Alternatives to Custody and Restorative Justice

Non-custodial sentences operate on a spectrum from the least to the most intensive community supervision. At the lower end, a conditional discharge allows an offender to avoid punishment if they commit no further offences within a set period. A fine imposes a financial penalty proportionate to the offence and, in some jurisdictions, to the offender's income. Suspended sentences impose a custodial term that is not served unless the offender reoffends within a defined period.

Community service orders, known in England and Wales as unpaid work requirements and in the United States as community service, require the offender to complete a set number of hours of supervised work for public benefit. Probation involves regular reporting to a supervising officer, compliance with conditions (such as not leaving the jurisdiction, attending treatment, or maintaining employment), and can include programme requirements targeting criminogenic needs. These two elements are often combined into a single community order.

Electronic monitoring (EM) entered mainstream use in the United States in the 1980s and has since been adopted in England and Wales, Sweden, the Netherlands, Germany, Australia, and India, among others. Ankle tags enforcing a curfew can substitute for pre-trial detention or short custodial sentences. GPS variants track location in real time and can enforce exclusion zones (preventing an offender from approaching a victim's home, for example). EM does not address criminogenic needs but reduces the costs and harms of incarceration for lower-risk offenders and is a condition that can be paired with supervision and treatment.

Diversion programmes intercept offenders before or during prosecution and redirect them into an alternative process. They are most common for young people, first-time offenders, and those whose offending is linked to specific needs such as substance misuse or mental illness. The rationale is that formal prosecution and a criminal record can damage employment prospects, housing, and social identity in ways that increase long-term offending risk, particularly for people whose original offence was minor. Avoiding conviction removes the label and its consequences.

Drug courts are the most studied specialist diversion model. The first drug court opened in Miami, Florida in 1989. By 2023 there were over 3,000 drug courts operating across the United States, and similar models exist in Canada, Australia, England and Wales, Scotland, Ireland, and several European countries. The model requires the offender to plead guilty (or admit the facts) and then enter a structured treatment programme under the active supervision of a judge who reviews progress at regular hearings, typically monthly. Rewards for compliance (praise, reduced reporting requirements) and sanctions for non-compliance (brief custody, increased reporting) are used throughout. Participants who complete the programme may have the charge withdrawn or the sentence reduced.

Mental health courts and homelessness courts follow a similar problem-solving model, diverting defendants with mental illness or housing instability into coordinated support rather than prosecution. Veterans courts, common in the United States, address the specific criminogenic needs of military veterans. All of these specialist courts share a logic: the criminal justice response should match the underlying driver of offending, not simply respond to the offence category.

The theoretical foundations of restorative justice are usually traced to Howard Zehr's 1990 book Changing Lenses, in which he contrasted retributive justice (which asks: what rule was broken, who broke it, and what punishment do they deserve?) with restorative justice (which asks: who was harmed, what are their needs, and whose obligation is it to address them?). John Braithwaite's theory of reintegrative shaming, developed in Crime, Shame and Reintegration (1989), provided a complementary framework: shame that stigmatises and excludes the offender increases reoffending, while shame that is followed by reintegration into the community reduces it. Restorative processes are designed to produce the latter.

Several distinct models have been developed. Victim-offender mediation (VOM) is the oldest formal model: a trained mediator facilitates a meeting or shuttle dialogue between victim and offender, resulting in a written agreement. VOM is used in adult criminal proceedings in Germany, Norway, Finland, Austria, and many US states. Family Group Conferences, originating in New Zealand, bring a wider group (family members, support persons, and sometimes a police representative) into the discussion. Sentencing circles, used in Indigenous communities in Canada and some Australian jurisdictions, involve a wider community group sitting with the court to determine an appropriate community-led response.

Restorative practices have also been applied in settings outside criminal prosecution. Schools use restorative circles to address bullying and conflict as an alternative to suspension. Prisons in several countries have developed restorative programmes between incarcerated people and victims who volunteer to participate. At the post-conflict scale, truth and reconciliation commissions in South Africa, Rwanda, Canada (the Truth and Reconciliation Commission on Indian Residential Schools), and elsewhere draw on restorative principles to address mass harm that the ordinary criminal justice system cannot adequately process.

Youth justice systems have been the primary site of restorative justice institutionalisation. New Zealand's Children, Young Persons, and Their Families Act 1989 made the Family Group Conference the standard response to offending by under-17s (other than murder and manslaughter). A trained youth justice coordinator convenes the conference, which includes the young person, their family (broadly defined, consistent with Maori concepts of whanau), the victim (or a victim representative), and a police youth aid officer. The group discusses the offending, its impact on the victim, and what the young person should do to repair the harm and avoid reoffending. The agreed plan, which can include community service, apology, financial reparation, and educational or therapeutic conditions, is filed with the Youth Court.

England and Wales introduced youth offending teams (YOTs) and restorative disposals through the Crime and Disorder Act 1998 and subsequent legislation. Restorative cautions and community resolutions can divert young people from prosecution entirely for minor first-time offences. The referral order, mandatory for most first-time offenders who plead guilty in the Youth Court, involves a youth offender panel with community volunteers, the young person, and their parents agreeing a contract that can include restorative elements. Referral orders have been evaluated positively on victim satisfaction and reoffending compared to conventional sentences.

Australia operates a mix of state-level models. Queensland and South Australia use conferencing models. South Australia's Juvenile Justice Act 1993 and subsequent Youth Court Act provisions allow conferencing for most juvenile offences. Victoria and New South Wales retain more traditional court-based responses with restorative elements available as diversionary tools. Canada's Youth Criminal Justice Act 2002 explicitly promotes extrajudicial measures and conferences as the preferred response to minor youth offending, consistent with the Act's stated preference for the least possible intervention consistent with public protection.

Evidence on whether alternatives to custody reduce reoffending is substantial but uneven. Drug court evaluations in the United States are the most methodologically rigorous body of research in this field. Meta-analyses by David Wilson, Ojmarrh Mitchell, and others consistently find reductions in recidivism of between 8 and 14 percentage points compared to standard prosecution. The effect is larger for high-risk participants and for courts that follow the drug court model with fidelity. A 2012 Campbell Collaboration systematic review found similar effects internationally.

Restorative justice has been evaluated through several randomised controlled trials, most notably the Reintegrative Shaming Experiments (RISE) conducted in Canberra, Australia in the 1990s by Lawrence Sherman and Heather Strang. RISE randomly assigned offenders to court or conference and found lower reoffending in conference groups for violent offences, though not consistently for property offending. A series of studies by the same team in England (funded by the UK Home Office and then the Ministry of Justice) found restorative justice reduced reoffending by around 14% compared to controls and saved approximately two pounds in criminal justice costs for every one pound spent on conferencing. Victim satisfaction in restorative processes is consistently higher than in conventional proceedings in most studies.

Community supervision outperforms short custodial sentences on reoffending in multiple European comparisons. A 2017 study from England and Wales (Ministry of Justice) found that offenders sentenced to custody of less than 12 months reoffended at a rate of around 64% within a year of release, compared to around 33% for those given community orders. However, custodial and community sentences are not randomly assigned; selection effects make direct comparison difficult. The most careful analyses, including those using natural experiments created by sentencing policy changes, generally support the finding that community sentences are not worse and are often better than short custody for lower-to-moderate risk offenders.

Alternatives to custody face several persistent criticisms. Net-widening is the most frequently cited: if diversion programmes draw in people who would otherwise have received no formal response at all (rather than those who would have gone to prison), they expand the reach of state control rather than reducing it. Stanley Cohen's classic critique in Visions of Social Control (1985) argued that community corrections could become a system of dispersed surveillance rather than a genuine alternative to imprisonment. Empirical studies have found net-widening effects in some programmes and not in others, depending on targeting and eligibility criteria.

Restorative justice faces separate concerns. Critics argue it can pressure victims to participate or to forgive, and that victim-centred rhetoric can mask coercive aspects of the process. Feminist criminologists, including Kathleen Daly, have raised questions about whether conferences adequately address the power imbalances in gendered violence and whether informal processes provide sufficient accountability for serious sexual and domestic offending. Most restorative justice systems now exclude the most serious violent and sexual offences from mandatory conferencing, though some offer voluntary processes as a supplement to conventional prosecution in those cases.

There is also a consistency problem. Community sentences and restorative outcomes vary considerably between cases, courts, and regions. Two offenders convicted of similar offences may receive very different community requirements depending on the local probation service's capacity or the conference facilitator's skill. Proportionality and fairness require some baseline consistency, which community-based systems can struggle to maintain at scale. Countries with large geographic and resource disparities, including India, Brazil, and South Africa, face particular challenges in extending alternative measures equitably beyond urban centres.