Download details

No bail, more jail? Breaking the nexus between community protection and escalating pre-trial detention

Glossary

Bail: the conditional release from custody of a person arrested and charged with a criminal offence.

Bail applicant: a person who is arrested and charged with an offence and who applies for bail.

Bail conditions: conditions imposed as part of the grant of bail.

Bail hearings: these involvea one-step or two-step process.

For less serious offences (not contained in Schedule 1 or 2 of the Bail Act), a one-step process applies; the prosecution must satisfy the decision-maker that the person applying for bail presents an unacceptable risk of endangering the safety or welfare of any person, committing an offence while on bail, interfering with a witness or obstructing the course of justice, or not attending court for the hearing of their case.

For more serious offences (mainly contained in Schedule 1 or 2 of the Bail Act), a two-step (reverse onus plus consideration of unacceptable risk) process applies. Depending on the offence with which they are charged, the person applying for bail must satisfy the decision-maker that either exceptional circumstances or a compelling reason exists to grant bail. If the person succeeds in this first step, the prosecution can then still argue at the second step that bail should be denied because the person presents an unacceptable risk of endangering the safety or welfare of any person, committing an offence while on bail, interfering with a witness or obstructing the course of justice, or not attending court for the hearing of their case.

Criminogenic: causing or likely to cause criminal behaviour.

Exceptional circumstances: requires the bail applicant to show exceptional circumstances as to why they should not be detained in custody until trial. Applies only to applicants charged with certain offences.

Remand: describes the situation where a person who is in custody and does not apply for bail, cannot meet a condition of bail or fails in their application for bail, will continue to be held in custody ('on remand') until their trial or until an application for bail is successful.

Reverse onus test: requires the bail applicant to show why they should not remain in custody until trial. There are two reverse onus tests in Victoria: 'compelling reason' (formerly 'show cause') and 'exceptional circumstances.'

Show cause: required the bail applicant to show cause as to why they should not remain in custody until trial. Applied only to applicants charged with certain offences. Now replaced by the 'show compelling reason' test.

Show compelling reason: requires the bail applicant to show compelling reason as to why they should not be detained in custody until trial. Applies only to applicants charged with certain offences. Replaced the 'show cause' test.

Surrounding circumstances: matters that a decision maker must take into account when determining whether a person should be granted bail. These matters include the nature and seriousness of the alleged offending, the strength of the prosecution case, the person's personal circumstances, associations, home environment, background and other matters listed in s 3AAA of the Bail Act 1977 (Vic).

Unacceptable risk: requires the prosecution to show why a bail applicant, if granted bail, may present an unacceptable risk by failing to attend court, committing an offence, endangering the safety or welfare of the community or interfering with justice.

Willie Horton effect: named after Willie Horton, a prisoner who escaped from prison in Massachusetts while on weekend leave. Horton subsequently raped a woman and assaulted her boyfriend. Horton's case was widely discussed in the U.S. presidential campaign in 1988 and used to justify punitive criminal justice policies in political debate.

 

1. Executive summary

Victoria is experiencing an 'incarceration crisis',[footnote 1] caused by an unprecedented growth in prisoner numbers. The key driver of this growth is the increasing number of persons who are denied bail and remanded into custody.[footnote 2] As of 31 May 2019, 38 per cent of adult prisoners in Victoria were being held on remand.[footnote 3]

The increase in the remand population is gendered, with higher rates for women—nearly half the women in Victorian prisons are now being held on remand.[footnote 4] It is also contributing to the incarceration crisis with indigenous offenders, as Aboriginal and Torres Strait Islander peoples are disproportionately represented among those remanded.

Moreover, there are signs that pre-trial detention in Victoria is increasing. In 2017–18, Victoria had the largest increase in prisoners held on remand of any Australian State or Territory. Recent reforms to bail law are likely to further contribute to this trend by further restricting eligibility for bail.

What lies behind the increased number of persons being held on remand, and the changes to bail law? A comparison of bail laws and parliamentary debates in Victoria in 1977 (when the Bail Act was introduced) and 2017–18 (the most recent substantive reforms) reveals escalating concern about community protection; that is, the need to protect the community by remanding into custody those persons deemed to present a risk of committing offences if released on bail.

However, the shift towards greater community protection has come at a cost. There are significant human, fiscal and legal consequences associated with pre-trial detention. The current framework for risk assessment is resulting in over-incarceration. The present legal framework for assessing risk, involving a two-step process for serious offences, is unduly complex. Troublingly, there is also evidence that the long-term consequences of remanding people into custody may actually decrease community safety, as even relatively short periods of incarceration (such as that experienced by many of those held on remand) are associated with higher rates of subsequent criminal offending.

While the desire to protect the community from those who are likely to commit serious offences if released on bail is grounded in legitimate political and community interests, it is necessary to simplify current law and to utilise modern research to target only those individuals who are most likely to commit serious violent offences if released on bail. Adopting a single 'unacceptable risk' test would simplify bail decision-making. Assessment instruments, such as those reviewed by the non-profit MacArthur Foundation in the United States, could assist bail decision-makers to remand into custody only those who present a serious risk of violent offending if released on bail.[footnote 5] Adopting these recommendations could satisfy the goal of community protection, while minimising pre-trial detention.

2. Introduction: Bail and the incarceration crisis

The increasing number of persons detained in prisons in Victoria is a matter of concern. What is often not recognised is that the growth is primarily due to the number of persons who are denied bail and detained in prison ('remanded into custody').[footnote 6]

The 'incarceration crisis' in Victoria is intimately related to changes in bail laws and associated practices.[footnote 7]

The changing nature of bail considerations

Traditionally, the key concern when considering whether to grant an application by a person seeking bail was whether they would attend court for the hearing of their matter; indeed, English legislation in the first part of the nineteenth century made this the sole risk criterion.[footnote 8] This continued to be the primary consideration in Australia until the 1970s and 1980s. However, new statutory laws introduced at that time also included three other risk factors to be taken into consideration: whether there was an 'unacceptable risk' that the person would, if released on bail, commit an offence; endanger the safety or welfare of members of the public; or interfere with witnesses or otherwise obstruct the course of justice.[footnote 9]

Concerns about future offending and endangering the safety or welfare of members of the public have now fused into concern about 'community safety' or 'community protection' and have dominated bail discussions in recent years. Driven by high-profile cases of heinous crimes committed by persons on bail,[footnote 10] and community concerns about crime and safety,[footnote 11] bail has become a focus of heightened political concern and restrictive practices. It increasingly functions as a site for crime prevention through preventive detention. As with other points in the criminal justice system,[footnote 12] liberty is increasingly being predicated on risk assessment and bail is viewed as a privilege rather than a right.[footnote 13] The ascendancy of community protection is reflected in recent reviews, law reform and prison numbers.[footnote 14]

Aims of this research

This report investigates the reasons behind the increase in the number of people being denied bail and detained in prison in Victoria. Through a comparison of bail laws and their associated parliamentary debates in Victoria in 1977 and 2017–18 (two periods that correspond to the introduction of the first comprehensive bail statute and the most recent reforms), it analyses how risk has been constructed, charting the rise of community protection as a key concern in bail law and decision-making.

This paper also identifies a narrative shift: from traditional concerns around persons failing to attend their court hearing, to current concerns, which centre upon community protection. It examines how this shift underpins a long-term trend towards more people being denied bail and held in prison.

While recognising that the goal of community protection is grounded in legitimate political and community concerns, the increasing number of people being incarcerated before trial raises concern. In addition to the human, fiscal and legal costs, there is evidence of adverse downstream effects of even short periods of imprisonment (such as that experienced by many people who are denied bail). These matters suggest a need to re-evaluate current approaches to community protection.

The report concludes by recommending strategies for simplifying risk assessment in bail decision-making through the adoption a single test of risk ('unacceptable risk') and improving risk prediction by the use of an appropriate actuarial instrument. These reforms have the potential to protect the community while substantially reducing the number of those held in prison before the hearing of their cases.

3. The human, fiscal and legal costs of pre-trial detention

At 31 May 2019, there were 3,044 persons held in custody in Victorian prisons who had not been sentenced;[footnote 15] this constituted 38 per cent of the Victorian adult prison population.[footnote 16]

These figures represent the culmination of a trend towards increased pre-trial detention that has been occurring in Victoria for nearly two decades, with a notable increase since 2014, as seen in Figure 1 below.

With an estimated current daily cost of nearly one million dollars, detaining these people in prison before the hearing of their cases is costly.[footnote 17] While broadly similar to other jurisdictions in Australia, the position in Victoria contrasts sharply with that in England and Wales, for example, where those held on remand comprise only ten per cent of the adult prison population.[footnote 18]

While all Australian States and Territories are detaining more people pre-trial, in recent years only South Australia and the Australian Capital Territory have experienced higher remand rates than Victoria (see Figure 2).[footnote 19] There is reason to believe that Victoria may soon have the highest proportion of prisoners on remand; in the 12 months preceding June 2018, Victoria had the greatest increase (22 per cent) in these prisoners.[footnote 20]

Notably, the frequency of pre-trial detention is gendered, with higher rates for women. In May 2019, nearly half (48 per cent) of adult women incarcerated in Victoria were being held on remand,[footnote 21] compared with 37 per cent of men[footnote 22] (see Figure 3). Interestingly, just over half of these women did not even apply for bail.[footnote 23] Their most common alleged offences were not violent crimes but property, drug and burglary offences.[footnote 24]

Pre-trial detention also contributes to the incarceration crisis with indigenous offenders. Nearly twenty years ago, the Royal Commission into Aboriginal Deaths in Custody found that more Aboriginal people had died while on remand than had died while serving a sentence.[footnote 25] More recently, the National Congress of Australia's First Peoples identified that bail laws contributed to the over-representation of Aboriginal and Torres Strait Islander people in Australian prisons through the 'overuse and abuse' of remand, particularly for low level, non-violent offences.[footnote 26] It is also notable that indigenous status interacts with gender, with indigenous women being particularly over-represented among those held on remand.[footnote 27]

There are also troublesome legal aspects of this trend. As has been pointed out many times, detaining people pre-trial:

§ challenges the foundational legal right to liberty and the presumption of innocence;

§ imposes significant adverse consequences on those detained: possible loss of employment; separation from families; and a reduced ability to prepare for legal proceedings and exposure to the dangers of a prison environment;[footnote 28] and,

§ presents multiple linked legal disadvantages. Pre-trial detention is associated with an increased likelihood of pleading guilty,[footnote 29] being more likely to be found guilty,[footnote 30] and being given a custodial sentence. While these outcomes could be attributed to the proper operation of a criminal justice system,[footnote 31] there is some evidence that pre-trial detention itself independently contributes to these adverse outcomes—that is, being held on remand itself contributes to the probability that a person will plead guilty, be convicted and be given a sentence of imprisonment.[footnote 32]

A further troublesome finding is that many people who are detained in prison prior to their trial will subsequently be found not guilty or, if convicted, will be given a non-custodial sentence. For instance, a report by the New South Wales Law Reform Commission indicated that, at the Local Court level, eight per cent of those on remand at the time of the hearing of their case were found not guilty and 34 per cent of those found guilty did not receive a custodial sentence. Similarly, another study reported that just over a quarter of persons who had been held in prison on remand subsequently received a community based (non-custodial) order.[footnote 33] These individuals receive no compensation for the time they have spent in prison.[footnote 34]

Taken together, this information indicates that there is a trend towards fewer people being granted bail; consequently, more people are being detained pre-trial and spending time in prison on remand. In essence, our prisons are increasingly being used to detain persons who are legally presumed to be innocent and who will experience multiple disadvantages through their incarceration.[footnote 35]

4. The rise of community protection

The increase in the number of people being denied bail and held on remand in Victoria has been accompanied by significant reforms to bail law that prioritise community protection.

As previously noted, traditionally the key concern when making a decision about bail was whether the person applying for bail would attend court for the hearing of their matter. Prior to the enactment of the Bail Act in Victoria in 1977, it appears that few persons on remand were detained specifically because of concerns about community safety. Information provided to a parliamentary committee in 1974 about persons held on remand in Pentridge Prison revealed that only two of the 38 men who had been refused bail (five per cent) had been detained specifically because they were perceived to be a 'danger to the community'.[footnote 36]

However, statutory formulations of bail law enacted in Victoria and other States and Territories of Australia in the 1970s and 1980s routinely included four risk factors that must be taken into consideration when deciding whether bail should be granted.[footnote 37] Under the Bail Act introduced in Victoria in 1977, when determining bail, a decision-maker was required to consider whether there was an 'unacceptable risk' that the person applying for bail would:

§ attend court for the hearing of their matter;

§ commit an offence while on bail;

§ endanger the safety or welfare of members of the public; or

§ interfere with witnesses or otherwise obstruct the course of justice.[footnote 38]

Concern that a person would commit an offence if released on bail or endanger the safety or welfare of members of the public gradually fused into concern about 'community safety' and the need for 'community protection'. This consideration has become much more prominent in discussions of bail in recent years.[footnote 39] Congruent with broader developments in criminal justice systems that favour a 'tougher' approach on crime, bail has become a site of increased political focus; adopting restrictive, risk-oriented practices. It increasingly functions as a mechanism of crime prevention through preventive detention. Similar to other points in the criminal justice system, liberty is increasingly being predicated on risk assessment. This development is clearly evident in the evolution and amendment of bail law and practice in Victoria.

5. Risk and bail: The Bail Act 1977 (Vic)

The Bail Act that was enacted in Victoria in 1977, like bail legislation introduced in other Australian jurisdictions in the 1970s and 80s, was primarily aimed at 'tidying up' bail law and reducing the use of monetary bail (i.e. release on bail that was primarily based on the ability of the person to pay a required amount of money).[footnote 40] In addition to a shift to non-financial conditions, the new Act continued many of the matters that had had developed under the common law—bail still primarily functioned as a mechanism to ensure that a person charged with a criminal offence would attend court for the hearing of their case and there was a presumption that bail would be granted.[footnote 41]

Both the report of the parliamentary committee that made recommendations about new bail legislation[footnote 42] and the parliamentary debates that took place following the Bill's introduction, revealed a concern for the rights and liberty of those charged with (but not convicted of) criminal offences. There was a desire to promote awareness of bail, and a disquiet about long periods being spent in pre-trial detention. A Member of Parliament who was formerly a justice of the peace (and therefore experienced in bail matters) observed that police opposed bail 'on far too many occasions'.[footnote 43] The Labor Opposition emphasised that denying bail should only occur in exceptional circumstances and media reports emphasised the rights of those accused, but not convicted, of crimes.[footnote 44]

The Bail Act was supported by all parties. The key consideration in determining eligibility for bail was whether a person would attend the hearing of their matter.[footnote 45] Victorian Attorney-General, Haddon Storey, stated that 'Bail is fixed primarily to ensure that an accused will attend his trial...'.[footnote 46] The presumption of innocence was acknowledged through the recognition that a person held in custody after being arrested had a 'prima facie right to be released'.[footnote 47]

Risk assessment

The Bail Act 1977 largely repeated the position of the common law in relation to bail, establishing a prima facie entitlement to be released before trial, but qualifying this by providing that bail would not be granted if there was an 'unacceptable risk.'[footnote 48] This mean that, if released on bail, the person would not appear in court for the hearing of their case, commit an offence, endanger the safety or welfare of the community, or interfere with witnesses or justice.[footnote 49] Where a bail decision maker was satisfied that a person presented an unacceptable risk in relation to any of these four matters, they were obliged to refuse bail and remand the person into custody.

Although the original version of the Act in 1977 provided that bail could be denied if the person presented an unacceptable risk of offending if granted bail, community protection did not loom large in the development of the Bail Act or the associated parliamentary debates.[footnote 50] These debates acknowledged the competing interests of legal principle and risk assessment. They tried to balance the conflicting rights of persons to be free pending their trial and the interests of the community in ensuring they committed no offences and did in fact appear when required. On the issue of the risk of offences being committed by an applicant if bail was granted, the Attorney-General stated that it was not easy to determine how many offences were committed by people on bail. He further observed that 'although this does happen from time to time, in the majority of cases it does not'.[footnote 51]

Nevertheless, concern about the dangers presented by some bail applicants (those charged with aggravated burglary, offences that involved the use of a weapon or who were charged with committing certain offences while on bail), led to the introduction of a special risk management strategy relevant to community protection.[footnote 52] A reverse onus provision required persons in these circumstances who applied for bail to 'show cause' as to why bail should be granted.[footnote 53] This unusual risk management mechanism removed the presumption of bail and placed the onus on the applicant to satisfy the bail decision maker as to why they should be released on bail. This shift was grounded in concerns about community protection.

It may be argued that this additional risk mechanism was redundant because the risks were already adequately addressed in the four basic risk criteria.[footnote 54] Yet, it began a process that was to become increasingly important in using bail decision-making as a means to protect the community— it articulated particular circumstances in which a presumption against bail applied, and placed the onus on the person applying for bail to demonstrate that any relevant risks could be managed and that bail should be granted.

The operation of the Bail Act 1977

The report of the parliamentary committee that preceded the new bail legislation, together with the Bail Act 1977 in its original form and the parliamentary debates that accompanied it, constructed bail applicants as rights-invested citizens and demonstrated a concern to preserve pre-trial liberty. Risk assessment focused on traditional concerns, except for the introduction of a very limited reverse onus provision that required some bail applicants to 'show cause' as to why they should be granted bail.

For the first 20 years of the operation of the Bail Act, relatively few people were refused bail; studies routinely estimated that more than 90 per cent of people who applied for bail by police were successful.[footnote 55] Taking into account subsequent applications to bail justices and the courts, this figure increased to 97.5 per cent of applicants being granted bail.[footnote 56]

Within Victoria Police, there was 'a cultural expectation that a person would be granted bail'.[footnote 57]Consequently, the proportion of the prison population being held on remand was relatively constrained. In 1998, it amounted to about 15 per cent of prisoners.[footnote 58] But, in the following two decades, as concerns about community protection grew, the provisions of the Bail Act were repeatedly amended and became increasingly restrictive. The use of reverse onus provisions as a risk management strategy increased, with the expansion of the 'show cause' test and of a new, even more restrictive test of 'exceptional circumstances' being introduced. Accompanying these changes, the proportion of unsentenced prisoners more than doubled over this period, rising to 32 per cent of all prisoners in Victoria in 2017.[footnote 59]

6. Risk and bail: 2019

In 2019, Victoria has essentially the same general framework for bail that was established in 1977. Some clarification has occurred—the Bail Act is now more clearly organised through the introduction of two schedules that list the offences to which the reverse onus tests apply, and the relationship between the unacceptable risk and reverse onus tests has been clarified.

However, there are also some significant changes. The 'show compelling reason' test has replaced the previous 'show cause' test and the range of offences to which the reverse onus tests apply has been extended. Most significantly, community protection has now become the primary consideration in relation to all applications for bail.

This change in the focus of risk assessment is the culmination of change brought about through the accumulation of amendments to the Bail Act 1977 and significant reforms introduced in 2017–18 following the review into bail conducted by the Hon. Paul Coghlan QC.

Although there is now concern about the number of persons held in prison as a consequence of these developments, initial responses to the tightening of bail were highly favourable. Premier Daniel Andrews initially treated the increase in the remand population as a positive indication that the community was being protected.[footnote 60] The Attorney-General approvingly noted that the Victorian bail system was arguably the most onerous in Australia and that further reforms would 'make it even stronger'.[footnote 61]

The growth of statutory amendments

From 1977 until the major reforms of 2017-18, the Bail Act was amended many times. Another reverse onus provision was added, which established an even higher hurdle to the granting of bail. Under this provision, persons charged with murder or treason are required to establish 'exceptional circumstances' if they are to be granted bail.[footnote 62] Subsequently, more offences were added to both this and the show cause reverse onus provisions. For example, certain serious drug offences, stalking in particular circumstances and other offences were added to the 'exceptional circumstances' provision.[footnote 63] Similarly, numerous drug offences were added to the 'show cause' (now replaced by 'show compelling reason') category.[footnote 64] These additions extend the presumption against bail and its inherent starting point—that risk can only be managed by detaining the person in prison prior to the hearing of their case.

Other statutory changes have also had impact. The secondary offences of contravening a conduct condition of bail and committing an indictable offence while on bail, that were enacted in 2013,[footnote 65] mean that a person who contravenes certain bail conduct conditions no longer simply breaches bail—they commit a criminal offence (in addition to the actual offence they have allegedly committed).[footnote 66] Similarly, a person who commits an indictable offence while on bail is not simply liable for that offence and for breaching bail, but is now also liable for the offence of committing an indictable offence while on bail.[footnote 67]

In addition to these reforms, other limitations on pre-trial liberty (albeit not involving detention) were introduced through the growth of bail conditions.[footnote 68] These conditions have been progressively extended and must now be imposed where they will reduce the likelihood that an accused will endanger the safety or welfare of any person; or fail to attend court; or commit an offence while on bail; or interfere with witnesses or otherwise obstruct the course of justice.[footnote 69]

The Coghlan Review: Managing risk and maximising community safety

In 2017, the Victorian Government commissioned the Hon. Paul Coghlan, to review the operation of bail laws and practice in Victoria, with the key aim of best managing risk and maximising community safety.[footnote 70] The immediate trigger for the review were the crimes committed by James Gargasoulas, who drove a stolen car into pedestrians in the Melbourne CBD in January 2017, killing six individuals and directly injuring 27 more. The fact that Gargasoulas was on bail when committing these offences was heavily emphasised throughout the extensive publicity generated by his crimes.

The terms of reference of the Coghlan Review focused on the relationship between bail and community safety, and required consideration of the risk management tests and the balance between protection of the community and the presumption of innocence. The Review was also required to consider whether bail decision-makers obtained sufficient information and the way in which relevant circumstances are taken into account.

The Coghlan Review contained a total of 37 recommendations for reform.[footnote 71] These favoured maintaining the presumption of bail but also recommended that the reverse onus tests should apply to a larger range of offences and that a new test should replace the 'show cause' test. The Review acknowledged that this extension was inconsistent with the presumption of innocence and the right to liberty (as protected by the common law and the Victorian Charter of Human Rights and Responsibilities Act 2006).[footnote 72] However, the proposed reforms were justified on the basis that the community favoured such an extension because it would enhance community protection. Coghlan also considered that 'greater emphasis should be placed on the assessment of risk'[footnote 73] and that reforms must recalibrate the balancing of the rights of the individual against those of the community to clearly favour community protection. Both the reports and associated parliamentary debates centred upon a belief that members of the community favoured such a shift,[footnote 74] with surveys of public opinion supporting this view.[footnote 75]

The Government accepted most of the recommendations, which were endorsed by the Opposition. The Attorney-General observed that the 'overhaul' of the bail system would improve community safety by making it 'harder than ever for people to get bail'.[footnote 76]

Changing the Bail Act, 2017–2018

Statutory amendments enacted in 2017 (the Bail (Stage One) Amendment Act 2017) and 2018 (the Bail (Stage Two) Amendment Act 2018) significantly increased the number of offences to which the reverse onus tests apply, and substituted a 'show compelling reason' test for the previous 'show cause' test.[footnote 77] The Guiding Principles of the Act now make community protection theprimary risk consideration in all bail matters, by specifying that all decision-making in relation to bail should recognise the importance of 'maximising the safety of the community and persons affected by crime to the greatest extent possible'[footnote 78] (other guiding principles only have to be taken into account or recognised as promoting certain values).

Parliamentary debates associated with the new bail legislation in 2017 and 2018 were heavily influenced by two cases where heinous crimes had been committed by individuals who were on bail at the time of offending. In addition to Gargasoulas,[footnote 79] discussion focused on Sean Price, who murdered Masa Vukotic while out on bail.[footnote 80] References in parliamentary debates to individuals who had committed serious offences while on bail were included with references to those who had been released on bail, and whose victims had reported fear (but no offending by the relevant person),[footnote 81] thereby creating an image of lawlessness and offending under the auspices of lax bail laws. Unlike the situation 20 years earlier, when the then Attorney-General stated that he did not know how many people committed offences while on bail, politicians now referred to a 'raft' of cases where this occurred,[footnote 82] claiming that 'bail laws have been broken far too often'.[footnote 83] During the debate, a Member argued that 'violent criminals are let out on bail and have been free to commit other offences',[footnote 84] and the solution was identified as being the adoption of restrictive bail laws and practices.[footnote 85] Though some alternative views occasionally appeared, they were limited in number.[footnote 86] A strong campaign by media outlets, especially The Herald Sun newspaper, also favoured more restrictive bail practices and law.[footnote 87]

Community protection dominated the parliamentary debates and media reporting, which overwhelmingly focused on the risk of future offending by those applying for bail and relied on refusal of bail and remand (incapacitation) as the appropriate strategy for risk management.[footnote 88] This differs from earlier concerns that centred on violent crimes committed against banks—violence and assaults now dominated parliamentary discussion.[footnote 89] There was increased focus on violent crimes such as rape, murder, gang violence, carjackings, home invasions and individual assaults committed against members of the community. Compared with parliamentary debates in 1977, in 2017–18 there was less concern with the rights of bail applicants. There were fewer references to principles such as the right to liberty, the presumption of innocence and there was little concern about the adverse consequences, or circumstances, of pre-trial detention.[footnote 90] Indeed, the presumption of innocence was rarely mentioned.[footnote 91] Some parliamentarians appeared to assume the guilt of bail applicants, viewing being detained on remand as the penalty[footnote 92] or punishment for 'offending'.[footnote 93]

The Bail Act 1977 (Vic), as now amended:

§ re-orders the matters to be taken into account when determining 'unacceptable risk' and makes community safety the primary risk to be considered for all bail applicants;

§ provides legislative guidance as to the matters that must be taken into account when assessing risk. In addition to special provisions in relation to children,[footnote 94] Aboriginal persons[footnote 95] and other vulnerable persons,[footnote 96] assessments of a bail applicant's risk must take into account at least 31 other 'surrounding circumstances', including the nature and seriousness of the alleged offending, the person's criminal history, their personal circumstances, etc.[footnote 97] There is no weighting of these factors and no guidance as to how the decision-making should occur; and

§ expands the range of offences to which the reverse onus tests apply. This has occurred in two ways:

o Adding offences to these categories.

Aggravated carjacking, aggravated home invasion, and 'an offence of conspiracy to commit, incitement to commit or attempting to commit' any of the offences listed in Schedule 1 of the Bail Act are now governed by the requirement that the bail applicant demonstrate exceptional circumstances.[footnote 98] This has been justified on the basis that the offences are 'serious'. Other offences, including rape, kidnapping, armed robbery and culpable driving causing death have similarly been added to the 'show compelling reason' category.[footnote 99]

o Adding individuals on conditional liberty to those categories.

Individuals already on conditional liberty through bail, summons, parole, being subject to an arrest warrant or subject to a supervision order under the Serious Sex Offenders (Detention and Supervision) Act 2009 (Vic) must 'show compelling reason' or 'exceptional circumstances' to obtain bail. The justification for this is that these persons, already subject to some restrictions on their liberty and then charged with more alleged offending, present a higher level of risk to the community.[footnote 100]

While the framework established by bail law remains essentially the same as that established 40 years ago, relevant Victorian law has been amended so significantly that the presumption of a right to bail no longer operates for more than 100 offences. No other jurisdiction has the presumption of bail hedged in by so many exceptions.[footnote 101] Unsurprisingly, it is generally acknowledged that Victoria now has the most restrictive regime governing bail in Australia.[footnote 102]

Community protection—more specifically, protecting the community from offences that a person might commit if released on bail—has driven the restrictions on bail in Victoria and underpins the increase in pre-trial (preventive) detention. But the increasing reliance on remand has not been restricted to those who present a risk of serious violent crime; only about 14 per cent of women held on remand in Victoria at the end of June 2017 were charged with assault offences (more than half were charged with property and drug offences rather than crimes of violence).[footnote 103] In this context, a more detailed analysis of how 'community protection' and the associated notion of 'risk' are understood is essential.

7. Community protection and the construction of risk


Bail decision-making has always been about risk. However, what has changed in recent years is the focus of the risk. Traditionally concerned with the likelihood that a person would not turn up at court for the hearing of their case, over the last 40 years statutory formulations of bail law in Victoria have increasingly identified other risks—especially the risk of the person offending and endangering public safety or welfare if released on bail. This risk has dominated discussion of bail in recent years and has underpinned the significant increase in the Victorian prison population.

Bail law in Victoria now focuses on assessments of the likelihood of future offending based on the offence charged, the liberty status of the individual and the 'surrounding circumstances' of the offence which is linked to their bail application. More specifically, where certain serious offences are involved, risk management in bail law is based on the consequences of specific offence recidivism (the term recidivism is used deliberately), irrespective of the probability of that future offending eventuating. This risk management mechanism is effectuated through the use of reverse onus provisions requiring some bail applicants to demonstrate ('show compelling reason' or 'exceptional circumstances') why they should be released on bail, and increasingly results in pre-trial detention—demonstrated by the increasing proportion of prison detainees who are being held on remand.

Given the human and fiscal costs associated with this approach, as well as its challenge to foundational legal principles, it is surprising that the evidence to justify its efficacy in enhancing community safety is limited. In particular, the limited empirical research on offending by persons on bail—a matter critical to the issue of community protection—has not been taken into consideration.

How common is offending while on bail?

More than 40 years ago, the then Victorian Attorney-General, Haddon Storey, commented on the lack of available information about offending by those released on bail,[footnote 104] a point that has been echoed in the Coghlan Review[footnote 105]and many other studies.[footnote 106]Nevertheless, many political commentators assume that such offending is common. This belief has driven policy and law reform. Thus, in political debates in 2017 and 2018 there were references to a 'huge amount of offending'[footnote 107] and a 'raft' of crimes[footnote 108] being committed by those on bail.

In the absence of reliable information, this perception appears to have been based on a small number of highly publicised, violent crimes committed by individuals who were on bail at the time of the offending. It is an example of the 'Willie Horton' effect: the influence on criminal justice policy of high profile cases that emphasise the perils of some form of clemency towards dangerous individuals. But what does more comprehensive research indicate about the issue?

While there are thousands of people on bail in Victoria at any given time, little is known about their criminal offending. There is a remarkable lack of information about the amount and types of crime committed by those released on bail. The few studies that have been conducted in Australia suggest that most persons on bail are not charged with an offence while on bail. For instance, a study in Western Australia reported that 24 per cent of their sample offended while on bail.[footnote 109] However, the scarcity of Australian studies means that most available information comes from other jurisdictions, involving different time periods, varying methodologies and derived from vastly different bail systems. Nevertheless, these studies provide some data indicative of the likelihood of offending while on bail.

Studies from England, Scotland, the United States and New Zealand indicate that most people (typically more than 80 per cent) do not commit offences while on bail.[footnote 110] For instance, research conducted on bail in five areas around London in 1993–94 found that less than 20 per cent of persons were convicted of offences committed while they were on bail. If a person did commit an offence while on bail, it was most likely to be similar to their original alleged offence; e.g. a person charged with burglary and granted bail was most likely, if they offended, to commit burglary.[footnote 111] Interestingly, those on bail for violent crimes were among the least likely to offend while on bail; only seven per cent of those on bail for assault, and six per cent of those on bail for sex offences, were convicted of similar offences while on bail.[footnote 112]

In Scotland, a study reported that, while a higher proportion of those released on bail were charged with another offence alleged to have been committed while they were on bail (29 per cent of the sample), only six per cent of the alleged offences involved violence; the majority involved relatively minor crimes of dishonesty.[footnote 113] In New York City, similar findings were reported: about 17 per cent of those released on bail were arrested during their bail period, with most arrests (over 80 per cent) being for non-violent offences; only ten per cent of that sub-group who were arrested were alleged to have committed serious violent offences (homicide, robbery, sexual offences, kidnapping, assault).[footnote 114]

In New Zealand—which has not experienced a long-term increase in the remand population similar to Australia—from 2011–2016, more than 80 per cent of those on bail did not offend.[footnote 115] For those who did offend, more than half (56 per cent) of their offences related to traffic offences, theft and offences against justice (mainly breaches of community work orders). The New Zealand study reported that only theft offences were committed more frequently on bail, and sexual and traffic offences less frequently, compared to general offending rates.[footnote 116]


Clearly, given the differences in law, legal procedures, demographic and cultural factors between different jurisdictions, and methodologies of the various research studies, care must be exercised before generalising these findings.[footnote 117] Nevertheless, collectively, these studies suggest that about one in five people commit an offence while released on bail, and only a small proportion of these offences involve serious violence or sexual offending.[footnote 118]

These data offer a number of challenges to current assumptions about offending while on bail:

§ the focus on a small number of heinous crimes committed by persons on bail has led to a disregard of the large number of persons on bail who do not offend;[footnote 119] available research indicates that most people do not commit an offence while they are on bail; and

§ most offences committed by persons on bail are non-violent offences—in particular, only a small proportion of those on bail after being charged with violent or sexual offences will commit sexual or violent offences while on bail.

This is not to dismiss the problems caused by those who commit serious offences while on bail. However, it does suggest that risk assessment must be more specifically targeted to identify this relatively small group who commit further serious offences and that the reverse onus provisions— based solely on the offence with which the person is charged—over-predict risk.

Reverse onus provisions

The reverse onus provisions incorporated in bail law in Victoria are the key mechanism for managing risk of offending while on bail. Essentially, these provisions do this by establishing categories of offences whereby if the offence is charged, there is a presumption that the applicant will not get bail. Note that the rationale for particular offences being captured by reverse onus provisions is not based on the likelihood of bail applicants committing these offences if they are released into the community on bail. Rather, the provisions are based on the severity of the consequences if the bail applicant was released and committed an offence of similar severity to that with which they were charged.[footnote 120] This reasoning was made explicit when a 'show cause' provision was introduced into the Bail Act in New South Wales in 2014. The architect of the reform, Attorney-General John Hatzigeros, explained that the change would apply to:

offenders whose alleged offences are such that in the ordinary course, the consequences of materialisation of the risk to the community and the administration of justice are such that they outweigh the likelihood of it occurring. ... Broad categories should more accurately reflect groups or types of offences that have such significant consequences to the community.[footnote 121]

This approach is very different to most forms of risk assessment, which consider the severity of the consequences of specified conduct as well as the likelihood of it occurring. The distinctive form of risk assessment inherent in reverse onus provisions does not consider the likelihood or probability of a crime being committed and therefore over-predicts the likelihood of offending. In essence, reverse onus provisions strongly favour community protection over the right to liberty/presumption of innocence. This approach is not based on the likelihood of a serious offence being committed if bail is granted. Rather, it assumes that such an offence will be committed and requires the bail applicant to satisfy the decision-maker that either 'exceptional circumstances' or 'compelling reason' exists to justify bail.

While this offence category-based risk assessment takes into account the individual characteristics of bail applicants (these are required to be considered as 'surrounding circumstances' under s 3AAA), and requires consideration of whether imposing bail conditions can mitigate the risk,[footnote 122] this is a cumbersome and unnecessarily complex approach. It is a very different approach to that adopted in reforming jurisdictions in the United States, which utilise actuarial and structured judgement risk-assessment instruments. While consistent with the high volume, time-limited bail hearings that take place in Australia, this approach to risk assessment clearly prioritises community protection and contributes to the increasing remand population. Whether it actually enhances community protection is another matter; the preceding review of research on bail and offending suggests that the approach over-predicts risk.[footnote 123]

The consequences of incarceration: remand is criminogenic

Using remand as a means of community protection offers some short-term benefits,[footnote 124] although, as previously noted, current risk assessment tests lead to too many people being detained. What is particularly concerning are the long-term effects of this strategy. A focus on pre-trail detention ignores the 'downstream effects' of this policy; that is, the long-term impact of pre-trial detention on future offending.

There is little research on this matter; indeed, the New South Wales Law Reform Commission specifically acknowledged the importance of the issue and noted the absence of information.[footnote 125] However, important research has recently been undertaken in the United States. Well-controlled studies suggest that even short-term, pre-trial detention contributes to future offending.

A recent study of persons in Harris County, Texas, compared two groups of 'similarly situated' people who had been arrested and charged with criminal offences; one group was granted bail and the other held on remand in prison. During an 18-month follow-up period, those who had been remanded into prison were more likely to be charged with offences; this difference persisted even after controlling for offence, defendant demographics, criminal history and legal representation. Pre-trial detention was associated with a 30 per cent increase in new serious (felony) charges and a 20 per cent increase in new minor (misdemeanor) charges. [footnote 126]

Another study, using data from Florida and Pennsylvania, adopted a different methodology but reached a similar conclusion: pre-trial release was associated with a reduced likelihood of long-term criminal offending.[footnote 127] This evidence suggests that pre-trial detention may itself have a criminogenic effect.[footnote 128] Paradoxically, the protective benefits of short-term, pre-trial incapacitation may ultimately compromise community safety.

The diminishing rights of the accused

A significant feature of contemporary debates about bail is the relative absence of consideration of the rights of the bail applicant. Compared with previous discussions and parliamentary debates, in Victoria in 2017–18 there was relatively little reference in the parliamentary debates on bail to principles such as the right to liberty, the presumption of innocence and the adverse consequences, or circumstances, of pre-trial detention.[footnote 129] Indeed, recent political and public discourse frequently assumes guilt, with remand being the penalty[footnote 130] or punishment for 'offending'.[footnote 131] This is more than linguistic slippage–it indicates a troublesome erosion of the presumption of innocence.

In summary, the focus on a small number of high-profile cases that involved serious crimes committed by persons on bail, reliance on reverse onus provisions for risk management, the lack of awareness of the criminogenic effects of even short periods of pre-trial incarceration, and a disregard of the rights of those accused—but not convicted—of a crime, suggest that the current approach to the bail system in Victoria is flawed.

The following section of this paper identifies some strategies for moving forward.

8. Breaking the nexus between community protection and increased incarceration

Assessing the risk that a person will commit an offence if they are granted bail is unnecessarily complex and also results in too many people being denied bail and remanded into custody. Three reforms could simplify bail decision-making, improve risk assessment and serve the twin goals of protecting the community and reducing the number of persons held on remand. These are:

§ removing relatively minor, non-violent offences for those for which bail is relevant;

§ making 'unacceptable risk' the sole risk criteria; and

§ using actuarial instruments to assist in the determination of risk.

Removing some offences from the category of 'bailable' offences

Although this report has focused on risk assessment relating to serious offending, a significant, pragmatic strategy for reducing the number of persons held on remand would be to adopt the recommendation in the Coghlan Review simply to remove some relatively minor, non-violent offences from any requirement of bail.[footnote 132] In effect, this reform involves an assessment that the minor nature of the alleged offending fails to raise concerns about community protection and other risks. This approach has been adopted in England and has contributed to the reduced rates of persons being incarcerated before their trial.[footnote 133]

Making unacceptable risk the sole risk criterion

The Court of Appeal of Victoria,[footnote 134] the Victorian Law Reform Commission[footnote 135] and the Law Institute of Victoria[footnote 136] have recommended that the reverse onus categories in the Bail Act 1977 should be abandoned, and that bail decision-making should be based on a single test—unacceptable risk. In practice, this would mean simply determining in all bail applications whether there was an 'unacceptable risk' of the person: failing to turn up for the hearing of their matter; offending if released; presenting a threat to the safety or welfare of the community; or interfering with the justice process. This approach not only simplifies bail decision-making but also avoids the breach of legal principle involved in the reverse onus provisions (which impose responsibility on the person applying for bail to satisfy the court as to why they should not be detained in custody).

However, recent reforms in Victoria have confirmed a two-step process for assessing risk and expanded the offences that fall into the 'exceptional circumstances' and 'show compelling reason' reverse onus categories. The process is cumbersome, with the same risk factors ('surrounding circumstances') being taken into consideration at both steps of the process. A better approach would be to adopt a streamlined process, whereby a bail decision-maker, assisted by appropriate risk assessment information (incorporating actuarial assessments) makes the decision as to whether releasing a person on bail constitutes an unacceptable risk.

Using actuarial instruments to assist with risk assessment

Currently, bail decision-makers are directed to take into account at least 31 'surrounding circumstances' when deciding whether a person applying for bail presents an 'unacceptable risk' of committing an offence if released. In circumstances where either of the reverse onus tests apply, these 31 factors must also be taken into account when determining whether the person has established 'exceptional circumstances' or has shown 'compelling reason' as to why they should be granted bail. Not all circumstances will be relevant in all cases, and some are related to the legal process rather than the bail applicant (for instance, the length of time the person is likely to spend in custody if bail is refused).

Nevertheless, there are 22 potential circumstances that relate to the person, their history and their alleged offending. The Bail Act does not give any guidance as to how this information should be weighed and integrated.[footnote 137] As most bail hearings take place at an early stage in the legal process with hearings of notoriously short duration,[footnote 138] the information available is often limited and of dubious reliability.[footnote 139] Getting and appropriately using relevant information is a difficult task. On this issue, much could be learnt from contemporary reforming jurisdictions in the United States.

In America, bail reformers in several states, including Kentucky,[footnote 140] New Jersey[footnote 141] and Washington D.C.,[footnote 142] have sought to improve bail decision-making by developing and using validated, standardised assessments. These scales are based on previous research that has empirically identified factors related to offending while on bail. Under these reforms, a person who applies for bail is assessed on specified criteria and given a score; this score is then used to identify the level of risk if they were released on bail. For example, the Laura and John Arnold Foundation funded the development of the Public Safety Assessment (PSA) scale, whereby persons applying for bail are assessed on nine risk factors.[footnote 143] Their score then generates a prediction of high, medium or low risk on three key issues: failure to appear; new criminal activity; and new violent criminal activity (see Figure 6 for the sub-scale for new violent criminal activity).

Figure 6: New Violent Criminal Activity Scale, Public Safety Assessment[footnote 144]

The issues covered in the PSA scale are matters that bail decision-makers in Victoria are already required to take into consideration as surrounding circumstances when determining risk;[footnote 145] the scale simply selectively utilises and weighs a small number of these factors that have been empirically identified as most likely to accurately predict risk.[footnote 146]

Although more elaborate bail risk assessments exist, the PSA is quick and easy-to-use, utilising information that is readily available to police, and consequently would be well suited to the brief bail hearings that occur in Victoria. While requiring testing and validation (and likely modification) for use in Victoria, employing a standardised, validated risk assessment scale would promote consistency, remove some of the subjectivity that police have identified as problematic in bail decision-making,[footnote 147] and focus attention on the surrounding circumstances most relevant to risk. Given that the general literature on risk assessment indicates that predictions are most accurate when the risk period is relatively short (as occurs for many bail applicants), there is good reason to explore empirically derived risk assessment instruments such as the PSA in future.

However, there are limits to the use of these actuarial scales.[footnote 148] Namely, they are designed to assist bail decision-makers, not to replace them and they cannot themselves provide answers to key policy questions relating to risk. For instance, should bail be denied when there is an unacceptable risk that the bail applicant will engage in any subsequent offending, or should the focus be restricted to serious violent and sexual offending?[footnote 149] Should bail be denied to those who are assessed as having a moderate risk of offending, or only those who present a high risk? Finally, an exclusive focus on risk assessment disregards legal rights and therefore provides a limited view of relevant interests.[footnote 150] As Andrew Ashworth, Professor of Law at Oxford University, has pointed out,

Assessing situations and persons from the point of view of perceived risk sits rather awkwardly with respecting the dignity of others as full, rights-bearing citizens.[footnote 151]

Nevertheless, the use of standardised risk assessment scales, using information already known to police, could assist bail decision-makers to more accurately identify those who present a risk of offending if released on bail.

9. Conclusion

Most people would accept that, in a liberal democracy, pre-trial detention through the denial of bail is sometimes necessary. However, determining when it is necessary is more difficult.

Currently in Victoria, denying bail and remanding into custody is being used as a crime prevention tool, to protect the community from the risk that a person would commit offences if released into the community on bail.[footnote 152] Indeed, the primary function of bail hearings in Victoria now is to assess the risk that a person seeking bail will, if released into the community, commit offences.

Although there is no simple and direct link between restrictive bail laws and remand rates,[footnote 153] it is clear that notions of community protection now dominates discussion of bail in Victoria. Restrictive bail laws have been accompanied by broader changes in practices that have led to escalating remand rates, generated by concerns about community safety. This development blurs the distinction between the presumed innocent and the proven guilty.[footnote 154] While satisfying immediate political and community imperatives, it comes at immediate fiscal and human costs, may itself be criminogenic, and have a paradoxical downstream effect. This approach is without adequate justification in legal principle or empirical validation.

Furthermore, an increasing remand population is not inevitable and there are strategies that can be employed to reduce it.[footnote 155] As discussed earlier, the current approach of incarcerating unprecedented numbers of persons by denying them bail actually presents long-term threats to community safety. It is recognised that Victoria's bail laws and practices need to be overhauled.[footnote 156] When that review and reform occurs, among the many other matters that should be taken into account, Victoria should consider adopting a more refined risk assessment of bail eligibility, specifically targeting those who present an unacceptable risk of committing serious violent or sexual offences if released on bail. This approach should be based on simplified law and utilise actuarial information, so that both short- and long-term community protection are considered and that only those persons who present a risk of perpetrating serious violent crime are detained pre-trial.

 

10. References

(2018) 'Note: Bail Reform and Risk Assessment: The Cautionary Tale of Federal Sentencing', Harvard Law Review, 131(1).

Allan, A. et al. (2003) 'The relationship between bail decision-making and legal representation in the criminal justice system', Edith Cowan University.

Andrews, D., Premier of Victoria (2018) Questions without Notice and Ministers Statements, Debates, Victoria, Legislative Assembly, 19 September.

Ashworth, A. (2006) 'Four threats to the presumption of innocence', International Journal of Evidence and Proof, 10(4).

Australian Bureau of Statistics (2014) Prisoners in Australia 2014: Victoria Snapshot, cat. no. 4517.0, Canberra, ABS.

Australian Bureau of Statistics (2018) Prisoners in Australia, 2018: State and Territory Profiles, Victoria, cat. no. 4517.0, Canberra, ABS.

Australian Bureau of Statistics (2018) Proportion of Unsentenced Prisoners, cat. no. 4517.0, Canberra, ABS.

Bagaric, M. et al. (2018) 'Mitigating America's mass incarceration crisis without compromising community protection: Expanding the role of rehabilitation in sentencing', Lewis & Clarke Law Review, 22(1).

Bamford, D., S. King and R. Sarre (1999) Factors Affecting Remand in Custody: A Study of Bail Practices in Victoria, South Australia and Western Australia, Canberra, Australian Institute of Criminology.

Bartels, L., K. Gelb et al. (2018) 'Bail, risk and law reform: A review of bail legislation across Australia', Criminal Law Journal, 42(1).

Bottoms, A. & J. McClean (1976) Defendants in the Criminal Process, London, Routledge & Kegan Paul.

Brown, D. (1987) 'Remanded in custody: A NSW refrain', Legal Services Bulletin, 12(4).

Brown, D. & J. Quilter (2014) 'Speaking too Soon: The sabotage of bail reform in New South Wales' International Journal for Crime, Justice and Social Democracy, 3(3), p. 89.

Brown, K. et al. (2004) Offending On Bail: An Analysis of the Use and the Impact of Aggravated Sentences for Bail Offenders, Edinburgh, Scottish Government.

Bushnell, A. (2017) Australia's Criminal Justice Costs: An International Comparison, Sydney, Institute of Public Affairs.

C. Yang (2017) 'Toward an optimal bail system', New York University Law Review, 92(5).

Cape York Institute (2019) The Indigenous Incarceration Crisis: The Queensland Productivity Response is Inadequate, CYI Further Submission to QPC Imprisonment and Recidivism Inquiry,April, Cairns, Cape York Institute.

Cavanagh, R. (2018) 'Teen Thug Released Despite Being on Bail When Sexually Assaulting Two Women', The Herald Sun, January 1.

Clark, R., MLA (2018) 'Second reading speech: Serious Offenders Bill 2018', Debates, Victoria, Legislative Assembly, 23 May.

Coghlan, P. (2017a) Bail Review: First Advice to the Victorian Government, report prepared for the Victorian government, Melbourne, Victorian Government.

Coghlan, P. (2017b) Bail Review: Second Advice to the Victorian Government, report prepared for the Victorian government, Melbourne, Victorian Government.

Connolly, Justice T. (2006) 'Golden Thread or Tattered Fabric: Bail and the Presumption of Innocence', paper presented to the Law Council of Australia, National Access to Justice and Pro Bono Conference, Melbourne, 11–12 August.

Corrections Victoria (2019a) Monthly Time Series Prisoner and Offender Data, Melbourne, Department of Justice and Community Safety.

Corrections Victoria (2019b) Women in the Victorian Prison System, Melbourne, Department of Justice and Community Safety, January.

Crime Statistics Agency (2018) 'Spotlight: Breaches of Orders – The Impact of Legislative Changes', Crime Statistics Agency website.

Desmarais, S. L. & E. M. Lowder (2019) Pretrial Risk Assessment Tools: A primer for judges, prosecutors, and defense attorneys, Chicago, MacArthur Foundation.

Dhami, M. & P. Ayton (2001) 'Bailing and jailing the fast and frugal way', Journal of Behavioral Decision Making, 14(2).

Dobbie, W., J. Goldin & C. Yang (2018) 'The effects of pre-trial detention on conviction, future crime, and employment', American Economic Review, 108(2).

Eddy, R. (1977) Second Reading Debate: Bail Bill, 19 April, Debates, Victoria, Legislative Council.

Ericson, M. & T. Vinson (2010) Young People on Remand in Victoria: Balancing Individual and Community Interests, Richmond,Jesuit Social Services.

Feeley, M. & J. Simon (1992) 'The New Penology: Notes on the Emerging Strategy of Corrections and Its Implications', Criminology, 30(4).

Fitzherbert, M., MLC (2017) 'Second reading speech: Bail Amendment (Stage Two) Bill 2017', Debates, Victoria, Legislative Council, 22 February

Galouzis, J. and S. Corben (2016) 'Judicial outcomes of remand inmates in New South Wales', Research Bulletin No.34, Sydney, Corrective Services NSW.

Gatt, W. (2017) 'Wayne Gatt: Bail Has Lost its Sense of Balance', The Herald Sun, March 8.

Gleeson, H. (2019) 'Australia Must 'Radically Re-Think' Its Prisons to Avoid Becoming Like America, U.S Activist Says', ABC News, 30 April.

Guy, M., MLA (2018) 'Questions without Notice and Ministers Statements', Debates, Victoria, Legislative Assembly, 19 September.

Hamilton, H. (1977) 'Second Reading Debate: Bail Bill', Debates, Victoria, Legislative Council, 19 April.

Hatzigeros, J. (2014) Review of the Bail Act 2013, report prepared for the Department of Justice, Sydney, NSW Government.

Heaton, P., S. Mayson & M. Stevenson (2017) 'The downstream consequences of misdemeanor pretrial detention', Stanford Law Review, 69(3).

Henderson, P. F. & T. Nichols (1992) Offending while on bail, Research Bulletin, no. 32, London, Home Office.

Heyerly, M. (2013) Pretrial Reform in Kentucky, report prepared for Pretrial Services, Frankfort, Kentucky Court of Justice.

Hucklesby, A. (2009) 'Keeping the lid on the prison remand population', Current Issues in Criminal Justice, 21(1).

Hucklesby, A. (1996) 'Bail or jail? The practical operation of the Bail Act 1976', Journal of Law and Society, 23(2).

Johnson, E. (1991) Royal Commission into Aboriginal Deaths in Custody, National Report, 1.

Kealy, E., MLA (2018) 'Condolences', Debates, Victoria, Legislative Assembly, 19 June.

Kellough, G. and S. Wortley (2002) 'Remand for plea', British Journal of Criminology, 42(1).

King, M. (1973) Bail or Custody, London, The Cobden Trust.

Koepke, J. & D. Robinson (2018) 'Danger ahead: Risk assessment and the future of bail reform', Washington Law Review, 93(1).

Lash, B. (2003) Trends in the Use of Bail and Offending While on Bail, 1990–1999, Wellington, New Zealand Ministry of Justice.

Laura and John Arnold Foundation (date unknown) 'Public Safety Assessment: Risk factors and formulas', Public Safety Assessment website.

Laura and John Arnold Foundation (2013) 'Pretrial Criminal Justice Research', Houston and New York, Laura and John Arnold Foundation.

Law Institute of Victoria (2017) Submission: Review of Victoria's Bail System, Melbourne, Law Institute of Victoria.

Matthau, D. (2017) 'Is NJ bail reform working, or just giving criminal suspects a free pass?' New Jersey 101.5, 8 February.

Miles, J. (1991) 'Bail Legislation: Objectives and Achievements', paper presented to the Australian Institute of Criminology Conference, Canberra, 29 Nov–1 Dec.

Ministry of Justice (2018) 'Offending on Bail', factsheet, New Zealand Government website.

Ministry of Justice (2019) Offender Management Statistics Bulletin, England and Wales, Quarterly July to Sept 2018, Prison population: 31 December 2018, London, Ministry of Justice.

Morgan, P. & P. Henderson (1998) Remand Decisions and Offending While on Bail, London, Home Office.

Morgan, P. (1992) Offending While on Bail: A Survey of Recent Studies, London, Home Office.

Morris, J., MLC (2018) 'Second reading speech: Serious Offenders Bill 2018', Debates, Victoria, Legislative Council, 7 June.

Morris, J., MLC (2018) 'Second reading speech: Victims and Other Legislation Amendment Bill 2018', Debates, Victoria, Legislative Council, 23 August.

Mueller-Smith, M. (2015) 'The criminal and labor market impacts of incarceration', unpublished working paper, August.

Myers, N. (2017) 'Eroding the presumption of innocence: Pre-trial detention and the use of conditional release on bail', British Journal of Criminology, 57(3).

National Congress of Australia's First Peoples (2017) Submission to the Australian Law Reform Commission on Incarceration Rates of Aboriginal and Torres Strait Islander Peoples, Strawberry Hills, National Congress, September.

New South Wales Law Reform Commission (2012) Bail Report 133, Sydney, New South Wales Law Reform Commission.

O'Malley, P. (2010) Crime and Risk, London, Sage.

Ondarchie, C., MLC (2018) 'Government Performance', Debates, Victoria, Legislative Council, 19 September.

Ondarchie, C., MLC (2018) 'Second reading speech: Bail Amendment (Stage Two) Bill 2017', Debates, Victoria, Legislative Council, 22 February.

Pakula, M., Attorney-General (2017) Community Safety the Priority in Bail System Overhaul, media release, 8 May.

Pakula, M., Attorney-General (2017) New Police Remand Powers Under Bail Reform, media release, 12 December.

Pretrial Justice Institute (2015) 'The D.C. Pretrial Services Agency: Lessons from five decades of innovation and growth', Case Studies, 2(1).

Productivity Commission (2019) 'Report on Government Services 2019', Productivity Commission website.

Rahman, S. (2019) 'The marginal effect of bail decisions on imprisonment, failure to appear and crime', No.224, Sydney, NSW Bureau of Crime Statistics and Research.

Ramsay, S., MLC (2018) 'Second reading speech: Bail Amendment (Stage Two) Bill 2017', Debates, Victoria, Legislative Council, 22 February.

Ramsay, S., MLC (2018) 'Second reading speech: Serious Offenders Bill 2018', Debates, Victoria, Legislative Council, 7 June.

Rich-Phillips, G., MLC (2018) 'Second reading speech: Justice Legislation Miscellaneous Amendment Bill 2018', Debates, Victoria, Legislative Council, 18 September.

Rich-Phillips, G., MLC (2018) 'Second reading speech: Bail Amendment (Stage Two) Bill 2017', Debates, Victoria, Legislative Council, 22 February.

Roulston, R. P.(1972) 'The quest for balance in bail: The New South Wales experience', in D. Chappell & P. Wilson (eds) The Australian Criminal Justice System, Sydney, Butterworths.

Roy Morgan (2018) 'Break Bail, Go to Jail' – 82% of Victorians Agree with Matthew Guy, public opinion poll, Finding no. 7813, Roy Morgan website.

Schwartzapfel, B. & B. Keller (2015) 'Willie Horton revisited', The Marshall Project, 13 May.

Sheed, S., MLA (2018) Justice Legislation Amendment (Unlawful Association and Criminal Appeals) Bill 2018, Debates, Victoria, Legislative Assembly, 4 September.

Siddiqi, Q. (2009) Predicting the Likelihood of Pretrial Failure to Appear and/or Re-Arrest for a Violent Offense Among New York City Defendants: An Analysis of the 2001 Dataset, New York, Criminal Justice Agency.

Snowball, L., L. Roth & D. Weatherburn (2010) Bail Presumptions and Risk of Bail Refusal: An Analysis of the NSW Bail Act, Bureau Brief, no. 49, Sydney, NSW Bureau of Crime Statistics and Research.

Statute Law Revision Committee (1975) Report From the Statute Law Revision Committee upon Bail Procedures, Melbourne, Parliament of Victoria.

Storey, H. (1977) 'Second Reading Speech: Bail Bill', Debates, Victoria, Legislative Council, 15 March.

Storey, H. (1977) 'Second Reading Speech: Bail Bill', Debates, Victoria, Legislative Council, 19 April.

The Leadership Conference Education Fund (2019) 'Pretrial Risk Assessments', The Leadership Conference Education Fund website.

The Police Association of Victoria (2017) Submission to the Review of Victoria's Bail System, East Melbourne, The Police Association of Victoria.

Victoria, H., MLA (2018) Second Reading Speech: Serious Offenders Bill 2018, Debates, Victoria, Legislative Assembly, 23 May.

Victorian Law Reform Commission (2007) Review of the Bail Act: Final Report, Melbourne, Victorian Law Reform Commission.

Wells, J. (1978) 'A Hell for Innocents', The Age, 7 October.

Yeong, S. & S. Poynton (2018) Did the 2013 Bail Act increase the rate of bail refusal?, Bulletin, no.212, Sydney, NSW Bureau of Crime Statistics and Research.

Legislation

Bail Act 1977 (Vic)

Bail (Amendment) Act 1981 (Vic)

Bail (Amendment) Act 1986 (Vic)

Bail Amendment Act 2013 (Vic)

Bail Act 1980 (Qld)

<href="#/view/act/2013/26">Bail Act 2013 (NSW)

Legal Aid, Sentencing and Punishment of Offenders Act 2012 (UK)

Cases

Woods v DPP[footnote 2014] VSC 1

Robinson v The Queen [footnote 2015] VSCA 161

 

Research & Inquiries Service

Research papers are produced by the Library's Research & Inquiries service. They provide in-depth coverage and detailed analysis of topics of interest to Members of Parliament. Research publications are prepared for Parliament in response to Members, and in anticipation of their requirements, in areas of interest to the Victorian legislature.

Information in this paper was current as at the time of printing. It should not be considered as a complete guide to the particular subject or legislation covered. While it is intended that all information provided is accurate, it does not represent professional legal opinion. Any views expressed are those of the author(s).

Some hyperlinks may only be accessible on the Parliament of Victoria's intranet. All links are current and available as at the time of publication.

Enquiries:
Jon Breukel

Coordinator, Research & Inquiries

Victorian Parliamentary Library & Information Service

Parliament House

Spring Street, Melbourne

Telephone (03) 9651 8633
www.parliament.vic.gov.au

 


[footnote 1] Several commentators have identified an incarceration crisis in Australia; their analyses are applicable to Victoria. See Baz Dreisinger, quoted by Hayley Gleeson (2019) 'Australia Must 'Radically Re-Think' Its Prisons to Avoid Becoming Like America, U.S Activist Says', ABC News, 30 April; Cape York Institute (2019) The Indigenous Incarceration Crisis: The Queensland Productivity Response is Inadequate, CYI Further Submission to QPC Imprisonment and Recidivism Inquiry, April, Cairns, Cape York Institute, p.14.

[footnote 2] Remanded prisoners are also known as unsentenced prisoners. They comprise those who await conviction and have been remanded into custody (remanded), convicted but awaiting sentencing (remanded for sentencing) and remanded pending appeal. A small number of persons awaiting deportation are also included in this category.

[footnote 3] Corrections Victoria (2019a) Monthly Time Series Prisoner and Offender Data, Melbourne, Department of Justice and Community Safety, Table 1 – Prisoner numbers.

[footnote 4] ibid.; see also Corrections Victoria (2019b) Women in the Victorian Prison System, Melbourne, Department of Justice and Community Safety, January, p. 4­5.

[footnote 5] S. L. Desmarais & E. M. Lowder (2019) Pretrial Risk Assessment Tools: A primer for judges, prosecutors, and defense attorneys, Chicago, MacArthur Foundation.

[footnote 6] Other contributors to increased pre-trial detention are heavier sanctions for breach of bail conditions (resulting in incarceration) and court backlogs which result in increased time being spent on remand.

[footnote 7] Gleeson (2019) op. cit., Cape York Institute (2019) op. cit.

[footnote 8] R.P. Roulston(1972) 'The quest for balance in bail: The New South Wales experience', in D. Chappell & P. Wilson (eds) The Australian Criminal Justice System, Sydney, Butterworths, p. 469.

[footnote 9] Bail statutes refer to an 'unacceptable risk' that the applicant will, if released from custody, 'commit a serious offence' or endanger the safety of victims, individuals or the community: Bail Act 2013 (NSW) s 17(2)(b), (c);) or 'commit an offence': Bail Act 1977 (Vic) s 4E(1)(a)(ii); Bail Act 1980 (Qld) s 16(1)(a)(ii)(A).

[footnote 10] Nearly all discussion of future offending by bail applicants is restricted to considering the issue of offending in the community if released on bail. There is almost no consideration of the issue of offending in custody if persons are remanded. While this article also focuses on the issue of offending in the community, I acknowledge that offending by those remanded into custody also warrants attention.

[footnote 11] When compared with citizens of other OECD countries, Australians report relatively high levels of fear of crime; see A. Bushnell (2017) Australia's Criminal Justice Costs: An International Comparison, Sydney, Institute of Public Affairs, pp. 14–15.

[footnote 12] M. Bagaric et al. (2018) 'Mitigating America's mass incarceration crisis without compromising community protection: Expanding the role of rehabilitation in sentencing', Lewis & Clark Law Review, 22(1), pp. 3–60.

[footnote 13] M. Feeley & J. Simon (1992) 'The New Penology: Notes on the Emerging Strategy of Corrections and Its Implications', Criminology, 30(4),pp. 452–455;P. O'Malley (2010) Crime and Risk, London, Sage, Chapter One.

[footnote 14] Similar developments have occurred in New South Wales. The former Attorney-General, John Hatzistergos, was commissioned in 2013 to review bail law in that State with community safety being paramount. By comparison, reform of bail in Western Australia, the Northern Territory, the Australian Capital Territory and Tasmania has been much more limited. For a summary of recent developments in bail law in the various States and Territories of Australia see L. Bartels, K. Gelb et al. (2018) 'Bail, risk and law reform: A review of bail legislation across Australia', Criminal Law Journal, 42(1), p. 91.

[footnote 15] Corrections Victoria (2019a) op. cit.

[footnote 16] ibid.

[footnote 17] The Productivity Commission estimated that the cost per prisoner per day in Victoria in 2018 was $323.82: Productivity Commission (2019) 'Report on Government Services 2019', Productivity Commission website, Chapter 8, Table 8A.18. With data from May 2019 indicating that 3,004 persons were detained on remand, this gives a total 'ball-park' cost per day of $972,755.28. However, not all of this will be additional costs; individuals who subsequently are convicted and given a custodial sentence will have time served on remand counted towards their sentences, so the costs are primarily associated with those who are subsequently found not guilty or who are given a non-custodial sentence.

[footnote 18] Ministry of Justice (2019) Offender Management Statistics Bulletin, England and Wales, Quarterly July to Sept 2018, Prison population: 31 December 2018, London, Ministry of Justice, p. 2.

[footnote 19] In the past Victoria had relatively low remand rates compared with other States and Territories in Australia: see M. Ericson & T. Vinson (2010) Young People on Remand in Victoria: Balancing Individual and Community Interests, Richmond,Jesuit Social Services, p. 17.

[footnote 20] Australian Bureau of Statistics (2018) Prisoners in Australia, 2018: State and Territory Profiles, Victoria, cat. no. 4517.0, Canberra, ABS.

[footnote 21] Corrections Victoria (2019a) op. cit.; Figure calculated from data presented in Table 1.

[footnote 22] ibid.; Figure calculated from data presented in Table 1.

[footnote 23]Corrections Victoria (2019b) op. cit., p. 11.

[footnote 24] The most common alleged offences of females on remand were drug offences (20 per cent), burglary (15 per cent) and other property offences (23 per cent): Corrections Victoria (2019b), ibid., p. 4.

[footnote 25] E. Johnson (1991) Royal Commission into Aboriginal Deaths in Custody, National Report, 1, [footnote 2.5.2].

[footnote 26] National Congress of Australia's First Peoples (2017) Submission to the Australian Law Reform Commission on Incarceration Rates of Aboriginal and Torres Strait Islander Peoples, Strawberry Hills, National Congress, September, pp. 6–7. However, the National Congress also endorsed the primacy of community protection in bail law, observing that 'the requirement to consider issues that arise due to the person's Aboriginality would not supersede considerations of community safety' (p. 6).

[footnote 27] Seventeen per cent of the women who entered remand in Victoria in 2017 were Aboriginal or Torres Strait Islander: Corrections Victoria (2019b), op. cit., p. 4.

[footnote 28] D. Brown & J. Quilter (2014) 'Speaking too Soon: The sabotage of bail reform in New South Wales', International Journal for Crime, Justice and Social Democracy, 3(3), p. 89. More generally, see New South Wales Law Reform Commission (2012) Bail Report 133, Sydney, New South Wales Law Reform Commission, [footnote 5.42-5.44] (hereafter 'NSWLRC'); P. Heaton et. al. (2017) 'The downstream consequences of misdemeanor pretrial detention', Stanford Law Review, 69(3), p. 722; A. Allan et al (2003) 'The relationship between bail decision-making and legal representation in the criminal justice system', Edith Cowan University, pp. 7–11; C. Yang (2017) 'Toward an optimal bail system', New York University Law Review, 92(5), p. 1399.

[footnote 29] This has been attributed to pre-trial detention weakening the person's bargaining position in pre-trial negotiations: W. Dobbie, J. Goldin & C. Yang (2018) 'The effects of pre-trial detention on conviction, future crime, and employment', American Economic Review, 108(2), p.201; see also Laura and John Arnold Foundation (2013) 'Pretrial Criminal Justice Research', Houston and New York, Laura and John Arnold Foundation.

[footnote 30] ibid.

[footnote 31] ibid. Concluding that these adverse consequences occur because evidence that was influential at a bail hearing was similarly influential at a subsequent criminal hearing is questionable, given that bail decisions are often made on the basis of very limited information: M. King (1973) Bail or Custody, London, The Cobden Trust. It also disregards the role of bail in plea negotiations—several researchers have described a process of 'bail bargaining' engaged in by police whereby they use the threat of refusal of bail to secure admissions or pleas of guilty by accused. See A. Bottoms & J. McClean (1976) Defendants in the Criminal Process, London, Routledge & Kegan Paul; G. Kellough and S. Wortley (2002) 'Remand for plea', British Journal of Criminology, 42(1), pp. 186–210; King (1973), ibid, p. 33.

[footnote 32] Laura and John Arnold Foundation (2013) op. cit.

[footnote 33] J. Galouzis and S. Corben (2016) 'Judicial outcomes of remand inmates in New South Wales', Research Bulletin No.34, Sydney, Corrective Services NSW, p.13.

[footnote 34] NSWLRC (2012) op. cit., [footnote 5.23-5.31]. While it might be contended that the figure in relation to non-custodial sentences overstates the situation because Magistrates might not impose a custodial sentence on the basis of time spent in custody on remand, proper sentencing practice requires the backdating of a custodial sentence to take into account the time spent on remand (rather than imposing a non-custodial sentence in lieu). The NSW Law Reform Commission observed that is 'reasonable to assume that this happens in most cases' (at [footnote 5.30]).

[footnote 35] The current report does not mean to suggest that bail is a simple, 'one-time' process but acknowledges that bail/remand into custody is typically a more complex process where bail may be granted or revoked at multiple points in a person's contact with the criminal justice system.

[footnote 36] Statute Law Revision Committee (1975) Report From the Statute Law Revision Committee upon Bail Procedures, Melbourne, Parliament of Victoria. The most common reason bail was denied was due to the seriousness of the offences with which the person was charged; while this could be viewed as a proxy measure of dangerousness, given the primacy of concern for the person's likelihood of attending court for the hearing of their case it was unlikely that it was related to risk of future offending. A later study by researchers in Western Australia, based on cases from 2001, reported an interesting finding: the likelihood of offending if released on bail was identified by Magistrates as a relevant factor in only four per cent of cases but when the issue was raised it usually resulted in bail being denied: Allan et al (2003) op. cit., p.42.

[footnote 37] For instance, several statutes refer to an 'unacceptable risk' that the applicant will, if released from custody, 'commit a serious offence' or endanger the safety of victims, individuals or the community (Bail Act 2013 (NSW) s 17(2)(b), (c);) or 'commit an offence': Bail Act 1977 (Vic) s 4E(1)(a)(ii); Bail Act 1980 (Qld) s 16(1)(a)(ii)(A).

[footnote 38] Legislation in Victoria expressly provided that no more onerous conditions were to be imposed on the accused than was warranted by the public interest: Bail Act 1977 (Vic) ss. 5(1), 9(3).

[footnote 39] See the Hon. P. Coghlan (2017a) Bail Review: First Advice to the Victorian Government, report prepared for the Victorian government, Melbourne, Victorian Government; the Hon. P. Coghlan (2017b) Bail Review: Second Advice to the Victorian Government, report prepared for the Victorian government, Melbourne, Victorian Government.

[footnote 40] Information provided to the Statute Law Revision Committee in 1975 revealed the pernicious effect of requiring applicants for bail to obtain a surety; of the 88 men then held on remand at Pentridge Prison, 50 (57 per cent) had actually had bail set but were in custody because they couldn't raise the money or contact a relevant person: Statute Law Revision Committee (1975) op. cit.

[footnote 41] Bail Act 1977 (Vic) s 4. In Woods v DPP [footnote 2014] VSC 1, [footnote 30] Bell J stated the traditional position: 'without in any way doubting the importance of the other considerations, the primary purpose of bail is to ensure the attendance of the accused at his or her trial and the associated preliminary hearings'. However, the situation changed with the enactment of the bail reforms of 2017 and 2018; see Part 4.

[footnote 42] Statute Law Revision Committee (1975) op. cit. p. 9.

[footnote 43] R. Eddy (1977) 'Second Reading Debate: Bail Bill', Debates, Victoria, Legislative Council, 19 April, p. 7462.

[footnote 44] For instance, contemporary newspaper reports cited experts such as the Reverend Charles Bailey, a chaplain at Pentridge Prison, who referred to 'the hypocrisy of a society which smugly presumes innocence of its citizens until proven guilty of a crime and yet brutally punishes men before they are even brought to trial': J. Wells (1978) 'A Hell for Innocents', The Age, 7 October.

[footnote 45] H. Storey (1977) 'Second Reading Speech: Bail Bill', Debates, Victoria, Legislative Council, 15 March, p. 6340.

[footnote 46] ibid.

[footnote 47] Statute Law Revision Committee (1975) op. cit. p. 7.

[footnote 48] Factors to be taken into account when determining 'unacceptable risk' were the nature and seriousness of the offence; the character, antecedents, associations, home environment and background of the accused person; the history of any previous grants of bail to the accused person; and the strength of the evidence against the accused: Bail Act 1977 (Vic) s 4(3).

[footnote 49] Bail Act 1977 (Vic) s 5(2).

[footnote 50] Background information provided to the Committee about persons held on remand in Pentridge Prison revealed that the most common reason was due to the seriousness of the offences with which they were charged; only two of the 38 remandees who had been refused bail (five per cent) were detained because they were a 'danger to the community:' Statute Law Revision Committee (1975) op. cit. p. 9 (Table 1).

[footnote 51] H. Storey (1977) 'Second Reading Speech: Bail Bill', Debates, Victoria, Legislative Council, 19 April, p. 7465. The Statute Law Revision Committee had previously acknowledged that 'there would be times when it would be quite wrong to release persons' but provided no further commentary on the matter: Statute Law Revision Committee (1975) op. cit.; H. Hamilton (1977) 'Second Reading Debate: Bail Bill', Debates, Victoria, Legislative Council, 19 April, p. 7464, and recommended that a court should exercise its discretion on this issue: Statute Law Revision Committee (1974) op. cit. p. 11. Future offending was particularly related to offences involving the use of firearms or violence. H. Storey (1977) 'Second Reading Speech: Bail Bill', Debates, Victoria, Legislative Council, 15 March, p. 6340.

[footnote 52] The reverse onus provision also applied where there was a heightened flight risk - where the bail applicant had been charged with an indictable offence and was not normally resident in Victoria. As this issue is not the focus of the current report, it is not further investigated.

[footnote 53] Bail Act 1977 (Vic) s 4(4). It appears that this provision was driven by then contemporary concerns about recidivist armed robbers targeting banks. A similar concern had arisen in New South Wales where the Bail Review Committee that developed the first comprehensive Bail Act in that state was influenced by the fatal shooting of a bank manager during a robbery committed by a person who was already on bail for armed robbery: J. Miles (1991) 'Bail Legislation: Objectives and Achievements', paper presented to the Australian Institute of Criminology Conference, Canberra, 29 Nov–1 Dec 1988, p. 37.

[footnote 54] Miles (1991) op. cit., p. 40.

[footnote 55] D. Bamford, S. King and R. Sarre (1999) Factors Affecting Remand in Custody: A Study of Bail Practices in Victoria, South Australia and Western Australia, Canberra, Australian Institute of Criminology, p. 49.

[footnote 56] ibid, p. 40.

[footnote 57] ibid.

[footnote 58] Corrections Victoria (2019a) op. cit., Table 1: End of Month Prisoner Numbers by Gender and Warrant Status.

[footnote 59] Australian Bureau of Statistics (2014) op. cit.

[footnote 60] In 2018, in response to questioning about the safety of Victorians, Premier Andrews observed that Victoria had 'more people on remand today than there has been at any point in the State's history'; D. Andrews, Premier of Victoria (2018) 'Questions without Notice and Ministers Statements', Debates, Victoria, Legislative Assembly, 19 September, p. 3405.

[footnote 61] M. Pakula, Attorney-General (2017) Community Safety the Priority in Bail System Overhaul, media release, 8 May.

[footnote 62] The original Act specified thatonly the Supreme Court could consider applications for bail for those charged with murder or treason: Bail Act 1977 (Vic) s 4(2)(a). Some commentators believed that it was unclear whether this provision simply restricted bail applications to this court or whether it also imposed a higher hurdle for bail. It constituted an exception ('a court shall refuse bail': s 4(2)) but no additional standard was required to be employed by the Supreme Court when considering these applications. This was later remedied when these offences were specified as requiring 'exceptional circumstances' for the granting of bail: Bail (Amendment) Act 1981 (Vic) s. 3.

[footnote 63] For instance, s 4(c) of the Bail (Amendment) Act 1986 (Vic) required applicants charged with certain serious drug offences to demonstrate exceptional circumstances in order to be granted bail and ss 4(d),(e) expanded the operation of the 'show cause' category.

[footnote 64] Bail (Amendment) Act 1981 (Vic). A person charged with an offence of growing, preparing, manufacturing, selling, dealing in or trafficking in opium, cocaine or cannabis was now required to 'show cause' to be granted bail.

[footnote 65] Bail Act 1977 (Vic) ss 30A, 30B, as inserted by Bail Amendment Act 2013 (Vic) s 8. Immediately prior to this period there were only two secondary offences related to bail: failing to answer bail and failing to notify of a change of address.

[footnote 66] Bail Act 1977 (Vic) s 30A.

[footnote 67] Bail Act 1977 (Vic) s 30B.

[footnote 68] Prior to the introduction of bail legislation, only the Supreme Court could impose conditions on bail (in addition to the basic requirement of a deposit). Bail decision makers can now impose conditions that include, inter alia, drug treatment, electronic monitoring, curfews and geographical exclusion zones. While these conditions must be 'reasonable' and not unnecessarily onerous (Bail Act 1977 (Vic) ss 5AA(2)(a), (b)) they are a potent risk management technique.

[footnote 69] Bail Act 1977 (Vic) s 5AA.

[footnote 70] Coghlan (2017a) op. cit., p. 13.

[footnote 71] Coghlan (2017a) op. cit., pp.6-12; Coghlan (2017b) op. cit., pp. 8–11.

[footnote 72] ibid.

[footnote 73] Coghlan (2017a) op. cit., p.3.

[footnote 74] For instance, see Coghlan (2017a) op. cit., p.25.

[footnote 75] For instance, a poll conducted in November 2018 reported that 82 per cent of Victorians agreed that a person who breached a condition of bail should immediately go to jail: Roy Morgan (2018) 'Break Bail, Go to Jail' – 82% of Victorians Agree with Matthew Guy, public opinion poll, Finding no. 7813, Roy Morgan website.

[footnote 76] M. Pakula, Attorney-General (2017) New Police Remand Powers Under Bail Reform, media release, 12 December.

[footnote 77] The Coghlan Review had recommended that the 'show cause' reverse onus test be replaced by another test; however, the test that was introduced was the 'show compelling reason' test: Bail Act 1977 (Vic) ss 4C, 4D.

[footnote 78] Bail Act 1977 (Vic) s 1B(1)(a) [footnote emphasis added].

[footnote 79] G. Rich-Phillips, MLC (2018) 'Second reading speech: Bail Amendment (Stage Two) Bill 2017', Debates, Victoria, Legislative Council, 22 February, p. 512; C. Ondarchie, MLC (2018) 'Second reading speech: Bail Amendment (Stage Two) Bill 2017', Debates, Victoria, Legislative Council, 22 February, p. 539.

[footnote 80] J. Morris, MLC (2018) 'Second reading speech: Serious Offenders Bill 2018', Debates, Victoria, Legislative Council, 7 June, p. 2521; S. Ramsay (2018) 'Second reading speech: Serious Offenders Bill 2018', Debates, Victoria, Legislative Council, 7 June, p. 2529; R. Clark, MLA (2018) 'Second reading speech: Serious Offenders Bill 2018', Debates, Victoria, Legislative Assembly, 23 May, p. 1525; H. Victoria, MLA (2018) 'Second reading speech: Serious Offenders Bill 2018', Debates, Victoria, Legislative Assembly, 23 May, p. 1553.

[footnote 81] For instance, Mr Mathew Guy (the leader of the opposition in Victoria) raised the case of Paulo Kele, charged with home invasion and theft, and released on bail: M. Guy, MLA (2018) 'Questions without Notice and Ministers Statements', Debates, Victoria, Legislative Assembly, 19 September, p. 3405.

[footnote 82]G. Rich-Phillips, MLC (2018) 'Second reading speech: Justice Legislation Miscellaneous Amendment Bill 2018', Debates, Victoria, Legislative Council, 18 September, p. 5009.

[footnote 83] C. Ondarchie, MLC (2018) 'Government performance', Debates, Victoria, Legislative Council, 19 September, p. 5075.

[footnote 84] ibid.

[footnote 85] ibid.

[footnote 86] Independent Member of the Legislative Assembly, Ms Suzanna Sheed, countered the narratives of 'extremely dangerous' by describing several cases, including that of a 37-year-old mother of five who spent 68 days on remand for allegedly stealing a handbag from Myer; she was subsequently convicted and released on a good behaviour bond: S. Sheed, MLA (2018) 'Justice Legislation Amendment (Unlawful Association and Criminal Appeals) Bill 2018', Debates, Victoria, Legislative Assembly, 4 September, p. 3107.

[footnote 87] For instance, see: W. Gatt (2017) 'Wayne Gatt: Bail Has Lost its Sense of Balance', The Herald Sun, 8 March; R. Cavanagh (2018) 'Teen Thug Released Despite Being on Bail When Sexually Assaulting Two Women', The Herald Sun, 1 January.

[footnote 88] The Attorney-General, Mr Pakula, commented: 'Strengthening the bail tests will ensure that risk to community safety is given a higher priority when deciding whether to grant bail'; M. Pakula, Attorney-General (2017) New Police Remand Powers Under Bail Reforms, media release, 12 December.

[footnote 89] Rich-Phillips (2018) op. cit., p. 5009.

[footnote 90] M. Fitzherbert, MLC (2017) 'Second reading speech: Bail Amendment (Stage Two) Bill 2017', Debates, Victoria, Legislative Council, 22 February, p. 543.

[footnote 91] C.f. Ms Sheed (2018) op. cit.

[footnote 92] Ms Kealy MLA stated: 'The penalty for raping and murdering an innocent woman should be harsh and strong, with no soft-touch approach of bail…': E. Kealy, MLA (2018) 'Condolences', Debates, Victoria, Legislative Assembly, 19 June, p. 1936.

[footnote 93] For instance, Mr Morris MLC criticised the granting of bail to an individual charged with a violent crime in Frankston and referred more generally to 'thugs' being granted bail: J. Morris, MLC (2018) 'Second reading speech: Victims and Other Legislation Amendment Bill 2018', Debates, Victoria, Legislative Council, 23 August, p. 4317. Mr Ramsay referred to inadequate 'punishment that is being applied through the bail system': S. Ramsay, MLC (2018) 'Second reading speech: Bail Amendment (Stage Two) Bill 2017', Debates, Victoria, Legislative Council, 22 February, p. 541.

[footnote 94] Bail Act 1977 (Vic) s 3 AAA(1)(h).

[footnote 95] ibid.

[footnote 96] ibid.

[footnote 97] ibid., s 3 AAA(1).

[footnote 98] ibid., Schedule 1.

[footnote 99] ibid., Schedule 2.

[footnote 100] The increased risk (demonstrated by the later charges) can be constituted by a summary or indictable offence and there is no requirement that this alleged offence has any particular nexus with the earlier offending. For instance, a person on a supervision order following conviction for a sex offence who is then charged with theft is deemed to present a 'higher level of risk' which warrants them having to show 'compelling reasons' to be granted bail. Similarly, there is no requirement of temporal proximity in the offences. So, a person who is a sex offender on a supervision order and who is now charged with a social security offence committed a decade ago will have to show 'compelling reasons' why he should not be detained in custody from the time the fraud charge was laid until the completion of his trial.

[footnote 101] While New South Wales has both 'show cause' and 'exceptional circumstances' tests, the latter is restricted to just one category of offence (terrorism charges under Commonwealth law): see the Bail Act 2013 (NSW) s 22A (exceptional circumstances); ss 16A, 16B (show cause). Tasmania has no presumptions against bail. Other jurisdictions either have the equivalent of the 'exceptional circumstances' or the 'compelling reasons' provisions and generally include fewer offences in the respective categories.

[footnote 102] Bartels et al (2018) op. cit.

[footnote 103] Sixty-three per cent of the women were charged with property or drug offences: Corrections Victoria (2019b) op. cit., p. 4.

[footnote 104] H. Storey, op. cit.

[footnote 105] Coghlan (2017b) op. cit., pp. 62–67.

[footnote 106] E.g. Bamford, King and Sarre (1999) op. cit., p. 104; A. Hucklesby (2009) 'Keeping the lid on the prison remand population', Current Issues in Criminal Justice, 21(1), p. 6; S Rahman (2019) The marginal effect of bail decisions on imprisonment, failure to appear and crime,Bulletin, no.224, Sydney, NSW Bureau of Crime Statistics and Research, p. 2.

[footnote 107] R. Clark, MLA (2018) 'Second reading speech: Serious Offenders Bill 2018', Debates, Victoria, Legislative Assembly, 23 May, p. 1523.

[footnote 108] G. Rich-Phillips, MLC (2018) 'Second reading speech: Justice Legislation Miscellaneous Amendment Bill 2018', Debates, Victoria, Legislative Council, 18 September, p. 5009.

[footnote 109] A. Allan et al (2003) op. cit., p. 59.

[footnote 110] Interestingly, research from England and New Zealand suggests that the proportion of people on bail who commit offences has not increased in those jurisdictions since the introduction of the modern bail statutes; consequently, increased concern about offending while on bail cannot simply be explained as a reaction to an escalating problem of crimes committed by those on bail. A study by Henderson and Nichols in England found that there had been little change in the proportion of those who offended while on bail between 1978 and 1988: P.F. Henderson & T. Nichols (1992) Offending while on bail, Research Bulletin, no. 32, London, Home Office, pp. 23–27. Similarly, a report from New Zealand showed that between 1990 and 1999, the proportion of people who committed an offence while on bail changed very little: B. Lash (2003) Trends in the Use of Bail and Offending While on Bail, 1990–1999, Wellington, New Zealand Ministry of Justice. However, it should be noted that legal reforms may impact on this issue. For example, in Victoria the introduction of the offence of breaching bail conditions in 2011 has made a potent contribution to the incidence of offending while on bail: see Crime Statistics Agency (2018) 'Spotlight: Breaches of Orders – The Impact of Legislative Changes', Crime Statistics Agency website.

[footnote 111] P. Morgan & P. Henderson (1998) Remand Decisions and Offending While on Bail, London, Home Office, pp. 44–55. A re-calculation of figures presented in Table 5.1 (p. 44) gives an overall rate of convictions while on bail of 12 per cent.

[footnote 112] P. Morgan (1992) Offending While on Bail: A Survey of Recent Studies, London, Home Office.

[footnote 113] K. Brown et al. (2004) Offending On Bail: An Analysis of the Use and the Impact of Aggravated Sentences for Bail Offenders, Edinburgh, Scottish Government, Chapter 4.

[footnote 114] Q. Siddiqi (2009) Predicting the Likelihood of Pretrial Failure to Appear and/or Re-Arrest for a Violent Offense Among New York City Defendants: An Analysis of the 2001 Dataset, New York, Criminal Justice Agency.

[footnote 115] Ministry of Justice (2018) 'Offending on Bail', factsheet, New Zealand Government website.

[footnote 116] ibid. This research also found that different risks may be unrelated; for instance, the risk of failing to appear was independent of the risk of offending during the bail period.

[footnote 117] M. Dhami & P. Ayton (2001) 'Bailing and jailing the fast and frugal way', Journal of Behavioral Decision Making, 14(2), pp. 141–168.

[footnote 118] ibid.

[footnote 119] Anthea Hucklesby noted a similar process in England where, during the 1990s, debates about bail and offending were dominated by 'specific atypical cases' where individuals on bail committed serious crimes that were extensively covered by the media: A. Hucklesby (1996) 'Bail or jail? The practical operation of the Bail Act 1976', Journal of Law and Society, 23(2), pp. 213–233. In relation to a similar problem in reforming jurisdictions in the United States (especially New Jersey), see: (2018) 'Note: Bail Reform and Risk Assessment: The Cautionary Tale of Federal Sentencing', Harvard Law Review, 131(1), pp. 1125–1146.

[footnote 120] J. Hatzigeros (2014) Review of the Bail Act 2013, report prepared for the Department of Justice, Sydney, NSW Government, p. 8.

[footnote 121] ibid, [footnote 227]–[footnote 228].

[footnote 122] Bail Act 1977 (Vic) s. 5AAA(1).

[footnote 123] The New South Wales Law Reform Commission, after noting the increase in remand rates in that State, observed that 'Such evidence as there is does not suggest an effect in reducing crime': NSWLRC (2012) op. cit., pp. 62–63.

[footnote 124] Pre-trial release has been associated with short-term increased offending: Dobbie, Goldin and Yang (2018) op. cit. However, as was noted in the preceding part of this report dealing with offending while on bail, most of the offending is non-violent and relatively minor.

[footnote 125] NSWLRC (2012) op. cit., [footnote 5.18–5.22].

[footnote 126] P. Heaton et al. (2017) op. cit., pp. 714–716. Controlling for these factors in the analyses allowed the researchers to conclude that any differences between the two groups were not simply due to factors unrelated to pre-trial detention. For instance, it is prima facie arguable that a higher recidivism rate could be expected for those who were detained compared to those who were released because the bail process accurately identified risk (and detained the high-risk individuals who subsequently affirmed their high-risk status by further offending). But the authors' analysis compared defendants who had been assessed of equal risk (their bail was set identically). Thus, the differences in recidivism rates of the two groups were for groups that had been matched on risk and differed in relation to whether they had obtained bail or been detained in custody.

[footnote 127] Dobbie, Goldin & Yang (2018) op. cit., pp. 234–235.

[footnote 128] P. Heaton et al. (2017) op. cit., p. 718. See also M. Mueller-Smith (2015) 'The Criminal and Labor Market Impacts of Incarceration', unpublished working paper, August.

[footnote 129] Cf M. Fitzherbert, MLC (2018) 'Second reading speech: Bail Amendment (Stage Two) Bill 2017', Debates, Victoria, Legislative Council, 22 February, p. 543.

[footnote 130] Ms Kealy, MLA stated: 'The penalty for raping and murdering an innocent woman should be harsh and strong, with no soft-touch approach of bail…' E. Kealy, MLA (2018) 'Condolences', Debates, Victoria, Legislative Assembly, 19 June, p 1936.

[footnote 131] For instance, Mr Morris MLC criticised the granting of bail to an individual charged with a violent crime in Frankston and referred more generally to 'thugs' being granted bail: J. Morris, MLC (2018) 'Second reading speech: Victims and Other Legislation Amendment Bill 2018', Debates, Victoria, Legislative Council, 23 August, p. 4317. And Mr Ramsay, MLC referred to inadequate 'punishment that is being applied through the bail system': S. Ramsay, MLC (2018) 'Second reading speech: Bail Amendment (Stage Two) Bill 2017' Debates, Victoria, Legislative Council, 22 February, p. 541.

[footnote 132] Coghlan (2017b) op. cit., pp. 14–28.

[footnote 133] The Legal Aid, Sentencing and Punishment of Offenders Act 2012 (UK) changed the way that people can be remanded to custody in England and Wales. It restricts the use of pre-trial detention, especially where there is no real prospect that the person would be given a custodial sentence if subsequently convicted of the offence with which they had been charged.

[footnote 134] Robinson v The Queen [footnote 2015] VSCA 161, [footnote 31].

[footnote 135] Victorian Law Reform Commission (2007) Review of the Bail Act: Final Report, Melbourne, VLRC, p. 52–54.

[footnote 136] Law Institute of Victoria (2017) Submission: Review of Victoria's Bail System, Melbourne, Law Institute Victoria, pp. 21–22.

[footnote 137] Most commentators note that the seriousness of the offence charged, the prior record of the bail applicant and any previous breaches of bail are key considerations. E.g. Bamford, King and Sarre (1999) op. cit.

[footnote 138] See Bamford, King and Sarre (1999) op. cit., p. 16. While the short duration might be expected in uncontested bail applications, even contested applications are usually of very brief duration.

[footnote 139] Dhami & Ayton (2001) op. cit.; Hucklesby (1996) op. cit.; Morgan & Henderson (1998) op. cit., p. 57.

[footnote 140] M. Heyerly (2013) Pretrial Reform in Kentucky, report prepared for Pretrial Services, Frankfort, Kentucky Court of Justice.

[footnote 141] D. Matthau (2017) 'Is NJ bail reform working, or just giving criminal suspects a free pass?', New Jersey 101.5, 8 February.

[footnote 142] Pretrial Justice Institute (2015) 'The D.C. Pretrial Services Agency: Lessons from five decades of innovation and growth', Case Studies, 2(1), pp. 1–8.

[footnote 143] Laura and John Arnold Foundation (date unknown) 'Public Safety Assessment: Risk factors and formula', Public Safety Assessment website.

[footnote 144] Laura and John Arnold Foundation (date unknown) op. cit.

[footnote 145] Bail Act 1977 (Vic) s 3AAA.

[footnote 146] It is noteworthy that the three individuals who committed serious offences while on bail and who have figured largely in media discussion of bail and risk – Sean Price, Adrian Bayley and James Gargasoulas – would, if the PSA had been administered at their last bail hearing prior to being released into the community on bail and committing their serious crimes, have been flagged as presenting a risk of new violent criminal activity (each of them would each have obtained a score higher than 4, the cut-off point for flagging new violent criminal activity risk).

[footnote 147] A survey of Victorian police officers reported that most were confused about the differences between the various risk tests and believed that there was too much variation in the way that bail decision makers interpreted and applied the tests: The Police Association of Victoria (2017) Submission to the Review of Victoria's Bail System, East Melbourne, The Police Association of Victoria.

[footnote 148] For a summary of the limitations and concerns associated with the use of actuarial approaches to risk assessment for bail see J. Koepke & D. Robinson (2018) 'Danger ahead: Risk assessment and the future of bail reform', Washington Law Review, 93(1), p.1725.

[footnote 149] On this point it is interesting to note that the related provision in s 19(2)(b) of the Bail Act 2013 in New South Wales, which refers to an unacceptable risk that the person will 'commit a serious offence' [footnote emphasis added].

[footnote 150] An outline of key concerns and relevant principles for the use of these actuarial instruments is provided in The Leadership Conference Education Fund (2019) 'Pretrial Risk Assessments', The Leadership Conference Education Fund website.

[footnote 151] A. Ashworth (2006) 'Four threats to the presumption of innocence', International Journal of Evidence and Proof, 10(4), p. 242.

[footnote 152] See N. Myers (2017) 'Eroding the presumption of innocence: Pre-trial detention and the use of conditional release on bail', British Journal of Criminology , 57(3), p. 664.

[footnote 153] More than 30 years ago, David Brown observed in relation to bail laws in New South Wales that problems in their operation may 'lie less in the Act than in police, legal and judicial practices, attitudes and interpretations which have combined to restrict its reformist potential…the reluctance of some magistrates to scrutinise critically police objections to bail and the specific problems of unrepresented defendants being cases in point': D. Brown (1987) 'Remanded in custody: A NSW refrain', Legal Services Bulletin, 12(4), p.186. Inspection of recent remand rates for the various States and Territories of Australia confirms the continuing validity of Brown's analysis (see Australian Bureau of Statistics (2018) Proportion of Unsentenced Prisoners, cat. no. 4517.0, Canberra, ABS.

Other researchers have pointed to the importance of matters other than precise statutory formulations in bail decision-making. For instance, Snowball et al.'s large-scale review of the operation of the Bail Act in New South Wales challenged the importance of the reverse onus provisions in that Act, concluding that 'legal factors pertaining to the characteristics of the case were significantly more relevant for predicting a bail decision than the presumption carried by the offence': L Snowball, L Roth & D. Weatherburn (2010) Bail Presumptions and Risk of Bail Refusal: An Analysis of the NSW Bail Act, Bureau Brief, no. 49, Sydney, NSW Bureau of Crime Statistics and Research. But this is not uncontested; a later study by Yeong and Poynton's of the same jurisdiction after subsequent restrictive reforms found that the legislative changes did increase the probability that courts would refuse bail where the bail applicant had been charged with non-minor offences: S. Yeong & S. Poynton (2018) Did the 2013 Bail Act increase the risk of bail refusal?, Bulletin, no.212, Sydney, NSW Bureau of Crime Statistics and Research; Justice T. Connolly (2006) 'Golden Thread or Tattered Fabric: Bail and the Presumption of Innocence', paper presented to the Law Council of Australia, National Access to Justice and Pro Bono Conference, Melbourne, 11–12 August.

[footnote 154] Myers (2017) op. cit., p. 682.

[footnote 155] This report has focused on risk assessment. There are multiple other strategies for reducing the number of persons held on remand, including the pragmatic suggestion that some offences should not require bail: see Coghlan (2017b) op. cit., pp. 14–28. Additionally, the increased provision of drug and alcohol treatment programs and other support services in the community could provide a stronger framework for granting bail.

[footnote 156] Coghlan (2017b) op. cit., p. 68.