Beyond sexual recidivism: A review of the sexual criminal career parameters of adult sex offenders

Highlights

• Sexual recidivism is an incomplete measure of sex offending.

• A criminal career framework is proposed to study sex offending.

• Several key criminal career parameters are defined and reviewed.

• Much heterogeneity is found in the criminal career of sex offenders.

• Current evidence suggests that most adult sex offending careers are short-lived.

Abstract

The current study claims that measures of sexual recidivism provide a distorted view of the criminal activity of adult sex offenders. To address this important limitation, the criminal career perspective is presented and key concepts are defined and described. The study also provides an up-to-date review of the scientific literature on various criminal career parameters of the sexual criminal activity of adult sex offenders. Hence, current empirical knowledge on the prevalence, age of onset, frequency, continuity, versatility, and desistance from sex offending is presented. The findings highlight the complexities of the sexual criminal career of adult sex offenders, and most importantly, its dynamic aspect, both of which are not captured by traditional measures of sexual recidivism. The review also underscores the importance of recognizing that sexual offending develops according to a series of stages, that, if not recognized, may lead to the underestimation of risk for some and over-estimation of risk for others. The review provides a framework to stimulate new areas of research as well as policy-development that is not limited to the identification of the “high-risk” convicted sex offenders.

Introduction

Interest in the criminal career is far from new and several commentaries and observations about sex offenders' criminal activity have in fact been made for quite some time (e.g., Adler, 1984, Amir, 1971, Radzinowicz, 1957). Most of these commentaries and observations were focused on the same underlying issue, that is, sex offenders' dangerousness. The issue of dangerousness has been addressed by examining sex offenders' likelihood of sexual recidivism using different methodologies. However, early descriptive studies of sex offenders' criminal records were not supported by an organizing conceptual framework which led to the emergence of controversies among researchers about the nature and extent of sex offenders' criminal behavior (e.g., Abel and Rouleau, 1990, Langevin et al., 2004, Marshall et al., 1991, Webster et al., 2006). These controversies certainly did not help to challenge common myths, stereotypes, and false beliefs about sex offenders' criminal behavior which have, in some instances, served as the foundation to develop new criminal justice policies to tackle the problem of sexual violence and abuse (e.g., La Fond, 2005, Quinn et al., 2004). The current review re-introduces the criminal career approach and, in doing so, aims to provide a common organizing framework for policymakers as well as researchers from various disciplines such as psychology, psychiatry, sociology, social work, criminal justice, and criminology. Although there is a long history of criminal career research (e.g., Wolfgang, Figlio, & Sellin, 1972), the seminal publication of “Criminal Careers and Career Criminals” by Blumstein et al. (1986) provided an organizing conceptual framework that has since been adopted primarily in the fields of criminal justice and criminology. This conceptual framework is mainly concerned with the description and explanation of the longitudinal sequence of offending over the life-course. We refer readers to several reviews of the state of empirical knowledge on criminal careers in the general population of offenders (e.g., Delisi and Piquero, 2011, Piquero et al., 2003). Nonetheless, while the criminal career approach has been prominent in criminological circles, it would take some time before this conceptual framework would be introduced more explicitly to the field of sexual violence and abuse (Blokland and Lussier, 2013, Lussier et al., 2005).

Building on the criminal career approach first proposed by Blumstein et al., and the pioneer work of Gebhard, Gagnon, Pomeroy, and Christenson (1965) with sex offenders, this review examines the current state of knowledge regarding the criminal activity of sex offenders. While the criminal career approach should not been seen as a panacea to the problem of sexual violence and abuse, the general aim is to provide a conceptual framework to organize the findings of existing studies, to guide future empirical research, and importantly, to help more clearly conceptualize sex offenders' sexual criminal behavior. More specifically, the three key goals of this review are to: 1) introduce researchers and practitioners in the field of sexual violence and abuse to the criminal career approach; 2) organize the empirical knowledge about the sexual criminal activity of sex offenders using the criminal career approach; and, 3) review the state of empirical knowledge on various dimensions of the sexual criminal careers of sex offenders.

Sexual recidivism has been conceptualized as the key outcome measure to describe sex offenders' criminal behavior for quite some time. There is a long tradition of empirical research on sexual recidivism among sex offenders, and this measure of reoffending has been applied in a variety of contexts examining the impact of penal and legal measures, investigating the efficacy of sex offender treatment programs, and creating risk assessment protocols, to name a few. Typically, studies along these lines have been conducted prospectively and longitudinally examining offenders' criminal behavior in a given period of time following release. These studies have examined the proportion of sex offenders having committed a new sex crime during a given follow-up period, using indicators such as a new arrest or a new conviction. More and more sophisticated techniques have also been gradually introduced to examine the rate of re-offending such as survival methods (e.g., Soothill & Gibbens, 1978) and actuarial prediction models (Quinsey, Rice, & Harris, 1995). However, while empirical techniques have evolved, the outcome measure has remained the same; a new arrest or a new conviction for a sex crime, while controlling for time at risk of reoffending. The innovation of using survival techniques to account for right-censoring of data was important as research was suggesting that with longer follow-up periods, more sex offenders were arrested (or convicted) again for a new sex crime. However, these results were often interpreted as indicating that sex offenders remained at-risk of sexual offending for long time periods. While a considerable amount of knowledge has been accumulated from these recidivism studies, some of the most fundamental questions about the criminal behavior of sex offenders have remained unanswered. To illustrate, Hanson and Morton-Bourgon (2005) reported that as of January, 2003, there were more than 100 recidivism studies on sex offenders' recidivism. This is more a decade ago and the number of studies has grown considerably since. Comparatively speaking, a recent review of the onset of sex offending in adult offenders by Lussier and Mathesius (2012) included less than 15 empirical studies on the issue. In effect, concerns over sex offenders' dangerousness and related policy responses to it have overshadowed the importance of research examining key etiological issues related to the onset and development of their criminal behavior. For example, the vast amount of research findings stemming from sexual recidivism studies do not provide insight to several critical questions such as:

• At what age does sex offending typically start?

• Does persistence in sexual offending vary according to patterns of onset?

• Does the frequency and rate of sex offending remain stable over time and across arrests?

• Is the frequency and rate of sexual offending stable across different stages of the criminal career?

• Are some sex offenders more prolific than others in terms of their volume of offending (e.g., number of victims and number of sex crime events)?

• Are some sex offenders more skilled at avoiding detection?

• Are sex offenders committing the same type of crime and offending against the same victim type across offenses?

• Is there an increase or decrease in the seriousness (e.g., level of intrusiveness and level of violence) of sex offending over time?

• At what age do sex offenders typically desist?

• Is desistance a process that takes place over time or it is best understood as an event?

• Do sex offenders become more specialized or diversified in the way they operate and commit their offense over time?

• Are there distinct patterns of sex offending and how many offending trajectories best describe the criminal behavior of sex offenders' over time?

In effect, the virtually exclusive focus on the likelihood of sexual recidivism has substantially diverted researchers' attention away from the basic description and understanding of the development of sex offenders' criminal activity over time (see Lussier & Davies, 2011).

Sexual recidivism as a criminal career indicator provides a limited description of the nature and extent of offending (e.g., Furby, Weinrott, & Blackshaw, 1989). In addition, because observed sexual recidivism rates are generally low, it is common practice among researchers to collapse all forms of sex crimes under an umbrella term — i.e., sexual recidivism. If sexual recidivism refers to a new charge or a new conviction for a new sex crime, it does not provide information about several key aspects of offending such as: the type of sex crime committed; the characteristics of victim; the number of times the offender offended against each victim(s); the level of violence used in each crime event; the type of sexual behaviors manifested; the time period elapsed between the first and the last event; whether the offending behavior changed between the first and the last event; and, whether the offender had desisted from sex offending before being apprehended, for example. In other words, the gap between unknown offending patterns and “a new charge” for a sex crime leaves us with minimal information about the sexual criminal activity of sex offenders. When considering some of these ‘unknowns’ and their relevance, it is unclear how risk assessors using research on sexual recidivism can make accurate predictions about the future offending behavior of an individual. Crucial information is lost when relying only on official indicators of offending, such as a charge or a conviction, and this may substantially distort the current state of knowledge pertaining to sex offenders. This point was illustrated in an empirical study by Lussier, Bouchard, and Beauregard (2011). One conviction for a sex crime may include individuals having committed a sex crime against one victim on a single occasion, but it may also include individuals having offended against the same victim on several hundreds of occasions, or, it may include individuals having offended against multiple victims on multiple occasions. These researchers demonstrated that the most active sex offenders, defined as those with the highest number of victims and the highest number of sexual crime events showed the lowest detection rate. Their ability to avoid detection for much longer periods (up to 40 years) was observed by their greater likelihood of not having a criminal history for a sex crime. In fact, the most successful sex offenders were also showing a lower-risk profile for sexual recidivism on static scales that rely on official criminal career indicators. These findings raise concerns about who is described as the sexual recidivist and the “dangerous offender” using traditional indicators out of the context of the over-arching criminal career approach.

For criminal career researchers, the importance of analyzing the longitudinal sequence of offending stems from the recognition that an offending career is best conceptualized as a series of stages. More specifically, criminal career researchers have argued that the longitudinal sequence of offending includes a beginning (i.e., onset), a quantitative and qualitative course of development (i.e., offending rate, volume of offending, aggravation, crime-switching, specialization), and a desistance stage (i.e., the slowing down of criminal activity and eventual termination of offending,). Breaking down offending into these different stages helps to better describe, understand, explain, and contextualize offenders' criminal activity at a particular point in time. To illustrate, criminal career researchers have raised the possibility that the factors responsible for someone initiating criminal activities might be distinct from those explaining its persistence or escalation. This approach stands in stark contrast from the recidivism perspective that is only concerned with the probability of recidivism, or the proportion of offenders who will reoffend after their prison release. By design, the recidivism approach adopts a static perspective to the criminal activity of offenders. This static approach suggests that the likelihood of offending remains the same irrespective of whether the offender is starting their criminal activity, whether their criminal activity is escalating, or whether it is slowing down over time.

Many clinicians and clinical researchers have struggled with the idea that offender's risk is static over time. This is apparent by the introduction of dynamic risk factors or factors associated with changes in the risk of reoffending. It is also apparent by the plethora of recent studies on the role of age, aging and sexual recidivism in the past decade (e.g., Barbaree et al., 2007, Doren, 2006, Lussier and Healey, 2009). These two lines of work, however, have not been integrated into a conceptual model of sex offending. In fact, while these lines of work provide information about possible correlates of changes in sex offending over time, they provide little information about the possible age-graded dynamic factors as well as the dynamic aspect of sex offending over time. For example, are changes in intimacy deficits or sexual regulation associated with changes in the offending behavior of sex offenders? Are these two set of factors important for 20 year olds sex offenders as well as for 50 year olds? Put differently, research on dynamic risk factors for sexual recidivism do not provide information pertaining to patterns of activation, course, and desistance from sex offending. If offenders do change over time and offending goes through stages over time, then the study of within-individual changes becomes pivotal to understanding sex offenders and predicting sexual crime. Recidivism studies have combined offenders into samples who are at different stages of their offending, that is, some offenders who potentially just began their offending, some who are at the peak of their offending, and even others who may be at the end of their criminal careers. By combining offenders at different stages of their criminal career, it is unclear whether the base rate of sexual recidivism applies to first-time offenders, escalators, de-escalators, or sex crime specialists. The aggregate base rate of sexual recidivism has been seriously criticized for not taking into account the offenders' age at their release from prison (e.g., Wollert, 2006). From a criminal career perspective, this limitation is only the tip of the iceberg. The iceberg is the sexual criminal career of sex offenders.