Speaking Too Soon: The Sabotage of Bail Reform in New South Wales

Within just over one month of coming into operation in May 2014, the new Bail Act 2013 (NSW), a product of long‐term law reform consideration, was reviewed and then amended after talk‐back radio ‘shock jock’ and tabloid newspaper outcry over three cases. This article examines the media triggers, the main arguments of the review conducted by former New South Wales (NSW) Attorney General John Hatzistergos, and the amendments, with our analysis of the judicial interpretation of the Act thus far providing relevant background. We argue that the amendments are premature, unnecessary, create complexity and confusion, and, quite possibly, will have unintended consequences: in short, they are a mess. The whole process of reversal is an example of law and order politics driven by the shock jocks and tabloid media, the views of which, are based on fundamental misconceptions of the purpose of bail and its place in the criminal process, resulting in a conflation of accusation, guilt and punishment. Other consequences of the review and amendments process recognised in this article include the denigration of judicial expertise and lack of concern with evidence and process; the disproportionate influence of the shock jocks, tabloids and Police Association of NSW on policy formation; the practice of using retired politicians to produce ‘quick fix’ reviews; and the political failure to understand and defend fundamental legal principles that benefit us all and are central to the maintenance of a democratic society and the rule of law. The article concludes with some discussion of ways in which media and political debate might be conducted to produce more balanced outcomes.


Introduction
In a recent article in the Alternative Law Journal, David Shoebridge MP (2014: 132)  … there is cause for real hope that, come mid-2014, we will see a new and even progressive bail regime in place in NSW. This could well be one of those rare occasions where the law and order debate in NSW has been hijacked by principle and, for once, the tabloids and shock jocks haven't even noticed.
Within just five weeks of operation of the new Act these words have unfortunately turned out to be a case of speaking too soon. The tabloid newspapers and the talk-back radio shock jocks, far from 'not noticing', led a media campaign which resulted in repudiation of key features of the new bail legislation. This article will examine how and why this happened and what it means for the conduct of criminal law reform.
To set the scene, we begin with the backstory, outlining the impetus for bail reform in NSW, as reflected in the Bail report of the NSW Law Reform Commission (LRC) (2012) and the subsequent passing of the new Act with the key features of this legislation highlighted. After little more than one month in operation, however, the new Act was called into question in a media storm centred on three cases and the NSW Liberal-National Party Coalition government initiated a review, presided over by former Labor Party (the main opposing party to the Coalition in NSW) Attorney General John Hatzistergos. The article then examines the media triggers for the review, its major findings and the government response in the form of amending legislation. It examines the merits of the review arguments accepted by the government, in the light of our detailed analysis of the reported bail decisions 1 in the Supreme Court and District Court of NSW and a selection of unreported decisions 2 since the Act came into force. Based on this analysis, we argue that the review and the amendments are premature, unnecessary, create complexity and confusion and, quite possibly, will have unintended consequences including the potential to make finding 'unacceptable risk' and thereby refusing bail more difficult. The article then considers some broader issues apparent in the unfolding of these events. In conclusion, we consider the question: how might this latest chapter in the history of bail reform have played out differently? In particular, how might the media and political debate around the 'notorious' cases at the heart of the media storm have been differently inflected and contextualised? How might this important debate have been infused with a stronger commitment to taking seriously both evidence of various forms, and important legal and democratic principles?

Background
When the Coalition came to power in NSW in 2011, then Attorney General Greg Smith SC took the laudable step of reforming bail laws. As a former prosecutor with detailed inside knowledge of the operation of the criminal justice system, Smith was particularly concerned with the way bail refusal impacted on young people (Gibson 2010) and on 8 June 2011 he gave the NSW LRC a reference to 'develop a legislative framework that provides access to bail in appropriate cases' (NSW LRC 2012: xv).
The need for bail law reform had arisen out of the unwieldy complexity of the Act's predecessor, the Bail Act 1978 (NSW). This Act was subject to constant amendments and contained a series of complex presumptions against bail (NSW LRC 2012: ch 3) 3 which led to a soaring remand population in NSW (25 per cent of all prisoners, which is also the figure nationally) (Brown 2013). 4 The new Act was the product of a rigorous law reform process. The NSW LRC, headed by the highly respected Judge James Wood and retired Supreme Court Judge Hal Sperling, engaged in a painstaking review over nearly 12 months. The process involved circulation of Questions for Discussion, 5 followed by extensive consultations with all interested parties, including NSW Police, and the circulation of a preliminary draft for comment. The final Report was then considered by the Attorney General and his Department and a response was published (NSW Government 2012), foreshadowing legislation which in part followed the NSW LRC recommendations. The new Act was passed with the support of all major parties in May 2013. The legislation was set aside for 12 months to enable time for significant training, including the Judicial Commission of NSW's bail scenarios 6 and 26 training sessions run by Legal Aid NSW across that state. 7 The government's aim was simplicity and clarity: The Government anticipates that dispensing with the system of presumptions will not only simplify the bail decision making process, but will also result in fewer amendments to the legislation enabling it to remain simple and clear, as was intended when the original bail laws were codified in 1978. (NSW Government 2012: 7) Four important features of the new Act designed to enhance these aims are highlighted in the next section because each came to have greater significance as events unfolded. These are the removal of the old complexity of presumptions; the introduction of a purposes section; introduction of a two-step unacceptable risk model; and the fact that conditions can only be imposed to mitigate unacceptable risks.

The unacceptable risk model
The most significant feature of the new Act is that it does away with the complexity of the old presumptions that had plagued the Bail Act 1978 (as discussed above). While the new Act did not adopt the NSW LRC's recommendation of an explicit presumption in favour of bail for all offences, it provides in the purposes section, that a bail authority in making a bail decision is to have regard to the presumption of innocence and an accused's general right to be at liberty (s 3 (2)). As a matter of statutory construction, when construing a provision of the Act, a 'construction that would promote' those purposes is to be 'preferred to a construction that would not': see Interpretation Act 1987 (NSW) s 33.
The Act dramatically simplifies the old regime by making central the model of 'unacceptable risk' (s 17 (1)). This model requires a two-stage assessment. First, a bail authority must consider whether an accused person poses an unacceptable risk of: failing to appear at any proceedings for the offence; committing a serious offence; endangering the safety of victims, individuals or the community; or interfering with witnesses or evidence (s 17(2)(a)-(d)). In deciding whether such an unacceptable risk exists the bail authority is to have regard only to the matters set out in s 17(3) (and 17 (4) where relevant). These matters include (in s 17 (3)): the accused's person's background (a); the nature and seriousness of the offence (b); the strength of the prosecution case (c); whether the accused person has a history of violence (d); the length of time the accused person is likely to spend in custody if bail is refused (g); and any special vulnerability or needs of the accused (j). These criteria are neither novel nor new and are broadly similar to those used in s 32 of the old Bail Act 1978 regime. If no such 'unacceptable risks' exist, bail must be granted (s 18).
If there are unacceptable risks, the 'second step' or assessment must be undertaken with the bail authority considering whether those risks can be mitigated by imposing conditions. Such conditions may be requirements as to: conduct (s 25); security (s 26); character acknowledgements (s 27); accommodation (s 28); pre-release (s 29); and enforcement (s 30). Again, such conditions are not new but what is very different is that conditions can now only be imposed for the purpose of mitigating an unacceptable risk (s 24 (1)). Furthermore, such conditions must be 'reasonable, proportionate to the offence for which bail is granted, and appropriate to the unacceptable risk in relation to which they are imposed' (s 24 (3)). This was specifically designed to target problems with the old Bail Act 1978 where conditions were often imposed in a pro-forma way leading to a proliferation of conditions not appropriate to the specific alleged offence or offender, leading to extensive breaching and often the arrest of the offender and the revocation of bail (NSW LRC 2012: 191-245;Brown 2013).
If conditions can mitigate the unacceptable risk, bail is granted (s 19); only if they cannot will bail be refused (s 20 (1)). (For a useful discussion of the operation of the new regime see also Sanders 2014.)

The triggers for review
On 20 (Hatzistergos 2014: 11-13). Three are the most significant and the most damaging to the new Act's regime. In this section of the paper we offer an analysis of these three recommendations drawing on an examination of all available reported decisions (ten in total) and a selection of unreported cases.

Presumption of innocence
The first recommendation is to omit the purposes section of the Act (s 3) which includes the need 'to have regard to the presumption of innocence and the general right to be at liberty' (s 3 (2)). Instead, the Review Report recommends this be relocated to a 'preamble' to the Act which would also 'note the importance of bail decisions to community safety' (Hatzistergos 2014: [11]; Rec 1). The stated reason for this recommendation, is 'the role played by section 3 (2) has been problematic or at least confusing, particularly in its interaction with section 17(3)' (Hatzistergos 2014: [96]). The only evidence offered for this alleged confusion is a reference to one unreported decision and an even vaguer reference to 'some Local Court determinations' (Hatzistergos 2014: [96]). Our review of the cases demonstrates there is no confusion over the inclusion of the 'presumption of innocence and the general right to liberty' or that decisions are giving it undue weight: indeed, the cases often make no mention of s 3 (2) or where they do it is to recognise that this is not a new consideration (Fesus: [8]; Lago: [13] We have not found any case where the role of the presumption of innocence amounts to being 'problematic or at least confusing' as asserted in the Review Report at [96]. We have located one decision where the interaction between the presumption of innocence and s 17 is emphasised, R v Morris (Unreported, NSWSC, McCallum J, 20 May 2014, as cited in Rodger 2014. This case involved an Aboriginal woman who was charged with one larceny offence and had been in custody for two months at the time of the bail application. McCallum J found that Ms Morris's background placed her in a category of person with a special vulnerability under the Act in s 17(3)(j) for she: … came from a background of severe deprivation including her subjection to violence, sexual abuse and movement between family and foster parents. In addition, her mother was murdered when she was a teenager and she suffered from depression together with a number of physical and mental conditions. (cited in Rodger 2014: 4) With regard to the relationship between the presumption of innocence and special vulnerability her Honour stated: In determining the application, I am required to have regard to the presumption of innocence and the general right to be at liberty: s 3 of the Act. The weight of that consideration is reinforced in the present case by relevant evidence of the applicant's background which, in my assessment, plainly places her in the category of a person with special vulnerability: cf s 17 ( McCallum J found in relation to any unacceptable risk of committing a serious offence, that Ms Morris's criminal history suggested that any offences committed would be relatively minor (such as shoplifting) and so found no unacceptable risk.
We consider that in no way could such a finding suggest that s 3(2) has been 'problematic or at least confusing, particularly in its interaction with section 17(3)': on the contrary, we would argue that McCallum J's assessment was entirely sound.

Unacceptable risk test too complex?
The second significant recommendation is to collapse the two-stage test for determining 'unacceptable risk' into one (Hatzistergos 2014: [14]; Rec 2). The recommended single-step test would involve determining unacceptable risk, with an expanded list of risk factors (Hatzistergos 2014: [15]; , and by reference also to the bail conditions that might be imposed (Hatzistergos 2014: [14]; Rec 2). Three reasons underpin this recommendation: the ability to impose conditions prior to finding 'unacceptable risk' (Hatzistergos 2014: [140]); alleged problems in the application of the unacceptable risk test; and the community finding it hard 'to appreciate how an accused who was found to present an "unacceptable risk" can be safely released, even with strict bail conditions' (Hatzistergos 2014: [14]).
In relation to the first reason, one case indicates some disquiet about a finding of unacceptable risk as a pre-requisite for imposing conditions (Lago: [28]). Aside from this one instance, our examination of the cases shows no evidence of a problem regarding the imposition of conditions and more importantly no confusion over the application of the two-stage unacceptable risk test. Rather the cases are developing a body of principles regarding 'unacceptable risk', albeit the repository of such cases is in its infancy. In other words, the cases are developing principles for interpreting the term 'unacceptable risk' as it is not further defined in the Act (Lago: [5]). The cases do have recourse to the decisions in Victoria and Queensland (Hawi: [15]; Lago: [10]- [12]) (which have similar, although not identical, models 9 ) which remind us that bail is not 'risk free' (Hawi: [41]; Alexandridis: [34]) and that a 'tenuous suspicion' or 'fears of the worst possibility' (Lago: [9]) if an offender is released is not sufficient to constitute 'unacceptable risk'. Some risk will always be present given the 'unpredictability of individual behaviour' but the evaluation is whether the risk is unacceptable (Hawi: [41]). A number of judges have quite reasonably made the observation that bail decisions are inevitably an exercise in prediction and, as Justice Hamill put it, 'Bail authorities do not have a crystal ball. They are not soothsayers' (Alexandridis: [33]). This is not the product of the 2013 reforms, it has always been thus. Indeed, making predictive evaluations is not foreign to the courts, this being undertaken in the context of various other statutory regimes. 10 In deciding whether there is an unacceptable risk by reference to the matters in s 17 (3), the cases demonstrate that judges are logically working through the matters. For example, in Hawi, one of the cases at the heart of the storm that triggered the review, Justice Harrison methodically worked through each of the s 17 (3) [25]);  evaluations of the 'nature and seriousness of the offence' (s 17(3)(b)), especially if it involves extreme violence such as murder, may alone be a basis for finding unacceptable risk (Lago: [15]), however, the 'nature' of the offence even if serious may arise from a relatively unique set of circumstances that are not likely to be replicated, thereby reducing risk (Rokhzayi: [30]);  emphasis is given to the length of time an accused person will spend in custody before trial if bail is refused (s 17(3)(g)) particularly where that is likely to be more than a year (Alexandridis: [11]- [12]; This brief summary indicates that judicial consideration of the Act is very much in its infancy with further time needed to appropriately develop the principles, including appellate decisions. This was envisaged by Parliament, expressly providing for a review 'to be undertaken as soon as possible after the period of 3 years' (s 101 (3), emphasis added) -not in less than two months as has occurred -at which point statistical evidence of how the new Act is operating would also be available from the NSW Bureau of Crime Statistics and Research (Weatherburn 2014). What the cases do not demonstrate, though, is that there are significant problems or confusion with interpreting the 'unacceptable risk' test as the Review Report stated.

Serious offences and show cause
The third and most problematic recommendation is the introduction of categories of 'serious offences' (Hatzistergos 2014: 12;Rec 6) for which an accused person would have to 'show cause' as to why the accused's detention is not justified (Rec 5). The stated reason for this recommendation is to 'provide a useful level of reassurance for the community in relation to serious offenders whilst also providing a greater level of consistency' (Hatzistergos 2014: [220]). The Review Report implies (at [188]- [191]) that this is required because while bail decisions in the Supreme Court appropriately weigh the relevant s 17 (3) factors (particularly 'nature and seriousness of the offence' (s 17 (3)(b)) and 'strength of the prosecution case' (s 17 (3)(c)), the majority of bail decisions are made by police, registrars and magistrates who may not (Hatzistergos 2014: [191]).
In evaluating this recommendation, it is instructive to return to the three cases mentioned above that triggered the review and which produced the supposed level of community anxiety around the new Act. It is important to recall that, when bail was granted in each of these matters, no judgments or reasons were made public. 12 This created a vacuum in which the media dramatised the idea that the Act was 'soft on crime' and the community was potentially at risk by having such accused persons on bail. Furthermore, with no published reasons, this meant there was no possible 'counter-argument' to explain the basis for the decisions. This is to be lamented, particularly as two of the judgments were later made available which provide detailed reasons as to why conditional bail was granted. 13 Thus, the decision in Fesus indicates that new forensic evidence became available at the bail application that impacted on the strength of the prosecution case (s 17(3)(c)) and this factor, taken together with the length of time since offending (17 years), led Adams J to grant conditional bail. Justice Harrison's judgment in Hawi sets out (as noted above) the reasoned basis for granting conditional bail.
Hawi's murder conviction appeal had also been upheld by the NSW Court of Criminal Appeal, with the Chief Justice favouring a complete acquittal and the two majority judges ordering a retrial. 14 Ultimately, Hawi pleaded guilty to the manslaughter of Zervas on 5 September 2014 and bail was not opposed by the Crown (Bibby 2014a), raising serious questions about what this whole scenario was actually about. In the final matter (Ibrahim), no first instance or appeal judgment has been made available but the DPP successfully appealed the original decision to grant bail and on 27 June 2014 bail was revoked by the Supreme Court.
At the heart of media reporting of these cases lies confusion over the purpose of bail which is not about whether or not an accused person is guilty of an offence. In recent media coverage and in the way the Government has handled the situation, bail has come to symbolise 'judgment' and serve as a proxy for guilt and punishment. Denying bail -putting a person in jail before trial -has become a way of expressing condemnation of the behaviour in which a person is alleged to have engaged (Quilter 2014a).

The Government response and the Bail Amendment Act 2014
The same day as the Review Report was made public (5 August  2014b). Just over one week later, on 13 August 2014, the Bail Amendment Bill 2014 was read for a second time. In the Second Reading speech, the Attorney General expressly noted that the Bill accepts all of the recommendations of the Review Report as they '… are common-sense changes' (Hazzard 2014: 9). The Second Reading speech concludes: The Government acknowledges that the NSW Police Force, courts and legal practitioners will need some time to digest these changes. Education and training will be required, along with changes to various information management systems and bail forms. The Government recognises, however, that the changes proposed in this bill must be implemented swiftly to ensure that the Bail Act is striking the right balance in protecting the community and the integrity of the justice system. I commend the bill to the House. (Hazzard 2014:11;emphasis added) While the Government's Amendment Bill suggests an urgent need to ensure the Act is striking the right balance, the evidence (as discussed above) suggests otherwise. But the Government may get more than it bargained for with some unintended effects of the amending provisions discussed below.
The Amendment Bill was passed by Parliament on 17 September 2014 and, at the time of publication of this article, was yet to commence operation. There are three main points to make about the Bail Amendment Act 2014.
First, the reference to the presumption of innocence which had been in the purposes section of the Act has been relocated to the Preamble along with statements regarding: the need to ensure the safety of victims, individuals and the community (a); and need to ensure the integrity of the justice system (b). Resort to a preamble is old fashioned and has generally been discontinued (Pearce & Geddes 2011: [1.32]). The clear purpose is to reduce the importance of a bail authority having regard to the presumption of innocence -a regard that has not featured strongly in the cases analysed -with the assumption that a preamble is not usually construed as part of the Act. 15 If this construction is upheld by the courts, a cornerstone of our criminal justice system -the presumption of innocence -is significantly downgraded. Ironically, given the rhetoric around the asserted need for the 2014 amendments, the legislation's expression of principles regarding community safety has also been consigned to the relatively innocuous location of the preamble.
Secondly, s 17 introduces a new concept of 'bail concern' not otherwise known to bail law in NSW or any other Australian jurisdiction, and certainly not in the equivalent Victorian/Queensland unacceptable risk models. Thus, the amending provisions require a bail authority to assess any 'bail concerns' before making a bail decision (s 17 (1)). A 'bail concern' is one that relates to the former factors used to assess 'unacceptable risks' in s 17(2) (failing to appear; commit a serious offence; endanger safety of victims, individuals or the community; or interfere with witnesses). In assessing the 'bail concern' the bail authority must take into account only the matters in s 18, which is comprised of an expanded list 16 of the matters previously used to assess unacceptable risk in s 17 (3). Importantly, these factors now also include any bail conditions that could reasonably be imposed to address any bail concerns (s 18(1)(p)) whereas previously such conditions could only be imposed to mitigate an 'unacceptable risk'. The clear intent is to allow bail conditions to be imposed at the lower threshold of a 'bail concern': that is, even if no 'unacceptable risk' is identified. This is in direct conflict with the NSW LRC Report on Bail (discussed above) which attempted to restrict the proliferation of pro-forma bail conditions and conduct requirements (see NSW LRC 2012: 191-245;Brown 2014) by making it a two-step process so that conditions only came into play where the person would otherwise be detained, or as provided in the legislation, where there had been an unacceptable risk finding. While this may be the intended effect, introducing a concept not otherwise known to Australian bail laws may also have unintended consequences. In particular, the 'unacceptable risk' assessment in s 19 still requires a 'second step' before bail can be refused and so the two steps are not fully collapsed. More importantly, the concept of 'unacceptable risk' seems to become a 'free floating' one attached now only to the factors in s 19(2)(a)-(d) (failing to appear; commit a serious offence; endanger the safety of victims, individuals or the community; or interfere with witnesses) but not by reference to the s 18 matters. 17 While s 19 (1) indicates that a bail authority is to refuse bail if satisfied, 'on the basis of an assessment of bail concerns under this Division, that there is an unacceptable risk', it is difficult to marry s 19(1) with s 19 (2). Thus, what is it about the 'assessment of bail concerns under this Division' that can raise it from a mere 'bail concern' to an 'unacceptable risk' when a 'concern' is clearly a lower threshold which may not even amount to a 'risk' let alone an 'unacceptable' one. Ironically, this could also have the other unintended consequence of making bail refusal more difficult since it is now unclear how the bail authority moves from an assessment of 'bail concerns' in s 18 to the higher threshold of 'unacceptable risk': that is, a bail concern is not even a 'risk' let alone an 'unacceptable' one.
If the Government was genuinely concerned that the two-step unacceptable risk regime in the original 2013 Act was causing confusion, it is surprising that they have now introduced what is in effect another layer of assessment and prediction, using the language of 'bail concern'. Finally, an additional layer of complexity has been added by the 'show cause' provision in s 16A, which returns us squarely to the old territory of the complexities of presumptions against bail in the former Bail Act 1978 (discussed above). While the Review Report argues for a distinction between a show cause test to offence-based 'presumptions' at [222], the distinction is poorly explained and appears to be one of semantics: The show cause test is different to the previous offence-based presumptions in the 1978 Act. Offence-based presumptions indicate the bail outcome that an accused is expected to receive. Where an accused cannot rebut the presumption, the presumption has operative force. This does not work within an unacceptable risk model. The show cause requirement is part of a process to reach the end decision which includes the risk assessment. ( preventable delays and an accused's special medical condition. 20 In the event that the accused does 'show cause' and the bail authority turns to the unacceptable risk assessment, does s 19 (3) mean that the factors taken into account for showing cause cannot be re-assessed for the making of a bail decision of 'unacceptable risk'? If that is so, this would be particularly problematic given that, when assessing the factors in s 18, specific reference to matters that may be used to 'show cause' are included (for example s 18(1)(c) (strength of the prosecution case), (h) the length of time the accused person is likely to spend in custody if bail is refused and (k) (special vulnerability/needs)). 21 While the Government has claimed that the changes are 'common sense', in its determination to look 'tougher' on crime and to give the electorate the impression that more people will be denied bail, they have rashly introduced complicated and unnecessary changes to a regime that had only just begun to become familiar to police, lawyers, magistrates and judges after a twelve month familiarisation and training period. It is unclear what status the emerging body of legal principles under the Act discussed in this article will now have. Indeed, will magistrates and judges yet again need to start from scratch? And clearly, the Attorney General has already indicated (see second reading speech cited above) that further training and education and timelag may well be needed to accommodate these amendments. What a mess!

Broader issues in the bail reform reversal
What are we to make more generally of the sudden media-driven reversal of bail reform outlined above? Do we just add it to the list as yet another example of the irrationality of public policy making? We suggest there are some wider features that go beyond the specific instance that are worth highlighting in the context of an outbreak of regressive criminal justice legislation in a range of jurisdictions, not least in Queensland, Victoria and NSW. These features include: the denigration of judicial expertise and lack of concern with evidence and process; the power of the 'shock jocks', tabloids and police; and political failure to understand and defend fundamental legal principles that benefit us all and are central to the maintenance of a democratic society and the rule of law (Brown 2014).
The denigration of judicial expertise and lack of concern with evidence and process A common feature is the denigration of judicial expertise and a lack of concern with evidence. In contrast to the NSW LRC Report and process, the Review Report involved little or no new research. Its arguments do not arise out of empirical analysis nor, as analysed above, was there evidence, after one month's operation, of any judicial concern over the interpretation of the Act. What is the point in investing time and expense in the extensive consultative inquiries if the results, having been scrutinised by the Attorney General's Department, reduced to legislation and subjected to one year's training for all who will administer it, can be overturned by a short term political process before the new Act has been in operation long enough to be evaluated? What message does this send about political regard for judicial expertise?
The events also highlight impatience with and lack of faith in legal processes and the ability of the system to correct error. The first recourse in response to arguments that a decision is wrong is to take it on appeal, not change the law. In recent years a feature of attacks on the judiciary as 'out of touch' and 'not reflecting community values' has been lack of preparedness to take contested cases on appeal or, where they are taken on appeal, to wait for the results before changing the law. A clear example was the introduction of the NSW 'one punch' law in response to the media and public outcry over the sentence handed down in 2013 to Kieran Loveridge for the manslaughter of Thomas Kelly (Quilter 2014b), even though the DPP had announced an intention to appeal the sentence. Legislation to introduce a new 'assault causing death' offence was rushed through Parliament in January 2014 (Quilter 2014b relevant sentencing principles and finding manifest inadequacy in the sentence which was, as a result, significantly increased (Loveridge [2014] NSWCCA 120). In addition, the opportunity for a more nuanced guideline judgment was lost as the DPP had to withdraw his guideline judgment application in the Loveridge appeal in light of the introduction of the new 'one punch' law.
It is easier for the Police Association of NSW (the NSW police union) to complain to the Daily Telegraph or selected shock jocks and initiate a media campaign than request the DPP to go through the normal criminal justice processes. During his campaign against the bail laws Ray Hadley, a high rating radio 'shock jock', 'claimed on his 2GB radio program to have received 400 emails from serving police officers complaining about magistrates decisions' (Olding 2014). Such recourse to the media often produces quicker results than reliance on legal processes, as politicians, unprepared to argue for principles or await established legal processes, rush to change laws in response to media criticism. In this way the level of public trust in the judiciary and the criminal justice system is constantly eroded, to the long-term detriment of legal integrity (Hindess and Sawyer 2004). 22 This process feeds into an increasingly uncivil public discourse. Instant experts abound. People who would not think of looking over the shoulder of, say, their electrician, plumber or mechanic, seem to think judicial decisions can be second guessed by anyone, irrespective of their (lack of) knowledge of all the specific facts or evidence (Ryan 2005;Loader 2006;Garland 2001). 23 Another feature of these events is the growing practice of using former politicians to produce quick 'reviews' of complex legislation, often in preference to considered judicial and other extensive reports (Tink and Wheelan 2013). 24 Such reviews lack the element of independence, the depth of research and reliance on evidence, and the time for widespread consultation among interested parties, that are characteristic of law reform and other inquiries. They are 'in-house' political exercises designed to produce a way for politicians to go through the appearances of consultation while securing the quick results they want so as to alleviate media and public pressure (Sentas and Cowdery 2013).
As the former Attorney General who presided over more 'punitive' amendments (Steel 2009) to the reformist Bail Act 1978 than any other, turning bail from a release mechanism that respected the presumption of innocence and the right to liberty into a mechanism for pre-trial punishment, Hatzistergos was possibly the most inappropriate person imaginable to carry out the review. It was precisely the process of constant undermining of the presumption in favour of bail which he championed, that led to the complexity and incoherence of the 1978 Act and to the reference to the NSW LRC to remedy it. Hatzistergos's hard-line position could not have been clearer than when he introduced the Bail Amendment Bill 2007, boasting that 'New South Wales now has the toughest bail laws in Australia'. He noted that: … part of those changes includes removing the presumption in favour of bail for a large number of crimes and introducing presumptions against bail for crimes including drug importation, firearm offences, repeat property offences, and riots, and an even more demanding exceptional circumstances test for murder and serious personal violence, including sexual assault.
Those types of offenders now have a much tougher time being granted bail under our rigorous system. These extensive changes have delivered results. There is no doubt that the inmate population, particularly those on remand, has risen considerably as a result of the changes. In fact the number of remand prisoners has increased by 20 per cent in the last three years alone and new jails are being opened to accommodate the increase. (Hatzistergos 2007 In short, access to bail should be heavily restricted for 'those types of offenders', referring to accused people who have not yet had their guilt determined by a court, a classic conflation of accusation and guilt which lies at the heart of the media and political responses outlined above. A similar nonchalance towards the presumption of innocence was shown more recently by NSW Attorney General Brad Hazzard appointed to this position in April 2014. Speaking about the shifting of the onus to defendants to prove that they should be granted bail under the amendments, he stated: 'I have no doubt that needed to be changed, and they [defendants] now have to convince the court on behalf of the community they should be allowed out, and I don't really see that as a big deal.' (Huntsdale 2014, emphasis added) If Attorneys General cannot respect the presumption of innocence and the role of bail in protecting the value of liberty, how are we to expect shock jocks and tabloid journalists to do so? And is an explosion in the remand population and the hugely expensive building of new prisons seriously suggested to be the primary indicator of successful criminal justice policy, especially in relation to the unconvicted?
Finally, the complete lack of concern with evidence for the change is in contrast to the constant political parroting of the need for 'evidence driven policy'. Here a reform based on years of intensive and expert consideration was sabotaged before it had had time to operate long enough to be evaluated. Why not wait to see how it was working? The answer can only be that the main concern was not with the evidence, the reality, but with the political imperative of appearing to placate the shock jocks and the Police Association of NSW in order to diminish media criticism of the government.

The power of the shock jocks, tabloids and Police Association
A further feature of the bail reform reversal is the power of the shock jocks (Mickler 2004) and the tabloids (Scalmer and Goot 2004), in this case Ray Hadley in particular. Hadley's animosity towards former Attorney General Smith was played out in the debate prior to the release of the NSW LRC Report as part of a pre-emptive attack on the Report, commenced even before the Commission's final recommendations had been settled. On one front page of the Daily Telegraph Greg Smith was pictured as turning from Rambo to a marshmallow, accompanied by a banner headline 'How DPP Greg Smith went from Rambo to cream puff with stance of sentencing in NSW' (Clennell 2012). This undermining of Greg Smith for not being tough enough on law and order was ultimately successful with Smith being removed from the portfolio.
Smith's former media advisor, legal journalist Michael Pelly, has noted that Smith's attempts to reform bail and sentencing and bring the prison population down was initially popular within the government, particularly on financial grounds, until pressure was applied. According to Pelly, 'the pressure came largely from one source: 2GB  (Needham 2014).
The initial downgrading of the position of the Attorney General vis-a-vis the Police Minister is another illustration of the power and influence wielded by the Police Association of NSW.
Gallacher was a former President of this organisation. Increasingly, NSW government responses to significant criminal justice issues have been driven by the Police Minister, and indirectly the Police Association, rather than the Attorney General. This came to a head in outbursts from Barry O'Farrell who claimed that the judiciary were 'out of touch' after one bail decision (Davies and Patty 2012) and again over the Thomas Kelly 'one punch' case where he called for more judges and magistrates to be selected from the police because the police were more in touch with community values. The Police Association sees its role not simply as protecting the industrial interests of its members but also in securing the widest possible police powers, irrespective of whether that is desirable in the broader public interest of maintaining democratic traditions of liberties and rights (Sentas and Cowdery 2013).
This is not a new development. Mark Finnane has traced the history of police union political power back to the 1920s, noting that 'the annual conferences of police unions in Australia have been a standing item in the diaries of ministers responsible for police, their shadows in opposition, and occasionally even a premier' (Finnane 2000: 5). Notable examples of political interventions occurred in NSW by way of selective non-enforcement of public order offences as part of a campaign against the repeal of the Summary Offences Act 1970 (NSW) (Egger and Findlay 1988), in Queensland in the wake of the Fitzgerald Inquiry (1989) and in the role the Police Association of Victoria played in bringing down three Police Commissioners: Comrie, Nixon and Overland (Bachelard and Munro 2011). Finnane (2000: 17) concluded that 'police unions by the end of the twentieth century had become major players in the organisation of criminal justice in Australia.' What is perhaps new is his suggestion that 'the media became at some point a captive of police union viewpoints -at what point and to what degree would await further research' (Finnane 2000: 16).
The political failure to defend rule of law principles Another feature of the events is the fickleness of politicians and their inability to stand up for principles. Amongst various rule of law principles are that citizens should have a right to personal liberty, enjoy a presumption of innocence and not suffer punishment without conviction after due process (NSW LRC 2012: ch 2). Detention before trial offends these principles and so it should be strictly limited. The laws relating to bail should ensure this, by providing that bail can only be denied if it is likely that the accused person will abscond, attempt to interfere with witnesses or, with the integrity of justice processes, threaten to harm the victim, individuals or the community, or commit further serious offences. These decisions are best made, in the first instance, by police and then, on review, by the judiciary, on the basis of evidence particular to the specific case and the individual circumstances. A person accused of homicide in relation to a mercy killing of a partner suffering long-term crippling pain should probably be released on bail pending the trial; an accused serial or contract killer should not. This is a matter for judicial discretion, not for politicians to decide in advance through creating complex categories of offence based presumptions which cut across the ability to assess the individual merits of cases. Are these principles too difficult for politicians to comprehend and defend?
A common mistake is to frame rule of law principles as individual interests, to be balanced against public or social interests. The error lies in seeing the interest in liberty, and indeed in the other fundamental principles of the law, such as the presumption of innocence and the right to a fair trial, as interests of the individual and in particular the individual defendant. Conceiving them in this way, within the familiar metaphor of balance, renders one far more likely to see them as of less weight than social, community and public interests.
[But] the interest in liberty and fundamental principles is correctly seen as a collective, social, public interest. The issue then is one of reconciling or evaluating the strength of competing public interests. (NSW LRC 2012: para 3.12;Brown, 2013: 87, emphasis in original) Finally, the NSW bail reform reversal offends two other important legal principles, those of generality and of reciprocity. Laws should be made after careful consideration in relation to general states of affairs, not in relation to individual cases. Constantly changing the law after media outcries over particular cases offends the principle of generality and produces distortions in the law for short-term political gain. The principle of reciprocity is a form of social glue fundamental to a safe and cohesive society. It requires that we should want the same laws and legal processes applied to others that we would wish to be in operation in relation to ourselves, our families and our friends. Shock jocks and journalists might usefully reflect on this principle.

Being otherwise?
The rather depressing state of affairs outlined in this article raises the question of how might it have been otherwise? Russell Hogg (2013) has suggested that populism should be taken more seriously as a political rationality and that the common characterisation of penal populism as a form of pathology or irrationalism is politically unproductive. He further suggests that the two terms should be uncoupled, paving the way for an attempt to take populism more seriously and open up spaces for the mobilisation of more progressive forms of populism. We have not utilised the notion of penal populism directly in this article but certainly an underlying theme has been the sheer irrationality of the events and processes we have outlined, as an exercise in law reform. A possible example of a more progressive populist response leading to non-punitive law reform may be found in the initial responses to the death from one punch of Thomas Kelly in July 2012, Kings Cross, Sydney. In this instance, Quilter (2014c) noted that community and government responses focussed on a broad range of regulatory conditions and a multi-faceted response to address alcohol-related violence rather than simply on the individual culpability of the offender. But, as with Shoebridge's (2014) comments on bail reform quoted at the outset of this article, this also turned out to be a case of 'speaking too soon'. The political response to the sentence originally given to Thomas Kelly's killer, Kieran Loveridge, turned suddenly in the direction of a new 'one punch law', replete with a mandatory sentencing regime, which was subsequently adopted in different forms in Victoria and Queensland. 26 Indeed, there seems little that is progressive in the recent rash of law and order politics and legislation across a number of Australian jurisdictions. Nevertheless, in the spirit of Hogg's suggestion, we will conclude this article with some thoughts and questions on how events might have unfolded otherwise.
The centrepiece of our critique has been the appeal to evidence as a requirement for considered public policy making and certainly for any sudden about-turns. As argued above, statistical evidence as to how the new Act was working was not yet available and our analysis of the case law evidence revealed no major problems in interpretation of the Act. In the absence of such evidence, two major arguments were mounted. First, the Review Report pitched its objections to the new legislation at the level of policy disagreement, claiming that 'the review has not been hampered by this lack of data, as it focuses on the underlying policy of the Act' (Hatzistergos 2014: [46]). The policy disagreement is that 'on no basis could the presumption of innocence as referenced in section 3(2) be regarded as a purpose of the Act' (Hatzistergos 2014: [103]). This amounted to a repudiation (by simple assertion) of the more detailed and careful arguments in the NSW LRC Report (NSW LRC 2012: Ch 2) previously accepted by the Government and the Parliament. Secondly, the new Attorney General, Brad Hazzard, in the absence of evidence to support the Government's position, fell back on that old favourite, 'common sense'; the Review Report recommendations were merely 'common sense changes' (Hazzard 2014: 9). Law and order 'commonsense' has a long pedigree. As argued by Hogg and Brown (1998: 19): Commonsense is partial rather than wrong. By its very nature it resists engagement with other, more systematic bodies of knowledge where these resist commonsense assumptions. Commonsense is what 'we all know' already. It embodies tacit judgments and assumptions about the world that are harboured prior to the evidence being gathered. This is what makes it so resistant to debate or dialogue which questions, rather than shares, its starting points. Hogg and Brown (1998: 21-41) identify a number of 'enduring themes' within law and order 'commonsense': soaring crimes rates; 'it's worse than ever' (law and order nostalgia); US comparisons (the shape of things to come); soft on crime (the criminal justice does not protect citizens); 'we need more police with greater powers'; 'we need tougher penalties'; and victims should be able to get revenge through the courts. They go on to argue that there is an 'uncivil' side to law and order commonsense and that it is to 'the fallibility of the system and the importance of not committing further crimes in the name of justice, punishment, sacrifice or vengeance, that criminal justice policy should be directed in a civil society' (Hogg and Brown 2014: 43).
Appeals to evidence, rationality and process, while central to our analysis above and, hopefully, to considered criminal justice public policy making (Hobbs and Hamerton 2014), must also be translated into and through the forms of argument and the register of popular media and politics, a realm in which emotion and affect play key roles. Criminal justice issues arouse strong emotions and lend themselves to a more individualised focus than other types of news stories. The focus is often on the victim, the language is emotive, constructing a virtual community around identification with particular high profile victims and a sense of cohesion that flows from widespread condemnation of the alleged offender (Quilter 2014c). 'Legal niceties' are often seen as just that, a formalistic overlay which does not reflect or give expression to the personal narratives of hurt, loss, outrage and revenge that swirl around specific crimes.
While bail decision-making is a procedural stage in the process of bringing someone to trial so that their culpability can be decided, it has increasingly become, as we argued earlier, a forum for the condemnation of the accused and of their alleged behaviour, a moment where accusation, guilt and punishment are conflated. The passage of the Act was an attempt to shift bail from a form of pre-trial preventive detention and return it to its place as a procedural forum focussed not on the nature of the alleged offence but largely on whether the defendant would present for trial and respect the integrity of the trial process, the evidence, witnesses, victims and the community. The media outrage over the three specific bail decisions was focussed rather on the nature of the alleged offences (in two cases, homicide), the gang affiliations of Mick Hawi and Hassan Ibrahim, and the views of relatives of a murder victim. 27 Earlier we noted the vacuum that occurred in the aftermath of the three contentious bail decisions, during which talk-back radio and the Daily Telegraph 28 were able to portray the decisions as putting the public at risk, doing a disservice to the victims and their relatives, and indicating the new Act and the government were 'soft on crime'. We noted that, without published reasons for the decisions being made available, the opportunity to mount counter arguments to the 'outrage' line was restricted. When the decisions did become available, after the damage had been done, they seemed cogent and explicable in two of the cases and in the other, an appeal was successful. If courts made decisions available more quickly, commentators would be better placed to respond and it is possible the public 'debate', such as it is, would be better informed and more nuanced.
Another question worth some consideration is how both politicians and the media might be brought into some engagement with the law reform process in ways that makes them better informed and more responsible? How was it possible for a Parliament that unanimously approved and passed the new Act in May 2013, giving its imprimatur to the long law reform process, to perform such an about turn a year later and repudiate the key features of the legislation, on precious little evidence save some media uproar over a small number of cases?
What does this say about the level of commitment to legislative integrity and responsibility, about the quality of information provided to parliamentarians, about their comprehension of the arguments for reform, and about their ability to articulate and defend key principles central to the protection of legal processes and liberties in a democratic society? What does it say about the way party whip and caucus systems operate and about the quality of leadership, of both government and opposition? Are there better Committee processes that might have been used to provide some greater scrutiny to the Review Report arguments?
The Parliament of NSW Legislative Review Committee commented on the Bail Amendment Bill that: 'In removing a requirement that the bail authority gives regard to the presumption of innocence and the general right of liberty when making bail decisions, the Bill impacts these rights. The Committee refers these matters to Parliament for further consideration.' (Legislation Review Committee 2014: vii) This referral to the parliament appears to have been ignored. Why? What exactly would have been wrong with referring the report to a parliamentary standing committee for further consideration, as proposed by the Australian Greens party? Why did the Australian Labor Party (ALP) opposition roundly criticise the Amendment Bill and yet still vote for it (Whitbourn 2014)? Shadow Attorney General Paul Lynch said in debate: 'The Opposition does not oppose the bill but it thinks the Government has not the slightest idea what it is doing' (Lynch 2014: 6). ALP MP Ron Hoenig echoed the refrain: 'the Opposition does not oppose this convoluted, dreadful bill, but it is bad public policy created by a panicked Government trying to curry favour with a reactionary media' (Hoenig 2014: 9). As Mr Jamie Parker, Australian Greens MP for Balmain noted: 'Opposition members have made some incredibly strong arguments as to why the bill is faulty, poorly developed and attacks fundamental rights, why it waters down the bedrock of our justice system -the presumption of innocence -and trashes judicial expertise, yet the Opposition essentially will be waving the bill through' (Parker 2014: 11). Why? As Mr Parker went on to point out: 'If Labor were to oppose this bill it would present us with an opportunity to conduct further negotiations to improve the bill as it stands' (Parker 2014: 12).
What do we make of the fact that the man asked to complete a hasty review of the Act had, in 2009, while Attorney General, in the then ALP government, rejected the overtures of the then Opposition (Coalition) Shadow Attorney General, Smith, to develop a bi-partisan, evidence driven, fairer, and less costly criminal justice policy, taking criminal justice out of the law and order auction (West 2009). How might the business of criminal justice policy making and law reform have been different if Smith's offer of a 'truce' had been accepted, rather than rejected (Robins 2009)? These questions deserve reflection, for the attempt to formulate answers might illuminate paths to a different process and a different outcome, challenging the notion that the sabotage of bail reform in NSW was somehow inevitable.
In relation to the law reform process itself it is evident that the existing extensive consultative exercise engaged in by the NSW LRC does not preclude certain parties from leaking to the media prior to the finalisation of the final report, nor from continuing to try to undermine recommendations they do not agree with once the report is released. In short, the consultation process does not seem to generate a sense of commitment to the various compromise positions inevitably reflected in a final report and in subsequent legislation. This raises the question of whether different forms of consultative process might produce a more collective outcome. Might a 'workshop' approach built around selected scenarios, as later developed by the NSW Sentencing Commission, 29 work better to focus discussion, highlight the empirical evidence and the various arguments, and generate a wider community consensus around compromise recommendations?
To push such thinking even further, might it be possible for sections of the media to be invited to put forward submissions to law reform bodies and processes or to be somehow involved in the deliberative process? Such a suggestion sounds a bit farfetched, but certain media commentators clearly have very strong views on how particular legal processes ought to operate and it might be of benefit if those views were able to be articulated in a general way, divorced from specific cases, and in an environment that encouraged them to be constructive. If the Ray Hadleys and Daily Telegraph editors of the world could somehow be given a voice during the official process of law reform policy formulation, the very process of engagement and of having to come to terms with some of the underlying principles and alternative arguments might make them a little less inclined to try to trump or undermine the results down the track, results in which they, as outsiders to the law reform consultation process, have no stake.
Another challenge to be confronted is how to generate good news stories in a media climate that thrives on highlighting dysfunction and failure. The outcome of the Greg Smith-initiated bail law reform process might have been covered in a very different way. With 25 per cent of the NSW prison population unconvicted at the time, between 2500 and 3000 people at any one time and more than 10,000 over the course of a year in NSW (72.4 per cent of total prison receptions in 2010: NSW LRC 2012: [4.6]), the system was ripe for reform. Not only is the system extremely costly in monetary terms, its social consequences are profoundly dysfunctional. They include: physical and psychological hardship; assaults and deaths in custody; financial implications for the accused and their family; deleterious effects on children; the criminogenic effect of mixing with sentenced prisoners and high risk remandees; the lack of access to any programs; a range of effects on the ability to receive a fair trial, and on conviction rates; pressure to plead guilty; and others (see NSW LRC: Ch 5; Grunseit et al. 2008).
The enactment of the Act was potentially a good news story: more people would be granted bail; prison numbers would be reduced; considerable financial savings would be achieved; Corrective Services would be freed up from processing so many people received into prison on remand for minor offences, enabling them to concentrate resources on programs for the convicted; better justice would be achieved for the 55 per cent of those on remand (5,218 of 10,342) who in 2010 in NSW were subsequently released to bail, received a non-custodial sentence or were acquitted (NSW LRC 2012: [4.13]); fewer lives would be disrupted; the criminogenic (crime producing) effect of being sent to prison would be reduced; and so on.
It is our understanding that at least one journalist was researching a story along these lines to coincide with the new Act coming into operation but decided not to go ahead. It would have been timely to produce a follow up story to the one written by Joel Gibson in the Sydney Morning Herald in 2010 under the headline 'No Bail Go to Jail', which focussed on the bashing of a subsequently acquitted remand prisoner, and which was a vehicle for explaining Greg Smith's reformist agenda (Gibson 2010). Journalistic traditions of 'more bad news' tend to gel with a widespread political reluctance to argue for progressive change. This is a version of 'don't mention the war'; or, in this case, don't mention that the government might be doing something other than taking a 'tough' (read punitive) stand on crime and punishment. Governments of all persuasions are happy to trumpet an increase in penalties but often prefer to remain silent about programs or initiatives which attempt to reduce imprisonment rates, provide rehabilitative programs or post release assistance, as if they are embarrassed by sound and constructive reform outcomes, or fear that they may be an electoral risk for being insufficiently draconian. These attitudes have led to the entrenching of a 'reform on the sly' approach, whereby more progressive social and welfare approaches to criminal justice issues, however much evidence can be marshalled in their favour and however successful, are not promoted lest they draw adverse attention and claims of being 'soft on crime', the automatic assumption being that the public are universally punitive, a 'common sense' notion challenged by research (Roberts et al. 2003). One effect of this approach is that little on-going public support is built for reformist measures. So when the spotlight is shone on a sound and carefully constructed reform initiative like the Act, and media and political criticisms emerge, there is little well informed and widespread public support to point out that, despite the specific 'weakness' that has been identified, the initiative is meritorious and beneficial (and at the very least, worth being allowed to 'bed down' before judgment is passed on it). Because the discursive ground for this sort of insulating strategy has not been prepared, punitive and exclusionary responses quickly swamp the field.
Unfortunately, this was the unfavourable environment into which the work of the NSW LRC and Greg Smith's bail reform legislation emerged. The fact that the legislation's aims were eminently sensible -making bail easier to obtain would both alleviate the injustice of pre-trial detention for many accused and reduce the remand population in NSW prisons -was lost amidst the shouting and the fear-mongering. The ease and rapidity with which these laudable goals of the new Act came to be characterised as undesirable and dangerous, effects 'proved' by the outcomes in three contentious cases, is troubling. Conspiratorial theories were floated including the suggestion that any increase in the proportion of accused persons being granted bail (characterised in the media as a 'bad' thing) could be attributable to deliberate manipulation by the NSW Police, as part of a strategy to undermine the new Act (Olding 2014). Police Association President Scott Weber countered that it was 'the judiciary not enforcing the community standards' that was the reason for any increase in the success rate of bail applications (Olding 2014). Could it be that if bail rates do increase and the remand population shrinks (it is far too soon to tell), that neither police sabotage nor judges being 'out of touch' is the cause? Might it simply be that police, magistrates and judges are just carrying out the intentions behind, and applying the principles in, the Act? 30 Worthwhile law reform takes time. It requires ongoing political commitment, constant justification and community and media engagement. It involves restatement of the underlying principles governing the criminal justice system, full recourse to the available evidence, and recognition that sound and effective reform cannot be done 'on the sly' if it is to command wide respect and be resilient in the face of anticipated attacks. Rather than 'speaking too soon' the key problem may be not speaking soon or often enough.

Postscript
The following developments took place after the completion of this article.
On 10 September 2014, the NSW Attorney General, bypassing the NSW Law Reform Commission, asked the NSW Sentencing Council to consider a proposal to make further amendments to the Bail Act 2013 (NSW) on whether a new 'show cause' offence should be added to the Act. The specific addition under consideration is with respect to an accused charged with a serious indictable offence committed:  while subject to a good behaviour bond, intervention program order, intensive correction order;  while serving a sentence in the community; or  while in custody. We made a submission to the NSW Sentencing Council arguing against the proposal.
The Sentencing Council is not due to report on this matter until May 2015, after the March 2015 NSW State election.
On 10 October 2014, Attorney General Hazzard announced the appointment of John Hatzistergos as a judge of the NSW District Court.
Finally, in spite of the Government's claim that there was a pressing need to amend the Bail Act 2013 the Bail Amendment Act 2014 passed on 17 September 2014 has still not commenced operation -almost three months later. We note that this 'waiting period' is longer than the period for which the Bail Act 2013 was in operation before an urgent review was announced by the Government.  16 The new factors (s 18(1)(f), (g), (n), (o) and (p)) emphasise the rhetoric of 'tough on crime' and protection of the community with an accused's criminal associations (g) and the views of victims and their families (n), (o) becoming factors. 17 We note a contra argument in Flowchart 2 in the Bail Amendment Act 2014 (NSW) which suggests the s 18 factors are taken into account. The Flowchart, however, is 'illustrative' only (s 16 (3)) and s 19 does not so provide. 18 In the debate on the Bill, Lynch (2014: 4) decscribed the proposition that a show-cause test can be distinguished from the former presumptions as: 'That is an interesting argument-almost as interesting as arguing about how many angels can fit on the head of a pin.' 19 Hatzistergos (2014: [223]) indicates that in Victoria there is a divergence of authority on whether the question of unjustifiable risk falls to be determined as part of showing cause or whether, it is an additional matter that needs to be determined if the accused successfully shows cause.  21 Although it is noted that this argument assumes the unacceptable risk assessment references back to s 18, which is not clear (as discussed above). 22 Part of a more general 'anti-elites' politics (Hindess and Sawyer 2004). 23 This is part of a more general phenomenon described by Ryan (2005) as 'the rise of the public voice', by Loader (2006) as the 'fall of the platonic guardians', and by Garland as the 'declining influence of social expertise' (Garland 2001: 150). 24 For a previous NSW example, see Tink andWhelan 2013. 25 R v Wakefield (1969)  should be noted that the Sydney Morning Herald (SMH) promoted the 'community anxiety about new laws' line, a theme that was particularly evident in some articles; for instance of Paul Bibby (see Bibby 2014b) as did radio talk back hosts not usually in the 'shock jock' mode, such as Richard Glover on ABC 702. It was only when the Government announced the Hatzistergos Review that the SMH became more critical and questioned its timing: see Bibby and Whitbourn (2014). 29 See note 4 above. 28 For example The NSW LRC [15:35] recommended the listing of responses available to police where they suspect a failure to comply with a conduct direction, as a response to suggestions that police had no options but to arrest. This was incorporated in the Act (s 77(1) with options to take no action, issue a warning, issue a Court Attendance Notice or arrest. The intention was that the proportion of revocations of bail for minor technical breaches be reduced. Brown (2013) suggests in relation to a drop in juvenile detentions that 'it may be that the debate surrounding the Law Reform Commission Inquiry and Report and the Attorney General's response has already had an effect on bail decision-makers and on the complex organisational and cultural climate, in the direction of a more "resilient" attitude to the grant of bail'. (Brown, 2013:95