Gender Inequality, Sexual Violence, and Legal system Responses: Comparative Sex Offender Governance in Seven Jurisdictions

1. Introduction

Sexual violence remains one of the most consequential forms of gender-based harm. It violates bodily autonomy, disrupts family and community life, and reproduces gender hierarchy through fear, coercion, and unequal access to justice. Feminist criminologists have long argued that sexual violence cannot be understood solely as an interpersonal act; it is embedded in social arrangements that normalize male entitlement, minimize women’s credibility, and fragment institutional responsibility (Daly & Chesney-Lind, 1988; Gelsthorpe, 2002). At the same time, criminal justice systems are repeatedly asked to solve a practical question: how should states supervise, treat, and restrain people convicted of sexual offenses after conviction and release?

These two lines of inquiry are often kept apart. One body of scholarship focuses on survivors, reporting barriers, rape myths, attrition, and gender inequality. Another centers on sex offender policy, including registration, risk assessment, cognitive-behavioral treatment, relapse prevention, and civil commitment. This separation can distort policy. A purely punitive approach may express social condemnation but fail to reduce reoffending. A purely therapeutic approach may neglect power, coercion, and the gendered social conditions that make sexual violence possible. The present article therefore treats sex offender governance as part of the broader state response to gender inequality and sexual violence.

This article compares sex offender governance across seven legal jurisdictions—the United States, the United Kingdom, Canada, Germany, India, Indonesia, and Taiwan—with Taiwan examined as a distinct legal-administrative jurisdiction. Because they represent different legal traditions, levels of institutional development, and degrees of investment in post-sentence management. The comparison is designed to answer three questions. First, how do jurisdictions differ in their governance of convicted sexual offenders, particularly in registration, risk assessment, treatment, and preventive detention? Second, what types of arrangements appear more coherent and more likely to contribute to the prevention of revictimization? Third, why do some jurisdictions build evidence-informed systems while others rely mainly on punitive or symbolic responses?

The title deliberately uses the term governance rather than policy. Governance captures the layered reality of this field: legislation, court doctrine, administrative practice, risk instruments, supervision routines, treatment models, and local implementation all matter. In many jurisdictions, statutory severity tells only a small part of the story. A state may enact harsh penalties while lacking validated tools, trained personnel, or sustained community treatment. Another may reject public shaming but maintain a dense system of probation, case review, and targeted treatment. The real question, then, is not who punishes most harshly, but which jurisdictions create durable structures capable of managing risk while preserving legality and proportionality.

The article proceeds as follows. It first outlines the conceptual framework, drawing on feminist criminology, comparative criminal justice, and evidence on recidivism reduction. It then explains the narrative comparative method. The following sections examine each of the seven jurisdictions, after which the article compares them across the four governance domains. Taiwan is then discussed as a distinctive case that links feminist advocacy, statutory design, and integrated treatment. The article closes by identifying implications for low- and middle-income jurisdictions, especially India and Indonesia, and by arguing that the governance of sexual offenders should be understood as a component of gender-responsive state capacity rather than as a stand-alone technical field.

2. Conceptual and Analytical Framework

Three propositions guide this analysis. The first comes from feminist criminology: sexual violence is inseparable from gendered structures of power, socialization, and institutional response. Therefore, state responses should not be judged only by conviction rates or sentence length, but also by whether they prevent revictimization and take women’s safety seriously over time. Policies that ignore recidivism risk, post-release support, and offender treatment may appear tough while leaving women and children exposed to repeated harm.

The second proposition is drawn from comparative criminal justice. Legal families matter, but they do not determine outcomes by themselves. Common-law jurisdictions such as the United States, the United Kingdom, Canada, and India often rely heavily on precedent, discretionary supervision, and layered administrative practices. Civil-law jurisdictions such as Germany and Taiwan rely more explicitly on statute and codified procedure. Yet the comparison in this article shows that common-law systems vary sharply among themselves, as do civil-law systems. Legal tradition shapes the form of regulation, but political learning, administrative capacity, and the legitimacy of expert knowledge shape whether systems become coherent.

The third proposition concerns effectiveness. Research on sexual recidivism consistently suggests that neither denunciatory punishment nor generic therapy is enough. Better outcomes are associated with structured risk assessment, interventions aligned with criminogenic needs, responsivity to learning style and motivation, and post-release supervision integrated with treatment (Hanson et al., 2009; Hanson et al., 2017). Static instruments are useful for baseline classification, but dynamic instruments are needed to capture change over time. Similarly, supervision without treatment can generate surveillance fatigue, while treatment without monitoring may fail to address acute relapse conditions.

For these reasons, the article compares jurisdictions across four governance domains. The first is registration and notification, including whether information is public, restricted, or unavailable to communities. The second is risk assessment, including the use of validated actuarial or dynamic tools. The third is treatment, including prison-based and community-based intervention. The fourth is preventive detention or its functional equivalent, including civil commitment, dangerous offender designations, and post-sentence custodial measures for individuals judged to remain at very high risk.

These domains are analytically linked. Registration without risk assessment may produce indiscriminate stigma. Risk assessment without treatment produces classification without change. Treatment without community supervision risks discontinuity at the point of release. Preventive detention without due process invites constitutional challenge. A coherent governance model is therefore one that connects these domains through law, professional training, and multi-agency implementation. This framework also makes visible a deeper argument: the quality of sex offender governance often reflects the seriousness with which a state addresses gender inequality itself. Where women’s political participation and institutional influence are greater, laws and practices are more likely to move beyond symbolic anger toward durable prevention.

The Taiwan case may also be situated within the broader literature on Asian criminology. Rather than treating Asian jurisdictions merely as sites for the transplantation of Western penal technologies, Asian criminology asks how local legal cultures, institutional capacities, relational norms, and state practices shape distinctive approaches to crime control and justice. This perspective is useful for the present comparison because Taiwan’s model combines imported actuarial tools, local treatment innovation, controlled disclosure, and post-sentence legal review in a way that is neither fully American nor fully European. The case therefore illustrates how Asian jurisdictions may adapt global criminological knowledge while also contributing new concepts and practices to comparative sex offender governance (Liu, 2009; 2017). By leveraging this perspective, the present analysis positions Taiwan, India, and Indonesia not merely as diverse administrative units, but as sites of knowledge production that contribute to the evolving Asian criminological paradigm (Liu, 2016; 2021)

3. Method and Comparative Design

This article uses a narrative comparative design combining doctrinal review, policy comparison, and secondary analysis of empirical studies. The goal is not to produce a formal ranking of all jurisdictions, but to identify recurring governance patterns and policy logics. The seven jurisdictions were selected for substantive variation: they include high-income and middle-income settings, common-law and civil-law traditions, stronger and weaker gender-equality environments, and both established and underdeveloped treatment systems.

The principal materials include statutes, court decisions, government reports, and research literature on risk assessment, treatment, supervision, and post-sentence detention. The article also draws on Taiwanese program and recidivism data cited in earlier studies and local evaluations, including work by Lin and Dong (2005), Hsu et al. (2023), Lin and Chen (2023), and Lin et al. (2024). These materials are used not to claim perfect comparability of recidivism rates across legal systems, which is not currently possible, but to show how governance arrangements shape what jurisdictions can monitor and learn.

The article adopts a structured comparative strategy. For each jurisdiction it asks: What is the legal model of registration or disclosure? Are validated risk tools used? Is treatment mandated in custody, in the community, or both? What high-risk post-sentence mechanisms exist? The structured questions permit comparison without erasing local context. They also help distinguish between punitive intensity and administrative capacity. Some jurisdictions impose severe sanctions but have thin treatment infrastructures. Others restrict public disclosure while investing heavily in professional assessment and case management.

Several limitations should be noted. First, legal rules change over time, and local implementation may differ from region-wide policy. Second, direct cross-jurisdictional comparison of recidivism is limited by different definitions, follow-up periods, and recording practices. Third, some claims in the literature remain contested, especially regarding the deterrent effect of public notification. These limitations do not defeat comparison, but they require caution. Accordingly, the article emphasizes governance architecture and institutional coherence rather than simplistic scorekeeping.

Despite these limitations, comparative analysis remains valuable. It reveals that the field is not reducible to a binary between harsh punishment and permissive rehabilitation. Real-world systems combine law, technology, professional judgment, and local politics in highly variable ways. The comparison also shows why low-capacity jurisdictions should not copy the most visible punitive measure from elsewhere. The more transferable lessons concern validated assessment, professional training, multi-agency coordination, and legal mechanisms that link custody, community treatment, and periodic review.

4. Comparative Findings Across Seven Jurisdictions

4.1 United States

The United States is the most influential and the most contested jurisdiction in the field of sex offender governance. The modern system developed through a sequence of high-profile child murder cases and corresponding legislative responses. Washington State began community notification in 1990. The Jacob Wetterling Act of 1994 created federal registration expectations, and Megan’s Law in 1996 required states to establish notification procedures. SORNA under the Adam Walsh Act later standardized federal expectations by organizing registry obligations into tiers tied to offense seriousness.

The most distinctive feature of the U.S. model is its public-facing orientation. In many states, registry websites provide names, addresses, and photographs. The rationale is protective transparency. Families, schools, and communities are expected to use information to avoid danger. Yet this model has also generated the strongest evidence of collateral harm. Studies reviewed by Lasher and McGrath (2012) and Levenson and Cotter (2005) suggest that public notification can intensify housing instability, unemployment, social isolation, and vigilantism. The empirical question is not whether public exposure expresses condemnation - it clearly does - but whether it improves long-term safety enough to justify those costs.

The U.S. nevertheless remains a leader in assessment technology. Static-99R, STABLE-2007, Acute-2007, MnSOST-R, VASOR-2, SOTIPS, and related instruments are used in different states and settings. This is a major strength of the system. Risk is not treated only as a moral label; it is also operationalized through tools used for sentencing support, supervision level, treatment placement, and post-release monitoring. The best American practice therefore lies not in public shaming but in its assessment infrastructure and the growth of containment-style community supervision.

Treatment in the United States is decentralized. Cognitive-behavioral therapy remains dominant, with relapse prevention and the Good Lives Model used to varying degrees. The U.S. experience also provides an important caution: not all treatment is equally effective. Marques et al. (2005) found that relapse prevention as delivered in the California SOTEP program did not significantly reduce sexual recidivism compared with controls. This finding contributed to a broader shift toward interventions more closely aligned with Risk-Need-Responsivity principles and with dynamic risk monitoring.

The final hallmark of the American system is civil commitment for sexually violent predators in a number of states. This mechanism allows high-risk offenders to be held after sentence completion if they meet statutory criteria, usually involving mental abnormality and serious risk. Kansas v. Hendricks (1997) upheld the constitutionality of such schemes, framing them as treatment rather than punishment. The practical result is a determinate-plus model in which some offenders move from prison to secure treatment custody. From a governance perspective, the United States combines sophisticated assessment and powerful incapacitative tools with a highly punitive, often public, notification regime.

4.2 United Kingdom

The United Kingdom presents a contrasting model in which information control is more selective, and multi-agency coordination plays a central role. ViSOR and MAPPA created an infrastructure in which police, probation, prisons, and related agencies share information and stratify cases according to seriousness and need. MAPPA is particularly important because it embodies a governance logic that is administrative rather than populist. The public is not asked to monitor everyone; instead, agencies coordinate around assessed risk.

The United Kingdom does not operate a U.S.-style public registry. Through Sarah’s Law, limited disclosures can be made when necessary to protect a child, but the model remains controlled and case-specific. This reflects a different balance between public protection and rehabilitation. British policy assumes that indiscriminate public exposure may undermine reintegration and provoke harm while adding little to actual prevention. The British system therefore seeks to preserve operational secrecy while increasing professional information sharing.

Risk assessment in the United Kingdom relies on tools such as Risk Matrix 2000 and SARN. These instruments support the gradation of supervision and case management. British practice illustrates an important lesson for comparative policy: the utility of a risk tool depends not only on predictive validity but also on whether agencies have a decision framework into which scores can meaningfully feed. MAPPA provides that framework.

Treatment has evolved considerably. The discontinuation of the old Core Sex Offender Treatment Programme after the Mews et al. (2017) evaluation is often cited as evidence of policy failure, but it also indicates institutional willingness to revise ineffective practice. Newer programs such as Horizon and Kaizen incorporate cognitive-behavioral, desistance, and responsivity elements. The United Kingdom therefore shows a form of governance that is less theatrical than the U.S. model but more bureaucratically integrated. Its weakness lies in persistent tension between surveillance and rehabilitation, yet its overall design is more discriminating and less reliant on community fear.

4.3 Canada

Canada occupies an intermediate position between the American and British models. It has a national registration framework through the Sex Offender Information Registration Act, but it does not embrace broad public notification. Canadian law has also shown significant constitutional sensitivity. In R v. Ndhlovu (2022), the Supreme Court invalidated mandatory lifetime or automatic registration provisions that were judged overbroad. Subsequent reform through Bill S-12 preserved registration for serious child-sex and repeat cases while restoring more individualized proportionality.

This rights-consciousness is important. Canadian policy demonstrates that public protection and constitutional review can coexist. Rather than treating registries as politically untouchable, Canadian law asks whether the burden is justified in relation to risk. That question is central to any gender-responsive and rights-conscious approach, because overbroad systems may satisfy public anger while misallocating state attention.

Canada also shares the North American investment in validated risk tools, including Static-99R, STABLE-2007, and Acute-2007. At the treatment level, Canadian practice is strongly shaped by Risk-Need-Responsivity principles. Hanson et al. (2009) concluded that the principles of effective correctional treatment apply to sexual offenders as well. Canada has also contributed one of the most notable community support models: Circles of Support and Accountability. COSA underscores the idea that risk reduction is not achieved only by surveillance but also through structured prosocial inclusion.

For very high-risk offenders, Canada uses dangerous offender designations and long-term supervision orders rather than American-style civil commitment. Functionally, this still allows indeterminate or extended control of serious cases, but through a different legal route. Compared with the United States, Canada is less public, more rights filtered, and more willing to treat reintegration as part of public protection itself.

4.4 Germany

Germany offers one of the clearest alternatives to the American public-notification model. German law strongly protects personal data and generally rejects open community disclosure of offender information. Risk management instead relies on court-imposed measures, treatment, probation, and preventive detention for serious high-risk cases. Information may be shared with relevant agencies or institutions under controlled conditions, but not as a general public warning tool.

The German approach is therefore anchored in a rehabilitative and privacy-oriented tradition. Yet this should not be misunderstood as leniency. Germany has used preventive detention, or Sicherungsverwahrung, for very dangerous offenders, and legal controversy has arisen over the extension of confinement. European human-rights review forced doctrinal and statutory refinements, showing again that durable governance requires not only control capacity but also lawful review mechanisms.

Germany also uses structured risk instruments, including Static-99R, STABLE-2007, and HCR-20 in adapted settings. These tools support a professionally mediated model of governance rather than public disclosure. Treatment is comparatively well developed, with therapeutic work in prisons and post-release support. This means that Germany is capable of combining privacy, assessment, and incapacitation without embracing populist exposure.

From the standpoint of gender-responsive state response, Germany’s significance lies in demonstrating that a state can reject public shaming while still taking sexual reoffending seriously. The real strength of the German model is not secrecy alone but secrecy plus professionalization. Public access is limited, yet case management and high-risk control remain substantial.

4.5 India

India’s response to sexual offending has been shaped by public outrage, colonial legal inheritance, and persistent institutional gaps. Sexual offense law was historically rooted in the Indian Penal Code, and despite major reforms after shocking sexual assault cases, the system remains primarily punitive in design. India established a national sex offender registry in 2018, but it is not public and appears to function mainly as a law-enforcement resource. Unlike in the U.K. or Canada, however, that confidentiality is not embedded within a broader, mature system of specialized supervision, risk assessment, and mandated treatment.

The most salient weakness in the Indian field is the absence of a validated national risk assessment system for sexual offenders. Decisions about seriousness and release conditions are often driven by judicial discretion, psychiatric opinion, or offense facts rather than structured actuarial and dynamic tools. This means that India lacks one of the key mechanisms by which jurisdictions allocate scarce supervision resources. Without validated classification, it is difficult to distinguish who requires intensive community management from who does not.

Treatment is also underdeveloped. India has psychiatric hospitals, NGOs, and scattered counseling resources, but no widely institutionalized prison-and-community continuum comparable to those in Taiwan, Canada, or parts of the United States. The result is a policy field in which condemnation is visible but rehabilitation infrastructure is thin. From a feminist perspective, this is especially problematic because severe punishment after major incidents does little to prevent the next offense if the system lacks sustained supervision, tailored treatment, and evidence production.

India therefore illustrates a broader comparative lesson: confidentiality alone does not make a system humane or effective. In the United Kingdom or Canada, limited public disclosure is paired with professional case management. In India, the lack of public notification does not necessarily signal rehabilitative sophistication; it often reflects administrative incompleteness. Reform priorities should thus include locally validated risk tools, trained probationary and therapeutic personnel, and pilot prison-community treatment programs informed by evidence rather than episodic outrage.

4.6 Indonesia

Indonesia’s response is marked by primarily punitive or symbolic responses and legal hybridity. Sexual offense regulation reflects both Dutch colonial inheritance and, in specific regions such as Aceh, the influence of Islamic legal norms. Public and political responses to child sexual abuse have often framed these offenses as extraordinary crimes requiring extraordinary penalties. This climate facilitated the authorization of chemical castration, identity disclosure, and electronic monitoring in child sexual abuse cases.

The Indonesian model is notable for the sharp divergence between adult and juvenile approaches. Juvenile justice increasingly uses restorative ideas, while adult sex offender policy remains overwhelmingly punitive. The legal power to impose chemical castration has attracted intense attention, but treatment infrastructures capable of changing dynamic risk appear limited. Indonesia therefore exemplifies a governance pattern in which punitive visibility substitutes for systematic management.

There is little evidence of nationwide use of validated sex offender risk tools such as Static-99R or dynamic instruments. This absence matters because it leaves the system without a robust basis for differentiating low-, moderate-, and high-risk cases. Harsh sanctions may communicate state seriousness, but they do not tell officials how to allocate treatment, monitoring, or review. They also risk human-rights and medical-ethics conflict, especially where medical associations resist implementation.

For comparative purposes, Indonesia shows why symbolic severity should not be mistaken for institutional strength. The key missing links are the same as in India: validated risk assessment, trained personnel, prison-and-community treatment, and legal pathways that connect punishment with supervised reintegration. Without these, policy remains reactive and highly dependent on public anger after notorious cases.

4.7 Taiwan

Taiwan is a distinctive jurisdiction because its contemporary framework emerged through the interaction of feminist mobilization, public shock after sexual homicide cases, and sustained legal and professional reform. After the murder of feminist activist Peng Wan-ju in 1996, Taiwan enacted the Sexual Assault Crime Prevention Act and built local sexual assault prevention committees. Later reforms after additional notorious cases added stronger post-incarceration treatment, reporting requirements, community supervision, and mechanisms for high-risk case management.

Unlike the United States, Taiwan did not fully adopt a Megan’s Law model of public exposure. Personal information is not broadly disclosed to the public, although specific persons or organizations may apply for information where safety concerns exist, and aggregate or localized information on high-risk cases may be published. This compromise reflects an attempt to increase vigilance without creating generalized hatred toward released offenders. The policy logic is closer to controlled disclosure than to open public registry.

Taiwan’s most important contribution lies in linking law to assessment and treatment. The system uses Static-99-based classification under the Sexual Assault Crime Prevention Act, and Taiwan also produced an indigenous actuarial scale, the Taiwan Sex Offender Static Risk Assessment Scale (TSOSRAS), developed by Lin and Dong (2005). SONAR was translated for community use, with stable factors reviewed periodically and acute factors monitored more frequently. These elements show that Taiwan did not merely import punishment; it imported and adapted professional tools.

Treatment is mandated both in prison and in the community. Lin and Chen (2023) described the Four-Key Therapy model as an integrated intervention combining cognitive-behavioral and relapse-prevention elements with trauma-informed and nutritional components. Whether every component of that model will be replicated elsewhere remains an open question, but the broader institutional point is clear: Taiwan built continuity across custody and release. Supervision may include regular reporting, home visits, community counseling, and electronic monitoring for serious cases. High-risk offenders may also face post-sentence custodial treatment under legal review since 2005.

Following a constitutional challenge regarding post-sentence treatment environments, Taiwan’s constitutional adjudication mechanism issued a 2020 ruling (Interpretation No. 799) declaring the use of prison-based facilities for such purposes unconstitutional. The Court ordered a transition period of two years for relocation; as a result, treatment is now conducted in designated medical wards within community hospitals to ensure a clearer distinction between punishment and therapy. Moreover, the release application review should be processed by court yearly after five-year treatment.

Taiwanese recidivism findings, while not perfectly comparable with international figures, are nevertheless significant. Lin and Dong reported that region-wide seven-year recidivism in Taiwan declined from 11.3% among offenders released in 1994-1996 to 6.7% among offenders released in 1997-1999 after the expansion of community treatment. Later county-level analyses cited by Hsu et al. (2023) and Lin et al. (2024) reported very low recidivism in several localities and a zero rate in Chiayi County across two observational periods. These figures should be interpreted cautiously, but they suggest that Taiwan’s integrated legal and treatment structure merits serious comparative attention rather than dismissal as a local anomaly.

5. Risk Assessment Across Jurisdictions

Risk assessment is the clearest dividing line between more developed and less developed governance infrastructures. In the United States, Canada, Germany, the United Kingdom, and Taiwan, structured instruments are used to classify cases, guide treatment, and allocate supervision. In India and Indonesia, comparable national infrastructures are absent or underdeveloped. This distinction matters more than whether a jurisdiction uses a public registry. Without structured assessment, a state cannot target scarce resources well.

Static instruments such as Static-99R remain influential because they provide a baseline estimate derived from historical risk markers. Their predictive validity is moderate rather than perfect, but they outperform unstructured intuition and support transparent decision-making (Hanson & Morton-Bourgon, 2009). Dynamic tools such as STABLE-2007, Acute-2007, SONAR, and SOTIPS go further by capturing changeable factors relevant to treatment progress and imminent concern. The strongest systems therefore combine static and dynamic assessment rather than relying on either alone.

The comparative lesson is not that every jurisdiction should copy a single tool. Tools must be culturally translated, legally understood, and professionally implemented. But the principle is portable: supervision intensity and treatment dosage should follow structured assessment. Taiwan’s development of an indigenous scale alongside imported tools is especially instructive because it shows that local adaptation is possible without abandoning evidence-based logic.

Risk tools also matter normatively. They allow states to avoid treating all convicted sexual offenders as equally dangerous forever. This is a due-process issue as much as a management issue. A rights-conscious system should restrict liberty as much as necessary but no more. Structured assessment helps operationalize that principle.

6. Treatment Models and Their Limits

The comparative record shows broad movement toward cognitive-behavioral and risk-responsive treatment, but with important differences. The United States and Canada have long histories of prison and community programs, though not all evaluated models have shown positive effects. The United Kingdom revised its treatment architecture after national evidence called the old program into question. Germany emphasizes therapeutic work linked to rehabilitation and reintegration. Taiwan mandates both prison and community treatment and has experimented with integrated approaches such as Four-Key Therapy. India and Indonesia remain comparatively underdeveloped.

Two cautions follow. First, treatment cannot be romanticized. Not every program works, and fidelity matters. Marques et al. (2005) remains a landmark reminder that a named intervention is not enough. Second, treatment cannot be detached from supervision and risk assessment. Some of the best meta-analytic evidence suggests that treatment is more effective when embedded in risk-informed systems rather than delivered as generic counseling (Hanson et al., 2009; Hanson et al., 2017).

At the same time, the evidence clearly supports rejecting a purely punitive model. Lösel and Schmucker (2005) found that psychological treatment for sexual offenders, on average, reduced recidivism. The exact magnitude varies across studies and designs, but the general implication is strong: states that fail to develop treatment capacity are leaving a preventable portion of risk untouched. This is why feminist concern for women’s safety should not be interpreted as hostility to treatment. On the contrary, if treatment reduces reoffending, then investment in treatment is part of protecting women and children.

Taiwan’s Four-Key Therapy warrants careful, critical attention. Some components - especially trauma-informed work and responsivity to health and addiction - align with emerging correctional interests. Other components require more independent evaluation before broad transfer. Still, the model is valuable at minimum for highlighting that sexual offending is often sustained by multiple pathways: trauma, distorted learning, cognition, self-regulation problems, and comorbid health or substance issues. The comparative field would benefit from more quasi-experimental and longitudinal tests of integrated models rather than continued dependence on either punishment rhetoric or narrow single-technique interventions.

7. Registration, Notification, and Preventive Detention

Registration and notification policies show the widest ideological divergence. The United States uses the most public model, while Germany rejects public notification and the United Kingdom and Taiwan rely on more controlled disclosure. Canada maintains registration but tempers it through constitutional review. India’s registry is not public, but this confidentiality exists without a corresponding mature treatment and supervision system. Indonesia may authorize disclosure in some cases but lacks a broader coherent framework.

The evidence does not support a simple equation between more publicity and more safety. Public notification can increase awareness, but it can also destabilize reintegration, intensify stigma, and sometimes provoke vigilantism (Levenson & Tewksbury, 2009; Lasher & McGrath, 2012). If housing and employment collapse, dynamic risk may rise rather than fall. Controlled disclosure models seek to avoid these harms by directing information to those who can use it meaningfully.

Preventive detention or its functional equivalent raises a different issue. Here the comparative question is not publicity but how states manage the very small group regarded as persistently dangerous after sentence completion. The United States uses civil commitment in a number of states; Canada relies on dangerous offender mechanisms and long-term supervision; Germany uses preventive detention; Taiwan authorizes post-sentence custodial treatment under statutory and judicial review. India and Indonesia generally lack parallel, mature systems. Across jurisdictions, the legitimacy of such confinement depends on periodic review, legal clarity, and genuine treatment opportunities.

In practice, the most coherent model is not public notification alone and not preventive detention alone. It is a layered system in which low- and moderate-risk offenders receive supervision proportionate to assessed need, while high-risk offenders are subject to more restrictive measures under reviewable legal standards. Such a system is demanding to build, but it is more defensible than permanent public exposure for everyone (see Table 1).

8. Why the Gaps Are So Large

Why do the seven jurisdictions differ so sharply? The first explanation is gender inequality itself. Jurisdictions with stronger women’s political participation and broader institutional concern for gender violence appear more likely to move from outrage to durable infrastructure. The comparison in the underlying manuscript associated stronger legislative and treatment systems with higher levels of gender equality and women’s legislative representation. This claim should not be overstated into a deterministic rule, but the pattern is difficult to ignore.

The second explanation is administrative capacity. Building a real sex offender governance system requires far more than passing a law. It requires validated tools, trained assessors, probationary or community-supervision infrastructure, treatment professionals, court review procedures, and data collection. Wealthier and more administratively consolidated jurisdictions are better positioned to create these layers. This helps explain why India and Indonesia have struggled despite intense public concern over sexual violence.

The third explanation is the politics of visibility. Severe sanctions and public notification are politically attractive because they are easy to communicate. Treatment quality, actuarial validity, and community case review are far less visible. Yet the latter may matter more for prevention. Jurisdictions can therefore become trapped in a politics of symbolic severity, where the most dramatic reform is rewarded even if it adds little long-term protection.

The fourth explanation is research culture. Systems improve when they can learn. The United Kingdom discontinued a flagship program after troubling evaluation. Taiwan built on local research by Lin and Dong (2005) and later county-level tracking. Canada constitutionalized proportionality questions rather than insulating policy from review. By contrast, where research traditions and evaluation systems are weak, policy may remain event driven and reactive.

9. Taiwan as a Distinctive Policy Case

Taiwan deserves separate discussion because it condenses several themes of this article. First, reform was politically driven by gender violence and feminist advocacy. Second, the reform did not stop at denunciation but moved toward institutionalization: local committees, treatment mandates, risk classification, reporting requirements, and reviewable high-risk confinement. Third, Taiwan adapted foreign tools while also producing local research, including the TSOSRAS and local recidivism studies.

This does not mean Taiwan has solved the problem of sexual violence. No jurisdiction has. But Taiwan demonstrates that a middle-sized jurisdiction can build a governance system that is neither purely American nor purely European. It combines mandatory treatment and intensive supervision with constrained disclosure rather than full public registry. It also shows that local innovation matters.

In Taiwan, Lin and Dong (2004) found that the recidivism rate for sex offenders released region-wide in 1994, 1995, and 1996 with the prison treatment beginning in 1994 after a college girl was sexually murdered by two released sex offenders, tracked over seven years, was 11.3% (see Table 2). The recidivism rate for sex offenders released region-wide in 1997, 1998, and 1999 with the community treatment beginning in 1997 after a famous feminist was sexually murdered, tracked for seven years, had decreased to 6.7% (see Table 3). The two cohorts show a substantially lower observed seven-year recidivism rate in the later period. Because the cohorts differ by release period and may also differ in supervision practices, offender composition, reporting, and treatment exposure, this finding should be interpreted as consistent with improved governance rather than as conclusive proof of a single program effect.

Hsu et al. (2023) and Lin et al. (2024) tracked recidivism ratio in a four-year period (2019 to 2022), and the recidivism ratio was the number of recidivism divided by the number of people who were required to report to treatment. Chiayi County was found to have zero recidivism rate in both periods and was the only county out of 20 counties on the island (Table 4). In Chiayi County, a subset of higher-need cases received intensive clinical intervention from a practitioner involved in the development and discussion of the Taiwan treatment model. This professional involvement provides valuable practice-based insight, but the county-level results should still be interpreted as observational rather than causal.

Lin and Chen (2023) and Lin et al. (2024) suggest that treatment development in Taiwan has not been limited to copying Western curricula; it has involved attempts to integrate trauma, self-regulation on learning and thought, and nutritional dimensions in ways tailored to local practice. Participants with adverse childhood or adulthood experiences reportedly described reductions in trauma-related symptoms after EFT-based exercises. Existing evidence on EFT may support its relevance to trauma symptoms, but further research is needed to determine whether this component contributes to reduced sexual recidivism. Participants also reported improved emotional regulation after nutrition-oriented intervention. This component is best presented as an exploratory and clinically suggestive element of the program, rather than as direct evidence of reduced sexual reoffending..

From an international policy perspective, Taiwan’s importance lies less in any single program element than in the architecture of continuity. Prison treatment is connected to release supervision. Static and dynamic assessment support case differentiation. Noncompliance has legal consequences. High-risk cases can be escalated. These design features are precisely what India and Indonesia currently lack. For that reason, Taiwan may be more relevant to reforming jurisdictions than the most visible aspects of the U.S. model.

The Taiwanese case also illustrates a deeper feminist point. Women’s safety is advanced not only when survivors receive services, but also when the state treats repeat sexual offending as preventable through structured intervention. This is not a departure from feminist goals. It is an extension of them into the post-conviction field.

10. Policy Implications

Several implications follow from this comparison. First, jurisdictions should stop treating sentencing severity as the main indicator of commitment to women’s safety. Severity may be justified in serious cases, but it is not a governance system. A credible system requires risk classification, treatment access, multi-agency coordination, and post-release continuity.

Second, jurisdiction(s), with thin infrastructure should prioritize validated assessment before expanding public notification. Publicly naming offenders without the ability to classify, monitor, and treat them is administratively shallow and normatively dangerous. The more urgent need is to identify which offenders present what kind of risk and what intervention they need.

Third, treatment should be mandated for appropriate cases both in prison and in the community. The prison-to-community gap is one of the most dangerous transition points in sexual-offense policy. Treatment completion in custody should not mean the end of intervention. Community follow-up, dynamic reassessment, and compliance review are indispensable.

Fourth, very high-risk post-sentence confinement should be narrowly defined, legally reviewable, and connected to actual treatment. Jurisdictions may differ in whether they use civil commitment, dangerous offender designations, or preventive detention, but whatever the legal vehicle, the safeguard principles should be similar: clear criteria, professional evidence, periodic review, and proportionality.

Fifth, low- and middle-income jurisdictions should pursue international cooperation and pilot models rather than wholesale transplantation. India and Indonesia do not need an exact copy of the U.S., U.K., or Taiwan. They need locally workable assessment tools, trained personnel, and demonstrable pilot programs in selected prisons and communities. This is where comparative research can have concrete value.

Sixth, prevention must extend beyond known offenders. Feminist scholarship is correct that school, family, and community norms matter. But primary prevention and offender governance should reinforce each other rather than compete for attention. States need both social norm change and effective management of convicted offenders.

Finally, the comparative experience of Taiwan, India, and Indonesia underscores the necessity of an Asian criminological lens. For jurisdictions such as India and Indonesia, the challenge lies in moving beyond the reactive politics of symbolic severity toward the development of 'institutional capacity' (Liu, 2017). Rather than adopting wholesale penal measures from the Global North, these jurisdictions may benefit from an incremental approach that prioritizes locally validated assessment tools and professional training—a strategy that aligns with the Asian criminological emphasis on 'locally grounded' practice. This suggests that the future of sex offender governance in Asia depends on the ability to synthesize global criminological knowledge with local institutional realities, thereby creating models that are both scientifically rigorous and culturally responsive (Liu, 2021).

11. Conclusion

This article has argued that sex offender governance should be understood as part of the state response to gender inequality and sexual violence. Across seven jurisdictions, the strongest systems are not the loudest or the most punitive. They are the ones that connect law, structured risk assessment, treatment, supervision, and high-risk review into a durable architecture of prevention.

The United States shows the power and the danger of public-facing control: advanced assessment and confinement capacity combined with collateral harms from broad notification. The United Kingdom and Canada offer more restrained disclosure and stronger institutional filtering. Germany demonstrates that privacy and serious risk management can coexist. India and Indonesia reveal the limitations of punitive visibility without assessment and treatment infrastructure. Taiwan stands out as a jurisdiction in which feminist mobilization, codified reform, local research, and mandated intervention were combined in a comparatively coherent way.

The broader lesson is straightforward. Women’s safety is not enhanced by punishment rhetoric alone. It is enhanced when states build evidence-based, reviewable, and gender-responsive systems capable of reducing reoffending over time. Comparative policy should move away from asking which jurisdiction is most punitive and toward asking which legal system has developed reviewable, risk-based, and institutionally coherent mechanisms for reducing sexual reoffending.