most serious career criminals have offending careers that include potentially hundreds of arrests, convictions, and various punishments. Their careers often span decades and chronicle events when defendants were frequently intoxicated on alcohol and illicit substances. For these and other reasons, the validity and internal consistency of self-reports from the worst offenders may be the least reliable (DeLisi, 2001, 2005; Simon, 1999). Therefore, self-reported criminal histories were supplemented with official records from the Interstate Identification Index (III) system. Under the III system, the FBI maintains an automated criminal record containing an FBI number and a state identification number (SID) for each state holding criminal history information on an individual. The III records are accessed using the National Crime Information Center (NCIC) telecommunications lines that retrieve criminal records from automated repositories. The NCIC was founded in 1967 and criminal history information of offenders in the current sample date to 1948. All 50 states and the District of Columbia participate in the III system which has 92,329,600 criminal history files of which 85,836,300 or 93% are automated (Bureau of Justice Statistics, 2009).²

Dependent Variables

Three dependent variables were used as criminal career outcomes. Total arrests include all police contacts and arrests in the offender’s entire criminal career (M = 59.76, SD = 30.64, range = 30-267). A binary measure of 1 standard deviation above the mean for career arrests was also created to capture the extreme tail of offending. The threshold for this measure was 90 arrests (0 = fewer than 90 arrests, 1 = 90 or more arrests). To capture velocity of offending, a lambda (λ) measure was created by taking total arrests divided by the career span in years into account. This produced the annual offending average (M = 3.33, SD = 2.39, range = 0.38-16.29).

Independent Variables

Onset offense type

Twelve forms of juvenile delinquency were used to examine the onset offense type. These included the eight traditional index offenses in addition to weapons violations, drug sales/trafficking, drug use, and sexual offenses other than rape. Univariate statistics for these measures appear in parentheses: juvenile murder (M = 0.01, SD = 0.14, range = 0-2, prevalence = 0.8%), juvenile rape (M = 0.03, SD = 0.23, range = 0-2, prevalence = 2.2%), juvenile robbery (M = 0.07, SD = 0.41, range = 0-5, prevalence = 4.2%), juvenile aggravated assault (M = 0.13, SD = 0.63, range = 0-7, prevalence = 5.6%), juvenile burglary (M = 0.85, SD = 2.44, range = 0-26, prevalence = 22.6%), juvenile larceny (M = 0.99, SD = 2.42, range = 0-26, prevalence = 26.4%), juvenile auto theft (M = 0.53, SD = 1.86, range = 0-22, prevalence = 16.6%), juvenile arson (M = 0.02, SD = 0.21, range = 0-3, prevalence = 1.6%), juvenile weapons offense (M = 0.07, SD = 0.46, range = 0-6, prevalence = 3.8%), juvenile sexual offense (M = 0.01, SD = 0.14, range = 0-2, prevalence = 0.8%), juvenile drug sales/trafficking (M = 0.05, SD = 0.43, range = 0-8, prevalence = 2.2%), and juvenile drug use (M = 0.38, SD = 1.30, range = 0-10, prevalence = 11.8%).

Controls

Four additional predictor variables were used based on their relation to habitual offending (DeLisi & Piquero, 2011). Onset measured the age of first arrest in the criminal career (M = 18.64, SD = 5.35, range = 8-57), age in years (M = 39.61, SD = 10.74, range = 18-74), sex (89% male, n = 445, 11% female, n = 55), and a dummy term for White race (M = 0.52, SD = 0.50). The sample was diverse with 28.8% Hispanic, 12.2% Black, 6% American Indian, and 0.6% Asian/other.

Statistical Analyses

Three analytical techniques were used: (a) total arrests, (b) 1 SD above the mean cutpoint for total arrests to indicate habitual offending, and (c) lambda to indicate annual offending rate.³ Arrests (output displayed in Table 1) are an example of count data that are positively skewed, assume only integer values, are bound by zero, and have heteroscedastic error terms. These conditions violate ordinary least squares (OLS) assumptions and are usually resolved with Poisson’s regression. However, the extremity of these offenders creates overdispersion in the distribution where the variance exceeds the mean. To correctly account for overdispersion, negative binomial regression is used and diagnostics confirm that negative binomial regression is appropriate (likelihood ratio [LR] test of α = 3,842.28, p < .001).

Table 1.

Negative Binomial Regression Model for Total Arrests (N = 500).

Predictor	b	SE	z	95% CI
Violent index offenses
Juvenile murder	0.13	0.16	0.81	[−0.18, 0.44]
Juvenile rape	0.01	0.09	0.08	[−0.16, 0.18]
Juvenile robbery	0.03	0.05	0.74	[−0.06, 0.13]
Juvenile assault	−0.02	0.03	−0.68	[−0.08, 0.05]
Property index offenses
Juvenile burglary	0.02	0.01	1.69	[−0.00, 0.04]
Juvenile larceny	0.01	0.01	0.53	[−0.02, 0.03]
Juvenile auto theft	0.01	0.01	0.83	[−0.01, 0.03]
Juvenile arson	−0.03	0.10	−0.35	[−0.23, 0.16]
Miscellaneous offenses
Juvenile weapons	−0.03	0.05	−0.62	[−0.12, 0.06]
Juvenile sexual offense	−0.14	0.14	−1.01	[−0.42, 0.13]
Juvenile drug sales	−0.02	0.04	−0.35	[−0.10, 0.07]
Juvenile drug use	0.06	0.01	4.01**	[0.03, 0.09]
Other covariates
Arrest onset	−0.01	0.00	−2.51*	[−0.02, −0.00]
Age	0.01	0.00	6.42**	[0.01, 0.02]
Sex	−0.16	0.06	−2.64*	[−0.28, −0.04]
White	0.02	0.04	0.53	[−0.05, 0.09]
LR χ²	89.88**
Log likelihood	−2263.84
Likelihood ratio test of α	3842.28**

Note. CI = confidence interval.

p < .01. ** p < .001.

The binary measure indicating arrests for 1 SD above the mean was estimated with logistic regression with odds ratios (OR) reported. Three predictors were automatically dropped from this equation (juvenile murder, juvenile sex offense, and sex) because they predicted failure perfectly. As a result, the analytical sample shown in Table 2 is n = 437. The lambda outcome (output in Table 3) was estimated with OLS regression based on its continuous distribution. Diagnostics indicated no problems with multicollinearity (variance inflation factors [VIFs] ranged from 1.06 to 2.51 with mean VIF = 1.44) with all values well below stringent and standard thresholds of VIF = 4 or VIF = 10 (O’Brien, 2007).

Table 2.

Logistic Regression Model for 1 SD Above Mean for Total Arrests (n = 437).

Predictor	OR	SE	z	95% CI
Violent index offenses
Juvenile murder	1 (omitted)
Juvenile rape	0.68	0.59	−0.44	[0.13, 3.74]
Juvenile robbery	1.53	0.46	1.41	[0.85, 2.75]
Juvenile assault	0.96	0.36	−0.10	[0.46, 2.01]
Property index offenses
Juvenile burglary	1.11	0.10	1.20	[0.93, 1.33]
Juvenile larceny	1.01	0.09	0.06	[0.84, 1.20]
Juvenile auto theft	1.10	0.11	1.04	[0.91, 1.33]
Juvenile arson	0.66	0.52	−0.53	[0.14, 3.05]
Miscellaneous offenses
Juvenile weapons	0.62	0.54	−0.55	[0.11, 3.42]
Juvenile sexual offense	1 (omitted)
Juvenile drug sales	0.77	0.46	−0.44	[0.24, 2.45]
Juvenile drug use	1.26	0.12	2.43*	[1.05, 1.51]
Other covariates
Arrest onset	0.99	0.03	−0.35	[0.92, 1.05]
Age	1.06	0.02	3.78**	[1.02, 1.10]
Sex 1	(omitted)
White	1.34	0.45	0.53	[−0.05, 0.09]
LR χ²	29.47*
Log likelihood	−140.68

Note. Juvenile murder, juvenile sex offense, and sex automatically dropped from the model because they predicted failure perfectly. OR = odds ratio; CI = confidence interval.

p < .01. ** p < .001.

Table 3.

OLS Regression Model for Annual Offending Frequency (λ; N = 500).

Predictor	b	SE	t	95% CI
Violent index offenses
Juvenile murder	1.04	0.69	1.51	[−0.31, 2.40]
Juvenile rape	0.01	0.38	0.03	[−0.73, 0.76]
Juvenile robbery	0.03	0.21	0.12	[−0.39, 0.44]
Juvenile assault	0.13	0.13	0.97	[−0.14, 0.41]
Property index offenses
Juvenile burglary	0.09	0.05	1.86	[−0.01, 0.19]
Juvenile larceny	0.12	0.05	2.41*	[0.02, 0.22]
Juvenile auto theft	0.19	0.05	3.63**	[0.09, 0.29]
Juvenile arson	0.02	0.41	0.04	[−0.79, 0.82]
Miscellaneous offenses
Juvenile weapons	0.32	0.20	1.55	[−0.08, 0.72]
Juvenile sexual offense	−0.70	0.60	−1.15	[−1.88, 0.49]
Juvenile drug sales	0.10	0.19	0.51	[−0.28, 0.47]
Juvenile drug use	0.23	0.07	3.48**	[0.10, 0.36]
Other covariates
Arrest onset	0.16	0.02	9.13**	[0.13, 0.20]
Age	−0.12	0.01	−14.18**	[−0.14, −0.11]
Sex	−0.44	0.26	−1.73	[−0.95, 0.06]
White	−0.06	0.16	−0.39	[−0.38, 0.25]
Model F (16, 483)	28.68**
R²	0.487
Adjusted R²	0.470

Note. CI = confidence interval; OLS = ordinary least squares.

p < .01. **p < .001.

Results

Of all the onset type of offenses, only juvenile drug use was significantly associated with all three of the criminal career outcomes. As shown in Table 1, juvenile drug use (b = .06, z = 4.01, p < .001) was associated with chronicity/career arrests and its effect was only exceeded by age in terms of z score. Arrest onset produced a significant, inverse relationship to total arrests. Older offenders and males were also significantly likely to accumulate more arrests. Overall model fit was significant (LR χ² = 89.88, p < .001).

As shown in Table 2, juvenile drug use and age were the only significant predictors of 90 or more career arrests, which were measured at 1 SD above the mean. Juvenile drug use increased the likelihood of accumulating 90 or more arrests by 26% (OR = 1.26, z = 2.43, p < .01). Older offenders were also more likely to accumulate 90 or more arrests. Overall model fit was significant (LR χ² = 29.47, p < .01).

In terms of lambda, three types of onset were significant as shown in Table 3. Juvenile larceny, juvenile auto theft, and juvenile drug use were significantly associated with a higher annual offending pace. These findings suggest an association between juvenile property offending and greater offending frequency. Onset was positively associated with lambda, suggesting that offenders who begin their offending careers later are more prolific per year.⁴ Finally, a strong inverse effect was found for age. Despite very few significant effects, the model explained 47% of variation in lambda.

Discussion

Across disciplines in the social and behavioral sciences, it is understood that the earlier the antisocial career emerges, the more negative its developmental course and severity (Le Blanc, 2012; Moffitt, 1993; Moffitt & Caspi, 2001; Patterson, 1986). Despite hundreds of studies on the effect of onset, surprisingly few studies have examined the offense type of criminal debut, and whether it impacts the resulting criminal career. Drawing on data from a sample of serious adult criminal offenders, the current study examined 12 forms of juvenile delinquency (murder, rape, robbery, aggravated assault, burglary, larceny, auto theft, arson, weapons, sexual offense, drug sales, and drug use) in addition to age at arrest onset, age, sex, race to explore their association with chronicity (total arrests), extreme chronicity (1 SD above the mean which was equivalent to 90 career arrests), and lambda (offending per year). The only onset offense type that was significantly associated with all criminal career outcomes was juvenile drug use. In fact, juvenile drug use was the only form of criminal debut that predicted total arrests in either specification. For lambda, juvenile drug use, juvenile larceny, and juvenile auto theft were significant in addition to arrest onset and age.

Although it is a small literature, there is now evidence based on offenders from Sweden (Svensson, 2002), Australia (Mazerolle et al., 2010), and the United States (Harris, 2013) indicating that property-related delinquent onset is most strongly predictive of a habitual criminal career. The importance of property onset perhaps underscores the developmental and nonlinear course that typifies criminal careers (Jennings & Reingle, 2012). Almost never does the antisocial career follow a direct progression that moves from very low-level offenses before incrementally graduating to more severe ones. In this way, it is possible that larceny, auto theft, and related crimes represent a general antisocial tendency that given opportunity and situational factors could escalate to violence. Intuitively, it is logical to believe that violent onset offenders would be most likely to accumulate the most arrests and convictions, but empirically that has not been the case (Harris, 2013; Mazerolle et al., 2010). Here, juvenile murder, rape, robbery, aggravated assault, and sexual offenses were not significantly associated with the three outcome variables.

That juvenile drug use was so consistently related to the subsequent criminal career was partially unexpected. On one hand, the role of juvenile drug use as the launching pad for a sustained offending career is unsurprising given the interconnections between substance abuse and crime. On the other hand, juvenile drug use is significantly lower in terms of offense seriousness than most of the other forms of onset in the models—yet it was robustly predictive of frequency and chronicity of offending, and the effects held across specifications and analytical strategies (e.g., negative binomial regression, logistic regression, and OLS regression). Surprising too was the importance of juvenile drug use relative to juvenile drug sales/trafficking. Recently, Shook, Vaughn, Goodkind, and Johnson (2011) compared institutionalized adolescents who sold either marijuana, hard drug such as cocaine, or prescription drugs to their non-drug-selling peers who were also in placement facilities. Overwhelmingly, juvenile drug sellers were significantly more antisocial than their non-drug-selling peers evidenced by self-reports of various violent, property, and weapons offenses. Moreover, drug sellers tended to have worse substance abuse problems and drug histories than non-drug-selling peers (Shook et al., 2011; also see Gallupe et al., 2011; Maxwell & Maxwell, 2000; Phillips, 2012; Sommers, Baskin, & Fagan, 1996). In the current models, juvenile drug selling was not associated with more extensive criminal careers.

The current findings are relevant to the literature that examines various trajectories of substance use careers (Phillips, 2012; Sullivan & Piquero, 2010; Ward et al., 2010). Unfortunately, just as few criminal career studies have examined onset offense type, it is also understudied in the drug careers literature. An interesting exception examined the early-life predictors of substance use initiation using a nationally representative dataset. Stogner and Gibson (2011) found that general strain and/or health strains were predictive of alcohol onset, marijuana onset, cocaine onset, and the initiation of other forms of drug use. In addition, negative emotionality was significantly associated with cocaine onset. In this way, early forms of substance use reflect personality and social-psychological risk factors that themselves are associated with externalizing behaviors (Vaughn et al., 2012; Vaughn et al., 2011). And when these risks are coupled with drug use, the likelihood of criminal behavior is even greater.

There are limitations of these data. The juvenile drug measures were aggregate measures of substance use that did not specify a particular substance. In other words, it is unknown if juvenile cocaine users, juvenile crack users, juvenile heroin users, juvenile inhalant users, juvenile methamphetamine users, or others were independently or uniquely likely to become habitual offenders. Recent research on drugs of choice, for instance, found that offenders who reported that alcohol was their preferred substance were more likely to commit person or violent offenses. Those for whom cocaine was the drug of choice were significantly likely to commit property offenses, but not person or drug offenses. And those for whom marijuana was the drug of choice were most likely to be arrested for drug violations (Clark et al., 2012). It is possible that substance use onset for a particular substance is differentially related to the offending career. Future research should explore this issue.

Another limitation of these data is the lack of theoretically relevant measures to better understand how and why delinquent onset is associated with later behavior. For example, analyses of a national sample of youth selected from the National Longitudinal Study of Adolescent Health found that low self-control and association with delinquent peers were predictive of early onset marijuana use (Ragan & Beaver, 2010). To extrapolate their findings to the current study, it is possible that the robust effects of juvenile drug use in the current models could be viewed as a proxy for other antisocial features, such as low self-control or antisocial environments, such as associating with delinquent peers. A more nuanced dataset is needed to clarify this.

Despite these limitations, the current study has several notable advantages. The sample not only contains a large number of extreme habitual offenders (recall the minimum arrest criterion was 30 arrests) but also is heterogeneous in terms of the onset, course, and versatility of their offending careers (DeLisi, 2006). These data included multiple measures of juvenile delinquency to examine onset type along with multiple outcome measures associated with the criminal career. Of course, given the extremity of the sample, it is important for future research to explore and potentially replicate these models on a community/nonclinical sample.

To conclude, it is already well known that antisocial onset, especially when it occurs early in the life course, is a harbinger of antisocial conduct to come. Presently, criminologists are examining the offense type of criminal debut to see how separate forms of crime unfold in the criminal career. To date, it is property onset offenders who appear to be at greatest risk for habitual criminal conduct. The current models also suggest that juvenile drug users are also prone to a life of crime, and these findings were gleaned from a sample of career offenders.

Footnotes

Declaration of Conflicting Interests

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

Funding

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

Notes

References

Amirault

Lussier

(2011). Population heterogeneity, state dependence and sexual offender recidivism: The aging process and the lost predictive impact of prior criminal charges over time. Journal of Criminal Justice, 39, 344-354.

Andersson

Torstensson Levander

(2013). Adult onset offending in a Swedish female birth cohort. Journal of Criminal Justice, 41, 172-177.

Barnes

J. C.

Boutwell

B. B.

(2012). On the relationship of past to future involvement in crime and delinquency: A behavior genetic analysis. Journal of Criminal Justice, 40, 94-102.

Behnken

M. P.

Caudill

J. W.

Berg

M. T.

Trulson

C. R.

DeLisi

(2011). Marked for death: An empirical criminal careers analysis of death sentences in a sample of convicted male homicide offenders. Journal of Criminal Justice, 39, 471-478.

Boyum

D. A.

Caulkins

J. P.

Kleiman

M. A. R.

(2011). Drugs, crime, and public policy. In Wilson

J. Q.

Petersilia

(Eds.), Crime and public policy (pp. 368-410). New York, NY: Oxford University Press.

Brame

Mazerolle

Piquero

A. R.

(2010). Criminal career progression among serious youthful offenders in Australia. Journal of Criminal Justice, 38, 340-347.

Bureau of Justice Statistics. (2009). Survey of state criminal history information systems, 2008. Washington, DC: US Department of Justice, Office of Justice Programs, Bureau of Justice Statistics.

Clark

B. C.

Perkins

McCullumsmith

C. B.

Islam

A. M.

Sung

Cropsey

K. L.

(2012). What does self-identified drug of choice tell us about individuals under community corrections supervision? Journal of Addiction Medicine, 6, 57-67.

De Li

. (2005). Race, self-control, and drug problems among jail inmates. Journal of Drug Issues, 35, 645-664.

10.

DeLisi

(2001). Extreme career criminals. American Journal of Criminal Justice, 25, 239-252.

11.

DeLisi

(2003). The imprisoned nonviolent drug offender: Specialized martyr or versatile career criminal? American Journal of Criminal Justice, 27, 167-182.

12.

DeLisi

(2005). Career criminals in society. Thousand Oaks, CA: SAGE.

13.

DeLisi

(2006). Zeroing in on early arrest onset: Results from a population of extreme career criminals. Journal of Criminal Justice, 34, 17-26.

14.

DeLisi

Neppl

T. K.

Lohman

B. J.

Vaughn

M. G.

Shook

J. J.

(2013). Early starters: Which type of criminal onset matters most for delinquent careers? Journal of Criminal Justice, 41, 12-17.

15.

DeLisi

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.

16.

Elliott

D. S.

Huizinga

Menard

(1989). Multiple problem youth: Delinquency, drugs, and mental health problems. New York, NY: Springer-Verlag.

17.

Farabee

Joshi

Anglin

M. D.

(2001). Addiction careers and criminal specialization. Crime & Delinquency, 47, 196-220.

18.

Frick

P. J.

Lahey

B. B.

Loeber

Tannenbaum

Van Horn

Christ

M. G.

. . . Hanson

(1993). Oppositional defiant disorder and conduct disorder: A meta-analytic review of factor analyses and cross-validation in a clinic sample. Clinical Psychology Review, 13, 319-340.

19.

Gallupe

Bouchard

(2013). Adolescent parties and substance use: A situational approach to peer influence. Journal of Criminal Justice, 41, 162-171.

20.

Gallupe

Bouchard

Caulkins

J. P.

(2011). No change is a good change? Restrictive deterrence in illegal drug markers. Journal of Criminal Justice, 39, 81-89.

21.

Garcia

C. A.

Lane

(2012). Dealing with the fall-out: Identifying and addressing the role that relationship strain plays in the lives of girls in the juvenile justice system. Journal of Criminal Justice, 40, 259-267.

22.

Hammersley

(2011). Pathways through drugs and crime: Desistance, trauma, and resilience. Journal of Criminal Justice, 39, 268-272.

23.

Harris

D. A.

(2013). Age and type of onset of offending: Results from a sample of male sexual offenders referred for civil commitment. International Journal of Offender Therapy and Comparative Criminology, 57, 1226-247.

24.

Higgins

G. E.

Mahoney

Ricketts

M. L.

(2009). Nonsocial reinforcement of the nonmedical use of prescription drugs: A partial test of social learning and self-control theories. Journal of Drug Issues, 39, 949-964.

25.

Hochstetler

DeLisi

Puhrmann

A. M.

(2007). Toward an integrated model of offending frequency: A replication study. Justice Quarterly, 24, 582-599.

26.

Jennings

W. G.

Piquero

A. R.

Farrington

D. P.

(2013). Does resting heart rate at age 18 distinguish general and violent offending up to age 50? Findings from the Cambridge Study in Delinquent Development. Journal of Criminal Justice, 41, 213-219.

27.

Jennings

W. G.

Reingle

J. M.

(2012). On the number and shape of developmental/life-course violence, aggression, and delinquency trajectories: A state-of-the-art review. Journal of Criminal Justice, 40, 472-489.

28.

Joshi

Grella

C. E.

Hser

Y.-I.

(2001). Drug use and treatment initiation patterns: Differences by birth-cohorts. Journal of Drug Issues, 31, 1039-1062.

29.

Le Blanc

. (2012). Twenty-five years of developmental criminology: What we know, what we need to know. In Loeber

Welsh

B. C.

(Eds.), The future of criminology (pp. 124-133). New York, NY: Oxford University Press.

30.

Lussier

Bouchard

Beauregard

(2011). Patterns of criminal achievement in sexual offending: Unravelling the “successful” sex offender. Journal of Criminal Justice, 39, 433-444.

31.

Magyar

M. S.

Edens

J. F.

Lilienfeld

S. O.

Douglas

K. S.

Poythress

N. G.

(2011). Examining the relationship among substance abuse, negative emotionality and impulsivity across subtypes of antisocial and psychopathic substance abusers. Journal of Criminal Justice, 39, 232-237.

32.

Marlowe

D. B.

Festinger

D. S.

Dugosh

K. L.

Caron

Podkopacz

M. R.

Clements

N. T.

(2011). Targeting dispositions from drug-involved offenders: A field trial for Risk and Needs Triage (RANT). Journal of Criminal Justice, 39, 253-260.

33.

Maxwell

S. R.

Maxwell

C. D.

(2000). Examining the criminal careers of prostitutes within the nexus of drug use, drug selling, and other illicit activities. Criminology, 38, 787-810.

34.

Mazerolle

Piquero

Brame

(2010). Violent onset offenders: Do initial experiences shape criminal career dimensions? International Criminal Justice Review, 20, 132-146.

35.

Menard

Mihalic

Huizinga

(2001). Drugs and crime revisited. Justice Quarterly, 18, 269-299.

36.

Mitchell

Wilson

D. B.

Eggers

MacKenzie

D. L.

(2012). Assessing the effectiveness of drug courts on recidivism: A meta-analytic review of traditional and non-traditional drug courts. Journal of Criminal Justice, 40, 60-71.

37.

Moffitt

T. E.

(1993). Adolescence-limited and life-course persistent antisocial behavior: A developmental taxonomy. Psychological Review, 100, 674-701.

38.

Moffitt

T. E.

Caspi

(2001). Childhood predictors differentiate life-course persistent and adolescence-limited antisocial pathways among males and females. Development and Psychopathology, 13, 355-375.

39.

O’Brien

R. M.

(2007). A caution regarding rules of thumb for variance inflation factors. Quality & Quantity, 41, 673-690.

40.

O’Connell

Visher

C. A.

Martin

Parker

Brent

(2011). Decide your time: Testing deterrence theory’s certainty and celerity effects on substance-using probationers. Journal of Criminal Justice, 39, 261-267.

41.

Patterson

G. R.

(1986). Performance models for antisocial boys. American Psychologist, 41, 432-444.

42.

Pedneault

Harris

D. A.

Knight

R. A.

(2012). Toward a typology of sexual burglary: Latent class findings. Journal of Criminal Justice, 40, 278-284.

43.

Pettus-Davis

Howard

M. O.

Roberts-Lewis

Scheyett

A. M.

(2011). Naturally occurring social support in interventions for former prisoners with substance use disorders: Conceptual framework and program model. Journal of Criminal Justice, 39, 479-488.

44.

Phillips

M. D.

(2012). Assessing the impact of drug use and drug selling on violent offending in a panel of delinquent youth. Journal of Drug Issues, 42, 298-316.

45.

Ragan

D. T.

Beaver

K. M.

(2010). Chronic offenders: A life-course analysis of marijuana users. Youth & Society, 42, 174-198.

46.

Schroeder

R. D.

Ford

J. A.

(2012). Prescription drug misuse: A test of three competing criminological theories. Journal of Drug Issues, 42, 4-27.

47.

Sealock

M. D.

Manasse

(2012). An uneven playing field: The impact of strain and coping skills on treatment outcomes for juvenile offenders. Journal of Criminal Justice, 40, 238-248.

48.

Sharp

S. F.

Peck

B. M.

Hartsfield

(2012). Childhood adversity and substance use of women prisoners: A general strain theory approach. Journal of Criminal Justice, 40, 202-211.

49.

Shook

J. J.

Vaughn

Goodkind

Johnson

(2011). An empirical portrait of youthful offenders who sell drugs. Journal of Criminal Justice, 39, 224-231.

50.

Simon

L. M. J.

(1999). Are the worst offenders the least reliable? Studies on Crime and Crime Prevention, 8, 210-224.

51.

Sommers

Baskin

Fagan

(1996). The structural relationship between drug use, drug dealing, and other income support activities among women drug sellers. Journal of Drug Issues, 26, 975-1006.

52.

Stogner

Gibson

C. L.

(2011). The influence of health strain on the initiation and frequency of substance use in a national sample of adolescents. Journal of Drug Issues, 41, 69-94.

53.

Sullivan

C. J.

Piquero

A. R.

(2010). Investigating stability and change in substance use and criminal activity using a synthesized longitudinal modeling approach. Journal of Drug Issues, 40, 63-92.

54.

Svensson

(2002). Strategic offenses in the criminal career context. British Journal of Criminology, 42, 395-411.

55.

Thornberry

T. P.

Krohn

M. D.

Freeman-Gallant

(2006). Intergenerational roots of early onset substance use. Journal of Drug Issues, 36, 1-27.

56.

Vaughn

(2011). Addiction and criminal justice: Empirical findings and theory for advancing public safety and health. Journal of Criminal Justice, 39, 205-206.

57.

Vaughn

M. G.

DeLisi

Beaver

K. M.

Perron

B. E.

Abdon

(2012). Toward a criminal justice epidemiology: Behavioral and physical health of probationers and parolees in the United States. Journal of Criminal Justice, 40, 165-173.

58.

Vaughn

M. G.

DeLisi

Gunter

Beaver

K. M.

Perron

B. E.

Howard

M. O.

(2011). The severe 5%: A latent class analysis of the externalizing behavior spectrum in the United States. Journal of Criminal Justice, 39, 75-80.

59.

Vaughn

M. G.

Wernet

S. J.

DeLisi

Beaver

K. M.

Perron

B. E.

Howard

M. O.

(2011). Characteristics of abstainers from substance use and antisocial behavior in the United States. Journal of Criminal Justice, 39, 212-217.

60.

Walters

G. D.

DeLisi

(2013). Antisocial cognition and crime continuity: Cognitive mediation of the past crime–future crime relationship. Journal of Criminal Justice, 41, 135-140.

61.

Ward

J. T.

Stogner

Gibson

C. L.

Akers

R. L.

(2010). A new trick for an old dog: Applying developmental trajectories to inform drug use progression. Journal of Drug Issues, 40, 755-782.

62.

Zhang

Wieczorek

W. F.

Welte

J. W.

(2011). Early onset of delinquency and the trajectory of alcohol-impaired driving among young males. Addictive Behaviors, 36, 1154-1159.