Predicting Specialization in Offending Over the Life Course

474262 CJBXXX10.1177/0093854812474262Crimin
al Justice and BehaviorBaker et al. / Specialization in Offending Over the Life Course
Predicting Specialization in Offending
Over the Life Course
Virginia Commonwealth University
Florida State University
University of South Florida
A number of methodological techniques and theoretical propositions have been used in the extant literature on specialization/
versatility. These various methodologies and theories have created an ongoing debate that has revealed several areas of focus
that need to be addressed. The current study applies Moffitt’s (1993) taxonomy and uses a series of random effects logistic
regression models to estimate the impact of a violent, property, drug, or other prior offense on the commission of a subsequent
offense across different types of criminal careers. Specifically, we attempt to determine the odds a prior offense has on predicting a subsequent offense among four offender trajectories, controlling for various aspects of the criminal career and
demographic characteristics. Findings suggest that the odds of committing any offense type over the life course are greater
if the prior offense was of the same type. This relationship remains consistent across offender trajectories.
Keywords: specialization; versatility; life course; trajectory modeling; offending
The ongoing debate between specialization and versatility in offending is both theoretical
and methodological. Scholars like Gottfredson and Hirschi (1990) contend that there is
great variability in the criminal acts that an individual engages in and it is unreasonable to
assume that offenders specialize in particular types of crime. Alternatively, positive learning
theories suggest that certain offenders will repeat specific forms of crime, insinuating the
possibility of specialization and distinctive subcultures of delinquency based around particular crimes (Cloward & Ohlin, 1960; Gottfredson & Hirschi, 1990; Moffitt, 1993). Early
research using transition matrices, factor analyses, and simple comparisons of crime
involvement demonstrate that there is a lot of versatility in offending with minor specialization (Bursik, 1980; Farrington, Snyder, & Finnegan, 1988; Hindelang, 1971; Hindelang,
Hirschi, & Weis, 1981; Kempf, 1987; Klein, 1984; Petersilia, 1980; Rojek & Erickson,
1982; Wolfgang, Figlio, & Sellin, 1972). Yet, current research using more advanced statistical methods, such as log-linear methods, latent class analysis, item response theory, and
AUTHORS’ NOTE: Correspondence concerning this article should be addressed to Thomas Baker, L.
Douglas Wilder School of Government and Public Affairs, Virginia Commonwealth University, 923 West
Franklin Street, Richmond, VA 23284, USA; email: [email protected]
CRIMINAL JUSTICE AND BEHAVIOR, Vol. 40, No. 8, August 2013, 909-932.
DOI: 10.1177/0093854812474262
© 2013 International Association for Correctional and Forensic Psychology
short-term data from smaller time periods, finds higher levels of specialization (Britt, 1996;
Francis, Liu, & Soothill, 2010; Francis, Soothill, & Fligelstone, 2004; McGloin, Sullivan,
& Piquero, 2009; Osgood & Schreck, 2007; Sullivan, McGloin, Pratt, & Piquero, 2006;
Sullivan, McGloin, Ray, & Caudy, 2009).
Because of this theoretical and methodological debate, specialization has become one
of the most widely studied topics in criminal career research, appealing to crime scholars
and public policy specialists alike (Mazerolle, Brame, Paternoster, Piquero, & Dean,
2000). Furthermore, its important implication for theory and policy has led scholars to
continue to refine their measures and explore nuances of specialization in both the short
term and the long term. These efforts have produced an extensive literature testing the
presence of specialization/versatility among various types of offenders using multiple
methods. A specialist is identified in the literature as having a greater tendency to repeat
the same crime or offense, whereas a nonspecialist commits different offenses on various
occasions, with no inclination to pursue a specific criminal act or pattern of criminal acts
(Britt, 1996; Gottfredson & Hirschi, 1990; Paternoster, Brame, Piquero, Mazerolle, &
Dean, 1998). Therefore, tests of specialization are designed to determine whether an
offender is more likely to commit the same offense or a different offense over a particular
time span.
Even among those studies that find greater versatility and less specialization, it is noted
that the highest probabilities in the transition from one offense to the next are the transitions between similar offenses (Bursik, 1980; Kempf, 1987; Wolfgang et al., 1972).
Considering that many tests of specialization using various methods are criticized for
random error in the calculations, confounding effects, unclear meanings, aggregation, and
the inability to identify the type of crime (Britt, 1996; Farrington, 1986; McGloin,
Sullivan, Piquero, & Pratt, 2007; Sullivan et al., 2009), these findings must be considered
with some degree of caution. However, what remains clear throughout these studies is that
when specialization is observed, it is because there is a larger probability for specialization, demonstrated either by the transition between offenses or the creation of latent
In order to advance the theoretical and methodological debate surrounding specialization/
versatility, this study sets out to expand upon prior methods of testing specialization using
the 1958 Philadelphia Birth Cohort Study (Tracy & Kempf-Leonard, 1996; Tracy,
Wolfgang, & Figlio, 1990). We consider the theories and methods that have been used to
test specialization/versatility and select a theoretical framework in which to study specialization that addresses some of the methodological concerns while also considering that
not all offenders have similar trajectories over the life course. Specifically, we consider
the propositions established by Moffitt (1993) and create offense trajectories to simulate
the groups delineated by her. We then disaggregate by these trajectories to determine
whether specialization is more likely for certain offenses depending on the type of
offender. To test for specialization, we use a series of random effects logistic regression
models to predict the likelihood that an offender’s subsequent offense is the same as the
prior offense he or she committed for violent, property, drug, and other crimes across the
life course. Both the theoretical framework and the actual model facilitate in accounting
for several limitations of prior studies, including the lack of control for other variables,
the inability to account for multiple types of offenses and offending patterns, and the
tendency to aggregate the data.
Gottfredson and Hirschi (1990) identify four elements in their notion of low self-control:
(1) stability of individual differences over time, (2) great variability in the kinds of criminal
acts committed, (3) conceptual equivalence of criminal and noncriminal acts, and (4) inability to predict forms of deviant engagement. Of central importance for the specialization/versatility debate is their emphasis on the variability in the offenses committed by individuals.
Because they believe that crime will be engaged in at a high rate by people with low selfcontrol, they conclude that there will be versatility in offending as a consequence. To bolster
their argument, they point to overwhelming evidence of versatility in the literature, including the Wolfgang et al. (1972) cohort study, the work by Hindelang and colleagues (1971,
1981), and the research of Rojek and Erickson (1982). According to Gottfredson and Hirschi
versatility even extends to analogous behaviors, including truancy, dropping out of school,
employment instability, alcohol use, drug use, child and spouse abuse, motor vehicle accidents, and unrestrained sexual activity (Gottfredson & Hirschi, 1994).
In the discussion of their theoretical propositions, Gottfredson and Hirschi (1990) address
the concept of the criminal career as well. They note that the career model implies a certain
terminology, including specialization, amount of time and effort devoted, level of accomplishment, productivity, current direction, overall shape, and time out for other activities.
According to the theorists, crime as a career is refuted by their concept of low self-control,
and it is not reasonable to assume criminals can specialize, crime escalates, or even that there
is intermittency or a “time out” between offenses (p. 266). Specifically, the labels placed on
offenders in relation to these careers, such as robber, burglar, drug dealer, rapist, and murderer, are “retrospective rather than predictive” and they ignore criminal behavior committed
by these offenders that is not consistent with the label they have been given (p. 92).
Gottfredson and Hirschi (1994) do recognize that specialization in offenses can occur
for some offenders. They argue that this apparent specialization that is sometimes
observed is a function of opportunity and circumstance and not necessarily a characteristic of the offender. As Osgood and Schreck (2007) note, Gottfredson and Hirschi claim
that a pattern whereby an individual commits a disproportionate number of identical
offenses “reflects erratic variation in opportunities rather than lasting individual traits or
preferences” (p. 301). If this view is taken, it is probable that offenders will occasionally
specialize, and therefore any finding of specialization is an artifact of this inherent probability and not necessarily an indicator of a specialized career.
While Gottfredson and Hirschi (1990) focus on low self-control and how individuals
low in self-control will experience versatile offending throughout the life course, Farrington’s
(2005) integrated cognitive antisocial potential (ICAP) theory focuses on antisocial potential and how high levels of antisocial potential can lead to versatile offending patterns.
According to his theory, people with high antisocial potential are more likely to commit
many different types of antisocial acts, including different types of offenses. This would
mean that offending and antisocial behavior are versatile, not specialized. Similar to
Gottfredson and Hirschi, Farrington recognizes that the commission of offenses and antisocial acts depends on the interaction between individuals and their social environments,
which means that the commission of delinquent acts is dependent on opportunity. Following
the logic of Gottfredson and Hirschi (1994), any observation of specialization would be an
artifact of opportunity.
Alternatively, Moffitt’s (1993) taxonomy is built around the criminal career model and
explicitly recognizes the possibility of specialization among certain types of offenders. She
identifies two groups of offenders: adolescence-limited (AL) and life course–persistent
offenders (LCP). ALs generally follow the age-crime curve, committing the majority of
their offenses in their teenage years and dropping out of crime shortly thereafter. These
offenders lack consistency in their criminal behavior, often exhibiting sporadic, crime-free
periods throughout their short criminal careers. AL offenders are expected to engage primarily in crimes that symbolize their adulthood, such as vandalism, public order offenses,
substance use, and theft. Unlike ALs, LCP offenders continuously offend throughout the
life course. These offenders fall into a pattern of criminal behavior that becomes inescapable as they continue to suffer the consequences of their behavior and fail to learn prosocial alternatives to crime. LCP offenders are expected to engage in a wide variety of
offenses, including violence and fraud, and they are expected to have lower levels of
self-control and would presumably have higher levels of antisocial potential. Therefore, it
is LCPs that resemble the offenders identified by Gottfredson and Hirschi (1990) and
Farrington (2005).
Based on Moffitt’s (1993) propositions, specialization becomes a plausible possibility in
the commission of offenses across the life course. In accordance with this possibility, recent
studies have looked at specialization for specific types of offenders and offense overlap
among offenders to determine if there is evidence of specialization. Specifically, Rosenfeld,
White, and Esbensen (2012) found a large amount of overlap among different types of
offending during adolescence and young adulthood and from adolescence into young adulthood, suggesting versatility over specialization, especially among drug dealers. However,
they found both versatility and specialization in sex and homicide offending, in that these
offenders were more likely to be arrested for another sex offense or homicide than were
those imprisoned for other types of offenses. The authors also noted that the modal offense
category for prior arrests of each crime type was the offenses for which the individual was
imprisoned. For example, about 88% of those who were in prison for a property crime had
a prior arrest for property crime.
In a similar fashion, Loeber, Farrington, Stouthamer-Loeber, and Raskin White (2008)
used data from the oldest and youngest cohorts of the Pittsburgh Youth Study, a self-report
study, to look at specialization and versatility, among other elements of the criminal career.
According to their results, there was a considerable amount of specialization in serious
violence and theft, with a little less than half of all violent offenders specialized in violent
offenses and one third of all theft offenders specialized in theft offenses. Similar to these
findings, trajectory models also showed more evidence for violence specialization than
theft specialization. In addition, they found partial confirmation that offense specialization
increases with age and that specialization may depend on the level of severity of the
offense. The authors concluded that “the data contradict the notion that specialized offenders are uncommon and generalized offending is the rule” (p. 128).
Essentially, Moffitt (1993) predicts that ALs specialize in the delinquent offenses they
commit, whereas LCPs are more versatile, committing multiple types of offenses over their
life course. In this sense, versatility is only characteristic of life course–persistent offenders. The switching between offenses that Moffitt refers to in these scenarios is the switch
that occurs between categories of offenses, and not necessarily within categories of offenses
(Piquero, Paternoster, Mazerolle, Brame, & Dean, 1999). Therefore, any specialization that
occurs among AL offenders should involve less serious forms of crime, such as property,
drug, and other lower forms of crime. Because LCPs are expected to commit more violent
forms of crime, specialization among violent offenses should be unlikely (Piquero,
Farrington, & Blumstein, 2007), although research to date seems to indicate otherwise
(Loeber et al., 2008; Rosenfeld et al., 2012).
Because specialization/versatility is such a highly debated topic theoretically, a methodological debate was initiated in an effort to find the best way to statistically capture
specialization/versatility. In the research to date, most studies incorporate probabilities and
indices built from probabilities to capture specialization/versatility. The original use of
probabilities dates back to the 1970s and 1980s with the initiation of transition matrices, or
aggregate probabilities of offending behavior (Kempf, 1987). Transition matrices reflect
the probability that an individual committed a certain offense on arrest k and then committed the same offense on arrest k + 1 (Bursik, 1980; Wolfgang et al., 1972). According to the
transition matrices, there was a lot of versatility among offenses, but as referenced previously, the highest probabilities were the transitions between similar offenses, suggesting
some specialization (Bursik, 1980; Kempf, 1987; Wolfgang et al., 1972).
These original transition matrices were further complicated by the introduction of the
Forward Specialization Coefficient (FSC) as an index of the degree of specialization on a
scale from 0 to 1, where 0 is complete versatility and 1 is perfect forward specialization
(Farrington, 1986; Farrington et al., 1988). Studies using the FSC have reached various
conclusions. Whereas Kempf (1987) found low overall specialization based on low FSC
values, Farrington et al. (1988) found low FSC values also, but these values were significantly different from zero, suggesting evidence of forward specialization. According to
their analysis, persistent offenders were more specialized, whereby the average FSC of
persistent offenders increased with each transition. Armstrong (2008) noted the influence
of age on this relationship and the tendency for higher specialization among persistent
offenders in drug and miscellaneous offenses, but not in violent and property offenses.
Piquero et al. (1999) also found that higher FSCs, which demonstrate greater forward specialization, are characteristic of offenders with a late age of onset. Overall, Farrington et al.
(1988) concluded that there seems to be some evidence of specialization among a lot of
versatility in offending.
Due to criticisms of the FSC, scholars have continued to use probabilities, but in
improved methods that take into account some of the issues related to the FSC and earlier
transition matrices, such as random error, unclear meanings, aggregation, and confounding
effects of frequency of offending and sample size (Britt, 1996; McGloin et al., 2007). For
instance, the FSC cannot determine whether specialization exists in an individual’s criminal career; it only addresses whether there is specialization within the offending patterns of
the entire sample (Sullivan et al., 2006). Britt (1996) proposed a measure of specialization
based on the log-linear analytic method. Offense specialization is when the log-odds of
repeating the same offense are greater than switching between offenses for each pair of
offense types. Using this model, the findings indicated that there was a greater likelihood
of repeating the same offense rather than switching offenses, providing stronger evidence
of specialization (Britt, 1996). There was also evidence that specialization varied across
groups of offenders based on race. Stemming from this log-linear method, Armstrong and
Britt (2004) used multinomial logistic regression for five offense types across 10 transitions
to find increases in the probability of repeating a given offense for miscellaneous offenses,
drug offenses, and violent offenses. They also found that when controlling for correlates of
offending, the average offender is not likely to repeat the same offense, but this varies when
considering age at time of arrest.
As another alternative method to the FSC, some researchers have used Agresti and
Agresti’s (1978) diversity index. The diversity index is the probability that any two
offenses drawn randomly from an individual’s set of offenses belongs to two different
offending categories (Piquero et al., 1999). The index ranges from a minimum value of 0,
representing the least diverse population (specialization), to a maximum value of (k – 1)/k,
representing the most diverse population that is evenly spread over k categories (Agresti
& Agresti, 1978). Because this is a probability based on an individual’s set of offenses, it
measures specialization at the individual level and captures the overall versatility of an
individual offender (Mazerolle et al., 2000).
Using the diversity index, scholars assess specialization in terms of age, onset age, gender,
persistent offending, and local life circumstances. Studies using the diversity index show
that versatility increases as age and age of onset decrease and as the number of offenses
increases (Mazerolle et al., 2000; McGloin et al., 2007; Piquero et al., 1999). Therefore,
younger early onsetters that are more persistent offenders have greater offending diversity
among both men and women (Mazerolle et al., 2000). It appears that early adulthood is
characterized by increased diversity in offending that decreases into adulthood (Nieuwbeerta,
Blokland, Piquero, & Sweeten, 2011). It has also been found that specialization varies
across offending trajectories, where sporadic offenders are more specialized and frequent
offenders are more diverse (Nieuwbeerta et al., 2011).
Also incorporating the use of probabilities, some researchers have turned their attention
to latent class analysis. In these models, specialization is classified depending on whether
the identified latent classes reflect subgroups around distinct offense types, indicating specialized groups (Francis et al., 2004; Sullivan et al., 2009). Other research has expanded
latent class analysis to latent transition analysis, “an autoregressive model that examines
change and stability in latent groups over time” (McGloin et al., 2009, p. 249). It provides
a probability of transition across latent categories and infers whether individuals show
change and the form of change in offending behavior (McGloin et al., 2009). Studies using
both models have found distinct latent classes involving drugs, burglary, theft, and shoplifting, with some tendency toward versatility depending on the age and gender of the offender
(Francis et al., 2004, 2010; McGloin et al., 2009).
A more recent study by Osgood and Schreck (2007) used item response theory to model
specialization as a latent variable. Through a multilevel analysis, they assessed the presence
of specialization in violence by determining the log-odds that an individual committed an
offense depending on (a) individual differences in the rate of offending or the tendency of
an individual to offend, (b) the base rate of that offense (with rare offenses having lower
values), and (c) the probability of committing each offense or specialization (the difference
between the log odds of committing a violent offense versus the log odds of committing a
nonviolent offense). This model distinguishes between offending frequency and versatility,
and it considers individual patterns of specialization relative to offense patterns of the
population (Sullivan et al., 2009). Using this model, specialization has been found among
violent offenses and intimate partner violence among women (Bouffard, Wright, Muftic, &
Bouffard, 2008; Osgood & Schreck, 2007).
Through the evolution of methods to study specialization/versatility, it has become clear
that specialization is more apparent than the early research suggests (Bouffard et al., 2008;
Francis et al., 2004, 2010; Osgood & Schreck, 2007; Sullivan et al., 2006). These studies
demonstrate that it is important to separate specialization from offense frequency (Farrington,
1986), address specialization at the individual level (McGloin et al., 2007; Sullivan et al.,
2006), consider the unique offending patterns of individuals (Osgood & Schreck, 2007;
Sullivan et al., 2009), and indicate the type of crimes for which offenders tend to specialize
(Sullivan et al., 2009). Measures of specialization that can be employed in regression
analyses are also beneficial, since the researcher can control for other factors that may be
impacting the likelihood of specialization (McGloin et al., 2007; Osgood & Schreck,
2007). Conclusions from these analyses reveal that specialization/versatility (a) varies
depending on the type of offender, specifically in relation to persistent offenders and
offense trajectories, and (b) age, onset age, race, offense rank, and gender are important
predictors of specialization/versatility and should be considered within regression analyses.
Addressing both the theoretical and methodological debate, this study uses a random
sample of 2,500 individuals with two or more offenses, responsible for a total of 14,649
offenses, from the 1958 Philadelphia Birth Cohort Study (Tracy et al., 1990; Tracy &
Kempf-Leonard, 1996). We limit the study to offenders with at least two offenses in order
to capture patterns of specialization (McGloin et al., 2007; Piquero et al., 1999) and
offenses committed after the age of 7, since 7 is generally considered to be the minimum
age at which individuals may be held accountable for their actions (Bernard, 1992, 2007;
Farrington & Welsh, 2007; Tanenhaus, 2004; Zimring, 2005). This data set contains various
measures for individuals from birth to age 26, who were originally born in Philadelphia in
1958. Until age 18, police contact data were recorded, along with records of “remedial, or
informal, handling of the youth by an officer” (Tracy & Kempf-Leonard, 1996, p. 65).
Once the cohort reached age 18, “court files included police reports, so data on adult crime
are comparable to that for delinquency,” but there were no remedial reports for adults
(Tracy & Kempf-Leonard, 1996, p. 65). The individual measures and police contact/arrest
information combine to form a nested data set of offenses within individuals.
The dependent variable consists of various offenses combined into offense categories
based on the nature of the offense. Specifically, four offense types were created: violent,
property, drug, and other. Violent offenses are operationalized as all offenses identified by
the police as homicide, rape, robbery, assault, weapons possession, and domestic violence.
Similar to Piquero et al. (1999), we classified robbery as a violent crime. Burglary, theft,
auto theft, arson, forgery, fraud, embezzlement, stolen property, and vandalism are considered
property offenses. Included among the drug offenses are narcotic and drug law violations.
Other offenses are comprised of prostitution, sex offenses, gambling, driving while intoxicated, liquor law violations, drunkenness, disorderly conduct, vagrancy, minor disturbances, and miscellaneous offenses. The miscellaneous offenses classification is rather
ambiguous and there is little additional information to help further elucidate this group of
offenses; still, we think it is important to include this classification of offenses even if we
are unable to identify a specific offense for which individuals came into contact with the
police. Each offense type was dichotomized accordingly into four variables: violent (1) and
nonviolent (0), property (1) and non-property (0), drug (1) and non-drug (0), and other (1)
and non-other (0) offenses.
We are aware that the broad categorization of offenses can potentially hide the switching
that occurs within offense types (Piquero et al., 1999). As recognized by Piquero et al.
(1999), Moffitt combines property and other offenses into a nonviolent category to compare
against violent offenses in her own research (Moffitt, Caspi, Dickson, Silva, & Stanton,
1996). Following this logic, it is clear that she is more concerned with the switch that occurs
between broad categories of offenses, not within categories of offenses. In terms of her
research, our categorizations are slightly less broad, considering that we create four different
categories of offenses, but they still fall into the broad conceptualization she advocates.
Also, because of the tendency of offenses to aggregate toward distinct groups and the desire
to maintain clarity and simplicity, Cohen (1986) and Spelman (1994) have recommended the
use of broad categories of offenses. We are aware that others, such as Farrington et al. (1988)
and Rosenfeld et al. (2012), have identified certain offenses within these categories as specialized offenses. However, following the theoretical foundation of this study, and in order
to create a clear and simplified analysis, we maintain the violent, property, drug, and other
crime distinctions.
The main independent variable in the analysis is type of prior offense. Four dichotomous
variables were created to denote whether the prior offense was violent (yes = 1), property
(yes = 1), drug (yes = 1), or other (yes = 1). The first offense was dropped for all models
because there is no prior offense (J) to predict the subsequent offense (J + 1). Separate
models were analyzed for each type of offense J predicting each type of offense J + 1 so
that, for example, the odds of having a recorded police contact for a violent offense were
predicted if the prior recorded police contact was for a violent offense, property offense,
drug offense, or other offense.
The gap between contacts, or intermittency, is operationalized as the months between
offenses. Recorded police contacts that occurred in the same month and police contacts that
occurred in consecutive months were scored 0, such that there were 0 months between
police contacts. The 1958 Birth Cohort Study only contains the month and year of each
contact, so we are limited to studying intermittency as the months between contacts. Recent
research has noted the importance of intermittency in the study of criminal careers and its
effect on offense seriousness (Baker, Falco Metcalfe, & Piquero, in press). Intermittency
is also a confounding factor in analyses including age, age of onset, sex, and socioeconomic status (SES), since it is dependent on each of these factors (Baker et al., in press;
Piquero, 2004; Piquero et al., 2007). Because of this, we control for intermittency in our
models predicting offense specialization.
Following prior research (Patterson et al., 1992; Piquero et al., 2007; Tibbetts &
Piquero, 1999), age of onset is a dichotomous variable coded early onset (1) for offenders
with an onset age before 14 years and late onset (0) for offenders with an onset age of 14
years or older. A dichotomous measure of onset age was chosen, as opposed to a continuous
measure, to align with the theoretical propositions of Moffitt (1993) and to address the
potential confounding effects of age and age of onset suggested by prior research (Piquero
et al., 1999). Age is a continuous variable measured in years, while SES,1
race, and sex are
dichotomized by high SES (1) and low SES (0), White (1) and non-White (0), and male (1)
and female (0).
The analysis in the current study involves a twofold process by first estimating trajectories for the random sample of 2,500 individuals with two plus offenses and then analyzing
the data using random effects logistic regression models disaggregated by the offense trajectories. The trajectories are designed to address the propositions of Moffitt (1993).
According to Moffitt, adolescence-limited offenders should exhibit patterns of specialization, especially among less serious offense types, while life course–persistent offenders
should demonstrate greater versatility. The trajectories also address the methodological
concern regarding consideration of the unique offending patterns of individuals within
models of specialization/versatility (Osgood & Schreck, 2007).
The choice of random effects models is meant to address some of the methodological
issues while also providing an alternative method of testing for specialization/versatility.
By using random effects logistic regression models, we are able to test the odds that police
contact J + 1 falls into the same offense category as police contact J for four types of
offenses: violent, property, drug, and other offenses. Using random effects logistic regression to predict the odds of coming into contact with the police for offense J + 1 extends
prior research exploring specialization, addressing several of the methodological flaws
identified by prior studies. Specifically, while past research predicting specialization uses
transition matrices (Wolfgang et al., 1972), forward specialization coefficients (Farrington,
1986), diversity indices (McGloin et al., 2007; Piquero et al., 1999; Sullivan et al., 2006),
latent classes (Francis et al., 2004, 2010; McGloin et al., 2009), and/or item response theory
(Osgood & Schreck, 2007), each is limited either from the inability to control for other
substantive variables of interest, account for more than one type of offense (i.e., violent vs.
nonviolent offenses), or measure specialization at the individual level. Acknowledging
these limitations, the current analysis provides a method for remedying these issues by
predicting the odds of coming into contact with the police for offense type J + 1 based on
prior police contact for offense type J, controlling for time-varying factors, such as type of
prior offense, intermittency, and age, and time-invariant factors, such as age of onset, SES,
sex, and race.
While other methods have their own advantages, the random effects logistic regression models applied in this study (a) adjust for the differing numbers of police contacts
per person, (b) provide an individual-level approach to studying specialization/versatility,
and (c) focus on different types of crime (Farrington, 1986; McGloin et al., 2007;
Osgood & Schreck, 2007; Sullivan et al., 2006, 2009). Furthermore, disaggregating by
offense trajectories accounts for individual offending rates, which Moffitt (1993) posits
and Osgood and Schreck (2007) demonstrate could influence specialization/versatility.
The offender trajectories are graphically displayed in Figure 1. Considering that the
dependent variable in this analysis was a count-based measure representing the number of
police contacts by age, and there was an overrepresentation of zeros in the data, a zeroinflated semi-parametric Poisson model was employed (Nagin, 2005). Utilizing the Proc
Traj macro available in SAS (Jones & Nagin, 2007), the final trajectory solution was determined through an iterative process based on maximum likelihood, whereby intercept only,
linear, quadratic, and cubic parameters were estimated across J + 1 trajectory groups until
the Bayesian Information Criterion (BIC) was maximized. In addition, the mean posterior
probabilities for trajectory group assignment were all well above the .70 cutoff argued by
Nagin (2005), suggesting that the final trajectory solution demonstrated a relatively high
degree of precision when assigning individuals to a particular group-based trajectory.
Before reviewing the trajectory results estimated in the current study, it is considerably
important to note here that the trajectory method has been criticized for what has been
referred to as “the reification of groups” (Sampson & Laub, 2005b). In response to this
criticism, Nagin and Tremblay (2005a, 2005b) have provided additional clarifications in
this regard by stating that the groups are labeled not as an exercise in reification, rather they
are assigned a label purely for “literary convenience.” Or in other words, these labels
facilitate the discussion of comparisons and contrasts that can be made between similar
and/or different observable patterns of behavior across trajectory groups. Furthermore,
“discovering” trajectory groups is a statistical approximation (e.g., sometimes you may
find two groups, sometimes three groups, etc.; for a review, see Jennings & Reingle, 2012),
and as such, the observed trajectories in any particular study with any sample are not meant
to be interpreted as factual representations of reality (a common limitation of all statistical
models). In addition, Sampson and Laub (2005c) even argued that they were not suggesting
that trajectory models should not be used or are not relevant, they just wanted to acknowledge the potential limitation that modelers may apply and misinterpret the results (e.g., or
reify groups), which is a mistake akin to researchers making errors in interpreting regression results as causal versus correlational (for further discussions, also see Maughan, 2005;
Nagin & Tremblay, 2005c, 2005d; Raudenbush, 2005; Sampson & Laub, 2005a). Finally,
Figure 1: 1958 Philadelphia Birth Cohort Trajectories: Random Sample of Offenders With Two Plus
Offenses (n = 2,500)
Skardhamar (2010; although also see Brame, Paternoster, & Piquero, 2012, in defense of
the method) provided a simulation experiment with trajectory models and concluded with
offering the following words of caution when applying trajectory models:
It might be that the direct interpretation of the trajectory groups is reasonable in some special
cases, but this finding will have to be justified explicitly in each application. Given the predominantly exploratory approach in criminological studies that apply SPGM, the findings
from these studies are typically equally in accordance with both general and typological
approaches and, therefore, are not a test of either. (p. 315)
Acknowledging the importance of the reification issue and related debate reviewed previously, though our trajectory solution produced four separate trajectories, they fall into
two distinct categories of offenders—adolescent peaked offenders and chronic offenders
(see Figure 1). We use these categorizations to stay consistent with prior trajectory analyses
of the 1958 Philadelphia Birth Cohort Study where four similar trajectories of offenders
were produced and labeled as such (D’Unger, Land, & McCall, 2002; D’Unger, Land,
McCall, & Nagin, 1998). To avoid confusion, however, it is important to note that chronic
refers to persistence in offending, not frequency. Frequency in offending is captured in the
delineation between low-rate and high-rate offending. Therefore, these two categorizations
resemble Moffitt’s (1993) ALs and LCPs. Finding multiple trajectories within these
offending categories is not uncommon (see e.g., Ferguson, Horwood, & Nagin, 2000;
Nagin, Farrington, & Moffitt, 1995). Though Moffitt originally posited two patterns of
offending, she concedes that the original AL/LCP groups were “defined several years ago
using cutoff methods that have recently been superseded by trajectory-detection methods
less susceptible to misclassification error (Nagin, 1999; Roeder et al., 1999). Our future
studies will adopt trajectory methods” (Moffitt, Caspi, Harrington, & Milne, 2002, p. 201). As
such, we have chosen the trajectory method and our four-group model to test the propositions of Moffitt’s (1993) original assertions regarding ALs and LCPs.
Descriptive statistics for each offender trajectory are presented in Table 1. Due to the fact
that the main independent variables are dichotomous, the means for each statistic are the
percentage of the total sample of each trajectory that came into contact with the police for
a particular offense. For example, a mean of .22 for prior violent offenses among low-rate
adolescent peaked offenders indicates that 22% of all prior police contacts of low-rate
adolescent peaked offenders were for violent offenses. This allows us to see which trajectory had the highest percentage of police contacts for each offense. High-rate chronic
offenders have the highest percentage of police contacts for prior violent offenses (30%
were violent), high-rate adolescent peaked offenders have the highest percentage of police
contacts for prior property offenses (45% were property), low-rate chronic offenders have
the highest percentage of police contacts for prior drug offenses (10% were drug), and lowrate adolescent peaked offenders have the highest percentage of police contacts for prior
other offenses (40% of the offenses). This pattern holds for the dependent variable as well,
with the exception of violent crime, since 33% and 32% of police contacts of low-rate
chronic and high-rate chronic offenders, respectively, are for violent offenses.
In terms of career attributes, the high-rate offenders have the shortest gaps between police
contacts, and the low-rate offenders have the longest gaps, with low-rate chronic offenders
experiencing the longest time between police contacts on average (23.45 months). The
majority of offenders are early onsetters, except for low-rate chronic offenders (35% early
TABLE 1: Descriptive Statistics by Offender Trajectory
Low-Rate Adolescent
Peaked (4,160, 4.04)a
High-Rate Adolescent
(1,949, 20.51)
Low-Rate Chronic
(4,066, 3.84)
High-Rate Chronic (4,474,
Mean SD Range Mean SD Range Mean SD Range Mean SD Range
Independent variables
Prior offense
Violent 0.22 0.41 0 to 1 0.24 0.43 0 to 1 0.28 0.45 0 to 1 0.30 0.46 0 to 1
Property 0.33 0.47 0 to 1 0.45 0.50 0 to 1 0.32 0.47 0 to 1 0.38 0.48 0 to 1
Drug 0.05 0.22 0 to 1 0.02 0.13 0 to 1 0.10 0.30 0 to 1 0.08 0.27 0 to 1
Other 0.40 0.49 0 to 1 0.30 0.46 0 to 1 0.30 0.46 0 to 1 0.24 0.43 0 to 1
Career attributes
Intermittency 10.25 13.26 0 to 114 5.66 10.17 0 to 110 23.45 27.58 0 to 191 9.13 12.73 0 to 113
Early onset 0.58 0.49 0 to 1 0.97 0.17 0 to 1 0.35 0.48 0 to 1 0.63 0.48 0 to 1
Male 0.88 0.33 0 to 1 1.00 0.00 1 0.90 0.30 0 to 1 0.96 0.20 0 to 1
White 0.27 0.44 0 to 1 0.23 0.42 0 to 1 0.23 0.42 0 to 1 0.18 0.38 0 to 1
Age 15.68 1.87 7 to 26 14.85 2.83 7 to 26 19.74 3.86 7 to 26 18.12 3.41 7 to 26
High socioeconomic status 0.33 0.47 0 to 1 0.24 0.43 0 to 1 0.29 0.45 0 to 1 0.21 0.41 0 to 1
Dependent variables
Violent 0.24 0.43 0 to 1 0.26 0.44 0 to 1 0.33 0.47 0 to 1 0.32 0.47 0 to 1
Property 0.32 0.47 0 to 1 0.44 0.50 0 to 1 0.31 0.46 0 to 1 0.38 0.48 0 to 1
Drug 0.06 0.24 0 to 1 0.02 0.14 0 to 1 0.12 0.33 0 to 1 0.08 0.27 0 to 1
Other 0.37 0.48 0 to 1 0.28 0.45 0 to 1 0.24 0.43 0 to 1 0.22 0.41 0 to 1
a. The number in parentheses denotes the number of police contacts within each trajectory and the average number of police contacts per offender in that trajectory,
onset). Demographically, the vast majority of the sample is male. There are actually no
females in the high-rate adolescent peaked trajectory, and only 4% of the high-rate chronic
trajectory is female. The majority of the sample is non-White and comes from a low SES
background. The mean ages range from approximately 15 to 20 years of age, with high-rate
adolescent peaked representing the youngest group (14.85) and low-rate chronics the oldest
(19.74) group.
Table 2 shows the odds of contact with the police for a particular offense type based on
the offense type of the prior contact, controlling for intermittency, early onset, age, SES,
race, and sex. Regardless of type of offender, the odds of contact with the police for a violent offense versus any other type of offense are more likely if an individual’s prior police
contact is for a violent offense. These odds are greater for the high-rate adolescent peaked
offenders and the high-rate chronic offenders than their low-rate offender counterparts.
High-rate chronic offenders have greater odds of repeat police contacts for violence (1.67)
than high-rate adolescent peaked (1.58) offenders. While these greater odds of specialization for the chronic life course offender may seem counterintuitive given prior literature
demonstrating the versatility of persistent offenders (Mazerolle et al., 2000; Nieuwbeerta
et al., 2011), these results are consistent with the findings exploring the odds of committing
a particular offense type following the commission of that same offense category (see
Bursik, 1980; Kempf, 1987; Wolfgang et al., 1972). Perhaps just as interesting is the finding that both high-rate adolescent peaked and high-rate chronic offenders are significantly
less likely to have a police contact for a property offense following a police contact for a
violent offense (0.76 and 0.70, respectively).
The results predicting property offending are similar to violent offending, with all
offender trajectories exhibiting specialization in offending. Again, high-rate chronics display
TABLE 2: The Odds of Being Contacted by the Police for Committing Offenses by Prior Offense and
Offender Trajectory Controlling for Intermittency, Sex, Race, Age of Onset, Age, and
Socioeconomic Status
Adolescent Peaked
Adolescent Peaked Low-Rate Chronic High-Rate Chronic
Violent → Violent 1.32** 1.58*** 1.54*** 1.67***
Violent → Property 0.87 0.76* 0.87 0.70***
Violent → Drug 0.89 0.65 0.70** 0.98
Violent → Other 0.98 0.95 0.96 0.86
Property → Property 1.25** 1.55*** 1.65*** 1.79***
Property → Violent 1.06 0.70** 0.80* 0.70***
Property → Drug 0.88 1.66 0.84 0.82
Property → Other 0.83* 0.75** 0.83* 0.77**
Drug → Drug 1.92* 0.40 2.31*** 2.11***
Drug → Violent 0.80 1.02 0.79 0.90
Drug → Property 1.09 0.70 0.76* 0.72*
Drug → Other 0.92 1.96 1.16 1.02
Other → Other 1.35*** 1.61*** 1.44*** 1.70***
Other → Violent 0.81* 0.79 1.00 0.90
Other → Property 0.88 0.94 0.93 0.68
Other → Drug 0.90 0.79* 0.75** 0.74***
*p < .05. **p < .01. ***p < .001.
the greatest odds of specialization in offending (1.79) and low-rate adolescent peaked have
the smallest odds (1.25). Police contacts for violent and other offenses are significantly less
likely following a police contact for a property offense among all four offender trajectories,
excluding property to violent for low-rate adolescent peaked offenders. Police contacts for
drug offending again provide evidence of specialized criminal careers with the highest odds
of repeat police contacts for all offender trajectories, excluding high-rate adolescent peaked
offenders. Both low-rate and high-rate chronic offenders display significant odds of repeat
police contacts for drugs greater than 2 (2.31 and 2.11, respectively), and low-rate adolescent peaked offenders display odds of repeating drug offenses just under 2 (1.92). Other
offenses closely parallel the findings of the violent, property, and drug offenses. High-rate
chronic offenders exhibit the greatest odds of repeat police contacts for other offenses
(1.70), with low-rate adolescent peaked offenders exhibiting the smallest odds (1.35).
Again, each trajectory shows significant findings supporting specialized criminal careers.
Excluding the relationship from a prior police contact for a drug offense to a subsequent
police contact for a drug offense among high-rate adolescent peaked offenders, all
offender trajectories display positive significant relationships supporting specialized
criminal careers. These findings alone are compelling, but just as compelling are the
negative significant relationships between police contacts for certain offenses and particular categories of unrelated offenses, or nonsignificant relationships among expectedly
linked crime categories.
Table 2 essentially shows a summary of figures for 64 detailed models, presenting only
the odds ratios for the effect of prior offense type on subsequent offense type, but while
intermittency, sex, race, age of onset, age, and SES are controlled. The effect of each on the
likelihood of being contacted by the police for committing the various types of offenses by
offender trajectory is not elucidated. Therefore, Tables 3 to 6 present these findings displaying the odds ratios for each exogenous variable. Based on the findings of Table 2, only the
tests of specialization are detailed further.
Table 3 presents the odds of contact by the police for committing a violent offense
among each offender trajectory. Race and age are significant predictors for all offender
TABLE 3: The Odds (OR) of Being Contacted by the Police for Committing a Violent Offense by Offender
Adolescent Peaked
Adolescent Peaked
Prior offense violent 1.32 0.14** 1.58 0.22*** 1.54 0.15*** 1.67 0.14***
Intermittency 0.999 0.003 0.96 0.01* 1.01 0.001*** 1.00 0.003
Male 1.26 0.18 1.56 0.23** 1.47 0.45
White 0.54 0.07*** 0.52 0.15* 0.73 0.08** 0.53 0.08***
Early onset 1.16 0.11 0.49 0.30 1.01 0.10 1.20 0.13
Age 1.13 0.03*** 1.24 0.03*** 1.11 0.01*** 1.11 0.01***
High socioeconomic status 0.88 0.10 0.60 0.16 0.88 0.09 0.93 0.13
Rho .10 .15 .10 .13
*p < .05. **p < .01. ***p < .001.
trajectories, with Blacks displaying higher odds of contact by the police for committing
violent offenses and higher odds of contact by the police for committing violent offenses
as age increases. Intermittency significantly predicts the odds of contact by the police for
committing a violent offense among high-rate adolescent peaked and low-rate chronic
offenders. However, the directions of the relationships are not the same. Intermittency has
a negative significant effect on the odds of a police contact for committing a violent offense
among high-rate adolescent peaked offenders (0.96), but a positive significant effect among
low-rate chronic offenders (1.01). Neither of these findings greatly differs from 1, so while
statistically significant, substantively these findings are not very compelling.
The odds of contact by the police for committing a property offense among each
offender trajectory are presented in Table 4. Unlike the consistency across offender groups
found among violent offenses, the odds of contact by the police for committing a property
offense differ among each trajectory. The odds that low-rate adolescent peaked offenders
are contacted by the police for committing a property offense are significantly higher for
males (1.32) and are lower as individuals age (0.95). High-rate adolescent peaked offenders
show similar odds for age (0.92), though sex is not measured for high-rate adolescent
peaked because no females from the sample fall into this offender trajectory. Higher SES
increases the odds of a property offense (1.52) for high-rate adolescent peaked offenders.
Black low-rate chronic offenders have higher odds of contact by the police for committing
a property offense, while White high-rate chronic offenders have higher odds of contact by
the police for committing a property offense.
Tables 5 and 6 show the odds of contact by the police for committing a drug offense
and the odds of contact by the police for committing other offenses, respectively, by
offender trajectory. Age is a positive and significant predictor of drug offenses for all
offender trajectories and a negative and significant predictor of other offenses for all four
trajectories. For low-rate adolescent peaked offenders, early onset increases the probability of contact by the police for committing a drug offense while it decreases the odds for
other offenses. Greater intermittency produce higher odds of contact by the police for
committing a drug offense among the low-rate adolescent peaked and low-rate
TABLE 4: The Odds (OR) of Being Contacted by the Police for Committing a Property Offense by
Offender Trajectory
Low-Rate Adolescent
High-Rate Adolescent
Prior offense property 1.25 0.11** 1.55 0.16*** 1.65 0.15*** 1.79 0.14***
Intermittency 0.997 0.003 1.01 0.01 0.995 0.001*** 0.999 0.003
Male 1.32 0.17* 0.80 0.11 1.42 0.45
White 1.09 0.11 0.86 0.17 0.77 0.08* 1.53 0.22**
Early onset 1.11 0.10 0.96 0.40 1.16 0.12 0.89 0.10
Age 0.95 0.02* 0.92 0.02*** 1.02 0.01 1.02 0.01
High socioeconomic status 0.88 0.09 1.52 0.29* 1.08 0.11 1.04 0.14
Rho .10 .07 .11 .15
*p < .05. **p < .01. ***p < .001.
chronic, though longer gaps in offending decrease the odds of contact by the police for
committing other offenses for low-rate chronic offenders.
The prior offense is the strongest nondemographic predictor of each offense type for all
offender trajectories, excluding high-rate adolescent peaked drug offenses. Including
demographics, the previous offense is the greatest predictor in 10 of the 16 models and the
second strongest predictor in 4 others. Contact by the police for committing the same
offense versus any other offense increases the chances of contact for that offense again by
as little as 25% for low-rate adolescent peaked offenders contacted for property offenses,
to a high of 131% for low-rate chronic offenders contacted by the police for a drug offense.
Together, these results present a strong case for the existence of specialization in offending
throughout the life course.
TABLE 5: The Odds (OR) of Being Contacted by the Police for Committing a Drug Offense by Offender
Prior offense drug 1.92 0.55* 0.40 0.34 2.31 0.41*** 2.11 0.37***
Intermittency 1.01 0.01* 0.98 0.01 1.01 0.002** 1.00 0.004
Male 0.85 0.21 2.12 0.51** 1.92 1.02
White 1.30 0.26 2.17 1.23 1.72 0.25*** 1.09 0.25
Early onset 1.83 0.32*** 0.00 0.00 1.21 0.17 1.19 0.21
Age 1.19 0.06*** 1.56 0.12*** 1.08 0.02*** 1.12 0.02***
High socioeconomic status 1.49 0.28* 0.79 0.46 0.81 0.12 .81 0.18
Rho .24 .28 .18 .23
*p < .05. **p < .01. ***p < .001.
TABLE 6: The Odds (OR) of Being Contacted by the Police for Committing an Other Offense by Offender
Prior offense other 1.35 0.11*** 1.61 0.19*** 1.44 0.15*** 1.70 0.16***
Intermittency 0.998 0.003 0.99 .01 0.995 0.002** 0.995 0.004
Male 0.69 0.08*** 0.53 0.08*** 0.38 0.01***
White 1.33 0.13** 1.75 0.28*** 1.49 0.17*** 1.12 0.15
Early onset 0.70 0.06*** 1.89 0.62 0.81 0.09 0.91 0.10
Age 0.91 0.02*** 0.87 0.02*** 0.81 0.01*** 0.78 0.01***
High socioeconomic status 1.12 0.10 1.00 0.16 1.24 0.14 1.17 0.15
Rho .06 .02 .10 .08
*p < .05. **p < .01. ***p < .001.
Using random logistic effects models disaggregated by trajectories, we were able to test
Moffitt’s (1993) theoretical propositions regarding specialization/versatility while remedying
some methodological limitations of prior research. The criticisms offered from both those
who argue for specialization and those who advocate versatility have been the lack of fully
specified models, the relatively few measures of specialization at the individual level, the
inability to account for multiple types of offenses, and the failure to consider the nature and
chronicity of the offenders (McGloin et al., 2007; Osgood & Schreck, 2007; Sullivan et al.,
2009). We sought to address these issues by predicting the odds of contact by the police for
committing a particular offense throughout the life course based on the offenses previously
resulting in a police contact while controlling for additional aspects of the criminal career
and demographic variables. To address the theoretical arguments presented by Moffitt, we
disaggregated the models by four distinct offender trajectories. Our findings lend support
to the recent literature demonstrating greater evidence of specialization in offending. We
find that the odds of contact by the police for a particular offense during the life course are
increased if the prior offense is of the same offense type across individuals, even when
controlling for other covariates. Additionally, we find no support for significantly increased
odds of switching between offense types.
Since it was our goal to test Moffitt (1993) and remedy some of the methodological
issues in the specialization/versatility literature, it is important to note the contribution of
our findings to both the theory and methods. The finding that a prior offense of the same
offense category as the subsequent offense is the strongest nondemographic predictor in 15
of the 16 elaborated models (see Table 2) provides evidence to challenge the idea that specialization is an artifact of opportunity (Gottfredson & Hirschi, 1990). The current analysis
provides support for stability in offending specialization, not versatility. The results demonstrate that the best predictor of the type of offense is an individual’s prior offense of the
same type, even when controlling for onset, age, productivity (offender trajectory), and
intermittency, all of which have been shown to significantly impact offending behavior
(Baker et al., in press; Laub & Sampson, 2001; Piquro et al., 2007; Wolfgang et al., 1972).
Moffitt (1993) posits versatility among LCPs while predicting specialization among
ALs. We find support for the latter prediction of specialization in the most limited offenders
(low-rate and high-rate adolescent peaked), but we also find specialization among the most
persistent offenders, which is contrary to Moffitt’s expectations. Rather than finding some
evidence of versatility between offense types for the most persistent offenders, we find
significantly reduced odds of switching between offense types in several categories. For
example, the odds of contact by the police for committing a property offense are significantly lower for high-rate chronic offenders whose prior police contact was for a violent or
a drug offense, and for low-rate chronic offenders, the odds of contact with the police for
a drug offense are significantly less likely if their prior contact was for either a violent or
other offense. Contrary to Moffitt’s predictions, it appears that offenders, despite their level
of chronicity, tend to offend within the same offense categories. We do acknowledge that
there may be versatility in the types of offenses committed within each of the delineated
categories—violent, property, drug, or other—but the findings suggest that offenders do
specialize within these broader categories. Someone who commits a violent act, whether it
be a homicide, robbery, or assault, is significantly more likely to commit another violent
act and even significantly less likely to commit other types of offenses, whether it is a
property, drug, or other offense. In this sense, the type of offense seems to matter more than
the type of offender, which is consistent with the findings of Loeber et al. (2008) and
Rosenfeld et al. (2012).
We also find similar inconsistencies in some of Moffitt’s (1993) other predictions concerning the relationship between age and offending behavior. While we should see an aging
out effect of all crime types for adolescent peaked offenders, increases in age only significantly reduce the odds of contact by the police for committing property and other offenses.
Alternatively, increases in age are positively and significantly related to the odds of contact
by the police for committing both violent and drug offenses. Other offenses are much more
representative of those offenses that would be committed as a result of the maturity gap, so
a negative relationship between age and other offenses is expected and supported by our
findings. However, greater odds of contact by the police for committing violent and drug
offenses as individuals get older is inconsistent with Moffitt’s predictions for the most
limited offenders. Not only should these offenders be aging out of crime, but they should
not have significant odds of engaging in violent crime. According to Moffitt, violent crime
is characteristic of LCPs, while ALs are expected to commit lesser forms of crime that
symbolize their adulthood, such as vandalism and public order offenses. Essentially, the
findings demonstrate that ALs and LCPs are not as different as Moffitt proposed in terms
of specialization/versatility and certain factors do not fit the profile of these two main types
of offenders. While some factors may distinguish these two groups, the tendency to specialize within these broad categories of offenses among chronic offenders and the increased
odds of contact by the police for committing violent and drug offenses at later ages for
adolescent peaked offenders does not.
Because we find specialization across different types of offenders, we fail to find support
for Moffitt’s (1993) contentions regarding the taxonomy. Taking this into consideration, we
can speculate as to whether other theories of offending may be better suited to explaining
specialized offending patterns. While strong findings of specialization seem to invalidate
concerns regarding opportunity, we are not able to specifically test this potential effect. The
tendency to commit an offense of the same type may be mediated by opportunity, a possibility suggested by both Gottfredson and Hirschi’s (1990) self-control theory and
Farrington’s (2005) ICAP theory. Thornberry’s (1987) interactional theory of delinquency
also focuses on the interaction between individuals and the social structure. While we control for SES of the individual, we do not account for other social-structural aspects that may
influence offending patterns and the tendency to specialize, such as social disorganization.
In other words, it is possible that the odds of being cited for the same offense depend on
additional factors, such as opportunity and social structure. Also following interactional
theory, the tendency to specialize may be contingent on association with delinquent peers
and the adoption of delinquent values. While we find evidence for specialization, the next
step would be to consider some of the factors mentioned and whether or not they influence
the tendency to specialize generally and to specialize in a particular type of offense.
Within the context of Moffitt’s (1993) theoretical framework, we also address some of the
methodological issues of prior research by applying random effects logistic regression. By
using random effects logistic regression models, we are able to account for the unbalanced
nature of most offense data, since individuals differ in the number of offenses committed
and the waves in which they commit these offenses (McGloin et al., 2007). These models
also include a parameter (rho) to account for unobserved or unmeasured heterogeneity that may
influence the likelihood of committing a violent, property, drug, or other offense. Since
random effects logistic regression is a variant of ordinary least squares regression analyses
designed to account for the nested structure of data, we are able to (a) use longitudinal data
(Sullivan et al., 2006), (b) measure specialization at the individual level, and (c) control for
both time-varying and time-invariant factors related to criminal careers, such as age, age of
onset, and intermittency, since “conditions in an individual’s life can fluctuate relatively
frequently” (Horney, Osgood, & Marshall, 1995, p. 658; Sullivan et al., 2006). Also, because
of the simple and clear interpretation of random effects models, we are able to consider
specialization/versatility among four different types of crime, as opposed to a single comparison of violent and nonviolent offenses (Osgood & Schreck, 2007).
In addition to these contributions, we also remedy issues with frequency of offending
and offending patterns by creating offender trajectories and disaggregating by these trajectories in our analyses. While these trajectories are mainly designed to account for the
propositions of Moffitt (1993), they do extend the methodological merit of the current
study. Since the trajectories are created based on police contact data, they approximate
frequency of offending among individuals. An offender can either be a low-rate adolescent
peaked, high-rate adolescent peaked, low-rate chronic, or high-rate chronic offender.
Additionally, disaggregating by each trajectory in the random effects models considers
offending patterns. Osgood and Schreck (2007) emphasize the focus of specialization
research on transitions between offenses, which often fails to capture the full pattern of
offending by an individual. By creating offender trajectories, we also indirectly address
what Osgood and Schreck refer to as the confounding between specialization and rate of
offending by considering the tendency to offend. After addressing some of the limitations
of previous methodologies, we find greater evidence of specialization, as has been the trend
among studies using methodologies that account for the potential problems of early specialization research.
While our findings are generally supportive and consistent with the recent specialization
literature, this study is not without limitations. While we acknowledge that our analyses
resolve several of the methodological issues identified by the specialization/versatility literature, the methodology used does not resolve all measurement issues and we are still
limited to the data we have available. For example, although we remedy the limitation of
including controls for other variables within models predicting specialization, we are still
unable to account for any psychological measures of the offenders, including self-control
(Gottfredson & Hirschi, 1990) and antisocial potential (Farrington, 2005). The 1958
Philadelphia Birth Cohort Study does not contain such measures for large portions of the
population, and so findings are limited to the extent that such variables would affect our
results. In addition, we are unable to account for the effect that punishment, or lack thereof,
may have on individuals’ choices of offenses. It is also possible that individual choices are
impacted by peers and the involvement of peers in the criminal justice system (Thornberry,
1987). These are factors to consider in future tests of specialization. Accounting for concepts
like self-control and the maturity gap may provide further information about offending
throughout the life course (Gottfredson & Hirschi, 1990; Moffitt, 1993).
Many of these limitations prevent us from being able to predict why individuals might
be specializing in each of the particular categories of offenses. For instance, we do not
know if violent specialization is a result of gang-related activity or if property specialization
is the result of economic need or simply thrill. Additionally, the fact that drug offending
had the highest odds of specialization may be due to biological factors such as addiction,
but not all drug offenses are necessarily drug use and could be drug sales as well. While
selling drugs may be a means for continued drug use, this is impossible to ascertain from
the data available. We are limited in that we can only establish if there is evidence of
specialization. Future research should explore the causal mechanisms behind why such
specialization occurs. Any attempts to do so as a result of this analysis would be pure
It is important to note that our results demonstrate specialization in offending among
transitions between police contacts, not as a description of the criminal career as a whole.
The findings are unlike a diversity index, in which an entire criminal career is examined
in aggregate to uncover a pattern of specialization. Rather, the findings in this analysis
demonstrate the propensity to specialize between any two offenses. As such, the purpose
of this study was not to label individuals as violent, property, drug, or other offenders, but
instead to examine the odds that an individual’s subsequent offense would be the same
(categorically) as their immediately prior offense. We were not attempting to describe
individuals’ careers but rather to identify the likelihood that an individual would experience a police contact for a property offense after a property offense or a violent offense
after a violent offense, and so on, throughout his or her life course. In addition, this
approach provided the opportunity to examine characteristics of individuals that could
have accounted for specialization/versatility in offending. This allowed us to move
beyond simply describing specialization/versatility in offending and actually trying to
predict and explain behavior among different types of offenders and offenses throughout
the life course.
Osgood and Schreck (2007) have also noted the need to separate specialization from
offense base rates. This means gauging specialization through the contrast of an individual’s offenses in certain offense categories to the overall rate of those offenses in the
population. For example, if half of all offenses in the data are property offenses, it is not
specialization if half of the offenses by any particular individual are property offenses,
which would be expected by chance alone (Osgood & Schreck, 2007). By the nature of
correcting for some of the other limitations, we are unable to address this base rate issue
using random logistic effects models. Because of this, it is unknown whether the lack of
switching across offense categories is likely due to the commonality of property and other
offenses in each of the four offender classifications, which could give an unfair advantage
to finding specialization. However, the intent of the current analyses is to determine
whether a subsequent offense is affected by the prior offense, whereby the subsequent
offense is likely to be the same as the prior offense. The fact that there is a greater likelihood of committing the same offense and a decreased probability of committing a different offense lends evidence toward specialization, regardless of the base rate, since we are
looking from one offense to the next throughout the life course, as opposed to specialization within a criminal career. Despite this, future research should consider incorporating
this aspect of specialization/versatility in their models.
Additionally, the study only followed individuals to age 26. The results show no indication that specialization varies with age or chronicity of offending, but it may be the case
that versatility increases past this point in the life course. The current study also uses police
contact information and is reliant on official data sources. These sources may be unreliable
as to the actual events that took place, the actual charges that were eventually levied against
the offenders, or eventual convictions or acquittals of the charges. With the use of official
data also comes the loss of additional crimes undoubtedly committed by each offender.
This limitation is especially vital given the use of trajectories to identify offenders. If the
full range and number of offenses is accounted for, it is certainly possible that offenders
would be differentially distributed, though our findings indicate specialization regardless
of trajectory. The use of life-event calendars for shorter more detailed periods of time may
be useful in this regard (cf. Horney et al., 1995; McGloin et al., 2007), as well as qualitative
analysis (cf. Laub & Sampson, 1993, 2003), that may be helpful in understanding the pathways and context that lead to specific types of offending among individuals. However,
current self-report data do not allow for analyses of a longer time span of the life course,
which is an added benefit of using official statistics. Additionally, even when using selfreport data, Loeber et al., (2008) found that 55% of reported and 41% of arrested violent
offenders in the youngest cohort were also theft offenders, and 54% of reported and 64%
of arrested violent offenders in the oldest cohort were also theft offenders. This means that
there were still a fairly large number of specialized offenders (45% reported and 59%
arrested in the youngest cohort and 46% reported and 36% arrested in the oldest cohort),
even when using self-report data.
Using official police contact data also increases the possibility that upon the commission
of a particular crime, the police are contacting the “usual suspects,” as opposed to individuals for whom there is evidence of a criminal offense. While this is true of all the aforementioned offenses, it may be particularly true of the other offense category. There is little
additional information about the miscellaneous classification of the other offense category,
and finding specialization in this category of offenses should be viewed with some trepidation. While we feel that this category broadly represents the analogous behaviors
Gottfredson and Hirschi (1990) describe, without further information about the offense
individuals were contacted by the police for and the ultimate disposition of the contact,
specialization within the “other” category may not accurately represent actual offending.
To the extent that this is true, some evidence of specialization may in fact be an artifact of
police practices, as opposed to actual criminal activity.
The results of this study can potentially inform future criminal justice policy. Given the
increased probability that an offender specializes in a particular type of offense, specialized
treatment programs or interventions may be beneficial in reducing recidivism. The findings
may also indicate where resources should be allocated. Anger management or drug treatment may not be as beneficial for a career property offender as it would be for a career
violent offender or drug offender, respectively. Similarly, for individuals who have criminal
histories marked by specialization in a particular offense type, voluntary completion of
certain programs may be a “signal” that the individual is reforming or desisting from crime
(Bushway & Apel, 2012). That is, if a career drug offender voluntarily seeks drug rehabilitation, a career violent offender seeks treatment for anger management, or a property
offender actively seeks out employment or educational training, this may be a signal for
probation officers, correctional officers, and other criminal justice practitioners that this
individual, given their criminal history, is reforming. Understanding individual specialization is important for elaborating on this signaling process. For example, voluntarily enlisting in drug rehabilitation programming for a career violent offender may not be a signal of
desistance from violent crime, though alternative benefits are no doubt possible. This area
of specialized treatment targeting is an important area for future research. To the extent that
desistance signals are taken into consideration, understanding an individual’s offending
propensity offers additional information about the seriousness and likelihood of desistance
given the signal (i.e., a career drug offender voluntarily seeking drug treatment as opposed
to a career violent offender voluntarily seeking drug treatment).
Ultimately, this study offers direction for future research in the area of specialization
by serving as a contribution to the theoretical and methodological debate surrounding
specialization/versatility. The findings have implications for both policy and a better
understanding of individuals with specialized criminal careers.
1. The dichotomy is based on a 10-item scale obtained from census tract data (see Kempf, 1983; Tracy, 1981).
Agresti, A., & Agresti, B. F. (1978). Statistical analysis of qualitative variation. Sociological Methodology, 10, 204-237.
Armstrong, T. A. (2008). Are trends in specialization across arrests explained by changes in specialization occurring with
age? Justice Quarterly, 25, 201-222.
Armstrong, T. A., & Britt, C. L. (2004). The effect of offender characteristics on offense specialization and escalation. Justice
Quarterly, 21, 843-876.
Baker, T., Falco Metcalfe, C. & Piquero, A. R. (in press). Measuring the intermittency of criminal careers. Crime and
Delinquency doi: 10.1177/0011128712466382.
Bernard, T. J. (1992). The cycle of juvenile justice. New York, NY: Oxford University Press.
Bernard, T. J. (2007). Serious delinquency: An anthology. Los Angeles, CA: Roxbury Publishing Company.
Bouffard, L. A., Wright, K. A., Muftic, L. R., & Bouffard, J. A. (2008). Gender differences in specialization in intimate
partner violence: Comparing the gender symmetry and violent resistance perspective. Justice Quarterly, 25, 570-594.
Brame, R., Paternoster, R., & Piquero, A. (2012). Thoughts on the analysis of group-based developmental trajectories in
criminology. Justice Quarterly, 29, 469-490.
Britt, C. L. (1996). The measurement of specialization and escalation in the criminal career: An alternative modeling strategy.
Journal of Quantitative Criminology, 12, 193-222.
Bursik, R. J. (1980). The dynamics of specialization in juvenile offenses. Social Forces, 58, 851-864.
Bushway, S., & Apel, R. (2012). A signaling perspective on employment-based reentry programming: Training completion
as a desistance signal. Criminology & Public Policy, 11, 21-50.
Cloward, R. A., & Ohlin, L. E. (1960). Delinquency and opportunity: A theory of delinquent gangs. New York, NY:
Free Press.
Cohen, J. (1986). Research on criminal careers: Individual frequency rates and offense seriousness. In A. Blumstein,
J. Cohen, J. A. Roth, & C. A. Visher (Eds.), Criminal careers and “career criminals” (pp. 292-418). Washington, DC:
National Academy Press.
D’Unger, A. V., Land, K. C., & McCall, P. L. (2002). Sex difference in age patterns of delinquent/criminal careers: Results
from Poisson latent class analyses of the Philadelphia cohort study. Journal of Quantitative Criminology, 18, 349-375.
D’Unger, A. V., Land, K. C., McCall, P. L., & Nagin, D. S. (1998). How many latent classes of delinquent/criminal careers?
Results from mixed Poisson regression analyses of the London, Philadelphia, and Racine Cohort Studies. American
Journal of Sociology, 103, 1593-1630.
Farrington, D. P. (1986). Age and crime. In M. Tonry & N. Morris (Eds.), Crime and justice: A review of research (Vol. 7,
pp. 189-250). Chicago, IL: University of Chicago Press.
Farrington, D. P. (2005). The integrated cognitive antisocial potential (ICAP) theory. In D. Farrington (Ed.), Integrated
developmental and life-course theories of crime (pp. 73-92). New Brunswick, NJ: Transaction Publishers.
Farrington, D. P., Snyder, H. N., & Finnegan, T. A. (1988). Specialization in juvenile court careers. Criminology, 26, 461-487.
Farrington, D. P., & Welsh, B. C. (2007). Saving children from a life of crime. Oxford, UK: Oxford University Press.
Ferguson, D. M., Horwood, L. J., & Nagin, D. S. (2000). Offending trajectories in a New Zealand birth cohort. Criminology,
38, 525-552.
Francis, B., Liu, J., & Soothill, K. (2010). Criminal lifestyle specialization: Female offending in England and Wales.
International Criminal Justice Review, 20, 188-204.
Francis, B., Soothill, K., & Fligelstone, R. (2004). Identifying patterns and pathways of offending behavior: A new approach
to typologies of crime. European Journal of Criminology, 1, 47-87.
Gottfredson, M., & Hirschi, T. (1990). A general theory of crime. Stanford, CA: Stanford University Press.
Gottfredson, M., & Hirschi, T. (1994). A general theory of adolescent problem behavior: Problems and prospects. In
R. Ketterlinus & M. Lamb (Eds.), Adolescent problem behaviors: Issues and research (pp. 41-56). Hillsdale, NJ:
Lawrence Erlbaum Associates, Inc.
Hindelang, M. J. (1971). Age, sex, and the versatility of delinquent involvements. Social Problems, 18, 522-535.
Hindelang, M. J., Hirschi, T., & Weis, J. (1981). Measuring delinquency. Beverly Hills, CA: Sage.
Horney, J., Osgood, D. W., & Marshall, I. H. (1995). Criminal careers in the short-term: Intra-individual variability in crime
and its relations to local life circumstances. American Sociological Review, 60, 655-673.
Jennings, W. G., & Reingle, J. (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.
Jones, B. L., & Nagin, D. S. (2007). Advances in group-based trajectory modeling and a SAS procedure for estimating them.
Sociological Methods and Research, 35, 542-572.
Kempf, K. L. (1983). Assessment of the relationship between socioeconomic status and delinquency using the 1958
Philadelphia birth cohort. Paper presented at the annual meeting of the American Society of Criminology, Denver, CO.
Kempf, K. L. (1987). Specialization and the criminal career. Criminology, 25, 399-420.
Klein, M. (1984). Offense specialization and versatility among juveniles. British Journal of Criminology, 24, 185-194.
Laub, J. H., & Sampson, R. J. (1993). Turning points in the life course: Why change matters to the study of crime.
Criminology, 31, 301-325.
Laub, J. H., & Sampson, R. J. (2001). Understanding desistance from crime. Crime and Justice, 28, 1-69.
Laub, J. H., & Sampson, R. J. (2003). Shared beginnings, divergent lives: Delinquent boys to age 70. Cambridge, MA:
Harvard University Press.
Loeber, R., Farrington, D. P., Stouthamer-Loeber, M., & Raskin White, H. (2008). Violence and serious theft: Development
and prediction from childhood to adulthood. New York, NY: Taylor & Francis Group, LLC.
Maughan, B. (2005). Developmental trajectory modeling: A view from developmental psychopathology. The ANNALS of the
American Academy of Political and Social Science, 602, 118-130.
Mazerolle, P., Brame, R., Paternoster, R., Piquero, A., & Dean, C. (2000). Onset age, persistence, and offending versatility:
Comparisons across gender. Criminology, 38, 1143-1172.
McGloin, J. M., Sullivan, C. J., & Piquero, A. R. (2009). Aggregating to versatility? Transitions among offender types in the
short term. British Journal of Criminology, 49, 243-264.
McGloin, J. M., Sullivan, C. J., Piquero, A. R., & Pratt, T. C. (2007). Local life circumstances and offending specialization/
versatility: Comparing opportunity and propensity models. Journal of Research in Crime and Delinquency, 44, 321-346.
Moffitt, T. E. (1993). Adolescence-limited and life-course-persistent antisocial behavior: A developmental taxonomy.
Psychological Review, 100, 674-701.
Moffitt, T. E., Caspi, A., Dickson, N., Silva, P., & Stanton, W. (1996). Childhood-onset versus adolescent-onset antisocial
conduct problems in males: Natural history from age 3 to 18 years. Developmental Psychopathology, 8, 399-424.
Moffitt, T. E., Caspi, A., Harrington, H., & Milne, B. J. (2002). Males on the life-course-persistent and adolescence-limited
antisocial pathways: Follow-up at age 26 years. Development and Psychopathology, 14, 179-207.
Nagin, D. S. (2005). Group-based modeling of development. Cambridge, MA: Harvard University Press.
Nagin, D. S., & Tremblay, R. E. (2005a). Developmental trajectory groups: Fact or a useful statistical fiction. Criminology,
43, 873-904.
Nagin, D. S., & Tremblay, R. E. (2005b). From seduction to passion: A response to Sampson and Laub. Criminology, 43,
Nagin, D. S., & Tremblay, R. E. (2005c). What has been learned from group-based trajectory modeling? Examples from
physical aggression and other problem behaviors. The ANNALS of the American Academy of Political and Social Science,
602, 82-117.
Nagin, D. S. & Tremblay, R. E. (2005d). Analyzing developmental trajectories: A response to Maughan and Raudenbush. The
ANNALS of the American Academy of Political and Social Science, 602, 145-154.
Nagin, D. S., Farrington, D. P., & Moffitt, T. E. (1995). Life-course trajectories of different types of offenders. Criminology,
33, 111-139.
Nieuwbeerta, P., Blokland, A. A. J., Piquero, A. R., & Sweeten, G. (2011). A life-course analysis of offense specialization across
age: Introducing a new method for studying individual specialization over the life course. Crime and Delinquency, 57, 3-28.
Osgood, D. W., & Schreck, C. J. (2007). A new method for studying the extent, stability, and predictors of individual specialization in violence. Criminology, 45, 273-312.
Paternoster, R., Brame, R., Piquero, A., Mazerolle, P., & Dean, C. W. (1998). The forward specialization coefficient:
Distributional properties and subgroup differences. Journal of Quantitative Criminology, 14, 133-154.
Patterson, G. R., Crosby, L., & Vuchinich, S. (1992). Predicting risk for early police arrest. Journal of Quantitative
Criminology, 8, 335-355.
Piquero, A. (2004). Somewhere between persistence and desistance: The intermittency of criminal careers. In S. Maruna
& R. Immarigeon (Eds.), After crime and punishment: Pathways to offender reintegration (pp. 102-125). Cullompton,
UK: Willan.
Piquero, A. R., Farrington, D. P., & Blumstein, A. (2007). Key issues in criminal career research: New analyses of the
Cambridge study in delinquent development. Cambridge, MA: Cambridge University Press.
Piquero, A., Paternoster, R., Mazerolle, P., Brame, R., & Dean, C. W. (1999). Onset age and offense specialization. Journal
of Research in Crime and Delinquency, 36, 275-299.
Petersilia, J. (1980). Criminal career research: A review of recent evidence. In N. Morris & M. Tonry (Eds.), Crime and
justice: A review of research (Vol. 2, pp. 321-379). Chicago, IL: University of Chicago Press.
Raudenbush, S. W. (2005). How do we study “what happens next”? The ANNALS of the American Academy of Political and
Social Science, 602, 131-144.
Rojek, D., & Erickson, M. (1982). Delinquent careers. Criminology, 20, 5-28.
Rosenfeld, R., White, H., & Esbensen, F. (2012). Special categories of serious and violent offenders: Drug dealers, gang
members, homicide offenders, and sex offenders. In R. Loeber & D. Farrington (Eds.), From juvenile delinquency to adult
crime: Criminal careers, justice policy, and prevention (pp. 118-149). New York, NY: Oxford University Press.
Sampson, R. J., & Laub, J. H. (2005a). A life-course view of the development of crime. The ANNALS of the American
Academy of Political and Social Science, 602, 12-45.
Sampson, R. J., & Laub, J. H. (2005b). Seductions of method: Rejoinder to Nagin and Tremblay’s “Developmental Trajectory
Groups: Fact or Fiction.” Criminology, 43, 905-914.
Sampson, R. J., & Laub, J. H. (2005c). When prediction fails: From crime-prone boys to heterogeneity in adulthood. The
ANNALS of the American Academy of Political and Social Science, 602, 73-79.
Skardhamar, T. (2010). Distinguishing facts and artifacts in group-based modeling. Criminology, 48(1), 295-320.
Spelman, W. (1994). Criminal incapacitation. New York, NY: Plenum.
Sullivan, C. J., McGloin, J. M., Pratt, T. C., & Piquero, A. R. (2006). Rethinking the “norm” of offender generality:
Investigating specialization in the short-term. Criminology, 44, 199-233.
Sullivan, C. J., McGloin, J. M., Ray, J. V., & Caudy, M. S. (2009). Detecting specialization in offending: Comparing analytic
approaches. Journal of Quantitative Criminology, 25, 419-441.
Tanenhaus, D. S. (2004). Juvenile justice in the making. Oxford, UK: Oxford University Press.
Thornberry, T. (1987). Toward an interactional theory of delinquency. Criminology, 25, 863-892.
Tibbetts, S. G., & Piquero, A. (1999). The influence of gender, low birth weight, and disadvantaged environment in predicting
early onset of offending: A test of Moffitt’s interactional hypothesis. Criminology, 37, 843-878.
Tracy, P. (1981). Ecology and delinquency: The development of a composite measure of social class. Unpublished manuscript.
Tracy, P. E., & Kempf-Leonard, K. (1996). Continuity and discontinuity in criminal careers. New York, NY: Plenum Press.
Tracy, P. E., Wolfgang, M. E., & Figlio, R. M. (1990). Delinquency in two birth cohorts. New York, NY: Plenum Press.
Wolfgang, M. E., Figlio, R. M., & Sellin, T. (1972). Delinquency in a birth cohort. Chicago, IL: University of Chicago Press.
Zimring, F. E. (2005). American juvenile justice. Oxford, UK: Oxford University Press.
Thomas Baker, PhD, is an assistant professor in the L. Douglas Wilder School of Government and Public Affairs at Virginia
Commonwealth University. His research interests include procedural justice, policing, courts, criminological theory, life
course criminology, and quantitative research methods. His work has appeared in various outlets including Journal of
Criminal Justice, Crime & Delinquency, and Social Science Quarterly.
Christi Falco Metcalfe is a PhD candidate in the College of Criminology and Criminal Justice at Florida State University.
Her research interests include organizational components of criminal justice systems and its effect on criminal justice outcomes,
developmental/life course criminology, and institutions and crime.
Wesley G. Jennings, PhD, is an assistant professor in the College of Behavioral and Community Sciences in the Department
of Criminology and has a Courtesy assistant professor appointment in the Department of Mental Health Law and Policy at
the University of South Florida. He received his doctorate degree in criminology from the University of Florida in 2007.

Get Professional Assignment Help Cheaply

Buy Custom Essay

Are you busy and do not have time to handle your assignment? Are you scared that your paper will not make the grade? Do you have responsibilities that may hinder you from turning in your assignment on time? Are you tired and can barely handle your assignment? Are your grades inconsistent?

Whichever your reason is, it is valid! You can get professional academic help from our service at affordable rates. We have a team of professional academic writers who can handle all your assignments.

Why Choose Our Academic Writing Service?

  • Plagiarism free papers
  • Timely delivery
  • Any deadline
  • Skilled, Experienced Native English Writers
  • Subject-relevant academic writer
  • Adherence to paper instructions
  • Ability to tackle bulk assignments
  • Reasonable prices
  • 24/7 Customer Support
  • Get superb grades consistently

Online Academic Help With Different Subjects


Students barely have time to read. We got you! Have your literature essay or book review written without having the hassle of reading the book. You can get your literature paper custom-written for you by our literature specialists.


Do you struggle with finance? No need to torture yourself if finance is not your cup of tea. You can order your finance paper from our academic writing service and get 100% original work from competent finance experts.

Computer science

Computer science is a tough subject. Fortunately, our computer science experts are up to the match. No need to stress and have sleepless nights. Our academic writers will tackle all your computer science assignments and deliver them on time. Let us handle all your python, java, ruby, JavaScript, php , C+ assignments!


While psychology may be an interesting subject, you may lack sufficient time to handle your assignments. Don’t despair; by using our academic writing service, you can be assured of perfect grades. Moreover, your grades will be consistent.


Engineering is quite a demanding subject. Students face a lot of pressure and barely have enough time to do what they love to do. Our academic writing service got you covered! Our engineering specialists follow the paper instructions and ensure timely delivery of the paper.


In the nursing course, you may have difficulties with literature reviews, annotated bibliographies, critical essays, and other assignments. Our nursing assignment writers will offer you professional nursing paper help at low prices.


Truth be told, sociology papers can be quite exhausting. Our academic writing service relieves you of fatigue, pressure, and stress. You can relax and have peace of mind as our academic writers handle your sociology assignment.


We take pride in having some of the best business writers in the industry. Our business writers have a lot of experience in the field. They are reliable, and you can be assured of a high-grade paper. They are able to handle business papers of any subject, length, deadline, and difficulty!


We boast of having some of the most experienced statistics experts in the industry. Our statistics experts have diverse skills, expertise, and knowledge to handle any kind of assignment. They have access to all kinds of software to get your assignment done.


Writing a law essay may prove to be an insurmountable obstacle, especially when you need to know the peculiarities of the legislative framework. Take advantage of our top-notch law specialists and get superb grades and 100% satisfaction.

What discipline/subjects do you deal in?

We have highlighted some of the most popular subjects we handle above. Those are just a tip of the iceberg. We deal in all academic disciplines since our writers are as diverse. They have been drawn from across all disciplines, and orders are assigned to those writers believed to be the best in the field. In a nutshell, there is no task we cannot handle; all you need to do is place your order with us. As long as your instructions are clear, just trust we shall deliver irrespective of the discipline.

Are your writers competent enough to handle my paper?

Our essay writers are graduates with bachelor's, masters, Ph.D., and doctorate degrees in various subjects. The minimum requirement to be an essay writer with our essay writing service is to have a college degree. All our academic writers have a minimum of two years of academic writing. We have a stringent recruitment process to ensure that we get only the most competent essay writers in the industry. We also ensure that the writers are handsomely compensated for their value. The majority of our writers are native English speakers. As such, the fluency of language and grammar is impeccable.

What if I don’t like the paper?

There is a very low likelihood that you won’t like the paper.

Reasons being:

  • When assigning your order, we match the paper’s discipline with the writer’s field/specialization. Since all our writers are graduates, we match the paper’s subject with the field the writer studied. For instance, if it’s a nursing paper, only a nursing graduate and writer will handle it. Furthermore, all our writers have academic writing experience and top-notch research skills.
  • We have a quality assurance that reviews the paper before it gets to you. As such, we ensure that you get a paper that meets the required standard and will most definitely make the grade.

In the event that you don’t like your paper:

  • The writer will revise the paper up to your pleasing. You have unlimited revisions. You simply need to highlight what specifically you don’t like about the paper, and the writer will make the amendments. The paper will be revised until you are satisfied. Revisions are free of charge
  • We will have a different writer write the paper from scratch.
  • Last resort, if the above does not work, we will refund your money.

Will the professor find out I didn’t write the paper myself?

Not at all. All papers are written from scratch. There is no way your tutor or instructor will realize that you did not write the paper yourself. In fact, we recommend using our assignment help services for consistent results.

What if the paper is plagiarized?

We check all papers for plagiarism before we submit them. We use powerful plagiarism checking software such as SafeAssign, LopesWrite, and Turnitin. We also upload the plagiarism report so that you can review it. We understand that plagiarism is academic suicide. We would not take the risk of submitting plagiarized work and jeopardize your academic journey. Furthermore, we do not sell or use prewritten papers, and each paper is written from scratch.

When will I get my paper?

You determine when you get the paper by setting the deadline when placing the order. All papers are delivered within the deadline. We are well aware that we operate in a time-sensitive industry. As such, we have laid out strategies to ensure that the client receives the paper on time and they never miss the deadline. We understand that papers that are submitted late have some points deducted. We do not want you to miss any points due to late submission. We work on beating deadlines by huge margins in order to ensure that you have ample time to review the paper before you submit it.

Will anyone find out that I used your services?

We have a privacy and confidentiality policy that guides our work. We NEVER share any customer information with third parties. Noone will ever know that you used our assignment help services. It’s only between you and us. We are bound by our policies to protect the customer’s identity and information. All your information, such as your names, phone number, email, order information, and so on, are protected. We have robust security systems that ensure that your data is protected. Hacking our systems is close to impossible, and it has never happened.

How our Assignment Help Service Works

1. Place an order

You fill all the paper instructions in the order form. Make sure you include all the helpful materials so that our academic writers can deliver the perfect paper. It will also help to eliminate unnecessary revisions.

2. Pay for the order

Proceed to pay for the paper so that it can be assigned to one of our expert academic writers. The paper subject is matched with the writer’s area of specialization.

3. Track the progress

You communicate with the writer and know about the progress of the paper. The client can ask the writer for drafts of the paper. The client can upload extra material and include additional instructions from the lecturer. Receive a paper.

4. Download the paper

The paper is sent to your email and uploaded to your personal account. You also get a plagiarism report attached to your paper.

smile and order essay GET A PERFECT SCORE!!! smile and order essay Buy Custom Essay

Place your order
(550 words)

Approximate price: $22

Calculate the price of your order

550 words
We'll send you the first draft for approval by September 11, 2018 at 10:52 AM
Total price:
The price is based on these factors:
Academic level
Number of pages
Basic features
  • Free title page and bibliography
  • Unlimited revisions
  • Plagiarism-free guarantee
  • Money-back guarantee
  • 24/7 support
On-demand options
  • Writer’s samples
  • Part-by-part delivery
  • Overnight delivery
  • Copies of used sources
  • Expert Proofreading
Paper format
  • 275 words per page
  • 12 pt Arial/Times New Roman
  • Double line spacing
  • Any citation style (APA, MLA, Chicago/Turabian, Harvard)

Our guarantees

Delivering a high-quality product at a reasonable price is not enough anymore.
That’s why we have developed 5 beneficial guarantees that will make your experience with our service enjoyable, easy, and safe.

Money-back guarantee

You have to be 100% sure of the quality of your product to give a money-back guarantee. This describes us perfectly. Make sure that this guarantee is totally transparent.

Read more

Zero-plagiarism guarantee

Each paper is composed from scratch, according to your instructions. It is then checked by our plagiarism-detection software. There is no gap where plagiarism could squeeze in.

Read more

Free-revision policy

Thanks to our free revisions, there is no way for you to be unsatisfied. We will work on your paper until you are completely happy with the result.

Read more

Privacy policy

Your email is safe, as we store it according to international data protection rules. Your bank details are secure, as we use only reliable payment systems.

Read more

Fair-cooperation guarantee

By sending us your money, you buy the service we provide. Check out our terms and conditions if you prefer business talks to be laid out in official language.

Read more
error: Content is protected !!
Open chat
Need assignment help? You can contact our live agent via WhatsApp using +1 718 717 2861

Feel free to ask questions, clarifications, or discounts available when placing an order.
  +1 718 717 2861           + 44 161 818 7126           [email protected]
  +1 718 717 2861         [email protected]