Sydney Law Review
The role of prosecution in achieving compliance with Occupational Health and Safety (OHS) legislation is a highly contentious issue, particularly in the mining industry. In New South Wales, what is perceived by some as a newly aggressive approach to prosecution has brought to the foreground a number of critically important questions concerning how best to regulate OHS. This article seeks to identify a number of principles relating to prosecution policy that would achieve more effective OHS outcomes. In particular, it seeks to steer a middle path that neither rejects prosecution as an important deterrent at the top of an enforcement pyramid, nor uses it in circumstances where it is likely to do more harm than good.
Prosecution? ‘Make it achievable and make it fair!’ (a NSW Mine Manager).
[I]t is fundamental that the criminal law must be administered in an appropriate fashion. The legislature has chosen to emphasise the importance of occupational health and safety matters by creating absolute offences. If the prosecution of offences is undertaken in an arbitrary, capricious and irresponsible fashion, the laws themselves are brought into disrepute for reasons that are obvious. This is especially so in the area of occupational health and safety prosecutions where it is the custom of the prosecutor to seek a moiety of the penalty, that is payment of one half of any amount imposed by way of penalty — Newcastle Wallsend Coal Company Pty Ltd v Inspector McMartin  NSWIR Comm 339 at  (Marks J).
The role of prosecution in achieving compliance with OHS legislation is a highly contentious issue, particularly in the mining industry. Nowhere is this more so than in New South Wales, where, following the Gretley disaster, the Department of Primary Industries developed a new found enthusiasm for punitive action, particularly following fatalities. It has, moreover, chosen to prosecute not just companies but also individual mine managers and other statutory duty holders.
The Department’s prosecution policy, and the approach of the independent Investigations Unit charged with investigating fatalities, has precipitated a seething dispute between the New South Wales Minerals Council and major mining companies on the one hand, and the mine safety regulator and the mining trade unions on the other. The companies argue that prosecution is counter-productive, inhibits adequate safety investigation, encourages a defensive rather than a proactive OHS culture, and drives away would-be mine managers at a time of severe labour shortage. The enactment of the Occupational Health and Safety Amendment (Workplace Deaths) Act 2005, introducing a higher penalty regime for workplace fatalities involving recklessness or intent, has fuelled their concerns. The trade unions, on the other hand, welcome these developments as providing effective deterrence to corporate law-breaking and urge regulators to expand their use of prosecution to a far wider range of circumstances.
Although a similar prosecution policy has not yet been adopted by regulators in Queensland or Western Australia, there are, nevertheless, glimmerings of prosecutorial activism in both states, notwithstanding a dearth of previous prosecutions by agencies with a strong philosophical commitment to an ‘advise and persuade’ approach. In consequence, many companies are watching anxiously, fearing that New South Wales’ new found enthusiasm for prosecution will infect their own states. The mining trade unions on the other hand, urge that this is precisely what is needed to achieve improved OHS outcomes.
The ongoing debate in a number of jurisdictions, concerning the virtues of introducing industrial manslaughter legislation, has added to employers’ concerns, not just in mining, but across industry generally. This has caused the Australian Chamber of Commerce and Industry (ACCI), amongst others, to call for a major rethink of OHS policy. What the ACCI argues for example, is that the current system of OHS regulation is seriously deficient, that a more enlightened approach is needed than a one-sided ‘employer blame-game’ and that current laws undermine a culture of shared responsibility.
At the heart of many of these questions (though rarely articulated in these terms) is the appropriate role of prosecution and the criminal law more generally, in enforcing OHS offences. On the one hand, penalties are imposed that are intended by the courts to serve the functions of individual and general deterrence. If successful, these deterrent messages will provide disincentives to non-compliance and reduce levels of work related injury and disease. On the other hand, criminal law often serves a much less pragmatic role, fulfilling moral, symbolic and retributive functions. As Hawkins argues:
Prosecution is a ceremonial restatement of norms by which people and individuals order social life. Its use sustains the moral world which the regulatory organization inhabits. One way it does this is through the satisfaction given by the prosecution of a blameworthy defendant that moral boundaries are being maintained and reinforced...In making public those standards of behaviour deemed proper, decent and desirable, prosecution can be cathartic, since it can sometimes satisfy a demand, whether from the victim, the victim’s family, the media or people generally, for a public statement of the worth of the victim and the culpability of the defendant.
Unfortunately, these instrumental and expressive roles are rarely well integrated and commonly the pursuit of one serves to undermine the other. It will be argued that much of the conflict between employers and unions over the role of prosecution can be understood in these terms. Reconciling these roles is not easy and in some circumstances is likely to prove impossible, making hard choices between prevention and retribution unavoidable.
This article describes the current approach to prosecution adopted by the mines inspectorates in New South Wales, Queensland and Western Australia. It also summarises the implications of the international evidence-based literature on the impact of ‘deterrence’ and ‘compliance’ approaches to enforcement, in order to identify the ingredients of a rational and effective approach to prosecution. In doing so it seeks to identify a number of principles that, if adopted, would be much more successful in preventing work-related injury and disease than current prosecution practice. These principles, it is suggested, should steer a middle path between an over-zealousness that may inhibit accident investigation and chill safety initiatives, and a timorousness which, by failing to deter the recalcitrant, may prove equally antithetical to safety.
The debate with which this article engages is by no means confined to the mining industry, although it finds expression in that industry in a particularly acute form. Accordingly, the principles proposed below have more general application both to occupational health and safety generally, and to related areas of social policy.
Until quite recently, none of the inspectorates of the three mining states engaged in prosecution to any significant extent. For example, in New South Wales, according to the report of the Gretley Inquiry, in the seven years before that disaster there had been 33 deaths in New South Wales coal mines without a single resulting prosecution. Further, the very few prosecutions that had taken place in the mining industry in other circumstances (relating to metalliferous mines) had involved low penalties, were poorly publicised, and failed to send any significant deterrent signal. This led to a general perception, particularly within the mining trade unions, that prosecution was a ‘dead duck’.
It was the unwillingness of the New South Wales mines inspectorate to prosecute, coupled with political pressure (especially from the Construction, Forestry, Mining and Energy Union (CFMEU)), that prompted the establishment of an independent Investigations Unit in 1998. Following the Gretley disaster, Justice Staunton’s call for the ‘timely prosecution’ of mining companies and senior officials, was particularly influential on enforcement policy. Two mine managers, a surveyor and a number of under-managers were prosecuted in relation to Gretley (as well as the companies involved). Convictions against the mine manager and former mine manager were upheld on appeal (although no conviction against the latter was formally recorded). Another mine manager suffered a similar fate following a death at Awaba. The employers associations maintain that in neither case did the circumstances merit prosecution, and that the prosecution of individuals was particularly inappropriate. A number of other cases involving death or serious injury (and a handful that did not) have also been prosecuted in the period since 2001.
Strikingly, behaviour that exposes workers to serious OHS risks, but which does not result in death or injury, only exceptionally attracts prosecution, even if it is calculated or reckless. This is a matter of some considerable concern, since the relevant legislation makes it clear that neither death nor injury is a prerequisite to prosecution. On the contrary, the offence lies in the act or omissions that led to the worker being exposed to potential death or injury.
Queensland and Western Australia have been, and remain, substantially less prosecutorial in orientation than New South Wales, although the trend in both states is towards more prosecution, albeit from a very low base. In Queensland, there is no evidence of there being any prosecutions in the coal mining industry and, until the last few years, there were only a relatively small number related to metalliferous mines. Even in relatively recent years (from around 2002 on) however, official reports identify less than a handful of prosecutions each year, including some against site senior executives. All of these have involved some serious injury or death. The level of fines themselves has been modest, with a maximum of $30,000 against a contractor and $3,500 against individuals. Prosecutions relating to fatalities have so far all produced guilty pleas with no substantial penalties imposed.
In Western Australia, it is very difficult to obtain information about the level of prosecutions, and there has been a considerable reluctance on the part of the inspectorate to disclose relevant information about prosecution practice. According to the 2004 Ritter Report:
[T]he level of prosecutions would appear (on the basis of the very limited information provided by DOIR) to be very low. Notwithstanding very frequent identification of breaches by the inspectorate, there is very little evidence of any formal action being taken, beyond the giving of directions on some occasions.
A few prosecutions, even in the absence of death or serious injury, can be identified from press reports, with companies being fined low five-figure sums, but these, so far as one can tell, are highly exceptional. This, however, may be changing, with responsibility for mine safety being shifted to the Department of Consumer and Employment Protection, following inquiries which expressed concern about the danger of the mine safety regulator being captured by the very industry it purported to regulate. Whether the 2006 prosecution of BHP Billiton in relation to a death in the Pilbara and the ensuing fine of $200,000 are symptomatic of a more general change of approach remains to be seen.
The paucity of prosecutions cannot be explained by any practical difficulties in gathering evidence, by the costs of mounting prosecutions, or by any lack of specialist expertise in conducting such proceedings (although all of these may be significant in particular circumstances). Rather, as a number of reports have pointed out, it reflects a cultural antipathy to prosecution, and the underlying belief that much more can be achieved by advice and persuasion than by coercion. In Queensland and Western Australia, the mines inspectorates continue to subscribe to an ‘advise and persuade’ philosophy which leads them to operate almost exclusively in the lower reaches of the enforcement pyramid outlined in Figure One below.
This cultural aversion to prosecution would not be apparent from a review of formal policy documents. For example, in Western Australia, the relevant policy states that:
Enforcement is an essential element in controlling or regulating activities and gaining compliance with statutory requirements. This is done by detecting breaches, bringing them to the attention of the alleged offender, requiring corrective or preventative action, applying penalties (directly or through the courts) and providing deterrence.
The Queensland Compliance Policy, whilst emphasising a co-operative approach, also refers to the capacity to initiate prosecutions against people or companies for failing to meet safety and health obligations. In these states, there is good reason to conclude, as the Ritter Report stated, that there is a ‘very large gap’ between official policy, and what inspectors do ‘on the ground’.
To summarise, in Queensland and Western Australia, a cultural antipathy to prosecution has resulted in a paucity of prosecutions and the failure to provide any credible tip to the enforcement pyramid described in Figure One below. In New South Wales, the situation is more complex. Prosecution policy has lurched from a pure ‘advise and persuade’ approach (which still prevails in the large majority of circumstances), to one in which prosecution has become routine in the case of fatalities, even where the culpability of the defendant is (at least as perceived by the industry and its associations) relatively low.
As will become apparent, neither this extreme reluctance to prosecute on the one hand, nor zealous prosecution in a limited range of circumstances, on the other, are rational or remotely optimal enforcement strategies. To understand why this might be the case, we turn next to the empirical evidence as to the strengths, limitations and potentially counter-productive consequences of deterrence in general, and to prosecution in particular, and to the even greater limitations of a ‘compliance’ strategy at the other end of the compliance-deterrence continuum.
Figure One: The Enforcement Pyramid
A useful heuristic in thinking about the role of prosecution as a component of overall enforcement policy is the widely recognised concept of the enforcement pyramid as conceived by Ayres and Braithwaite. This pyramid, which involves advisory and persuasive measures at the bottom, mild administrative sanctions in the middle, and punitive sanctions at the top, is intended to assist in determining what enforcement tools to use in any given case. According to its proponents, regulators should begin by assuming virtue (to which they should respond by offering co-operation), but when their expectations are disappointed, they respond with progressively punitive and deterrent-oriented strategies until the regulated group conforms.
Central to this model are the need for (i) gradual escalation up the face of the pyramid and (ii) the existence of a credible peak or tip which, if activated, will be sufficiently powerful to deter even the most egregious offender. The former (rather than any abrupt shift from low to high interventionism) is desirable because it facilitates the ‘tit for tat’ response on the part of regulators which forms the basis for responsive regulation. Although the concept of an enforcement pyramid has its critics, it provides a valuable conceptual framework for thinking about the various tools available to enforcement officers, and how they might be used to optimal effect.
In the past, regulators in all the mining states have prosecuted and obtained legal sanctions against violators in only a very small number of cases, if at all. They have dealt with most detected violations by means of advice, warnings and demands for remedial action at the bottom of the ‘pyramid of sanctions’, or at its middle levels through various forms of administrative action and directions. They have adopted what the regulatory literature terms a ‘compliance’ strategy: one which relies heavily upon advice and persuasion to the virtual exclusion of more punitive policies and sanctions, and which rejects deterrence and prosecution almost entirely.
Whilst there are understandable reasons in terms of politics and history as to why such an approach evolved, there are no convincing reasons as to why it should continue. The empirical evidence is both compelling and well known and need not be rehearsed at length. In essence, there is little if any evidence that a policy of de facto non-prosecution (or even a combination of minimal prosecution and small penalties in the few cases that get to court) will achieve improved OHS outcomes. Under such policies, regulators have no credible means of dissuading the recalcitrant from exposing their workforce to unacceptable risk. Unsurprisingly, in a number of well-documented cases, such an approach has failed to prevent high levels of work-related injury, disease and death.
Such policies have also aroused considerable criticism, not just from the trade union movement, but also from a range of official inquiries. For example, the Laing Report took the view that it was fundamentally important that ‘those who are unwilling to comply with their safety and health obligations understand that prosecutions will be taken if there is a failure to comply with the Act’, while the Ritter Report was also heavily critical of what it regarded as a fundamental failure of enforcement. The 1997 New South Wales Mine Safety Review also expressed concern regarding inadequate levels of enforcement, as did a 2005 review of the Queensland Inspectorate.
The failure of a pure compliance policy has led some commentators to emphasise the virtues of deterrence and to suggest that prosecution and the imposition of substantial legal penalties are crucial to achieving improved OHS outcomes. There is indeed considerable evidence to suggest that deterrence can and should play an important role in an overall enforcement policy, but unfortunately that role is complex. Further, as we will see, a ‘pure’ deterrence strategy may prove almost as unsuited to achieving improved OHS outcomes as a ‘pure’ compliance strategy.
At the risk of simplifying a difficult subject, there is evidence that both specific and general deterrence play variable but important roles in any credible enforcement policy. From this it might appear, as one overview of the literature states, that deterrence is demonstrably ‘an effective means of securing compliance.’
However, upon closer examination, it becomes apparent that the problem is not so much with a generalised belief in the value of deterrence (and prosecution as a vehicle to achieve it), but with how deterrence and prosecution are invoked in practice. For just as there is evidence of the positive benefits of prosecution, so there is also evidence of its negative impact when it is inappropriately used. For example, there is evidence that a confrontational style of enforcement may diminish the willingness of firms to cooperate and learn from past experience as well as make them reluctant to share information and unwilling to consult regulators for fear that their disclosures may be used against them. Similarly, it may inhibit in-firm accident investigation, prevention and remedial action. Individual prosecutions against statutory office holders may make it difficult to attract well qualified applicants to such positions and reduce the skills base of the industry. In sum, the evidence in favour of deterrence described above, must be weighted against evidence suggesting that its indiscriminate use can be counter-productive.
A further challenge is that in so far as general deterrence ‘works’, it does so only in some contexts and to some extent. Different types of firms, different sizes of firms, and different types of office holders, are all likely to react differently to the signals sent by prosecutions. For example, there is some evidence to suggest that deterrence is likely to have much greater impact in relation to small and medium sized enterprises than in relation to large ones. The simpler management structures of small firms and the relative incapacity of key decision-makers within them to avoid personal liability, also make them much easier targets for prosecution. The size of penalty may also be an important consideration: mega-penalties tend to penetrate corporate consciousness in a way that other penalties do not.
Deterrence is also likely to have a different impact on firms who are differently motivated. It is likely to work best in relation to those who deliberately choose non-compliance for pragmatic reasons (usually because they perceive that they can make a greater profit by non-compliance than by compliance). But whilst deterrence may be effective in causing ‘reluctant compliers’ and the recalcitrant to improve their behaviour, the prosecution of firms who perceive themselves as ‘good guys’ (for example, committed compliers and those who go ‘beyond compliance’) may be counter-productive, causing resentment and generating a culture of ‘regulatory resistance.’
Deterrence also works better in relation to individuals than to organisations, but once again much depends upon the context. It is one thing to prosecute key decision-makers in small organisations (where they are both readily identifiable and amenable to incentives to comply) and another to prosecute senior officers in large organisations whom, while appropriate targets in principle, are exceptionally hard to convict in practice. It is yet another to target middle management (who often lack decision-making power and are widely perceived as innocent scapegoats). Prosecuting ‘wrong’ individuals — those who are vulnerable and easy targets for example, rather than senior decision-makers — creates a considerable sense of injustice and damages the legitimacy of the entire regulatory regime.
Specific deterrence, it seems, has a substantially greater impact than general deterrence, although this impact is more theoretical than real. This is because it is practicable to prosecute only a small number of enterprises and individuals, with the consequence that the ‘reach’ of specific deterrence will necessarily be extremely limited. Prosecutions have been, and will continue to be, rare events in the experience of individual firms.
From this evidence, it may reasonably be concluded that prosecution should be used sparingly, and carefully targeted to appropriate circumstances, and to actors who are most likely to respond positively to it. But what precisely does this mean and what would an optimal prosecution policy look like? How can the positive impact of deterrence be enhanced while minimising its adverse and counter-productive impacts? How can a balance be achieved between the use of deterrence and the use of other less coercive strategies?
On the one hand, an under-investment in deterrence (manifested through the relative infrequency or absence of prosecutions) may send insufficient signals to some duty holders to focus on safety and improve their OHS performance, and in extreme cases, may result in a complete collapse of a credible enforcement strategy. On the other hand, an inappropriate, misdirected or indiscriminate use of prosecution can create a culture of regulatory resistance, destroy trust between regulators and the regulated, send perverse incentives, inhibit OHS information flow and chill corporate OHS initiatives.
At the time of writing, none of the three mining states have managed to steer a middle course between the perils of an extreme compliance strategy on the one hand, and the heavy handed use of deterrence following death or serious injury on the other. Nor do the three mining states appear to have developed any strategies or principles to assist them in this regard, beyond some rather loosely worded enforcement and compliance policies to which we will return.
Against this backdrop, the following sections articulate a series of design principles intended to achieve a more balanced and effective prosecution strategy. To foreshadow the detailed discussion below, the first two of these principles suggest means of avoiding the counter-productive consequences of both under-prosecution and over-prosecution. The third identifies criteria to determine what sorts of circumstances should justify prosecution, emphasising the importance of culpability, risk and track record. The fourth suggests that prosecutions should generally take place irrespective of whether harm results, while preserving a role for ‘event based’ prosecutions at the tip of the enforcement pyramid. The fifth principle identifies circumstances in which deterrence is likely to be effective as well as those where it is not and suggests how and when individual decision-makers should be targeted, and who they should be. The last two principles concern the appropriate role of retribution and the circumstances in which more might be achieved by applying the techniques of restorative justice.
The first two of these principles may be seen as precursors to the others. Since they flow directly from the discussion in the previous section, they can be briefly stated. The first, which flows from the well-documented limitations of a pure compliance strategy, and from its demonstrable failure to provide incentives to the recalcitrant to improve their safety performance, is as follows:
However, it is also clear from the analysis in the previous section that prosecution does not work across the board and the available evidence suggests that prosecution should be used sparingly — carefully targeted to appropriate circumstances and to actors who are most likely to respond positively to it. For just as there is evidence of the positive benefits of prosecution, so also there is evidence of its negative impact when it is inappropriately used. A confrontational style of enforcement or enforcement against ‘good’ enterprises at a low level of culpability may diminish the willingness of firms to cooperate and learn from past experience, as well as generating a variety of defensive behaviour which impedes preventative action. This leads to a second principle:
However, this begs the question of what is meant by ‘inappropriate’. This is a question on which there is a diversity of competing views. Trade unions and mining communities, especially following a fatality or serious injury, argue in favour of prosecution, even against those whose culpability is quite low. On the other hand, mining companies, managers and other statutory position holders are inclined to suggest that prosecution should be reserved for ‘bad apples’ which they tend to equate with the reckless and willful. Accordingly, prosecutors, in determining which cases to prosecute, and in seeking to identify an acceptable basis for prosecution, find themselves between a rock and a hard place. They will inevitably either offend those who demand retribution or those who put prevention first and argue that retribution, particularly against those whose culpability is relatively low, is seriously counter-productive. As a result, politics, rather than rational decision-making, often holds sway.
In later sections, it will be argued that to some extent this conflict can be minimised, and a number of suggestions will be made as to how this might best be achieved. However, there will remain a range of circumstances where it is probably impossible to identify any commonly agreed position. The best that can be done here is to identify the writer’s own value position and then attempt to develop a set of principles which would most likely achieve those values.
The value position of this article is that the primary purpose of prosecution is preventative: to reduce the level of work-related injury and disease. Although it does not reject retribution in its entirety, it suggests that, to the extent that the two principles are in conflict, prevention should be given precedence. Those who believe that the principal role of the criminal law is retribution will likely disagree with the analysis that follows. Its virtue, however, is to identify principles which, if followed, will send a set of signals that deter ‘bad actors’ from wrongdoing without inhibiting ‘good actors’ — or even those capable of becoming good actors under the right circumstances — from pursuing strategies conducive to improved workplace safety and health. Overall, it is submitted that implementation of the principles identified below will prove effective in preventing accidents, whilst also (to the extent that it is consistent with achieving prevention) recognising the moral and expressive needs of victims and their families for ‘justice’.
Accepting for present purposes that the primary role of prosecution is prevention, how should prosecutions be targeted? What general principles and approach should shape enforcement policy? What criteria should determine the sort of non-compliance which merit action at the top of the enforcement pyramid and which do not? These are hardly new questions, but they are ones which have not been satisfactorily addressed under current prosecution policy in the mining industry.
One critical question to ask in this context is: what degree of mental culpability on the part of the defendant is sufficient to justify, either in law or as a matter of administrative practice, the regulator defaulting from an ‘advise and persuade’ approach, or gentle (and increasingly less gentle) prodding at lower levels of an enforcement pyramid, to a punitive, prosecution-oriented approach? This question is a complex one, which will be answered first by describing the current legal approach and identifying its limitations, and then by proposing an alternative strategy.
Under the legal and regulatory status quo, the body of law which is relied upon to define culpability, and form the basis for prosecution, is OHS legislation. This includes the mainstream OHS law and specialist mining statutes which together form the legal regime relating to mining OHS in New South Wales and the specialist and mine-specific OHS legislation of Queensland and Western Australia.
Under that legislation, the employer and various other duty holders must comply with a set of general duties (in all jurisdictions except Queensland) or risk-based standards (in Queensland), with this obligation being either (a) one of absolute liability (subject to certain defences including that of reasonable practicability), as in New South Wales, or (b) subject to a ‘reasonable practicability’ qualification, as elsewhere.
The main distinction between these two approaches is that under (a) above, the onus is on the defendant to invoke one of the defences (including that it was not ‘reasonably practicable’ to comply) on the balance of probabilities, whereas under (b) it is for the prosecution to prove all aspects of its case (including that the defendant failed to do what was reasonably practicable) beyond reasonable doubt.
Under both versions, the mental element required to establish guilt is in effect the standard of negligence in civil law (‘reasonable practicability’ being seen by the courts to be in effect a codification of the negligence calculus). Negligence, on this standard, involves a failure to live up to the standards of the ‘reasonable person’ acting in the circumstances of the case. It is how the negligence standard (or in legal terms under New South Wales law, the ‘reasonable practicability’ and ‘due diligence’ defences) has been interpreted in the Gretley case, summarised below, that have given rise to considerable controversy and angst within the mining industry.
The facts of the Gretley disaster and the subsequent judicial findings are well known and can be stated briefly. Four miners at Gretley colliery punched into old and flooded mine workings. There was an in-rush of water and the miners were drowned. An inquiry into the incident by former Justice James Staunton made recommendations concerning prosecution and charges were subsequently brought in the New South Wales Industrial Commission, both against the two former operating companies and against a number of individuals. Commissioner Justice Patricia Staunton found that the corporate defendants had failed to ensure the health, safety and welfare of their employees, and two former mine general managers and a mine surveyor were ‘[d]eemed to have committed the same offences as the corporations, having failed to satisfy the onus placed upon them’ to exercise due diligence to protect workers (McMartin v Newcastle Wallsend Coal Company Pty Ltd  NSWIRComm 202 at 979). Although the defendants argued that they were entitled to rely on old plans of the old workings supplied by the relevant government agency, Justice Staunton found that this:
[D]oes not excuse the defendants from their independent statutory obligation ... to ensure a safe system of work. Nor does it relieve the defendants of their obligation to satisfy themselves by way of their own research as to the accuracy of ... [the Department of Minerals and Resources plans which] [o]n any considered view ... were seriously deficient in purporting to depict old coal workings in a way that one could be confident of their accuracy ( NSWIRComm 202 at ).
On appeal to the Full Bench of the Industrial Court of New South Wales, the conviction against the two companies was affirmed, as was that against the mine manager and former mine manager. The conviction of the surveyor was overturned on the basis that he was not ‘concerned in the management’ of either company. His role was ‘not managerial, but rather more akin to that of an advisor or consultant to mine management in relation to surveying’ (Newcastle Wallsend Coal Company Pty Ltd v McMartin  NSWIRComm 339 at ).
Because prosecutions under OHS legislation take place at a relatively low point in the culpability hierarchy (that is, they are usually based on negligence rather than on intent or recklessness), the penalties imposed themselves have tended to be low, particularly against individuals. This sends out the unfortunate signal that breaches of OHS law are ‘not really criminal’. Low penalties are also ‘indicative of the inherent difficulty associated with assessing the appropriate penalty ... where conviction is not the result of individual criminal culpability in the normally understood sense’.
However, in New South Wales, recent political pressure for increased levels of prosecution and higher penalties has resulted, particularly, but not exclusively, in the Gretley decision described above, in substantial penalties being imposed both on the operators and owners and on an individual manager. The fine of $42,000 imposed on the mine manager in Gretley was a substantial one for an individual. But even if it had been less, an individual mine manager (who is unlikely to fall foul of the criminal law in any other context) is likely to experience such prosecution as a traumatic event. As a result, fear of such prosecution is in the forefront of many managers’ minds.
If such penalties serve to send the message that OHS offences really are criminal, then this is no bad thing. But the way that negligence was interpreted in the Gretley case had caused not only considerable angst within the mining industry (which in itself might only suggest that the law is now finally having ‘bite’), but also a number of related adverse consequences for preventative OHS. For example, much energy is expended by companies on defensive training on matters such as how to avoid self-incrimination and disclosure of the circumstances relating to an alleged breach and on a number of other strategies described earlier that are antithetical to OHS.
The mining industry’s position is that the prosecutions (both against the companies and against particular individuals) were inappropriate in the circumstances of the case. However, their greatest concern relates to the prosecution of the individual statutory duty holders (for example mine managers). The industry’s perception is not only that these individuals have been held liable in circumstances where their level of culpability is relatively low (negligence to civil standard), but that the interpretation of the negligence standard by the courts is an unreasonable one, making the prosecutions wholly unjustified. Indeed, Hopkins goes so far as to claim that we have arrived at a point where individuals are held culpable ‘for failure to live up to an idealised reasonable person standard’. Similarly, one industry association has argued that ‘the concepts of ’reasonably practicable’, ’foreseeable’ and ’control’ have been significantly distorted ... to the point where they no longer reflect what is reasonable, practicable and achievable’.
The industry’s position may in some respects be overstated. As regards prosecution of the companies in Gretley, the focus of the courts in recent years (in interpreting ‘reasonable practicability’) has been on the necessity for systematic occupational health and safety management and risk management. This is entirely consistent with industry OHS initiatives and best practice. In the circumstances of Gretley, where if the risk of water in-rush eventuated, the likely consequences included multiple deaths, then it seems ‘reasonably practicable’ for the company to have had systems and risk management arrangements which would have led them to challenge the information contained in problematic government plans. However, whether individual mine managers were negligent, is more doubtful. Certainly it is arguable that a mine manager who relied on the judgment of the surveyor (that there was no need to check those plans, which the government had supplied) had discharged his obligation to demonstrate ‘due diligence’, although it is far less obvious that the surveyor himself could have done so. The situation is muddied in this regard by the fact that the original surveyor (who was terminally ill by the time of the court case) was not prosecuted, and by the finding of the appeal court that the other surveyor was not ‘concerned in the management’ of the enterprise and so could not be liable.
Nevertheless, although the culpability of the mine managers in the Gretley case is a matter on which reasonable people can disagree, there is little disagreement (i) that the industry’s perception of the Gretley decision is that it was grossly unfair; (ii) that as a result the regulatory system has lost legitimacy in the eyes of many duty holders; and (iii) that they (including leading companies who by and large aspire to go beyond compliance) have responded by invoking a number of defensive strategies described above, many of which are antithetical to the cause of preventative OHS.
One may conclude that the current approach to prosecution is counter-productive, and there is a strong argument to be made for developing a very different approach in the interests of improved OHS. But what would this involve? At what point in the culpability hierarchy should prosecutions take place? Answering this question involves a delicate balancing act between the virtues of deterrence on the one hand, and encouraging open reporting and investigation, nurturing a safety culture and maintaining the industry skills-base, on the other.
Perhaps some guidance as to where this balance should be struck is to be found in James Reason’s well known argument in favour of nurturing a ‘just culture’ in relation to OHS. Reason emphasises that ‘valid feedback on the local and organisational factors promoting errors and incidents is far more important [to improving safety] than assigning blame to individuals’. However, he also recognises that an undiscriminating, across-the-board ‘no blame’ culture is neither feasible nor desirable. A small proportion of human unsafe acts are egregious and warrant sanctions, so what is needed is not a blanket amnesty on all unsafe acts, but a just culture which generates:
... an atmosphere of trust in which people are encouraged, even rewarded, for providing essential safety-related information ¯ but in which they are also clear about where the line must be drawn between acceptable and unacceptable behaviour. [Emphasis added.]
Whilst this advice was intended to apply to internal corporate management, it may also prove useful in a regulatory context, where it can be argued that the line should be drawn at a point that would encourage reporting and avoid the sorts of defensive individual and corporate behaviour documented earlier, whilst making clear that behaviour which departed substantially from reasonable expectations, would not be countenanced. This has been the approach of a number of high reliability organisations which are distinguished by their exemplary OHS performance. For example, British Airways Flight Crew Order 608 suggests that disciplinary action should only be taken against an employee where they have taken action or risks which, in the Company’s opinion, no reasonably prudent employee with his or her training and experience would have taken.
To ensure that the line is drawn so as to only encompass serious departures from reasonable expectations and to avoid the sorts of borderline decisions that caused such counter-productive consequences following Gretley, it is submitted that only cases involving a substantial falling short of reasonable expectations should merit prosecution. That is, prosecutorial discretion should only be exercised when there is at least this degree of culpability. To avoid ambiguity, a broader elaboration of what is contemplated might be provided in prosecution guidelines. Crucially, these would make clear that prosecutions would be contemplated where there has been system failure (that is, a failure by the enterprise to engage systematically with OHS issues) as well as in the case of individual failings. Under such a test, it is suggested that the employers in the circumstances of Gretley would still be culpable (for there was indeed a substantial falling short of what was required to prevent systemic failure) whilst the mine managers would not.
However, it is not suggested that culpability alone should be sufficient to justify prosecution. For reasons which will be explored below, two other considerations should be weighted in the balance: the degree of risk and the defendant’s past OHS performance.
Risk is important because some failures on the part of a duty holder expose others (usually workers) to a substantial risk of serious harm, while other failures are of far less significance. Inspectoral resources are scarce and prosecution is a particularly expensive and time consuming activity, which must be reserved for those cases where it is likely to have most effect. Cases involving serious risk with potentially severe consequences fall within this category. This category extends to circumstances where, although the degree of risk is not particularly high in any individual identified circumstance, nevertheless, the general quality of occupational health and safety management demonstrated by the enterprise is poor. Systemic failure on the duty holder’s part is a matter for serious concern, particularly where ‘a duty holder’s standard of managing health and safety is found to be far below what is required by health and safety law and to be giving rise to significant risk’. Thus, a focus on risk enables substantial consideration to be given to prosecutions which target the failure to deal with crucial issues such as management systems and risk control, and which are geared to promote the proactive and preventative ‘systems-based’ aspects of OHS management.
The final factor which should be weighted in the decision to prosecute is the past OHS record and approach of the defendant. All else being equal, there is a stronger case for prosecuting those whose past OHS record has been poor (as evidenced by a history of incidents, warnings and other documented action by the inspectorate) than those whose previous performance has been a positive one. Where, in the words of the UK Health & Safety Commission, ‘there have been repeated breaches which give rise to significant risk, or persistent and significant poor compliance’ or failures to comply with improvement and prohibition notices or their equivalent, or ‘a breach which gives rise to significant risk has continued despite relevant warnings from employees or their representatives, or from other affected by a work activity’, then this should weigh substantially in the decision to prosecute.
Of course, real world cases do not present themselves in neat, clearly labeled categories, and neither the prospective defendant nor the circumstances of the case are likely to enable categorisation at the extreme of culpability, risk severity or leader-laggard continuums. Faced with shades of grey rather than black and white, regulatory decision-makers must weigh competing considerations. Does an employer who is (seemingly), seriously negligent but not reckless, who has a reasonable (but not impressive) past safety performance, and who exposed a single employee to a fleeting but serious risk merit prosecution? What of someone who was reckless and exposes multiple employees to a serious risk with potentially severe consequences, but has an exemplary past record? Does an uncharacteristic failure (even involving substantial negligence) merit action at some lower point on the pyramid, and does one which seems part of a pattern of neglect going back some time, deserve prosecution?
The best that can be done when making hard choices is to use a calculus as a guide to enable a preliminary decision (that is, the higher the composite score in terms of culpability, risk and poor past record, the higher the chance of prosecution). However, mechanistic approaches to decision making have their limitations and numerical calculations can give rise to a spurious impression that a scientific and objective judgment has been reached. In practice, putting things to numbers provides useful guidance and a frame of reference on which to base provisional decision-making, but not complete answers. There is no alternative but to rely upon human judgment, and the best that can be hoped for are consistent, transparent decisions, made in accordance with clearly stated prosecution guidelines, made at arms length by an expert committee.
The approach proposed, involving a weighing of three different considerations, does not preclude prosecution where serious injury or death has eventuated, but nor does it privilege what Johnstone describes as ‘event’-based prosecution (that is, those that follow particular incidents, usually involving injury or death) over ‘pure risk’ prosecution (that is, those that focus on the risk of injury or death rather than upon consequences). It also ensures that some prosecutions will be undertaken in large part as a result of the defendant’s unsatisfactory past OHS record. For example, a duty holder with a history of non-compliance with OHS regulation, who has been negligent and exposed a worker to a substantial risk, might justifiably be prosecuted applying the calculus, whereas one with a better past record might not.
The importance of taking account of a duty holder’s track record in a substantial proportion of prosecutions cannot be over-emphasised. The essence of responsive regulation and the enforcement pyramid referred to above is that, where persuasion fails, then a tit-for-tat strategy involving a gradual escalation must ultimately result in the prosecution of the recalcitrant at the top of the pyramid, otherwise action at lower levels will lose its credibility and the entire strategy is likely to fail. That is, it is only if OHS duty holders believe that persuasive and administrative enforcement mechanisms at the lower levels of the pyramid are being backed up by big sticks at the top of the pyramid, that specific and general deterrence are credible, and that ‘rational calculators’, in particular, are given sufficient incentive to comply with their legal obligations.
Of course, under the sort of calculus proposed above, not all prosecutions will be influenced by track record. A duty holder who intentionally or recklessly exposed a worker to a serious risk of severe harm might be prosecuted, even in the absence of a poor past safety record, but one who was less culpable (and for this reason had a lower aggregate score) might not. For example, permitting work underground without adequate roof support would merit prosecution even if no harm eventuated and even if this was seemingly out of character. The point is not that such prosecutions are inappropriate, but rather that they must be complemented by prosecutions in which track record is a substantial consideration, for it is these which give the enforcement pyramid credibility.
What must be emphasised above all else about the above calculus is that it would result in the prosecution of only the worst cases, or in Reason’s terms the ‘small proportion of human unsafe acts [that] are egregious, and warrant sanctions’. By doing so, it would emphasise that OHS offences really are criminal and should be punished as such, that the enforcement pyramid really does have ‘bite’ and provides a credible deterrent, but that at the same time, the large majority of duty holders, even when they break the law, can be dealt with effectively at much lower points in the pyramid and without need for prosecution. In this manner, it should be possible to demonstrate the fairness of the prosecutorial approach, to preserve the legitimacy of the law and to avoid the counter-productive consequences that have flowed from the industry’s response to the Gretley decision.
Strikingly, it is doubtful whether many of the principal cases prosecuted to date would have justified prosecution under the calculus proposed above (although there are other cases that would — and should, in the writer’s view — be prosecuted under this calculus). In Gretley, for example, there was, at most, negligence to the civil standard, and the defendants would have ranked low in terms of track record. As has been emphasised:
[T]he judge found that the defendants, both corporate and personal, were generally safety conscious; the company had an effective safety management system; there was “an active workplace safety culture among employees and corporate defendants”; and workers were encouraged to cease work when they encountered a hazard.
Thus, it is only in terms of risk and potential severity of consequence that the case would have received a high scoring. It is reasonable to suggest that the low scoring it would have received on the other two variables go a substantial way towards explaining the strength of feelings the case has aroused and to the decision’s perceived lack of legitimacy in the eyes of employers.
Finally, in terms of policy reform, the approach proposed above could be achieved without any change to legislation, by adopting formal prosecution guidelines and an enforcement policy incorporating these principles. It is, in fact, not wildly different from some of the existing state enforcement policies. However, many of these, while ‘touching the right bases’ do so in ways which provide relatively little guidance in actual decision-making. For example, the Queensland Department of Natural Resources, Mines and Water Compliance Policy (2001) identifies the need for enforcement measures which are commensurate with the seriousness of a situation and the need to escalate where previous measures have been ineffective. It also emphasises the need to assess the level of risk, the seriousness of the situation and the immediacy of the problems detected. These are helpful parameters which are coupled with assessment-response guidelines and procedures for assessing recommendations. However, insufficient attention is given to the relative weight of different factors, leaving very large discretion to decision-makers. This in turn has enabled an ‘advise and persuade’ policy to prevail in almost all circumstances.
Under the calculus set out above (and in contrast to a number of existing enforcement policies), it was made clear that the actual severity of harm caused — or the absence of harm — should not be regarded as a material factor in determining whether to prosecute. There are a number of reasons for this.
The OHS legislation clearly provides for prosecution irrespective of whether injury eventuates. Its intent is to prevent exposure to the risk of harm (for example through an unsafe work system) and it is the seriousness of the risk, rather than the actual consequences of the breach, that are its concern. There is considerable judicial support for this view. For example, the Full Court of the Industrial Court has stated, in no uncertain terms, that:
The general duties created by the Occupational Health and Safety Act, such as in ss 15 and 16 (now sections 8 and 9), are clearly directed, we think, at obviating “risks” to safety at the workplace; it would be therefore wrong in considering whether an alleged breach of those general duties had been made out in a particular case to reason from the actual occurrence of an accident ... The accident may well, and probably does, manifest the existence of a detriment to safety and will, no doubt, be some measure of the degree of severity of the detriment; but it seems to us, it is to the essential ingredients of the offence charged which one must attend by assessing the objective facts causing the detriment to safety and the causal connection therewith of the employer.
Yet in New South Wales, prosecutions have largely been undertaken, and at one stage exclusively, as a reaction to a work-related death or serious injury. This is not only contrary to the spirit of the legislation, it is also undesirable on broader policy grounds. Johnstone in particular has argued that what he terms ‘the reactive and event-focused emphasis of OHS prosecutions’ does little to take account of the importance of systems of work or OHS management systems, but instead constructs OHS contraventions as a chain of specific actions leading to a specific injury or death:
Consequently, arguments in mitigation of penalty use ‘isolation techniques’ which shift the sentencing court’s attention away from an analysis of the failure of the OHS system, to scrutinising the minute details of the events leading to the inquiry. This enables defendants to shift blame onto workers and others and facilitate uncontested claims to be good corporate citizens; coupled often with the allegation that the accident was a “freak” or “one off”.
As was argued earlier, it is only when an event-based focus is largely replaced by ‘risk-based’ prosecution, in accordance with the calculus set out above, and under a genuinely pyramidal approach, that prosecution can be used to optimal preventative effect.
This is not to suggest that events and their consequences will invariably be unimportant as regards the decision to prosecute. There remains an exceptional category — offences where the duty holder’s state of mind — coupled with the consequences of their act or omission — are so heinous that the full weight of the criminal law should be brought to bear upon them.
While many would argue that the principal justifications for prosecution in these circumstances are ‘moral, symbolic and retributive, and show society’s intolerance for organisational behaviour causing workplace deaths’, there is also a preventative rationale for prosecution in these circumstances: namely that it is desirable to identify a class of offences at the top of the enforcement pyramid that are so heinous that the full weight of the ‘real’ criminal law can be applied to them. Prosecution here is also seen to address ‘public disquiet... about the leniency afforded to workplace deaths in comparison to other forms of homicide occurring outside the workplace’.
According to the McCallum Committee, the failure of sentencing patterns to keep pace with legislated increases in maximum penalties, particularly in relation to cases involving workplace death, and the apparent associated failure of general deterrence, makes special legislative provision for such cases essential. This sort of reasoning has led to demands for the introduction of an additional tier of liability for offences which are ‘really criminal’ (that is, involve intent or recklessness coupled with serious consequences: severe injury or death). A number of jurisdictions have been exploring this general approach in recent years, particularly with regard to the introduction of a new offence of ‘industrial manslaughter’.
Consistent with this general approach, New South Wales enacted the Occupational Health and Safety Amendment (Workplace Deaths) Act 2005. This legislation amended the OHS Act 2000 (NSW), the Occupational Health and Safety Regulation 2001 (NSW) and the Criminal Appeal Act 1912 to include a new offence with a maximum penalty of $1.6 million for corporations and $165,000 and/or imprisonment of 5 years for individuals, where a breach of safety legislation results in death at a workplace. According to the Minister, this targets the small minority of employers (so called ‘rogue employers’) who demonstrate little or no regard for the safety of their workers — and are reckless or intentional in their behaviour. However, the introduction of this legislation does not (at least in principle) diminish the role of prosecution under the OHS Act with regard to reckless conduct in the absence of death (or injury).
Nevertheless, the relationship between OHS regulation and the ‘real’ criminal law (such as the Workplace Deaths legislation) is an uncomfortable one, with relatively light penalties under the former being juxtaposed with potentially heavy penalties under the latter in such a way as to suggest that OHS offences are minor offences, and not ‘really criminal’.
While the tension between criminal and OHS law cannot be fully resolved, the use of both types of prosecutions in tandem will at least avoid the suggestions that corporate offenders are not subject to mainstream criminal law. The fact that the provisions of the Occupational Health and Safety Amendment (Workplace Deaths) Act 2005 form part of the OHS Act 2000 (as amended) can only assist such integration. Moreover, if prosecutions under OHS statutes are only taken (as proposed above) where there is (at the very least) substantial negligence, coupled with serious risk and poor track record, then prosecutions will only take place in circumstances involving serious wrongdoing. If so, then one can reasonably expect penalties to increase to reflect this fact. This too would serve to reduce the perception that such behaviour is ‘not really criminal’.
Beyond this, and in terms of the framework of this article, the introduction of ‘Workplace Deaths’ or similar legislation, at the very least serves to integrate the mainstream criminal law into the pyramid and to maximise the deterrent effect of the top of the pyramid.
As indicated earlier, prosecution also has a greater deterrent impact in relation to individuals than to organisations, but much depends upon the context. Accordingly, although there are sound arguments in favour of prosecutions against individuals in some circumstances, it remains crucially important that the appropriate level of management is targeted for prosecution; otherwise the outcome may not only be unjust but may also fail to send the deterrent message to the right audience.
In New South Wales, the practice has been to target mine managers, surveyors, and under-managers on the basis that they are ‘concerned in the management’ of the corporation. This may often be inappropriate, because mine managers are usually (but not invariably) too low in the managerial hierarchy to be responsible for the major decisions that often contribute to death or injury at the workplace. For example, it is usually senior personnel at the corporate level who have the capacity to make the necessary decisions concerning catastrophic hazards or to grant or deny the resources necessary to address a major OHS issue at site level. In contrast, mine managers, under-managers and surveyors operate under constraints and incentives set by mine owners, boards and chief financial officers. One mine manager encapsulated a common view when he told us ‘I have real issues because I don’t agree with the layout plan, but I have been told by my superior that that’s the way it is. Yet the reality is that I am the statutory person who would take the fall, for a system I don’t agree with’.
Certainly there are exceptions and Gretley was arguably one of them. With regard to day to day management matters, the mine manager (though far less than the surveyor) is well placed to ensure that various safety rules and procedures are discharged, and could fairly be held responsible if, for example, personal protective equipment is not being used, or support rules were not being complied with. But for the most part, it will be more appropriate to prosecute senior corporate management than site level officials, since major OHS decisions are the responsibility of higher level management.
However, if the regulator targets directors and senior managers (or in Queensland, ‘executive officers’), then another difficulty arises. For, although the legislation deems directors and managers of corporations personally to have committed an offence, if a corporation breaches the Act, it also provides them with a defence if they can establish ‘due diligence’, or that they were not in a position to influence the conduct of the corporation in relation to the offence. The problem here, as experience both in OHS and in the comparable area of environmental regulation makes clear, is the organisational complexity of many corporations, which will often provide a shield for directors and managers which is difficult to penetrate. It is no coincidence that the few cases under safety, health and environmental legislation that have involved substantial penalties against individuals have all involved very small enterprises — the only ones where the ‘corporate veil’ can credibly be broken down. However, in Inspector Kumar Ken v David Aylmer Richie, a CEO was prosecuted and found guilty because he did not have detailed knowledge of work procedures to be used when cleaning out containers- albeit that this was in circumstances where he was far removed from the business in terms of this level of detail and relied on others to enforce such procedures. Whether this case is an exception or the beginning of a new trend remains to be seen.
As the McCallum Committee has pointed out, this problem might be overcome by differently designed statutory provisions intended to establish relevant occupational health and safety standards for responsible risk management, integrated with appropriate liability principles. Exploring options in relation to the Workplace Deaths legislation described above, they propose the adoption of a Code of Practice creating benchmarks ‘against which liability issues can then be evaluated in terms of culpability and the scope for escaping liability but only in circumstances where it can be demonstrated that a manager/director has been relevantly proactive’ (Emphasis added.) Under this approach a complex hierarchy would no longer provide de facto immunity. On the contrary, in order to establish a defence to personal liability, directors and managers would need to establish that they have relevantly discharged their individual responsibilities in implementing the relevant safety management system in such a manner as to ensure compliance with an objective and measurable code to prevent a corporate culture of risk from developing.
It might appear that the New South Wales and Queensland mine-specific legislation is already heading in this direction, in so far as there are statutory responsibilities relating to the creation and implementation of both OHS management systems and with regard to hazard management plans, and requirements to establish a management structure and register persons occupying positions. However, most specific responsibilities are imposed at too low a level in the corporate hierarchy. For example, under the 1999 Queensland legislation, particular statutory responsibilities are placed on the Senior Site Executive. Certainly identification of a responsible officer gives such a person an incentive to request the enterprise itself to take any necessary steps to protect OHS, including providing sufficient resources to maintain compliance. Furthermore, failure to take such steps would provide evidence of lack of due diligence.
However, a limitation nevertheless remains in that the Senior Site Executive in a large enterprise may have only very limited decision-making power and it is more senior managers and Board members that remain the most appropriate targets for individual liability. The identification of responsible individuals at corporate level will not be easy, and the successful prosecution of directors and senior managers remains problematic under the present system.
Although the above analysis has focused on the role of prosecution in preventing work-related injury and disease, (in particular by providing specific and general deterrence), some would argue that prosecution can and should fulfill a further role — that of retribution. Retribution is sometimes seen as appropriate where society seeks to exact vindication by punishing acts considered egregious, to express moral outrage and to reaffirm a commitment to the maintenance of legal and moral standards. For example, according to Kruse & Wilkinson, there are:
[A] very small number of occasions when prosecution must take place. These are when the seriousness of the breach of the law and or consequences of that breach is such that there is widespread public opprobrium that demands public retribution.
Thus, the goal for those who seek retribution is not an instrumental concern to improve future OHS performance, but rather to satisfy feelings of revenge and to achieve ‘justice’ in the victim’s (or their family’s) terms.
Retribution is widely seen as a significant objective of criminal law, and judges frequently invoke it as an important consideration in sentencing. The argument of this chapter is not that retribution is never appropriate, but rather that it is only appropriate in a limited number of circumstances. Applied beyond these circumstances, it will be argued, its use tends to be antithetical to prevention and for this reason, undesirable.
So in what circumstances is retribution appropriate? Because moral and symbolic rationales underlie retribution as a justification for criminal punishment, the defendant’s culpability should be a crucial consideration. For this reason, those who were reckless in their approach to the health and safety of the workforce, and whose behaviour results in serious injury or death, are appropriate targets for retribution. Theirs is the sort of egregious behaviour targeted by the New South Wales Workplace Deaths legislation (confined to offences involving recklessness or intent), by manslaughter under the mainstream criminal law, or contemplated in most proposals for an offence of industrial manslaughter. However, when retribution is extended substantially beyond these circumstances for instance, to cases where the culpability of the defendant is low then, as described earlier in this chapter, it produces results which are widely seen to be unjust, undermines the general belief in the legitimacy of regulatory requirements, and has consequences which are antithetical to preventative OHS.
There is now considerable evidence that there is a better means than retribution in meeting the legitimate needs of victims or their families and communities for justice in the aftermath of a disaster: restorative justice.
John Braithwaite, who pioneered this approach, argues with considerable empirical support that approaches to regulation that seek to identify important problems and fix them work better than those which focus on imposing the right punishment or ‘just desserts’. For example, as was argued in the previous section, beyond a very limited range of circumstances, retribution does not ‘work well’, both because it is widely perceived to be unfair and because it has counter-productive consequences for prevention.
Yet at the same time, if prevention trumps prosecution and retribution is rejected, then the legitimate concerns of victims and their families for justice, may be ignored. Braithwaite recognises this, and suggests that there is a need for others to ‘listen to the stories of our hurts’ before we can move on to solve the problem. In his view, restorative justice is ‘a process whereby all the parties with a stake in a particular offence come together to resolve collectively how to deal with the aftermath of the offence and its implications for the future’ thus showing us the practical paths for moving from healing to problem solving.
Now is not the place for a detailed analysis of restorative justice, but it is apposite to draw from Braithwaite’s own work on the enforcement of coal mine safety in the USA, to suggest the specific application of restorative justice techniques in the mining context. Braithwaite argues that what is needed is the creation of restorative justice mechanisms such as community conferences in which workers, victims and their families participate with management (including senior management) in a dialogue about what went wrong and what should be done to make sure it never happens again. He points to the experience in British pits where he found that safety leaders were companies that ‘not only thoroughly involve everyone concerned after a serious accident to reach consensual agreement on what must be done to prevent recurrence but also did this after ‘near accidents’ as well as discussing safety audit results with workers even when there was no near accident.’ He concludes:
After mine disasters... so long as there had been an open public dialogue among all those affected, the families of the miners cared for, and a credible plan to prevent recurrence put in place, criminal punishment served little purpose. The process of the public inquiry and helping the families of the miners for whom they were responsible seemed such a potent general deterrent that a criminal trial could be gratuitous and might corrupt the restorative justice process that I found in so many of the thirty-nine disaster investigations I studied.
In terms of the themes of this article, Braithwaite also connects the role of restorative justice with the enforcement pyramid. He argues that what persuades offenders to participate in restorative justice dialogues and processes at lower levels of such a pyramid is their awareness that the alternative is escalation to more punitive responses. In his view ‘offenders who know that they will benefit from ... mercy so long as they participate in a high-integrity process of truth-seeking and take active responsibility for the hurts they have caused can help us to learn from the truth they tell’. The result is that by persuading offenders to embrace restorative justice techniques in the lower parts of the pyramid, future preventative safety is substantially enhanced and the need for retribution obviated.
This article, which is part of a broader project on regulation and enforcement in the mining industry, has sought to identify a number of principles relating to prosecution policy that would achieve more effective OHS outcomes. In particular, it has sought to steer a middle path that neither rejects prosecution as an important deterrent at the top of the enforcement pyramid, nor uses it in circumstances where it is likely to do more harm than good.
Achieving such a balanced approach is not easy. On the one hand, the evidence suggests that the sort of extreme ‘advise and persuade’ policy that the Queensland and Western Australian inspectorates have favoured will fail to send appropriate deterrent signals to the recalcitrant. On the other hand, the sort of zealous prosecution policy that New South Wales has applied to fatalities will also fail in preventative terms. As we have seen, vengeful prosecution against those who neither intended harm nor were reckless in their behaviour (epitomised in the Gretley decision) is widely perceived to be unjust, and this has caused the law to lose its legitimacy in the eyes of duty holders. It has also generated a defensiveness on the part of duty holders that results in an unwillingness to examine the root causes of accidents and incidents for fear of being prosecuted.
This article has proposed an alternative approach to prosecution which (i) focuses on risk rather than consequences; (ii) which takes previous track record seriously (and makes escalation up an enforcement pyramid credible); and (iii) which emphasises that prosecution should not take place in the absence of culpability. For these purposes, it has been argued that culpability should mean a substantial falling short of reasonable expectations (a form of negligence), recklessness or intent. The actual decision to prosecute, it has been suggested, should be based on a calculus which takes account of all three of the above factors.
This approach would ensure that prosecution takes place even where no injury results (exposure to risk, irrespective of consequences, being at the heart of OHS regulation). It would also enable the inspectorate to target failures of risk management, and on general patterns of failure to attend to risk despite warnings, while also reserving the right to take action in the absence of poor past history if there was high culpability (intent or recklessness) coupled with a high degree of risk or potential for extreme consequences. Such an approach would do much to restore legitimacy to the prosecution process, whilst ensuring that serious breaches of OHS legislation, and those who did not give serious attention to complying with OHS law, were firmly dealt with.
This approach need not imply the need for multiple prosecutions, because the literature suggests that a distinction must be made between the actual chances of detection and punishment, and the perceptions thereof. What is important is the belief that duty holders have of the likelihood and degree of punishment, even if, in actual fact, that belief is overstated. Even a handful of prosecutions in the course of a year can achieve this effect provided the ‘right’ cases are chosen. That handful of prosecutions will, however, play a crucially important role at the tip of an enforcement pyramid, for without them less coercive policies at the lower levels of the pyramid lose their credibility.
The argument so far has assumed that the principal role of prosecution is as an instrument for achieving prevention, and that it will achieve this outcome largely because of its capacity to provide an effective deterrent. Against this, it might be argued that deterrence does not apparently work across the board, and that it is not necessary in all cases. However, this is not an argument against the need for prosecution, but rather for targeting it to circumstances and actors where it is most likely to be effective. Because the calculus approach described above pays particular attention to track record, it is well equipped to achieve such targeting. It is particularly important that it does so because prosecuting those who aspire to go beyond compliance rather than reluctant compliers or the recalcitrant, can be particularly counter-productive. The incompetent (usually small and medium sized enterprises) present particular problems because prosecution is not well suited to bring about basic levels of competence. On the other hand, license removal for the seriously incompetent, is a strategy well justified in preventative terms.
Finally, it has been suggested that prevention and retribution are not comfortable bedfellows. Prevention is instrumental, while retribution is moral, expressive and symbolic. Sometimes, what satisfies calls for retribution will be antithetical to prevention. Certainly there are circumstances where retribution is appropriate, namely egregious breach coupled with severe consequences (especially death). If the use of retribution is confined to these circumstances then it can co-exist comfortably with prevention. If it is extended beyond them, it cannot. The Gretley decision was used to illustrate precisely this point. In circumstances such as Gretley, where the degree of culpability of the defendants was low, then the legitimate concerns of victims, their families, and communities can more constructively be addressed through applying the techniques of restorative justice.
An optimal prosecution policy must take account of all the above factors and achieve a trade off between competing considerations, and in some circumstances it must choose between prevention and deterrence. Implementing such a policy, particularly in the emotionally charged atmosphere of the mining industry will not be easy. On the other hand, the adverse implications for preventative safety of the prosecutorial status quo, suggest that there is no credible alternative.
[∗] Neil Gunningham is Director of the National Research Centre for Occupational Health and Safety Regulation within the Regulatory Institutions Network, Australian National University.
 I am grateful for the comments of Neil Foster, Richard Johnstone and Darren Sinclair on an earlier draft. They are of course in no way responsible for the views expressed or for any remaining errors. This research was conducted with the support of an Australian Research Council Linkage Grant.
 NSW Department of Mineral Resources, Gretley Inquiry Report: Final Report (1998) (‘Gretley Report’).
 Formerly the Department of Mineral Resources.
 NSW Department of Primary Industry, The Enforcement of Health and Safety Standards in Mines (1999) <http://www.dpi.nsw.gov.au/__data/assets/pdf_file/82612/Enforcement-of-Health-and-Safety-Standards-in-Mines.pdf> accessed 2 Aug 2007.
 The Investigation Unit was established following the recommendations of the 1997 Mine Safety Review (ACIL 1997) but it was the Gretley Report, above n2, that promoted a reconsideration of the role of prosecution, particularly of individuals. The Unit is also responsible for investigating a range of ‘prescribed matters’. See NSW Department of Mineral Resources, The Investigative Unit – general facts (2005) <http://www.minerals.nsw.gov.au/_data/page/22/Investigations_Unit_FINAL.pdf> accessed 20 October 2006.
 On the different perspectives relating to the criminalisation of organisational conduct causing the deaths of workers, see the special issue of the Flinders Journal of Law Reform: ‘Industrial Manslaughter’ (2005) 8 FJLR at 1–113; See also Neil Foster, ‘Manslaughter by Managers: The Personal Liability of Company Officers for Death flowing from Company Workplace Safety Breach’ (2006) 9 Flinders Journal of Law Reform at 79–111.
 Australian Chamber of Commerce and Industry, Modern Workplaces: Safety Workplaces – An Industry Blueprint for Improving OHS in Australia (2005) <http:www.acci.asn.au/text_files/review/r123.pdf> accessed 19 October 2006.
 Id at 3.
 For a nuanced historical account of the evolution of present enforcement practices and the relationship of OHS offences to the ‘read’ criminal law, see W G Carson, ‘The Conventionalisation of Early Factory Crime’ (1979) 7 International Journal of Sociology of Law 37. See also Neil Gunningham & Richard Johnstone, Safeguarding the Worker (1984) at Chapter 4.
 Keith Hawkins, Law as Last Resort: Prosecution Decision Making in a Regulatory Agency (2002) at 416–417.
 Principles for more effective OHS prosecutions, with a somewhat different focus (how to create a ‘big stick’ at the apex of an enforcement pyramid) were proposed in Neil Gunningham & Richard Johnstone, above n9, at Chapter 8.
 Gretley Report above n2 at 694.
 In the absence of public records for much of this period I am indebted to information provided by Graham Terrey, former Chief Inspector of Mines, Department of Mineral Resources, NSW.
 CFMEU National Office, Submission to the NSW Review of Mine Safety (2004) at 71 <http://www.cfmeu.asn.au/> accessed: 19 October 2006.
 McMartin v Newcastle Wallsend Coal Company Pty Ltd and other matters  NSWIRComm 202.
 Newcastle Wallsend Coal Company Pty Ltd and Ors v McMartin  NSWIRComm 339.
 Morrison v Powercoal Pty Ltd  NSWIRComm 297.
 See for example, Morrison v Peter Dale Ross; Morrison v Glennies Creek Coal Management Pty Ltd  NSWIRComm 205.
 See for example Department of Mineral Resources (Chief Inspector McKensey) v Berrima Coal Pty Ltd & Anor  NSWIRComm 130; Rodney Morrison v Powercoal Pty Limited  NSWIRComm 298; Rodney Morrison v Tahmoor Coal Pty Ltd  NSWIRComm 327; Rodney Dale Morrison v Coal Operations Australia Limited  NSWIRComm 239; Morrison v Powercoal Pty Ltd  NSWIRComm 342; Morrison & Powercoal Pty Ltd & Anor (No 3)  NSWIRComm 61; Rodney Morrison v Tecrete Industries Pty Ltd  NSWIRComm 371; Rodney Morrison v Anglo Coal (Dartbrook Management Pty Ltd)  NSWIRComm 397; Rodney Morrison v Gregory Alan Gardner  NSWIRComm 440; Rodney Morrison v Waratah Engineering Pty Limited  NSWIRComm 38; Rodney Morrison v Akula Pty Limited formerly known as RaiseBore Australia  NSWIRComm 41; Inspector Morrison v Cumnock No 1 Colliery Pty Ltd  NSWIRComm 151; Morrison v Wambo Coal Pty Ltd  NSWIRComm 189; Rodney Morrison v Eureka Opals Pty Limited  NSWIRComm 437; Morrison v Glennies Creek Coal Management Pty Ltd and anor  NSWIRComm 205.
 That this is the current policy, was made clear by the NSW Chief Inspector, in evidence before the Gretley Inquiry, above n2.
 See further pp23–25 below.
 See for example, Queensland Government Natural Resources Mines and Water, Queensland Mines and Quarries Safety Performance and Health Report (2004–2005) <http://www.nrm.qld.gov.au/mines/publications/safety_health/mqsafe05.html> accessed 18 October 2006.
 ‘Prosecutions Following Mining Fatality’ Minesafe News 15 Dec 2004 <http://minesafe.org/minesafe_news> accessed 10 Nov 2005> accessed 18 October 2006.
 See generally Lisa Mathews, ‘OHS Crime Seen as ‘lesser’ Offence?’ 42 Inside OHS at 42.
 Mark Ritter, Ministerial Inquiry Occupational Health and Safety Systems and Practices of BHP Billiton Iron Ore and Boodarie Iron Sites in Western Australia and Related Matters (2004) <http://www.premier.wa.gov.au/docs/features/BHP_Ministerial_Inquiry_Vol1.pdf> accessed 13 December 2006 (‘Ritter Report’). Note the power to issue directives under section 22 of MSIA to mine management to take certain corrective actions or to stop unsafe activity/equipment and withdraw persons from potentially hazardous areas. Directives are recorded in Mine Record Books and in the Mine Record Book database.
 Id at Appendix 4.
 Michelle Lam, ‘BHP Fine Alerts WA Mining Sector’ (2006) Occupational Health News 692 at 1.
 Robert Laing, Review of the Mines Safety and Inspection Act 1994: Final Report (2004) (‘Laing Report’).
 The Gretley Report, above n2; Ritter, above n25 at Appendix 4; Neil Gunningham, ‘Negotiated Non-compliance: A Case Study of Regulatory Failure’ (1987) 9 Law and Policy 69–87.
 Richard Johnstone, From Fiction to Fact — Rethinking OHS Enforcement (Working Paper No 11, National Research Centre for Occupational Health and Safety Regulation, 2003) at 5–7 available: <www.ohs.anu.edu.au> accessed 20 December 2006.
 See generally The Ritter Report, above n25 and Queensland ACIL Tasman, Queensland Mines Inspectorate Review: Final Report (2005).
 Department of Consumer and Employment Protection (WA) Resources Safety Enforcement and Prosecution Policy (2005) available at <http://www.docep.wa.gov.au/ResourcesSafety/Sections/Generic_PDFs/RS_Generic_EnforcProsec_policy_16Nov05.pdf> accessed 20 December 2006.
 Natural Resources and Mines (Qld), Compliance Policy (2001) available at <http://www.minesafe.org/library/files/compliance_policy.pdf> accessed 23 January 2007; currently under review – Natural Resources and Mines (Qld), Reforms to the Queensland Mines Inspectorate (2005) available at <http://www.nrw.qld.gov.au/mines/inspectorate/pdf/reforms_inspectorate.pdf> accessed 23 January 2007.
 Ritter, above n25.
 Ian Ayres & John Braithwaite, Responsive Regulation: Transcending the Deregulation Debate (1992).
 Under this strategy, the regulatory agency approaches each firm in a cooperative, flexible manner, but turns to punishment if and when the firm clearly defects from cooperation. Once the firm begins to cooperate again, the agency does so too. It should be noted that the enforcement pyramid is based on a repeat player prisoner’s dilemma, under which the regulator’s response (up or down the pyramid) depends upon the previous response of the regulatee.
 Colin Scott, ‘Regulation in the Age of Governance: The Rise of the Post Regulatory State’ in Jacinta Jordana & David Levi-Faur (eds), The Politics of Regulation: Institutions and Regulatory Reforms for the Age of Governance (2004). A central criticism is that the enforcement pyramid is based on game theory and in particular on a repeat player prisoner’s dilemma (or in simplistic terms, a ‘tit-for tat’ approach). See also Ayres & Braithwaite, above n35 at Chapters 2 and 3. Note in many regulatory contexts, encounters between regulators and regulated are so infrequent as to make such an approach implausible. This is not the case in the mining industry, where in relation to larger mine sites at least such regulatory encounters take place on a regular and routine basis.
 For an overview see Gunningham, above n29 at 69.
 See generally Gunningham and Johnstone, above n9 at Chapter 4; Hawkins, above n10.
 Johnstone, above n30 at 9.
 As regards the NSW mines inspectorate, see in particular Gunningham, above n41.
 Significantly, a rise in the incidence of fatalities and major injuries in the UK in the first half of the 1980s occurred at a time when inspection and enforcement activity fell significantly, due to a combination of increased workload and staff cuts. Phil James, ‘Reforming British Health and Safety Law: A Framework for Discussion’ (1992) 21 (2) Industrial Law Journal 87 at 97. For an Australian critique, documenting the failure of this approach in New South Wales, see Gunningham, above n41.
 See generally Industry Commission, Work, Health and Safety – An Inquiry into Occupational Health and Safety, Report No 47 (1995).
 Laing Report, above n28 at 771.
 Ritter Report, above n25 at Appendix 4.
 ACIL (1997) Final Review of Mine Safety in NSW, A Report to the Minister for Mineral Resources and Fisheries, ACIL Tasman, New Horizon Consulting Pty Ltd, Shaw Idea Pty Ltd.
 Richard Johnston, Review of Mine Safety in NSW (1997).
 Steve Tombs, ‘Law, Resistance and Reform’ (1995) 4 Social and Legal Studies at 343–365.
 See for example Robert Baldwin & Richard Anderson, Rethinking Regulatory Risk (2002) at 11; Wayne Gray & John Scholz, ‘Analysing the Equity and Efficiency of OHS Enforcement’ (1991) 3 Law and Policy 3 at 185.
 A distinction is made between general deterrence (premised on the notion that punishment of one enterprise will discourage others from engaging in similar proscribed conduct) and specific deterrence (premised on the notion that an enterprise that has experienced previous legal sanctions will be more inclined to make efforts to avoid future penalties). See generally Sally Simpson, Corporate Crime and Social Control (2002).
 See generally Gunningham & Johnstone 1999, above n9 at Chapter 6.
 Michael Wright, ‘Building an Evidence Based Case for the Health and Safety Commission Strategy to 2010’, Health and Safety Executive Research Report (2004) at 196.
 Eugene Bardach & Robert A Kagan, Going By The Book: The Problem of Regulatory Unreasonableness (1982); Robert Kagan, Adversarial Legalism: The American Way of Law (2001); John Scholz, ‘Voluntary Compliance and Regulatory Enforcement’ (1984) 6 Law and Policy 385; Andrew Hopkins, ‘The Gretley Coal Mining Disaster: Reflections on the Finding that Mine Managers Were to Blame’ National Research Centre for OHS Regulation Working Paper 39 (2005).
 See for example Submission from the Mine Managers Association of Australia to the Mine Safety Review (2004) at 9. However, there is as yet, little empirical support for this assertion.
 Fiona Haines, Corporate Regulation: Beyond Punish and Persuade (1997). However, see also Robyn Fairman & Charlotte Yapp, ‘Enforced Self-Regulation, Prescription and Conceptions of Compliance within Small Businesses’ (2005) 27 Law and Policy 4 – this suggests SMEs are rational and that until they are detected their conception of compliance allows them to feel compliant, therefore, being deterred from non-compliance is irrelevant.
 See Neil Gunningham, Shades of Green: Business, Regulation and Environment (2003).
 See John Braithwaite, ‘Restorative and Responsive Regulation of OHS’ in Elizabeth Bluff (ed), OHS Regulation for a Changing World of Work (2004).
 Bardach & Kagan, above n53.
 Much greater impact is likely to be achieved by the strategic (and relatively frequent) use of administrative penalties, improvement and prohibition notices and on the spot fines at in the lower reaches of the pyramid. See Wayne B Gray & John T Scholz, ‘OHSA Enforcement and Workplace Injuries: A Behavioural Approach to Risk Assessment’ (1990) 3 Journal of Risk & Uncertainty 283.
 Ken Phillips Monday, ‘The Politics of a Tragedy: The Gretley Mine Disaster and NSW OHS’, Occasional Paper (2006).
 Neil Gunningham, ‘Safety, Regulation and the Mining Industry’ (2006) 19 Australian Journal of Labour Law 30.
 See for example Chugg v Pacific Dunlop Ltd  VicRp 49;  VR 411 at 414–416; Richard Johnstone, Occupational Health and Safety Law and Policy (2004a) at Chapter 4.
 See Bankstown Foundry Pty Ltd v Braistina (1986) 160 CLR 301.
 Kit Carson & Richard Johnstone, ‘The Dupes of Hazard: Occupational Health and Safety and the Victorian Sanctions Debate’ (1990) 26 Australian and New Zealand Journal of Sociology 126.
 Ron McCallum, ‘Advice in Relation to Workplace Death, Occupational Health and Safety Legislation and Other Matters’, Report to the WorkCover Authority of New South Wales (2004) at 10, available at <http://www.workcover.nsw.gov.au/NR/rdonlyres/CA20497D-E6E8-4CDA-8649-D58A691E51B9/0/final_report_4481.pdf> accessed 23 February 2007.
 Andrew Hopkins, An Evaluation of Certain Criticisms of the NSW OHS Act, Working Paper 40, National Research Centre for Occupational Health and Safety Regulation at 7 <http://www.ohs.anu.edu.au/> accessed 18 December 2006.
 Australian Chamber of Commerce and Industry, Modern Workplace: Safer Workplace — An Industry Blueprint for Improving OHS in Australia (2005) at 21 available <http://www.acci.asn.au/text_files/review/r123.pdf> accessed 20 December 2006.
 See generally Elizabeth Bluff & Richard Johnstone, ‘The Relationship Between “Reasonably Practicable” and Risk Management Regulation’ (2005) 18 Australian Journal of Labour Law 197.
 McMartin v Newcastle Wallsend Coal Company Pty Ltd  NSWIRComm 31; Newcastle Wallsend Coal Company Pty Ltd and Ors v McMartin  NSWIRComm 339.
 James Reason, Managing the Risks of Organisational Accidents (1997) at 198.
 Id at 195.
 Reason, above n70 at 199.
 Health and Safety Executive, Enforcement Management Model (2002) <http://www.hse.gov.uk/enforce/emm.pdf> accessed 18 December 2006.
 Id at 34.
 Health and Safety Executive, above n74 at 7.
 Neil Gunningham & Richard Johnstone, Regulating Workplace Safety: Systems and Sanctions (1999) at 206.
 Id at Chapter 6.
 Reason, above n70 at 195.
 Andrew Hopkins, ‘The Gretley Coal Mine Disaster: Reflections on the Finding that Mine Managers were to Blame’ (Working Paper No 39, National Research Centre for Occupational Health and Safety Regulation, Australian National University, 2005) at 11, available at <http://www.ohs.anu.edu.au/> accessed 19 December 2006.
 Natural Resources and Mines, above n33 at 3.
 ACIL (2005) Final Report on the Queensland Mines Inspectorate Review, ACIL Tasman, New Horizon Consulting Pty Ltd, Shaw Idea Pty Ltd at 28–30.
 Haynes v CI and D Manufacturing Pty Ltd  NSWIRComm 230;  60 IR 149 at 158–159.
 Johnstone, above n30 at 5–7.
 Gunningham & Johnstone, above n78 at 212.
 Id at 213.
 McCallum, above n65.
 Andy Hall, Richard Johnstone & Alexa Ridgway, ‘Reflection on Reforms: Developing Criminal Accountability for Industrial Deaths’ (Working Paper No 26, National Research Centre for Occupational Health and Safety Regulation, Australian National University, 2004) available at <http://www.ohs.anu.edu.au/> accessed 19 December 2006.
 Ibid. See also Foster, above n6 at 79.
 See Foster, above n6 at 107–110.
 Carson & Johnstone, above n64 at 81.
 Andrew Hopkins, Making Safety Work: Getting Management Commitment to Occupational Health and Safety (1995a) at 130.
 See for example, Queensland Department of Mines and Energy and the Queensland Mining Council, Safety and Health Management for Queensland Mines and Quarries: Information Paper (1998) at 8.
 McMartin v Newcastle Wallsend Coal Company Pty Ltd  NSWIRComm 31 at 895
 See for example Department of Mineral Resources v Berrima Coal Pty Ltd  NSWIR Comm 130.
 Occupational Health and Safety Act 2000 (NSW) s 26(1).
 Occupational Health and Safety Act 2000 (NSW) s 26(1)(b); for a detailed analysis see Foster, above n6 at 79.
 Occupational Health and Safety Act 2000 (NSW) s 26(1)(a).
  NSWIRComm 323.
 McCallum, above n65 at 56.
 Id at 58.
 Coal Mining Safety and Health Act 1999 (Qld) s 33.
 See Coal Mine Health and Safety Act 2002 (NSW) s 37.
 For an alternative approach see Gunningham & Johnstone, above n78 at 33; See also Andrew Hopkins, ‘Holding Corporate Leaders Responsible’ (Working Paper No 48, National Research Centre for Occupational Health and Safety Regulation, Australian National University, 2006) available at <http://www.ohs.anu.edu.au/> accessed 19 December 2006.
 Gunningham & Johnstone, above n78 at 194.
 S Kruse & P Wilkinson, ‘A Brave New World: Less Law, More Safety?’ Paper delivered to New South Wales Minerals Council Health and Safety Conference, 15–18 May 2005 (Leura, NSW) at 5.
 Dan Dobbs, ‘Ending Punishment in “Punitive” Damages: Deterrence Measured Remedies’ (1989) 40 Alabama Law Review 831 at 844.
 Andy Hall, Richard Johnstone & Alexa Ridgway, ‘Reflection on Reforms: Developing Criminal Accountability for Industrial Deaths’ (Working Paper No 26, National Research Centre for Occupational Health and Safety Regulation, Australian National University, 2004), available at <http://www.ohs.anu.edu.au/> accessed 19 December 2006.
 Tony Marshall in John Braithwaite, Restorative Justice and Responsive Regulation (2002) at 11.
 Id at 64.
 Id at 298.
 Id at 29.
 Simpson, above n50.