In a five-to-eight-year study of 140 men convicted of sexual crimes against children, researchers at Université de Montréal discovered something that reshapes how we think about rehabilitation: the quality of the relationship between a person in treatment and their therapist is the single most powerful factor in preventing reoffending.
The finding, published in April in the Journal of Child Sexual Abuse, matters because it offers a clear pathway forward for a question that has puzzled researchers since the 1980s. While extensive studies have examined rehabilitation programs and clinician roles, understanding how these elements work together to prevent reoffending after treatment ends has remained elusive. Étienne Garant, a postdoctoral researcher supervised by psychology professor Tamsin Higgs, set out to untangle these interactions by following men through incarceration-based treatment and into group therapy sessions conducted after their release, often as a parole condition.
Garant began by measuring cognitive distortions—the rationalizations and false beliefs offenders use to justify their crimes, such as minimizing harm, blaming victims, or dismissing consequences. He found these distortions notably decreased by the end of treatment. The real insight came when he investigated what actually drove that change. Three factors stood out: a person's motivation to change, the atmosphere of group therapy, and something called therapeutic alliance—the quality of trust, collaboration, and openness that develops between therapist and patient.
The therapeutic alliance emerged as the decisive factor. Using the California Psychotherapy Alliance Scale, a standardized 24-point assessment tool, Garant measured how participants rated their bond with their therapist, their receptiveness to feedback, and their perception of support. The results were clear: this human connection mattered more than the other variables.
The implications point toward a humanistic approach to therapy—one that sees a person as distinct from their offense. Garant observed that the dominant discourse surrounding these individuals often reduces them to labels: sex offender, child molester. "In a therapeutic setting, this attitude can hinder the therapeutic alliance, and if a person is constantly reduced to their offense, they may eventually internalize this identity and regard it as their inescapable fate," he explained.
Practical techniques can strengthen this alliance. Reflecting and reframing rather than confronting, avoiding dynamics that center guilt, and—most importantly—declining to reduce a person to their deed all help build the foundation for change. Garant is careful to note that this humanistic stance does not minimize the gravity of the crimes. The majority of participants had committed sexual offenses against family members. These are serious offenses, and the system has already held offenders accountable through trial and incarceration.
"But the person has been tried, served a prison sentence and undergone treatment," Garant said. "Now we have to focus on their potential."
This reframing does not excuse harm or abandon accountability. Instead, it recognizes that once a person has served their sentence and engaged in the hard work of treatment, their capacity for change becomes the focus. The research suggests that building genuine therapeutic relationships—where a person feels heard, respected, and seen as more than their worst act—is what actually keeps children safer by reducing the likelihood of reoffending. It is a reminder that healing and protection are not opposing forces, but intertwined.