Reducing Recidivism: Impact of Legislative Revisions

Law and Human Behavior | 2024, Vol. 48, No. 3, p. 214-227

How the Risk Principle Reduces Recidivism: The Impact of Legislative Revisions on the Release and Reoffense Rates of Individuals Convicted of Sexual Offenses

Martin Rettenberger; Centre of Criminology (Kriminologische Zentralstelle–KrimZ), Wiesbaden, Germany; Department of Psychology, Johannes Gutenberg-University Mainz

Reinhard Eher; Federal Evaluation Centre for Violent and Sexual Offenders, Austrian Ministry of Justice, Vienna, Austria

Objective: The present study examined the relationship between legislative revisions regarding sexual offenses and the release decisions and recidivism rates of individuals convicted of sexual offenses. In 2008, the Austrian government passed a package of revised criminal laws aiming to decrease incarceration rates. At the same time, connecting recidivism risk to professional risk management efforts was expected to increase public safety. Hypotheses: Given the strong empirical background of the implemented risk assessment and management efforts, we expected both an increase in the percentage of conditional release decisions and a decrease in recidivism rates.

Method: We analyzed the data of 2,610 male individuals convicted of sexual offenses who were released from the Austrian Prison System between 2001 and 2016 within a natural experiment using a prospective-longitudinal quasi-experimental study design. Results: The results indicated that the percentage of conditional releases of individuals convicted of sexual offenses increased substantially since 2008. Additionally, within the same period, the recidivism rates of individuals convicted of sexual offenses decreased further.

Conclusion: Even if both developmental processes are only correlational and a causal relationship cannot be examined, the present results supported the empirical evidence of the risk principle—at least if it is based on scientifically sound risk assessment and management methods.

risk assessment, sexual offenses, sexual recidivism, Static-99, risk principle

“The sentencing, incarcerating, and releasing of individuals convicted of sexual offenses represent an important societal and political issue with high public interest in most countries worldwide. Societies generally agree on the need to protect the public from further possible offenses committed by individuals already convicted of a sexual offense. Because those individuals are already identified as persons who have committed crimes in the past, the public usually expects that the judicial authorities are particularly attentive when it comes to release decisions in these cases… Legislation in almost every Anglo-American and European country responded to this societal climate by implementing different measures to control, treat, and supervise individuals convicted of sexual offenses” (p. 214 - 215). 

“The main aim of the present study was to evaluate a possible relationship between legislative revisions in Austria in 2008 leading to changes in the release decision practice of prison-sentenced individuals having committed sexual offenses. Since January 2008, the new sexual offending laws in Austria require the systematic consideration of different risk management aspects: (a) conducting a comprehensive risk assessment at the beginning of the client’s imprisonment and preparing an individual intervention plan based on treatment needs; (b) treating and managing those individuals during and after their prison term; and (c) conditionally releasing all imprisoned individuals ) after the first half or at least after two-thirds of their prison sentence, if their relative risk for reoffending is not higher when conditionally released compared with the release at the end of their prison term” (p. 215).

“The core research question was whether an increase in the conditional release rate is related to an increase or decrease in reoffense rates… The present investigation can be regarded as a natural experiment using a prospective-longitudinal quasi-experimental study design. The database of the present study consisted of all incarcerated male individuals convicted of a sexual offense in Austria and released between 2002 and 2016 … The final sample of the present study contained 2,610 individuals. The data of the present study were collected at the FECVSO” (217 - 218). 

“The results of the present study show that, first, the conditional release rates of individuals convicted and incarcerated because of a sexually motivated offense in Austria substantially increased since the new law was implemented in 2008. Second, this increase was accompanied by substantial annual savings of public funds, which steadily increased to approximately €5 million in 2015. Third, and most importantly, reoffense rates substantially and significantly decreased within the corresponding timeframe. The current results indicate the positive influence of the new law on reoffense rates even after controlling for time on risk and for the baseline risk level of the participants in the present study. Taken together, the findings suggest a reduction in the number of imprisoned individuals convicted of sexual offenses by conditional release policies, which can be accompanied by a reduction in sexual and violent recidivism rates and, thus, by an increase in public safety” (p. 222).

“Generally, it is important to note that the penal law reform in Austria in 2008 must be considered and interpreted against the background of other criminological, societal, historical, political, and psychological developments. The two most important aspects in this context are the phenomenon of a general decrease in crime and (sexual and violent) recidivism rates and the general increase in public, media, and political concern about sexual offending in the past decades in all Western countries. Regarding the latter, it should be considered that in Austria—as in many other countries—sexual offending in general and sexual recidivism in particular (individuals who are already known to the law enforcement authorities) are an intensively discussed issue of concern. Because of tragic single cases in the 1990s, the assessment and management of sexual recidivism risk have become a general societal subject. Consequently, many policymakers and researchers interpreted the legislative changes as indicators of a punitive turn, leading to a substantial increase in the number of incarcerated individuals convicted of sexual offenses” (p. 222). 

“Putting together the results of these previous studies about the effects of the SORN laws and the findings of the present study, we can conclude from the existing state of research that public safety could be increased when individuals convicted of sexual offenses are released early (conditionally) from prison, under the circumstances that the release decisions are, first, based on empirically derived and psychometrically sound risk assessment procedures and, second, directly related to professional and effective risk management strategies, which are mandated by criminal law courts. Evaluation studies conducted in Austria and Germany about the effectiveness of treatment programs, which were provided by professional aftercare outpatient clinics for individuals convicted of sexual offenses and released from prison, exhibited significant and substantial treatment effects when scientifically based treatment programs and rehabilitative models, such as the RNR model, were used. If such treatment programs are used together with appropriate risk assessment approaches, the sexual recidivism risk of individuals already known by law enforcement authorities can be sustainably and substantially reduced. However, such risk-reducing effects cannot be expected if sexual offending policies are only or predominantly repressive, deterrent, and punitive. 

In this context, it is important to note that in the present study, only the risk principle was investigated in detail. However, the other two components of the RNR model also have some relevance in the discussion of the results of the present study. First, we discuss the clinical, treatment-related, and assessment-related aspects of the need and responsivity principles in the context of the present study; in the following section, we discuss the theoretical implications. As previously published meta-analytical findings indicated, the risk-reducing effect can be regarded as a linear function of the number of implemented RNR principles—that is, the more the three RNR principles are considered and implemented systematically, the higher the decrease in recidivism risk. In the present study, the empirical examination was limited to the application of static (i.e., unchangeable) risk factors captured by the Static-99. However, the systematic assessment of stable- and acute-dynamic risk factors, which supports the additional consideration of the need and responsivity principles, would probably contribute to a further decrease in recidivism (risk) by increasing the effectiveness of intervention programs. 

Another aspect related to the assessment and the treatment planning process, which could lead to an additional reduction in recidivism rates, would be the systematic inclusion of protective and strength-based factors over and above dynamic risk factors. The legislative revisions described in the present study included only the systematic and direct consideration of the assessment of static risk factors, whereas the inclusion of dynamic risk and protective factors occurred in an unsystematic and unstructured way. This means that some treatment facilities and aftercare institutions include systematic assessment, documentation, and implementation of dynamic, protective, and responsivity-related factors, whereas other institutions have not considered them (because of resource constraints or a lack of appropriate assessment training opportunities). For the present study, no data about dynamic risk, protective, and responsivity-related factors were available, so we could only hypothesize about their potential (additional) influence. At the current state of research, we have substantial reason to believe that the systematic assessment of (static) risk using a scientifically proven standardized instrument has contributed to a risk-based allocation of resources and a systematic consideration of recidivism risk in the course of treatment, supervision, and surveillance planning” (p. 223). 

“Although the results of the present study can be interpreted as further support for the effectiveness of the risk principle, some limitations of the present study need to be addressed. First, the recidivism data used in the present study were based solely on one single data source: the computerized database of the Austrian Ministry of Internal Affairs. Given the fact that in rights-oriented countries, only officially proven offenses should count as reoffending, and although these data are comparatively reliable and are the most commonly used operationalization for recidivism internationally, there is an inherent restriction to one source of outcome measurement. Reconviction rates are—as all official crime rates— inevitably an underestimation of the actual recidivism, and previous research indicated that this limitation was especially relevant for empirical investigations of sexually motivated reoffenses. However, compared with the criminal justice systems of Anglo-American countries, in which a substantial proportion of sexually motivated offenses would remain unrecognized because the official conviction would imply a nonsexually motivated (e.g., a nonsexual violent) offense because of, for example, plea bargaining, German-speaking countries have a lower frequency of plea bargaining, particularly for serious delinquent behaviors such as sexual offenses. This means that this and related sources of biases regarding the measurement of recidivism in terms of reconvictions presumably play a smaller role than in studies from North American countries. Second, there are different risk-relevant variables that could not be considered in the present study because they were not available in the FECVSO database (e.g., race or education level). Third, the results of the present study were based on a sample of male participants and are therefore not generalizable to female individuals convicted of sexual offenses. Fourth, it is important to note that in Austria mandated aftercare is available only via conditional release; in contrast, in other countries such as the United States or Germany mandated aftercare can be ordered without conditional release—at least under certain juridical conditions such as the Sexually Violent Predator commitment laws. Therefore, the generalizability of the results of the present study to other countries may be limited because of different juridical opportunities and backgrounds. Fifth, even if the total sample of the present study is comparatively large, the effect sizes for some analyses are small, which inevitably tempers the extrapolation of the findings. Sixth, as explained above, only static (i.e., unchangeable) risk factors captured by an actuarial (statistical, mechanical, nomothetic) risk assessment instrument were considered, whereas stable- and acute-dynamic risk factors were neglected. Seventh, the economic calculations suffer from several uncertainties and unknown variables, which reduces the informative value of this part of the analysis. For example, no information was available about the (average) costs of running a conditional release program (i.e., the annual costs for an aftercare treatment program per individual) and about the additional financial expenditures induced by the new risk assessment system. However, even if exact numbers were not available, we could estimate that at least the latter amount was rather low. Taken together and as explained earlier, the savings are overestimated as well as underestimated because not all saved costs due to the prevention of further offenses could have been considered. Another limitation refers to the research design of the present study: By using a prospective-longitudinal quasi-experimental study design, we were limited in the ability to differentiate recidivism-reducing effects because of the legislative changes in 2008 and other possible criminological, societal, historical, political, and psychological developments within the same time frame. In future studies, the Austrian recidivism data could be compared with data from a similar country, and both data sources could be analyzed comparatively using a nonequivalent control group interrupted time-series design. For the present study, such recidivism data from another country were not available for comparison purposes. Finally, the representativeness of the present sample is slightly limited because it included only individuals who were convicted and incarcerated because of the commission of a sexual offense, so low-risk participants (who had not served a prison sentence because of their sexual offense) could be underrepresented” (p. 224 - 225).