Sex Offender Recidivism: Some Lessons Learned From Over 70 Years of Research

• Published date: 01 December 2024
• DOI: http://doi.org/10.1177/07340168231157385
• Author: Patrick Lussier,Stéphanie Chouinard Thivierge,Julien Fréchette,Jean Proulx
• Date: 01 December 2024

Sex Offender Recidivism: Some

Lessons Learned From Over 70

Years of Research

Patrick Lussier

1,2

,

Stéphanie Chouinard Thivierge

1

,

Julien Fréchette

1

, and Jean Proulx

1,3

Abstract

Sex offender recidivism (SOR) has beenthe subject of research for over 70 years. Myths,miscon-

ceptions, and erroneous conclusions about SOR, however, remain widespread, impeding the devel-

opment of evidence-based policies aimed at preventing sexual offenses. To address the rich but

uneven literature, a comprehensive review was conducted making it possible to provide a contex-

tualized overview of scientific knowledge against the backdrop of methodological issues, challenges,

and shortcomings. Over the years, researchers have been asked to provide a simple answer to a

seemingly simple question: what are the recidivism rates for sexual offending? In response, the

field has produced a widerangeof findings making it difficult to draw firm conclusions, leaving

room for interpretation and personal biases. The variations in recidivism rates are attributable to

offender and methodological characteristics, both of which are embedded in a particular sociolegal

context. As a result, the base rate of SOR is more effectively considered in terms of a series of

questions that should include the type of recidivism, with whom, over what period, and in what

context. Issues and debates that have marked the field and fueled its growth are highlighted.

Research innovations and important areas of research are also discussed.

Keywords

sex crimes, corrections, crime policy, evaluation research, quantitative methods, sexual recidivism

Every generation seems to define particular sets of behaviors as social problems (e.g., Blumer, 1971)

and identify the groups involved as folk devils (Cohen, 1972). In North America, since the 1940s,

these behaviors and their perpetrators have included sexual offenses and sexual offenders, defined

in ways that led to three waves of sociolegal responses: the clinical-medical model, the sociolegal-

feminist model, and the community risk-protection model (e.g., Lieb et al., 1998; Petrunik, 2002).

1

School of Social Work and Criminology, Université Laval, Quebec, Canada

2

Centre International de Criminologie Comparée, Quebec, Canada

3

School of Criminology, Université de Montréal, Montreal, Quebec, Canada

Corresponding Author:

Patrick Lussier;

Email: patrick.lussier@tsc.ulaval.ca

Article

Criminal Justice Review

2024, Vol. 49(4) 413-452

© 2023 Georgia State University

Article reuse guidelines:

sagepub.com/journals-permissions

DOI: 10.1177/07340168231157385

journals.sagepub.com/home/cjr

We have now entered a fourth wave, sparked in part by the #MeToo movement and its focus on

certain sexual offenses (e.g., sexual misconduct, sexual harassment, sexual trafficking) by individu-

als in positions of power (e.g., Lussier et al., 2021b). Sexual offending phenomena include a broad

spectrum of behaviors but at any given time only some are defined as deserving of government atten-

tion and intervention. In the 1990s, following a series of rare but brutal events, the existence of “the

sexual recidivist”began to be seen as an urgent social problem in need of immediate attention and

intervention by policymakers (e.g., Simon, 1998) and laws tailored toward prevention of sexual

offending by individuals previously convicted of a sex crime became part of the agenda of many

criminal justice agencies. This attention led to legal changes and policy developments in the criminal

justice system that affected individuals labeled as “sex offenders.”Research in this area has increased

over the years, and researchers are now closer to answering questions raised some 30 years ago by

policymakers and stakeholders, prosecutors and lawyers, psychologists and psychiatrists, victims,

victims’rights advocates, and journalists. However, despite current social movements and media

attention to certain sexual offenses, the question of how to define and deal with “sexual recidivists”

is still at issue.

In North America, while the 1990s was a pivotal point for the development of policies and

measures aimed at managing the risk of recidivism by sex offenders, the issue had caught

researchers’attention several decades earlier (e.g., Tappan, 1950). Research in this area has

been uneven, limited in scope, and plagued by methodological issues and limitations, but its

history, evolution, and current status merit more attention. First, there are still misconceptions,

myths, and erroneous beliefs about sex offender recidivism that are not limited to the general

public but affect criminal justice practitioners, professionals, and policymakers. These myths

and misconceptions contribute to the idea that sex offenders constitute a distinct group

whose members are more unpredictable, more irrational,andlesslikelytochangethanother

offenders (e.g., Weekes et al., 1995). Such individuals are often thought of as on a path of

increasingly serious and violent sexual offenses, culminating in sexual homicide. Second,

despite the existence of a large body of research, its uneven quality has led to study findings

that are unclear and ambiguous. For instance, several researchers over the years have stressed

that sex offender recidivism (SOR) rates are overestimated by the general public and policy-

makers. Third, while there have been other reviews of research on sex offender recidivism

(e.g., Furby et al., 1989), these have generallybeenorientedaroundtwomainquestions:(a)

What is the base rate of sexual recidivism among individuals convicted of a sex crime? (b)

Are treatments for sex offenders capable of reducing these rates? We believe that this approach

is not only too limited and restrictive but is based on assumptions about the risk of sexual recid-

ivism that need to be examined. Therefore, given the backdrop of myths and misconceptions,

seemingly contradictory research findings, and a relatively narrow approach to a complex

issue, we feel that a narrative review of the scientific literature on this subject can make a sig-

nificant contribution to the field.

Purpose of the Study

The criminal justice system expects criminal justice professionals, expert witnesses, and researchers

to have at least some understanding of the danger posed by an individual convicted of sexual

offenses. However, clinical assessment of risk is often based on criteria that stem not from scientific

consensus but from legal considerations (e.g., Donaldson & Wollert, 2008). Conducting a risk

assessment and making some form of risk prediction have, over the years, become almost routine

for courts, correctional institutions, parole/review boards, and in civil commitment situations (e.g.,

Frechette & Lussier, 2021). The stakes of these risk assessments are high, for both the individual

and society (e.g., Janus, 2000). When a convicted offender sexually reoffends, the legitimacy of

414 Criminal Justice Review 49(4)

the criminal justice system is scrutinized, questioned, and criticized (see Sutherland, 1950). The pres-

sure to make such assessments exist irrespective of whether it is based on realistic expectations about

what professionals can and cannot do and is sometimes greeted with doubt, skepticism, and criticism,

especially when professional risk assessments seem insufficiently decisive or precise, or prove to be

inaccurate. Risk assessment is heavily dependent on the quality of the scientific knowledge and

instruments available to professionals conducting these assessments as well as the ability to use

research on these issues with professional judgment and discernment. A central part of the assess-

ment of risk rests on a key, although imperfect, measure: sexual recidivism.

Sexual recidivism has been investigated for years by researchers from various backgrounds, cre-

ating an impressive amount of scientific knowledge. The scientific literature on SOR now spans over

7 decades and, in the process, numerous issues, challenges, and debates have characterized this field

of research. While several policy analyses, reviews, and meta-analyses (e.g., Hanson & Bussière,

1998; Soothill, 2010; Lussier et al., 2022; Lussier et al., 2023a) have been conducted on SOR and

related issues, we believe that a comprehensive and critical overview on SOR research and its evo-

lution is lacking. The need for a comprehensive overview of SOR research is supported by the fact

that this literature is: (a) dense and complex, producing seemingly contradictory findings that may be

difficult to translate into criminal justice policies; (b) fragmented across academic disciplines,

research interests, and research focus; (c) evolving over time, characterized by a unique and cumu-

lating knowledge base which needs to be properly situated. The purpose of this study, therefore, is to

conduct a comprehensive review of the scientific literature on SOR by examining what we know,

what we do not know, and should know about SOR. We believe that an understanding of this knowl-

edge requires a thorough examination of SOR research and how it has evolved over time and context.

Review Search and Selection Process

We agree with Soothill (2010) that an effective review of the scientific literature on SOR requires a

legal and sociohistorical context that is beyond the scope of a systematic review. While a systematic

review is useful to address a specific research question, a narrative review can provide a wider over-

view of the research on a given topic. This type of review, however, often involves the loss of

detailed information about particular methodology and study findings. Our narrative review is there-

fore supported by information collected as part of a systematic review (e.g., see Lussier et al., 2022),

a combination of review methods that has been shown to be of value in addressing complex research

questions (e.g., Contandriopoulos et al., 2010). Using Gold Rush software, a library content compar-

ison tool, the following databases, chosen because they minimized overlaps, were selected to conduct

the literature search: Academic search premier (EBSCO), Criminal Justice Abstracts (EBSCO/

Proquest), Embase (Elsevier), ERIC (EBSCO), International Bibliography of the Social Sciences

(ProQuest), Law Journal Library (HeinOnline), Legal Source (EBSCO), Medline (Ovid), NCJRS

(ProQuest), PsycInfo (Ovid), Sociological Asbtract (ProQuest), and Web of Science. Because of

the importance of “gray literature”(i.e., unpublished material that is not included in typical comput-

erized retrieval systems), the search was expanded to several additional databases

1

to minimize the

impact of publication bias and its impact on our findings (e.g., Conn et al., 2003)

Using a combination of search terms (see Appendix I), a total of 24,417 references were retrieved.

Removal of duplicates reduced the number to 23,632 (see Figure 1). At this stage, research assistants

were instructed to identify documents that potentially included an empirical assessment of recidivism

rates for a sample of offenders. They were also instructed to screen studies that potentially included a

sample of individuals with histories of sexual offending. To do so, research assistants examined the

document’s title, the abstract, and all keywords to determine if a publication was potentially an

empirical study dealing with SOR. In doubt, research assistants were instructed to keep the document

for the next stage which involved a complete examination of the document. This stage revealed how

Lussier et al.415