Challenges of estimating the number of crimes averted through incapacitation

By Enes Al Weswasi

Photo by Quino Al on Unsplash
Photo by Quino Al on Unsplash

The understanding of the specific deterrent effects of prisons is growing, and some argue that the failure of custodial sanctions to decrease recidivism after release is now a “criminological fact” (Petrich et al., 2021). While many studies have rejected the specific deterrent effect of imprisonment, one could still argue that although incarceration may not deter crime, it might prevent it through incapacitation. This raises the question of the magnitude of this crime-control effect and whether it is substantial enough to justify an increased use of incarceration. In a study published in the European Journal of Criminology, I investigate the extent of the incapacitation effect of first-time incarcerated offenders (Al Weswasi 2024).

Trying to estimate incapacitation effects poses methodological challenges as it involves exploring how many crimes an incarcerated offender would have committed if they had not been imprisoned. This hypothetical scenario requires counterfactual reasoning and the use of a comparison group that is not incarcerated, and whose behavior can be analyzed to infer the impact of incapacitation. The difficulty arises when comparing offenders sentenced to prison with those receiving non-custodial sanctions, as these two groups systematically differ in characteristics that influence both the likelihood of receiving a prison sentence and the risk of recidivism.

Using a comprehensive collection of both time-varying and time-invariant covariates picked from multiple Swedish registers, this study employed a matching technique to pair offenders sentenced to incarceration in Sweden in 2018 with “statistical twins.” These statistical twins were individuals who were convicted in the same year but received a sentence other than incarceration. The non-incarcerated offender was subsequently observed for the exact number of incarcerated days as their incarcerated “twin”, and their criminal behavior during this period served as the counterfactual for the incarcerated group, allowing for the estimation of the incapacitation effect.

The findings reveal that for the average Swedish sentence of a first-time incarcerated offender, the incapacitation effect is 0.3 in terms of prevented convictions and 0.7 in terms of the number of convicted offenses. Most of these averted offenses, accounting for almost two-thirds, consist of narcotics or traffic-related offenses. Additionally, for first-time incarcerated offenders in the highest risk group, an average of 0.9 convictions and 1.8 offenses that would have resulted in convictions are prevented. In comparison, for offenders in the medium-risk groups, the corresponding figures are approximately 0.2 convictions and 0.5 convicted offenses prevented.

It is important to emphasize that the estimated incapacitation effect should not be interpreted as the overall net effect of incarceration. Previous studies have demonstrated that choosing incarceration over a non-custodial sanction can have criminogenic effects on certain offenders. Consequently, any crime-reducing benefits resulting from incapacitation may be offset in the long run when considering net effects. In addition, utilizing a matching approach is associated with some limitations since the method is restricted to finding suitable ‘twins’ based only on the available covariates.

As demonstrated by this study, it is evident that incarceration does have an incapacitation effect, which is not surprising considering the physical confinement of inmates within prisons. However, what stands out is that the incapacitation effect observed in this study is generally modest on average, with significant heterogeneity observed when offenders are classified into risk groups. From a policy standpoint, the suggested incapacitation effect identified for low-risk, first-time incarcerated offenders may still justify the imposition of a prison sentence. Nevertheless, it is worth noting that Sweden has one of the highest per-inmate expenditure levels, and the increased reliance on incarceration has raised concerns regarding fiscal constraints, prison overcrowding, and security issues. These concerns, in turn, have prompted apprehensions about the ability to effectively rehabilitate offenders as the inmate population grows. The relatively limited crime-preventive effect associated with incapacitating low-risk, first-time incarcerated offenders raises a crucial question: Could non-custodial sanctions be a viable alternative for certain inmates without giving rise to a risk for substantial costs in terms of recidivism?


Al Weswasi, E. (2024). Estimating the incapacitation effect among first-time incarcerated offenders. European Journal of Criminology.

Petrich, D. M., Pratt, T. C., Jonson, C. L., & Cullen, F. T. (2021). Custodial sanctions and reoffending: A meta-analytic review. Crime and Justice, 50(1), 353-424.

About the author

Enes Al Weswasi is a PhD student at The Department of Criminology, Stockholm University.

His area of research is the effects of custodial and non-custodial punishments on recidivism, labor market attachment, and health.