Remand in custody: critical factors and key issues

Foreword | Between 1984 and 2004 the proportion of remanded prisoners rose from 12 to 20 percent of the total prisoner population, and the rate of prisoners remanded into custody tripled. In an attempt to identify the factors associated with high and low remand rates the researchers undertook a detailed study of Victoria (which has comparatively low remand rates) and South Australia (which has comparatively high remand rates). Factors associated with increased remand rates included increasing levels of drug and mental health issues, and the informal and formal rules that influence police, police custody sergeants and court bail authorities in their decision to grant bail. Factors associated with lower remand rates included 'enhanced police accountability for bail refusal, improved feedback loops between courts and police, higher transaction costs for custodial remand, and longer bail hearings'. The authors have concluded that the key to good practice in bail decision-making is to ensure that pre-court and non-judicial processes are given due consideration, and they point to the need for enhanced performance monitoring, data collection and research.

Toni Makkai


Remanding a person in custody is a serious matter. Remandees awaiting trial enjoy a presumption of innocence, yet they remain incarcerated, often for months at a time. With almost 5000 Australian prisoners currently in custody because they have been refused bail, there has been a renewed focus on custodial remand policies and procedures by academics and practitioners.

This research task was threefold:

  • to search for critical factors that determine and affect the rates at which people enter Australian corrections systems as remandees
  • to study the effects of any remand strategy on remandees and the wider justice environment, and
  • to consider principles of good practice and policy implications arising from that consideration (King, Bamford & Sarre 2006).

The study included:

  • a literature review
  • a statistical analysis of correctional data on defendants remanded into custody over a three year period (2001-03) in Victoria and South Australia
  • qualitative and quantitative analyses of the effects of custodial remand on justice outcomes
  • a series of courtroom observations
  • a case file study of changes in bail status, and
  • interviews with key decision-makers concerning policy implications.

This paper contains a summary of its conclusions.

The remand in custody population

Of the 24,171 persons in custody in Australia at June 2004, 19,231 had been sentenced and 4935 had been remanded in custody awaiting trial or sentence (ABS 2004b). The proportion of prisoners held in custody on remand in Australia (20.4%) is similar to rates in comparable countries such as England and Wales (16.9%), New Zealand (18.3%), the USA (20%), Canada (21.1%) and Germany (21.2%: ICPS).

Like sentenced prisoners, remandees are overwhelmingly male. They are also relatively young, although, according to this research, the average age of remandees in Australia is rising at a faster rate than the average age of the sentenced prisoner population. The distribution of charge types for which remandees are in custody is generally similar to the distribution of offences in the sentenced prisoner population.

In remand populations, females are more highly represented than in sentenced populations. Nationally, approximately 25 percent of female prisoners are remandees, compared with 19 percent of male prisoners (ABS 2004a). While Indigenous prisoners make up about 20 percent of the prison population and also about 20 percent of the remand population (ABS 2004b), there are significant regional differences. In Victoria, for example, around 4.5 percent of all prisoners are Indigenous, compared with around 20 percent of remandees. In South Australia, Indigenous prisoners comprise about 17 percent of the total prisoner population, but between 35 percent and 40 percent of remandees.

According to evidence from the United Kingdom, remandees are more likely than sentenced prisoners to be homeless, unemployed or have some form of mental disorder (Morgan & Henderson 1998). Remandees are also more likely than sentenced prisoners to die in custody, according to recent data released as part of the ongoing Deaths in Custody study. The death in custody rate of unsentenced prisoners, while trending slightly downwards, remains consistently higher than the rate of deaths of sentenced prisoners. In 2004, there were 3 deaths per 1000 unsentenced prisoners compared with 1.2 deaths per 1000 sentenced prisoners (Joudo & Veld 2005).

The total number of prisoners in Australia has increased by around 20 percent since 1995, but remandee numbers have jumped almost 150 percent over the same period, from 1999 remand prisoners at 30 June 1995 to 4934 remand prisoners at 30 June 2004 (ABS 2004b). Remandee numbers have increased in the states and territories by between 50 and 270 percent. Numbers continue to rise substantially in some jurisdictions (New South Wales and Queensland) although in South Australia, Western Australia, Tasmania, the Northern Territory and the Australian Capital Territory, the numbers appear to have remained relatively steady for the past four years.

Factors critical to remand rates

The decision to remand a person in custody emerges from a complex array of legal and social dynamics that vary between jurisdictions and over time. It is difficult to isolate precisely what causes remand populations to change and what factors influence the rise and fall of remand rates. Building on earlier work (Bamford, King & Sarre 1999) however, it was determined that changes in remandee numbers are usually the result of one or more of the following factors:

  • changes in the volume of persons appearing before the courts, as a result of changes in crime rates, apprehensions or charging practices
  • changes in bail practices and policies that affect the probability of bail being granted or denied
  • changes in defendant characteristics that make custodial remand more likely
  • court delays, which affect the average period spent on remand.

Research on remand patterns in Australia has been principally concerned with the first category of explanations. For example, Fitzgerald's (2000) research indicated that the principal cause of the increase in the New South Wales remand population between 1994 and 2000 was the rise in the number of persons charged with offences that have high bail refusal rates, such as robbery, and break and enter. Similarly, the South Australian Office of Crime Statistics (2002) found that the increase in remand numbers after 1999 could have been influenced by increased apprehensions of offenders for crimes of burglary, motor vehicle theft, and major drug offences.

Other research has gone beyond crime rates, apprehensions and charging practices to look at broader political factors as well. In Victoria, one study concluded that the rise in sentenced prisoner numbers was the result of shifts in criminal justice policy that increased the number of persons imprisoned much faster than the rate of increase in the general population (Freiberg & Ross 1999). The authors speculated that the rise in the Victorian remand rate proceeded from the same cause.

This research indicates that defendant characteristics are changing, and that these changes are influencing remand in custody trends. For example, there appear to be increasing levels of drug and mental health issues affecting those in custodial remand populations. Victorian data collected as part of this study indicate that, over the three years studied, remandees demonstrated statistically significant declines in seriousness of criminal history at the same time as there were indications of increasingly severe drug and alcohol abuse and mental health problems. Key persons interviewed for this study nominated these changes in defendant characteristics as one of the significant reasons for rises in custodial remand numbers in recent years.

Corrections data on Indigenous remandees, however, highlight the fact that personal characteristics on their own are an insufficient explanation for remand rate differences. In South Australia, over the three years 2000-03, there were 1782 remand receptions of Indigenous Australians from an Indigenous population of 23,425 (ABS 2002a), whereas in Victoria there were only 450 remand receptions in the same period from a very similar sized Indigenous population (25,078: ABS 2002b).

Finally, there is always a suspicion that delays in prosecutorial processes or court delays might affect upward trends in remand rates in a significant way. In this analysis, however, time on remand, while important in influencing overall remand rates, was found to be an unreliable predictor of rates. For example, South Australia has, and has consistently had, one of the shortest average times on remand. Over the past three years, the mean time of remand in South Australia was around 56 days whereas in Victoria it was over 85 days. Yet Victoria has a remand rate about one-third that found in South Australia.


Figure 1: Australian remand rates, 1998-2004 [see attached PDF for graph]



The search for critical factors: comparing South Australia with Victoria

Since 2001, South Australia and Victoria have had, respectively, the highest and the lowest remand rates of the Australian states. The remand rate in South Australia continued to climb steeply from 1998 (22.7 per 100,000 adult population) to 2002 (42.2), while the Victorian rate climbed only from 12.2 to 16.5 over the same period. In 2003, these rates converged slightly, with South Australia coming down to 39.4 and Victoria rising to 18.2. By 2004, this had shifted again, with the South Australian rate rising to 43.5 and the Victorian rate dropping to 15.9 (ABS 2004b).

South Australia also has a much higher percentage than Victoria of remand prisoners as a proportion of all prisoners. Over one-third of all prisoners in South Australia are remandees, compared with less than a fifth of all prisoners in Victoria.

These differences pointed to the possibility that these two jurisdictions might reveal useful information about what it is that drives remand rates up and down, and whether there is anything that policy-makers can do to influence those movements directly or indirectly.

Analysis of Victorian and South Australian processes and policies, along with statistical data and interviews with key personnel, indicate that the critical factors are to be found in the personal characteristics (discussed earlier) that draw certain individuals to the attention of police and hence into the criminal justice system, but more significantly, in the decision-making processes once those individuals have been selected. That is, the policies and practices of police, police custody sergeants and court bail authorities in relation to bail (and the formal and informal rules that empower and constrain them) are crucial to the determination of remand trends. Four critical factors are: bail legislation, accountability for decision-making, agency procedures, and an emphasis upon therapeutic justice.

Differences in bail legislation

Bail decision-makers rely on bail legislation to guide their decisions. Legislation in different jurisdictions is similar but not identical. The Victorian Bail Act 1977 has a number of significant differences from the South Australian Bail Act 1985. The Victorian Act distinguishes between grounds for remand in custody and the information to be used in determining whether those grounds exist. The South Australian Act is less constraining, with grounds for bail merged with information to be used in determining the granting of bail. The Victorian Act also contains reverse onus provisions that require defendants in certain circumstances to overcome a presumption against bail, provisions that are currently under investigation by the Victorian Law Reform Commission. The Victorian legislation also provides for immediate review of police bail decisions, either by a court or, if out of court hours, by a Bail Justice. The approach taken by the courts and some police to the reverse onus provisions, and the extensive use of Bail Justices, reflects a culture that either promotes bail or does not discourage the granting of bail.

The accountability of bail authorities and review of remand decisions

There is significantly greater transparency of remand decisions in Victoria than in South Australia. The opportunity for review of police remand decisions is more available in Victoria with its Bail Justice process, compared with the more piecemeal telephone review process in South Australia. Those opposing bail for accused persons in Victoria (operational police) are required to attend the Bail Justice and any subsequent court hearing to give sworn evidence. This ensures that the information upon which a court assesses the risk of a defendant not complying with bail is more likely to be closely scrutinised. Moreover, the increased level of scrutiny of information is reflected in the time taken to consider contested bail applications. The court observation study, conducted between November 2002 and October 2003, showed that the percentage of contested matters is about the same in both jurisdictions (40%), but the median time taken for a contested hearing in South Australia is 5 minutes, while in Victoria it is 18 minutes.

Agency operational procedures

Bail decision-making occurs in a time-pressured context and in accordance with the policy and cultural constraints of the various bail authorities, especially the police. In South Australia custodial remand was closely linked to operational policing objectives and strategies. For example, it is not uncommon to find operational policies that encourage arrest even where a summons could also be appropriate, or that use custodial remand as an incapacitation strategy to achieve crime reduction goals. That same operational ethos was not as apparent in Victoria.

Therapeutic justice and court resources

The research identified a trend in Victoria towards new perceptions of judicial roles that have not been identified in South Australia to the same extent. While both jurisdictions boast a wide variety of diversionary courts, some Victorian magistrates have, in addition, adapted what has been described as the therapeutic jurisprudence or therapeutic justice model to the bail process. Under this model, the court (through the magistrate or judge) explores opportunities to act as a therapeutic agent using mental health and related disciplines (Birgden 2004). This trend appears to be, in part, a response to the changing characteristics of defendants appearing in the remand system, and is designed to focus on their needs. The therapeutic emphasis has enabled Victorian courts to attract a greater range of resources than those available in South Australia to help defendants with alternatives to remand in custody.

What are the effects of custodial remand?

The study was also designed to identify the effects of any remand strategy on remandees and the wider justice environment. There are arguably, three broad goals to be achieved in bail decision-making. Bail authorities will make their decisions with the following aims in mind:

  • to ensure the integrity and credibility of the justice system
  • to protect the community, and
  • to assist in the care and protection of the rights of the defendant.

There are limited data collected by agencies that allow the measurement of the extent to which current bail and remand practices achieve these aims. The availability of data on issues such as failure to appear, the reasons for failure to appear, offending on bail, interference with witnesses and victims is so poor that the effectiveness of the remand in custody system cannot be analysed with any degree of accuracy.

There is a clear tension between the goals of custodial remand. It is arguable that, in the past 40 years, there have been legislative and operational policy changes that have shifted the emphasis away from the first and third of these goals in order to elevate the importance of the second.

Studying the effect of having been remanded in custody on the final sentence of accused persons subsequently found guilty fell beyond the scope of this research. However, custodial remandees who are not subsequently convicted or imprisoned were examined briefly. If custodial remandees typically receive sentences of imprisonment, this may indicate that the decision to remand anticipates (or perhaps influences) subsequent sentencing decisions. Conversely, a low remand-to-sentence rate might suggest that bail authorities are too often wrong in their assessments, or that custodial remand is being used inappropriately as a substitute punishment. This research could not make a judgement about the relationship between custodial remand and subsequent sentence. Nevertheless, in South Australia only about 30 percent of those remanded in custody serve additional time in prison following sentencing, whereas in Victoria, with its lower remand rate, about 60 percent of remandees spend additional time in custody after sentence.

Another part of the study reviewed the significant number of defendants each year who fail to attend court as required. Explanations from the qualitative research suggest that this is largely due to poor self-management skills, or a lack of respect for the court's authority, rather than deliberate attempts to abscond, although technically all attract the same outcome.

Good practice in bail and remand

The final purpose of this research was to identify desirable good practice characteristics of a remand system. A preferred system of bail and custodial remand will establish, develop and maintain:

  • statements of principles, objectives and criteria guiding bail decision-making
  • clear definitions of the roles of bail authorities and their responsibilities
  • adequate resourcing
  • quality assurance mechanisms.

Within both South Australia and Victoria differences in interpretation of the principles, objectives and criteria guiding bail decision-making were found. For example, while both jurisdictions appear to use the statutory criteria for custodial remand in similar ways, there is evidence of cultural differences underpinning practice and policy. The Victorian process appears to be willing to go beyond the legislation in certain areas and to take a pro-bail approach to the reverse onus provisions in some situations, and to facilitate a review of police decisions to refuse bail.

Good practice in remand requires clearly defined roles for decision-makers. The judicial power of bail was initially exercised only by the sheriffs in medieval England, but it has now also become a power exercised by police. It became clear from this study that it is very easy for police to merge their role as bail decision-makers with their role as crime preventers and crime investigators, and that custodial remand can be employed as a tool to achieve other police objectives such as crime reductions. Clearer definitions would remove these ambiguities.

Resourcing is crucial to good practice. Resources are required to ensure that support services that have been established are maintained, such as the South Australian Bail Advocacy Unit, and the Victorian Bail Advocacy and Support Programme which is available in some parts of Melbourne. Such services are designed to focus on the issues that gave rise to the risks on which decisions to remand are based.

Adequate resourcing is also required to address appropriately the needs of victims. Concerns about victims and their interests were clearly evident in both jurisdictions. Some bail decision-makers indicated that they take special care when dealing with bail applications from defendants charged with domestic violence offences since the risks and consequences of re-offending were particularly significant in this situation.

However, in other offence situations, there appears to be very little by way of formalised processes in either jurisdiction that allow victims to either have a voice or receive information. Analysis of court files indicated that the remand period might involve several changes of bail status, and victims' representatives reported that their clients were shocked on occasion to come face to face in public with a defendant whom they believed to be in prison. In today's justice policy context, greater sensitivity to the rights and needs of victims is expected by the community and within the justice system itself. Resourcing of victim support services is axiomatic in such an environment.

To facilitate quality assurance, reliable data must be collected and made publicly available. To understand the contribution of each set of decision-makers to remand rates requires systematic, comparable and accessible data on remand hearings and remand outcomes. The lack of data currently available may explain the significant number of key interviewees who had little or no idea whether remand rates were changing and if so, in what direction. Better statistical services are required within and between jurisdictions, using common terms and collection and collation processes that can allow data to be compared and trends determined.

Good practice would be enhanced by the inclusion of performance indicators relating to bail decisions and processes in agency reporting. Inter- and intra-agency liaison between bail decision-makers encourages the identification and addressing of problems as they arise, along with the development of innovative practices to meet the changing context in which decision-making occurs.

Policy implications

This research identified that lower remand rates are associated with enhanced police accountability for bail refusal, improved feedback loops between courts and police, higher transaction costs for custodial remand, and longer bail hearings. In the current political environment, the high level of interest in law and order policy has made remand decision-making a matter of public interest and comment. Community interest in justice processes is an important element of sustained legitimacy of the justice system. However, positioning bail decision-making as an indicator of the strength or weakness of any criminal justice policy, as often happens in a politically charged debate, moves remand from its role as a tool to ensure effective court administration to something that it was never designed to be.

The research also suggests that many defendants' failures to attend court as required are simply the result of disorganised lives or indifferent attitudes. Alternatives to custody for this group of defendants range from increased social support and case worker intervention to the use of surveillance technologies, for example, electronic monitoring of home detention defendants under curfew. The report also identifies a number of other specific examples where efficiency could be improved, for example, eliminating delays in forensic analyses and court listings (King, Bamford & Sarre 2006).


This research indicates that remand rates, resulting from the interaction of personal characteristics of defendants and the policies and practices of decision-makers, can be influenced over time by strategic provision of resources and a change in a jurisdiction's philosophical approach to remand.

Furthermore, it suggests that the key to understanding the remand in custody process is for researchers to move outside the courtroom and to focus on issues that arise prior to the judicial hearing. This is not to say that the courts have no influence. Several magistrates argued that an important future direction in remand decision-making would be for them to be more active in claiming the time needed to give more detailed consideration to the bail/remand decision and to influence the development of more creative options for achieving the objectives of custodial remand. However, a focus on, and analyses of, the decisions made by the non-judicial participants in the process, especially police decision-making and information they provide to the courts, is the key to isolating the critical factors affecting remand in custody trends. With this in mind, policy-makers should be far better equipped to be able to drive custodial remand policy in a manner that is consistent, efficient and fair.


This project is supported by a grant from the Criminology Research Council. The views expressed are the responsibility of the authors and are not necessarily those of the Council.


All URLs were correct at April 2006

  • Australian Bureau of Statistics (ABS) 2004a. Corrective services Australia, December quarter 2003. ABS cat. no. 4512.0. Canberra: Australian Bureau of Statistics
  • Australian Bureau of Statistics (ABS) 2004b. Prisoners in Australia, 30 June 2003. ABS cat. no. 4517.0. Canberra: Australian Bureau of Statistics
  • Australian Bureau of Statistics (ABS) 2002a. Census of population and housing: selected social and housing characteristics for statistical and local areas, South Australia 2001. ABS cat. no. 2015.4. Canberra: Australian Bureau of Statistics
  • Australian Bureau of Statistics (ABS) 2002b. Census of population and housing: selected social and housing characteristics for statistical and local areas, Victoria 2001. ABS cat. no. 2015.4. Canberra: Australian Bureau of Statistics
  • Bamford D, King S & Sarre R 1999. Factors affecting remand in custody: a study of bail practices in Victoria, South Australian and Western Australia. Research and public policy series no. 23
  • Birgden A 2004. Therapeutic jurisprudence: the role of forensic psychology. In Sarre R & Tomaino J (eds) Key issues in criminal justice. Adelaide: Australian Humanities Press: 166-191
  • Fitzgerald J 2000. Increases in the NSW remand populationCrime and justice statistics: bureau brief November
  • Freiberg A & Ross C 1999. Sentencing reform and penal change. Sydney: Federation Press
  • International Centre for Prison Studies (ICPS). World prison brief
  • Joudo J & Veld M 2005. Deaths in custody in Australia: National Deaths in Custody Program annual report 2004. Technical and background paper series no. 19. Canberra: Australian Institute of Criminology
  • King S, Bamford D & Sarre R 2006. Factors that influence remand in custody. (forthcoming)
  • Morgan P M & Henderson P F 1998. Remand decisions and offending on bail: evaluation of the Bail Process Project. Home Office Research Study 184. London: Home Office
  • South Australian Office of Crime Statistics 2002. Increasing remand rates in South Australia: some explanatory factors. Adelaide: South Australian Office of Crime Statistics

About the authors

Rick Sarre is at the School of Commerce, University of South Australia.

Sue King is at the School of Social Work and Social Policy, University of South Australia.

David Bamford is at the School of Law, Flinders University.